Mraacnaie PDF

Ofer Aluf
Microwave
RF Antennas
and Circuits
Nonlinearity Applications in Engineering
Microwave RF Antennas and Circuits
Ofer Aluf
Microwave RF Antennas
and Circuits
Nonlinearity Applications in Engineering
123
Ofer Aluf
Netanya
Israel
ISBN 978-3-319-45425-2 ISBN 978-3-319-45427-6 (eBook)

DOI 10.1007/978-3-319-45427-6
Library of Congress Control Number: 2016950418
© Springer International Publishing Switzerland 2017

This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part
of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations,
recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission
or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar
methodology now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this
publication does not imply, even in the absence of a specific statement, that such names are exempt from
the relevant protective laws and regulations and therefore free for general use.
The publisher, the authors and the editors are safe to assume that the advice and information in this
book are believed to be true and accurate at the date of publication. Neither the publisher nor the
authors or the editors give a warranty, express or implied, with respect to the material contained herein or
for any errors or omissions that may have been made.
Printed on acid-free paper
This Springer imprint is published by Springer Nature

The registered company is Springer International Publishing AG
The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland
Preface
This book on microwave RF circuits: nonlinearity applications in engineering

covers and deals with two separate engineering and scientific areas and what
between. It gives advance analysis methods for Microwave RF Circuits which
represent many applications in engineering. Microwave RF Circuits come in many
topological structures and represent many specific implementations which stand the
target engineering features. Microwave RF Circuits include RFID antenna systems,
microwave elements, microwave semiconductor amplifiers, small-signal
(SS) amplifiers and matching networks, power amplifiers, oscillators, filters,
antennae systems, and high-power transistor circuit. The basic Microwave RF
Circuits can be characterized by some models and the associated equations. The
Microwave RF Circuits include RFID ICs and antennas, microstrip, circulators,
cylindrical RF network antennas, tunnel diode (TD), bipolar transistors, field-effect
transistors, IMPATT amplifiers, small-signal (SS) amplifiers, Bias-T circuits, PIN
diode, power amplifiers, LNAs, oscillators, resonators, filters, N-turn antennae, dual
spiral coils antennae, Helix antennas, linear dipole and slot array, and hybrid
translinear circuit. The Microwave RF Circuits analyze as linear and nonlinear
dynamical systems and their dynamics under parameter variations. This book is
aimed at newcomers to linear and nonlinear dynamics and chaos Microwave RF
Circuits. The presentation stresses analytical and numerical methods, concrete
examples, and geometric intuition. The Microwave RF Circuits analysis is devel-
oped systematically, starting with first-order differential equations and their bifur-
cation, followed by phase plane analysis, limit cycles and their bifurcations, chaos,
iterated maps, period doubling, renormalization, and strange attractors.
Additionally, the book is dealt with delayed Microwave RF Circuits which char-
acterized by overall variables delayed with time. Each variable has specific delay
parameter and can be inspected for dynamics. More realistic Microwave RF
Circuits models should include some of the past states of Microwave RF Circuits
and systems; that is, ideally, a real Microwave RF Circuits should be modeled by
differential equations with time delays. The use of delay differential equations
(DDEs) in the modeling of Microwave RF Circuits dynamics is currently very
active, largely due to progress achieved in the understanding of the dynamics of
v
vi Preface
several classes of delayed differential equations and Microwave RF Circuits and

systems. This book is designed for advanced undergraduate or graduate students in
electronics, RF and electronic engineering, physics, and mathematics who are
interested in Microwave RF Circuits dynamics and innovative analysis methods. It
is also addressed to electrical and RF engineers, physics experts and researchers in
physics, electronics, engineering and mathematics who use dynamical systems as
modeling tools in their studies. Therefore, only a moderate mathematical and
electronic semiconductor background in geometry, linear algebra, analysis, and
differential equations is required. Each chapter includes various Microwave RF
Circuits drawing and their equivalent analyses circuits. Microwave RF Circuits
fixed points and stability analysis done by using much estimation. Various bifur-
cations of Microwave RF Circuits are discussed.
In this book, we try to provide the reader with explicit procedures for application
of general Microwave RF Circuits mathematical representations to particular
research problems. Special attention is given to numerical implementation of the
developed techniques.
Let us briefly characterize the content of each chapter.
Chapter 1. RFID Antenna Systems Descriptions and Analysis. In this chapter,
RFID antenna systems are described and analyzed. RFID is a dedicated short-range
communication (DCRC) technology. RFID system consists of an antenna and a
transceiver, which read the radio frequency, and transfers the information to a
processing device (reader) and a transponder, or RFID tag. Active RFID tag system
includes energy source (battery), and it consumes energy. The active RFID tag
system is analyzed as an excitable linear bifurcation system. RFID tag-dimensional
parameters are optimized to get the best performances. Under delayed electro-
magnetic interferences, there are delays in some RFID tag coil variables and we
analyze it for stability optimization. There is a unique structure of RFID system,
semi-passive RFID tags with double-loop antennae arranged as a shifted gate. The
structure is optimized under delayed electromagnetic interferences. RFID tag
detector circuit is implemented by using schottky diode, and stability is analyzed for
parameter values variation. RFID system burst switch is a very important element,
and its behavior in time is inspected. The analysis fills the gap of analytical methods
for RFID systems analysis, concrete examples, and geometric examples. One of the
crucial RFID system optimization is in electromagnetic environmental which faced
RFID system variables delay in time. In some cases, RFID system can be repre-
sented as delayed differential equations, which depends on variable parameters and
delays. There are practical guidelines that combine graphical information with
analytical work to effectively study the local stability of RFID system models
involving delay-dependent parameters.
Chapter 2. Microwave Element Description and Stability Analysis. In this
chapter, microwave element stability is discussed. There are three types of
microwave circuits which include microwave elements. The first is a discrete cir-
cuit, packaged diodes/transistors mounted in coax and waveguide assemblies. The
second is Hybrid MIC (microwave integrated circuit), diodes/transistors and
microstrip fabricated separately and then assembled. The third is MMIC
Preface vii
(monolithic microwave integrated circuit), diodes, transistors, and microstrip

circuits, and other circuit elements, such as lumped capacitors and resistors, which
have parasitic effects influenced on overall system stability behavior. Microwave
transmission lines are delayed in time and are integral part of power limiter; the
stability is inspected for optimization. Reflection-type phase shifter (RTPS)
employs a circulator. The RTPS circuit includes microstrip transmission lines with
three-port active circulator and analyzes for stability optimization under time
delayed. Cylindrical RF network antennas for coupled plasma sources include
copper legs. They run as large-volume plasma sources and have stability switching
due to system’s copper leg parasitic effects. Tunnel diode (TD) is the p-n junction
device that exhibits negative resistance. Tunnel diode (TD) can be a microwave
oscillator. Transient is in the resonant cavity after turning the bias voltage ON. The
resonant circuit with NDR can oscillate. The Tunnel diode (TD) microwave
oscillator has parasitic effects in time and delay variables. The stability is optimized
when implementing tunnel diode (TD) in microwave oscillator.
Chapter 3. Microwave Semiconductor Amplifiers Analysis. In this chapter,
microwave semiconductor amplifier circuit analysis is discussed. Microwave
semiconductor amplifiers are widely used, and stability analysis is needed.
Microwave semiconductors can be bipolar transistors which operate at microwave
frequencies, and microwave field-effect transistors (FETs) minimize the adverse
effects of transit time and internal capacitance and resistance, IMPATT
(impact-ionization avalanche transit time) amplifier which widely used at the high
end of the microwave band. Stability of these microwave amplifiers is affected by
internal parameter variation and circuit microstrip parasitic effects. IMPATT diodes
which are a form of high-power diode are used in high-frequency electronic and
microwave devices. FET-combined biasing and matching circuit has many stability
issues which must be taken for every RF design, and analysis is done for best
performances.
Chapter 4. Small Signal (SS) Amplifiers and Matching Network Stability
Analysis. In this chapter, small-signal (SS) amplifiers and matching network
structures are analyzed for best performances. There are some types of amplifiers.
Amplifiers types are zero-frequency amplifiers (DC amplifiers), low-frequency
amplifiers (audio amplifiers), and high-frequency amplifiers (RF amplifiers).
Amplifiers come in three basic flavors: common base (CB) amplifiers, common
collector (CC) amplifiers, and common emitter (CE) amplifiers. It depends whether
the base, collector, or emitter is common to both the input and output of the
amplifier. When an amplifier’s output impedance matches the load impedance,
maximum power is transferred to the load and all reflections are eliminated. When
an amplifier’s output impedance unmatched the load impedance, there are reflec-
tions and less than maximum power is transferred to the load. There are instability
behaviors in these three types of amplifiers caused by circuit microstrip delays in
time parasitic effects. We use RF matching network in our design. There are typical
amplifiers matching networks: L matching network, T matching network, and PI
matching network. In design of microwave matching network, device parasitic
effects of length on RF circuit matching and stability. Bias-T three-port network
viii Preface
also suffers from instability under delayed microstrip in time. A PIN diode is
suitable for many applications and operates under high level of injection. The PIN
diode suffers from instability under parameter variations.
Chapter 5. Power Amplifier (PA) System Stability Analysis. In this chapter,
power amplifiers (PAs) are analyzed for best performances, and stability was also
discussed. Large-signal or power amplifiers (PAs) are used in the output stages of
audio amplifier systems to derive a load speaker. There are different types
of amplifiers which classified according to their circuit configurations and method
of operation. The classification of amplifiers ranged from linear operation with very
low efficiency to nonlinear operation but with a much higher efficiency, while
others are a compromise between the two. There are two basic amplifier class
groups. The first are the classically controlled conduction angle amplifiers forming
the more common amplifier classes (A, B, AB, and C). The second set of amplifiers
are the newer so-called switching amplifier classes (D, E, F, G, S, T). The most
commonly structured amplifier classes are those that are the most common type of
amplifier class mainly due to their simple design. We analyze the stability of these
amplifiers by inspecting the equivalent circuit differential equations. BJT transistor
is replaced by large-signal model in our analysis. The BJT model is known as the
Gummel–Poon model. The Ebers–Moll BJT model is a good large signal. We use
nonlinear dynamic in our analysis for amplifiers that feed by inputs/outputs exceed
certain limits. LNAs are used in many microwave and RF applications. We analyze
the stability of wideband low-noise amplifier (LNA) with negative feedback under
circuit’s parameter variation.
Chapter 6. Microwave/RF Oscillator Systems Stability Analysis. In this
chapter, our oscillator systems are discussed and their stability behavior is analyzed.
Oscillators can be classified into two types: relaxation and harmonic oscillators.
A microwave oscillator is an active device to generate power and a resonator to
control the frequency of the microwave signal. Important issues in oscillators are
frequency stability, frequency tuning, and phase noise. A phase-shift oscillator is a
linear electronic oscillator circuit that produces a sine wave output. The feedback
network “shifts” the phase of the amplifier output by 180° at the oscillation fre-
quency to give positive feedback, total phase shift of 360°. Phase-shift resonator
circuit stability analysis is done by considering BJT small-signal (SS) equivalent
circuit model. Closed-loop functioning oscillator can be viewed as feedback sys-
tem. The oscillation is sustained by feeding back a fraction of the output signal,
using an amplifier to gain the signal, and then injecting the energy back into the
tank. Closed-loop functioning oscillator stability is inspected and analyze. There are
types of transistor oscillators which use feedback and lumped inductance and
capacitance resonators. There are three types of transistor LC oscillators, Colpitts,
Hartley, and Clapp. In the Hartley oscillator, the feedback is supplied by the
inductive divider formed by two inductors. We apply the stability criterion of
Liapunov to our system. Colpitts oscillator is the same as Hertley oscillator and
instead of using a tapped inductance, Colpitts oscillator uses a tapped capacitance.
Colpitts oscillator circuit stability analysis is done by criterion of Liapunov.
Preface ix
Chapter 7. Filter Systems Stability Analysis. In this chapter, filter systems in

many circuits are inspected for dynamical behavior and stability analysis. The target
of analog and RF filtering is to modify the magnitude and phase of signal frequency
components. Many analog or radio frequency (RF) circuits perform filtering on the
signals passing through them. The analog and RF filter types are defined on the
criteria how they modify the magnitude and/or phase of sinusoidal frequency
components. Microwave and RF filters pass a range of frequencies and reject other
frequencies. A diplexer is a passive device that implements frequency-domain
multiplexing. Two ports are multiplexed onto a third port. A diplexer multiplexes
two ports onto one port, but more than two parts may be multiplexed. We analyze
BPF diplexer circuit stability by using geometric stability switch criteria in delay
differential systems. A diplexer filters to pass two bands to separate ports, and
stability analysis under parameter variation. The standard local stability analysis
about any one of the equilibrium points of dual-band diplexer filter circuit is done.
We use crystal in place of LC filter for low-frequency applications. There are lattice
crystal filter, half lattice, and cascaded half lattice filters. The standard local stability
analysis about any one of the equilibrium point of lattice crystal filter circuit is
done. A tunable BPF employing varactor diodes is ideal for many diverse wireless
applications. There are two types of tunable BPF employing varactor diodes: top
inductively coupled variable BPF and capacitively coupled variable band-pass
filter. BPF (varactor diodes) circuit involving N variables and stability behavior is
inspected.
Chapter 8. Antenna System Stability Analysis. In this chapter, we discussed
various antenna systems and behaviors for different conditions for best perfor-
mances. An antenna is a conductor or group of conductors used for radiating
electromagnetic energy into space or collecting electromagnetic energy from space.
There are many types of antennas and we discussed those antennas that operate at
microwave frequencies. Microwave refer to radio waves with wavelength ranging
from as long as one meter to as short as one millimeter with frequencies between
300 MHz and 300 GHz. Another antenna area is for RFID applications.
A complete RFID system includes RFID reader and transponder units. N-turn
multilayer circular-coil antennas can be integrated with RFID IC for complete RFID
tags. We investigate the system stability optimization under delayed electromag-
netic interference and parasitic effects. The system is constructed from two
antennas: each one N-turn multilayer circular antenna. The standard local stability
analysis about any one of the equilibrium points (fixed points) of N-turn multilayer
circular-coil antenna RFID system is done. We analyze circuit stability where there
is a delay in the first and second RFIDs’ N-turn multilayer-coil antenna voltages
and antenna voltage derivatives. A double-rectangular spiral antenna is constructed
from two antennas, each antenna is a rectangular spiral antenna. Antennas are
connected in series with microstrip line and to the RFID IC. The standard local
stability analysis about any one of the equilibrium points of RFID tags with double
rectangular spiral antenna system is done. A system of single-turn square planar
straight thin-film inductor antenna (four segments) is constructed from four straight
thin-film inductors which are connected in a single-turn square structure. There are
x Preface
delays in time for the microstrip line parasitic effects, and stability switching is
inspected for different values of delay variables. A helical antenna is an antenna
consisting of a conducting wire wound in the form of a helix. The helical antennas
are mounted over a ground plane. Helical antennas can operate in one of two
principal modes: normal mode or axial mode. Helix antenna system stability is
inspected under parameter variation.
Chapter 9. Microwave RF Antennas and Circuits Bifurcation Behavior,
Investigation, Comparison and Conclusion. In this chapter, we summarized the
main topics regarding microwave and RF antennas and systems, inspect behavior,
dynamics, stability, comparison, and conclusion. Microwave RF antennas are an
integral part of every RF or microwave system. An antenna is an electrical device
which converts electric power into radio waves, and vice versa. In many wireless
applications, antennas are required by radio receiver or transmitter to couple its
electrical connection to the electromagnetic field. When we inspect system stability
which includes radio waves, we inspect electromagnetic waves which carry signals
through the space (or air) at the speed of light with almost no transmission loss.
There are mainly two categories of antennas. The first is omnidirectional antenna
which receives and/or radiates in all directions. The second is directional antenna
which radiates in a particular direction or pattern. Antennas are characterized by a
number of parameters, radiation pattern, and the resulting gain. Antenna’s gain is
dependent on its power in the horizontal directions, and antenna’s power gain takes
into account the antenna’s efficiency (figure of merit). The physical size of an
antenna is a practical issue, particularly at lower frequencies. Stability analysis
includes a complete RF system with antennas and matching networks.
Netanya, Israel Ofer Aluf

Contents
1 RFID Antennas Systems Descriptions and Analysis . . . . . . . . . . . . . . 1

1.1 Active RFID TAGs System Analysis of Energy Consumption
as Excitable Linear Bifurcation System . . . . . . . . . . . . . . . . . . . . . 2
1.2 RFID TAG’s Dimensional Parameters Optimization
as Excitable Linear Bifurcation Systems . . . . . . . . . . . . . . . . . . . . . 14
1.3 RFID TAGs Coil’s System Stability Optimization Under
Delayed Electromagnetic Interferences . . . . . . . . . . . . . . . . . . . . . . 22
1.4 Semi-Passive RFID Tags with Double Loop Antennas
Arranged as a Shifted Gate System for Stability
Optimization Under Delayed Electromagnetic Interferences . . . . . . 44
1.5 RFID TAGs Detectors Stability Analysis Under Delayed
Schottky Diode’s Internal Elements in Time . . . . . . . . . . . . . . . . . 72
1.6 RFID System Burst Switch Stability Analysis Under
Delayed Internal Diode Circuitry Parasitic Effects in Time . . . . . . . 104
Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 144
2 Microwave Elements Description and Stability Analysis . . . . . . .... 155
2.1 Microstrip Transmission Lines Delayed in Time Power
Limiters Stability Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . .... 156
2.2 Three Ports Active Circulator’s Reflection Type Phase
Shifter (RTPS) Circuit Transmission Lines Delayed
in Time System Stability Analysis . . . . . . . . . . . . . . . . . . . . . .... 171
2.3 Cylindrical RF Network Antennas for Coupled Plasma
Sources Copper Legs Delayed in Time System Stability
Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .... 196
2.4 Tunnel Diode (TD) as a Microwave Oscillator System
Cavity Parasitic Elements Stability Analysis . . . . . . . . . . . . . . .... 221
Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .... 267
xi
xii Contents
3 Microwave Semiconductor Amplifiers Analysis. . . . . . . . . . . . . . .... 279

3.1 Bipolar Transistor at Microwave Frequencies Description
and Stability Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .... 279
3.2 Field Effect Transistor (FETs) at Microwave Frequencies
Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .... 299
3.3 Field Effect Transistor (FETs) at Microwave Frequencies
Stability Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .... 318
3.4 IMPATT Amplifier Stability Analysis . . . . . . . . . . . . . . . . . . .... 333
3.5 Multistage IMPATT Amplifier System Microstrip Delayed
in Time Stability Switching Analysis . . . . . . . . . . . . . . . . . . . .... 373
3.6 FET Combined Biasing and Matching Circuit Stability
Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .... 382
Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .... 392
4 Small Signal (SS) Amplifiers and Matching Network
Stability Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .... 405
4.1 Small Signal (SS) Amplifiers and Matching Network . . . . . . . .... 406
4.2 Small Signal (SS) Amplifiers PI & T’s Matching Network
and Transformation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .... 422
4.3 Small Signal (SS) Amplifiers Matching Network Stability
Analysis Under Microstrip Parasitic Parameters Variation . . . .... 435
4.4 Bias—T Three Port Network Stability Switching Under
Delayed Micro Strip in Time . . . . . . . . . . . . . . . . . . . . . . . . . .... 460
4.5 PIN Diode Stability Analysis Under Parameters Variation . . . .... 489
Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .... 501
5 Power Amplifier (PA) System Stability Analysis . . . . . . . . . . . . . .... 513
5.1 Class AB Push-Pull Power Amplifiers Stability Analysis
Under Parameters Variation . . . . . . . . . . . . . . . . . . . . . . . . . . .... 514
5.2 Class C Power Amplifier (PA) with Parallel Resonance
Circuit Stability Analysis Under Parameters Variation . . . . . . .... 528
5.3 Single Ended Class B Amplifier Gummel-Poon Model
Analysis Under Parameters Variation . . . . . . . . . . . . . . . . . . . .... 559
5.4 Wideband Low Noise Amplifier (LNA) with Negative
Feedback Circuit Stability Analysis Under Circuit’s
Parameters Variation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .... 573
Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .... 587
6 Microwave/RF Oscillator Systems Stability Analysis . . . . . . . . . . . . . 601
6.1 A Resonator Circuit 180° Phase Shift from Its Input
to Output Stability Analysis Under Delayed Variables
in Time . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 602
6.2 Closed Loop Functioning Oscillator Stability Analysis
Under Parameters Variations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 617
Contents xiii
6.3 Hartley Oscillator Stability Analysis . . . . . . . . . . . . . . . . . . . . . . . . 630

6.4 Colpitts Oscillator Stability Analysis . . . . . . . . . . . . . . . . . . . . . . . 644
Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 661
7 Filters Systems Stability Analysis . . . . . . . . . . . . . . . . . . . . . . . . . .... 673
7.1 BPF Diplexer Without a Series Input Stability Analysis . . . . . .... 674
7.2 Dual Band Diplexer Filter Stability Analysis Under
7.3 A Crystal-Lattice BPF Circuit Stability Analysis . . . . . . . . . . .... 711
7.4 A Tunable BPF Employing Varactor Diodes Stability
Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .... 745
Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .... 768
8 Antennas System Stability Analysis . . . . . . . . . . . . . . . . . . . . . . . .... 783
8.1 N-Turn Multilayer Circular Coil Antennas Transceiver
System Stability Optimization Under Delayed
Electromagnetic Interferences . . . . . . . . . . . . . . . . . . . . . . . . . .... 785
8.2 Double Rectangular Spiral Coils Antennas System Stability
Optimization Under Delayed Electromagnetic Interferences
and Parasitic Effects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .... 818
8.3 Single-Turn Square Planar Straight Thin Film Inductors
Antenna System Stability Optimization Under Microstrip
Delayed in Time . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .... 849
8.4 Helix Antennas System Stability Analysis Under
Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .... 885
9 Microwave RF Antennas and Circuits Bifurcation
Behavior, Investigation, Comparison and Conclusion . . . . . . . . . . . . 903
Appendix A: RFID LF TAG 125 kHz/134 kHz Design
and Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 913
Appendix B: RF Amplifiers Basic and Advance Topics
and Design Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 941
Appendix C: BJT Transistor Ebers-Moll Model and MOSFET
Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 993
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1035
Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1041
Introduction
Microwave RF antenna products are currently in a widely use in all aspects of

engineering designs. Microwave RF antenna products are transmission lines,
coaxial cables, waveguide, strip line and microstrip, microwave semiconductors
(PIN diode, RF bipolar transistor, RF FET, varactor, schottky diode, LDMOS,
DMOS, GaN devices, etc.), RF combiner and couplers, isolators and circulator,
filters, attenuators, switches, phase shifter, detectors, amplifiers, oscillators, tubes,
microwave antennas (dipole, slot, horn, spiral, helix, arrays, parabolic dish, phased
arrays), low-power communication antennas (ZigBee, RFID/NFC, Bluetooth,
Wi-Fi, GPS, etc.). The basic structure of Microwave RF antenna product contains
de/multiplexed amplifiers, filters, mixers, etc. The Microwave RF antenna system
ports are RF inputs, RF outputs, oscillators, and input control lines). The below
figure demonstrates the basic structure of Microwave RF antenna system:
Control
lines
RF inputs RF outputs
( VjiIN , IjiIN ) ( VjiOUT , IjiOUT )
Microwave
system
Oscillators
The input control lines can be connected to additional embedded system by

many kinds of interfaces (RS232, RS485, UART, SPI, SDIO, etc.). RF inputs can
be connected to various antennas and other RF systems. RF outputs can be
xv
xvi Introduction
connected to additional RF systems and devices. The overall Microwave RF

antenna system equation can be represented as below:
i¼1 Y
X j¼1
d n VjiIN d n IjiIN
fVjiOUT ; IjiOUT g ¼ f ðVji; IjiÞ ¼ kji ðVjiIN ; IjiIN ; ; Þ
n i
dtn dtn
The Microwave RF systems can be represented as RF input circuit and RF

output circuit. The RF system function contains high-degree derivatives of input
and output voltages.
The primary purpose of a Microwave and RF system application functionality on
input variables. Many topological Microwave/RF system schematics give a variety
of nonlinear behaviors which can be implemented in many engineering areas. Each
Microwave/RF system can be represented by a set of differential equations which
depend on Microwave/RF system’s variable parameters. The investigation of
Microwave and RF system’s differential equation bifurcation theory, the study of
possible changes in the structure of the orbits of a differential equation depending
on variable parameters. The book illustrates certain observations and analyzes local
bifurcations of an appropriate arbitrary scalar differential equation. Since the
implicit function theorem is the main ingredient used in these generalizations,
include a precise statement of this theorem. Additional analyze the bifurcations of a
Microwave/RF system’s differential equation of the circle. The bifurcation behavior
of specific differential equations can be encapsulated in certain pictures called
bifurcation diagrams. All of that for optimization of Microwave/RF system’s
parameter optimization—to get the best performance. Dynamics (Chaos, fractals)
change with systems that evolute in time. There are two types of dynamical sys-
tems: differential equations and iterated map (difference equations). Differential
equation has described the evolution of systems in continuous time. Iterated map is
arising in problems where the time is discrete. Differential equations can be divided
into two main groups: ordinary differential equations and partial differential equa-
tions. The differential equation system can be represented as below:
dX1
dt ¼ f1 ðx1 ; . . .. . .; xn Þ
:
:
dXn
dt ¼ fn ðx1 ; . . .. . .; xn Þ
:
Xi ¼ dX dt
i
Some of the Microwave/RF systems can be represented as an equation in

dimension “one.” Basic notions of Microwave/RF systems, circuit stability, and
bifurcations of vector fields are easily explained for scalar autonomous equations
dimension one—because their flows are determined from the equilibrium points.
Numerical solutions of such equations lead to scalar maps and show some of the
“anomalies” albeit profound and exciting that may arise when numerical approxi-
mation is a poor period doubling bifurcation, chaos, etc.
Introduction xvii
Microwave/RF systems equations can be turned to the dynamics and bifurca-

tions of periodic solutions of no autonomous equations with periodic coefficients’
dimension one and one half, where scalar maps reappear naturally as Poincare
maps. Microwave/RF system investigates the dynamics of planar autonomous
equations—dimension two—where, in addition to equilibria, new dynamical
behavior, such as periodic and homoclinic orbits, appears. Microwave/RF system
schematic stability of an equilibrium point, subtle topological aspects of linear
systems as well as the standard theory of Liapunov functions. Center manifolds and
the method of Liapunov–Schmidt to make a reduction to a scalar autonomous
equation. Periodic orbit—Poincare—Andronov—Hopf bifurcation—and its analy-
sis can be reduced to that of a nonautonomous periodic equation.
Additionally, we discussed Microwave RF antenna systems with delay elements
(parasitic effects, circuit component delays, microstrip delays, etc.). Our
Microwave RF antenna system delay differential and delay different model can be
analytically used with delay differential equations in dynamically. The need of the
incorporation of a time delay is often of the existence of any stage structure. It is
often difficult to analytically study models with delay-dependent parameters, even if
only a single discrete delay is present. There are practical guidelines that combine
graphical information with analytical work to effectively study the local stability of
models involving delay-dependent parameters. The stability of a given steady state
is simply determined by the graphs of some function of s1 ; . . .; sn ; n 2 N which can
be expressed explicitly and thus can be easily depicted by MATLAB and other
popular software. We need only look at one such function and locate the zero. This
function often has only two zeros, providing thresholds for stability switches. As
time delay increases, stability changes from stable to unstable to stable. We
emphasize the local stability aspects of some models with delay-dependent
parameters. Additionally, there is a general geometric criterion that, theoretically
speaking, can be applied to models with many delays, or even distributed delays.
The simplest case of a first-order characteristic equation provides more
user-friendly geometric and analytic criteria for stability switches. The analytical
criteria provided for the first- and second-order cases can be used to obtain some
insightful analytical statements and can be helpful for conducting simulations.
Chapter 1
RFID Antennas Systems Descriptions
and Analysis
RFID is short for radio frequency identification; RFID is a dedicated short range
communication (DSRC) technology. The term RFID is used to describe various
technologies that use radio waves to automatically identify people or objects.
With RFID, the electromagnetic or electrostatic coupling in the RF (radio fre-
quency) portion of the electromagnetic spectrum is used to transmit signals. RFID
system consists of an antenna and a transceiver, which read the radio frequency and
transfers the information to a processing device (reader) and a transponder, or RF
tag, which contains the RF circuitry and information to be transmitted. The antenna
provides the means for the integrated circuit to transmit its information to the reader
that converts the radio waves reflected back from the RFID tag into digital infor-
mation that can then be passed on to computers that can analyze the data. In RFID
systems, the tags that hold the data are broken down into two different types.
Passive tags use the radio frequency from the reader to transmit their signal and
Active tags. Passive tags use the radio frequency from the reader to transmit their
signal. Passive tags will generally have their data permanently burned into the tag
when it is made, although some can be rewritten. Active tags are much more
sophisticated and have an on-board battery for power to transmit their data signal
over a greater distance and power random access memory (RAM) giving them the
ability to store up to 32,000 bytes of data. RFID systems can use a variety of
frequencies to communicate, but because radio waves work and act differently at
different frequencies, a frequency for a specific RFID system is often dependent on
its application. An RFID system is always made up of two components:
transponder, which is located on the object to be identified, detector or reader,
which, depending upon design and the technology used, may be a read or write/read
device. There is a need to analyzing RFID systems. The analysis is based on
nonlinear dynamics and chaos models and shows comprehensive benefits and
results. The dynamics of RFID systems provides several ways to use them in a
variety of applications covering wide areas. The analysis fills the gap of analytical
methods for RFID systems analysis, concrete examples, and geometric examples.
The RFID systems analysis is developed systematically, starting with basic passive
© Springer International Publishing Switzerland 2017 1
O. Aluf, Microwave RF Antennas and Circuits,
DOI 10.1007/978-3-319-45427-6_1
2 1 RFID Antennas Systems Descriptions and Analysis
and active RFID systems, differential equations and their bifurcations, followed by
fixed point analysis, limit cycles and their bifurcations. One of the crucial RFID
system optimization is in electromagnetic environmental which faced RFID system
variable delay in time. In some cases RFID system can represent as a delayed
differential equations which, depending on variable parameters and delays. There
are practical guidelines that combine graphical information with analytical work to
effectively study the local stability of models involving delay dependent parame-
ters. The stability of a given steady state is determined by the graphs of some
function [2–4, 85].
1.1 Active RFID TAGs System Analysis of Energy

Consumption as Excitable Linear Bifurcation System
Active RFID Tags have a built in power supply, such as a battery, as well as
electronics that perform specialized tasks. By contrast, passive RFID TAGs do not
have a power supply and must rely on the power emitted by an RFID Reader to
transmit data. Thus, if a reader is not present, the passive TAGs can’t communicate
a data. Active TAGs can communicate in the absence of a reader. Active RFID
Tags system energy consumption can be a function of many variables: q(m), u(m), z
(m), t(m), tms(m), when m is the number of TAG IDs which are uniformly dis-
tributed in the interval [0,1). It is very important to emphasize that basic
Active RFID TAG, equivalent circuit is Capacitor (Cic), Resistor (Ric), L (RFID’s
Coil inductance as a function of overall Coil’s parameters) all in parallel and
Voltage generator Vs(t) with serial parasitic resistance. The Voltage generator and
serial parasitic resistance are in parallel to all other Active RFID TAG’s elements
(Cic, Ric, and L (Coil inductance)). The Active RFID TAG equivalent circuit can
be represented as a differential equation which depending on variable parameters.
The investigation of Active RFID’s differential equation based on bifurcation
theory, the study of possible changes in the structure of the orbits of a differential
equation depending on variable parameters. We first illustrate certain observations
and analyze local bifurcations of an appropriate arbitrary scalar differential equa-
tion. Finally, investigate Active RFID TAGs system energy for the best perfor-
mance using an excitable bifurcation diagram. Active RFID Tags have a built in
power supply, such as a battery. The major advantages of an active RFID Tags are:
It can be read at distances of one hundred feet or more, greatly improving the utility
of the device. It may have other sensors that can use electricity for power. The
disadvantages of an active RFID Tags are: The TAG cannot function without
battery power, which limits the lifetime of the TAG. The TAG is typically more
expensive. The TAG is physically larger, which may limit applications. The long
term maintenance costs for an active RFID tag can be greater than those of a passive
Tag if the batteries are replaced. Battery outages in an active TAGs can result in
expensive misreads. Active RFID TAGs may have all or some of the following
1.1 Active RFID TAGs System Analysis of Energy Consumption … 3
features: Longest communication range of any TAG, the capability to perform an

independent monitoring and control, the capability of initiating communications,
the capabilities of performing diagnostics, and the highest data bandwidth. The
active RFID TAGs may even be equipped with autonomous networking; the TAGs
autonomously determine the best communication path. Mainly active RFID TAGs
have a built in power supply, such as battery, as well as electronics that perform
specialized tasks. By contrast, passive RFID TAGs do not have a power supply and
must rely on the power emitted by an RFID Reader to transmit data. There is an
arbitration while reading TAGs (TAGs anti-collision problem). First, identify and
then read data stored on RFID Tags [85] (Fig. 1.1).
It is very important to read TAG IDs of all. The Anti-collision protocol based on
two methods: ALOHA and its variants and Binary tree search. ALOHA protocol,
reducing collisions by separating TAG responds by time (probabilistic and simple).
TAG ID may not read for a very long time. The Binary tree search protocol is
deterministic in nature. Read all TAGs by successively querying nodes at different
levels of the tree with TAG IDs distributed on the tree based on their prefix.
Guarantee that all TAGs IDs will be read within a certain time frame. The binary
tree search procedure, however, uses up a lot of reader queries and TAG responses
by relying on colliding responses of TAGs to determine which sub tree to query
next. Higher energy consumption in readers and tags (If they are active TAGs).
TAGs can’t be assumed to be able to communicate with each other directly. TAGs
may not be able of storing states of the arbitration process in their memory. There
are three anti-collision protocols: All’s include and combine the ideas of a binary
tree search protocol with frame slotted ALOHA, deterministic schemes, and energy
aware. The first anti-collision protocol is a Multi Slotted (MS) scheme, multiple
slots per query to reduce the chances of collision among the TAG responses. The
second anti-collision protocol is a Multi Slotted with Selective sleep (MSS) scheme;
using sleep commands to put resolved TAGs to sleep during the arbitration process.
Both MS and MSS have a probabilistic flavor, TAGs choose a reply slot in a query
frame randomly. The third anti-collision protocol is a Multi Slotted with Assigned
slots (MAS), assigning tags in each sub tree of the search tree to a specific slot of
the query frame. It’s a deterministic protocol, including the replay behavior of tags.
All three protocols can adjust the frame size used per query. Maximize energy
savings at the reader by reducing collisions among TAG responses. The frame size
is also chosen based on a specified average time constraint within which all TAGs
IDs must be read. The binary search protocols are Binary Tree (BT) and Query Tree
Fig. 1.1 Reader TAG

interrogation diagram TAG 0
Reader Interrogation
Unit signal (query)
TAG n
(QT). Both work by splitting TAG IDs using queries from the reader until all tags
are read. Binary Tree (BT) relies on TAGs remembering the results of previous
inquiries by the readers. TAGs susceptible to their power supply. Query Tree
(QT) protocol, is a deterministic TAG anti-collision protocol, which is memory loss
with TAGs requiring no additional memory except that required to store their ID
(Fig. 1.2).
The approach to energy aware anti-collision protocols for RFID systems is to
combine the deterministic nature of binary search algorithms along with the sim-
plicity of frame slotted ALOHA to reduce the number of TAG response collisions.
The QT protocol relies on colliding responses to queries that are sent to internal
nodes of a tree to determine the location of TAG ID. Allow tags to transmit
responses within a slotted time frame and thus, try to avoid collisions with
responses from other tags. The energy consumption at the reader is a function of the
number of queries it sends, and number of slots spent in the receive mode. Energy
consumption at an active TAG is a function of the number of queries received by
the TAG and the number of responses it sends back. Neglect the energy spent in
modes other than transmit and receive for simplicity. Assumption: Time slot in
which a reader query or message is sent is equal to the duration as that of a TAG
response. The energy model of the reader is based upon a half-duplex operation.
Reader transmits energy, and its query for a specific period and then wait in
receiving mode with no more energy transmission until the end of the frame. The
flow chart for reader query and TAGs: (Fig. 1.3).
Response mechanism is as below: (Fig. 1.4).
Pulse based half duplex operation is termed as sequential (SEQ) operational
(Fig. 1.5).
Query
(prefix)
TAGn+1 (no
Reader Perfix)
TAG1
(Perfix)
TAG2
(Perfix)
Responds (Perfix
+ TAG ID)
TAGn
(Perfix) TAGn+k (no
Perfix)
Fig. 1.2 Reader TAGs system query and responds

Reader Wait time

query (Receive mode)
Energy No Energy
consumption consumption
One Frame
Start of End of
Frame Frame
Fig. 1.3 One frame reader query and wait time
Start
n=1
Reader query
(specific prefix)
TAG n, TAG n+1

….. Respond
No. of TAGs respond YES

to a specific prefix
query (reader) > 1
NO
Reader extends
TAG is resolved and the prefix by
uniquely identified ‘0’ or ‘1’ bit
and continues
the query with
n=n+1 this longer
Fig. 1.4 Flow chart for reader query and TAGs

One slot for a query F slots reader wait

from reader for response
Fig. 1.5 One slot for a query and F slots reader wait for a response
The power required by the reader to transmit The power required by the reader to receive
PRtx PRrx
The power required by an active TAG to The power required by an active TAG to
transmit receive
PTtx PTrx
Reader energy consumption: q(m) (PRtx + PRrx F) when q(m) is the number
of queries for reading m TAGs. The energy consumption of all active TAGs:
q(m) PTrx + u(m) PTtx when q(m) is the number of reader queries, u(m) is the
number of TAG responses. For MSS scheme (include sleep command) the reader
energy consumption is q(m) (PRtx + PRrx F) + z(m) PRtx. The total energy
consumption for all active TAGs is q(m) PTrx + u(m) PTtx + z(m) PTrx,
when z(m) is the number of sleep commands issued by the reader. The average
analysis of energy consumption:

qðmÞaverage number of reader queires:

uðmÞaverage number of TAG responses:
zðmÞaverage number of sleep commands
issued by the reader ðonly for MSS SchemeÞ
tðmÞaverage number of time slots required to read all TAGs:
tMS ðmÞaverage number of time slots required to read m TAGs
m TAG IDs are uniformly distributed in the interval [0.1]. We get the expression for
one active RFID TAG total energy consumption U(m) = u(m):
1
TAG Power ¼ ½qðmÞ PTrx þ UðmÞ PTtx þ ZðmÞ PTrx
m
Active RFID TAG can represent as a parallel Equivalent Circuit of Capacitor

and Resistor in parallel with Supply voltage source (internal resistance) (Fig. 1.6).
The Active RFID TAG Antenna can be represented as Parallel inductor to the
basic Active RFID Equivalent Circuit. The simplified complete equivalent circuit of
the label is as below: (Fig. 1.7)
Active RFID TAG

LA
LB Voltage
source
Antenna
Fig. 1.6 Active RFID TAG system
V(t)
Rs Active RFID Antenna

C1 L1
R1
Vs(t)
Active RFID's Equivalent circuit
Fig. 1.7 Active RFID TAG’s equivalent circuit
Zt1
dI 1
VL1 ¼ L1 ) I L1 ¼ VL1 dt;
dt L1
0
dVc1 X
4
IC1 ¼ C1 ; Ij ¼ 0
dt j¼1
V ¼ Vc1 ¼ VL1 ¼ VR1

Zt1
V dV 1 V VsðtÞ
þ C1 þ V dt þ ¼0
R1 dt L1 Rs
0
1 dV d2V 1 dV 1 dVsðtÞdt !eð0\e1Þ

f þ C1 2 þ Vþ g ! 0
R1 dt dt L1 dt Rs
:: 1 1 : 1 1
e 1 ) V C1 þ ð þ Þ V þ V ¼ V_ S ðtÞ
R1 Rs L1 Rs
1 dV d2 V 1 dV dVsðtÞ 1
þ C1 2 þ V þ½ ¼0
R1 dt dt L1 dt dt Rs
dV1 dV
V2 ¼ ¼ ; V1 ¼ V
dt dt
dV1 dV2 1 1 1 1 dVsðtÞ
¼ V2 ; ¼ ½ þ V2 V1 þ
dt dt C1 R1 Rs C1 C L Rs C1 dt
! ! 1 1 !
dV1
0 1 V1 0
dt
¼ þ dVsðtÞ
dV2 C11L1 ½C11R1 þ RsC
1
V2 RsC1 dt
1
dt
hl i
1
Lcalc ¼ 0 ½X1 þ X2 X3 þ X4 Ncp

p
L1 = Lcalc !
2 Aavg Bavg
X1 ¼ Aavg ln pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
d ðAavg þ Aavg2 þ Bavg2 Þ
!
2 Aavg Bavg
X2 ¼ Bavg ln pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
d ðBavg þ Aavg2 þ Bavg2 Þ
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
ffi
X3 ¼ 2 Aavg þ Bavg Aavg2 þ Bavg2 ; X4 ¼ ðAavg þ BavgÞ=4
The RFID’s coil calculation inductance expression is the definition of limits,

Estimations: Track thickness t, Al and Cu coils (t > 30 lm). The printed coils as
high as possible. Estimation of turn exponent p is needed for inductance calculation.
Coil manufacturing technology P

Wired 1.8–1.9
Etched 1.75–1.85
Printed 1.7–1.8
Active RFID can be considered as a Van der Pol’s system. Van der Pol’s
equation provides an example of an oscillator with nonlinear damping, energy
being dissipated at large amplitudes and generated at low amplitudes. Such systems
typically possess limit cycles, sustained oscillations around a state at which energy
generation and dissipation balance. The basic Van der Pol’s equation can be written
:: :
in the form: X þ a /ðxÞ X þ X ¼ b qðtÞ.
:: 1 1 : 1 1
e 1 ) V C1 þ ð þ ÞV þ V ¼ V_ S ðtÞ
R1 Rs L1 Rs
:: 1 1 1 : 1 1
e1)V þ ð þ ÞV þ V ¼ V_ S ðtÞ
C1 R1 Rs L1 C1 Rs C1
1 1 1 1 1
X ! V; a /ðxÞ ! ð þ Þ; ! 1; V_ S ðtÞ ! b qðtÞ
C1 R1 Rs L1 C1 Rs C1
Let’s define:
:: 1 1 : 1 1
fs ðtÞ ¼ V_ S ðt Þ ) e 1 ) V C1 þ ð þ ÞV þ V ¼ fs ðtÞ
R1 Rs L1 Rs
“f” is a “T” periodic function of the independent variable t, and k ¼ Rs 1
The term k fs ðtÞ ¼ Rs V_ S ðtÞ is called the forcing function k ! 0 ) Rs

1 1
!0)
Rs ! 1 there is no forcing and the system act as Van Der Pol Oscillator. It is
necessary to examine the trajectories (V1, V2, and t) of the non-autonomous Active
RFID system in R2 xR rather than the orbits in R2 . Equivalently, we may consider
the orbits of the Active RFID TAGs three dimensional autonomous systems.
dV1
¼ V2
dt
dV2 1 1 1 1
¼ ½ þ V2 V1 þ fs ðtÞ 8 fs ðtÞ ¼ V_ S ðtÞ
dt C1 R1 Rs C1 C1 L1 Rs C1
dV3
¼ 1 8 ðV3 ðtÞ ¼ tÞ
dt
First examine the case of k ¼ 0 ) Rs C1 ! 1; C1 ¼ const; then Rs ! 1

The limit cycle, the isolated periodic orbit, of the unforced oscillator of Van Der
Pol becomes a cylinder; that is, topologically it is a homomorphism to S1 R. The
cylinder is an invariant manifold in the sense that any solution starting on the
cylinder remains on it for all positive time. This invariant cylinder attracts all nearby
solutions. For k ¼ 0, k ! 0 Rs ! 1 the Active RFID TAG invariant cylinder is
filled with a family of periodic solutions. The cylinder under the projection R2
R ! R2 simply becomes the limit cycle. Actually active RFID TAGs act as peri-
odic forcing with small amplitude, that jkj small. In this case, there is still a cylinder
in R2 R close to the invariant cylinder of the unforced oscillator. This new
cylinder is an invariant manifold of solutions of the forced equation and attracts all
nearby solutions. The flow on the invariant cylinder of the forced equation can be
quite different from the one of the unforced oscillators. In Active RFID TAG
concern to the Van Der Pol’s equation, we get the equation:
:: :
X þ a /ðxÞ X þ X ¼ k fs ðtÞ
:: 1 1 1 : 1 1
e 1 ) V þð þ Þ V þ V ¼ fs ðtÞ
R1 Rs C1 L1 C1 Rs C1
:: 1 1 1 : 1 1
e 1 ) V þð þ Þ V þ V ¼ V_ S ðtÞ
1 1 1 1
then /ðxÞ ¼ 1; a ¼ ½ð þ Þ ; ! 1ðL1 C1 1Þ
R1 Rs C1 L1 C1
/ðxÞ ¼ 1 [ 0 8 jtj [ 1 sec; fs ðt) is T periodic and a; b are non
1 1 1
negative parameters: a ¼ ð þ Þ C1 ; b ¼
R1 Rs Rs C1
Unforced investigation: k ¼ 0 ) Rs 1
! 0 ) Rs ! 1 then we return to
Passive RFID TAG since the battery has a very high serial resistance—discon-
nected status. Active RFID equivalent circuit total TAG power is a summation of all
power elements.
X
N X
N
1
Ptotal ¼ pi ¼ TAG Power, pi ¼ ½qðmÞ PTrx þ UðmÞ PTtx þ ZðmÞ PTrx
i¼1 i¼1
m
X
N Zt Zt
0 0
pi ¼ pRs þ pC1 þ pR1 þ pL1 ; energy ) Wðt0 ; tÞ , pðt Þdt ¼ vðt0 Þ iðt0 Þdt0
i¼1
t0 t0
dWðt0 ; tÞ d X N
1
PðtÞtotal ¼ ¼ ½ wi ; energy ) winductor ¼ L I 2
dt dt i¼1 2
Q2
energy ) wcapacitor ¼ ; Presistor ¼ I 2 R; PR1 ¼ IR21 R1 ; PRs ¼ IRs
2
Rs
2C
1 d :
energy ) wL1 ¼ L1 IL21 ) PL1 ¼ wL1 ¼ L IL1 IL1
2 dt
:
Q2 d Q C QC 1
energy ) wC1 ¼ C1 ) PC1 ¼ wC1 ¼ 1
2 C1 dt C1
C1 VC21 d :
energy ) wC1 ¼ ) PC1 ¼ wC1 ¼ C1 VC1 VC1
2 dt
Zt :
VL1 X p QC 1 QC 1
N
1 : :
IL 1 ¼ VL1 dt ) IL1 ¼ ; i ¼ IR1 R1 þ IRs Rs þ L IL1 IL1 þ
2 2
L1 L1 i¼1 C1
0
X
N 2
½V VsðtÞ2 : :
pi ¼ V þ þ L IL1 IL1 þ C1 VC1 VC1
i¼1
R1 Rs
X Zt
N
1 1 2 V VsðtÞ ½VsðtÞ2 V :
pi ¼ V ½ þ
2
þ þ Vdt þ C1 V V
i¼1
R1 Rs Rs Rs L1
0
1 1 1
½qðmÞ PTrx þ UðmÞ PTtx þ ZðmÞ PTrx ¼ V 2 ½ þ
m R1 Rs
2 Zt
2 V VsðtÞ ½Vs ðtÞ V :
þ þ Vdt þ C1 V V
Rs Rs L1
0
dV1 dV2 1 1 1 1
¼ V2 ; ¼ ½ þ V2 V1 þ V_ S ðtÞ
dt dt C1 R1 Rs C1 C1 L1 Rs C1
Now we consider linear system: dV dt ¼ f ðV1 ; V2 Þ; dt ¼ gðV1 ; V2 Þ

1 dV2
And suppose that ðV1 ; V2 Þ is a fixed point: f ðV1 ; V2 Þ ¼ 0; gðV1 ; V2 Þ ¼ 0

Let U1 ¼ V1 V1 ; U2 ¼ V2 V2 Denote the components of a small disturbance
from the fixed point. To see whether the disturbance grows or decays, we need to
derive differential equations of U1 and U2. Let’s do the U1 equation first: dU

dt ¼ dt
1 dV1
Since V1 being constant.
dU1 dV1
¼ ¼ f ðU1 þ V1 ; U2 þ V2 Þ ¼ f ðV1 ; V2 Þ
dt dt
@f @f
þ U1 þ U2 þ OðU12 ; U22 ; U1 U2 Þ
@V1 @V2
@f @f
(Taylor series expansion). To simplify the notation, we have written @V 1
and @V2

these partial derivatives are to be evaluated at the fixed point ðV1 ; V2 Þ; thus they are
numbers, not functions. Also the shorthand notation OðU12 ; U22 ; U1 U2 Þ denotes
quadratic termss in U1 and U2. Since U1 and U2 are small, these quadratic terms are
extremely small. Similarly, we find
@g @g
dt ¼ U1 @V1 þ U2 @V2 þ OðU1 ; U2 ; U1 U2 Þ, Hence the disturbance (U1, U2)
dU2 2 2
0 dU 1
1 ! !
@f @f
B dt C @V1 @V2 U1
evolves according to @ A ¼ @g @g þ Quadratic terms.
dU2 @V1 @V2 U2
dt !
@f @f
@V1 @V2
The Matrix A ¼ @g @g is called the Jacobian matrix at the fixed
@V1 @V2 ðV1 ;V2 Þ
point ðV1 ; V2 Þ and the Quadratic terms are tiny, it’s tempting to neglect them
altogether. If we do that, we obtain the linearized system.
0 dU 1
1 ! !
@f @f
B dt C @V1 @V2 U1
@ A¼ @g @g
dU2 @V1 @V2 U2
dt
Who’s dynamic can be analyzed by the general methods.
1 1 1 1
f ðV1 ; V2 Þ ¼ V2 ; gðV1 ; V2 Þ ¼ ½
þ V2 V1 þ V_ S ðt)
C1 R1 Rs C1 C1 L1 Rs C1
@f @f @g 1 @g 1 1
¼ 0; ¼ 1; ¼ ; ¼ ð þ Þ
@V1 @V2 @V1 C1 L1 @V2 C1 R1 Rs C1
0 dU 1
1 !
B dt C 0 1 U1
@ A¼ C11L1 ½C11R1 þ

RC1
1
dU2 U2
dt
The basic Active RFID Forced Van Der Pol’s equation

:: 1 1 1 : 1 1
e 1 ) V þð þ Þ V þ V ¼ V_ S ðtÞ
1 1 1 1 1
then /ðxÞ ¼ 1; a ¼ ½ð þ Þ ; ! 1ðL1 C1 1Þ; b ¼
In our case /ðVÞ ¼ 1; /ðVÞ [ 0 for jVj [ 1 and V_ S ðtÞ is T periodic

and,ðR11 þ Rs
1
Þ C11 , RsC
1
1
is non-negative parameters. It is convenient to rewrite the
Active RFID forced Van Der Pol’s equation as an autonomous system.
dh : 1 1 1 : 1
h¼t) ¼ 1; V ¼ Y ð þ Þ /ðVÞ; Y ¼ V þ V_ S ðhÞ
dt R1 Rs C1 R1 C1
:
h ¼ 1; ðV; Y; hÞ 2 R2 x S1 :
/ðVÞ ¼ 1 remain strictly positive as jVj ! 1 for unforced system, R11C1

V_ S ðhÞ ! 0 but R11C1 6¼ 0 then V_ S ðhÞ ¼ 0 no energy is supplied to the Active RFID
TAG, become Passive RFID TAG. First, we suppose that a 1ððR11 þ Rs 1
Þ C11 1Þ
is a small parameter, so the autonomous system is a perturbation of linear oscillator.
: :
V ¼ Y; Y ¼ V Has a phase plane filled with circular periodic orbits each period
of 2 p. Using regular perturbation or averaging methods, we can show that pre-
cisely one of these orbits is preserved under the perturbation. Selecting the
invertible transformation:
! !
n1 cosðtÞ sinðtÞ V
¼
n2 sinðtÞ cosðtÞ Y
Which “freezes” the unperturbed system and

The autonomous system become:
: 1 1 1
n1 ¼ ð þ Þ cos t ½ðn1 cosðtÞ n2 sinðtÞÞ3 =3 ðn1 cosðtÞ n2 sinðtÞÞ
R1 Rs C1
: 1 1 1
n2 ¼ ð þ Þ sin t ½ðn1 cosðtÞ n2 sinðtÞÞ3 =3 ðn1 cosðtÞ n2 sinðtÞÞ
R1 Rs C1
This transformation is orientation reversing approximations the function n1; n2

: :
which vary slowly because n1 ; n2 being small. Integrating each function with
respect to time (t) from 0 to T ¼ 2 p, holding n1; n2 fixed we obtain:
: 1 1 1
n1 ¼ ð þ Þ n ½1 ðn1 2 þ n2 2 Þ=4=2
R1 Rs C1 1
: 1 1 1
n2 ¼ ð þ Þ n ½1 ðn1 2 þ n2 2 Þ=4=2
R1 Rs C1 2
2
This system is correct at first order, but there is an error of Oð½ðR11 þ RsÞ
1
C11 Þ .
In polar coordinates, we therefore have
: r 1 1 1 r2 1 1 1 2
r¼ ð þ Þ ð1 Þ þ Oð½ð þ Þ Þ
2 R1 Rs C1 4 R1 Rs C1
: 1 1 1 2
u ¼ 0 þ Oð½ð þ Þ Þ
R1 Rs C1
2
Neglecting the Oð½ðR11 þ Rs
1
Þ C11 Þ terms this system has an attracting circle of
fixed points at r = 2 reflecting the existence of a one parameter family of almost
sinusoidal solutions: V ¼ rðtÞ cosðt þ uðtÞÞ with slowly varying amplitude
1 1 1 2 1 1 1 2
rðtÞ ¼ 2 þ Oð½ð þ Þ Þ; uðtÞ ¼ u0 þ Oð½ð þ Þ Þ
R1 Rs C1 R1 Rs C1
2
1 1 1
uðtÞ ¼ u0 þ Oð½ð þ Þ Þ
R1 Rs C1
Constant u0 is being determined by initial conditions.

When the value of ðR11 þ Rs 1
Þ C11 is not small the averaging procedure no longer
works and other methods must be used. The investigation can be done for Active
RFID’s system forced Van Der Pole. Let’s consider V_ S ðtÞ 6¼ 0 we suppose
ðR11 þ Rs
1
Þ C11 ; RsC
1
1
1 and use the same transformation as we use in the unforced
_
system V S ðtÞ 6¼ 0 . When our interest in the periodic forced response we use the 2p x
periodic transformation [2–4].
! !
n1 cosðxtÞ x1 sinðxtÞ V
¼
n2 sinðxtÞ x1 cosðxtÞ Y
: 1 1 1 x2 1
n1 ¼ ð þ Þ /ðVÞ cosðx tÞ ð Þ V sinðx tÞ
R1 Rs C1 x
1
sinðx t V_ S ðtÞÞ
Rs C1 x
: 1 1 1 x2 1
n2 ¼ ð þ Þ /ðVÞ sinðx tÞ ð Þ V cosðx tÞ
R1 Rs C1 x
1
cosðx t V_ S ðtÞÞ
Rs C1 x
1
C1 L1 ! 1; /ðVÞ ¼ 1 in our case.
: 1 1 1 x2 1 1
n1 ¼ ð þ Þ cosðx tÞ ð Þ V sinðx tÞ sinðx t V_ S ðtÞÞ
R1 Rs C1 x Rs C1 x
: 1 1 1 x2 1 1
n2 ¼ ð þ Þ sinðx tÞ ð Þ V cosðx tÞ cosðx t V_ S ðtÞÞ
R1 Rs C1 x Rs C1 x
Active RFID TAG system can be represented as Voltage source (internal

resistance), Parallel Resistor, Capacitor, and Inductance circuit. Linear bifurcation
system explains Active RFID TAG system behavior for any initial condition V(t)
and dV(t)/dt . Active RFID’s Coil is a very critical element in Active RFID TAG
functionality. Optimization can be achieved by Coil’s parameters inspection and
System bifurcation controlled by them. Spiral, Circles, and other Active RFID
phase system behaviors can be optimized for better Active RFID TAG performance
and actual functionality. Active RFID TAG losses also controlled for best perfor-
mance and maximum efficiency.
1.2 RFID TAG’s Dimensional Parameters Optimization

as Excitable Linear Bifurcation Systems
RFID Equivalent circuits of a Label can be represented as Parallel circuit of

Capacitance (Cpl), Resistance (Rpl), and Inductance (Lpc). The Label measurement
principle is as follows: Label positioned in defining distance to measurement coil,
Low current or voltage source, Measuring of |Z| and Teta of measurement coil,
Resonance frequency fro at Teta = 0, Calculation of unloaded quality factor Q0 out
of measured bandwidth B0. The Coil design procedure is based on three important
steps. The RFID equivalent circuit can be represented as a differential equation
which, depending on variable parameter. The investigation of RFID’s differential
equation based on bifurcation theory [1], the study of possible changes in the
structure of the orbits of a differential equation depending on variable parameters.
We first illustrate certain observations and analyze local bifurcations of an appro-
priate arbitrary scalar differential equation [2]. Since the implicit function theorem
is the main ingredient used in these generalizations, include a precise statement of
this theorem. Additional analyze the bifurcations of a RFID’s differential equation
on the circle. The bifurcation behavior of specific differential equations can be
encapsulated in certain pictures called bifurcation diagrams. Analysis is done for
optimization of RFID TAG’s dimensional parameters to get the best performance.
RFID TAG can be represented as a parallel Equivalent Circuit of Capacitor and
Resistor in parallel. For example, see below NXP/PHILIPS ICODE IC, Parallel
equivalent circuit and simplified complete equivalent circuit of the label (L1 is the
antenna inductance) [7, 85] (Fig. 1.8).
1.2 RFID TAG’s Dimensional Parameters Optimization … 15
I-CODE RFID
TAG
LA LB
Fig. 1.8 NXP/PHILIPS ICODE IC, Parallel equivalent circuit and simplified complete equivalent
circuit of the label (L1 is the antenna inductance)
C1 ¼ Cic þ Ccon þ Cc; R1 ¼ ðRic RpcÞ=ðRic þ RpcÞ:

Zt1
dIl dVc 1
Vl1 ¼ L ; Ic1 ¼ C ; Il1 ¼ Vl1 dt
dt dt L1
0
X
i¼3 Zt¼t1
V dV 1
Ii ¼ 0; þ C1 þ V dt ¼ 0
i¼1
R1 dt L1
t¼0
1 dV d2V 1
þ C1 2 þ V ¼0
R1 dt dt L1
We get differential equation of RFID TAG system which describe the evolution
of the system in continues time. V = V(t).
Now I define the following Variable setting definitions:, And get the dynamic
dt ¼ V2 , dt ¼ C1 R1 V2 C1 L1 V1
equation system: dV 1 dV2 1 1
The system shape is as nonlinear system equations:
dV1 dV2
¼ f1 ðV1 ; V2 . . .; VnÞ;
dt dt
The V1 and V2 variables are the phase space dimension two. Now Let’s Move to
three variable system—which the time (t) is the third variable, V3 = t (Fig. 1.9).
dV1 dV2 1 1 dV3

¼ V2; ¼ V1 V2 ; ¼1
dt dt C1 L1 C1 R1 dt
d ¼ 2 ðt þ wÞ=p; Aavg ¼ a0 Nc ðg þ wÞ; Bavg ¼ b0 Nc ðg þ wÞ
a0, b0—Overall dimensions of the coil. Aavg, Bavg—Average dimensions of the

coil. t—Track thickness, w—Track width, g—Gap between tracks. Nc—Number of
turns, d—Equivalent diameter of the track. Average coil area; −Ac = Aavg Bavg.
Integrating all those parameters gives the equations for inductance calculation:
Aavg
A0
w B0 Bavg
Fig. 1.9 RFID’s coil dimensional parameters
!
2 Aavg Bavg
!
2 Aavg Bavg
ffi
X3 ¼ 2 Aavg þ Bavg Aavg þ Bavg ; X4 ¼ ðAavg þ BavgÞ=4
2 2
The RFID’s coil calculation inductance expression is

hl i
Lcalc ¼ 0
½X1 þ X2 X3 þ X4 Ncp ; L1 ¼ Lcalc
p
Definition of limits, Estimations: Track thickness t, Al and Cu coils (t > 30 lm).

The printed coils as high as possible. Estimation of turn exponent p is needed for
inductance calculation.

Wired 1.8–1.9
Etched 1.75–1.85
Printed 1.7–1.8
Now I integrate the Lcalc value inside the differential equations which charac-
terize the RFID system with the Coil inductance.
dV1
¼ 0 V1 þ 1 V2 þ 0 V3
dt
dV2 1 1
¼ V1 V2 þ 0 V3
dt C1 lp0 ½X1 þ X2 X3 þ X4 Ncp C1 R1
dV3
¼ 0 V1 þ 0 V2 þ 0 V3 þ 1
dt
The above differential equations can be represented as Matrix formulation:

2 3 2 3 2 3 2 3
dV1 0 1 0 V1 0
6 dV
dt
7 4 f l 1
g f C11R1 g 0 5 4 V2 5 þ 4 0 5
4 dt2 5¼ C1 ½ p0 ½X1 þ X2 X3 þ X4 Ncp
dV3 V3 1
dt 0 0 0
2 3
2 3 0 1 0 2 3 2 3
dV1
6 g 07 V1 0
7 6 f f C11R1 g
1
6 dV
dt
P 7 4 5 4 5
4 dt2 5¼6 7 V2 þ 0
4
l0
4 C1 p ðX3 þ Xk ÞNc p
5
dV3 k¼1;k6¼3 V3 1
dt
0 0 0
And denote the matrix’s elements as functions K1 and K2 of Coil overall

parameters.
1
K1 ¼ K1 ða0 ; b0 ; w; g; d; NC ; t; p; C1 ; R1 Þ ¼ f l0 g
C1 p ½X1 þ X2 X3 þ X4 Ncp
1
K2 ¼ K2 ða0 ; b0 ; w; g; d; NC ; t; p; C1 ; R1 Þ ¼ f g
C1 R1
Denote the RFID Matrix systems with those K1, K2 parameter function gives:
dV1
0 1 V1
dt
dV2 ¼
dt
K1 K2 V2
Now the consideration of trajectories of the form [3]: VðtÞ ¼ ekt S, Where
S <> 0 is some fixed vector to be determined, and k is a growth rate, also to be
determined. If a such solution exists, they correspond to exponential motion along
the line spanned by the vector S. To find the condition on S and k, we substitute
: 0 1
VðtÞ ¼ ekt S into V ¼ A V; A ¼ and obtain k ekt S ¼ ekt A S
K1 K2
and cancellation the nonzero scalar factor ekt yields to k S ¼ A S which state that
the desired straight line solutions exist if S is an eigenvector of A with corre-
sponding eigenvalue k and the solution is Eigen solution. The eigenvalues of a
matrix A are given by the characteristic equation detðA k IÞ ¼ 0 when I is the
1 0
identity matrix I ¼ , we get
0 1
D ¼ detðAÞ ¼ 0 K1 ¼ K1
s ¼ traceðAÞ ¼ 0 þ K2 ¼ K2
k2 s k þ D ¼ 0
k2 K2 k K1 ¼ 0
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
1 1
k1;2 ¼ K2
K22 þ 4 K1
2 2
The above k1;2 is a quadratic solution. The typical solution is for the eigenvalues
are distinct k1 6¼ k2 . In this case, a theorem of linear algebra states that the cor-
responding eigenvectors S1 and S2 are linearly independent, and hence span the
entire plane. Any initial condition V0 can be written as a linear combination of
eigenvectors, V0 ¼ C1 S1 þ C2 S2 . Then the general solution for V(t) it is simply
VðtÞ ¼ C1 ek1 t S1 þ C2 ek2 t S2 . By insertion quadratic solutions into the last
V(t) equation we get
pffiffiffiffiffiffiffiffiffiffiffiffiffiffi pffiffiffiffiffiffiffiffiffiffiffiffiffiffi
VðtÞ ¼ C1 e½2K2 þ 2 K2 þ 4K1 t S1 þ C2 e½2K2 2 K2 þ 4K1 t S2
1 1 2 1 1 2
RFID TAG which gives the best performance is one that his equivalent circuit
(Capacitor, Resistor, and Inductance (Antenna) in parallel), and his Voltage/Voltage
derivative respect to time phase plane converge (Spiral converge, fixed point
respect to the origin, etc.,)
k2 \k1 \0
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
1 1 1 1
K2 K22 þ 4 K1 \ K2 þ K22 þ 4 K1 \0
2 2 2 2
k1 \0. . . ! . . . K2 [ K22 þ 4 K1
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
2
1 1 1
f g [ f g þ 4 f l0 g
C1 R1 C1 R1 C1 p ½X1 þ X2 X3 þ X4 Ncp
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
2
1 1 4
[ f g f g
C1 R1 C1 R1 C1 lp0 ½X1 þ X2 X3 þ X4 Ncp
k2 \k1
1 1 1 1
K2 K22 þ 4 K1 \ K2 þ K22 þ 4 K1
2 qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
2 2 2
0\ K22 þ 4 K1 ! K22 þ 4 K1 [ 0 ! K22 [ 4 K1
2
1 1
f g [ 4 f l g
C1 R1 C1 p ½X1 þ X2 X3 þ X4 Ncp
0
2
1 4
f g [ l0
C1 R1 C1 p ½X1 þ X2 X3 þ X4 Ncp
Then both Eigen solutions decay exponentially. The fixed point is a stable node,
except eigenvectors are not mutually perpendicular, in general. Trajectories typi-
cally approach the origin tangent to the slow Eigen direction, defined as the
direction spanned by the eigenvector with the smaller jkj . In backward time t ! 1
the trajectories become parallel to the fast Eigen direction [2–4] (Fig. 1.10).
If we reverse all the arrows in the above figure, we obtain a typical phase portrait
for an unstable node. Now I investigate the case when eigenvalues are complex
number. If the eigenvalues are complex, the fixed point is either a center or a spiral.
The origin is surrounded by a family of closed orbits. Note that centers are neutrally
stable, since nearby trajectories are neither attracted to nor repelled from the fixed
point. A spiral would occur if the RFID system were lightly damped. Then the
trajectory would just fail to close, because the RFID system loses a bit of energy on
each cycle. To justify these statements, recall that the eigenvalues are
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
k1;2 ¼ 12 K2
12 K22 þ 4 K1 ; K22 þ 4 K1 \0
To simplify the notation, let’s write the eigenvalues as
1 1
k1;2 ¼ a
i x ; a ¼ K2 ; x ¼ K22 þ 4 K1
2 2
x 6¼ 0 9 VðtÞ ¼ C1 e½2 2 2 2
K þ 1
S1 þ C2 e½2K2 2 K2 þ 4K1 t S2
1 1
K 2 þ 4K t 1 1 2
C0 s; S0 s complex, since; k0 s complex

VðtÞ ¼ C1 e½a þ ixt S1 þ C2 e½aixt S2
Euler’s formula ! e½ixt ¼ cos½x t þ i sin½x t
Hence V(t) is a combination of terms involving
eat cos½x t; eat sin½x t
Such terms represent exponentially decaying oscillations if a ¼ ReðkÞ\0, And

growing if a [ 0 . The corresponding fixed points are stable and unstable spirals,
Fig. 1.10 Voltage/Voltage V2

derivative respect to time
converge after the reader
carrier signal end Case 1
V1
Fig. 1.11 For both centers

and spirals, rotation is
clockwise or Counter
clockwise
respectively. If the eigenvalues are pure imaginary a ¼ 0, then all the solutions are
periodic with period T ¼ 2px . The oscillators have fixed amplitude and the fixed
point is centered. For both centers and spirals, it’s easy to determine whether the
rotation is clockwise or counterclockwise (Fig. 1.11).
1 1
a¼ K2 ¼ f g
2 2 C1 R1
1 1
Decaying oscillators 8 a\0 ! f g\0 ! [0
2 C1 R1 2 C1 R1
1 1
Growing oscillators 8 a [ 0 ! f g[0 ! \0
2 C1 R1 2 C1 R1
C1, R1 > 0 always then only the first behavior, decaying oscillator can exist in
our RFID system. In all analysis until now, we have been assuming that the
eigenvalues are distinct. What happens if the eigenvalues are equal? Suppose
eigenvalues are equal k1 ¼ k2 ¼ k, then there are two possibilities: either there are
two independent eigenvectors corresponding to k, or there’s only one. If there are
two independent eigenvectors, then they span the plane and so every vector is an
eigenvector with this same eigenvalue k . To see this, let’s write an arbitrary vector
X0 as a linear combination of the two eigenvectors: X0 = C1 S1 + C2 S2.
Then A X0 ¼ A ðC1 S1 þ C2 S2 Þ ¼ C1 k S1 þ C2 k S2 ¼ k X0
X0 is also an eigenvector with eigenvalue k . Since the multiplication by A
simply stretches
every vector by a factor k, the matrix must be a multiple of the
k 0
identity: A ¼ then if k 6¼ 0, all trajectories are straight lines through the
0 k
origin XðtÞ ¼ ekt X0 and the fixed point is a star node. On the other hand, if k = 0
the whole plane is filled with fixed points. Let’s now sketch the above options with
RFID Overall parameter restriction. k1 ¼ k2 ¼ k 6¼ 0 then (Fig. 1.12)
1 1 1 1
K2 þ K22 þ 4 K1 ¼ K2 K22 þ 4 K1
2 ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
q 2 2 2
K22 þ 4 K1 ¼ 0 ! K22 þ 4 K1 ¼ 0 ! K22 ¼ 4 K1
l0
½X1 þ X2 X3 þ X4 Ncp ¼ C1 4 R21
p
Fig. 1.12 Voltage/Voltage V2

derivative respect to time
Converge after the reader
Carrier signal end Case 2
V1
Now let’s summarize the classification of fixed points in RFID system based on
all investigation I did. It is easy to show the type and stability of all the different
fixed points on a single diagram [4] (Figs. 1.13 and 1.14).
pffiffiffiffiffiffiffiffiffi
s2 4 D ¼ K22 þ 4 K1 ¼ 0; ! K2 ¼ 2 K1
s ¼ traceðAÞ ¼ K2 ; D ¼ detðAÞ ¼ K1
1 h pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffii
k2 K2 k K1 ¼ 0; k1;2 ¼ s
s2 4 D
2
s ¼ k1 þ k2 ¼ K2 ; D ¼ k1 k2 ¼ K1
Charecteristic equation : ðk k1 Þ ðk k2 Þ ¼ k2 s k þ D ¼ 0
Fig. 1.13 Stable/Unstable Unstable

diagram nodes
Unstable
spirals
Saddle
points Centers
stable
spirals
stable
nodes
Δ < 0....( K1 > 0) Δ > 0....( K1 < 0) Δ = 0....( K1 = 0)
The eigenvalues are The eigenvalues are At least one of the

real and have opposite either real with the same eigenvalues is zero.
sign hence the fixed sign (nodes), or complex Then the origin is not an
point is a saddle point, conjugate (spiral & isolated fixed point.
centers). There is either a whole
line of a fixed point , or a
plane of fixed point
Fig. 1.14 Stable Unstable diagram as function of K1
Nodes satisfy s2 4 D [ 0 and spirals satisfy s2 4 D\0 . The parabola

s 4 D ¼ 0 is the borderline between nodes and spirals. Star nodes and
2
degenerate nodes live on this parabola. The stability of the nodes and spirals is
determined by s value. When s\0, both eigenvalues have negative real parts, so
the fixed point is stable. Unstable spirals and nodes have s [ 0 . Neutrally stable
centers live on the borderline s ¼ 0, where eigenvalues are purely imaginary [5].
1.3 RFID TAGs Coil’s System Stability Optimization

Under Delayed Electromagnetic Interferences
The RFID TAG system has two main variables TAG’s voltage and TAG’s voltage
derivative respect to time. Due to electromagnetic interferences those variables have
delays in time domain. We define s1 as a time delay respect to TAG’s voltage and
s2 as a time delay respect to TAG’s voltage derivative. RFID Equivalent circuits of
a Label can be represented as Parallel circuit of Capacitance (Cpl), Resistance
(Rpl), and Inductance (Lpc). Our RFID TAG system delay differential and delay
different model can be analytically by using delay differential equations in
dynamically. The need of the incorporation of a time delay is often of the existence
of any stage structure. It is often difficult to analytically study models with delay
dependent parameters, even if only a single discrete delay is present. There are
practical guidelines that combine graphical information with analytical work to
ters. The stability of a given steady state is simply determined by the graphs of
some function of s1, s2 which can be expressed, explicitly and thus can be easily
depicted by Matlab and other popular software. We need only look at one such
function and locate the zero. This function often has only two zeroes, providing
thresholds for stability switches. As time delay increases, stability changes from
stable to unstable to stable. We emphasize the local stability aspects of some models
with delay dependent parameters. Additionally, there is a general geometric
1.3 RFID TAGs Coil’s System Stability … 23
criterion that, theoretically speaking, can be applied to models with many delays, or
even distributed delays. The simplest case of a first order characteristic equation,
providing more user friendly geometric and analytic criteria for stability switches.
The analytical criteria provided for the first and second order cases can be used to
obtain some insightful analytical statements and can be helpful for conducting
simulations. RFID TAG can be represented as a parallel Equivalent Circuit of
Capacitor and Resistor in parallel. For example, see below NXP/PHILIPS
ICODE IC, Parallel equivalent circuit and simplified complete equivalent circuit
of the label (L1 is the antenna inductance) [6, 8] (Fig. 1.15).
1 dV d2 V 1
þ C1 2 þ V ¼0
R1 dt dt L1
We get differential equation of a RFID TAG system which describe the evo-
lution of the system in continues time. V = V(t). Now I define the following
Variable setting definitions: V2 ¼ dV dt ¼ dt ; V1 ¼ V. The dynamic equation system:
1 dV
dt ¼ V2 ; dt ¼ C1 R1 V2 C1 L1 V1 (Fig. 1.16)

dV1 dV2 1 1
a0, b0—Overall dimensions of the coil. Aavg, Bavg—Average dimensions of the

coil. t—Track thickness, w—Track width, g—Gap between tracks. Nc—Number of
turns, d—Equivalent diameter of the track. Average coil area; −Ac = Aavg . Bavg.
Integrating all those parameters gives the equations for inductance calculation:
Fig. 1.15 NXP/PHILIPS

ICODE IC, Parallel I-CODE RFID TAG
equivalent circuit and
simplified complete LA LB
equivalent circuit of the label
(L1 is the antenna inductance)
Antenn
Fig. 1.16 RFID’s coil Aavg

dimensional parameters
A0
w B0 Bavg
!
2 Aavg Bavg
!
2 Aavg Bavg
ffi
X3 ¼ 2 Aavg þ Bavg Aavg2 þ Bavg2 ; X4 ¼ ðAavg þ BavgÞ=4

The RFID’s coil calculation inductance is Lcalc ¼ lp0 ½X1 þ X2 X3 þ X4 Ncp
L1 = Lcalc. Definition of limits, Estimations: Track thickness t, Al and Cu coils
(t > 30 lm).
dV1 " #
0 1 V1
dt
dV2 ¼ f C l0
1
g f C11R1 g V2
dt ½
1 p ½X1 þ X2 X3 þ X4 Ncp
Due to electromagnetic interferences, we get RFID TAG’s voltage and voltage

derivative with delays s1 and s2 respectively V1(t) ! V1(t − s1); V2(t) !
V2(t − s1). We consider no delay effect on dV1/dt and dV2/dt. The RFID TAG’s
differential equations under electromagnetic interference effects (we consider
electromagnetic interferences (delay terms) influence only RFID TAG voltage V1(t)
and voltage derivative V2(t) respect to time, there is no influence on dV1(t)/dt and
dV2(t)/dt):
dV1
¼ V2 ðt s2 Þ
dt
dV2 1 1
¼ f g V1 ðt s1 Þ V2 ðt s2 Þ
dt C1 lp0 ½X1 þ X2 X3 þ X4 Ncp C1 R1
To find the Equilibrium points (fixed points) of the RFID TAG system is by
lim V1 ðt s1 Þ ¼ V1 ðtÞ; lim V2 ðt s2 Þ ¼ V2 ðtÞ;

t!1 t!1
dV1 ðtÞ dV2 ðtÞ
¼ 0; ¼0
dt dt
8 t s1 ; t s2 9 ðt s1 Þ t; ðt s2 Þ t; t ! 1
ð0Þ ð0Þ
We get two equations and the only fixed point is E ð0Þ ðV1 ; V2 Þ ¼ ð0; 0Þ.
Stability analysis: The standard local stability analysis about any one of the equi-
librium points of RFID TAG system consists in adding to coordinate [V1 ; V2
arbitrarily small increments of exponential form, and retaining the first order terms
in v1, v2. The system of two homogeneous equations leads to a polynomial char-
acteristic equation in the eigenvalues. The polynomial characteristic equations
accept by set the below voltage and voltage derivative respect to time into two
RFID TAG system equations. RFID TAG system fixed values with arbitrarily small
increments of exponential form ½v1 v2 ekt are: i = 0 (first fixed point), i = 1
(second fixed point), i = 2 (third fixed point).
ðiÞ ðiÞ
V1 ðtÞ ¼ V1 þ v1 ekt ; V2 ðtÞ ¼ V2 þ v2 ekt ;
ðiÞ
V1 ðt s1 Þ ¼ V1 þ v1 ekðts1 Þ
ðiÞ
V2 ðt s2 Þ ¼ V2 þ v2 ekðts2 Þ 8 i ¼ 0; 1; 2
We choose the above expressions for our V1 ðtÞ; V2 ðtÞ as small displacement
[v1 ; v2 ] from the system fixed points in time t = 0.
ðiÞ ðiÞ
V1 ðt ¼ 0Þ ¼ V1 þ v1 ; V2 ðt ¼ 0Þ ¼ V2 þ v2
for k\0; t [ 0 the selected fixed point is stable otherwise k [ 0; t [ 0 is

Unstable. Our system tends to the selected fixed point exponentially for
k\0; t [ 0 otherwise go away from the selected fixed point exponentially. k Is the
eigenvalue parameter which establishes if the fixed point is stable or Unstable,
additionally his absolute value (jkj) establish the speed of flow toward or away from
the selected fixed point [1, 2].
k<0 k>0
t=0 V1(t = 0) = V(i)
1 + v1 V1(t = 0) = V(i)
1 + v1
V2(t = 0) = V(i)
2 + v2 V2(t = 0) = V(i)
2 + v2
t>0 V1(t) = V(i)
1 + v1e−|k|t V1(t) = V(i)
1 + v1e|k|t
−|k|t
V2(t) = V(i)
2 + v2e V2(t) = V(i)
2 + v2e
|k|t
t<0 V1(t ! ∞) = V1 (i)

V1(t ! ∞, k > 0) * v1e|k|•t
V2(t ! ∞) = V(i) 2 V2(t ! ∞, k > 0) * v2e|k|•t
The speeds of flow toward or away from the selected fixed point for RFID TAG
system voltage and voltage derivative respect to time are
ðiÞ ðiÞ
dV1 ðtÞ V1 ðt þ DtÞ V1 ðtÞ V þ v1 ekðt þ DtÞ ½V1 þ v1 ekt
¼ lim ¼ lim 1
dt Dt!0 Dt Dt!0 Dt
v1 ekt ½ekDt 1 ekDt 1 þ kDt v1 ekt ½1 þ k Dt 1
¼ lim ! lim ¼ k v1 ekt
Dt!0 Dt Dt!0 Dt
ðiÞ ðiÞ
dV2 ðtÞ V2 ðt þ DtÞ V2 ðtÞ V þ v2 ekðt þ DtÞ ½V2 þ v2 ekt
¼ lim ¼ lim 2
dt Dt!0 Dt Dt!0 Dt
v2 ekt ½ekDt 1 ekDt 1 þ kDt v2 ekt ½1 þ k Dt 1
¼ lim ! lim ¼ k v2 ekt
Dt!0 Dt Dt!0 Dt
And the time derivative of the above equations:
dV1 ðtÞ dV2 ðtÞ

¼ v1 k ekt ; ¼ v2 k ekt ;
dt dt
dV1 ðt s1 Þ
¼ v1 k ekðts1 Þ ¼ v1 k ekt es1 k
dt
dV2 ðt s2 Þ
¼ v2 k ekðts2 Þ ¼ v2 k ekt es2 k
dt
dV1
First, we take the RFID TAG’s voltage (V1) differential equation: ¼ V2 and
dt
adding to its coordinates [V1V2]. Arbitrarily small increments of exponential form
½v1 v2 ekt and retaining the first order terms in v1, v2.
ðiÞ ði¼0Þ
k v1 ekt ¼ V2 þ v2 ekt ; V2 ¼ 0 ; k v1 þ v 2 ¼ 0
dV2
Second, we take the RFID TAG’s voltage (V2) differential equation: ¼
dt
1 1
f hl ig V1 ðtÞ V2 ðtÞ and adding to its
C1 0
½X1 þ X2 X3 þ X4 N c
p C1 R1
p
coordinates [V1V2] arbitrarily small increments of exponential form ½v1 v2 ekt and
retaining the first order terms in v1, v2.
dV2 1 1
¼ f hl ig V1 ðtÞ V2 ðtÞ
dt C1 0
½X1 þ X2 X3 þ X4 N pc C1 R1
p
1
kt
k v2 e ¼ f hl ig ðV1ðiÞ þ v1 ekt Þ
C1 0
½X1 þ X2 X3 þ X4 N cp
p
1 ðiÞ
ðV2 þ v2 ekt Þ
C1 R1
ðiÞ ðiÞ
V1 ¼ 0; V2 ¼ 0
1 1
k v2 f l0 g v1 v2 ¼ 0
C1 ½ ½X1 þ X2 X3 þ X4 Ncp C1 R1
p
We can summery our system eigenvalues equations: k v1 þ v2 ¼ 0
1 1
f l0 g v1 k v2 v2 ¼ 0
C1 ½ ½X1 þ X2 X3 þ X4 Nc p C 1 R1
p
0 1 !
k 1
v
@ f l 1
g k C11R1 A
1
¼0
C1 ½
0
½X1 þ X2 X3 þ X4 Ncp v2
p
0 1
k 1
A k I ¼ @ f l 1
g k C11R1 A ;
C1 ½
0
½X1 þ X2 X3 þ X4 Ncp
p
detðA k IÞ ¼ 0
1 1
k ðk þ Þþ l0 ¼0
C1 R1 C1 ½ ½X1 þ X2 X3 þ X4 Ncp
p
We get two eigenvalues: k1 ; k2 . If k1 \0; k2 \0 then we have stable node.

If k1 [ 0; k2 [ 0 then we have unstable node. If k1 k2 \0 then we have saddle
point.
If k1 ¼ k2 \0 then we have attracting focus. If we have k1 \k2 ¼ 0 then we
have attracting line. If we have k1 ¼ 0\k2 then we have repelling line. If we have
0\k1 ¼ k2 then we have repelling focus. If k1 ; k2 are complex conjugate and the
real part is negative than we have attracting spiral otherwise (positive real part)
repelling spiral. If the real part is zero then we have a center. We define
ðiÞ ðiÞ
V1 ðt s1 Þ ¼ V1 þ v1 ekðts1 Þ ; V2 ðt s2 Þ ¼ V2 þ v2 ekðts2 Þ :
Then we get two delayed differential equations respect to adding to its coordi-
nates [V1V2] arbitrarily small increments of exponential form½v1 v2 ekt .
ðiÞ ði¼0Þ
v1 k ekt ¼ V2 þ v2 ekðts2 Þ ; V2 ¼ 0 ) v1 k ekt ¼ v2 ekðts2 Þ
1 1
k v2 ekt ¼ f l0 g V1ðiÞ V
ðiÞ
C1 p ½X1 þ X2 X3 þ X4 Nc p C1 R1 2
1
þ f l g v1 ekðts1 Þ
C1 p ½X1 þ X2 X3 þ X4 Ncp
0
1
v2 ekðts2 Þ
C1 R1
ði¼0Þ ði¼0Þ
In the equilibrium fixed point V1 ¼V2 ¼ 0 and in the equilibrium fixed
point
ði¼0Þ ði¼0Þ 1 1
V1 ¼ V2 ¼ 0; f l0 g V1ðiÞ ðiÞ
V2 ¼ 0
C1
p ½ X1 þ X2 X 3 þ X 4 Nc p C 1 R 1
1 1
k v2 ekt ¼ f g v1 e kðts 1 Þ
v2 ekðts2 Þ
C1 lp0 ½X1 þ X2 X3 þ X4 Ncp C 1 R1

We define f# ðX1 ; X2 ; etc:. . .Þ ¼ lp0 ½X1 þ X2 X3 þ X4 Ncp . The small
increments Jacobian of our RFID TAG system:
" # !
k eks2 v1
¼ 0;
C11f# eks1 C11R1 eks2 k v2
" #
k eks2
AkI ¼
C11f# eks1 C11R1 eks2 k
1 1
det jA k Ij ¼ 0; Dðk; s1 ; s2 Þ ¼ k2 þ k eks2 þ ekðs1 þ s2 Þ
C1 R1 C 1 f#
We have three stability analysis cases: s1 ¼s; s2 ¼ 0 or s2 ¼s; s1 ¼ 0 or s1 ¼

s2 ¼ s otherwise s1 6¼ s2 . We need to get characteristics equations as all above
stability analysis cases. We study the occurrence of any possible stability switching,
resulting from the increase of the value of the time delay s for the general char-
acteristic equation Dðk; sÞ. Dðk; sÞ ¼ Pn ðk; sÞ þ Qm ðk; sÞ eks
Pn
The expression for Pn ðk; sÞ is Pn ðk; sÞ ¼ Pk ðsÞ k ¼ P0 ðsÞ þ
k
k¼0
P1 ðsÞ k þ P2 ðsÞ k þ P3 ðsÞ k þ . . .:
2 3
P
m
The expression for Qm ðk; sÞ is Qm ðk; sÞ ¼ qk ðsÞ kk ¼ q0 ðsÞ þ
k¼0
q1 ðsÞ k þ q2 ðsÞ k2 þ . . ..
First, we analyze RFID Tag system second order characteristic equation for. The
first case we analyze is when there is a delay in RFID Label voltage and no delay in
voltage time derivative [4, 5].
1 1
Dðk; s1 ¼ s; s2 ¼ 0Þ ¼ k2 þ k þ eks1 ;
C1 R1 C1 f#
Dðk; sÞ ¼ Pn ðk; sÞ þ Qm ðk; sÞ eks
The expression for Pn ðk; sÞ:
X
n
1
Pn ðk; sÞ ¼ Pk ðsÞ k ¼ P0 ðsÞ þ P1 ðsÞ k þ P2 ðsÞ k ¼ k þ k ;
k 2 2
k¼0
C1 R1
1
P2 ðsÞ ¼ 1; P1 ðsÞ ¼ ; P0 ðsÞ ¼ 0
C1 R1
P
m
The expression for Qm ðk; sÞ: Qm ðk; sÞ ¼ qk ðsÞ kk ¼ q0 ðsÞ ¼ C 1f . Our RFID sys-
1 #
k¼0
tem second order characteristic equation: Dðk; sÞ ¼ k2 þ aðsÞ k þ bðsÞ k eks þ
cðsÞ þ dðsÞ eks
Then aðsÞ ¼ C11R1 ; bðsÞ ¼ 0; cðsÞ ¼ 0; dðsÞ ¼ C11f# s 2 R þ 0 and aðsÞ; bðsÞ;
cðsÞ; dðsÞ : R þ 0 ! R are differentiable functions of the class C1 ðR þ 0 Þ, such that
cðsÞ þ dðsÞ ¼ C11f# 6¼ 0 for all s 2 R þ 0 and for any s; bðsÞ; dðsÞ are not simulta-
neously zero. We have
1
Pðk; sÞ ¼ Pn ðk; sÞ ¼ k2 þ aðsÞ k þ cðkÞ ¼ k2 þ k
C1 R1
1
Qðk; sÞ ¼ Qm ðk; sÞ ¼ bðsÞ k þ dðsÞ ¼
C1 f#
We assume that Pn ðk; sÞ ¼ Pn ðkÞ and Qm ðk; sÞ ¼ Qm ðkÞ can’t have common
imaginary roots. That is, for any real number; C11f# x2 þ i x C11R1 6¼ 0
Fðx; sÞ ¼ jPði x; sÞj2 jQði x; sÞj2 ¼ ðc x2 Þ2 þ x2 a2 ðx2 b2 þ d 2 Þ
Fðx; sÞ ¼ x4 þ x2 ðC 1
2 ðC 1
2 ; Hence Fðx; sÞ ¼ 0 implies
1 R1 Þ 1 f# Þ
x4 þ x2 ðC 1
2 ðC 1
2 ¼ 0 and its roots are given by
1 R1 Þ 1 f# Þ
1 pffiffiffiffi 1 pffiffiffiffi 1
x2þ ¼ fðb2 þ 2 c a2 Þ þ Dg ¼ f D g;
2 2 ðC1 R1 Þ2
x2 ¼ fðb2 þ 2 c a2 Þ Dg ¼ f D þ g
2 2 ðC1 R1 Þ2
x2 ¼ fðb2 þ 2 c a2 Þ Dg ¼ f D þ g;
2 2 ðC1 R1 Þ2
1 2 1
D ¼ ðb2 þ 2 c a2 Þ 4 ðc2 d 2 Þ ¼ ½ð Þ2 2
C12 f# R1
pffiffiffiffi
Therefore the following holds: 2 x2þ = ðb2 þ 2 c a2 Þ ¼
D;
pffiffiffiffi
2 x2þ = þ 1
ðC1 R1 Þ2
¼
D
Furthermore
1
PR ði x; sÞ ¼ cðsÞ x2 ðsÞ ¼ x2 ðsÞ; PI ði x; sÞ ¼ xðsÞ aðsÞ ¼ xðsÞ
C1 R1
1
QR ði x; sÞ ¼ dðsÞ ¼ ; QI ði x; sÞ ¼ xðsÞ bðsÞ ¼ 0
C1 f#
Hence
PR ði x; sÞ QI ði x; sÞ þ PI ði x; sÞ QR ði x; sÞ
sin hðsÞ ¼
jQði x; sÞj2
PR ði x; sÞ QR ði x; sÞ þ PI ði x; sÞ QI ði x; sÞ
cos hðsÞ ¼
jQði x; sÞj2
ðc x2 Þ x b þ x a d f#
sin hðsÞ ¼ ¼x ;
x b þd
2 2 2 R1
ðc x2 Þ d þ x2 a b
cos hðsÞ ¼ ¼ x2 C1 f#
x 2 b2 þ d 2
Which jointly with x4 þ x2 ðC 1

2 ðC 1
2 ¼0 Defines the maps
1 R1 Þ 1 f# Þ
Sn ðsÞ ¼ s sn ðsÞ; s 2 I; n 2 N0 , that are continuous and differentiable in s based

on Lemma 1.1. Hence we use theorem 1.2. This proves the theorem 1.3 and
theorem 1.4. Remark: a, b, c, d parameters are independent of delay parameter s
even we use aðsÞ; bðsÞ; cðsÞ; dðsÞ. Second, we analyze RFID Tag system second
order characteristic equation for s1 ¼ 0; s2 ¼ s. The second case we analyze is
when there is no delay in RFID Label voltage and there is a delay in voltage time
derivative.
1 1
Dðk; s1 ¼ 0; s2 ¼ sÞ ¼ k2 þ k eks2 þ eks2
C1 R1 C1 f#
1 1
Dðk; s1 ¼ 0; s2 ¼ sÞ ¼ k2 þ ðk þ Þ eks ;
C1 R1 C1 f#
P
n
Pn ðk; sÞ Pn ðk; sÞ ¼ Pk ðsÞ k ¼ P0 ðsÞ þ
k
The expression for is
k¼0
P1 ðsÞ k þ P2 ðsÞ k ¼ k
2 2
P2 ðsÞ ¼ 1; P1 ðsÞ ¼ 0; P0 ðsÞ ¼ 0. The expression for Qm ðk; sÞ is
X
m
1 1
Qm ðk; sÞ ¼ qk ðsÞ kk ¼ k þ ;
k¼0
C1 R1 C1 f#
1 1
q0 ðsÞ ¼ ; q1 ðsÞ ¼ ; q2 ðsÞ ¼ 0
C1 f# C1 R1
Our RFID system second order characteristic equation:
Dðk; sÞ ¼ k2 þ aðsÞ k þ bðsÞ k eks þ cðsÞ þ dðsÞ eks

1 1
aðsÞ ¼ 0; bðsÞ ¼ ; cðsÞ ¼ 0; dðsÞ ¼
C1 R1 C1 f#
And in the same manner like our previous case analysis:
Pðk; sÞ ¼ Pn ðk; sÞ ¼ k2 þ aðsÞ k þ cðkÞ ¼ k2 ;

1 1
Qðk; sÞ ¼ Qm ðk; sÞ ¼ bðsÞ k þ dðsÞ ¼ k þ
C1 R1 C1 f#
imaginary roots. That is, for any real number x; pn ðk ¼ i x; sÞ þ
Qm ðk ¼ i x; sÞ 6¼ 0
1 1
x2 þ i x 6¼ 0;
C 1 f# C1 R1
Fðx; sÞ ¼ x4 x2 ðC 1
2 ðC 1
2
1 R1 Þ 1 f# Þ
Hence Fðx; sÞ ¼ 0 implies x x2 ðC 4 1

2 ðC 1
2 ¼ 0 And its roots are
1 R1 Þ 1 f# Þ
given by
x2þ ¼ fðb2 þ 2 c a2 Þ þ Dg ¼ f D þ g;
2 2 ðC1 R1 Þ2
x2 ¼ fðb2 þ 2 c a2 Þ Dg ¼ f D þ g
2 2 ðC1 R1 Þ2
1 2 1
D ¼ ðb2 þ 2 c a2 Þ 4 ðc2 d 2 Þ ¼ 2 ½ð Þ2 þ 2
C1 f# R1
Therefore the following holds:

pffiffiffiffi 1 pffiffiffiffi
2 x2þ = ðb2 þ 2 c a2 Þ ¼
D; 2 x2þ = þ 2
¼
D
ðC1 R1 Þ
Furthermore
PR ði x; sÞ ¼ cðsÞ x2 ðsÞ ¼ x2 ðsÞ;
1
PI ði x; sÞ ¼ xðsÞ aðsÞ ¼ 0; QR ði x; sÞ ¼ dðsÞ ¼
C 1 f#
1
QI ði x; sÞ ¼ xðsÞ bðsÞ ¼ xðsÞ ;
C1 R1
sin hðsÞ ¼
jQði x; sÞj2
cos hðsÞ ¼ ;
jQði x; sÞj2
ðc x2 Þ x b þ x a d x3 C1 R1
sin hðsÞ ¼ ¼
x2 b2 þ d 2 x2 þ ðRf 1 Þ2 #
ðc x Þ d þ x a b
2 2 x C1 Rf#1
2
cos hðsÞ ¼ ¼
x 2 b2 þ d 2 x2 þ ðRf#1 Þ2
Which jointly with

1 1
x4 x2 2
¼0
ðC1 R1 Þ ðC1 f# Þ2
Defines the maps Sn ðsÞ ¼ s sn ðsÞ; s 2 I; n 2 N0

Defines the maps Sn ðsÞ ¼ s sn ðsÞ; s 2 I; n 2 N0 are continuous and differen-
tiable in s based on Lemma 1.1. Hence we use theorem 1.2. This proves the
theorem 1.3 and theorem 1.4. Remark: a, b, c, d parameters are independent of
delay parameter s even we use aðsÞ; bðsÞ; cðsÞ; dðsÞ [4, 5].
Third, we analyze RFID Tag system second order characteristic equation for
s1 ¼s; s2 ¼ s. The third case we analyze is when there is delay both in RFID Label
voltage and voltage time derivative [4, 5].
1 1
Dðk; s1 ¼ s; s2 ¼ sÞ ¼ k2 þ k eks þ eks2 ;
C1 R1 C1 f#
1 1
Dðk; s1 ¼ s2 ¼ sÞ ¼ k2 þ ðk þ eks Þ eks
C1 R1 C1 f#
The expression for Pn ðk; sÞ is

Xn
Pn ðk; sÞ ¼ Pk ðsÞ k ¼ P0 ðsÞ þ P1 ðsÞ k þ P2 ðsÞ k ¼ k ; P2 ðsÞ ¼ 1;
k 2 2
k¼0
P1 ðsÞ ¼ 0; P0 ðsÞ ¼ 0:
P
m
The expression for Qm ðk; sÞ is Qm ðk; sÞ ¼ qk ðsÞ k k
¼ k C11R1 þ 1
C1 f# eks
k¼0
Taylor expansion: eks 1 k s þ k2 s2
2 since we need n > m [BK] analysis
we choose eks 1 k s.
X
m
1 1 s 1
Qm ðk; sÞ ¼ qk ðsÞ kk ¼ k ð Þþ ;
k¼0
C1 R1 f# C1 f#
1 1 1 s
q0 ðs; kÞ ¼ ; q1 ðsÞ ¼ ð Þ; q2 ðsÞ ¼ 0
C1 f# C1 R1 f#
Our RFID system second order characteristic equation:

1 1 s 1
aðsÞ ¼ 0; bðsÞ ¼ ð Þ; cðsÞ ¼ 0; dðsÞ ¼
C1 R1 f# C1 f#
And in the same manner like our previous case analysis: Pðk; sÞ ¼ Pn ðk; sÞ ¼
k2 ; Qðk; sÞ ¼ Qm ðk; sÞ ¼ k C11 ðR11 fs# Þ þ C11f#
We assume that Pn ðk; sÞ ¼ Pn ðkÞ and Qm ðk; sÞ can’t have common imaginary
roots. That is, for any real number x; pn ðk ¼ i x; sÞ þ Qm ðk ¼ i x; sÞ 6¼ 0
1 1 s 1
x2 þ i x ð Þþ 6¼ 0;
C1 R1 f# C1 f#
Fðx; sÞ ¼ jPði x; sÞj2 jQði x; sÞj2 ; Pði x; sÞ ¼ x2
PR ði x; sÞ ¼ x2 ; PI ði x; sÞ ¼ 0;
1 1 s 1
Qðk ¼ i x; sÞ ¼ i x ð Þ þ
C1 R1 f# C1 f#
1 1 s 1
QI ðk ¼ i x; sÞ ¼ x ð Þ; QR ðk ¼ i x; sÞ ¼
C1 R1 f# C 1 f#
jPði x; sÞj2 ¼ P2I þ P2R ; jQði x; sÞj2 ¼ Q2I þ Q2R ;
jPði x; sÞj2 ¼ P2I þ P2R ¼ x4
1 1 s 1
jQði x; sÞj2 ¼ x2 2 ð Þ2 þ ;
C1 R1 f# ðC1 f# Þ2
1 1 s 1
Fðx; sÞ ¼ x4 x2 2 ð Þ2
C1 R1 f# ðC1 f# Þ2
Hence Fðx; sÞ ¼ 0 implies x4 x2 C12 ðR11 fs# Þ2 ðC 1

2 ¼0
1 1 f# Þ
1 1 s 1 1 s
Fx ¼ 4 x3 2 x ð Þ2 ¼ 2 x ½2 x2 2 ð Þ2 ;
C12 R1 f# C 1 R 1 f#
2 x2 1 s
Fs ¼ ð Þ
C12 f# R1 f#
PIx ¼ 0; PRx ¼ 2 x;
1 1 s
QIx ¼ ½ ; QRx ¼ 0; PIs ¼ 0; PRs ¼ 0
C1 R1 f#
x
QIs ¼ ; QRs ¼ 0
C1 f#
The expressions for U, V can be derived easily [BK]:
U ¼ ðPR PIx PI PRx Þ ðQR QIx QI QRx Þ;

V ¼ ðPR PIx PI PRx Þ ðQR QIx QI QRx Þ
x 1 s 1 Fs
V ¼ 2 2 ;U ¼ 2 ½ ; xs ¼
C1 f# C1 f# f# R1 Fx
x
C12 f#
ðR11 fs# Þ
And we get the expression: xs ¼
2 x2 C12 ðR11 fs# Þ2
1
Defines the maps Sn ðsÞ ¼ s sn ðsÞ; s 2 I; n 2 N0
Defines the maps Sn ðsÞ ¼ s sn ðsÞ; s 2 I; n 2 N0 that are continuous and dif-
ferentiable in s based on Lemma 1.1 (see Appendix A). Hence we use theorem 1.2.
This proves the theorem 1.3 and theorem 1.4 (see Appendix D).
Remark Taylor approximation for eks 1 k s giving us a good stability
analysis, approximation only for a restricted delay time interval.
Now we discuss RFID TAG system stability analysis under delayed variables in
time. Our RFID homogeneous system for v1, v2 leads to a characteristic equation
for the eigenvalue k having the form PðkÞ þ QðkÞ eks ¼ 0; first case
s1 ¼s; s2 ¼ 0. Dðk; s1 ¼ s; s2 ¼ 0Þ ¼ k2 þ k C11R1 þ C11f# eks1 . We use different
parameters terminology from our last characteristics parameters definition:
k ! j; pk ðsÞ ! aj ; qk ðsÞ ! cj ; n ¼ 2; m ¼ 0
P
2
Additionally Pn ðk; sÞ ! PðkÞ; Qm ðk; sÞ ! QðkÞ then PðkÞ ¼ aj k j and
j¼0
P
0
QðkÞ ¼ cj k j .
j¼0
1 1
PðkÞ ¼ k2 þ k ; QðkÞ ¼ n; m 2 N0 ; n [ m
C1 R1 C1 f#
And aj ; cj : R þ 0 ! R are continuous and differentiable function of s such that

a0 þ c0 6¼ 0. In the following “−” denotes complex and conjugate. PðkÞ; QðkÞ
Are analytic functions in k and differentiable in s. And the coefficients:
faj ðC1 ; R1 Þ; cj ðC1 ; antenna parametrsÞg 2 R
Depend on RFID C1, R1 values and antenna parameters but not on s.

a0 ¼ 0; a1 ¼ C11R1 ; a2 ¼ 1; a3 ¼ 0; c0 ¼ C11f# ; c1 ¼ c2 ¼ 0
Unless strictly necessary, the designation of the varied arguments
ðR1 ; C1 ; antenna parametrs) will subsequently be omitted from P, Q, aj, cj. The
coefficients aj, cj are continuous, and differentiable functions of their arguments, and
direct substitution shows that a0 þ c0 6¼ 0; C11f# 6¼ 0
8 C1 ; antenna parameters 2 R þ I.e. k ¼ 0 is not a root of the characteristic
equation. Furthermore PðkÞ; QðkÞ are analytic functions of k for which the fol-
lowing requirements of the analysis (see Kuang 1993, Sect. 3.4) can also be verified
in the present case [4, 5].
(a) If k ¼ i x, x 2 R then Pði xÞ þ Qði xÞ 6¼ 0, i.e. P and Q have no common
imaginary roots. This condition was verified numerically in the entire
ðR1 ; C1 ; antenna parametrs) domain of interest.
(b) jQðkÞ=PðkÞj Is bounded for jkj ! 1, Rek 0. No roots bifurcation from 1.
Indeed, in the limit jQðkÞ=PðkÞj ¼ j C f ðk21þ k 1 Þ j
1 # C1 R1
2 2
(c) FðxÞ ¼ jPði xÞj jQði xÞj ; Fðx; sÞ
1 1
¼ x4 þ x2 2

Has at most a finite number of zeros. Indeed, this is a bi-cubic polynomial in x

(second degree in x2 ).
(d) Each positive root xðC1 ; R1 ; antenna parametrs) of FðxÞ ¼ 0 is continuous
and differentiable with respect to C1 ; R1 ; antenna parametrs.
The condition can only be assessed numerically.
In addition, since the coefficients in P and Q are real, we have, and Qði xÞ ¼
Qði xÞ thus, x [ 0 may be an eigenvalue of the characteristic equation. The
analysis consists in identifying the roots of the characteristic equation situated on
the imaginary axis of the complex k-plane, whereby increasing the parameters
C1 ; R1 ; antenna parametrs and delay s, Rek may, at the crossing, Change its sign
ð0Þ ð0Þ
from (−) to (+), i.e. from stable focus Eð0Þ ðV1 ; V2 Þ ¼ ð0; 0Þ to an unstable one, or
vice versa.
This feature may be further assessed by examining the sign of the partial
derivatives with respect to C1 ; R1 and antenna parameters.
@Rek
^1 ðC1 Þ ¼ ð Þ ; R1 ; antenna parameters ¼ const
@C1 k¼ix
@Rek
^1 ðR1 Þ ¼ ð Þ ; C1 ; antenna parameters ¼ const
@R1 k¼ix
@Rek @Rek
^1 ðf# Þ ¼ ð Þ ; C1 ; R1 ¼ const; ^1 ðsÞ ¼ ð Þ ; C1 ; R1 ;
@f# k¼ix @s k¼ix
antenna parameters ¼ const; where x 2 R þ :
In the first case s1 ¼s; s2 ¼ 0 we get the following results
1
PR ði xÞ ¼ a2 x2 þ a0 ¼ x2 ; PI ði xÞ ¼ a3 x3 þ a1 x ¼ x
C1 R1
1
QR ði xÞ ¼ c2 x2 þ c0 ¼ ; QI ði xÞ ¼ c1 x ¼ 0 ; FðxÞ ¼ 0
C1 f#
vffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
u
u 1 1 1 1 1 1
x¼
t
þ4 ; þ4 [0
2 4 2 4
2 ðC1 R1 Þ 2 ðC1 R1 Þ ½C1 f# ðC1 R1 Þ ½C1 f# 2
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
Always and additional for x 2 R; 1
2ðC1 R1 Þ2

12 1
ðC R Þ4
þ 4 ½C f1 2 [ 0
1 1 1 #
And there are two options: first always exist

2 þ 2 þ 4 ½C f1 2 [ 0 12
1 1 1 1
2ðC1 R1 Þ ðC1 R1 Þ4
Second 2ðC1 R1 Þ2
1 #
1
ðC R Þ4
þ 4 1
½C f 2
[ 0
1 1 1 #
Not exist and always negative for any RFID TAG overall parameter values.
Writing PðkÞ ¼ PR ðkÞ þ i PI ðkÞ and QðkÞ ¼ QR ðkÞ þ i QI ðkÞ, and inserting
k¼ix
Into the RFID characteristic equation, x must satisfy the following:
PR ði xÞ QI ði xÞ þ PI ði xÞ QR ði xÞ
sin x s ¼ gðxÞ ¼
jQði xÞj2
PR ði xÞ QR ði xÞ þ PI ði xÞ QI ði xÞ
cos x s ¼ hðxÞ ¼
jQði xÞj2
Where jQði xÞj2 6¼ 0 in view of requirement (a) above, and ðg; hÞ 2 R.

Furthermore, it follows above sin x s and cos x s equations that, by squaring and
adding the sides, x must be a positive root of FðxÞ ¼ jPði xÞj2 jQði xÞj2 ¼ 0.
Note that FðxÞ is independent of s. Now it is important to notice that if
s 62 I (assume that I R þ 0 is the set where xðsÞ is a positive root of FðxÞ and for
s 62 I , xðsÞ is not defined. Then for all s in I xðsÞ is satisfied that Fðx; sÞ ¼ 0)
Then there are no positive xðsÞ solutions of Fðx; sÞ ¼ 0, and we cannot have
stability switches. For any s 2 I, where xðsÞ is a positive solution of Fðx; sÞ ¼ 0
We can define the angle hðsÞ 2 ½0; 2 p as the solution of
sin hðsÞ ¼
jQði xÞj2
cos hðsÞ ¼
jQði xÞj2
And the relation between the argument hðsÞ and xðsÞ s for s 2 I must be
xðsÞ s ¼ hðsÞ þ n 2 p 8 n 2 N0 . Hence we can define the maps sn : I ! R þ 0
given by sn ðsÞ ¼ hðsÞxðsÞ þ n2p
; n 2 N0 ; s 2 I. Let us introduce the functions I ! R;
Sn ðsÞ ¼ s sn ðsÞ; s 2 I; n 2 N0 that is continuous and differentiable in s. In the
following, the subscripts k; x; R1 ; C1 and RFID TAG antenna parameters
ðw; g; B0 ; A0 ; Aavg ; Bavg ; etc:; Þ indicate the corresponding partial derivatives. Let us
first concentrate on ^ðxÞ, remember in kðR1 ; C1 ; w; g; B0 ; A0 ; Aavg ; Bavg ; etc:; Þ and
xðR1 ; C1 ; w; g; B0 ; A0 ; Aavg ; Bavg ; etc:; Þ, and keeping all parameters except one
(x) and s. The derivation closely follows that in reference [BK]. Differentiating
RFID characteristic equation PðkÞ þ QðkÞ eks ¼ 0 with respect to specific
parameter (x), and inverting the derivative, for convenience, one calculates:
Remark x ¼ R1 ; C1 ; w; g; B0 ; A0 ; Aavg ; Bavg ; etc:;
@k 1 Pk ðk; xÞ Qðk; xÞ þ Qk ðk; xÞ Pðk; xÞ s Pðk; xÞ Qðk; xÞ

ð Þ ¼
@x Px ðk; xÞ Qðk; xÞ Qx ðk; xÞ Pðk; xÞ
Where Pk ¼ @P
@k ; . . .: etc., substituting k ¼ i x, and bearing
i Pði xÞ ¼ Pði xÞ, Qði xÞ ¼ Qði xÞ then i Pk ði xÞ ¼ Px ði xÞ and i
Qk ði xÞ ¼ Qx ði xÞ and that on the surface jPði xÞj2 ¼ jQði xÞj2 , one obtains:
@k 1 i Px ði x; xÞ Pði x; xÞ þ i Qk ði x; xÞ Qðk; xÞ s jPði x; xÞj2

ð Þ jk¼ix ¼ ð Þ
@x Px ði x; xÞ Pði x; xÞ Qx ði x; xÞ Qði x; xÞ
Upon separating into real and imaginary parts, with P ¼ PR þ i PI ;

Q ¼ QR þ i QI ; Px ¼ PRx þ i PIx ;Qx ¼ QRx þ i QIx ; Px ¼ PRx þ i PIx ; Qx ¼
QRx þ i QIx P2 ¼ P2R þ P2I . When (x) can be any RFID TAG parameters R1, C1, and
time delay s etc., Where for convenience, we have dropped the arguments ði x; xÞ,
and where
Fx ¼ 2 ½ðPRx PR þ PIx PI Þ ðQRx QR þ QIx QI Þ;
Fx ¼ 2 ½ðPRx PR þ PIx PI Þ ðQRx QR þ QIx QI Þ
xx ¼ Fx =Fx
We define U and V:
U ¼ ðPR PIx PI PRx Þ ðQR QIx QI QRx Þ
We choose our specific parameter as time delay x = s. PIs ¼ 0; PRs ¼ 0;

QIs ¼ 0; QRs ¼ 0 ) V ¼ 0
x2 1
U¼ ; P2 ¼ x4 þ x2 ;
C1 R1 ðC1 R1 Þ2
@F 1
Fs ¼ 0; ¼ Fx ¼ 2 ½2 x3 þ x
@x ðC1 R1 Þ2
Fðx; sÞ ¼ 0
And differentiating with respect to s and we get

@x @x Fs @Rek
Fx þ Fs ¼ 0; s 2 I ) ¼ ; ^1 ðsÞ ¼ ð Þ
@s @s Fx @s k¼ix
h i
2 U þ s jPj2 þ i Fx 2 x2 þ ðC R
1
2
1 1 1Þ
^ ðsÞ ¼ Ref h i g ;
Fs þ i 2 V þ x jPj2 x þ x ðC R Þ2
4 2 1
1 1
@Rek
signf^1 ðsÞg ¼ signfð Þ g
@s k¼ix
@x U @x
@s þ V
signf^1 ðsÞg ¼ signfFx g signfs þxþ g;
@s jPj2
@x Fs @x
¼ xs ¼ ; Fs ¼ 0 ) ¼0
@s Fx @s
Then we get signf^1 ðsÞg ¼ signf2 x ½2 x2 þ 1

ðC1 R1 Þ2
g signfxg
1 1
Result: ^ ðsÞ [ 0 for all x; R1 ; C1 values. The sign of ^ ðsÞ is independent of
s values, then in the first case s1 ¼s; s2 ¼ 0 there is no stability switch for different
values s. We now inspect the third interesting case when s1 ¼s; s2 ¼ s. The third
case we analyze is when there are delays both in RFID Label voltage and voltage
time derivative [4, 5].
1 1
Dðk; s1 ¼ s; s2 ¼ sÞ ¼ k2 þ k eks þ eks2
C1 R1 C1 f#
Taylor expansion:
k2 s 2
eks 1 k s þ
2
Since we need n > m [BK] analysis, we choose eks 1 k s then we get our
RFID system second order characteristic equation:

1 1 s 1
aðsÞ ¼ 0; bðsÞ ¼ ð Þ; cðsÞ ¼ 0; dðsÞ ¼
C1 R1 f# C 1 f#
1 1
Fðx; sÞ ¼ x4 x2 2

Hence Fðx; sÞ ¼ 0 implies x4 x2 C12 ðR11 fs# Þ2 ðC 1

2 ¼ 0 and its roots are
1 1 f# Þ
given by
1 pffiffiffiffi 1 pffiffiffiffi 1 1 s
x2þ ¼ fðb2 þ 2 c a2 Þ þ Dg ¼ f D þ 2 ð Þ2 g
2 2 C1 R1 f#
1 p ffiffiffi
ffi 1 p ffiffiffi
ffi 1 1 s
x2 ¼ fðb2 þ 2 c a2 Þ Dg ¼ f D þ 2 ð Þ2 g
2 2 C1 R1 f#
1 1 s 4
D ¼ ðb2 þ 2 c a2 Þ 4 ðc2 d 2 Þ ¼ 2 ð Þ2 þ
C1 R1 f# ðC1 f# Þ2
pffiffiffiffi
Therefore the following holds: 2 x2þ = ðb2 þ 2 c a2 Þ ¼
D
sin hðsÞ ¼
jQði x; sÞj2
cos hðsÞ ¼
jQði x; sÞj2
ðc x2 Þ x b þ x a d x3 C11 ðR11 fs# Þ

sin hðsÞ ¼ ¼
x2 b2 þ d 2 x2 C12 ðR11 fs# Þ2 þ ðC 1
2
1 1 f# Þ
ðc x2 Þ d þ x2 a b x2 C11f#
cos hðsÞ ¼ ¼
x 2 b2 þ d 2 x2 1 1
ð fs# Þ2 þ 1
C12 R1 ðC1 f# Þ2
For our stability switching analysis, we choose typical RFID parameter values:
C1 ¼ 23 pF; R1 ¼ 100 kX ¼ 105 ; Lcalc ¼ f# ¼ 2:65 mH
Then
Fig. 1.17 RFID TAG F (x, s)

function for s1 = s2 = s
Fig. 1.18 RFID TAG

stability switch diagram based
on different delay values of
our RFID TAG system
1 1
¼ 1:89 1021 ; 2 2 ¼ 2:69 1026
C12 C1 f#
We find those x; s values which fulfill Fðx; sÞ ¼ 0. We ignore negative,

complex, and imaginary values of x for specific s values. The table gives the list.
s 2 ½0:001::10. And can be expressed by straight line (x = s 1.64 1013)
(Fig. 1.17).
s x
0.001 1.64 1010
0.01 1.64 1011
0.05 8.2 1011
0.1 1.64 1012
(continued)
(continued)
s x
0.2 3.28 1012
1 1.64 1013
5 8.2 1013
10 1.64 1014
Remark In the above figure x variable is 1010 units.

MATLAB: [w,t] = meshgrid(1:1:1640,0:0.01:10);
f = w. * w. * w. * w − w. * w. * 1.89 * 10^21. * (10^−5 − (t./(2.65 * 10^
−3))).^2−2.69 * 10^26; meshc(f); %x ! w; s ! t
We plot the stability switch diagram based on different delay values of our RFID
TAG system (Fig. 1.18).
@Rek 2 ½U þ s jPj2 þ i Fx
^1 ðsÞ ¼ ð Þk¼ix ¼ Ref g
@s Fs þ i 2 ½V þ x jPj2
@Rek 2 fFx ðV þ x P2 Þ Fs ðU þ s P2 Þg
^1 ðsÞ ¼ ð Þk¼ix ¼
@s Fs2 þ 4 ðV þ x P2 Þ2
@Rek
gðTauÞ ¼ ^1 ðsÞ ¼ ð Þ
@s k¼ix
The stability switch occurs only on those delay values (s) which fit the equation:
s ¼ xh þþ ðsÞ
ðsÞ and h þ ðsÞ is the solution of
x3 C11 ðR11 fs# Þ x2 C11f#

sin hðsÞ ¼ ; cos hðsÞ ¼
x2 C12 ðR11 fs# Þ2 þ 1
ðC1 f# Þ2
x2 C12 ðR11 fs# Þ2 þ 1
ðC1 f# Þ2
1 1
When x ¼ x þ ðsÞ if only x þ is feasible. Additionally When all RFID TAG

parameters are known and the stability switch due to various time delay values s is
described in the below expression:
signf^1 ðsÞg ¼ signfFx ðxðsÞ; sÞg signfs xs ðxðsÞÞ

UðxðsÞÞ xs ðxðsÞÞ þ VðxðsÞÞ
þ xðsÞ þ g
jPðxðsÞÞj2
Remark we know Fðx; sÞ ¼ 0 implies its roots xi ðsÞ and finding those delays
values s which xi is feasible. There are s values which xi is a complex or imag-
inary number, then unable to analyze stability [4, 5].
RFID TAGs environment is characterized by electromagnetic interferences
which can influence the RFID TAGs stability in time. There are two main RFID
TAGs variables which are affected by electromagnetic interferences, the voltage

developed on the RFID Label and his voltage time derivative respectively.
Each RFID Label variable under electromagnetic interferences is characterized by
time delay respectively. The two time delays are not the same, but can be cate-
gorized to some sub cases due to interferences behavior. The first case is when there
is RFID Label voltage time delay, but no voltage derivative time delay. The second
case is when there is no RFID Label voltage time delay, but there is a voltage
derivative time delay. The third case is when both RFID Label voltage time delay
and voltage derivative time delay exist. For simplicity of our analysis we consider
the third case, two delays are the same (there is a difference but it is neglected in our
analysis). In each case we derive the related characteristic equation. The charac-
teristic equation is dependent on RFID Label overall parameters and interferences
time delay. Upon mathematics manipulation and [BK] theorems and definitions we
derive the expression which gives us a clear picture on RFID Label stability
map. The stability map gives all possible options for stability segments, each
segment belongs to different time delay value segment. RFID Label stability
analysis can be influenced either by TAG overall parameter values. We left this
analysis and do not discuss it in the current chapter.
Lemma 1.1 Assume that xðsÞ is a positive and real root of Fðx; sÞ ¼ 0
Defined for s 2 I, this is continuous and differentiable. Assume further that if
k ¼ i x, x 2 R, then Pn ði x; sÞ þ Qn ði x; sÞ 6¼ 0; s 2 R hold true. Then the
functions Sn ðsÞ; n 2 N0 , are continuous and differentiable on I.
Theorem 1.2 Assume that xðsÞ is a positive real root of Fðx; sÞ ¼ 0 defined for
s 2 I; IR þ 0 , and at some s 2 I, Sn ðs Þ ¼ 0 for some n 2 N0 then a pair of
simple conjugate pure imaginary roots k þ ðs Þ ¼ i xðs Þ; k ðs Þ ¼ i xðs Þ of
Dðk; sÞ ¼ 0 exist at s ¼ s which crosses the imaginary axis from left to right if
dðs Þ [ 0 and cross the imaginary axis from right to left if dðs Þ\0 where
dRek dSn ðsÞ

dðs Þ ¼ signf jk¼ixðs Þ g ¼ signfFx ðxðs Þ; s Þg signf j g
ds ds s¼s
n ðsÞ
The theorem becomes signfd Re k
ds jk¼ix
g ¼ signf
D
1=2
g signfdSds js¼s g
Theorem 1.3 The characteristic equation: s1 ¼ s; s2 ¼ 0; s1 ¼ 0; s2 ¼ s
Dðk; sÞ ¼ k2 þ aðsÞ k þ bðsÞ k eks þ cðsÞ þ dðsÞ eks ;

1 1
Dðk; s1 ; s2 Þ ¼ k2 þ k eks2 þ ekðs1 þ s2 Þ
C1 R1 C1 f#
Has a pair of simple and conjugate pure imaginary roots k ¼

xðs Þ; xðs Þ
Real at s 2 I if Sn ðs Þ ¼ s sn ðs Þ ¼ 0 for some n 2 N0 . If xðs Þ ¼ x þ ðs Þ,
this pair of simple conjugate pure imaginary roots crosses the imaginary axis from
left to right if d þ ðs Þ [ 0 and crosses the imaginary axis from right to left if
dSn ðsÞ
d þ ðs Þ\0 where d þ ðs Þ ¼ signfd Re k
ds jk¼ix þ ðs Þ g ¼ signf ds js¼s g. If
xðs Þ ¼ x ðs Þ, this pair of simple conjugate pure imaginary roots cross the
imaginary axis from left to right if d ðs Þ [ 0 and crosses the imaginary axis from
right to left If d ðs Þ\0 where d ðs Þ ¼ signfd Re k
ds jk¼ix ðs Þ g ¼
n ðsÞ
signfdSds js¼s g. If x þ ðs Þ ¼ x ðs Þ ¼ xðs Þ then dðs Þ ¼ 0 and
signfdRek j
ds k¼ixðs Þ
0
g ¼ 0, the same is true when S ðs Þ ¼ 0. The following result can
n
be useful in identifying values of s where the stability switches happened.
Theorem 1.4 Assume that for all s 2 I, xðsÞ is defined as a solution of Fðx; sÞ ¼
0 then d
ðsÞ ¼ signf
D1=2 ðsÞg signD
ðsÞ
D
ðsÞ ¼ x2
½ðx2
b2 þ d 2 Þ þ a0 ðc x2
Þ þ b d 0 b0 d a c0
þ x
x0
½s ðx2
b2 þ d 2 Þ b d þ a ðc x2
Þ þ 2 x2
a
daðsÞ 0 dbðsÞ 0 dcðsÞ 0 ddðsÞ

a0 ¼ ;b ¼ ;c ¼ ;d ¼
ds ds ds ds
1.4 Semi-Passive RFID Tags with Double Loop Antennas

Arranged as a Shifted Gate System for Stability
Optimization Under Delayed Electromagnetic
Interferences
A semi-passive tags operate similarly to passive RFID tags. However, they contain
a battery that enables long reading distance and also enables the tag to operate
independently of the reader. Semi-Passive TAGs with double loop antennas ar-
ranged as a shifted gate system influence by electromagnetic interferences which
effect there stability behavior. Semi-Passive RFID TAGs system with a battery is
like a Reader unit and aimed to improve the communication performance by using
double loop antennas in walk-through gate arrangement in various TAGs orienta-
tions of the RFID system operating mainly in the LF band. The below figure
describes the double loop antennas as a shifted gate in x-direction [8] (Fig. 1.19).
The antenna gate is shifted to avoid cancellation of magnetic fields between two
TAGs, and to improve the magnetic-field distribution. The RFID system at Low
Frequency (LF) band has been widely adopted. The RFID tags for this application
have usually installed in applications such that the orientation of tag id difficult to
fix for transferring data with RFID reader. Most of the LF-RFID reader antennas are
rectangular or circular loops, but these antennas cannot generate sufficient field
1.4 Semi-Passive RFID Tags with Double Loop Antennas Arranged … 45
Fig. 1.19 Double loop D

antennas arranged as a shifted
gate in x-direction
d1
strengths in some locations and/or tag antenna orientations. The double loop
antennas arranged as a shifted gate improves magnetic-field distribution in a region
of interest suitable for communication with various tag orientations, and enhance
the communication distance [1]. The antenna gate is shifted to avoid cancellation of
magnetic fields between two gate antennas. The gate antenna consists of two sides
of rectangular loops with two types of excitations; i.e., in phase and 180° out of
phase. When two antennas are excited in phase, the directions of the currents
flowing in two loops are in the same direction, resulting in the cancellation of
magnetic fields in the x-direction in the middle region of the gate. When two
antennas are excited 180° out of phase, the direction of the currents flowing in two
antennas is in the opposite direction, resulting in the cancellation of magnetic fields
in the y-direction in the middle region of the gates. Thus, the gate antennas are
arranged as a shifted gate to maintain magnetic fields in the middle region. The
double loop antenna is employed due to the fact that this antenna consists of two
parallel loops (primary and secondary loops). The shape of the primary loop is
rectangular for generating the magnetic field in the y-direction. The secondary loop
is always within the primary loop, and is optimized such that the magnetic fields in
x- and z-directions are strongly generated. D is the separation distance between gate
antennas, and d1 is the shifted distance in the x-direction. Due to electromagnetic
interferences there are differences in time delays with respect to gate antenna’s first
and second loop voltages and voltages derivatives. The delayed voltages are
Vi1(t − s1) and Vi2(t − s2) respectively (s1 6¼ s2) and delayed voltages derivatives
are dVi1(t − D1)/dt, dVi2(t − D2)/dt respectively ðD1 6¼ D2 ; s1 0; s2 0;
D1 ; D2 0Þ. The Semi-Passive RFID TAG with double loop antennas equivalent
circuit can represent as a delayed differential equations which depending on vari-
able parameters and delays. Our Semi-Passive RFID TAG system delay differential
and delay different model can be analytically by using delay differential equations
in dynamically. The need of the incorporation of a time delay is often of the
existence of any stage structure. It is often difficult to analytically study models with
delay dependent parameters, even if only a single discrete delay is present. There
are practical guidelines that combine graphical information with analytical work to
some function of s1, s2 which can be expressed, explicitly and thus can be easily
depicted by Matlab and other popular software. We need only look at one such
function and locate the zeros. This function often has only two zeros, providing
thresholds for stability switches. As time delay increases, stability changes from
stable to unstable to stable. We emphasize the local stability aspects of some models
with delay dependent parameters. Additionally, there is a general geometric crite-
rion that, theoretically speaking, can be applied to models with many delays, or
even distributed delays. The simplest case of a first order characteristic equation,
providing more user friendly geometric and analytic criteria for stability switches.
The analytical criteria provided for the first and second order cases can be used to
obtain some insightful analytical statements and can be helpful for conducting
simulations [5, 6]. Semi-Passive RFID TAG with double loop antenna can be
represented as a two inductors in series (L11 and L12 for the first double loop gate
antenna) with parasitic resistance rP1. The double loop antennas in series are con-
nected in parallel to Semi-Passive RFID TAG. The Equivalent Circuit of
Semi-Passive RFID TAG is Capacitor (C1) and Resistor (R1) in parallel with
voltage generator Vs1(t) and parasitic resistance rS1. In case we have Passive RFID
TAG switch S1 is OFF otherwise is ON (Reader/Active RFID system) and long
distance is achievable. The second double loop gate antenna is defined as two
inductors in series L21 and L22 with series parasitic resistor rP2. Vs2(t) and parasitic
resistance rS2 are belong to the second gate antenna system with another
Semi-Passive RFID TAG [1].
L11 and L12 are mostly formed by traces on planar PCB. 2 Lm element represents
the mutual inductance between L11 and L12. We consider that the double loop
antennas parameter values (La1, La2, Lb1, Lb2, a1, a2) are the same in the first and second
gates. Since two inductors (L11, L12) are in series and there is a mutual inductance
between L11 and L12, the total antenna inductance LT: LT = L11 + L12 + 2 Lm and
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
Lm ¼ K L11 L12 . Lm is the mutual inductance between L11 and L12. K is the
coupling coefficient of two inductors 0 K 1. We start with the case of passive
RFID TAG which switch S1 is OFF. I(t) is the current that flow through a double loop
antenna. V11 and V12 are the voltages on L11 and L12 respectively. Vm is the voltage on
double loop antenna mutual inductance element.
dI dI dI
V11 ¼ L11 ; V12 ¼ L12 ; VCD ¼ I rp1 ; Vm ¼ 2 Lm ;
dt dt dt
dVC1
VAB ¼ VR1 ¼ VC1 ¼ V11 þ V12 þ VCD þ Vm ; IC1 ¼ C1
dt
dVC1
VAB ¼ VR1 ¼ VC1 ¼ V11 þ V12 þ VCD þ Vm ; IC1 ¼ C1 ;
dt
dVC1 VC
IC1 þ IR1 þ I ¼ 0 ) C1 þ 1 þ I ¼ 0; L11 6¼ L12
dt R1
dVC1 dV11 dV12 dVCD dVm

¼ þ þ þ ;
dt Zdt dt dtZ dt
1 1
I¼ V11 dt ¼ V12 dt;
L11 L12
Z Z
rp1 rp1
VCD ¼ I rp1 ¼ V11 dt ¼ V12 dt
L11 L12
dVCD rp1 rp1 L11 L12

¼ V11 þ V12 ; V11 ¼ V12 ; V12 ¼ V11 ;
dt L11 L12 L12 L11
Z Z
1 1 dI 1 1
I¼ V11 dt ¼ V12 dt ) ¼ V11 ¼ V12
L11 L12 dt L11 L12
rffiffiffiffiffiffiffi
dI pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi 1 L12
¼ 2 K L11 L12
Vm ¼ 2 Lm V11 ¼ 2 K V11 ;
dt L11 L11
dVm L12 dV11
¼2K
dt L11 dt
We get the following differential equation respect to V11(t) variable, g1 ; g2 ; g3

are global parameters.
d 2 V11 dV11
g1 þ g2 þ V11 g3 ¼ 0
dt2 dt rffiffiffiffiffiffiffi
L12 L12
g1 ¼ C1 ð1 þ þ2 K Þ;
L11 L11
C1 rp1 1 L12 L12
g2 ¼ þ ð1 þ þ2 K Þ;
L11 R1 L11 L11
C1 rp1 1 g1 1 rp1
g2 ¼ þ ;g ¼ ð1 þ Þ
L11 R1 C1 3 L11 R1
g1 ¼ g1 ðC1 ; L12 ; L11 ; KÞ; g2 ¼ g2 ðC1 ; rp1 ; L12 ; L11 ; K; R1 Þ;
0
dV11 dV11
0 d 2 V11
g3 ¼ g3 ðL11 ; rp1 ; R1 Þ; V11 ¼ ; ¼
dt dt dt2
0
dV11 0 g g dV 11 0
¼ V11 2 V11 3 ; ¼ V11 :
dt g1 g1 dt
In the same manner we find our V12 differential equation. We get the following
differential equation respect to V12(t) variable, n1 ; n2 ; n3 are global parameters.
d 2 V12 dV12 L11 L11
n1 þ n2 þ V12 n3 ¼ 0; n1 ¼ C1 ð1 þ þ2 K Þ;
dt2 dt L12 L12
C1 rp1 1 L11 L11 1 rp1
n2 ¼ þ ð1 þ þ2 K Þ; n3 ¼ ð1 þ Þ
L12 R1 L12 L12 L12 R1
0
C1 rp1 1 n1 0 dV12 dV12 d 2 V12
n2 ¼ þ ;V ¼ ; ¼ ;
L12 R1 C1 12 dt dt dt2
n1 ¼ n1 ðC1 ; L12 ; L11 ; KÞ; n2 ¼ n2 ðC1 ; rp1 ; L12 ; L11 ; K; R1 Þ
0
dV12 0 n n dV12 0
n3 ¼ n3 ðL12 ; rp1 ; R1 Þ; ¼ V12 2 V12 3 ; ¼ V12 :
dt n1 n1 dt
Summary: We get our RFID double loop antennas system’s four differential
equations.
0 0
dV11 0 g g dV11 0 dV12 0 n n dV12 0
¼ V11 2 V11 3 ; ¼ V11 ; ¼ V12 2 V12 3 ; ¼ V12
dt g1 g1 dt dt n1 n1 dt
0 dV 0 1
11 0 1 0 V0 1
B dt C C11 . . . C14 11
B dV11 C B C BV C g2 g3
B dt C B .. .. .. C B 11 C
B dV 0 C ¼ @ . . . AB C; C11 ¼ ; C12 ¼ ;
B 12 C @ V12 A
0 g1 g1
@ dt A
dV
C41 C44 V
12 12
dt
n2 n
C33 ¼ ; C34 ¼ 3 ; C21 ¼ C43 ¼ 1
n1 n1
C13 ¼ C14 ¼ C22 ¼ C23 ¼ C24 ¼ C31 ¼ C32 ¼ C41 ¼ C42 ¼ C44 ¼ 0
The RFID double loop antennas system’s primary and secondary loops are
composed of a thin wire or a thin plate element (Fig. 1.20). Units are all in cm, and
a1, a2 are radiuses of the primary and secondary wires in cm. There inductances can
be calculated by the following formulas (Fig. 1.21):
2 A1 2 A1
L11 ¼ 4 fLb1 ln½ þ La1 ln½ þ 2 ½a1 þ lc1
a1 ðLb1 þ lc1 Þ a1 ðLb1 þ lc1 Þ
ðLa1 þ Lb1 Þg
Fig. 1.20 Double loop

La1
antennas in series with
parasitic resistance and
Semi-Passive RFID TAG La2
2a1
2a2 Lb2 Lb1
Semi-Active
RFID / Reader
Fig. 1.21 Equivalent circuit

of double loop antennas in
series with Semi-
Passive RFID TAG
2 A2 2 A2
L12 ¼ 4 fLb2 ln½ þ La2 ln½ þ 2 ½a2 þ lc2
a2 ðLb2 þ lc2 Þ a2 ðLb2 þ lc2 Þ
ðLa2 þ Lb2 Þg
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
lc1 ¼ L2a1 þ L2b1 ; A1 ¼ La1 Lb1 ; lc2 ¼ L2a2 þ L2b2 ; A2 ¼ La2 Lb2
Due to electromagnetic interferences, we get a shifted gate RFID system’s pri-

mary and secondary antennas loops voltages with delays s1 and s2 respectively.
Additionally, we get antennas loop voltages derivatives with delays D1 and D2
0 0
respectively. V11 ðtÞ ! V11 ðt s1 Þ; V12 ðtÞ ! V12 ðt s2 Þ; V11 ðtÞ ! V11 ðt D1 Þ
0 0 dV 0 dV 0
V12 ðtÞ ! V12 ðt D2 Þ. We consider no delay effect on dVdt11 ; dVdt12 ; dt11 ; dt12 .
The RFID shifted gate system, differential equations under electromagnetic
interferences (delays terms) influence only RFID double loop voltages V11(t),
0 0
V12(t) and voltages derivatives V11 ðtÞ and V12 ðtÞ respect to time, there is no
0 0
dV11 ðtÞ dV12 ðtÞ dV11 ðtÞ dV12 ðtÞ
influence on dt ; dt ; dt ; dt .
0 0 1
dV11
B dt C 0 0 1
B C
B dV11 C 0 1 V11 ðt D1 Þ
B C C11 . . . C14 B C
B dt C B . .. .. C B V11 ðt s1 Þ C
B C
B dV 0 C ¼ @ .. . . A B 0
@ V12
C
ðt D2 Þ A
B 12 C
B C C41 C44
B dt C V12 ðt s2 Þ
@ dV A
12
dt
To find equilibrium points (fixed points) of the RFID shifted gate system is by
lim V11 ðt s1 Þ ¼ V11 ðtÞ; lim V12 ðt s2 Þ ¼ V12 ðtÞ;

t!1 t!1
0 0 0 0
lim V11 ðt D1 Þ ¼ V11 ðtÞ; lim V12 ðt D2 Þ ¼ V12 ðtÞ
t!1 t!1
dV11 ðtÞ dV12 ðtÞ dV 0 ðtÞ dV 0 ðtÞ
¼ 0; ¼ 0; 11 ¼ 0; 12 ¼ 0: 8 t s1 ; t s2 ; t D1 ; t D2
dt dt dt dt
9 ðt s1 Þ t; ðt s2 Þ t; ðt D1 Þ t; ðt D2 Þ t; t ! 1:
0 ð0Þ ð0Þ 0 ð0Þ ð0Þ

We get four equations and the only fixed point is E ð0Þ ðV11 ; V11 ; V12 ; V12 Þ ¼
ð0; 0; 0; 0Þ since
g3 6¼ 0 & g1 6¼ 0 ) C12 6¼ 0; n3 6¼ 0 & n1 6¼ 0 ) C34 6¼ 0

Stability analysis: The standard local stability analysis about any one of the
equilibrium points of RFID shifted gate system consists in adding to coordinate
0 0
½V11 V11 V12 V12 arbitrarily small increments of exponential form½v011 v11 v012 v12 ekt ,
0 0
and retaining the first order terms in V11 V11 V12 V12 . The system of four homogeneous
equations leads to a polynomial characteristic equation in the eigenvalues k. The
polynomial characteristic equations accept by set the below voltages and voltages
derivative respect to time into two RFID shifted gate system equations.
RFID shifted gate system fixed values with arbitrarily small increments of
exponential form ½v011 v11 v012 v12 ekt are: i = 0 (first fixed point), i = 1 (second
fixed point), i = 2 (third fixed point), etc.,
0 0ðiÞ ðiÞ
V11 ðtÞ ¼ V11 þ v011 ekt ; V11 ðtÞ ¼ V11 þ v11 ekt
0 0ðiÞ ðiÞ
V12 ðtÞ ¼ V12 þ v012 ekt ; V12 ðtÞ ¼ V12 þ v12 ekt
0 0
We choose the above expressions for our V11 ðtÞ; V11 ðtÞ and V12 ðtÞ; V12 ðtÞ as
0 0
small displacement ½v11 v11 v12 v12 from the system fixed points in time t = 0.
0 0ðiÞ ðiÞ
V11 ðt ¼ 0Þ ¼ V11 þ v011 ; V11 ðt ¼ 0Þ ¼ V11 þ v11
0 0ðiÞ ðiÞ
V12 ðt ¼ 0Þ ¼ V12 þ v012 ; V12 ðt ¼ 0Þ ¼ V12 þ v12
For k\0; t [ 0 the selected fixed point is stable otherwise k [ 0; t [ 0 is

Unstable. Our system tends to the selected fixed point exponentially to k\0; t [ 0
otherwise go away from the selected fixed point exponentially. k is the eigenvalue
parameter which establishes if the fixed point is stable or Unstable, additionally his
absolute value (jkj) establish the speed of flow toward or away from the selected
fixed point [2, 3].
k<0 k>0
0 0ðiÞ 0ðiÞ
t=0 V11 ðt ¼ 0Þ ¼ V11 þ v011 0
V11 ðt ¼ 0Þ ¼ V11 þ v011
ðiÞ ðiÞ
V11 ðt ¼ 0Þ ¼ V11 þ v11 V11 ðt ¼ 0Þ ¼ V11 þ v11
0 0ðiÞ 0ðiÞ
V12 ðt ¼ 0Þ ¼ V12 þ v012 0
V12 ðt ¼ 0Þ ¼ V12 þ v012
ðiÞ ðiÞ
V12 ðt ¼ 0Þ ¼ V12 þ v12 V12 ðt ¼ 0Þ ¼ V12 þ v12
0 0ðiÞ 0ðiÞ
t>0 V11 ðtÞ ¼ V11 þ v011 ejkjt 0
V11 ðtÞ ¼ V11 þ v011 ejkjt
ðiÞ ðiÞ
V11 ðtÞ ¼ V11 þ v11 ejkjt V11 ðtÞ ¼ V11 þ v11 ejkjt
0 0ðiÞ 0ðiÞ
V12 ðtÞ ¼ V12 þ v012 ejkjt 0
V12 ðtÞ ¼ V12 þ v012 ejkjt
ðiÞ ðiÞ
V12 ðtÞ ¼ V12 þ v12 ejkjt V12 ðtÞ ¼ V12 þ v12 ejkjt
(continued)
(continued)
k<0 k>0
0ðiÞ 0
t>0 0
V11 ðt ! 1Þ ¼ V11 V11 ðt ! 1; k [ 0Þ v011 ejkjt
t!∞
V11 ðt ! 1Þ ¼
ðiÞ
V11 V11 ðt ! 1; k [ 0Þ v11 ejkjt
0
0
V12 ðt ! 1Þ ¼
0ðiÞ
V12 V12 ðt ! 1; k [ 0Þ v012 ejkjt
V12 ðt ! 1Þ ¼
ðiÞ
V12 V12 ðt ! 1; k [ 0Þ v11 ejkjt
The speeds of flow toward or away from the selected fixed point for RFID
shifted gate system voltages and voltages derivatives respect to time are
0 0ðiÞ 0ðiÞ
dV11 ðtÞ V 0 ðt þ DtÞ V11 0
ðtÞ V þ v011 ekðt þ DtÞ ½V11 þ v011 ekt
¼ lim 11 ¼ lim 11
dt Dt!0 Dt Dt!0 Dt
v011 ekt ½ekDt 1
¼ lim ! ekDt 1 þ k Dtk v011 ekt
Dt!0 Dt
dV11 ðtÞ dV12 ðtÞ
¼ k v11 ekt ; ¼ k v12 ekt ;
dt dt
0
dV12 ðtÞ dV 0 ðt D1 Þ
¼ k v012 ekt ; 11 ¼ k v011 ekt ekD1
dt dt
0
dV11 ðt D1 Þ dV11 ðt s1 Þ
¼ k v011 ekt ekD1 ; ¼ k v11 ekt eks1
dt dt
dV12 ðt s2 Þ dV 0 ðt D2 Þ
¼ k v12 ekt eks2 ; 12 ¼ k v012 ekt ekD2
dt dt
First, we take the RFID shifted gate voltages V11, V12 differential equations:
0 dV12 0 0 0
dV11
dt ¼ V11 ; dt ¼ V12 and adding coordinates ½V11 V11 V12 V12 arbitrarily small
0 0 kt
increments of exponential terms ½v11 v11 v12 v12 e and retaining the first order
terms in v011 v11 v012 v12 .
0ðiÞ 0ðiÞ
k v11 ekt ¼ V11 þ v011 ekt ; V11 ¼ 0; k v11 ¼ v011 ) k v11 þ v011 ¼ 0
0ðiÞ 0ðiÞ
k v12 ekt ¼ V12 þ v012 ekt ; V12 ¼ 0; k v12 ¼ v012 ) k v12 þ v012 ¼ 0
0 0
Second, we take the RFID shifted gate’s voltages derivative V11 ; V12 differential
equations:
0
dV11 0 dV 0 0
¼ C11 V11 þ C12 V11 ; 12 ¼ C33 V12 þ C34 V12
dt dt
0 0
And adding coordinates ½V11 V11 V12 V12 arbitrarily small increments of expo-
nential terms ½v11 v11 v12 v12 e and retaining the first order terms in v011 v11 v012 v12 .
0 0 kt
0 0
0 ðiÞ 0 ðiÞ ði¼0Þ ði¼0Þ
k v11 ekt ¼ C11 ½V11 þ v11 ekt þ C12 ½V11 þ v11 ekt ; V11 ¼ 0 ; V11 ¼0
0 0 0 0
k v11 ¼ C11 v11 þ C12 v11 ) k v11 þ C11 v11 þ C12 v11 ¼ 0
0 0
0 ðiÞ 0 ðiÞ ði¼0Þ ði¼0Þ
k v12 ekt ¼ C33 ½V12 þ v12 ekt þ C34 ½V12 þ v12 ekt ; V12 ¼ 0 ; V12 ¼0
0 0 0 0
k v12 ¼ C33 v12 þ C34 v12 ) k v12 þ C33 v12 þ C34 v12 ¼ 0
0
We can summery our eigenvalues equations: ðk þ C11 Þ v11 þ C12 v11 ¼ 0
0 0 0
v11 k v11 ¼ 0 ; ( k þ C33 Þ v12 þ C34 v12 ¼ 0 ; v12 k v12 ¼ 0
0 0 1
0 1 v11
X11 ... X14 B C
B . B v11 C
B . .. .. C
C B C
@ . . . AB
B v0 C
C ¼ 0 ; X11 ¼ k þ C11 ; X12 ¼ C12 ; X13 ¼ 0 ; X14 ¼ 0
@ 12 A
X41 X44
v12
X21 ¼ 1 ; X22 ¼ k ; X23 ¼ 0 ; X24 ¼ 0 ; X31 ¼ 0 ; X32 ¼ 0 ; X33 ¼ k þ C33 ; X34 ¼ C34
X41 ¼ 0 ; X42 ¼ 0 ; X43 ¼ 1 ; X44 ¼ k

0 1
X11 . . . X14
B . .. .. C
A k I ¼ @ .. . . A ; detðA k IÞ ¼ 0
X41 X44
0 1 0 1
k 0 0 1 0 0
B C B C
detðA k IÞ ¼ ðk þ C11 Þ det@ 0 k þ C33 C34 A C12 det@ 0 k þ C33 C34 A
0 1 k 0 1 k
detðA k IÞ ¼ ðk þ C11 Þ ðkÞ ½ðk þ C33 Þ ðkÞ C34 C12 ½ðk þ C33 Þ ðkÞ C34
k4 k3 ðC33 þ C11 Þ þ k2 ðC11 C33 C34 C12 Þ þ k ðC11 C34 þ C12 C33 Þ þ C12 C34 ¼ 0
Eigenvalues stability discussion: Our Semi-passive RFID tags with double loop
antenna system involving N variables ðN [ 2; N ¼ 4Þ, the characteristic equation is
of degree N ¼ 4 and must often be solved numerically. Expect in some particular
cases, such an equation has (N ¼ 4) distinct roots that can be real or complex.
These values are the eigenvalues of the 4 4 Jacobian matrix (A). The general rule
is that the Steady State (SS) is stable if there is no eigenvalue with positive real
part. It is sufficient that one eigenvalue is positive for the steady state to be unstable.
0 0
Our 4-variables (V11 ; V11 ; V12 ; V12 ) system has four eigenvalues. The type of
behavior can be characterized as a function of the position of these eigenvalues in
the Re/Im plane. Five non-degenerated cases can be distinguished: (1) the four
eigenvalues are real and negative (stable steady state), (2) the four eigenvalues are
real, three of them are negative (unstable steady state), (3) and (4) two eigenvalues
are complex conjugates with a negative real part and the other eigenvalues are real
negative (stable steady state), two cases can be distinguished depending on the
relative value of the real part of the complex eigenvalues and of the real one,
(5) two eigenvalues are complex conjugates with a negative real part and other
eigenvalues real are positive (unstable steady state) [2–4].
We define
0 0ðiÞ ðiÞ
V11 ðt D1 Þ ¼ V11 þ v011 ekðtD1 Þ ; V11 ðt s1 Þ ¼ V11 þ v11 ekðts1 Þ
0 0ðiÞ ðiÞ
V12 ðt D2 Þ ¼ V12 þ v012 ekðtD2 Þ ; V12 ðt s2 Þ ¼ V12 þ v12 ekðts2 Þ
Then we get four delayed differential equations with respect to coordinates

0 0
½V11 V11 V12 V12 arbitrarily small increments of exponential ½v011 v11 v012 v12 ekt .
k ekt v011 ¼ C11 ekðtD1 Þ v011 þ C12 ekðts1 Þ v11 ; k ekt v11 ¼ ekðtD1 Þ v011
k ekt v012 ¼ C33 ekðtD2 Þ v012 þ C34 ekðts2 Þ v12 ; k ekt v12 ¼ ekðtD2 Þ v012
0 ði¼0Þ ði¼0Þ 0 ði¼0Þ ði¼0Þ

In the equilibrium fixed point V11 ¼ 0; V11 ¼ 0,V12 ¼ 0; V12 ¼ 0.
The small increments Jacobian of our RFID shifted gate system is as bellow:
11 ¼ k þ C11 ekD1 ; 12 ¼ C12 eks1 ; 13 ¼ 0; 14 ¼ 0;

21 ¼ ekD1 ; 22 ¼ k; 23 ¼ 0; 24 ¼ 0; 31 ¼ 0; 32 ¼ 0
!33 ¼ k þ C33 ekD2 ; !34 ¼ C34 eks2 ; !41 ¼ 0; !42 ¼ 0;

0 0 1
0 1 v11
!11 . . . !14 B C
B . . . C B B v11 CC
B . .. .. C C ¼ 0; !43 ¼ ekD2 ; !44 ¼ k
@ . AB B v0 C
@ 12 A
!41 !44
v12
0 1
!11 . . . !14
B . .. .. C
AkI ¼B @ .. .
C
. A; det jA k Ij ¼ 0
!41 !44
P
2 P
2
k½ si þ Dj
Dðk; s1 ; s2 ; D1 ; D2 Þ ¼ k þ C12 C34 e
4 i¼1 j¼1
P
2 P
2
k½s2 þ Dj k½s1 þ Dj
þ k fC11 C34 e j¼1
þ C33 C12 e j¼1
g
P
2
k Dj
kðD2 þ s2 Þ kðD1 þ s1 Þ
þ k fC34 e
2
C12 e þ C11 C33 e j¼1
g
k3 fC33 ekD2 þ C11 ekD1 g
We have three stability cases: s1 ¼ s2 ¼ s & D1 ¼ D2 ¼ 0 or s1 ¼ s2 ¼

0 & D1 ¼ D2 ¼ D or s1 ¼ s2 ¼ D1 ¼ D2 ¼ sD otherwise s1 6¼ s2 & D1 6¼ D2 and
they are positive parameters. There are other possible simple stability cases: s1 ¼
s; s2 ¼ 0; D1 ¼ D2 ¼ 0 or s1 ¼ 0; s2 ¼ s; D1 ¼ D2 ¼ 0. s1 ¼ s2 ¼ 0; D1 ¼
D; D2 ¼ 0 or s1 ¼ s2 ¼ 0; D1 ¼ 0; D2 ¼ D. We need to get characteristics equa-
tions for all above stability analysis cases. We study the occurrence of any possible
stability switching, resulting from the increase the value of the time delays s; D; sD
for the general characteristic equation Dðk; s=D=sD Þ. If we choose s parameter, then
Dðk; sÞ ¼ Pn ðk; sÞ þ Qm ðk; sÞ eks . The expression for Pn ðk; sÞ; Pn ðk; sÞ ¼
P n
Pk ðsÞ k ¼ P0 ðsÞ þ P1 ðsÞ k þ P2 ðsÞ k þ P3 ðsÞ k þ . . .:
k 2 3
k¼0
P
m
The expression for Qm ðk; sÞ is Qm ðk; sÞ ¼ qk ðsÞ kk ¼ q0 ðsÞ þ q1 ðsÞ k þ
k¼0
q2 ðsÞ k2 þ . . .:
First, we discuss RFID shifted gate system fourth order characteristic equation
for s1 ¼ s; s2 ¼ 0; D1 ¼ D2 ¼ 0. The first case we analyze is when there is a delay
in RFID first gate’s primary loop antenna voltage and no delay in secondary loop
antenna voltage. Additionally, there is no delay in the gate’s primary and secondary
loop antennas voltages derivatives [5, 6]. The general characteristic equation D(k,
s) is ad follow:
Dðk; sÞ ¼ k C11 C34 þ k2 ðC11 C33 C34 Þ k3 ðC33 þ C11 Þ

þ k4 þ fC12 C34 þ k C33 C12 k2 C12 g eks
Dðk; sÞ ¼ Pn ðk; sÞ þ Qm ðk; sÞ eks ; n ¼ 4; m ¼ 2; n [ m:
X
n
Pn ðk; sÞ ¼ Pk ðsÞ k ¼ P0 ðsÞ þ P1 ðsÞ k þ P2 ðsÞ k þ P3 ðsÞ k
k 2 3
k¼0
þ P4 ðsÞ k4 ¼ k C11 C34 þ k2 ðC11 C33 C34 Þ k3 ðC33 þ C11 Þ þ k4

P0 ðsÞ ¼ 0; P1 ðsÞ ¼ C11 C34 ; P2 ðsÞ ¼ C11 C33 C34 ; P3 ðsÞ ¼ ðC33 þ C11 Þ; P4 ðsÞ ¼ 1
The expression for Qm ðk; sÞ is

X
m
Qm ðk; sÞ ¼ qk ðsÞ kk ¼ q0 ðsÞ þ q1 ðsÞ k þ q2 ðsÞ k2 ¼ C12 C34 þ k C33 C12 k2 C12
k¼0
q0 ðsÞ ¼ C12 C34 ; q1 ðsÞ ¼ C33 C12 ; q2 ðsÞ ¼ C12
0 0
The homogeneous system for V11 V11 V12 V12 leads to a characteristic equation
for the eigenvalue k having the form PðkÞ þ QðkÞ eks ¼ 0; PðkÞ ¼
P4 P
2
aj k j ; QðkÞ ¼ cj k j
j¼0 j¼0
And the coefficients faj ðqi ; qk Þ; cj ðqi ; qk Þg 2 R depend on qi ; qk , but not on s.
qi ; qk are any two shifted gate system’s parameters, other parameters keep as a
constant. a0 ¼ 0; a1 ¼ C11 C34 ; a2 ¼ C11 C33 C34 ; a3 ¼ ðC33 þ C11 Þ; a4 ¼ 1
c0 ¼ C12 C34 ; c1 ¼ C33 C12 ; c2 ¼ C12 . Unless strictly necessary, the designa-
tion of the varied arguments ðqi ; qk Þ will subsequently be omitted from P, Q, aj, cj.
The coefficients aj, cj are continuous, and differentiable functions of their argu-
ments, and direct substitution shows that a0 + c0 6¼ 0 for 8 qi ; qk 2 R þ , i.e.
k = 0 is not a of PðkÞ þ QðkÞ eks ¼ 0. Furthermore, P(k), Q(k) are analytic
functions of k, for which the following requirements of the analysis [BK] can also
be verified in the present case:
(a) If k ¼ i x; x 2 R, then Pði xÞ þ Qði xÞ 6¼ 0.
(b) jQðkÞ=PðkÞj is bounded for jkj ! 1, Rek 0. No roots bifurcation from ∞.
(c) FðxÞ ¼ jPði xÞj2 jQði xÞj2 Has a finite number of zeros. Indeed, this is a
polynomial in x.
(d) Each positive root xðqi ; qk Þ of F(x) = 0 is continuous and differentiable re-
spect to qi ; qk .
We assume that Pn ðk; sÞ ¼ Pn ðkÞ and Qm ðk; sÞ ¼ Qm ðkÞ. It can’t have common
imaginary roots. That is, for any real number x;
pn ðk ¼ i x; sÞ þ Qm ðk ¼ i x; sÞ 6¼ 0;
pn ðk ¼ i x; sÞ ¼ i x C11 C34 þ i x3 ðC33 þ C11 Þ x2 ðC11 C33 C34 Þ þ x4
Qm ðk ¼ i x; sÞ ¼ i x C33 C12 þ C12 C34 þ x2 C12
pn ðk ¼ i x; sÞ þ Qm ðk ¼ i x; sÞ ¼ C12 C34 þ x2 ½C12 C11 C33 þ C34 þ x4
þ i x ½C33 C12 þ C11 C34 þ i x3 ½C33 þ C11 6¼ 0
jPði x; sÞj2 ¼ x2 C211 C234 þ x4 f2 C11 C34 ðC33 þ C11 Þ þ ðC11 C33 C34 Þ2 g
þ x6 fðC33 þ C11 Þ2 2 ðC11 C33 C34 Þg þ x8
jQði x; sÞj2 ¼ C212 C234 þ x2 C212 ð2 C34 þ C233 Þ þ x4 C212
Fðx; sÞ ¼ jPði x; sÞj2 jQði x; sÞj2 ¼ C212 C234 þ x2 fC211 C234 C212 ð2 C34 þ C233 Þg
þ x4 f2 C11 C34 ðC33 þ C11 Þ þ ðC11 C33 C34 Þ2 C212 g
þ x6 fðC33 þ C11 Þ2 2 ðC11 C33 C34 Þg þ x8
We define the following parameters for simplicity:
N0 ¼ C212 C234 ; N2 ¼ C211 C234 C212 ð2 C34 þ C233 Þ;

N4 ¼ 2 C11 C34 ðC33 þ C11 Þ þ ðC11 C33 C34 Þ2 C212
N6 ¼ ðC33 þ C11 Þ2 2 ðC11 C33 C34 Þ; N8 ¼ 1
Fðx; sÞ ¼ jPði x; sÞj2 jQði x; sÞj2 ¼ N0 þ N2 x2
X
4
þ N4 x4 þ N6 x6 þ N8 x8 ¼ N2k x2k
k¼0
P
4
Hence Fðx; sÞ ¼ 0 implies N2k x2k ¼ 0 and its roots are given by solving
k¼0
the above polynomial. Furthermore PR ði x; sÞ ¼ x2 ðC11 C33 C34 Þ þ x4
PR ði x; sÞ ¼ x2 ðC11 C33 C34 Þ þ x4

PI ði x; sÞ ¼ x fC11 C34 þ x2 ðC33 þ C11 Þg;
QR ði x; sÞ ¼ C12 C34 þ x2 C12 ; QI ði x; sÞ ¼ x C33 C12
Hence
sin hðsÞ ¼
jQði x; sÞj2
cos hðsÞ ¼
jQði x; sÞj2
fC11 C33 C34 x2 g x3 C33 C12

þ x fC11 C34 þ x2 ðC33 þ C11 Þg fC12 C34 þ x2 C12 g
sin hðsÞ ¼
C212 C234 þ x2 C212 ð2 C34 þ C233 Þ þ x4 C212
x2 fC34 C11 C33 þ x2 g fC12 C34 þ x2 C12 g

þ x2 fC11 C34 þ x2 ðC33 þ C11 Þg C33 C12
cos hðsÞ ¼
C212 C234 þ x2 C212 ð2 C34 þ C233 Þ þ x4 C212
P
4
Which jointly with Fðx; sÞ ¼ 0 ) N2k x2k ¼ 0 that is continuous and
k¼0
differentiable in s based on Lemma 1.1. Hence we use theorem 1.2. This proves the
theorem 1.3. Remark: RFID shifted gate system parameters are independent of the
delay parameter s.
Second, we discuss RFID shifted gate system fourth order characteristic equation
for s1 ¼ s2 ¼ s & D1 ¼ D2 ¼ 0. The second case we analyze is when there is a
delay in RFID gate’s primary and secondary loop antenna voltages (s1 ¼ s2 ¼ s)
and no delay in the gate’s primary and secondary loop antennas voltages derivatives
[5, 6]. The general characteristic equation D(k, s) is ad follow:
Dðk; sÞ ¼ k4 k3 ðC33 þ C11 Þ þ k2 C11 C33

þ fC12 C34 eks þ k ðC11 C34 þ C12 C33 Þ k2 ðC34 þ C12 Þg eks
Under Taylor series approximation: eks 1 k s þ 12 k2 s2 . The

Maclaurin series is a Taylor series expansion of a eks function about zero (0). We
get the following general characteristic equation D(k, s) under Taylor series
approximation:
1 2 2
eks 1 k s þ k s :
2
Dðk; sÞ ¼ k4 k3 ½C33 þ C11 þ k2 C11 C33
þ fC12 C34 þ k ½C11 C34 þ C12 C33 C12 C34 s
1
þ k2 ½ C12 C34 s2 C34 C12 g eks
2
X
n
Pn ðk; sÞ ¼ Pk ðsÞ k ¼ P0 ðsÞ þ P1 ðsÞ k þ P2 ðsÞ k
k 2
k¼0
þ P3 ðsÞ k3 þ P4 ðsÞ k4 ¼ k4 k3 ½C33 þ C11 þ k2 C11 C33

P0 ðsÞ ¼ 0; P1 ðsÞ ¼ 0 ; P2 ðsÞ ¼ C11 C33 ; P3 ðsÞ ¼ ½C33 þ C11 ; P4 ðsÞ ¼ 1:
P
m
The expression for Qm ðk; sÞ is Qm ðk; sÞ ¼ qk ðsÞ kk ¼ q0 ðsÞ þ q1 ðsÞ k þ
k¼0
q2 ðsÞ k2
X
m
Qm ðk; sÞ ¼ qk ðsÞ kk ¼ q0 ðsÞ þ q1 ðsÞ k þ q2 ðsÞ k2
k¼0
Xm
Qm ðk; sÞ ¼ qk ðsÞ kk ¼ C12 C34 þ k ½C11 C34 þ C12 C33 C12 C34 s
k¼0
1
þ k2 ½ C12 C34 s2 C34 C12 ; q0 ðsÞ ¼ C12 C34
2
1
q1 ðsÞ ¼ C11 C34 þ C12 C33 C12 C34 s; q2 ðsÞ ¼ C12 C34 s2 C34 C12
2
0 0
The homogeneous system for V11 V11 V12 V12 leads to a characteristic equation
for the eigenvalue k having the form
X
4 X
2
Pðk; sÞ þ Qðk; sÞ eks ¼ 0; PðkÞ ¼ aj k j ; QðkÞ ¼ cj k j
j¼0 j¼0
And the coefficients faj ðqi ; qk ; sÞ; cj ðqi ; qk ; sÞg 2 R depend on qi ; qk and delay
s. qi ; qk are any two shifted gate system’s parameters, other parameters keep as a
constant.
a0 ¼ 0; a1 ¼ 0; a2 ¼ C11 C33 ; a3 ¼ ½C33 þ C11 ; a4 ¼ 1
c0 ¼ C12 C34 ; c1 ¼ C11 C34 þ C12 C33 C12 C34 s;
1
c2 ¼ C12 C34 s2 C34 C12
2
Unless strictly necessary, the designation of the varied arguments ðqi ; qk Þ will
subsequently be omitted from P, Q, aj, cj. The coefficients aj, cj are continuous, and
differentiable functions of their arguments, and direct substitution shows that
a0 + c0 6¼ 0 for 8 qi ; qk 2 R þ , i.e. k = 0 is not a of Pðk; sÞ þ Qðk; sÞ eks ¼ 0.
We assume that Pn ðk; sÞ and Qm ðk; sÞ can’t have common imaginary roots. That is,
for any real number x:
pn ðk ¼ i x; sÞ þ Qm ðk ¼ i x; sÞ 6¼ 0; pn ðk ¼ i x; sÞ
¼ x4 þ i x3 ðC33 þ C11 Þ x2 C11 C33
Qm ðk ¼ i x; sÞ ¼ C12 C34 þ i x ½C11 C34 þ C12 C33 C12 C34 s

1
x2 ½ C12 C34 s2 C34 C12
2
1
pn ðk ¼ i x; sÞ þ Qm ðk ¼ i x; sÞ ¼ x x2 ½ C12 C34 s2 C34 C12 þ C11 C33
4
2
þ C12 C34 þ i x3 ðC33 þ C11 Þ þ i x ½C11 C34
þ C12 C33 C12 C34 s 6¼ 0; jPði x; sÞj2
¼ x8 þ x6 fðC33 þ C11 Þ2 2 C11 C33 g þ x4 C211 C233
jQði x; sÞj2 ¼ C212 C234 þ x2 f½C11 C34 þ C12 C33 C12 C34 s2
1
2 C12 C34 ½ C12 C34 s2 C34 C12 g
2
1
þ x4 ½ C12 C34 s2 C34 C12 2
2
We need to find the expression for Fðx; sÞ ¼ jPði x; sÞj2 jQði x; sÞj2
Fðx; sÞ ¼ jPði x; sÞj2 jQði x; sÞj2 ¼ x8 þ x6 fðC33 þ C11 Þ2 2 C11 C33 g

1
þ x4 fC211 C233 ½ C12 C34 s2 C34 C12 2 g
2
x2 f½C11 C34 þ C12 C33 C12 C34 s2
1
2 C12 C34 ½ C12 C34 s2 C34 C12 g C212 C234
2
N0 ¼ C212 C234 ; N2 ¼ ½C11 C34 þ C12 C33 C12 C34 s2

1
þ 2 C12 C34 ½ C12 C34 s2 C34 C12
2
1
N4 ¼ C211 C233 ½ C12 C34 s2 C34 C12 2 ;
2
2
N6 ¼ ðC33 þ C11 Þ 2 C11 C33 ; N8 ¼ 1
Fðx; sÞ ¼ jPði x; sÞj2 jQði x; sÞj2
X
4
¼ N0 þ N2 x2 þ N4 x4 þ N6 x6 þ N8 x8 ¼ N2k x2k
k¼0
P
4
k¼0
the above polynomial. Furthermore PR ði x; sÞ ¼ x4 x2 C11 C33
PR ði x; sÞ ¼ x4 x2 C11 C33
P1 ði x; sÞ ¼ x3 ðC33 þ C11 Þ
1
QR ði x; sÞ ¼ C12 C34 x2 ½ C12 C34 s2 C34 C12 ;
2
QI ði x; sÞ ¼ x ½C11 C34 þ C12 C33 C12 C34 s
Hence
sin hðsÞ ¼
jQði x; sÞj2
cos hðsÞ ¼
jQði x; sÞj2
fx4 x2 C11 C33 g x ½C11 C34 þ C12 C33 C12 C34 s

1
þ x3 ðC33 þ C11 Þ fC12 C34 x2 ½ C12 C34 s2 C34 C12 g
sin hðsÞ ¼ 2
C212 C234 þ x2 f½C11 C34 þ C12 C33 C12 C34 s2
1
2 C12 C34 ½ C12 C34 s2 C34 C12 g
2
1
þ x4 ½ C12 C34 s2 C34 C12 2
2
1
fx4 x2 C11 C33 g fC12 C34 x2 ½ C12 C34 s2
2
C34 C12 g þ x4 ðC33 þ C11 Þ ½C11 C34 þ C12 C33
C12 C34 s
cos hðsÞ ¼
C212 C234 þ x2 f½C11 C34 þ C12 C33 C12 C34 s2
1
2 C12 C34 ½ C12 C34 s2 C34 C12 g
2
1
þ x4 ½ C12 C34 s2 C34 C12 2
2
That is a continuous and differentiable in s based on Lemma 1.1. Hence we use

theorem 1.2. This proves the theorem 1.3. Third, we discuss RFID shifted gate
system fourth order characteristic equation for s1 ¼ s2 ¼ D1 ¼ D2 ¼ sD . The third
case we analyze is when there is a delay in RFID gate’s primary and secondary loop
antenna voltages (s1 ¼ s2 ¼ D1 ¼ D2 ¼ sD ) and delay in the gate’s primary and
secondary loop antennas voltages derivatives [5, 6]. The general characteristic
equation D(k, s) is as follows:
Dðk; sD Þ ¼ k4 þ fC12 C34 ek3sD þ k ðC11 C34 þ C12 C33 Þ ek2sD
þ k2 ðC34 þ C11 C33 C12 Þ eksD k3 ðC33 þ C11 Þg eksD
The Maclaurin series is a Taylor series expansion of eks ; e2ks ; e3ks func-
tions about zero (0). We get the following general characteristic equation D(k, s)
under Taylor series approximation:
eks 1 k s; ek2s 1 k 2 s
ek3s 1 k 3 s
.
Dðk; sD Þ ¼ k4 þ fC12 C34 þ k ðC11 C34 þ C12 C33 C12 C34 3 sD Þ

þ k2 ðC11 C33 C12 C34 ½C11 C34 þ C12 C33 2 sD Þ
þ k3 ð½C34 C11 C33 þ C12 sD C33 C11 Þg eksD
The expression for Pn ðk; sÞ being
X
n
Pn ðk; sÞ ¼ Pk ðsÞ k ¼ P0 ðsÞ þ P1 ðsÞ k þ P2 ðsÞ k þ P3 ðsÞ k
k 2 3
k¼0
þ P4 ðsÞ k4 ¼ k4
P0 ðsÞ ¼ 0; P1 ðsÞ ¼ 0 ; P2 ðsÞ ¼ 0 ; P3 ðsÞ ¼ 0 ; P4 ðsÞ ¼ 1
The expression for Qm ðk; sÞ being
X
m
Qm ðk; sÞ ¼ qk ðsÞ kk ¼ q0 ðsÞ þ q1 ðsÞ k þ q2 ðsÞ k2 þ q3 ðsÞ k3
k¼0
Qm ðk; sÞ ¼ C12 C34 þ k ðC11 C34 þ C12 C33 C12 C34 3 sD Þ
þ k3 ð½C34 C11 C33 þ C12 sD C33 C11 Þ
q0 ðsÞ ¼ C12 C34 ; q1 ðsÞ ¼ C11 C34 þ C12 C33 C12 C34 3 sD ;
q2 ðsÞ ¼ C11 C33 C12 C34 ½C11 C34 þ C12 C33 2 sD
q3 ðsÞ ¼ ½C34 C11 C33 þ C12 sD C33 C11
0 0
A homogeneous system for V11 V11 V12 V12 leads to a characteristic equation for
P4
the eigenvalue k having the form Pðk; sÞ þ Qðk; sÞ eks ¼ 0; PðkÞ ¼ aj
j¼0
P
3
k j ; QðkÞ ¼ cj k j and the coefficients faj ðqi ; qk ; sÞ; cj ðqi ; qk ; sÞg 2 R depend on
j¼0
qi ; qk and delay s. qi ; qk are any two shifted gate system’s parameters, other
parameters kept as a constant.
a0 ¼ 0; a1 ¼ 0; a2 ¼ 0; a3 ¼ 0; a4 ¼ 1; c0 ¼ C12 C34 ;
c1 ¼ C11 C34 þ C12 C33 C12 C34 3 sD
c0 ¼ C12 C34 ; c1 ¼ C11 C34 þ C12 C33 C12 C34 3 sD ;
c2 ¼ C11 C33 C12 C34 ½C11 C34 þ C12 C33 2 sD
c3 ¼ ½C34 C11 C33 þ C12 sD C33 C11 :
a0 + c0 6¼ 0 for 8 qi ; qk 2 R þ , i.e.
k = 0 is not a Pðk; sÞ þ Qðk; sÞ eks ¼ 0. We assume that Pn ðk; sÞ and
Qm ðk; sÞ can’t have common imaginary roots. That is, for any real number x:
pn ðk ¼ i x; sÞ þ Qm ðk ¼ i x; sÞ 6¼ 0
pn ðk ¼ ix; sÞ ¼ x4 ; Qm ðk ¼ i x; sÞ ¼
C12 C34 x2 ðC11 C33 C12 C34 ½C11 C34 þ C12 C33 2 sD Þ
þ i x ðC11 C34 þ C12 C33 C12 C34 3 sD Þ
i x3 ð½C34 C11 C33 þ C12 sD C33 C11 Þ
pn ðk ¼ i x; sÞ þ Qm ðk ¼ i x; sÞ ¼
C12 C34 x2 ðC11 C33 C12 C34 ½C11 C34 þ C12 C33 2 sD Þ
þ x4 þ i x ðC11 C34 þ C12 C33 C12 C34 3 sD Þ
i x3 ð½C34 C11 C33 þ C12 sD C33 C11 Þ 6¼ 0; jPði x; sÞj2 ¼ x8
jQm ðk ¼ i x; sÞj2 ¼ C212 C234 þ x2 fðC11 C34 þ C12 C33 C12 C34 3 sD Þ2
2 C12 C34 ðC11 C33 C12 C34 ½C11 C34 þ C12 C33 2 sD Þg
þ x4 fðC11 C33 C12 C34 ½C11 C34 þ C12 C33 2 sD Þ2
2 ðC11 C34 þ C12 C33 C12 C34 3 sD Þ ð½C34 C11 C33
þ C12 sD C33 C11 Þg
þ x6 ð½C34 C11 C33 þ C12 sD C33 C11 Þ2
Fðx; sÞ ¼ jPði x; sÞj2 jQði x; sÞj2 ¼ x8 x6 ð½C34 C11 C33 þ C12 sD C33 C11 Þ2
x4 fðC11 C33 C12 C34 ½C11 C34 þ C12 C33 2 sD Þ2
2 ðC11 C34 þ C12 C33 C12 C34 3 sD Þ ð½C34 C11 C33 þ C12 sD C33 C11 Þg
x2 fðC11 C34 þ C12 C33 C12 C34 3 sD Þ2
2 C12 C34 ðC11 C33 C12 C34 ½C11 C34 þ C12 C33 2 sD Þg C212 C234
N0 ¼ C212 C234 ; N2 ¼ fðC11 C34 þ C12 C33 C12 C34 3 sD Þ2

N4 ¼ fðC11 C33 C12 C34 ½C11 C34 þ C12 C33 2 sD Þ2
2 ðC11 C34 þ C12 C33 C12 C34 3 sD Þ ð½C34 C11 C33 þ C12 sD C33 C11 Þg
N6 ¼ ð½C34 C11 C33 þ C12 sD C33 C11 Þ2 ; N8 ¼ 1
X
4
Fðx; sÞ ¼ jPði x; sÞj2 jQði x; sÞj2 ¼ N0 þ N2 x2 þ N4 x4 þ N6 x6 þ N8 x8 ¼ N2k x2k
k¼0
P
4
k¼0
the above polynomial. Furthermore
PR ði x; sÞ ¼ x4 ; PI ði x; sÞ ¼ 0
QR ði x; sÞ ¼ C12 C34 x2 ðC11 C33 C12 C34 ½C11 C34 þ C12 C33 2 sD Þ
QI ði x; sÞ ¼ x fðC11 C34 þ C12 C33 C12 C34 3 sD Þ
x2 ð½C34 C11 C33 þ C12 sD C33 C11 Þg
Hence
sin hðsÞ ¼
jQði x; sÞj2
cos hðsÞ ¼
jQði x; sÞj2
x5 fðC11 C34 þ C12 C33 C12 C34 3 sD Þ

x2 ð½C34 C11 C33 þ C12 sD C33 C11 Þg
sin hðsÞ ¼
C212 C234 þ x2 fðC11 C34 þ C12 C33 C12 C34 3 sD Þ2
2 C12 C34 ðC11 C33 C12 C34 ½C11 C34 þ C12 C33
2 sD Þg þ x4 fðC11 C33 C12 C34 ½C11 C34
þ C12 C33 2 sD Þ2 2 ðC11 C34 þ C12 C33 C12 C34 3 sD Þ
ð½C34 C11 C33 þ C12 sD C33 C11 Þg þ x6 ð½C34 C11 C33
þ C12 sD C33 C11 Þ2
x4 fC12 C34 x2 ðC11 C33 C12 C34 ½C11 C34 þ C12 C33 2 sD Þg
cos hðsÞ ¼
þ x4 fðC11 C33 C12 C34 ½C11 C34 þ C12 C33 2 sD Þ2 2 ðC11 C34
þ C12 C33 C12 C34 3 sD Þ ð½C34 C11 C33 þ C12 sD C33 C11 Þg
þ x6 ð½C34 C11 C33 þ C12 sD C33 C11 Þ2
It is continuous and differentiable in sD based on Lemma 1.1. Hence we use

theorem 1.2. This proves the theorem 1.3. Next we analyze RFID shifted gate
system stability analysis under delayed variables in time. Our RFID shifted gate
homogeneous system for v011 v11 v012 v12 leads to a characteristic equation for the
eigenvalue k having the form PðkÞ þ QðkÞ eks ¼ 0; Second case s1 ¼ s2 ¼ s;
D1 ¼ D2 ¼ 0.
Dðk; s1 ¼ s2 ¼ s; D1 ¼ D2 ¼ 0Þ ¼ k4 k3 ðC33 þ C11 Þ þ k2 C11 C33


Maclaurin series is a Taylor series expansion of a eks function about zero (0). We
get the following general characteristic equation D(k, s) under Taylor series
approximation: eks 1 k s þ 12 k2 s2 .
Dðk; sÞ ¼ k4 k3 ½C33 þ C11 þ k2 C11 C33 þ fC12 C34

1
þ k ½C11 C34 þ C12 C33 C12 C34 s þ k2 ½ C12 C34 s2 C34 C12 g eks
2
We use different parameters terminology from our last characteristics parameters

definition: k ! j; pk ðsÞ ! aj ; qk ðsÞ ! cj ; n ¼ 4; m ¼ 2; n [ m
P4
Additionally Pn ðk; sÞ ! PðkÞ; Qm ðk; sÞ ! QðkÞ then PðkÞ ¼ aj k j ;
j¼0
P
2
QðkÞ ¼ cj k j
j¼0
Pk ¼ k k ½C33 þ C11 þ k C11 C33 ;

4 3 2
Qk ¼ C12 C34 þ k ½C11 C34 þ C12 C33 C12 C34 s

1
þ k2 ½ C12 C34 s2 C34 C12
2
n; m 2 N0 ; n [ m and aj ; cj : R þ 0 ! R are continuous and differentiable func-

tion of s such that a0 þ c0 6¼ 0. In the following “—” denotes complex and con-
jugate. PðkÞ; QðkÞ are analytic functions in k and differentiable in s. The coefficients
faj ðC1 ; R1 ; gate antenna parametrsÞ and cj ðC1 ; R1 ; s; gate antenna parametrs Þg 2 R
depend on RFID shifted gate system’s C1, R1,s values and antenna parameters.
a0 ¼ 0; a1 ¼ 0; a2 ¼ C11 C33 ; a3 ¼ ½C33 þ C11 ; a4 ¼ 1
c0 ¼ C12 C34 ; c1 ¼ C11 C34 þ C12 C33 C12 C34 s;
1
c2 ¼ C12 C34 s2 C34 C12
2
ðR1 ; C1 ; s; gate antenna parametrs) will subsequently be omitted from P, Q, aj, cj.
ments, and direct substitution shows that a0 þ c0 6¼ 0; C12 C34 6¼ 0.
r
g3 n3 ð1 þ Rp11 Þ2
1
L11 L12
¼ rffiffiffiffiffiffiffi 6¼ 0
g1 n1 L12 L12 L11
C12 ð1 þ þ2 K Þ ð1 þ
L11 L11 L12
L11
þ2 K Þ
L12
8 C1 ; gate antenna parameters 2 R þ
i.e. k ¼ 0 is not a root of the characteristic equation. Furthermore PðkÞ; QðkÞ are
analytic functions of k for which the following requirements of the analysis (see
Kuang 1993, Sect. 3.4) can also be verified in the present case [5, 6].
ðR1 ; C1 ; antenna parametrs) domain of interest.
(b) jQðkÞ=PðkÞj is bounded for jkj ! 1, Rek 0. No roots bifurcation from 1.
Indeed, in the limit.
QðkÞ fC12 C34 þ k ½C11 C34 þ C12 C33 C12 C34 s þ k2 ½12 C12 C34 s2 C34 C12 g
j j¼j j
PðkÞ k4 k3 ½C33 þ C11 þ k2 C11 C33
(c) FðxÞ ¼ jPði xÞj2 jQði xÞj2

1
þ x4 fC211 C233 ½ C12 C34 s2 C34 C12 2 g
2
x2 f½C11 C34 þ C12 C33 C12 C34 s2
1
2 C12 C34 ½ C12 C34 s2 C34 C12 g C212 C234
2
Has at most a finite number of zeros. Indeed, this is a polynomial in x (Degree
in x8 ).
(d) Each positive root xðR1 ; C1 ; s; gate antenna parametrs) of FðxÞ ¼ 0 being
continuous and differentiable with respect to R1 ; C1 ; s; gate antenna parametrs.
This condition can only be assessed numerically.
In addition, since the coefficients in P and Q are real, we have

Pði xÞ ¼ Pði xÞ, and Qði xÞ ¼ Qði xÞ thus k ¼ i x, x [ 0 maybe on
eigenvalue of characteristic equations. The analysis consists in identifying the roots
of the characteristic equation situated on the imaginary axis of the complex k-plane,
whereby increasing the parameters R1 ; C1 ; s; gate antenna parametrs, Rek may, at
the crossing, Change its sign from (−) to (+), i.e. from a stable focus
E ð0Þ ðV11 ; V11 ; V12 ; V12 Þ ¼ ð0; 0; 0; 0Þ to an unstable one, or vice versa. This
feature may be further assessed by examining the sign of the partial derivatives with
respect to C1 ; R1 ; s and gate antenna parameters.
@Rek
^1 ðC1 Þ ¼ ð Þ ; R1 ; s; gate antenna parametrs ¼ const
@C1 k¼ix
@Rek
^1 ðR1 Þ ¼ ð Þ ; C1 ; s; gate antenna parametrs ¼ const
@R1 k¼ix
@Rek @Rek
^1 ðL11 Þ ¼ ð Þ ; C1 ; R1 ; s ¼ const; ^1 ðL12 Þ ¼ ð Þ ; C1 ; R1 ; s ¼ const
@L11 k¼ix @L12 k¼ix
@Rek
^1 ðsÞ ¼ ð Þ ; C1 ; R1 ; gate antenna parametrs ¼ const; where x 2 R þ :
@s k¼ix
In the case s1 ¼ s2 ¼ s; D1 ¼ D2 ¼ 0 we get the following results:
PR ði x; sÞ ¼ x4 x2 C11 C33 ; PI ði x; sÞ ¼ x3 ðC33 þ C11 Þ;

1
QR ði x; sÞ ¼ C12 C34 x2 ½ C12 C34 s2 C34 C12
2
N0 ¼ C212 C234 ; N2 ¼ ½C11 C34 þ C12 C33 C12 C34 s2

1
þ 2 C12 C34 ½ C12 C34 s2 C34 C12
2
1
N4 ¼ C11 C33 ½ C12 C34 s2 C34 C12 2 ;
2 2
2
N6 ¼ ðC33 þ C11 Þ2 2 C11 C33 ; N8 ¼ 1
Fðx; sÞ ¼ jPði x; sÞj2 jQði x; sÞj2 ¼ N0 þ N2 x2

X
4
þ N4 x4 þ N6 x6 þ N8 x8 ¼ N2k x2k
k¼0
P
4
Hence Fðx; sÞ ¼ 0 implies N2k x2k ¼ 0 When writing PðkÞ ¼ PR ðkÞ þ i
k¼0
PI ðkÞ and QðkÞ ¼ QR ðkÞ þ i QI ðkÞ, and inserting k ¼ i x Into RFID Gate sys-
tem’s characteristic equation, x must satisfy the following:
jQði xÞj2
jQði xÞj2

Note that FðxÞ is dependent of s. Now it is important to notice that if s 62 I (assume
that IR þ 0 is the set where xðsÞ is a positive root of FðxÞ and for s 62 I ,xðsÞ is
not defined. Then for all s in I xðsÞ is satisfied that Fðx; sÞ ¼ 0). Then there are no
positive xðsÞ solutions for Fðx; sÞ ¼ 0, and we cannot have stability switches. For
any s 2 I where xðsÞ is a positive solution of Fðx; sÞ ¼ 0, we can define the angle
hðsÞ 2 ½0; 2 p as the solution of
sin hðsÞ ¼
jQði xÞj2
cos hðsÞ ¼
jQði xÞj2
given by
hðsÞ þ n 2 p
sn ðsÞ ¼ ; n 2 N0 ; s 2 I
xðsÞ
Let us introduce the functions I ! R ; Sn ðsÞ ¼ s sn ðsÞ; s 2 I; n 2 N0 (187)

That is a continuous and differentiable in s. In the following, the subscripts
k; x; R1 ; C1 and RFID Gate antenna parameters ðLa1 ; La2 ; Lb1 ; Lb2 ; a1 ; a2 Þ indicate
the corresponding partial derivatives. Let us first concentrate on ^ðxÞ remembering
in kðLa1 ; La2 ; Lb1 ; Lb2 ; a1 ; a2 Þ and xðLa1 ; La2 ; Lb1 ; Lb2 ; a1 ; a2 Þ, and keeping all
parameters except one (x) and s. The derivation closely follows that in reference
[BK]. Differentiating RFID characteristic equation PðkÞ þ QðkÞ eks ¼ 0 with
respect to specific parameter (x), and inverting the derivative, for convenience, one
calculates:
Remark
x ¼ R1 ; C1 ; La1 ; La2 ; Lb1 ; Lb2 ; a1 ; a2 ; etc:;

ð Þ ¼
Where Pk ¼ @P
@k ; . . .: etc., Substituting k ¼ i x, and bearing i
Pði xÞ ¼ Pði xÞ, Qði xÞ ¼ Qði xÞ then i Pk ði xÞ ¼ Px ði xÞ and
i Qk ði xÞ ¼ Qx ði xÞ. That on the surface jPði xÞj2 ¼ jQði xÞj2 , one obtains

Upon separating into real and imaginary parts, with
P ¼ PR þ i PI ; Q ¼ QR þ i QI ; Px ¼ PRx þ i PIx
Qx ¼ QRx þ i QIx ; Px ¼ PRx þ i PIx
Qx ¼ QRx þ i QIx ; P2 ¼ P2R þ P2I
When (x) can be any RFID Gate parameters R1, C1, And time delay s etc. Where
for convenience, we have dropped the arguments ði x; xÞ, and where

Fx ¼ 2 ½ðPRx PR þ PIx PI Þ ðQRx QR þ QIx QI Þ; xx ¼ Fx =Fx :
We define U and V:
We choose our specific parameter as time delay x = s.
PRx ¼ 2 x ½2 x2 C11 C33 ; PIx ¼ 3 x2 ðC33 þ C11 Þ;

PRs ¼ 0; PIs ¼ 0; QRs ¼ x2 C12 C34 s; QIs ¼ x C12 C34
PRs ¼ 0; PIs ¼ 0; QRs ¼ x2 C12 C34 s; QIs ¼ x C12 C34 ;
PRx PR ¼ 2 x3 ½2 x4 3 x2 C11 C33 þ C211 C233
PIx PI ¼ 3 x5 ðC33 þ C11 Þ2 ; xs ¼ Fs =Fx ;
1
QRx ¼ 2 x ½ C12 C34 s2 C34 C12
2
QIx ¼ C11 C34 þ C12 C33 C12 C34 s
1
QRx QR ¼ 2 x ½ C12 C34 s2 C34 C12 ½C12 C34
2
1
x ð C12 C34 s2 C34 C12 Þ
2
2
QIx QI ¼ x ½C11 C34 þ C12 C33 C12 C34 s2 ;
Fs ¼ 2 ½ðPRs PR þ PIs PI Þ ðQRs QR þ QIs QI Þ
1
Fs ¼ 2 x2 C12 C34 ½C11 C34 þ C12 C33 s x2 ð C12 C34 s2 C34 C12 Þ
2
PR PIx ¼ 3 x4 ðx2 C11 C33 Þ ðC33 þ C11 Þ;
PI PRx ¼ 2 x4 ðC33 þ C11 Þ ð2 x2 C11 C33 Þ
1
QR QIx ¼ ½C12 C34 x2 ð C12 C34 s2 C34 C12 Þ
2
½C11 C34 þ C12 C33 C12 C34 s
QI QRx ¼ 2 x2 ðC11 C34 þ C12 C33 C12 C34 sÞ
1
ð C12 C34 s2 C34 C12 Þ
2
V ¼ ðPR PIs PI PRs Þ ðQR QIs QI QRs Þ;
PR PIs ¼ 0; PI PRs ¼ 0
1
QR QIs ¼ x C12 C34 ½C12 C34 x2 ð C12 C34 s2 C34 C12 Þ;
2
QI QRs ¼ x3 C12 C34 s ½C11 C34 þ C12 C33 C12 C34 s
Fðx; sÞ ¼ 0:
Differentiating with respect to s and we get
@x @x Fs @Rek
Fx þ Fs ¼ 0; s 2 I ) ¼ ; ^1 ðsÞ ¼ ð Þ
@s @s Fx @s k¼ix
2 ½U þ s jPj2 þ i Fx @x Fs
^1 ðsÞ ¼ Ref g; ¼ xs ¼
Fs þ i 2 ½V þ x jPj @s
2 Fx
@Rek
signf^1 ðsÞg ¼ signfð Þ g;
@s k¼ix
@x U @x
@s þ V
signf^1 ðsÞg ¼ signfFx g signfs þxþ g
@s jPj2
We shall presently examine the possibility of stability transitions (bifurcations)

in a shifted gate, double loop RFID system, about the equilibrium point
E ð0Þ ðV11 ; V11 ; V12 ; V12 Þ as a result of a variation of delay parameter s. The
analysis consists in identifying the roots of our system characteristic equation sit-
uated on the imaginary axis of the complex k-plane, Whereby increasing the delay
parameter s, Re k may at the crossing, changes its sign from − to +, i.e. from a
stable focus E(*) to an unstable one, or vice versa. This feature may be further
assessed by examining the sign of the partial derivatives with respect to s,
^1 ðsÞ ¼ ð@ Re k
@s Þk¼ix
@Rek
^1 ðsÞ ¼ ð Þ ; C1 ; R1 ; gate antenna parametrs ¼ const where x 2 R þ :
@s k¼ix
For our stability switching analysis, we choose typical RFID shifted gate parameters
values: L11 = 4.5 mH, L12 = 2.5 mH, C1 = 23 pF, R1 = 100 kX = 105, rp1 = 100 X,
K = 0.6, 2Lm = 0.004 (2 Lm ¼ 2 K L11 L12 ). g1 ¼ 56:22 1012
g2 ¼ 2:49 105 ; g3 ¼ 222:42; n1 ¼ 101:2 1012 ; n2 ¼ 4:492 105

g g
n3 ¼ 400:4; C11 ¼ 2 ¼ 4:42 105 ; C12 ¼ 3 ¼ 3:95 1012 ;
g1 g1
n2 n 3
C33 ¼ ¼ 4:43 105 ; C34 ¼ ¼ 3:95 1012 :
n1 n1
C21 ¼ C43 ¼ 1; C13 ¼ C14 ¼ C22 ¼ C23 ¼ C24 ¼ 0;
C31 ¼ C32 ¼ C41 ¼ C42 ¼ C44 ¼ 0
Then we get the expression Fðx; sÞ for a typical RFID shifted gate parameters
values.
Fðx; sÞ ¼ jPði x; sÞj2 jQði x; sÞj2 ¼ x8 þ x6 39:16 1010

þ x4 f383:17 1020 ½7:8 1024 s2 þ 7:9 1012 2 g
x2 f½34:94 1017 15:6 1024 s2
31:2 1024 ½7:8 1024 s2 þ 7:9 1012 g 243:39 1048

complex, and imaginary values of x for specific s values. s 2 ½0:001::10
And we can be express by 3D function Fðx; sÞ ¼ 0. Since it is a very complex
function, we recommend to solve it numerically rather than analytic.
double gate system. Since it is a very complex function, we recommend to solve it
numerically rather than analytic.
^1 ðsÞ ¼ ð Þk¼ix ¼ Ref g;
s ¼ xh þþ ðsÞ
ðsÞ and h þ ðsÞ is the solution of sin hðsÞ ¼ . . .; cos hðsÞ ¼ . . . when x ¼
x þ ðsÞ if only x þ is feasible. Additionally, when all RFID double gate system’s
described in the following expression:

þ xðsÞ þ g
jPðxðsÞÞj2
Remark: we know Fðx; sÞ ¼ 0 implies its roots xi ðsÞ and finding those delays
values s which xi is feasible. There are s values which are xi complex or imaginary
numbered, then unable to analyze stability [5, 6].
Semi-passive RFID Tags with the double loop antennas environment is char-
acterized by electromagnetic interferences which can influence the shifted gate
system stability in time. There are four main RFID double loop antenna variables
which are affected by electromagnetic interferences, first and second loop antenna
voltages and voltages derivatives respectively. Each loop antennas voltage variable
under electromagnetic interferences are characterized by time delay respectively.
The two time delays are not the same, but can be categorized to some sub cases due
to interferences behavior. The first case we analyze is when there is a delay in RFID
first gate’s primary loop antenna voltage and no delay in secondary loop antenna
voltage. The second case we analyze is when there is a delay in RFID gate’s
primary and secondary loop antenna voltages (s1 ¼ s2 ¼ s) and no delay in the
gate’s primary and secondary loop antennas voltages derivatives [5, 6]. The third
case we analyze is when there is a delay in RFID gate’s primary and secondary loop
antenna voltages (s1 ¼ s2 ¼ D1 ¼ D2 ¼ sD ) and delay in the gate’s primary and
secondary loop antennas voltages derivatives [4, 5]. For simplicity of our analysis
we consider in the third case all delays are the same (there is a difference but it is
neglected in our analysis). In each case we derive the related characteristic equation.
The characteristic equation is dependent on double loop antennas overall parame-
ters and interferences time delay. Upon mathematics manipulation and [BK] the-
orems and definitions we derive the expression which gives us a clear picture on
double loop antennas stability map. The stability map gives all possible options for
stability segments, each segment belongs to different time delay value segment.
Double loop antennas arranged as a shifted gate’s stability analysis can be influ-
enced either by system overall parameter values [5, 6].
1.5 RFID TAGs Detectors Stability Analysis Under

Delayed Schottky Diode’s Internal Elements in Time
The RFID market is growing and several cost, size and DC power constraints in the
TAG itself have forced designers to abandon super heterodyne receivers for older
and simpler crystal video receiver. Consisting of a simple detector circuit and a
printed antenna, this receiver can face a stability issues due to delay elements in
time. The Schottky diode detector demodulates the signal and sends the data on to
the digital circuit of the TAG; this is the so-called “wake up” signal. A simple
RFID TAG receiver block diagram includes input antenna signal with series
1.5 RFID TAGs Detectors Stability Analysis … 73
resistance, inductor (choke), Schottky diode, and output capacitor. At a small signal
(RF Input) levels, the Schottky diode can be represented by a linear equivalent
circuit. Due to Schottky parasitic delayed in time, there is a stability issue by
analyzing the detector operation. We include two parasitic delay elements in the
Schottky equivalent circuit. We define s1, s2 as delays in time, respectively, for the
Schottky equivalent circuit. We consider first those two delays in time are not equal
s1 6¼ s2 then another three cases s1 = s & s2 = 0, s2 = s & s1 = 0, s1 = s2 = s. The
RFID receiver detector delayed in time equivalent circuit can represent as a delayed
differential equations which depending on variable parameters and delays. The
investigation of our RFID receiver detector system, differential equation based on
bifurcation theory [1], the study of possible changes in the structure of the orbits of
a delayed differential equation depending on variable parameters. We first illustrate
certain observations and analyze local bifurcations of an appropriate arbitrary scalar
delayed differential equation [2]. RFID receiver detector stability analysis is done
under different time delays respect to currents and currents derivative. All of that for
optimization of RFID receiver detector equivalent circuit parameter analysis to get
the best performance. RFID system, the reader or interrogator sends a modulated
RF signal which is received by the TAG. The Schottky diode detector demodulates
the signal and sends the data on to the digital circuits of the TAG. The reader stops
sending modulated data and illuminates the TAG with continuous wave (CW) or
un-modulated signal. The TAG’s FSK encoder and switch driver switch the load
placed on the TAG’s antenna from one state to another, causing the radar cross
section of the TAG to be changed. The weak signal reflected from the TAG is
modulated; this signal is then detected by the reader’s receiver. In this way the
reader and TAG can communicate using RF generated only in the reader. The key
performance parameter for RFID TAG detector diode is operating in the square law
region in voltage sensitivity. For incoming RF small signal from the RFID reader to
the TAG, we can use Schottky diode which represented by a linear equivalent
circuit. Rj is the junction resistance (Rv or video resistance) of the diode, where RF
power is converted into video voltage output. For maximum output, all the
incoming RF voltages should ideally appear across Rj. Cj is the junction capaci-
tance of the diode chip itself. It is a parasitic element which shorts out the junction
resistance, shunting the RF energy to the series resistance Rs. Rs is a parasitic
resistance representing losses in the diode’s bond wire, the bulk silicon at the base
of the chip and other loss mechanisms. The RF voltage appearing across Rs results
in power lost as heat. Lp and Cp are package parasitic inductance and capacitance,
respectively. Unlike the two chips parasitic, they can easily be tuned out with an
external impedance matching network. The package parasitic inductance Lp has a
parasitic delay element in time (s1). The resistance losses in the diode’s bond wire
have a parasitic delay element in time (s2). V(t) represents the RFID tag antenna
voltage in time, the incoming RF small signal from the RFID reader. We consider
ideal delay lines (TAU1, TAU2), Vs1 ! e1 Vs1 ! e2 ; e1 ; e2 e [ 0 [85]
(Fig. 1.22).
Fig. 1.22 RFID TAG receiver detector equivalent circuit
VðtÞ VA dIL
¼ IRin ; IRin ¼ IL1 þ ILP ; Vs1 ! e1 ; Vs1 ! e2 ; e1 ; e2 e [ 0; VA VB ¼ LP P
Rin dt
VB VC dIL1 dðVB VD Þ VC VD
ILP ¼ ICP þ IRS ; IRS ¼ ; VA ¼ L1 ; ICP ¼ CP ; IRj ¼
RS dt dt Rj
dðVC VD Þ dVD
ICj ¼ Cj ; IRS ¼ IRj þ ICj ; IC1 ¼ C1 ; IC1 ¼ ICP þ IRj þ ICj
dt dt
dVD IC1 dðVC VD Þ dVC dVD dVC dVD
¼ ) ICj ¼ Cj ¼ Cj ½ ) ICj ¼ Cj ½
dt C1 dt dt dt dt dt
dðVB VD Þ dVB dVD dVB IC1 VðtÞ VA
ICP ¼ CP ¼ CP ½ ¼ CP ½ ; ¼ IRin ¼ IL1 þ ILP
dt dt dt dt C1 Rin
VðtÞ L1 dIL1 VðtÞ VA

¼ IL1 þ ILP ; IRin ¼ IL1 þ ILP ) IL1 ¼ IRin ILP ¼ ILP
Rin Rin dt Rin
VðtÞ L1 d VðtÞ VA dVB ILP
½ ILP ¼ IL1 þ ILP ¼ IRin ; ICP ¼ CP ½
Rin Rin dt Rin dt C1
VC VD dðVC VD Þ dðIRj Rj Þ dIR
I Rj ¼ ) IRj Rj ¼ VC VD ; ICj ¼ Cj ¼ Cj ¼ Cj Rj j
Rj dt dt dt
ILP ¼ ICP þ IRS ) ICP ¼ ILP IRS ; IC1 ¼ ICP þ IRj þ ICj ¼ ILP IRS þ IRj þ ICj
IRS ¼ IRj þ ICj ) IC1 ¼ ILP IRS þ IRj þ ICj ¼ ILP ðIRj þ ICj Þ þ IRj þ ICj ¼ ILP
VðtÞ VA VðtÞ 1 dIL 1 dIL
IRin ¼ ¼ L1 1 ¼ ½VðtÞ L1 1 ; IL1 ¼ IRin ILP ; ICP ¼ ILP IRS
Rin Rin Rin dt Rin dt
dVC IC1 d dIR

IC1 ¼ ILP ; IRS ¼ IRj þ ICj ; ICj ¼ Cj ½ ¼ Cj ½IRj Rj ¼ Cj Rj j
dt C1 dt dt
dVB IC1 dILP dIL1 dILP dIL dIL
ICP ¼ CP ½ ; VA VB ¼ LP ) L1 VB ¼ LP ) VB ¼ L 1 1 L P P
dt C1 dt dt dt dt dt
dVB d 2 IL1 d 2 ILP dVB IC1 d 2 IL1 d 2 ILP IC1
¼ L1 2 LP 2 ; ICP ¼ CP ½ ¼ CP ½L1 2 LP 2
dt dt dt dt C1 dt dt C1
VðtÞ VA VðtÞ L1 dIL1
IL1 ¼ IRin ILP ¼ ILP ¼ ILP ; ICP ¼ ILP IRS ; IC1 ¼ ILP
Rin Rin Rin dt
VðtÞ L1 dIL1 dIL 1 dVðtÞ L1 d 2 IL1 dILP
IL1 ¼ ILP ) 1 ¼
Rin Rin dt dt Rin dt Rin dt2 dt
L1 d 2 IL1 1 dVðtÞ dILP dIL1 d 2 IL 1 dVðtÞ Rin dILP Rin dIL1

2 ¼ ) 21 ¼
Rin dt Rin dt dt dt dt L1 dt L1 dt L1 dt
VB VC
IRS ¼ IRj þ ICj ) ¼ I R j þ I Cj ;
RS
dIR dIR
ICj ¼ IRS IRj ; ICj ¼ Cj Rj j ) IRS IRj ¼ Cj Rj j
dt dt
dVðtÞ dILP dIL1 d 2 I L P I C1
I CP ¼ C P ½ Rin Rin LP 2
dt dt dt dt C1
2
dVðtÞ dIL dIL d IL IC
ILP IRS ¼ CP ½ Rin P Rin 1 LP 2 P 1
dt dt dt dt C1
dVðtÞ dILP dIL1 d 2 IL IL
IC1 ¼ ILP ) ILP IRS ¼ CP ½ Rin Rin LP 2 P P
dt dt dt dt C1
VB VC
IR S ¼ ) VB VC ¼ IRS RS ;
RS
Z
dðVB VD Þ IC d 1
I CP ¼ C P ) P ¼ ðVB VD Þ ) VB VD ¼ ICP dt
dt CP dt CP
Z
dðVC VD Þ ICj dðVC VD Þ 1
ICj ¼ Cj ) ¼ ) VC VD ¼ ICj dt
dt Cj dt Cj
Z Z
1 1
ð ÞVB VD ¼ ICP dt; ð ÞVC VD ¼ ICj dt
CP Cj
Z Z
1 1
ð Þ ð Þ ! VB VC ¼ ICP dt ICj dt ) IRS RS
CP Cj
Z Z
1 1
¼ ICP dt ICj dt
CP Cj
Z Z
1 1 dIR 1 1
IRS RS ¼ ICP dt ICj dt ) RS S ¼ I CP I Cj
CP Cj dt CP Cj
dIR 1 1 dIR 1 1
RS S ¼ ICP ICj ) RS S ¼ ðILP IRS Þ ðIRS IRj Þ
dt CP Cj dt CP Cj
dIRS 1 1 1 1 1 1
RS ¼ ðILP IRS Þ ðIRS IRj Þ ¼ ILP þ IRj IRS ð þ Þ
dt CP Cj CP Cj CP Cj
dIRS 1 1 1 1 dIRS
RS ¼ IL þ IRj IRS ð þ Þ )
dt CP P Cj CP Cj dt
1 1 1 1
¼ IL þ IRj IRS ð þ Þ
RS CP P RS Cj RS CP RS Cj
dIRS
We define Y ¼ ILP ) dt ¼ RS1CP Y þ 1
RS Cj IRj IRS ðRS1CP þ RS Cj Þ
1
dIRS 1 1 1 1
Y ¼ I LP ) ¼ Yþ IRj IRS ð þ Þ
dt RS CP RS Cj RS CP RS Cj
VðtÞ L1 dIL1 VðtÞ
¼ I L1 þ I LP ) I L1 I LP
Rin Rin dt Rin
L1 dIL1 VðtÞ Rin Rin dIL
¼ ) I L1 I LP ¼ 1
Rin dt L1 L1 L1 dt
dVðtÞ dIL dIL d 2 IL IL

ILP IRS ¼ CP ½ Rin P Rin 1 LP 2 P P
dt dt dt dt C1
dVðtÞ dILP VðtÞ Rin Rin d 2 IL IL
ILP IRS ¼ CP ½ Rin Rin ð IL1 ILP Þ LP 2 P P
dt dt L1 L1 L1 dt C1
dVðtÞ dIL Rin VðtÞ R2in R2 d 2 IL

ILP IRS ¼ CP ½ Rin P þ IL1 þ in ILP LP 2 P
dt dt L1 L1 L1 dt
ILP

C1
dVðtÞ dIL CP Rin VðtÞ

ILP þ IRS þ CP CP Rin P
dt dt L1
R2
R 2 2
d IL CP ILP
þ CP in IL1 þ CP in ILP CP LP 2 P ¼0
L1 L1 dt C1
d 2 ILP dIL R2 CP
CP LP 2
CP Rin P þ ILP ½CP in 1
dt dt L1 C1
R2 CP Rin VðtÞ dVðtÞ
þ IRS þ CP in IL1 þ CP ¼0
L1 L1 dt
dILP d 2 ILP dILP

We define: Y ¼ ILP ; X ¼ dt ; dX
dt ¼ dt2 ; dY
dt ¼ dt ¼X then we get the
expression:
dX R2 CP
CP LP CP Rin X þ Y ½CP in 1
dt L1 C1
L1 L1 dt
dX R2 CP
CP LP ¼ CP Rin X þ Y ½CP in 1
dt L1 C1
L1 L1 dt
dX Rin R2 1 1
¼ X þ Y ½ in
dt LP L1 LP C1 LP CP LP
1 R2in Rin VðtÞ 1 dVðtÞ
þ IR S þ IL þ
CP LP L1 LP 1 L1 LP LP dt
dY dIL VðtÞ Rin Rin dIR 1 1

¼ X; 1 ¼ I L1 Y; j ¼ IR S IRj
dt dt L1 L1 L1 dt Cj Rj Cj Rj
dIRS 1 1 1 1 1
¼ Yþ IRj IRS ð þ Þ
dt RS CP RS Cj RS CP Cj
We have five variables in our system: X; Y; IL1 ; IRj ; IRS and we can represent our
system as the following set of differential equations matrix representation.
0 1
dX
0 1
B dt C Rin 0 1
B C 0 1 1
B dY C X B L1 LP C
B C B C B C
B dt C 0 1 B YC B C B LP C
B C N . . . N1n B C B 0 C B 0C
B dIL1 C B .11 .. .. C B C B C B C
B C¼@ . B C B 1 C VðtÞ þ B C dVðtÞ
B dt C . . . A B IL1 C þ B C B 0 C dt
B C B C B C B C
B dIR C Nm1 Nmn n¼m¼5 @ IRj A B L1 C B 0C
B j C B C @ A
B C IRS @ 0 A
B dt C 0
@ dI A 0
RS
dt
Rin R2in 1 1 1 R2in 1 1 R2in

N11 ¼ ; N12 ¼ ¼ ð Þ; N13 ¼ ; N14 ¼ 0
LP L1 LP C1 LP CP LP LP L1 C1 CP L1 LP
1 Rin Rin
N15 ¼ ; N21 ¼ 1; N22 ¼ N23 ¼ N24 ¼ N25 ¼ 0; N31 ¼ 0; N32 ¼ ; N33 ¼
CP LP L1 L1
1 1 1
N34 ¼ N35 ¼ 0; N41 ¼ N42 ¼ N43 ¼ 0; N44 ¼ ; N45 ¼ ; N51 ¼ 0; N52 ¼
Cj Rj Cj Rj RS CP
1 1 1 1
N53 ¼ 0; N54 ¼ ; N55 ¼ ð þ Þ:
RS Cj RS CP Cj
We consider RF in signal VðtÞ ¼ A0 þ f ðtÞ; jf ðtÞj\1 & A0 jf ðtÞj then

df ðtÞ
VðtÞjA0 jf ðtÞj VðtÞjA0 jf ðtÞj ¼ A0 þ f ðtÞ A0 ; dVðtÞ
dt jA0 jf ðtÞj ¼ dt ! e. We can our
matrix representation: e ! 0. Due to parasitic delay elements in Schottky
equivalent circuit, s1 for the current flow through Schottky diode’s package para-
sitic inductance (LP) and s2 for the current flow through Schottky diode’s parasitic
resistance (RS).
YðtÞ ¼ ILP ðtÞ ! Yðt s1 Þ ¼ ILP ðt s1 Þ; IRS ðtÞ ! IRS ðt s2 Þ:
dI ðtÞ dI dI
And XðtÞ ¼ LdtP ; IL1 ðtÞ; IRj ðtÞ. We consider no delay effects on dY
dt ¼ dt ; dt .
LP RS
To find equilibrium points (fixed points) of the RFID tag detector, we define
lim Yðt s1 Þ ¼ YðtÞ; lim ILP ðt s1 Þ ¼ ILP ðtÞ; lim IRS ðt s2 Þ ¼ IRS ðtÞ
t!1 t!1 t!1
0 1
dX 1 0 R 1
0
B dY C 0
dt 1 X L Lin
B dt C N . . . N1n BY C B0 1 PC
B dIL1 C B .11 .. .. C B C B C
B dt C ¼ @ . B C B1 C A0 þ e
B dIR C . . . A B IL1 C þ B L1 C
B jC @ IR A @ 0 A
@ dt A Nm1 Nmn n¼m¼5 j
dIRS IRS 0
dt
In equilibrium points (fixed points)
dY dILP dIR
¼ ¼ 0; S ¼ 0 8 t s1 ; t s2
dt dt dt
9 ðt s1 Þ t; ðt s2 Þ t; t ! 1
We get five equations:
Rin R2 1 1 1
X þ Y ½ in þ IR S
LP L1 LP C1 LP CP LP CP LP
R2in Rin VðtÞ 1 dVðtÞ
þ I þ ¼0
L 1 L P L1 L1 LP LP dt
VðtÞ Rin Rin 1 1

X ¼ 0; I Y ¼ 0; I I ¼ 0
L1 L1 L1 L1 C j Rj R S C j Rj R j
1 1 1 1 1
Y þ IR j IR S ð þ Þ ¼ 0
RS CP RS Cj RS C P C j
Since X ¼ 0 then
R2in 1 1 1
Y ½ þ IR S
L1 LP C1 LP CP LP CP LP
þ IL 1 þ ¼ 0:
L1 LP L1 LP LP dt
VðtÞ Rin Rin VðtÞ
I Y ¼ 0 ) Y ¼ IL 1 :
L1 L1 L1 L1 Rin
Then
1 VðtÞ 1 1 1 1
ð IL 1 Þ þ I IR S ð þ Þ ¼ 0
RS CP Rin RS Cj Rj RS CP Cj
VðtÞ R2 1 1 1
ð IL 1 Þ ½ in þ IR S
Rin L1 LP C1 LP CP LP CP LP
þ I þ ¼0
L1 LP L1 L1 LP LP dt
We get three equations: 1

Cj Rj IR S Cj1Rj IR j ¼ 0
1 1
I I ¼ 0
Cj Rj RS Cj Rj Rj
1 VðtÞ 1 1 1 1
ð IL 1 Þ þ I IR S ð þ Þ ¼ 0
RS CP Rin RS Cj Rj RS CP Cj
VðtÞ R2 1 1 1
þ I þ ¼0
1 1
IR S I ¼ 0 ) IR j ¼ IR S
Cj Rj Cj Rj Rj
We get two equations:
1 VðtÞ 1 1 1 1
ð IL 1 Þ þ IR S IR S ð þ Þ ¼ 0
RS CP Rin RS Cj RS CP Cj
VðtÞ R2 1 1 1
R2in R VðtÞ 1 dVðtÞ
I
in
þ þ ¼0
By mathematic manipulation, we get the following two equations:
VðtÞ VðtÞ
IL 1 IR S ¼ 0 ) IR S ¼ IL 1
Rin Rin
1 1 1
IL 1 ð þ Þ þ IR S
C1 CP CP
1 R2 1 1 Rin dVðtÞ
þ VðtÞ f ½ in gþ ¼0
Rin L1 C1 CP L1 dt
We define for simplicity:
1 R2in 1 1 Rin
X¼ ½
Rin L1 C1 CP L1
VðtÞ 1 1 1 dVðtÞ
IR S ¼ IL 1 ; IL 1 ð þ Þ þ IR S þ VðtÞ X þ ¼0
Rin C1 CP CP dt
1 1 VðtÞ 1 dVðtÞ
IL 1 ð þ Þþð IL 1 Þ þ VðtÞ X þ ¼0
C1 CP Rin CP dt
1 1 dVðtÞ
) IL 1 þ VðtÞ ½ þ X þ ¼0
C1 Rin CP dt
1 1 dVðtÞ
IL 1 þ VðtÞ ½ þ X þ ¼0
C1 Rin CP dt
1 dVðtÞ
) IL 1 ¼ C1 fVðtÞ ½ þ X þ g
Rin CP dt
VðtÞ 1 dVðtÞ
IR S ¼ þ C1 fVðtÞ ½ þ X þ g
Rin Rin CP dt
1 1 dVðtÞ
¼ VðtÞ f þ C1 ½ þ Xg þ C1
Rin Rin CP dt
We define:
1 1 dVðtÞ
X1 ¼ þ C1 ½ þ X; IR S ¼ VðtÞ X1 þ C1
Rin Rin CP dt
dVðtÞ
IR j ¼ IR S ) IR j ¼ VðtÞ X1 þ C1 ; X ¼ 0
dt
VðtÞ 1 1 dVðtÞ
Y ¼ IL 1 ¼ VðtÞ f þ C1 ½ þ Xg þ C1
Rin Rin Rin CP dt
We can summery our system fixed points in the next table:
Fixed point coordinates Fixed points expression VðtÞjA0 jf ðtÞj ¼ A0 þ f ðtÞ A0
E ðX ; Y ; VðtÞ ¼ A0 þ f ðtÞ dVðtÞ
¼ dfdtðtÞ ! e
dt jA0 jf ðtÞj
IL 1 ; IR j ; IR S Þ jf ðtÞj\1 & A0 jf ðtÞj

X 0 0
Y VðtÞ fR1in þ C1 ½Rin1CP þ Xg þ C1 dVðtÞ A0 fR1in þ C1 ½Rin1CP þ Xg
dt
IL 1 C1 fVðtÞ ½Rin1CP þ X þ dVðtÞ C1 A0 ½Rin1CP þ X
dt g
IR j VðtÞ X1 þ C1 dVðtÞ A 0 X1
dt
IR S VðtÞ fR1in þ C1 ½Rin1CP þ Xg þ C1 dVðtÞ A0 fR1in þ C1 ½Rin1CP þ Xg
dt
equilibrium points of the RFID tag detector system consists in adding to coordinate
½X; Y; IL1 ; IRj ; IRs arbitrarily small increments of exponential form ½x; y; iL1 ; iRj ; iRs
ekt and retaining the first order terms in X; Y; IL1 ; IRj ; IRs . The system of five ho-
mogeneous equations leads to a polynomial characteristic equation in the eigen-
values. The polynomial characteristic equations accept by set the below currents
and currents derivative with respect to time into RFID tag detector system equa-
tions. RFID tag detector system fixed values with arbitrarily small increments of
exponential form ½x; y; iL1 ; iRj ; iRs ekt are: j = 0 (first fixed point), j = 1 (second
fixed point), j = 2 (third fixed point), etc.
XðtÞ ¼ X ðjÞ þ x ekt ; YðtÞ ¼ Y ðjÞ þ y ekt ;

ðjÞ
Yðt s1 Þ ¼ Y ðjÞ þ y ekðts1 Þ ; IL1 ðtÞ ¼ IL1 þ iL1 ekt
ðjÞ ðjÞ
IRj ðtÞ ¼ IRj þ iRj ekt ; IRS ðtÞ ¼ IRS þ iRS ekt ;
ðjÞ
IRS ðt s2 Þ ¼ IRS þ iRS ekðts2 Þ :
We choose these expressions for ourselves XðtÞ; YðtÞ; IL1 ðtÞ and IRj ðtÞ; IRS ðtÞ as a
small displacement ½x; y; iL1 ; iRj ; iRs from the RFID tag detector system fixed points
in time t = 0.
Xðt ¼ 0Þ ¼ X ðjÞ þ x; Yðt ¼ 0Þ ¼ Y ðjÞ þ y;
ðjÞ ðjÞ
IL1 ðt ¼ 0Þ ¼ IL1 þ iL1 ; IRj ðt ¼ 0Þ ¼ IRj þ iRj
ðjÞ
IRS ðt ¼ 0Þ ¼ IRS þ iRS
For k\0; t [ 0, the selected fixed point is stable otherwise k [ 0; t [ 0 is

unstable. Our system tends to the selected fixed point exponentially for k\0; t [ 0
otherwise go away from the selected fixed point exponentially. k Is the eigenvalue
parameter which is established if the fixed point is stable or unstable; additionally,
his absolute value (jkj) establishes the speed of flow toward or away from the
selected fixed point (Yuri 1995; Jack and Huseyin 1991). The speeds of flow toward
or away from the selected fixed point for Schottky detector system currents and
currents derivatives with respect to time are
dXðtÞ Xðt þ DtÞ XðtÞ X ðjÞ þ x ekðt þ DtÞ ½X ðjÞ þ x ekt ekDt 1 þ kDt
¼ lim ¼ lim ¼ ! k x ekt
dt Dt!1 Dt Dt!1 Dt
dYðtÞ Yðt þ DtÞ YðtÞ Y ðjÞ þ y ekðt þ DtÞ ½Y ðjÞ þ y ekt ekDt 1 þ kDt
¼ lim ¼ lim ¼ ! k y ekt
dt Dt!1 Dt Dt!1 Dt
dIL1 ðtÞ dIRj ðtÞ dIRS ðtÞ dYðt s1 Þ
¼ k iL1 ekt ; ¼ k iRj ekt ; ¼ k iRS ekt ; ¼ k y ekt eks1
dt dt dt dt
dIRS ðt s2 Þ
¼ k iRS ekt eks2
dt
First, we take Schottky detector variable X; Y; IL1 ; IRj ; IRs differential equations
and adding to coordinate ½X; Y; IL1 ; IRj ; IRs arbitrarily small increments of expo-
nential terms ½x; y; iL1 ; iRj ; iRs ekt and retaining the first order terms in
x; y; iL1 ; iRj ; iRs (VðtÞ ! e; dVðtÞ
dt ! e) then
E ðX ; Y ; IL 1 ; IR j ; IR S Þ ¼ ð0; 0; 0; 0; 0Þ;
ðj¼0Þ ðj¼0Þ ðj¼0Þ
X ðj¼0Þ ¼ 0; Y ðj¼0Þ ¼ 0; IL1 ¼ 0; IRj ¼ 0; IRS ¼ 0:
We can see that our fixed point is a saddle node. We define Yðt s1 Þ ¼
ðjÞ
Y þ y ekðts1 Þ and IRS ðt s2 Þ ¼ IRS þ iRS ekðts2 Þ . Then we get five delayed
ðjÞ
differential equations with respect to coordinates ½X; Y; IL1 ; IRj ; IRs arbitrarily small
increments of exponential ½x; y; iL1 ; iRj ; iRs ekt . We consider no delay effects on
dYðtÞ dIRS ðtÞ
dt and dt . We get the following equations:
k\0 k[0
ðjÞ
t=0 Xðt ¼ 0Þ ¼ X þx Xðt ¼ 0Þ ¼ X ðjÞ þ x
ðjÞ
Yðt ¼ 0Þ ¼ Y þy Yðt ¼ 0Þ ¼ Y ðjÞ þ y
ðjÞ ðjÞ
IL1 ðt ¼ 0Þ ¼ IL1 þ iL1 IL1 ðt ¼ 0Þ ¼ IL1 þ iL1
ðjÞ ðjÞ
IRj ðtÞ ¼ IRj þ iRj IRj ðtÞ ¼ IRj þ iRj
ðjÞ ðjÞ
IRS ðt ¼ 0Þ ¼ IRS þ iRS IRS ðt ¼ 0Þ ¼ IRS þ iRS
t>0 XðtÞ ¼ X ðjÞ þ x ejkjt XðtÞ ¼ X ðjÞ þ x ejkjt
ðjÞ jkjt
YðtÞ ¼ Y þy e YðtÞ ¼ Y ðjÞ þ y ejkjt
ðjÞ ðjÞ
IL1 ðtÞ ¼ IL1 þ iL1 ejkjt IL1 ðtÞ ¼ IL1 þ iL1 ejkjt
ðjÞ ðjÞ
IRj ðtÞ ¼ IRj þ iRj ejkjt IRj ðtÞ ¼ IRj þ iRj ejkjt
ðjÞ ðjÞ
IRS ðtÞ ¼ IRS þ iRS ejkjt IRS ðtÞ ¼ IRS þ iRS ejkjt
t > 0; t ! ∞ Xðt ! 1Þ ¼ X ðjÞ Xðt ! 1; k [ 0Þ ¼ x ejkjt
ðjÞ
Yðt ! 1Þ ¼ Y Yðt ! 1; kÞ ¼ y ejkjt
ðjÞ
IL1 ðt ! 1Þ ¼ IL1 IL1 ðt ! 1; k [ 0Þ ¼ iL1 ejkjt
ðjÞ
IRj ðt ! 1Þ ¼ IRj IRj ðt ! 1; k [ 0Þ ¼ iRj ejkjt
ðjÞ
IRS ðt ! 1Þ ¼ IRS IRS ðt ! 1; k [ 0Þ ¼ iRS ejkjt
Rin R2 1 1
k x ekt ¼ ½X ðjÞ þ x ekt þ ½Y ðjÞ þ y ekðts1 Þ ½ in
LP L1 LP C1 LP CP LP
ðjÞ 1 R2in ðjÞ
þ ½IRS þ iRS ekðts2 Þ þ ½I þ iL1 ekt
CP LP L1 LP L1
Rin VðtÞ 1 dVðtÞ dVðtÞ
þ ; VðtÞ; !e
L1 LP LP dt dt
Rin ðjÞ Rin R2 1 1

k x ekt ¼ X x ekt þ Y ðjÞ ½ in
LP LP L1 LP C1 LP CP LP
R2 1 1 ðjÞ 1
þ y ½ in ekðts1 Þ IRS
L1 LP C1 LP CP LP CP LP
1 R2in ðjÞ R2in
þ iRS ekðts2 Þ þ I L1 þ iL ekt
CP LP L1 LP L1 LP 1
Rin ðjÞ R2 1 1
k x ekt ¼ X þ Y ðjÞ ½ in
ðjÞ 1 R2in ðjÞ Rin
þ IRS þ I x ekt
CP LP L1 LP L1 LP
R2 1 1
þ y ½ in ekðts1 Þ
L1 LP C1 LP CP LP
1 R2in
þ iRS ekðts2 Þ þ iL ekt
CP LP L1 LP 1
At fixed point:
Rin ðjÞ R2 1 1 ðjÞ 1 R2in ðjÞ
X þ Y ðjÞ ½ in þ IRS þ I ¼0
LP L1 LP C1 LP CP LP CP LP L1 LP L1
Then
Rin R2 1 1
x ekt ½k þ þ y ½ in ekðts1 Þ
1 R2in
þ iRS ekðts2 Þ þ iL ekt ¼ 0
CP LP L1 LP 1
dY
¼ X ) k y ekt ¼ X ðjÞ þ x ekt :
dt
At fixed point X ðjÞ ¼ 0 ) x þ k y ¼ 0

VðtÞ Rin ðjÞ Rin
k iL1 ekt ¼ ½I þ iL1 ekt ½Y ðjÞ þ y ekðts1 Þ ; VðtÞ ! e
L1 L1 L1 L1
Rin ðjÞ Rin ðjÞ Rin kt Rin kðts1 Þ
k iL1 ekt ¼ I Y i L1 e y e :
L1 L1 L1 L1 L1
ðjÞ
At fixed point RLin1 IL1 RLin1 Y ðjÞ ¼ 0 then
Rin kt Rin kðts1 Þ

k iL1 ekt iL1 e y e ¼0
L1 L1
1 ðjÞ 1 ðjÞ
k iRj ekt ¼ ½IRS þ iRS ekðts2 Þ ½I þ iRj ekt
Cj Rj Cj Rj Rj
1 1
k iRj ekt iRj ekt þ iRS ekðts2 Þ
Cj Rj Cj Rj
1 ðjÞ 1 ðjÞ
þ I I ¼0
Cj Rj RS Cj Rj Rj
ðjÞ ðjÞ
At fixed point 1
Cj Rj IRS Cj1Rj IRj ¼ 0 then iRj ekt ½k þ Cj Rj þ
1
iRS Cj1Rj ekðts2 Þ ¼ 0
1 1 ðjÞ
k iRS ekt ¼ ½Y ðjÞ þ y ekðts1 Þ þ ½I þ iRj ekt
RS CP RS Cj Rj
ðjÞ 1 1 1
½IRS þ iRS ekðts2 Þ ð þ Þ
RS CP Cj
1 1 1 ðjÞ
k iRS ekt ¼ Y ðjÞ þ y ekðts1 Þ þ I
RS CP RS CP RS Cj Rj
1 ðjÞ 1 1 1
þ iRj ekt IRS ð þ Þ
RS Cj RS CP Cj
1 1 1
iRS ð þ Þ ekðts2 Þ
RS CP Cj
1 1 ðjÞ ðjÞ 1 1 1
k iRS ekt ¼ Y ðjÞ þ I IRS ð þ Þ
RS CP RS Cj Rj RS CP Cj
1 1
þy ekðts1 Þ þ iRj ekt
RS CP RS Cj
1 1 1
iRS ð þ Þ ekðts2 Þ
RS CP Cj
ðjÞ ðjÞ
At fixed point 1
RS CP Y ðjÞ þ 1
RS Cj IRj IRS R1S ðC1P þ C1j Þ ¼ 0 then
1 1 1
iRS ekt ½k þ ð þ Þ eks2
RS CP Cj
1 1
þy ekðts1 Þ þ iRj ekt ¼ 0
RS CP RS Cj
We can summarize our last results:
Rin R2 1 1
x ½k þ þ y ½ in eks1
R2in 1
þ iL þ iRS eks2 ¼ 0
L1 LP 1 CP LP
xky¼0
Rin ks1 Rin
y e i L1 ½ þ k ¼ 0
L1 L1
1 1
iRj ½k þ þ iRS eks2 ¼ 0
Cj Rj Cj Rj
1 1 1 1 1
y eks1 þ iRj iRS ½k þ ð þ Þ eks2 ¼ 0
RS CP RS Cj RS CP Cj
The small increments Jacobian of our RFID Schotky detector system is as

follows:
0 1
0 1 x
!11 . . . !15 B C
By C
B . .. .. C B C
B . C B iL1 C ¼ 0; !11 ¼ Rin k;
@ . . . A B C
B C LP
!51 !55 @ iRj A
iRS
R2in 1 1
!12 ¼ ½ eks1
L1 LP C1 LP CP LP
R2in 1
!13 ¼ ; !14 ¼ 0; !15 ¼ eks2 ;
L1 LP CP LP
!21 ¼ 1; !22 ¼ k; !23 ¼ !24 ¼ !25 ¼ 0
Rin ks1 Rin
!31 ¼ 0; !32 ¼ e ; !33 ¼ k;
L1 L1
!34 ¼ 0; !35 ¼ 0; !41 ¼ !42 ¼ !43 ¼ 0
1 1
!44 ¼ k; !45 ¼ eks2 ;
Cj Rj Cj Rj
1
!51 ¼ 0; !52 ¼ eks1 ; !53 ¼ 0
RS CP
1 1 1 1
!54 ¼ ; !55 ¼ ð þ Þ eks2 k
RS Cj RS CP Cj
0 1
!11 ... !15
B .. .. .. C
jA k Ij ¼ @ . . . A; detjA k Ij ¼ 0
!51 !55
We define for simplicity the following parameters:

Rin R2in 1 1
r1 ¼ ; r2 ¼ ;
R2in 1 Rin
r3 ¼ ; r4 ¼ ; r5 ¼
L1 LP CP LP L1
1 1 1 1 1 1
r6 ¼ ; r7 ¼ ; r8 ¼ ; r9 ¼ ð þ Þ
C j Rj RS CP RS Cj RS CP Cj
11 ¼ r1 k; 12 ¼ r2 eks1 ; 13 ¼ r3 ; 14 ¼ 0;
15 ¼ r4 eks2 ; 21 ¼ 1; 22 ¼ k; 23 ¼ 24 ¼ 25 ¼ 0
31 ¼ 0; 32 ¼ r5 eks1 ; 33 ¼ r5 k;
34 ¼ 0; 35 ¼ 0; 41 ¼ 42 ¼ 43 ¼ 0
44 ¼ r6 k; 45 ¼ r6 eks2 ; 51 ¼ 0;
52 ¼ r7 eks1 ; 53 ¼ 0; 54 ¼ r8 ; 55 ¼ r9 eks2 k
We need to find Dðs1 ; s2 Þ for the following cases: (A) s1 ¼ s; s2 ¼ 0

(B) s1 ¼ 0; s2 ¼ s (C) s1 ¼ s2 ¼ s. We need to get characteristics equations for all
above stability analysis cases. We study the occurrence of any possible stability
switching, resulting from the increase of the value of the time delays s1 ; s2 for the
general characteristic equation Dðs1 ; s2 Þ. If we choose s as a parameter, then the
expression: Dðk; sÞ ¼ Pn ðk; sÞ þ Qm ðk; sÞ eks ; n; m 2 N0 ; n [ m [12, 18, 19].
0 1 0 1
!11 . . . !15 r5 k 0 0
B . .. C
.. C B C
B
det@ .. r6 eks2
. . A ¼ ðr1 kÞ ðkÞ det@ 0 ðr6 þ kÞ A
ks2
!51 !55 0 r8 ðr 9 e kÞ
0 1
r5 k 0 0
B C
r2 eks1 det@ 0 ðr6 þ kÞ r6 eks2 A
0 r8 ðr9 eks2 kÞ
0 1
r5 eks1 0 0
B C
þ r3 fdet@ 0 ðr6 þ kÞ r6 eks2 A
ks1 ks2
r7 e r8 ðr9 e kÞ
0 1
0 0 0
B C
þ k det@ 0 ðr6 þ kÞ r6 eks2 Ag
0 r8 ðr9 eks2 kÞ
0 ks1
1
r5 e r5 k 0
B C
þ r4 eks2 fdet@ 0 0 ðr6 þ kÞ A
ks1
r7 e 0 r8
0 1
0 r5 k 0
B C
þ k det@ 0 0 ðr6 þ kÞ Ag
0 0 r8
0 1 0 1
0 0 0 0 r5 k 0
det@ 0 ðr6 þ kÞ r6 eks2 A ¼ 0; det@ 0 0 ðr6 þ kÞ A ¼ 0
0 r8 ðr9 eks2 kÞ 0 0 r8
We get the following expression:

0 1 0 1
!11 ... !15 r5 k 0 0
B . .. C
.. C B C
B
det@ .. r6 eks2
ks2
!51 !55 0 r 8 ðr 9 e kÞ
0 1
r5 k 0 0
B C
0 ðr9 eks2 kÞ
r8
0 1
r5 eks1 0 0
B C
þ r3 det@ 0 ðr6 þ kÞ r6 eks2 A
ks1 ks2
r7 e r8 ðr9 e kÞ
0 1
r5 eks1 r5 k 0
B C
þ r4 eks2 det@ 0 0 ðr6 þ kÞ A
r7 eks1 0 r8
First expression:
0 1
r5 k 0 0 !
B ks2 C ðr6 þ kÞ r6 eks2
det@ 0 ðr6 þ kÞ r6 e A ¼ ðr5 kÞ det
r8 ðr9 eks2 kÞ
0 r8 ðr9 eks2 kÞ
¼ ðr5 kÞ fðr6 þ kÞ ðr9 eks2 kÞ r8 r6 eks2 g
¼ ðr5 kÞ fr6 r9 eks2 þ r6 k k r9 eks2 þ k2 r8 r6 eks2 g
¼ ðr5 kÞ fr6 k þ k2 ½r6 r9 þ r8 r6 þ k r9 eks2 g
0 1
r5 k 0 0
B C
det@ 0 ðr6 þ kÞ r6 eks2 A
0 r8 ðr9 eks2 kÞ
!
ðr6 þ kÞ r6 eks2
¼ ðr5 kÞ det
r8 ðr9 eks2 kÞ
¼ ðr5 kÞ fðr6 þ kÞ ðr9 eks2 kÞ r8 r6 eks2 g
¼ ðr5 kÞ fr6 r9 eks2 þ r6 k k r9 eks2 þ k2 r8 r6 eks2 g
¼ ðr5 kÞ fr6 k þ k2 ½r6 r9 þ r8 r6 þ k r9 eks2 g
¼ r5 r6 k þ r5 k2 r5 ½r6 r9 þ r8 r6 þ k r9 eks2
r6 k2 k3 þ k ½r6 r9 þ r8 r6 þ k r9 eks2
¼ r5 r6 k þ r5 k2 ½r5 r6 r9 þ r5 r8 r6 þ k r5 r9 eks2
r6 k2 k3 þ ½k ðr6 r9 þ r8 r6 Þ þ k2 r9 eks2
¼ r5 r6 k þ ðr5 r6 Þ k2 k3
þ fr5 r6 ðr9 þ r8 Þ þ k ðr6 r9 þ r8 r6 r5 r9 Þ þ k2 r9 g eks2
We define for simplicity:
w1 ¼ r5 r6 ; w2 ¼ r5 r6 ;
w3 ¼ r5 r6 ðr9 þ r8 Þ
w4 ¼ r6 r9 þ r8 r6 r5 r9
Then we define
0 1
r5 k 0 0
ks2
0 r8 ðr9 e kÞ
¼ w1 k þ w2 k2 k3 þ fw3 þ k w4 þ k2 r9 g eks2
Second expression:
0 1
r5 eks1 0 0
B C
r7 eks1 r8 ðr9 eks2 kÞ
!
ðr6 þ kÞ r6 eks2
¼ r5 eks1 det
r8 ðr9 eks2 kÞ
¼ r5 eks1 fðr6 þ kÞ ðr9 eks2 kÞ r8 r6 eks2 g
¼ r5 eks1 fr6 r9 eks2 þ r6 k k r9 eks2 þ k2 r8 r6 eks2 g
¼ r5 eks1 fr6 k þ k2 ½r6 r9 þ r8 r6 þ k r9 eks2 g
¼ ðr6 k þ k2 Þ r5 eks1 r5 ½r6 r9 þ r8 r6 þ k r9 ekðs2 þ s1 Þ ;
w5 ¼ r6 r9 þ r8 r6
0 1
r5 eks1 0 0
B C
r7 eks1 r8 ðr9 eks2 kÞ
¼ ðr6 k þ k2 Þ r5 eks1 r5 ½w5 þ k r9 ekðs2 þ s1 Þ
Third expression:
0 1
r5 eks1 ðr5 kÞ 0
B C
det@ 0 0 ðr6 þ kÞ A
r7 eks1 0 r8

ks1 0 ðr 6 þ kÞ
¼ r5 e det
0 r8

0 ðr6 þ kÞ
ðr5 kÞ det
r7 eks1 r8

0 ðr6 þ kÞ
¼ ðr5 kÞ det ¼ ðr5 kÞ r7 eks1 ðr6 þ kÞ
r7 eks1 r8
¼ ðr5 kÞ r7 ðr6 þ kÞ eks1 ¼ r7 ðr5 r6 r5 k þ k r6 þ k2 Þ eks1
¼ r7 ðr5 r6 þ k ½r6 r5 þ k2 Þ eks1
w1 ¼ r5 r6 ; w2 ¼ r5 r6 ) w2 ¼ r6 r5
0 1
r5 eks1 ðr5 kÞ 0
det@ 0 0 ðr6 þ kÞ A ¼ r7 ðw1 k w2 þ k2 Þ eks1
r7 eks1 0 r8
We integrate our expression in below Dðs1 ; s2 Þ expression.

0 1 0 1
!11 ... !15 r5 k 0 0
B . .. C
.. C B C
B
det@ .. r6 eks2
ks2
!51 !55 0 r 8 ðr 9 e kÞ
0 1
r5 k 0 0
B C
0 r8 ðr9 eks2 kÞ
0 1
r5 eks1 0 0
B C
þ r3 det@ 0 ðr6 þ kÞ r6 eks2 A
ks1 ks2
r7 e r8 ðr9 e kÞ
0 1
r5 eks1 r5 k 0
B C
þ r4 eks2 det@ 0 0 ðr6 þ kÞ A
r7 eks1 0 r8
0 1
!11 ... !15
B . .. .. C
detB
@ .. .
C
. A ¼ ðr1 kÞ ðkÞ ½w1 k þ w2 k k
2 3
!51 !55

þ fw3 þ k w4 þ k2 r9 g eks2
r2 eks1 ½w1 k þ w2 k2 k3
þ r3 ½ðr6 k þ k2 Þ r5 eks1
r5 ½w5 þ k r9 ekðs2 þ s1 Þ
þ r4 eks2 ½r7 ðw1 k w2 þ k2 Þ eks1
0 1
!11 ... !15
B . .. .. C
B
det@ .. C
. A ¼ ðk r1 kÞ ½w1 k þ w2 k k
2 2 3
.
!51 !55
½ðw1 k þ w2 k2 k3 Þ r2 eks1
þ r2 fw3 þ k w4 þ k2 r9 g ekðs1 þ s2 Þ
þ r3 ðr6 k þ k2 Þ r5 eks1
r3 r5 ½w5 þ k r9 ekðs2 þ s1 Þ
þ ðw1 r4 r7 k w2 r4 r7 þ k2 r4 r7 Þ ekðs1 þ s2 Þ
0 1
!11 ... !15
B . .. .. C
B
det@ .. C
. A ¼ w1 k þ w2 k k
3 4 5
.
!51 !55
þ fw3 k2 þ k3 w4 þ k4 r9 g eks2
r1 w1 k2 r1 w2 k3 þ r1 k4
þ fr1 w3 k r1 w4 k2 r1 r9 k3 g eks2
ðw1 k þ w2 k2 k3 Þ r2 eks1
r2 fw3 þ k w4 þ k2 r9 g ekðs1 þ s2 Þ
þ ðr3 r6 k þ r3 k2 Þ r5 eks1
0 1
!11 ... !15
B . .. .. C
B
det@ .. C
. A ¼ r1 w1 k þ ðw1 r1 w2 Þ k þ ðw2 þ r1 Þ k k
2 3 4 5
.
!51 !55
ðw1 k þ w2 k2 k3 Þ r2 eks1
þ ðr3 r6 k þ r3 k2 Þ r5 eks1
r2 fw3 þ k w4 þ k2 r9 g ekðs1 þ s2 Þ
0 1
!11 ... !15
B . .. .. C
B
det@ .. C
2 3 4 5
.
!51 !55
þ ðw1 r2 k w2 r2 k2 þ r2 k3 Þ eks1
þ ðr3 r6 r5 k þ r3 r5 k2 Þ eks1
þ fr2 w3 k r2 w4 k2 r2 r9 g ekðs1 þ s2 Þ
þ ½r3 r5 w5 k r3 r5 r9 ekðs2 þ s1 Þ
0 1
!11 . . . !15
B . .. .. C
B
det@ .. C
2 3 4 5
.
!51 !55
þ fðr3 r6 r5 w1 r2 Þ k
þ ðr3 r5 w2 r2 Þ k2 þ r2 k3 g eks1
þ fr1 w3 k þ ðw3 r1 w4 Þ k2
þ ðw4 r1 r9 Þ k3 þ k4 r9 g eks2
þ fr2 w3 r3 r5 w5 w1 r4 r7
ðw2 r4 r7 þ r2 w4 þ r3 r5 r9 Þ k
þ ðr4 r7 r2 r9 Þ k2 g ekðs1 þ s2 Þ
We define for simplicity the following parameters:

H2 ¼ r1 w1 ; H3 ¼ w1 r1 w2 ; H4 ¼ w2 þ r1 ; H5 ¼ 1
A1 ¼ r3 r6 r5 w1 r2 ; A2 ¼r3 r5 w2 r2 ; A3 ¼r2
B1 ¼ r1 w3 ; B2 ¼ w3 r1 w4 ; B3 ¼ w4 r1 r9 ; B4 ¼ r9
C0 ¼ r2 w3 r3 r5 w5 w1 r4 r7 ;
C1 ¼ ðw2 r4 r7 þ r2 w4 þ r3 r5 r9 Þ
C2 ¼ r4 r7 r2 r9
0 1
!11 . . . !15
B . .. .. C X5 X 3
detB
@ .. .
C
. A¼ H l kl þ ½ Ak kk eks1
l¼2 k¼1
!51 !55
X4 X2
þ½ Bk kk eks2 þ ½ Ck kk ekðs1 þ s2 Þ
k¼1 k¼0
X
5 X
3
Dðs1 ; s2 Þ ¼ Hl k l þ ½ Ak kk eks1
l¼2 k¼1
X4 X
2
þ½ Bk kk eks2 þ ½ Ck kk ekðs1 þ s2 Þ
k¼1 k¼0
Three cases: (A) s1 ¼ s; s2 ¼ 0 (B)s1 ¼ 0; s2 ¼ s (C) s1 ¼ s2 ¼ s.
X
5 X4
s1 ¼ s; s2 ¼ 0; DðsÞ ¼ H l kl þ ½ B k kk
l¼2 k¼1
ðAÞ
X
3 X
2
þ½ Ak kk eks þ ½ Ck kk eks
k¼1 k¼0
X
5 X4
Dðs1 ¼ s; s2 ¼ 0Þ ¼ H l kl þ ½ B k kk
l¼2 k¼1
X
3 X2
þ½ Ak kk eks þ ½ Ck kk eks
k¼1 k¼0
X
4
Dðs1 ¼ s; s2 ¼ 0Þ ¼ B1 k þ ðHl þ Bl Þ kl
l¼2
X
2
þ H5 k5 þ ½C0 þ ðAl þ Cl Þ kl þ A3 k3 eks
l¼1
Dðk; sÞ ¼ Pn ðk; sÞ þ Qm ðk; sÞ eks ; n; m 2 N0 ; n [ m

X
4
Pn ðk; sÞ ¼ B1 k þ ðHl þ Bl Þ kl þ H5 k5 ; n ¼ 5;
l¼2
X
2
Qm ðk; sÞ ¼ ½C0 þ ðAl þ Cl Þ kl þ A3 k3 ; m ¼ 3
l¼1
X
n
Pn ðk; sÞ ¼ Pk ðsÞ k ¼ P0 ðsÞ þ P1 ðsÞ k þ P2 ðsÞ k þ P3 ðsÞ k þ . . .;
k 2 3
k¼0
X m
Qm ðk; sÞ ¼ qk ðsÞ kk ¼ q0 ðsÞ þ q1 ðsÞ k þ q2 ðsÞ k2 þ . . .
k¼0
Dðk; sÞ ¼ Pn ðk; sÞ þ Qm ðk; sÞ eks ; n ¼ 5; m ¼ 3; n [ m

Xn
Pn ðk; sÞ ¼ Pk ðsÞ k ¼ P0 ðsÞ þ P1 ðsÞ k þ P2 ðsÞ k
k 2
k¼0
þ P3 ðsÞ k3 þ P4 ðsÞ k4 þ P5 ðsÞ k5

P0 ¼ 0; P1 ¼ B1 ; P2 ¼ H2 þ B2 ; P3 ¼ H3 þ B3 ; P4 ¼ H4 þ B4 ; P5 ¼ H5
X
m
Qm ðk; sÞ ¼ qk ðsÞ kk ¼ q0 ðsÞ þ q1 ðsÞ k þ q2 ðsÞ k2 þ q3 ðsÞ k3 ;
k¼0
q0 ðsÞ ¼ C0 ; q1 ðsÞ ¼ A1 þ C1 ; q2 ðsÞ ¼ A2 þ C2
q3 ðsÞ ¼ A3 :
The homogeneous system for X; Y; IL1 ; IRj ; IRS leads to a characteristic equation
P5
for the eigenvalue k having the form Pðk; sÞ þ Qðk; sÞ eks ¼ 0; PðkÞ ¼ aj
j¼0
P
3
k j ; QðkÞ ¼ cj k j . The coefficients faj ðqi ; qk ; sÞ; cj ðqi ; qk ; sÞg 2 R depend on
j¼0
qi ; qk and delay s. qi ; qk are any Schottky detector’s global parameters, other

parameters kept as a constant.
a0 ¼ 0; a1 ¼ B1 ; a2 ¼ H2 þ B2 ; a3 ¼ H3 þ B3
a4 ¼ H4 þ B4 ; a5 ¼ H5 ; c0 ðsÞ ¼ C0 ;
c1 ðsÞ ¼ A1 þ C1 ; c2 ðsÞ ¼ A2 þ C2 ; c3 ðsÞ ¼ A3 :
subsequently be omitted from P, Q, aj, and cj. The coefficients aj, cj are continuous,
and differentiable functions of their arguments, and direct substitution shows that
a0 + c0 6¼ 0 for 8 qi ; qk 2 R þ ; that is, k = 0 is not of PðkÞ þ QðkÞ eks ¼ 0.
Furthermore, P(k), Q(k) are analytic functions of k, for which the following
requirements of the analysis (Kuang J and Cong Y 2005; Kuang Y 1993) can also
be verified in the present case:
(c) FðxÞ ¼ jPði xÞj2 jQði xÞj2 has a finite number of zeros. Indeed, this is a
polynomial in x.
(d) Each positive root xðqi ; qk Þ of F(x) = 0 is continuous and differentiable with
respect to qi ; qk .
We assume that Pn ðk; sÞ and Qm ðk; sÞ cannot have common imaginary roots.
That is for any real number x;
pn ðk ¼ i x; sÞ þ Qm ðk ¼ i x; sÞ 6¼ 0:
X
4
pn ðk ¼ i x; sÞ ¼ B1 i x þ ðHl þ Bl Þ ði xÞl þ H5 ði xÞ5
l¼2
X
4
¼ i x B1 þ ðHl þ Bl Þ il xl þ i H5 x5
l¼2
X
4
ðHl þ Bl Þ il xl ¼ ðH2 þ B2 Þ x2 þ ðH2 þ B2 Þ x4 ðH2 þ B2 Þ x3 i
l¼2
pn ðk ¼ i x; sÞ ¼ ðH2 þ B2 Þ x2 þ ðH2 þ B2 Þ x4
þ i ½x B1 ðH2 þ B2 Þ x3 þ H5 x5
X
2
Qm ðk ¼ i x; sÞ ¼ C0 þ ðAl þ Cl Þ ði xÞl i A3 x3 ;
l¼1
X
2
ðAl þ Cl Þ ði xÞl ¼ i x ðA1 þ C1 Þ ðA2 þ C2 Þ x2
l¼1
Qm ðk ¼ i x; sÞ ¼ C0 ðA2 þ C2 Þ x2 þ i ½x ðA1 þ C1 Þ A3 x3
pn ðk ¼ i x; sÞ þ Qm ðk ¼ i x; sÞ
¼ C0 ðH2 þ B2 Þ x2 ðA2 þ C2 Þ x2
þ ðH2 þ B2 Þ x4 þ i ½x B1 þ x ðA1 þ C1 Þ
ðH2 þ B2 Þ x3 A3 x3 þ H5 x5 6¼ 0
¼ C0 ðH2 þ B2 þ A2 þ C2 Þ x2
þ ðH2 þ B2 Þ x4 þ i ½x ðA1 þ C1 þ B1 Þ
ðH2 þ B2 þ A3 Þ x3 þ H5 x5 6¼ 0
jPði x; sÞj2 ¼ ½ðH2 þ B2 Þ x2 þ ðH2 þ B2 Þ x4 2

þ ½x B1 ðH2 þ B2 Þ x3 þ H5 x5 2
¼ ðH2 þ B2 Þ2 x4 þ ðH2 þ B2 Þ2 x8
2 ðH2 þ B2 Þ2 x6 þ x2 B21 B1 ðH2 þ B2 Þ x4
þ B1 H5 x6 ðH2 þ B2 Þ B1 x4
þ ðH2 þ B2 Þ2 x6 ðH2 þ B2 Þ H5 x8 þ H5 B1 x6
H5 ðH2 þ B2 Þ x8 þ H25 x10
jPði x; sÞj2 ¼ x2 B21 þ ½ðH2 þ B2 Þ 2 B1 ðH2 þ B2 Þ x4

þ ½2 B1 H5 ðH2 þ B2 Þ2 x6 þ ½ðH2 þ B2 Þ 2 H5 ðH2 þ B2 Þ x8 þ H25 x10
jQði x; sÞj2 ¼ ½C0 ðA2 þ C2 Þ x2 2 þ ½x ðA1 þ C1 Þ A3 x3 2 ¼ C02 þ ðA2 þ C2 Þ2 x4

2 C0 ðA2 þ C2 Þ x2 þ x2 ðA1 þ C1 Þ2 þ A23 x6 2 ðA1 þ C1 Þ A3 x4
jQði x; sÞj2 ¼ C02 þ ½ðA1 þ C1 Þ2 2 C0 ðA2 þ C2 Þ x2

þ ½ðA2 þ C2 Þ2 2 ðA1 þ C1 Þ A3 x4 þ A23 x6
Fðx; sÞ ¼ jPði x; sÞj2 jQði x; sÞj2 ¼ x2 B21

þ ½ðH2 þ B2 Þ 2 B1 ðH2 þ B2 Þ x4
þ ½2 B1 H5 ðH2 þ B2 Þ2 x6 þ ½ðH2 þ B2 Þ
2 H5 ðH2 þ B2 Þ x8 þ H25 x10
fC02 þ ½ðA1 þ C1 Þ2 2 C0 ðA2 þ C2 Þ x2
þ ½ðA2 þ C2 Þ2 2 ðA1 þ C1 Þ A3 x4 þ A23 x6 g
Fðx; sÞ ¼ jPði x; sÞj2 jQði x; sÞj2 ¼ C02

þ fB21 ½ðA1 þ C1 Þ2 2 C0 ðA2 þ C2 Þg x2
þ f½ðH2 þ B2 Þ 2 B1 ðH2 þ B2 Þ
½ðA2 þ C2 Þ2 2 ðA1 þ C1 Þ A3 g x4
þ f½2 B1 H5 ðH2 þ B2 Þ2 A23 g x6
þ ½ðH2 þ B2 Þ 2 H5 ðH2 þ B2 Þ x8 þ H25 x10
We define the following parameters for simplicity: P0 ; P2 ; P4 ; P6 ; P8 ; P10
P0 ¼ C02 ; P2 ¼ B21 ½ðA1 þ C1 Þ2 2 C0 ðA2 þ C2 Þ

P4 ¼ ½ðH2 þ B2 Þ 2 B1 ðH2 þ B2 Þ ½ðA2 þ C2 Þ2 2 ðA1 þ C1 Þ A3
P6 ¼ ½2 B1 H5 ðH2 þ B2 Þ2 A23 ; P8
¼ ½ðH2 þ B2 Þ 2 H5 ðH2 þ B2 Þ; P10 ¼ H25
P
5
Hence Fðx; sÞ ¼ 0 implies P2k x2k ¼ 0 and its roots are given by solving
k¼0
the above polynomial.
PR ði x; sÞ ¼ ðH2 þ B2 Þ x2 þ ðH2 þ B2 Þ x4
PI ði x; sÞ ¼ x B1 ðH2 þ B2 Þ x3 þ H5 x5 ;
QR ði x; sÞ ¼ C0 ðA2 þ C2 Þ x2
QI ði x; sÞ ¼ x ðA1 þ C1 Þ A3 x3
Hence
sin hðsÞ ¼
jQði x; sÞj2
cos hðsÞ ¼
jQði x; sÞj2

P
5
Additionally Pn ðk; sÞ ! PðkÞ; Qm ðk; sÞ ! QðkÞ Then PðkÞ ¼ aj k j ; QðkÞ ¼
j¼0
P
3
cj k j
j¼0
Pk ¼ a0 þ a1 k þ a2 k2 þ a3 k3 þ a4 k4 þ a5 k5 ;
Qk ¼ c0 þ c1 k þ c2 k2 þ c3 k3
tion of s such that a0 þ c0 6¼ 0. In the following “—” denotes complex and con-
jugate. PðkÞ; QðkÞ are analytic functions in k and differentiable in s.
The coefficients faj ðLP ; L1 ; Cf ; Rin ; Rs ; CP ; Rj ; s; . . .Þ and cj ðLP ; L1 ; Cf ; Rin ; Rs ;
CP ; Rj ; s; . . .Þg 2 R depend on RFID TAG detector system’s LP ; L1 ; Cf ; Rin ; Rs ;
CP ; Rj ; s; . . . values. Unless strictly necessary, the designation of the varied argu-
ments: ðLP ; L1 ; Cf ; Rin ; Rs ; CP ; Rj ; s; . . .Þ will subsequently be omitted from P, Q, aj,
cj. The coefficients aj, cj are continuous, and differentiable functions of their ar-
guments, and direct substitution shows that a0 þ c0 6¼ 0; a0 ¼ 0; c0 ðsÞ ¼ C0
C0 ¼ r2 w3 r3 r5 w5 w1 r4 r7
! r2 w3 r3 r5 w5 w1 r4 r7 6¼ 0:
R2 1 1
½ in w3
L1 LP C1 LP CP LP
R2in Rin 1 1
þ w5 w1 6¼ 0
L1 LP L1 CP LP RS CP
8 LP ; L1 ; Cf ; Rin ; Rs ; CP ; Rj ; s; . . . 2 R þ
i.e. k ¼ 0 is not a root of the characteristic equation. Furthermore PðkÞ; QðkÞ are
analytic functions of k for which the following requirements of the analysis (see
Kuang 1993, Sect. 3.4) can also be verified in the present case [6, 7].
ðLP ; L1 ; Cf ; Rin ; Rs ; CP ; Rj ; s; . . .Þ domain of interest.
Indeed, in the limit:
QðkÞ c 0 þ c 1 k þ c 2 k2 þ c 3 k3
j j¼j j
PðkÞ a0 þ a1 k þ a2 k2 þ a3 k3 þ a4 k4 þ a5 k5
(c) FðxÞ ¼ jPði xÞj2 jQði xÞj2 ;

X
5
Fðx; sÞ ¼ jPði x; sÞj2 jQði x; sÞj2 ¼ P2k x2k
k¼0

in x10 ).
(d) Each positive root xðLP ; L1 ; Cf ; Rin ; Rs ; CP ; Rj ; s; . . .Þ of FðxÞ ¼ 0 is continu-
ous and differentiable with respect to LP ; L1 ; Cf ; Rin ; Rs ; CP ; Rj ; s; . . ..
The condition can only be assessed numerically.

Pði xÞ ¼ Pði xÞ, and Qði xÞ ¼ Qði xÞ thus, x [ 0 maybe on eigenvalue
of characteristic equations. The analysis consists in identifying the roots of the
characteristic equation situated on the imaginary axis of the complex k-plane,
whereby increasing the parameters LP ; L1 ; Cf ; Rin ; Rs ; CP ; Rj ; s; . . ., Rek may, at the
crossing, change its sign from (−) to (+), i.e. from a stable focus E ð0Þ ðX ð0Þ ;
ð0Þ ð0Þ ð0Þ
Y ð0Þ ; IL1 ; IRj ; IRS Þj VðtÞj ¼ A þ f ðtÞ A ¼ ð0; 0; 0; 0; 0Þ to an unstable
A0 jf ðtÞj 0 0
dVðtÞ df ðtÞ
dt jA0 jf ðtÞj ¼ dt !e
A0 ! e
one, or vice versa. This feature may be further assessed by examining the sign of
the partial derivatives with respect to LP ; L1 ; Cf ; Rin ; Rs ; CP ; Rj ; s; . . . and gate
antenna parameters.
@Rek
^1 ðLP Þ ¼ ð Þ ; L1 ; Cf ; Rin ; Rs ; CP ; Rj ; s; . . . ¼ const;
@LP k¼ix
@Rek
^1 ðL1 Þ ¼ ð Þ ; LP ; Cf ; Rin ; Rs ; CP ; Rj ; s; . . . ¼ const
@L1 k¼ix
@Rek
^1 ðCf Þ ¼ ð Þ ; LP ; L1 ; Rin ; Rs ; CP ; Rj ; s; . . . ¼ const;
@Cf k¼ix
@Rek
^1 ðRin Þ ¼ ð Þ ; LP ; L1 ; Cf ; Rs ; CP ; Rj ; s; . . . ¼ const
@Rin k¼ix
@Rek
^1 ðRin Þ ¼ ð Þ ; LP ; L1 ; Cf ; Rs ; CP ; Rj ; s; . . . ¼ const;
@Rin k¼ix
@Rek
^1 ðRs Þ ¼ ð Þ ; LP ; L1 ; Cf ; Rin ; CP ; Rj ; s; . . . ¼ const
@Rs k¼ix
@Rek
^1 ðCP Þ ¼ ð Þ ; LP ; L1 ; Cf ; Rin ; Rs ; Rj ; s; . . . ¼ const;
@CP k¼ix
@Rek
^1 ðRj Þ ¼ ð Þ ; LP ; L1 ; Cf ; Rin ; Rs ; CP ; s; . . . ¼ const
@Rj k¼ix
@Rek
^1 ðsÞ ¼ ð Þ ; LP ; L1 ; Cf ; Rin ; Rs ; CP ; Rj ; . . . ¼ const; x 2 R þ :
@s k¼ix
When writing PðkÞ ¼ PR ðkÞ þ i PI ðkÞ and QðkÞ ¼ QR ðkÞ þ i QI ðkÞ, and
inserting k ¼ i x into active RFID Schottky detector system’s characteristic
equation x must satisfy the following:
sin ðx sÞ ¼ gðxÞ ¼
jQði xÞj2
cos ðx sÞ ¼ hðxÞ ¼
jQði xÞj2

Note: FðxÞ is dependent on s. Now it is important to notice that if s 62 I (assume
that IR þ 0 is the set where xðsÞ is a positive root of FðxÞ and for, s 62 I , xðsÞ is
s 2 I where xðsÞ is a positive solution of Fðx; sÞ ¼ 0, we can define the angle
sin hðsÞ ¼ . . .; cos hðsÞ ¼ . . .;

sin hðsÞ ¼ ;
jQði xÞj2
cos hðsÞ ¼ :
jQði xÞj2
; n 2 N0 ; s 2 I. Let us introduce the functions I ! R ;
Sn ðsÞ ¼ s sn ðsÞ; s 2 I; n 2 N0 that is continuous and differentiable in s. In the
following, the subscripts k; x; LP ; L1 ; Cf ; Rin ; Rs ; CP ; Rj ; . . . indicate the corre-
sponding partial derivatives. Let us first concentrate on ^ðxÞ, remember in
kðLP ; L1 ; Cf ; Rin ; Rs ; CP ; Rj ; . . .Þ and xðLP ; L1 ; Cf ; Rin ; Rs ; CP ; Rj ; . . .Þ, and keeping
all parameters except one (x) and s. The derivation closely follows that in reference
[BK]. Differentiating RFID TAG detector system characteristic equation
PðkÞ þ QðkÞ eks ¼ 0 with respect to specific parameter (x), and inverting the
derivative, for convenience, one calculates:
Remark
x ¼ LP ; L1 ; Cf ; Rin ; Rs ; CP ; Rj ; . . .; etc:;

ð Þ ¼
Where Pk ¼ @P
@k ; . . .: etc., substituting k ¼ i x and bearing Pði xÞ ¼ Pði xÞ;
Qði xÞ ¼ Qði xÞ then i Pk ði xÞ ¼ Px ði xÞ; i Qk ði xÞ ¼ Qx ði xÞ and that
on the surface jPði xÞj2 ¼ jQði xÞj2 , one obtains:


Q ¼ QR þ i QI
Px ¼ PRx þ i PIx ; Qx ¼ QRx þ i QIx ; Px ¼ PRx þ i PIx ;

Qx ¼ QRx þ i QIx ; P2 ¼ P2R þ P2I :
When (x) can be any RFID Schottky detector parameter’s LP ; L1 ; Cf ; Rin ; . . . and
time delay s etc. Where for convenience, we have dropped the arguments ði x; xÞ,
and where

Fx ¼ 2 ½ðPRx PR þ PIx PI Þ ðQRx QR þ QIx QI Þxx ¼ Fx =Fx :
We define U and V:

QI ¼ x ðA1 þ C1 Þ A3 x3
PR ¼ ðH2 þ B2 Þ x2 þ ðH2 þ B2 Þ x4 ; PI ¼ x B1 ðH2 þ B2 Þ x3 þ H5 x5 ;
QR ¼ C0 ðA2 þ C2 Þ x2
PRx ¼ 4 ðH2 þ B2 Þ x3 2 ðH2 þ B2 Þ x ¼ 2 ðH2 þ B2 Þ x ð2 x2 1Þ
PIx ¼ B1 3 ðH2 þ B2 Þ x2 þ 5 H5 x4 ;
QRx ¼ 2 ðA2 þ C2 Þ x; QIx ¼ ðA1 þ C1 Þ 3 A3 x2
PRs ¼ 0; PIs ¼ 0; QRs ¼ 0; QIs ¼ 0; xs ¼ Fs =Fx
PRx PR ¼ 2 ðH2 þ B2 Þ x ð2 x2 1Þ ½ðH2 þ B2 Þ x4 ðH2 þ B2 Þ x2
¼ 2 ðH2 þ B2 Þ x ð2 x2 1Þ ðH2 þ B2 Þ x2 ½x2 1
¼ 2 ðH2 þ B2 Þ2 x3 ð2 x2 1Þ ½x2 1
PRx PR ¼ 2 ðH2 þ B2 Þ2 x3 ð2 x2 1Þ ½x2 1;
QRx QR ¼ 2 ðA2 þ C2 Þ x ½C0 ðA2 þ C2 Þ x2
Fs ¼ 2 ½ðPRs PR þ PIs PI Þ ðQRs QR þ QIs QI Þ ¼ 0;
PR PIx ¼ ðH2 þ B2 Þ x2 ðx2 1Þ ½B1 3 ðH2 þ B2 Þ x2 þ 5 H5 x4
PI PRx ¼ 2 x2 ½B1 ðH2 þ B2 Þ x2 þ H5 x4 ðH2 þ B2 Þ ð2 x2 1Þ:
QR QIx ¼ ½C0 ðA2 þ C2 Þ x2 ½ðA1 þ C1 Þ 3 A3 x2 ;
QI QRx ¼ 2 x2 ½ðA1 þ C1 Þ A3 x2 ðA2 þ C2 Þ
V ¼ ðPR PIs PI PRs Þ ðQR QIs QI QRs Þ ¼ 0: Fðx; sÞ ¼ 0
Differentiating with respect to s and we get
@x @x Fs @Rek @x Fs
Fx þ Fs ¼ 0; s 2 I ) ¼ ; ^1 ðsÞ ¼ ð Þ ; ¼ xs ¼
@s @s Fx @s k¼ix @s Fx
2 ½U þ s jPj2 þ i Fx @Rek
^1 ðsÞ ¼ Ref g; signf^1 ðsÞg ¼ signfð Þ g
Fs þ i 2 ½V þ x jPj 2 @s k¼ix
@x U @x
@s þ V
signf^1 ðsÞg ¼ signfFx g signfs þxþ g:
@s jPj2

RFID TAG detector system, about the equilibrium point E ð0Þ ðX ð0Þ ; Y ð0Þ ;
ð0Þ ð0Þ ð0Þ
IL1 ; IRj ; IRS Þ ¼ ð0; 0; 0; 0; 0Þ as a result of a variation of delay parameter s. The
uated on the imaginary axis of the complex k-plane.
Where by increasing the delay parameter s, Re k may at the crossing, changes its
sign from − to +, i.e. from a stable focus E(*) to an unstable one, or vice versa. This
respect to s,
@Rek
^1 ðsÞ ¼ ð Þ
@s k¼ix
@Rek
^1 ðsÞ ¼ ð Þ ; LP ; L1 ; Cf ; Rin ; Rs ; CP ; Rj ; . . . ¼ const; x 2 R þ :
@s k¼ix
¼ ðH2 þ B2 Þ x2 ðx2 1Þ ½B1 3 ðH2 þ B2 Þ x2 þ 5 H5 x4
2 x2 ½B1 ðH2 þ B2 Þ x2 þ H5 x4 ðH2 þ B2 Þ ð2 x2 1Þ
½C0 ðA2 þ C2 Þ x2 ½ðA1 þ C1 Þ 3 A3 x2
2 x2 ½ðA1 þ C1 Þ A3 x2 ðA2 þ C2 Þ
The single diode detector, RL is the video load resistance which not seen in
RFID TAG receiver detector equivalent circuit. L1, the shunt inductance, provides a
current return path for the diode, and is chosen to be large compared to diode
impedance at the input or RF frequency. C1, the bypass capacitance, is chosen to be
sufficiently large that is capacitive reactance is small compared to the diode
impedance, but small enough to avoid having it resistance load the video circuit. Pin
is the RF input power applied to the detector circuit and VO is the output voltage
appearing across RL. LP is packaged parasitic inductance (Schottky linear equiva-
lent circuit). CP is package parasitic capacitance. RS is the diode’s parasitic series
resistance. Cj is junction parasitic capacitance, and Rj is the diode’s junction
resistance. LP, CP, and RL are constants. RS has some small variation with tem-
perature, but that variation is not a significant parameter in this analysis. Cj is a
function of both temperature and DC bias, but this analysis concerns itself with the
zero bias detectors and the variation with temperature is not significant. Rj is a key
element in equivalent circuit—its behavior clearly will affect the performance of the
detector circuit. For our stability switching analysis, we choose typical Schottky
detector parameter values: LP = 2 nH, RS = 1.5 X, CP = 0.08 pF, Cj = 0.2 pF,
Rj = 500 X, RL = 100 KX, Rin = 1 KX, L1 = 1 mH, C1 = 1 lF
r1 ¼ 5 1011 ; r2 ¼ 6:2492 1021 ; r3 ¼ 5 1017 ;

r4 ¼ 6:25 1021 ; r5 ¼ 106 ; r6 ¼ 1010
r7 ¼ 8:33 1012 ; r8 ¼ 3:33 1012 ; r9 ¼ 1:155 1013 ;
w1 ¼ 1016 ; w2 ¼ 1:0001 1010
w3 ¼ 8:22 1028 ; w4 ¼ 8:2212 1022 ; w5 ¼ 8:22 1022 ;
H2 ¼ 5 1027 ; H3 ¼ 5:0005 1021
H4 ¼ 5:1 1011 ; H5 ¼ 1; A1 ¼ 6:2497 1037 ;
A2 ¼ 6:2498 1031 ; A3 ¼ 6:2492 1021
B1 ¼ 4:11 1040 ; B2 ¼ 4:1106 1034 ;
B3 ¼ 5:8572 1024 ; B4 ¼ 1:155 1013
C0 ¼ 6:8997 1048 ; C1 ¼ 6:9178 1042 ;
C2 ¼ 2:0116 1034 ; P0 ¼ 4:7606 1097
P2 ¼ 4:8132 1085 ; P4 ¼ 3:3789 1075 ;
P6 ¼ 1:6897 1069 ; P8 ¼ 1:6897 1069 ; P10 ¼ 1
Then we get the expression for Fðx; sÞ Schottky diode detector parameter
values. We find those x; s values which fulfill Fðx; sÞ ¼ 0. We ignore negative,
complex, and imaginary values of x for specific s values. s 2 ½0:001. . .10, we can
be express by 3D function Fðx; sÞ ¼ 0. We plot the stability switch diagram based
on different delay values of our Schottky diode detector.
s ¼ xh þþ ðsÞ
x þ ðsÞ if only x þ is feasible. Additionally, when all Schottky diode detector’s
described in the following expression:

þ xðsÞ þ g
jPðxðsÞÞj2
Remark we know Fðx; sÞ ¼ 0 implies its roots xi ðsÞ and finding those delays
values s which xi is feasible. There are s values which give complex xi or
imaginary number, then unable to analyze stability [6, 7]. F function is independent
on s the parameter FðxÞ ¼ 0.
The results: We find those x; s values which fulfill Fðx; sÞ ¼ 0. We ignore
negative, complex, and imaginary values of x. We define new MATLAB script
parameters: p2k ! G2k (k = 0…5). Running a MATLAB script to find x values,
gives the following results:
FðxÞ ¼ 0 ) x1 ¼ 1:0e þ 034 ; x2 ¼ 0 þ 4:1106i;
x3 ¼ 0 4:1106i; x4 ; . . .; x11 ¼ 0
MATLAB script: G0 = −4.7606 * 1e97; G2 = −4.8132 * 1e85;

G4 = -3.3789 * 1e75; G6 = −1.6897 * 1e69; G8 = 1.6897 * 1e69; G10 = 1;
p = [G10 0 G8 0 G6 0 G4 0 G2 0 G0]; r = roots(p).
Next is to find those x, s values which fulfil sin hðsÞ ¼ . . .; sinðx sÞ ¼
PR QI þ PI QR
jQj2
and cos hðsÞ ¼ . . .; cosðx sÞ ¼ ðPR QRjQjþ2PI QI Þ ; jQj2 ¼ Q2R þ Q2I .
Finally, we plot the stability switch diagram
@Rek
gðsÞ ¼ ^1 ðsÞ ¼ ð Þ
@s k¼ix
gðsÞ ¼ ^1 ðsÞ ¼ ð Þk¼ix ¼
@Rek
sign½gðsÞ ¼ sign½^1 ðsÞ ¼ sign½ð Þ
@s k¼ix
2 fFx ðV þ x P2 Þ Fs ðU þ s P2 Þg
¼ sign½
Fs2 þ 4 ðV þ x P2 Þ2
Since Fs2 þ 4 ðV þ x P2 Þ2 [ 0 then

sign½^1 ðsÞ ¼ signfFx ðV þ x P2 Þ Fs ðU þ s P2 Þg

Fs
sign½^1 ðsÞ ¼ signf½Fx ½ðV þ x P2 Þ ðU þ s P2 Þg;
Fx
Fs @x @F=@x
xs ¼ ; xs ¼ ð Þ1 ¼
Fx @s @F=@s
sign½^1 ðsÞ ¼ signf½Fx ½V þ xs U þ x P2 þ xs s P2 g;
1 V þ xs U
sign½^1 ðsÞ ¼ signf½Fx ½ 2 ½ þ x þ xs sg
P P2
1 V þ xs U
sign½ 2 [ 0 ) sign½^1 ðsÞ ¼ signf½Fx ½ þ x þ xs sg
P P2
V þ xs U
sign½^1 ðsÞ ¼ sign½Fx sign½ þ x þ xs s;
P2
We check the sign of ^1 ðsÞ according the following rule:
sign½Fx sign½V þPx2 s U þ x þ xs s sign½^1 ðsÞ

± ± +
± –
If sign[K−1(s)] > 0 then the crossing proceeds from (−) to (+) respectively
(stable to unstable). If sign[K−1(s)] < 0 then the crossing proceeds from (+) to (−)
respectively (unstable to stable). Anyway the stability switching can occur only for
x = 1:0e þ 034 or x = 0 [30, 32].
1.6 RFID System Burst Switch Stability Analysis Under

Delayed Internal Diode Circuitry Parasitic Effects
in Time
There are systems which converting Radio Frequency (RF) energy into a Direct
Current (DC). In other areas, the circuit has been used to provide DC power to
operate remote autonomous devices that have no on-board power supply. In the
case of the part, a battery controlled by the burst switch is used to power the device.
CMOS (silicon) devices are equipped with a form of sleep circuitry with a current
draw at a minimum during sleep. An external input signal is used to wake-up the
device. The use of the switch requires considerably more design and analysis to
avoid false wake-up states and to ensure functionality under adverse conditions.
A simple generic burst switch is constructed from input RFID rectangular spiral
antenna, matching network, voltage doubler and load. The voltage doubler unit is
constructed from two diodes D1 and D2 with parasitic effects, delay in time. One of
1.6 RFID System Burst Switch Stability Analysis Under Delayed … 105
Fig. 1.23 A simple generic RFID burst switch diagram
the difficulties with the simple RF wake-up circuit is that spurious RF energy
(noise) could potentially waken the sleeping device. Thus, it may be necessary to
interface a low power or passive circuit (essentially a filter) between the RF switch
and the higher power consuming receiver. Spurious RF energy is presented in our
system as delay RFID antenna voltage and voltage derivative in time. The low
power circuit (filter) could be any low-power device that can be turned on for a
short period of time, increment a counter(s) and go back to sleep. In effect, this
device acts like a receiver. A watchdog timer may be used to reset the device after
extended noisy periods or after long intervals of inactivity. VA is the voltage on the
RFID rectangular spiral antenna.
dVðtÞ dVA ðtÞ

VðtÞ ¼ VA ðtÞ; ¼
dt dt
dV1 ðtÞ
We define: V1 ðtÞ ¼ VðtÞ; V2 ðtÞ ¼ dVðtÞ
dt ¼ dt . Tau1 (s1 ) and Tau2 (s2 ) delay
lines represent our diodes D1 and D1 parasitic effect delay In time, Vs1 ; Vs2 ! e
(Fig. 1.23)
ID1 ðtÞ ! ID1 ðt s1 Þ; ID2 ðtÞ ! ID2 ðt s2 Þ:
RFID burst switch matching network design: The matching network match
between RFID rectangular spiral antenna impedance to our load impedance. First,
we need to calculate our matching network input impedance Zin [85] (Fig. 1.24).
Rectangular spiral RFID antenna length calculation and Inductance/resistance
We have the following rectangular spiral RFID antenna and first we need to
calculate the total length.
Aavg
A0
w B0 Bavg
A0, B0—Overall dimensions of the coil. Aavg, Bavg—Average dimensions of

the coil. w—Track width, g—Gap between tracks, t—Track thickness, Average coil
Fig. 1.24 A simple generic RFID burst switch equivalent circuit

area: Ac = Aavg∙Bavg. Nc—Number of turns. L0 is the length of the first turn

l0 ¼ 2 ðA0 þ B0 Þ ðw þ gÞ. lk is the length of turn k + 1 [7, 8].
k ¼ 1 ) L1 ¼ A0 ðw þ gÞ þ B0 2 ðw þ gÞ þ A0 2 ðw þ gÞ þ B0 3 ðw þ gÞ
k ¼ 2 ) L2 ¼ A0 3 ðw þ gÞ þ B0 4 ðw þ gÞ þ A0 4 ðw þ gÞ þ B0 5 ðw þ gÞ
X
N c 1
LT ¼ L0 þ fA0 ½1 þ ðk 1Þ 2 ðw þ gÞ þ B0 ½2 þ ðk 1Þ 2 ðw þ gÞ
k¼1
þ A0 ½2 þ ðk 1Þ 2 ðw þ gÞ þ B0 ½3 þ ðk 1Þ 2 ðw þ gÞg
X
N c 1
fA0 ½1 þ ðk 1Þ 2 ðw þ gÞ þ B0 ½2 þ ðk 1Þ 2 ðw þ gÞ
k¼1
NX
C 1
¼ f2 ðA0 þ B0 Þ 8 k ðw þ gÞg
k¼1
NX
C 1
¼2 fðA0 þ B0 Þ 4 k ðw þ gÞg

k¼1
NX
C 1
¼ 2 ðA0 þ B0 Þ ðNC 1Þ 2 ½4 k ðw þ gÞ

k¼1
NX
C 1 NX
C 1
¼ 2 ðA0 þ B0 Þ ðNC 1Þ 8 ðw þ gÞ k; k ¼ NC 1
k¼1 k¼1
X
N c 1
k¼1
¼ 2 ðA0 þ B0 Þ ðNC 1Þ 8 ðw þ gÞ ðNC 1Þ
¼ 2 ðNC 1Þ ½A0 þ B0 4 ðw þ gÞ
LT ¼ L0 þ 2 ðNC 1Þ ½A0 þ B0 4 ðw þ gÞ

½A0 þ B0 4 ðw þ gÞ ¼ 2 ðA0 þ B0 Þ ðw þ gÞ
¼ L0 þ 2 ðA0 þ B0 Þ ð1 þ NC 1Þ ðw þ gÞ ½1 þ 8 ðNC 1Þ
¼ L0 þ 2 ðA0 þ B0 Þ NC ðw þ gÞ ð8 NC 7Þ
Final result: LT ¼ L0 þ 2 ðA0 þ B0 Þ NC ðw þ gÞ ð8 NC 7Þ

LT ¼ 2 ðA0 þ B0 Þ ðw þ gÞ þ 2 ðA0 þ B0 Þ NC ðw þ gÞ ð8 NC 7Þ
LT ¼ 2 ðA0 þ B0 Þ ð1 þ NC Þ ðw þ gÞ ½1 þ 8 NC 7
LT ¼ 2 ðA0 þ B0 Þ ð1 þ NC Þ ðw þ gÞ ½8 NC 6
LT ¼ 2 ðA0 þ B0 Þ ð1 þ NC Þ 2 ðw þ gÞ ½4 NC 3
LT ¼ 2 fðA0 þ B0 Þ ð1 þ NC Þ ðw þ gÞ ½4 NC 3g
The DC resistance of rectangular spiral RFID antenna:
LT LT
RDC ¼ ¼
r S r p a2
LT—total length of the wire. r—conductivity of the wire (mohm/m). S—Cross

section area p a2 . a—radius of the wire.
LT LT 2 fðA0 þ B0 Þ ð1 þ NC Þ ðw þ gÞ ½4 NC 3g
RDC ¼ ¼ ¼
r S r p a2 r p a2
RDC jS¼wt ¼ ¼ ¼
rS rpwt rpwt
l0 X
4
2 Aavg Bavg
Lcalc ¼ ð Xk X3 Þ NcP ; X1 ¼ Aavg lnð qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi Þ
p k¼1;k6¼3 d ðA þ A2 þ B2 Þ
avg avg avg
2 Aavg Bavg
X2 ¼ Bavg lnð qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi Þ;
d ðBavg þ A2avg þ B2avg Þ
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi ðAavg þ Bavg Þ
X3 ¼ 2 ðAavg þ Bavg A2avg þ B2avg Þ; X4 ¼
4
2 ðt þ wÞ
d¼ ; Aavg ¼ A0 Nc ðg þ wÞ; Bavg ¼ B0 Nc ðg þ wÞ
p
x—Angular frequency.
Zant ¼ RDC þ j x Lcalc ; Zin ¼ Zant k Zc1 þ j x L1

Zant Zc1 1
¼ þ j x L1 ; Zc1 ¼
Zant þ Zc1 j x C1
ðRDC þ j x Lcalc Þ jxC
1
Zin ¼ 1
þ j x L1
RDC þ j x Lcalc þ jxC 1
1
RDC þ j x Lcalc
¼ þ j x L1
½1 x2 Lcalc C1 þ j x C1 RDC
RDC þ j x Lcalc
Zin ¼
½1 x Lcalc C1 þ j x C1 RDC
2
½1 x2 Lcalc C1 j x C1 RDC
þ j x L1
½1 x2 Lcalc C1 j x C1 RDC
RDC ½1 x2 Lcalc C1 j x C1 R2DC þ j x Lcalc ½1 x2 Lcalc C1 þ x2 Lcalc C1 RDC

Zin ¼ þ j x L1
½1 x2 Lcalc C1 2 þ x2 C12 R2DC
RDC ½1 x Lcalc C1 þ x Lcalc C1 RDC þ j x fLcalc ½1 x Lcalc C1 C1
2 2 2
R2DC g
Zin ¼ þ j x L1
RDC ½1 x2 Lcalc C1 þ x2 Lcalc C1 RDC
Zin ¼
fLcalc ½1 x2 Lcalc C1 C1 R2DC g
þj x þ j x L1
RDC ½1 x Lcalc C1 þ x Lcalc C1 RDC
2 2
Zin ¼
þj x ½ þ L1
RDC ½1 x2 Lcalc C1 þ x2 Lcalc C1 RDC
A1 ¼ ;
B1 ¼ x ½ þ L1
Zin ¼ A1 þ j B1
If we neglect voltage doubler unit’s parasitic elements (Inductance, capacitance,

and resistances) then it is transparent and the load is connected directly to a
matching unit.
jxCload RL
1
RL
Zload ¼ ¼
jxCload þ RL
1 1 þ j x Cload RL
RL ð1 j x Cload RL Þ
¼
1 þ j x Cload RL ð1 j x Cload RL Þ
RL ð1 j x Cload RL Þ RL j x Cload RL RL
Zload ¼ ¼
1 þ j x Cload RL ð1 j x Cload RL Þ 1 þ x2 Cload
2 R2L
RL x Cload RL RL
Zload ¼ j
1 þ x Cload RL
2 2 2 1 þ x2 Cload
2 R2L
RL x Cload RL RL
A2 ¼ ; B2 ¼ ; Zload ¼ A2 j B2
1 þ x2 Cload RL
2 2 1 þ x2 Cload
2 R2L

If Zin ¼ Zload (complex conjugate) then maximum power is transferred from the
RFID rectangular spiral antenna to the load (no power reflections). For perfect
match A1 = A2 and B1 = B2.
RDC ½1 x2 Lcalc C1 þ x2 Lcalc C1 RDC RL

A1 ¼ A2 ) ¼
½1 x2 Lcalc C1 2
þ x2
R2DC C12 1 þ x2 Cload
2 R2L
fLcalc ½1 x2 Lcalc C1 C1 R2DC g x Cload RL RL
B1 ¼ B2 ) x ½ þ L1 ¼
½1 x2 Lcalc C1 2 þ x2 C12 R2DC 1 þ x2 Cload
2 R2L
Remark our matching is dependent on the angular frequency value, x = 2pf.

Stability analysis: we need to write our system, differential equations and ana-
lyze our system behavior under parameter variation.
IRDC ¼ ILcalc ; ILcalc ¼ I1 þ I2 ; I1 ¼ IC1 ; I2 ¼ IL1 ;

I2 ¼ I3 þ I4 ; I4 ¼ ID2 ; I3 ¼ ID1 ; I4 ¼ I5 þ I6
VC
I5 ¼ ICload ; I6 ¼ IRL ¼ ; VC1 ¼ VðtÞ; VL1 ¼ VA VB ;
RL
VD2 ¼ VB VC ; VB ¼ VD1 ; VC ¼ VCload ¼ VRL
dVðtÞ
VA ¼ VC1 ; VA ¼ VðtÞ ¼ VLcalc þ VRDC ; IC1 ¼ C1 ;
dt
dVC dIL
ICload ¼ Cload ; VL1 ¼ L1 1
dt dt
dILcalc
VLcalc ¼ Lcalc ; VRDC ¼ RDC IRDC ;
dt
ID ID
VD1 ¼ Vt ln½ 1 þ 1; VD2 ¼ Vt ln½ 2 þ 1
I0 I0
dVðtÞ dIL
ILcalc ¼ I1 þ I2 ; ILcalc ¼ C1 þ IL1 ; VL1 ¼ L1 1 ;
dt dt
VL1 ¼ VA VB ¼ VLcalc þ ILcalc RDC þ VD1
ID
VL1 ¼ VLcalc þ ILcalc RDC þ Vt ln½ 1 þ 1
I0
dIL1 dILcalc ID
) L1 ¼ Lcalc þ ILcalc RDC þ Vt ln½ 1 þ 1
dt dt I0
dIL1 d dVðtÞ
L1 ¼ Lcalc ½C1 þ IL1
dt dt dt
dVðtÞ ID
þ ½C1 þ IL1 RDC þ Vt ln½ 1 þ 1
dt I0
dIL1 d 2 VðtÞ dIL1
L1 ¼ Lcalc C1 þ Lcalc
dt dt2 dt
dVðtÞ ID
þ C1 RDC þ IL1 RDC þ Vt ln½ 1 þ 1
dt I0
dIL1 d 2 VðtÞ dIL1
L1 ¼ Lcalc C1 þ Lcalc
dt dt2 dt
dVðtÞ ID
þ C1 RDC þ IL1 RDC þ Vt ln½ 1 þ 1
dt I0
dIL1 d 2 VðtÞ dVðtÞ

ð Þ½L1 Lcalc ¼ Lcalc C1 2
þ C1 RDC þ IL1 RDC þ Vt
dt dt dt
ID
ln½ 1 þ 1
I0
I2 ¼ I3 þ I4 ) IL1 ¼ ID1 þ ID2 ; VD2 ¼ VB VC ¼ VD1 VCload ;

ID ID
VCload ¼ VC ; Vt ln½ 2 þ 1 ¼ Vt ln½ 1 þ 1 VCload
I0 I0
dVC VC
I4 ¼ I5 þ I6 ) ID2 ¼ Cload þ ;
dt RL
dIL dID1 dID2
IL1 ¼ ID1 þ ID2 ) 1 ¼ þ
dt dt dt
ID2 ID1 ID ID
Vt ln½ þ 1 ¼ Vt ln½ þ 1 VCload ) Vt ln½ 2 þ 1 ¼ Vt ln½ 1 þ 1 VC
I0 I0 I0 I0
ID2 ID1 ID1 ID
Vt ln½ þ 1 ¼ Vt ln½ þ 1 VC ) VC ¼ Vt ln½ þ 1 Vt ln½ 2 þ 1
I0 I0 I0 I0
dVC 1 1 dID1 1 1 dID2

¼ Vt ID V t ID
dt ½ þ 1 0
1 I
I0
dt ½ þ 1 0 dt
2 I
I0
dVC 1 dID1 1 dID2
) ¼ Vt Vt
dt ½ID1 þ I0 dt ½ID2 þ I0 dt
1 dID1 1 dID2
ð ÞID2 ¼ Cload fVt Vt g
½ID1 þ I0 dt ½ID2 þ I0 dt
1 ID ID
þ fVt ln½ 1 þ 1 Vt ln½ 2 þ 1g
RL I0 I0
dIL1 d 2 VðtÞ dVðtÞ

ð Þ½L1 Lcalc ¼ Lcalc C1 þ C1 RDC
dt dt2 dt
ID1
þ IL1 RDC þ Vt ln½ þ 1
I0
dID1 dID2 d 2 VðtÞ dVðtÞ

½L1 Lcalc ½ þ ¼ Lcalc C1 þ C1 RDC
dt dt dt2 dt
ID1
þ ½ID1 þ ID2 RDC þ Vt ln½ þ 1
I0
We define the following new variables:
dID2 dID1 dVðtÞ dZ d 2 VðtÞ

X¼ ;Y ¼ ;Z ¼ ; ¼
dt dt dt dt dt2
dZ
½L1 Lcalc ½Y þ X ¼ Lcalc C1 þ C1 RDC Z
dt
ID
þ ½ID1 þ ID2 RDC þ Vt ln½ 1 þ 1
I0
1 1
ð ÞID2 ¼ Cload fVt Y Vt Xg
½ID1 þ I0 ½ID2 þ I0
1 ID ID
þ fVt ln½ 1 þ 1 Vt ln½ 2 þ 1g
RL I0 I0
Cload Vt Cload Vt Vt ID ID
ID2 ¼ Y X fln½ 1 þ 1 þ ln½ 2 þ 1g
½ID1 þ I0 ½ID2 þ I0 RL I0 I0
Cload Vt Cload Vt Vt ID ID
ID2 ¼ Y X lnf½ 1 þ 1 ½ 2 þ 1g
Cload Vt Cload Vt Vt ID1 ID 2
X ¼ ID2 Y lnf½ þ 1 ½ þ 1g
ID ½ID2 þ I0 ½ID2 þ I0 ½ID þ I0 ID ID
X¼ 2 Y 2 lnf½ 1 þ 1 ½ 2 þ 1g
Cload Vt ½ID1 þ I0 RL Cload I0 I0
dID2 dID2 ID2 ½ID2 þ I0 ½ID2 þ I0
X¼ ) ¼
dt dt Cload Vt ½ID1 þ I0
½ID þ I0 ID ID
Y 2 lnf½ 1 þ 1 ½ 2 þ 1g
RL Cload I0 I0
dZ
ð Þ½L1 Lcalc ½Y þ X ¼ Lcalc C1 þ C1 RDC Z
dt
ID1
þ ½ID1 þ ID2 RDC þ Vt ln½ þ 1
I0
dZ
Lcalc C1 ¼ ½L1 Lcalc ½Y þ X C1 RDC Z
dt
ID
½ID1 þ ID2 RDC Vt ln½ 1 þ 1
I0
dZ ½L1 Lcalc RDC
¼ ½Y þ X Z
dt Lcalc C1 Lcalc
RDC Vt ID
½ID1 þ ID2 ln½ 1 þ 1
Lcalc C1 Lcalc C1 I0
We can summerize our RFID burst switch system, differential equations:
dZ ½L1 Lcalc RDC

¼ ½Y þ X Z ½ID1 þ ID2
dt Lcalc C1 Lcalc
RDC Vt ID
ln½ 1 þ 1
dZ ½L1 Lcalc ID2 ½ID2 þ I0 ½ID2 þ I0
¼ ½Y Y
dt Lcalc C1 Cload Vt ½ID1 þ I0
½ID þ I0 ID ID RDC
2 lnf½ 1 þ 1 ½ 2 þ 1g Z
RL Cload I0 I0 Lcalc
RDC Vt ID
dID2 ID2 ½ID2 þ I0 ½ID2 þ I0 ½ID2 þ I0 ID ID dID
¼ Y lnf½ 1 þ 1 ½ 2 þ 1g; 1 ¼ Y
dt Cload Vt ½ID1 þ I0 RL Cload I0 I0 dt
We have four variables in our system: Z; ID1 ; ID2 ; Y and we can represent our
dt ¼ n1 ðZ; ID1 ; ID2 ; YÞ
system as the following set of differential equations: dZ
dID2 dID
¼ n2 ðZ; ID1 ; ID2 ; YÞ; 1 ¼ n3 ðZ; ID1 ; ID2 ; YÞ;
dt dt
n1 ¼ n1 ðZ; ID1 ; ID2 ; YÞ; n2 ¼ n2 ðZ; ID1 ; ID2 ; YÞ
n3 ¼ n3 ðZ; ID1 ; ID2 ; YÞ
½L1 Lcalc ID ½ID2 þ I0 ½ID2 þ I0
n1 ¼ ½Y 2 Y
Lcalc C1 Cload Vt ½ID1 þ I0
2 lnf½ 1 þ 1 ½ 2 þ 1g Z
RDC Vt ID
ID2 ½ID2 þ I0 ½ID2 þ I0
n2 ¼ Y
Cload Vt ½ID1 þ I0
½ID þ I0 ID ID
2 lnf½ 1 þ 1 ½ 2 þ 1g; n3 ¼ Y
RL Cload I0 I0
RFID system burst switch’s voltage doubler unit is constructed from two diodes
D1 and D2 with parasitic effects, delay in time. D1 current delay in time
ID1 ðtÞ ! ID1 ðt s1 Þ and D2 current delay in time ID2 ðtÞ ! ID2 ðt s2 Þ.
Spurious RF energy is presented in our system as delay RFID antenna voltage (V
(t)) and voltage derivative (dV(t)/dt) in time. We neglect voltage delayed in time
and consider only voltage derivative delay in time (D).
dVðtÞ
ZðtÞ ¼ ; ZðtÞ ! Zðt DÞ
dt
dI dI
dt ¼ dt ; dt . To find equilibrium points (fixed
We consider no delay effects on dZ D1 D2
points) of the RFID system burst switches, we define lim ID1 ðt s1 Þ ¼ ID1 ðtÞ
t!1
lim ID2 ðt s2 Þ ¼ ID2 ðtÞ; lim Zðt DÞ ¼ ZðtÞ

t!1 t!1
In equilibrium points (fixed points)
dID1 dID dZ
¼ 0; 2 ¼ 0; ¼ 0 8 t s1 ; t s2 ;
dt dt dt
t D 9 ðt s1 Þ t; ðt s2 Þ t; ðt DÞ t; t ! 1:
dID1 dID I ½I þ I0
¼ 0 ) Y ¼ 0; 2 ¼ 0 ) D2 D2
dt dt Cload Vt

½I þ I0 I I
D2 lnf½ D1 þ 1 ½ D2 þ 1g ¼ 0
RL Cload I0 I0

I ½I þ I0 ½ID2 þ I0 I I
D2 D2 lnf½ D1 þ 1 ½ D2 þ 1g ¼ 0
Cload Vt RL Cload I0 I0
½ID 2 þ I0 ID 2 1 I
I
) ð þ lnf½ D1 þ 1 ½ D2 þ 1gÞ ¼ 0
Cload Vt RL I0 I0
Case I:
½ID 2 þ I0 dZ
¼ 0 ) ID 2 ¼ I0 ) ¼0
Cload dt
RDC Vt I
) ½ID 1 þ I0 RDC Z ln½ D1 þ 1 ¼ 0
C1 C1 I0
Case II:
ID 2 1 I I RL I I
þ lnf½ D1 þ 1 ½ D2 þ 1g ¼ 0 ) ID2 þ lnf½ D1 þ 1 ½ D2 þ 1g ¼ 0
Vt RL I0 I0 Vt I0 I0
ID 1 ID 2 RL R I
I
þ 1g ¼ ID 2 ) e Vt ID2 ¼ ½ D1 þ 1 ½ D2 þ 1:
L
lnf½ þ 1 ½
I0 I0 Vt I0 I0
Let us define
RL / X 1
/n /2 /3
/¼ ID2 ; e ¼ ¼ 1þ/þ þ þ
Vt n¼0
n! 2! 3!
RL X
1 ð RL I Þn
) e Vt ID2 ¼
V D2 t
n¼0
n!
RL X
1 ð RL ID 2 Þn X
1 ð1Þn ðRL I Þn
e Vt ID2 ¼ Vt V D2
¼ t
n¼0
n! n¼0
n!
RL ID 2 Þ2 ðRVLt ID 2 Þ3
ðRVLt
¼ 1 ID2 þ þ
Vt 2! 3!
VLt ID
R RL 1 RL 1 RL
e 2 ¼ 1 ID 2 þ ð Þ2 ðID 2 Þ2 ð Þ3 ðID 2 Þ3 þ
Vt 2 Vt 6 Vt
R
VLt ID
For easy investigation, we take e 2 ¼ 1 RVLt ID 2 ) 1 RVLt ID 2 ¼
I I
½ ID01 þ 1 ½ ID02 þ 1
VLt ID
R RL RL I I
e 2 ¼1 ID2 ) 1 ID 2 ¼ ½ D1 þ 1 ½ D2 þ 1
Vt Vt I0 I0
RL ID 1 ID 2 ID 1 ID 2 ID 1 I RL
1 ID2 ¼ ½ þ 1 ½ þ 1 ) þ þ D2 þ I ¼0
Vt I0 I0 2
I0 I0 I0 Vt D2
dZ ½L1 Lcalc I ½ID2 þ I0
¼0) ½ D2
dt Lcalc C1 Cload Vt
½ID 2 þ I0 ID 1 I RDC
lnf½ þ 1 ½ D2 þ 1g Z
RDC Vt I
½ID 1 þ ID 2 ln½ D1 þ 1 ¼ 0
Remark Our system, equilibrium points (fixed points) can be calculated numeri-
cally rather than analytically (Case I and Case II). For both cases Y* = 0,
Y* = dID1/dt = 0. At equilibrium no current is flowing through D1, D1 is in OFF
state (VB > 0).
The standard local stability analysis about any one of the equilibrium points of
the RFID system burst switch consists in adding to coordinate ½Z; ID1 ; ID2 ; Y
arbitrarily small increments of exponential form ½z; iD1 ; iD2 ; y ekt and retaining the
first order terms in Z; ID1 ; ID2 ; Y. The system of three homogeneous equations leads
to a polynomial characteristic equation in the eigenvalues. The polynomial char-
acteristic equations accept by set the below currents and currents derivative with
respect to time into RFID system burst switch equations. RFID system burst
switches fixed values with arbitrarily small increments of exponential form
½z; iD1 ; iD2 ; y ekt are: j = 0 (first fixed point), j = 1 (second fixed point), j = 2
(third fixed point), etc.
ZðtÞ ¼ Z ðjÞ þ z ekt

ðjÞ ðjÞ
ID1 ðtÞ ¼ ID1 þ iD1 ekt ; ID2 ðtÞ ¼ ID2 þ iD2 ekt ;
ðjÞ
YðtÞ ¼ Y ðjÞ þ y ekt ; ID1 ðt s1 Þ ¼ ID1 þ iD1 ekðts1 Þ
ðjÞ
ID2 ðt s2 Þ ¼ ID2 þ iD2 ekðts2 Þ ; Zðt DÞ ¼ Z ðjÞ þ z ekðtDÞ :
We choose these expressions for our ZðtÞ; ID1 ðtÞ; ID2 ðtÞ; YðtÞ as a small dis-
placement ½z; iD1 ; iD2 ; y from the RFID system burst switch fixed points in time
t = 0.
ðjÞ ðjÞ
ID1 ðt ¼ 0Þ ¼ ID1 þ iD1 ; ID2 ðt ¼ 0Þ ¼ ID2 þ iD2 ; Yðt ¼ 0Þ ¼ Y ðjÞ þ y; Zðt ¼ 0Þ
¼ Z ðjÞ þ z:
For k\0; t [ 0, the selected fixed point is stable otherwise k [ 0; t [ 0 is

unstable. Our system tends to the selected fixed point exponentially for k\0; t [ 0
otherwise go away from the selected fixed point exponentially. k Is the eigenvalue
his absolute value (jkj) establishes the speed of flow toward or away from the
selected fixed point (Yuri 1995; Jack and Huseyin 1991). The speeds of flow toward or
away from the selected fixed point for RFID system burst switch diodes (D1 and D2)
currents and antenna voltage derivative with respect to time are
dZðtÞ Zðt þ DtÞ ZðtÞ Z ðjÞ þ z ekðt þ DtÞ ½Z ðjÞ þ z ekt
¼ lim ¼ lim
dt Dt!1 Dt Dt!1 Dt
ekDt 1 þ kDt
¼ ! k z ekt
ðjÞ ðjÞ
dID1 ðtÞ ID ðt þ DtÞ ID1 ðtÞ I þ iD1 ekðt þ DtÞ ½ID1 þ iD1 ekt
¼ lim 1 ¼ lim D1
dt Dt!1 Dt Dt!1 Dt
ekDt 1 þ kDt
¼ ! k iD1 ekt
ðjÞ ðjÞ
dID2 ðtÞ ID ðt þ DtÞ ID2 ðtÞ I þ iD2 ekðt þ DtÞ ½ID2 þ iD2 ekt
¼ lim 2 ¼ lim D2
dt Dt!1 Dt Dt!1 Dt
ekDt 1 þ kDt
¼ ! k iD2 ekt
dYðtÞ Yðt þ DtÞ YðtÞ Y ðjÞ þ y ekðt þ DtÞ ½Y ðjÞ þ y ekt
¼ lim ¼ lim
dt Dt!1 Dt Dt!1 Dt
ekDt 1 þ kDt
¼ ! k y ekt
dID1 ðt s1 Þ dID2 ðt s2 Þ
¼ k iD1 ekt eks1 ; ¼ k iD2 ekt eks2 ;
dt dt
dZðt DÞ
¼ k z ekt ekD
dt
First, we take RFID system burst switch variable Z; ID1 ; ID2 ; Y differential equa-
tions and adding to coordinate ½Z; ID1 ; ID2 ; Y arbitrarily small increments of expo-
nential terms ½z; iD1 ; iD2 ; y ekt and retaining the first order terms in z; iD1 ; iD2 ; y.
dZðtÞ
ZðtÞ ¼ Z ðjÞ þ z ekt ) ¼ z k ekt
dt
dZ ½L1 Lcalc ID ½ID2 þ I0 ½ID2 þ I0
¼ ½Y 2 Y
dt Lcalc C1 Cload Vt ½ID1 þ I0
2 lnf½ 1 þ 1 ½ 2 þ 1g Z
RDC Vt ID
LcalcC1 Lcalc C1 I0
ðjÞ ðjÞ
½L1 Lcalc ðI þ iD2 ekt Þ ½ID2 þ iD2 ekt þ I0
z k ekt ¼ ½ðY ðjÞ þ y ekt Þ D2
Lcalc C1 Cload Vt
ðjÞ
½ID2 þ iD2 ekt þ I0
ðjÞ
ðY ðjÞ þ y ekt Þ
ðjÞ ðjÞ ðjÞ
½ID2 þ iD2 ekt þ I0 I þ iD1 ekt I þ iD2 ekt
lnf½ D1 þ 1 ½ D2 þ 1g
RL Cload I0 I0
RDC
ðZ ðjÞ þ z ekt Þ
Lcalc
ðjÞ ðjÞ RDC
½ðID1 þ iD1 ekt Þ þ ID2 þ iD2 ekt
Lcalc C1
ðjÞ
Vt I þ iD1 ekt
ln½ D1 þ 1
Lcalc C1 I0
½L1 Lcalc
z k ekt ¼ ½ðY ðjÞ þ y ekt Þ
Lcalc C1
ðjÞ ðjÞ ðjÞ ðjÞ
ID2 ðID2 þ I0 Þ þ ID2 iD2 ekt þ ðID2 þ I0 Þ iD2 ekt þ i2D2 e2kt

Cload Vt
ðjÞ ðjÞ
½ðID2 þ I0 Þ þ iD2 ekt ½ðID1 þ I0 Þ iD1 ekt
ðjÞ
ðjÞ ðY ðjÞ þ y ekt Þ
kt
½ðID1 þ I0 Þ þ iD1 e ½ðID1 þ I0 Þ iD1 e kt
ðjÞ ðjÞ ðjÞ

lnf½ D1 þ 1 ½ D2 þ 1g
RL Cload I0 I0
RDC
Lcalc
ðjÞ ðjÞ RDC
½ID1 iD1 ekt þ ID2 þ iD2 ekt
Lcalc C1
ðjÞ
Vt I þ iD1 ekt
ln½ D1 þ 0
Lcalc C1 I0
½L1 Lcalc
Lcalc C1
ID2 ðID2 þ I0 Þ þ ID2 iD2 ekt þ ðID2 þ I0 Þ iD2 ekt þ i2D2 e2kt

Cload Vt
ðI þ I0 Þ ðID1 þ I0 Þ ðID2 þ I0 Þ iD1 ekt þ ðID1 þ I0 Þ iD2 ekt iD2 iD1 e2kt
f D2 ðjÞ
g
ðID1 þ I0 Þ2 i2D1 e2kt
ðY ðjÞ þ y ekt Þ
ðjÞ ðjÞ ðjÞ
lnf½ D1 þ 1 ½ D2 þ 1g
RL Cload I0 I0
RDC
Lcalc
ðjÞ ðjÞ RDC
Lcalc C1
ðjÞ
Vt I þ iD1 ekt
ln½ D1 þ 0
Lcalc C1 I0
We consider i2D1 ! e 0; i2D2 ! e 0; iD2 iD1 ! e 0

ID ðID þ I0 Þ þ ID iD2 ekt þ ðID þ I0 ÞiD2 ekt
z k ekt ¼ ½LL1calc
Lcalc ðjÞ kt
C1 ½ðY þ y e Þ
2 2 2
Cload Vt
2
ðjÞ ðjÞ ðjÞ kt ðjÞ kt

ðID þ I0 ÞðID þ I0 ÞðID þ I0 ÞiD1 e þ ðID þ I0 ÞiD2 e
f 2 1 2
ðjÞ
1
g ðY ðjÞ þ y ekt Þ
ðID þ I0 Þ2
ð£Þ ðjÞ ðjÞ
1
ðjÞ
½ID þ iD2 ekt þ I0 I þ iD1 e kt
I þ iD2 ekt
2
RL Cload lnf½ D1 I0 þ 1 ½ D2 I0 þ 1g LRcalc
DC
ðjÞ
ðjÞ ðjÞ I þ iD1 ekt
½ID1 iD1 ekt þ ID2 þ iD2 ekt Lcalc
RDC
C1 Lcalc C1 ln½
Vt D1
I0 þ 1
Calculation No. 1:
ðjÞ ðjÞ
ID1 þ iD1 ekt I þ iD2 ekt
lnf½ þ 1 ½ D2 þ 1g
I0 I0
ðjÞ ðjÞ
ID 1 1 I 1
¼ ln½ð½ þ 1 þ iD1 ekt Þ ð½ D2 þ 1 þ iD2 ekt Þ
I0 I0 I0 I0
ðjÞ ðjÞ ðjÞ
ID 1 I I 1
¼ lnf½ þ 1 ½ D2 þ 1 þ ½ D1 þ 1 iD2 ekt
I0 I0 I0 I0
ðjÞ
ID2 1 1
þ½ þ 1 iD1 ekt þ 2 iD1 iD2 e2kt g
I0 I0 I0
iD1 iD2 0
ðjÞ ðjÞ
ID1 þ iD1 ekt ID 2 þ iD2 ekt
lnf½ þ 1 ½ þ 1g
I0 I0
ID1 I I 1 I 1
¼ lnf½ þ 1 ½ D2 þ 1 þ ½ D1 þ 1 iD2 ekt þ ½ D2 þ 1 iD1 ekt g
I0 I0 I0 I0 I0 I0
We define:
ðjÞ ðjÞ
ðjÞ ðjÞ I 1 I 1
X1 ¼ X1 ðID1 ; ID2 ; iD2 ; iD1 ; kÞ ¼ ½ D1 þ 1 iD2 ekt þ ½ D2 þ 1 iD1 ekt
I0 I0 I0 I0
I þ iD1 ekt I þ iD2 ekt I I
lnf½ D1 þ 1 ½ D2 þ 1g ¼ lnf½ D1 þ 1 ½ D2 þ 1 þ X1 g
I0 I0 I0 I0
ðjÞ ðjÞ
I þ iD1 ekt I þ iD2 ekt
lnf½ D1 þ 1 ½ D2 þ 1g
I0 I0
ðjÞ ðjÞ
I I X1
¼ lnf½ D1 þ 1 ½ D2 þ 1 ð1 þ ðjÞ ðjÞ
Þg
I0 I0 ID ID
½ I0 þ 1 ½ I0 þ 1
1 2
The above is assuming
ðjÞ ðjÞ
ID1 I
½ þ 1 ½ D2 þ 1 [ 0
I0 I0
ðjÞ ðjÞ
ID1 þ iD1 ekt I þ iD2 ekt
lnf½ þ 1 ½ D2 þ 1g
I0 I0
ðjÞ ðjÞ
I I X1
¼ lnf½ D1 þ 1 ½ D2 þ 1g þ lnf1 þ ðjÞ ðjÞ
g
I0 I0 ID ID
½ I0 þ 1 ½ I0 þ 1
1 2
ðjÞ ðjÞ
ID I
X1 ½ I0 þ 1 I10 iD2 ekt þ ½ ID02 þ 1 I10 iD1 ekt
1
ðjÞ ðjÞ
¼ ðjÞ ðjÞ
I I I I
½ ID01 þ 1 ½ ID02 þ 1 ½ ID01 þ 1 ½ ID02 þ 1
iD2 iD1 1 kt
¼f ðjÞ
þ ðjÞ
g e
ID ID I0
½ I02 þ 1 ½ I01 þ 1
x2 x3 x4 X1
xn
lnð1 þ xÞ ¼ x þ þ... ¼ ð1Þn þ 1 . . . ) lnð1 þ xÞ x
2 3 4 n¼1
n
X1 X1
lnf1 þ ðjÞ ðjÞ
g ðjÞ ðjÞ
ID ID ID I
½ I01 þ 1 ½ I02 þ 1 ½ I01 þ 1 ½ ID02 þ 1
iD2 iD1 1 kt
¼f ðjÞ
þ ðjÞ
g e
ID ID I0
½ I02 þ 1 ½ I01 þ 1
ðjÞ ðjÞ
I þ iD1 ekt I þ iD2 ekt
lnf½ D1 þ 1 ½ D2 þ 1g
I0 I0
ðjÞ ðjÞ
I I iD iD 1
¼ lnf½ D1 þ 1 ½ D2 þ 1g þ f ðjÞ 2 þ ðjÞ 1 g ekt
I0 I0 I I I0
½ ID02 þ 1 ½ ID01 þ 1
Calculation No. 2:
ðjÞ ðjÞ ðjÞ

lnf½ D1 þ 1 ½ D2 þ 1g
RL Cload I0 I0
ðjÞ ðjÞ ðjÞ
½ID2 þ iD2 ekt þ I0 I I iD iD 1
¼ ðlnf½ D1 þ 1 ½ D2 þ 1g þ f ðjÞ 2 þ ðjÞ 1 g ekt Þ
RL Cload I0 I0 I I I0
½ ID02 þ 1 ½ ID01 þ 1
ðjÞ ðjÞ ðjÞ

lnf½ D1 þ 1 ½ D2 þ 1g
RL Cload I0 I0
ðjÞ ðjÞ ðjÞ
ID2 þ I0 iD ekt I I iD iD 1
¼f þ 2 g ðlnf½ D1 þ 1 ½ D2 þ 1g þ f ðjÞ 2 þ ðjÞ 1 g ekt Þ
RL Cload RL Cload I0 I0 ID ID I0
½ I02 þ 1 ½ I01 þ 1
ðID2 þ I0 Þ I I ðI þ I0 Þ iD iD 1
¼ lnf½ D1 þ 1 ½ D2 þ 1g þ D2 f ðjÞ 2 þ ðjÞ 1 g ekt
RL Cload I0 I0 RL Cload ID ID I 0
½ I02 þ 1 ½ I01 þ 1
ðjÞ ðjÞ
iD2 ekt I I e2kt iD iD2 iD iD1
þ lnf½ D1 þ 1 ½ D2 þ 1g þ f ðjÞ2 þ ðjÞ2 g
RL Cload I0 I0 RL Cload I0 I I
½ ID02 þ 1 ½ ID01 þ 1
e2kt iD iD 2 iD iD1
iD2 iD2 0; iD2 iD1 0 ) f ðjÞ2 þ ðjÞ2 g!e
RL Cload I0 ID I
½ I02 þ 1 ½ ID01 þ 1
ðjÞ ðjÞ ðjÞ

lnf½ D1 þ 1 ½ D2 þ 1g
RL Cload I0 I0
ðID2 þ I0 Þ I I ðI þ I0 Þ iD iD 1
¼ lnf½ D1 þ 1 ½ D2 þ 1g þ D2 f ðjÞ 2 þ ðjÞ 1 g ekt
RL Cload I0 I0 RL Cload ID ID I 0
½ I02 þ 1 ½ I01 þ 1
ðjÞ ðjÞ
iD2 ekt I I
þ lnf½ D1 þ 1 ½ D2 þ 1g
RL Cload I0 I0
Calculation No. 3:
ðjÞ ðjÞ ðjÞ
ID1 þ iD1 ekt I iD ekt I iD1 ekt
ln½ þ 1 ¼ ln½ð D1 þ 1Þ þ 1 ¼ ln½ð D1 þ 1Þ f1 þ ðjÞ
g
I0 I0 I0 I0 I
I0 ð ID01 þ 1Þ
ðjÞ ðjÞ
ID1 þ iD1 ekt I iD1 ekt
ln½ þ 1 ¼ lnð D1 þ 1Þ þ lnf1 þ ðjÞ
g
I0 I0 I
I0 ð ID01 þ 1Þ
x2 x3 x4 X1
xn
lnð1 þ xÞ ¼ x þ þ ¼ ð1Þn þ 1 ) lnð1 þ xÞ x
2 3 4 n¼1
n
ðjÞ ðjÞ
ID1 þ iD1 ekt I iD1 ekt
ln½ þ 1 ¼ lnð D1 þ 1Þ þ lnf1 þ ðjÞ
g
I0 I0 I
I0 ð ID01 þ 1Þ
ðjÞ
ID 1 iD1 ekt
¼ lnð þ 1Þ þ ðjÞ
I0 I
I0 ð ID01 þ 1Þ
Integrating last results in the next expression:

ID ðID þ I0 Þ þ ID iD2 ekt þ ðID þ I0 ÞiD2 ekt
z k ekt ¼ ½LL1calc
Lcalc ðjÞ kt
C1 ½ðY þ y e Þ
2 2 2
Cload Vt
2
ðjÞ ðjÞ ðjÞ kt ðjÞ kt

ðID þ I0 ÞðID þ I0 ÞðID þ I0 ÞiD1 e þ ðID þ I0 ÞiD2 e
f 2 1 2
ðjÞ
1
ðID þ I0 Þ2
ð£Þ ðjÞ ðjÞ
1
ðjÞ
½ID þ iD2 ekt þ I0 I þ iD1 e kt
I þ iD2 ekt
2
RL Cload lnf½ D1 I0 þ 1 ½ D2 I0 þ 1g LRcalc
DC
ðjÞ
ðjÞ ðjÞ I þ iD1 ekt
½ID1 iD1 ekt þ ID2 þ iD2 ekt Lcalc
RDC
C1 Lcalc C1 ln½
Vt D1
I0 þ 1
½L1 Lcalc
Lcalc C1
ID2 ðID2 þ I0 Þ þ ID2 iD2 ekt þ ðID2 þ I0 Þ iD2 ekt

Cload Vt
ðID2 þ I0 Þ ðID1 þ I0 Þ ðID2 þ I0 Þ iD1 ekt þ ðID1 þ I0 Þ iD2 ekt
f ðjÞ
ðID1 þ I0 Þ2
ðjÞ ðjÞ ðjÞ
ðID2 þ I0 Þ I I
f lnf½ D1 þ 1 ½ D2 þ 1g
RL Cload I0 I0
ðjÞ
ðID2 þ I0 Þ iD iD 1
þ f ðjÞ 2 þ ðjÞ 1 g ekt
RL Cload ID ID I 0
½ I02 þ 1 ½ I01 þ 1
ðjÞ ðjÞ
iD2 ekt I I RDC
þ lnð½ D1 þ 1 ½ D2 þ 1Þg ðZ ðjÞ þ z ekt Þ
ðjÞ ðjÞ RDC
Lcalc C1
ðjÞ
Vt I iD1 ekt
flnð D1 þ 1Þ þ g
Lcalc C1 I0 I
ðjÞ
I0 ð ID01 þ 1Þ
The condition of our system fixed points:
dZ
j ðjÞ ðjÞ ¼0
dt @ID1 ;ID2 ;Y ðjÞ ;Z ðjÞ
ðjÞ ðjÞ ðjÞ
½L1 Lcalc I ðID2 þ I0 Þ ðI þ I0 Þ
½Y ðjÞ D2 f DðjÞ2 g Y ðjÞ
Lcalc C1 Cload Vt ðI þ I0 Þ D1
ðjÞ ðjÞ ðjÞ
ðID2 þ I0 Þ ID1 ID2
lnf½ þ 1 ½ þ 1g
RL Cload I0 I0
ðjÞ
RDC ðjÞ ðjÞ ðjÞ RDC Vt I
Z ½ID1 þ ID2 lnð D1 þ 1Þ ¼ 0
Lcalc Lcalc C1 Lcalc C1 I0
ðjÞ ðjÞ
½L1 Lcalc I iD2 ekt þ ðID2 þ I0 Þ iD2 ekt
z k ekt ¼ ½y ekt D2
Lcalc C1 Cload Vt
ðjÞ
ðID2 þ I0 Þ
ðjÞ
y ekt
ðID1 þ I0 Þ
ðjÞ ðjÞ
ðID2 þ I0 Þ ðID2 þ I0 Þ
þ ðjÞ
i ekt Y ðjÞ þ ðjÞ
2 D1
iD1 ekt y ekt
ðID1 þ I0 Þ ðID1 þ I0 Þ2
iD2 ekt Y ðjÞ iD2 ekt y ekt
ðjÞ
ðjÞ
ðjÞ
f ðjÞ 2 þ ðjÞ 1 g ekt
RL Cload I I I0
½ ID02 þ 1 ½ ID01 þ 1
ðjÞ ðjÞ
iD2 ekt I I
lnð½ D1 þ 1 ½ D2 þ 1Þ
RL Cload I0 I0
RDC RDC
z ekt ½iD1 ekt þ iD2 ekt
Lcalc Lcalc C1
Vt iD1 ekt
f g
Lcalc C1 I
ðjÞ
I0 ð ID01 þ 1Þ
ðjÞ
ðID2 þ I0 Þ
iD1 y 0 ) ðjÞ
iD1 ekt y ekt ! e;
ðID1 þ I0 Þ2
iD2 ekt y ekt
iD2 y 0 ) ðjÞ
!e
ðID1 þ I0 Þ
ðjÞ ðjÞ
½L1 Lcalc I iD2 ekt þ ðID2 þ I0 Þ iD2 ekt
Lcalc C1 Cload Vt
ðjÞ
ðID2 þ I0 Þ
ðjÞ
y ekt
ðID1 þ I0 Þ
ðjÞ
ðID2 þ I0 Þ iD2 ekt Y ðjÞ
þ ðjÞ
iD1 ekt Y ðjÞ ðjÞ
ðID1 þ I0 Þ2 ðID1 þ I0 Þ
ðjÞ
ðID2 þ I0 Þ iD2 i D1 1
f þ g ekt
RL Cload ðjÞ
ID
ðjÞ
ID I0
½ I02 þ 1 ½ I01 þ 1
ðjÞ ðjÞ
iD2 ekt I I RDC
lnð½ D1 þ 1 ½ D2 þ 1Þ z ekt
RDC Vt iD1 ekt
½iD1 ekt þ iD2 ekt f g
Lcalc C1 Lcalc C1 I
ðjÞ
I0 ð ID01 þ 1Þ
dID2 ID ½ID2 þ I0 ½ID2 þ I0 ½ID þ I0 ID ID

¼ 2 Y 2 lnf½ 1 þ 1 ½ 2 þ 1g
dt Cload Vt ½ID1 þ I0 RL Cload I0 I0
ðjÞ ðjÞ
½ID2 þ iD2 ekt ½ID2 þ iD2 ekt þ I0
k iD2 ekt ¼
Cload Vt
ðjÞ
ðjÞ
½Y ðjÞ þ y ekt
ðjÞ ðjÞ ðjÞ
½ID2 þ iD2 ekt þ I0 ðI þ iD1 ekt Þ ðI þ iD2 ekt Þ
lnf½ D1 þ 1 ½ D2 þ 1g
RL Cload I0 I0
ðjÞ ðjÞ
½ID2 þ iD2 ekt ½ðID2 þ I0 Þ þ iD2 ekt
k iD2 ekt ¼
Cload Vt
ðjÞ
½ðID2 þ I0 Þ þ iD2 ekt
ðjÞ
½Y ðjÞ þ y ekt
½ðID1 þ I0 Þ þ iD1 ekt
ðjÞ ðjÞ ðjÞ
½ðID2 þ I0 Þ þ iD2 ekt ðI þ iD1 ekt Þ ðI þ iD2 ekt Þ
lnf½ D1 þ 1 ½ D2 þ 1g
RL Cload I0 I0

ID2 ðID2 þ I0 Þ þ ID2 iD2 ekt þ ðID2 þ I0 Þ iD2 ekt þ iD2 iD2 e2kt
k iD2 ekt ¼
Cload Vt
ðjÞ ðjÞ
ðjÞ
ðjÞ ½Y ðjÞ þ y ekt
ðjÞ ðjÞ ðjÞ
lnf½ D1 þ 1 ½ D2 þ 1g
RL Cload I0 I0

ID2 ðID2 þ I0 Þ þ ID2 iD2 ekt þ ðID2 þ I0 Þ iD2 ekt þ iD2 iD2 e2kt
k iD2 ekt ¼
Cload Vt
ðID2 þ I0 Þ ðID1 þ I0 Þ ðID2 þ I0 Þ iD1 ekt þ ðID1 þ I0 Þ iD2 ekt iD2 iD1 e2kt
f ðjÞ 2
g
ðID1 þ I0 Þ i2D1 e2kt
½Y ðjÞ þ y ekt
ðjÞ ðjÞ ðjÞ
lnf½ D1 þ 1 ½ D2 þ 1g
RL Cload I0 I0
iD2 iD2 0; iD2 iD1 0; i2D1 0


k iD2 ekt ¼
Cload Vt
ðI þ I0 Þ ðID1 þ I0 Þ ðID2 þ I0 Þ iD1 ekt þ ðID1 þ I0 Þ iD2 ekt
f D2 ðjÞ
g
ðID1 þ I0 Þ2
½Y ðjÞ þ y ekt
ðjÞ ðjÞ ðjÞ
lnf½ D1 þ 1 ½ D2 þ 1g
RL Cload I0 I0

k iD2 ekt ¼
Cload Vt
Y ðjÞ ðID2 þ I0 Þ ðID1 þ I0 Þ Y ðjÞ ðID2 þ I0 Þ iD1 ekt þ Y ðjÞ ðID1 þ I0 Þ iD2 ekt
f ðjÞ
g
ðID1 þ I0 Þ2
ðID2 þ I0 Þ ðID1 þ I0 Þ y ekt ðID2 þ I0 Þ iD1 y e2kt þ ðID1 þ I0 Þ iD2 y e2kt
f ðjÞ
g
ðID1 þ I0 Þ2
ðjÞ ðjÞ ðjÞ
lnf½ D1 þ 1 ½ D2 þ 1g
RL Cload I0 I0
iD1 y 0; iD2 y 0
k iD2 ekt ¼
Cload Vt
f ðjÞ
g
ðID1 þ I0 Þ2
ðjÞ ðjÞ
ðID2 þ I0 Þ ðID1 þ I0 Þ y ekt
f ðjÞ
g
ðID1 þ I0 Þ2
ðjÞ ðjÞ ðjÞ
lnf½ D1 þ 1 ½ D2 þ 1g
RL Cload I0 I0
We have already approved in calculation No. 1

I þ iD1 ekt I þ iD2 ekt I I
lnf½ D1 þ 1 ½ D2 þ 1g ¼ lnf½ D1 þ 1 ½ D2 þ 1g
I0 I0 I0 I0
iD2 iD 1
þ f ðjÞ þ ðjÞ 1 g ekt
ID ID I0
½ I02 þ 1 ½ I01 þ 1

k iD2 ekt ¼
Cload Vt
f ðjÞ 2
g
ðID1 þ I0 Þ
ðjÞ ðjÞ
f ðjÞ
g
ðID1 þ I0 Þ2
ðjÞ ðjÞ ðjÞ
ðID2 þ I0 Þ iD ekt I I
f þ 2 g flnf½ D1 þ 1 ½ D2 þ 1g
RL Cload RL Cload I0 I0
iD2 iD 1 1 kt
þ f ðjÞ þ ðjÞ g e g
I I I0
½ ID02 þ 1 ½ ID01 þ 1

ID2 ðID2 þ I0 Þ ID2 iD2 ekt þ ðID2 þ I0 Þ iD2 ekt
k iD2 ekt ¼
Cload Vt Cload Vt
ðjÞ ðjÞ
Y ðjÞ ðID2 þ I0 Þ Y ðjÞ ðID2 þ I0 Þ iD1 ekt Y ðjÞ iD2 ekt
f ðjÞ
ðjÞ 2
þ ðjÞ
g
ðID1 þ I0 Þ ðID1 þ I0 Þ ðID1 þ I0 Þ
ðjÞ ðjÞ
ðI þ I0 Þ ðID1 þ I0 Þ y ekt
f D2 ðjÞ
g
ðID1 þ I0 Þ2
ðjÞ ðjÞ ðjÞ
ðI þ I0 Þ iD ekt I I
f D2 þ 2 g flnf½ D1 þ 1 ½ D2 þ 1g
RL Cload RL Cload I0 I0
iD 2 iD 1 1 kt
þ f ðjÞ þ ðjÞ g e g
I I I0
½ ID02 þ 1 ½ ID01 þ 1

k iD2 ekt ¼
Cload Vt Cload Vt
ðjÞ ðjÞ
f ðjÞ
ðjÞ 2
þ ðjÞ
g
ðID1 þ I0 Þ ðID1 þ I0 Þ ðID1 þ I0 Þ
ðjÞ ðjÞ
ðI þ I0 Þ ðID1 þ I0 Þ y ekt
f D2 ðjÞ
g
ðID1 þ I0 Þ2
ðjÞ ðjÞ ðjÞ
ðID2 þ I0 Þ ID 1 ID 2
lnf½ þ 1 ½ þ 1g
RL Cload I0 I0
ðjÞ
RL Cload ID ID I 0
½ I02 þ 1 ½ I01 þ 1
ðjÞ ðjÞ
iD2 ekt I I
lnf½ D1 þ 1 ½ D2 þ 1g
RL Cload I0 I0
1 iD2 iD2 iD1 iD2 1
f ðjÞ þ ðjÞ g e2kt
RL Cload ID ID I0
½ I0 þ 1 ½ I0 þ 1
2 1
iD2 iD2 0; iD1 iD2 0

k iD2 ekt ¼
Cload Vt Cload Vt
ðjÞ ðjÞ
f ðjÞ
ðjÞ
þ ðjÞ
g
ðID1 þ I0 Þ ðID1 þ I0 Þ2 ðID1 þ I0 Þ
ðjÞ ðjÞ
f ðjÞ
g
ðID1 þ I0 Þ2
ðjÞ ðjÞ ðjÞ
ðID2 þ I0 Þ I I
lnf½ D1 þ 1 ½ D2 þ 1g
RL Cload I0 I0
ðjÞ
RL Cload I I I0
½ ID02 þ 1 ½ ID01 þ 1
ðjÞ ðjÞ
iD2 ekt I I
lnf½ D1 þ 1 ½ D2 þ 1g
RL Cload I0 I0
dID2
At fixed point dt ¼0
ðjÞ ðjÞ ðjÞ

dID2 I ðID2 þ I0 Þ Y ðjÞ ðID2 þ I0 Þ
¼ 0 ) D2
dt Cload Vt ðjÞ
ðI þ I0 Þ D1
ðjÞ ðjÞ ðjÞ
ðID2 þ I0 Þ ID1 ID2
lnf½ þ 1 ½ þ 1g
RL Cload I0 I0
ðjÞ ðjÞ
kt
I iD2 ekt þ ðID2 þ I0 Þ iD2 ekt
k iD2 e ¼ D2
Cload Vt
ðjÞ
Y ðjÞ ðID2 þ I0 Þ iD1 ekt Y ðjÞ iD2 ekt
f ðjÞ 2
þ ðjÞ
g
ðjÞ ðjÞ
f ðjÞ
g
ðID1 þ I0 Þ2
ðjÞ
ðI þ I0 Þ iD iD 1
D2 f ðjÞ 2 þ ðjÞ 1 g ekt
RL Cload ID ID I 0
½ I02 þ 1 ½ I01 þ 1
ðjÞ ðjÞ
iD ekt I I
2 lnf½ D1 þ 1 ½ D2 þ 1g
RL Cload I0 I0
Remark: it is reader exercise to build the system Jacobian matrix and analyze the
dynamic and stability of the system based on eigenvalues investigation.
ðjÞ ðjÞ
We define ID1 ðt s1 Þ ¼ ID1 þ iD1 ekðts1 Þ ; ID2 ðt s2 Þ ¼ ID2 þ iD2 ekðts2 Þ and
Zðt DÞ ¼ Z ðjÞ þ z ekðtDÞ . Then we get three delayed differential equations with
respect to coordinates ½Z; ID1 ; ID2 ; Y arbitrarily small increments of exponential
dZðtÞ dID1 ðtÞ dID2 ðtÞ
½z; iD1 ; iD2 ; y ekt . We consider no delay effects on dt ; dt ; dt ; Y
ðjÞ
¼ 0.
ðjÞ ðjÞ ðjÞ

½L1 Lcalc I iD2 ekt þ ðID2 þ I0 Þ iD2 ekt ðID2 þ I0 Þ
z k ekt ¼ ½y ekt D2 ðjÞ y ekt
Lcalc C1 Cload Vt ðI þ I0 Þ D1
ðjÞ ðjÞ
ðID2 þ I0 Þ iD2 ekt Y ðjÞ ðI þ I0 Þ iD iD 1
þ iD1 ekt Y ðjÞ þ D2 f ðjÞ 2 þ ðjÞ 1 g ekt
ðjÞ
ðID1 þ I0 Þ2
ðjÞ
ðID1 þ I0 Þ RL Cload ID ID I 0
½ I02 þ 1 ½ I01 þ 1
ðiÞ
ðjÞ ðjÞ
iD2 ekt I I RDC RDC
lnð½ D1 þ 1 ½ D2 þ 1Þ z ekt ½iD1 ekt þ iD2 ekt
RL Cload I0 I0 Lcalc Lcalc C1
Vt iD1 ekt
f g
Lcalc C1 I
ðjÞ
I0 ð ID01 þ 1Þ
ðiiÞ k iD1 ekt ¼ Y ðjÞ þ y ekt

ðjÞ ðjÞ
ID2 iD2 ekt þ ðID2 þ I0 Þ iD2 ekt
k iD2 ekt ¼
Cload Vt
ðjÞ
f ðjÞ
þ ðjÞ
g
ðjÞ ðjÞ
ðiiiÞ f ðjÞ
g
ðID1 þ I0 Þ2
ðjÞ
RL Cload ID ID I 0
½ I02 þ 1 ½ I01 þ 1
ðjÞ ðjÞ
iD2 ekt I I
lnf½ D1 þ 1 ½ D2 þ 1g
RL Cload I0 I0
iD1 ekt ! iD1 ekt eks1 ; iD2 ekt ! iD2 ekt eks2 ;
z ekt ! z ekt ekD ; Y ðjÞ ¼ 0
Remark: left side of below equation doesn’t affect by delay parameter.

ðjÞ ðjÞ
½L1 Lcalc I iD2 ekt eks2 þ ðID2 þ I0 Þ iD2 ekt eks2
Lcalc C1 Cload Vt
ðjÞ
ðID2 þ I0 Þ
ðjÞ
y ekt
ðID1 þ I0 Þ
ðjÞ
ðID2 þ I0 Þ iD2 eks2 iD eks1 1 kt
þ f ðjÞ þ 1ðjÞ g e
RL Cload I I I0
ðiÞ ½ ID02 þ 1 ½ ID01 þ 1
ðjÞ ðjÞ
iD2 ekt eks2 I I
lnð½ D1 þ 1 ½ D2 þ 1Þ
RL Cload I0 I0
RDC kt kD
ze e ½iD1 e eks1 þ iD2 ekt eks2
kt
Lcalc
RDC Vt iD eks1 ekt
f 1 g
Lcalc C1 Lcalc C1 I
ðjÞ
I0 ð ID01 þ 1Þ
Divide above equations two sides by ekt :

ðjÞ ðjÞ
½L1 Lcalc f2 ID2 þ I0 g iD2 eks2 ðID2 þ I0 Þ
zk¼ ½y ðjÞ y
Lcalc C1 Cload Vt ðID1 þ I0 Þ
ðjÞ
ðID2 þ I0 Þ iD2 eks2 iD eks1 1
þ f ðjÞ þ 1ðjÞ g
RL Cload I I I0
½ ID02 þ 1 ½ ID01 þ 1
ðjÞ ðjÞ
iD2 eks2 I I RDC
lnð½ D1 þ 1 ½ D2 þ 1Þ z ekD
RDC Vt iD eks1
½iD1 eks1 þ iD2 eks2 f 1 ðjÞ g
Lcalc C1 Lcalc C1 I
I0 ð ID01 þ 1Þ
ðjÞ
½L1 Lcalc ½L1 Lcalc f2 ID2 þ I0 g iD2 eks2
zk y
Lcalc C1 Lcalc C1 Cload Vt
ðjÞ
½L1 Lcalc ðID2 þ I0 Þ
y
Lcalc C1 ðI ðjÞ þ I0 Þ
D1
ðjÞ
½L1 Lcalc ðID2 þ I0 Þ iD2 eks2 iD eks1 1
þ f ðjÞ þ 1ðjÞ g
Lcalc C1 RL Cload I I I0
½ ID02 þ 1 ½ ID01 þ 1
ðjÞ ðjÞ
½L1 Lcalc iD2 eks2 I I
lnð½ D1 þ 1 ½ D2 þ 1Þ
Lcalc C1 RL Cload I0 I0
RDC kD ks1 ks2 RDC
ze ½iD1 e þ iD 2 e
Lcalc Lcalc C1
Vt iD eks1
f 1 ðjÞ g¼0
Lcalc C1 I
I0 ð ID01 þ 1Þ
ðjÞ
½L1 Lcalc ½L1 Lcalc f2 ID2 þ I0 g iD2 eks2
zk y
Lcalc C1 Lcalc C1 Cload Vt
ðjÞ
½L1 Lcalc ðID2 þ I0 Þ
y
Lcalc C1 ðI ðjÞ þ I0 Þ
D1
ðjÞ
½L1 Lcalc ðID2 þ I0 Þ iD2 eks2
þ
Lcalc C1 RL Cload I0 IDðjÞ2
½ I0 þ 1
ðjÞ
½L1 Lcalc ðID2 þ I0 Þ iD1 eks1
þ
Lcalc C1 RL Cload I0 IDðjÞ1
½ I0 þ 1
ðjÞ ðjÞ
½L1 Lcalc iD2 eks2 I I
lnð½ D1 þ 1 ½ D2 þ 1Þ
RDC kD ks1 ks2 RDC
ze ½iD1 e þ iD2 e
Lcalc Lcalc C1
Vt iD eks1
f 1 ðjÞ g¼0
Lcalc C1 ID
I0 ð I0 þ 1Þ
1
ðjÞ
RDC kD ½L1 Lcalc ðID2 þ I0 Þ eks1
f e kg z þ f ðjÞ
Lcalc Lcalc C1 RL Cload I0 ID1
½ I0 þ 1
Vt eks1 RDC
f gþ eks1 g iD1
Lcalc C1 ðjÞ
ID Lcalc C1
I0 ð I01 þ 1Þ
ðjÞ
½L1 Lcalc ðID2 þ I0 Þ eks2
þf ðjÞ
Lcalc C1 RL Cload I0 ID2
½ I0 þ 1
ðjÞ
½L1 Lcalc f2 ID2 þ I0 g eks2

Lcalc C1 Cload Vt
ðjÞ ðjÞ
½L1 Lcalc eks2 I I
lnð½ D1 þ 1 ½ D2 þ 1Þ
ðjÞ
RDC ½L1 Lcalc ðI þ I0 Þ
eks2 g iD2 f1 þ DðjÞ2 gy¼0
Lcalc C1 Lcalc C1 ðID1 þ I0 Þ
ðjÞ
RDC kD ½L1 Lcalc ðID2 þ I0 Þ 1
f e kg z þ f
Lcalc Lcalc C1 RL Cload I0 ðjÞ
ID
½ I01 þ 1
Vt 1 RDC
f gþ g eks1 iD1
Lcalc C1 ðjÞ
ID Lcalc C1
I0 ð I01 þ 1Þ
ðjÞ ðjÞ
½L1 Lcalc ðID2 þ I0 Þ 1 ½L1 Lcalc f2 ID2 þ I0 g
þf
Lcalc C1 RL Cload I0 I
ðjÞ
Lcalc C1 Cload Vt
½ ID02 þ 1
ðjÞ ðjÞ
½L1 Lcalc 1 I I RDC
lnð½ D1 þ 1 ½ D2 þ 1Þ g eks2 iD2
Lcalc C1 RL Cload I0 I0 Lcalc C1
ðjÞ
½L1 Lcalc ðI þ I0 Þ
f1 þ DðjÞ2 gy¼0
Lcalc C1 ðID1 þ I0 Þ
We define for simplicity the following global parameters:
ðjÞ
½L1 Lcalc ðID2 þ I0 Þ 1 Vt 1 RDC
!1 ¼ f gþ
Lcalc C1 RL Cload I0 ID
ðjÞ
Lcalc C1 ðjÞ
ID Lcalc C1
½ I01 þ 1 I0 ð I01 þ 1Þ
ðjÞ ðjÞ
½L1 Lcalc ðID2 þ I0 Þ 1 ½L1 Lcalc f2 ID2 þ I0 g
!2 ¼
Lcalc C1 RL Cload I0 I
ðjÞ
Lcalc C1 Cload Vt
½ ID02 þ 1
ðjÞ ðjÞ
½L1 Lcalc 1 I I RDC
lnð½ D1 þ 1 ½ D2 þ 1Þ
Lcalc C1 RL Cload I0 I0 Lcalc C1
ðjÞ
½L1 Lcalc ðI þ I0 Þ
!3 ¼ f1 þ DðjÞ2 g
Lcalc C1 ðID1 þ I0 Þ
RDC kD
f e kg z þ !1 eks1 iD1 þ !2 eks2 iD2 !3 y ¼ 0
Lcalc
Remark left side of below equation doesn’t affect by delay parameter.
k iD1 ekt ¼ Y ðjÞ þ y ekt jY ðjÞ ¼0 ) k iD1 ekt ¼ y ekt ) k iD1 þ y ¼ 0

ðiiÞ ðjÞ ðjÞ
ID2 iD2 ekt þ ðID2 þ I0 Þ iD2 ekt
k iD2 ekt ¼
Cload Vt
ðjÞ
f ðjÞ
þ ðjÞ
g
ðjÞ ðjÞ
f ðjÞ
g
ðID1 þ I0 Þ2
ðiiiÞ ðjÞ
RL Cload ID ID I 0
½ I02 þ 1 ½ I01 þ 1
ðjÞ ðjÞ
iD2 ekt I I
lnf½ D1 þ 1 ½ D2 þ 1g
RL Cload I0 I0
Remark left side of below equation doesn’t affect by delay parameter.
ðjÞ ðjÞ
ID2 iD2 ekt eks2 þ ðID2 þ I0 Þ iD2 ekt eks2
k iD2 ekt ¼
Cload Vt
ðjÞ ðjÞ
f ðjÞ
g
ðID1 þ I0 Þ2
ðjÞ
ðID2 þ I0 Þ iD2 eks2 iD eks1 1 kt
f ðjÞ þ 1ðjÞ g e
RL Cload ID ID I0
½ I0 þ 1
2
½ I0 þ 1
1
ðjÞ ðjÞ
iD2 ekt eks2 I I
lnf½ D1 þ 1 ½ D2 þ 1g
RL Cload I0 I0
kt
We divide above two sides by e term.
ðjÞ ðjÞ ðjÞ

I iD2 eks2 þ ðID2 þ I0 Þ iD2 eks2 ðID2 þ I0 Þ
k iD2 ¼ D2 ðjÞ y
Cload Vt ðID1 þ I0 Þ
ðjÞ
ðID2 þ I0 Þ iD2 eks2 iD eks1 1
f ðjÞ þ 1ðjÞ g
RL Cload ID ID I0
½ I0 þ 1
2
½ I0 þ 1
1
ðjÞ ðjÞ
iD2 eks2 I I
lnf½ D1 þ 1 ½ D2 þ 1g
RL Cload I0 I0
ðjÞ ðjÞ ðjÞ
½2 ID2 þ I0 iD2 eks2 ðID2 þ I0 Þ ðID2 þ I0 Þ iD2 eks2
k iD2 ¼ ðjÞ y
Cload Vt ðID1 þ I0 Þ RL Cload I0 IDðjÞ2
½ þ 1 I0
ðjÞ ðjÞ ðjÞ
ðID2þ I0 Þ iD1 e iD e ks1 ks2 ID 1 ID 2
ðjÞ 2 lnf½ þ 1 ½ þ 1g
RL Cload I0 I RL Cload I0 I0
½ ID01 þ 1
ðjÞ
ðID2 þ I0 Þ 1
eks1 iD1 k iD2
RL Cload I0 ðjÞ
ID
½ I0 þ 11
ðjÞ ðjÞ
½2 ID2 þ I0 ðID2 þ I0 Þ 1
ð þ
Cload Vt RL Cload I0 ðjÞ
ID
½ I02 þ 1
ðjÞ ðjÞ ðjÞ
1 I I ðI þ I0 Þ
þ lnf½ D1 þ 1 ½ D2 þ 1gÞ iD2 eks2 DðjÞ2 y¼0
RL Cload I0 I0 ðID1 þ I0 Þ
We define the following global parameters for simplicity.
ðjÞ
ðID2 þ I0 Þ 1
!4 ¼ ;
RL Cload I0 ðjÞ
ID
½ I0 þ 1
1

½2 þ I0
ID 2 ðID2 þ I0 Þ 1 1 I I
!5 ¼ þ þ lnf½ D1 þ 1 ½ D2 þ 1g
Cload Vt RL Cload I0 I
ðjÞ
RL Cload I0 I0
½ ID02 þ 1
ðjÞ
ðID2 þ I0 Þ
!6 ¼ ðjÞ
ðID1 þ I0 Þ
!4 eks1 iD1 k iD2 !5 iD2 eks2 !6 y ¼ 0
ðjÞ ðjÞ
!k ¼ !k ðZ ðjÞ ; ID1 ; ID2 ; Y ðjÞ ; L1 ; Lcalc ; RL ; Cload ; RDC ; I0 ; VÞ 8 k ¼ 1; 2; 3; 4; 5; 6:
ðjÞ ðjÞ
In the equilibrium fixed points: Z ðjÞ ; ID1 ; ID2 ; Y ðjÞ ¼ 0
The small increments Jacobian of our RFID burst switch system is as follows:
RDC kD
f e kg z þ !1 eks1 iD1
Lcalc
þ !2 eks2 iD2 !3 y ¼ 0
k iD1 þ y ¼ 0
!4 eks1 iD1 ½k þ !5 eks2 iD2 !6 y ¼ 0
dID1 dYðtÞ d 2 ID1
¼Y ) ¼
dt dt dt2
d 2 ID1 dYðtÞ
We consider dt2 ! e then dt ¼0
dYðtÞ
YðtÞ ¼ Y ðjÞ þ y ekt ) ¼ y k ekt ;
dt
dYðtÞ
¼ 0 ) y k ekt ¼ 0 ) jekt 6¼0 y k ¼ 0 ) y k ¼ 0
dt 0 1
0 1 z
N11 . . . N14 B C
B . B iD1 C
B . .. .. C
CB C
@ . . . A B C
B ¼ 0;
C
@ iD2 A
N41 N44
y
RDC kD
N11 ¼ e k; N12 ¼ !1 eks1 ; N13 ¼ !2 eks2
Lcalc
N14 ¼ !3 ; N21 ¼ 0; N22 ¼ k; N23 ¼ 0; N24 ¼ 1;
N31 ¼ 0; N32 ¼ !4 eks1 ; N33 ¼ k !5 eks2
N34 ¼ !6 ; N41 ¼ 0; N42 ¼ 0; N43 ¼ 0; N44 ¼ k
0 1
N11 . . . N14
B . .. .. C
AkI ¼B @ .. .
C
N31 N34
RDC kD RDC kðD þ s2 Þ
Dðk; s1 ; s2 ; DÞ ¼ k4 þ k3 ½ e þ !5 eks2 þ k2 !5 e
Lcalc Lcalc
We have three sub cases: (I) s2 ¼ s; D ¼ 0(II) s2 ¼ 0; D [ 0(III) s2 ¼ D ¼ sD
RDC RDC
ðIÞ Dðk; s2 ¼ s; D ¼ 0Þ ¼ k4 þ k3 þ ½k3 !5 þ k2 !5 eks
Lcalc Lcalc
RDC RDC
ðIIÞ Dðk; s2 ¼ 0; D [ 0Þ ¼ k4 þ k3 !5 þ ½k3 þ k2 ! 5 ekD
Lcalc Lcalc
RDC
Dðk; s2 ¼ sD ; D ¼ sD Þ ¼ k4 þ k3 ½ þ !5 eksD
Lcalc
ðIIIÞ
RDC ksD ksD
þ k2 ! 5 e e
Lcalc
Under Taylor series approximation: eksD 1 k sD . The Maclaurin series is
a Taylor series expansion of a eksD function about zero (0). We get the following
general characteristic equation D(k, sD) under Taylor series approximation:
eksD 1 k sD [5, 6].
RDC
Dðk; s2 ¼ sD ; D ¼ sD Þ ¼ k4 þ fk3 ½ þ !5
Lcalc
RDC
þ k2 ! 5 ð1 k sD Þg eksD
Lcalc
Possible characteristic equations: (I) Dðk; sÞ ¼ Pn ðk; sÞ þ Qm ðk; sÞ eks 8 n [ m

(II) Dðk; DÞ ¼ Pn ðk; sÞ þ Qm ðk; sÞ ekD (III) Dðk; sD Þ ¼ Pn ðk; sÞ þ Qm ðk; sÞ eksD .
We summary, our results in the following table:
s2 ¼ s; D ¼ 0(n > m) s2 ¼ 0; D [ 0(n > m) s2 ¼ sD ; D ¼ sD (n > m)

n 4 4 4
m 3 3 3
Pn k4 þ k3 LRcalc
DC
k4 þ k3 !5 k4
Qm k3 !5 þ k2 !5 LRcalc
DC
k3 LRcalc
DC
þ k2 !5 LRcalc
DC
k3 ½
RDC
þ !5
Lcalc
RDC
þ k2 !5 ð1 k sD Þ
Lcalc
Our RFID bursts switch homogeneous system for z; iD1 ; iD2 ; y leads to a char-
acteristic equation for the eigenvalue k having the form PðkÞ þ QðkÞ eks ¼ 0.
First case s2 ¼ s; D ¼ 0. The general characteristic equation D(k, s) is ad follow:
RDC RDC
Dðk; s2 ¼ s; D ¼ 0Þ ¼ k4 þ k3 þ ½k3 !5 þ k2 !5 eks
Lcalc Lcalc

X
n
Pn ðk; sÞ ¼ Pk ðsÞ k ¼ P0 ðsÞ
k
k¼0
þ P1 ðsÞ k þ P2 ðsÞ k2 þ P3 ðsÞ k3 þ P4 ðsÞ k4

P0 ðsÞ ¼ 0; P1 ðsÞ ¼ 0; P2 ðsÞ ¼ 0;
RDC
P3 ðsÞ ¼ ; P4 ðsÞ ¼ 1
Lcalc
The expression for Qn ðk; sÞ is
X
M
Qn ðk; sÞ ¼ qk ðsÞ kk ¼ q0 ðsÞ þ q1 ðsÞ k þ q2 ðsÞ k2 þ q3 ðsÞ k3 q0 ðsÞ
k¼0
RDC
¼ 0; q1 ðsÞ ¼ 0; q2 ðsÞ ¼ !5 ; q3 ðsÞ ¼ !5
Lcalc
The homogeneous system for z; iD1 ; i0D2 ; y leads to a characteristic equation for
the eigenvalue k having the form
X
4 X
3
Pðk; sÞ þ Qðk; sÞ eks ¼ 0; PðkÞ ¼ aj k j ; QðkÞ ¼ cj k j
j¼0 j¼0
And the coefficients faj ðqi ; qk ; sÞ; cj ðqi ; qk ; sÞg 2 R depend on qi ; qk and delay
qi ; qk is any RFID burst switching parameters, other parameters keep as a constant.
RDC
a0 ¼ 0; a1 ¼ 0; a2 ¼ 0; a3 ¼ ;
Lcalc
RDC
a4 ¼ 1; c0 ¼ 0; c1 ¼ 0; c2 ¼ !5 ; c3 ¼ ! 5
Lcalc
a0 + c0 6¼ 0 (not in sub case I) for 8 qi ; qk 2 R þ , i.e. k = 0 is not a of
PðkÞ þ QðkÞ eks ¼ 0. Furthermore, P(k), Q(k) are analytic functions of k, for
which the following requirements of the analysis [BK] can also be verified in the
present case:
polynomial in x.
(d) Each positive root xðqi ; qk Þ of F (x) = 0 is continuous and differentiable
We assume that Pn ðk; sÞ and Qm ðk; sÞ can’t have common imaginary roots. That
is, for any real number x,
pn ðk ¼ i x; sÞ þ Qm ðk ¼ i x; sÞ 6¼ 0:
RDC
pn ðk ¼ i x; sÞ ¼ x4 i x3 ;
Lcalc
RDC
Qm ðk ¼ i x; sÞ ¼ i x3 !5 x2 !5
Lcalc
RDC RDC
¼ x4 i x3 i x3 !5 x2 !5 6¼ 0
Lcalc Lcalc
RDC 2
jPði x; sÞj2 ¼ P2R þ P2I ¼ x8 þ x6 ½ ;
Lcalc
RDC 2
jQði x; sÞj2 ¼ Q2R þ Q2I ¼ x6 !25 þ x4 !25 ½
Lcalc
RDC 2 RDC 2
¼ x8 þ x6 ½ x6 !25 x4 !25 ½
Lcalc Lcalc
RDC 2 RDC 2
¼ x8 þ x6 f½ !25 g x4 !25 ½
Lcalc Lcalc
U0 ; U2 ; U4 ; U6 ; U8
RDC 2 RDC 2
U0 ¼ 0; U2 ¼ 0; U4 ¼ !25 ½ ; U6 ¼ ½ !25 ; U8 ¼ 1
Lcalc Lcalc
P
4
Hence Fðx; sÞ ¼ 0 implies U2k x2k ¼ 0. And its roots are given by solving
k¼0
the above polynomial. Furthermore
RDC RDC
PR ði x; sÞ ¼ x4 ; PI ði x; sÞ ¼ x3 ; QR ði x; sÞ ¼ x2 !5
Lcalc Lcalc
QI ði x; sÞ ¼ x3 !5 ; sin hðsÞ ¼
jQði x; sÞj2
cos hðsÞ ¼
jQði x; sÞj2
DC 2
x7 !5 þ x5 ½LRcalc !5 ½ LRcalc
DC
þ Lcalc x !5
RDC 6
sin hðsÞ ¼ ; cos hðsÞ ¼ ¼0
DC 2 DC 2
x6 !25 þ x4 !25 ½LRcalc x6 !25 þ x4 !25 ½LRcalc
DC 2 DC 2
x7 !5 þ x5 ½LRcalc !5 x5 !5 fx2 þ ½LRcalc g x
sin hðsÞ ¼ 2
¼ ¼ ; cos hðsÞ ¼ 0
DC 2
x6 !5 þ x4 !5 ½Lcalc
2 2 RDC
x4 !5 fx2 þ ½Lcalc
2 R
g !5

definition: k ! j; pk ðsÞ ! aj ; qk ðsÞ ! cj ; n ¼ 4; m ¼ 3; n [ m. Additionally
P4 P
3
Pn ðk; sÞ ! PðkÞ; Qm ðk; sÞ ! QðkÞ then PðkÞ ¼ aj k j ; QðkÞ ¼ cj k j
j¼0 j¼0
PðkÞ ¼ k4 þ k3 LRcalc
DC
; QðkÞ ¼ k3 !5 þ k2 !5 LRcalc DC
n; m 2 N0 ; n [ m and
aj ; cj : R þ 0 ! R are continuous and differentiable function of s such that a0 þ c0 6
¼ 0 (not in sub case I). In the following “—” denotes complex and conjugate.
PðkÞ; QðkÞ are analytic functions in k and differentiable in s. The coefficients
faj ðLcalc ; RDC ; C1 ; L1 ; Cload ; RL ; s; . . .Þ and cj ðLcalc ; RDC ; C1 ; L1 ; Cload ; RL ; s; . . .Þg 2
R depend on RFID burst switch system’s Lcalc ; RDC ; C1 ; L1 ; Cload ; RL ; s; . . . values.
RDC RDC
a0 ¼ 0; a1 ¼ 0; a2 ¼ 0; a3 ¼ ; a4 ¼ 1; c0 ¼ 0; c1 ¼ 0; c2 ¼ !5 ; c3 ¼ ! 5
Lcalc Lcalc

ðLcalc ; RDC ; C1 ; L1 ; Cload ; RL ; s; . . .Þ will subsequently be omitted from P, Q, aj, cj.
ments. 8 Lcalc ; RDC ; C1 ; L1 ; Cload ; RL ; s; . . . 2 R þ I.e. k ¼ 0 is not a root of the
characteristic equation. Furthermore PðkÞ; QðkÞ are analytic functions of k for
which the following requirements of the analysis (see Kuang 1993, Sect. 3.4) can
also be verified in the present case [6, 7].
ðLcalc ; RDC ; C1 ; L1 ; Cload ; RL ; s; . . .Þ domain of interest.
R
k3 !5 þ k2 !5 L DC
Indeed, in the limit j QðkÞ j ¼ j
PðkÞ R
calc
j
k4 þ k3 L DC
calc

ðcÞ RDC 2 RDC 2
¼ x8 þ x6 f½ !25 g x4 !25 ½
Lcalc Lcalc
Has at most a finite number of zeroes. Indeed, this is a polynomial in x (Degree
in x8 ).
(d) Each positive root xðLcalc ; RDC ; C1 ; L1 ; Cload ; RL ; s; . . .Þ of FðxÞ ¼ 0 is con-

tinuous and differentiable with respect to Lcalc ; RDC ; C1 ; L1 ; Cload ; RL ; s; . . ..
In addition, since the coefficients in P and Q are real, we have, and Qði xÞ ¼
Qði xÞ thus, x [ 0 may be an eigenvalue of the characteristic equation. The
analysis consists in identifying the roots of characteristic equation situated on the
imaginary axis of the complex k-plane, where by increasing the parameters
Lcalc ; RDC ; C1 ; L1 ; Cload ; RL ; s; . . ., Rek may, at the crossing,Change its sign from
ðjÞ ðjÞ
(−) to (+), i.e. from a stable focus E ðjÞ ðZ ðjÞ ; ID1 ; ID2 ; Y ðjÞ ¼ 0Þ to an unstable one, or
vice versa. This feature may be further assessed by examining the sign of the partial
derivatives with respect to Lcalc ; RDC ; C1 ; L1 ; Cload ; RL ; s; . . . and gate antenna
parameters.
@Rek
^1 ðLcalc Þ ¼ ð Þ ; RDC ; C1 ; L1 ; Cload ; RL ; s; . . . ¼ const
@Lcalc k¼ix
@Rek
^1 ðRDC Þ ¼ ð Þ ; Lcalc ; C1 ; L1 ; Cload ; RL ; s; . . . ¼ const
@RDC k¼ix
@Rek
^1 ðC1 Þ ¼ ð Þ ; Lcalc ; RDC ; L1 ; Cload ; RL ; s; . . . ¼ const
@C1 k¼ix
@Rek
^1 ðL1 Þ ¼ ð Þ ; Lcalc ; RDC ; C1 ; Cload ; RL ; s; . . . ¼ const
@L1 k¼ix
@Rek
^1 ðCload Þ ¼ ð Þ ; Lcalc ; RDC ; C1 ; L1 ; RL ; s; . . . ¼ const
@Cload k¼ix
@Rek
^1 ðsÞ ¼ ð Þ ; Lcalc ; RDC ; C1 ; L1 ; Cload ; RL ; . . . ¼ const
@s k¼ix
x 2 Rþ :
X
4
¼ U0 þ U2 x2 þ U4 x4 þ U6 x6 þ U8 x8 ¼ U2k x2k
k¼0
P
4
Hence Fðx; sÞ ¼ 0 implies U2k x2k ¼ 0 When writing PðkÞ ¼ PR ðkÞ þ i
k¼0
PI ðkÞ and QðkÞ ¼ QR ðkÞ þ i QI ðkÞ, and inserting k ¼ i x into RFID burst switch
system’s characteristic equation, x must satisfy the following :
jQði xÞj2
jQði xÞj2

that IR þ 0 is the set where xðsÞ is a positive root of FðxÞ and for, s 62 I , xðsÞ is
any s 2 I , where xðsÞ is a positive solution of Fðx; sÞ ¼ 0, we can define the angle
sin hðsÞ ¼ ;
jQði xÞj2
cos hðsÞ ¼
jQði xÞj2
n 2 N0 ; s 2 I. Let us introduce the functions
I ! R; Sn ðsÞ ¼ s sn ðsÞSn ðsÞ ¼ s sn ðsÞ; s 2 I; n 2 N0 that are continuous and
differentiable in s. In the following, the subscripts k; x; Lcalc ; RDC ; C1 ; L1 ;
Cload ; RL ; . . . indicate the corresponding partial derivatives. Let us first concentrate
on ^ðxÞ, remember in kðLcalc ; RDC ; C1 ; L1 ; Cload ; RL ; . . .Þ and xðLcalc ; RDC ; C1 ; L1 ;
Cload ; RL ; . . .Þ, and keeping all parameters except one (x) and s. The derivation
closely follows that in reference [BK]. Differentiating RFID burst switch charac-
teristic equation PðkÞ þ QðkÞ eks ¼ 0 with respect to specific parameter (x), and
inverting the derivative, for convenience, one calculates:
Remark
x ¼ Lcalc ; RDC ; C1 ; L1 ; Cload ; RL ; . . .; etc:;

ð Þ ¼
Where Pk ¼ @P
@k ; . . .: etc., Substituting k ¼ i x, and bearing i
Pði xÞ ¼ Pði xÞ, Qði xÞ ¼ Qði xÞ then i Pk ði xÞ ¼ Px ði xÞ; i Qk ði
xÞ ¼ Qx ði xÞ and that on the surface jPði xÞj2 ¼ jQði xÞj2 , one obtains


Q ¼ QR þ i QI ; Px ¼ PRx þ i PIx ; Qx ¼ QRx þ i QIx ; Px ¼ PRx þ i PIx ;
Qx ¼ QRx þ i QIx ;P2 ¼ P2R þ P2I . When (x) can be any RFID burst switch
parameters and time delay s etc. Where for convenience, we have dropped the
arguments ði x; xÞ, and where

xx ¼ Fx =Fx . We define U and V:

V ¼ ðPR PIx PI PRx Þ ðQR QIx QI QRx Þ:
We choose our specific parameter as time delay x = s .
RDC RDC
PR ¼ x4 ; PI ¼ x3 ; QR ¼ x2 !5 ;
Lcalc Lcalc
QI ¼ x3 !5 ; PRs ¼ 0; PIs ¼ 0
RDC
QRs ¼ 0; QIs ¼ 0; PRx ¼ 4 x3 ; PIx ¼ 3 x2 ;
Lcalc
RDC
QRx ¼ 2 x !5 ; QIx ¼ 3 x2 !5
Lcalc
RDC RDC
QI QRx ¼ 2 x4 !25 ; PR PIx ¼ 3 x6 ;
Lcalc Lcalc
RDC RDC
PI PRx ¼ 4 x6 ; QR QIx ¼ 3 x4 !25
Lcalc Lcalc
RDC 2
PRx PR ¼ 4 x ; QRx QR ¼ 2 x !5 ½
7 3 2
;
Lcalc
V ¼ ðPR PIs PI PRs Þ ðQR QIs QI QRs Þ ¼ 0
RDC
U ¼ ðPR PIx P1 PRx Þ ðQR QIx QI QRx Þ ¼ 3 x6
Lcalc
RDC RDC RDC
þ 4 x6 ð3 x4 !25 2 x4 !25 Þ
Lcalc Lcalc Lcalc
RDC RDC RDC
U ¼ 3 x6 þ 4 x6 3 x4 !25
Lcalc Lcalc Lcalc
RDC RDC RDC
þ 2 x4 !25 ¼ x6 x4 !25
Lcalc Lcalc Lcalc
6 RDC 2 RDC RDC
QIx QI ¼ 3:x !5 ; U ¼ x
5 2
x !5
4
¼ x4 ½x2 !25
Lcalc Lcalc Lcalc
Fs ¼ 2 ½ðPRs PR þ PIs PI Þ ðQRs QR þ QIs QI Þ ¼ 0;
RDC 2
PIx PI ¼ 3 x5 ½
Lcalc

RDC 2 RDC 2
¼ 2 f4 x7 þ 3 x5 ½ ð2 x3 !25 ½ þ 3 x5 !25 Þg
Lcalc Lcalc
RDC 2 RDC 2
Fx ¼ 2 f4 x7 3 x5 !25 þ 3 x5 ½ 2 x3 !25 ½ g
Lcalc Lcalc
RDC 2 3
Fx ¼ 2 fx5 ð4 x2 3 !25 Þ þ ½ x ð3 x2 2 !25 Þg
Lcalc

@x @x Fs 1 @Rek @x Fs
Fx þ Fs ¼ 0; s 2 I ) ¼ ; ^ ðsÞ ¼ ; ¼ xs ¼
@s @s Fx @s k¼ix @s Fx
8 h i9
@x Fs < 2 U þ s jP j2
þ i F x =
¼ xs ¼ jFs ¼0 ¼ 0; ^1 ðsÞ ¼ Re h i
@s Fx : F þ i 2 V þ x jPj2 ;
s
@Rek
@s k¼ix
@x U @x
@s þ V
@s jPj2
@x U @x
@s þ V
signf^1 ðsÞg ¼ signfFx g signfs þxþ gj@x¼0
@s jPj2 @s
¼ signfFx g signfxg
RDC 2
signf^1 ðsÞg ¼ signf4 x7 þ 3 x5 ½
Lcalc
RDC 2
ð2 x3 !25 ½ þ 3 x5 !25 Þg signfxg
Lcalc

ðjÞ ðjÞ
RFID burst switch system, about the equilibrium point E ðjÞ ðZ ðjÞ ; ID1 ; ID2 ; Y ðjÞ ¼ 0Þ
as a result of a variation of delay parameter s. The analysis consists in identifying
the roots of our system characteristic equation situated on the imaginary axis of the
complex k-plane. Where by increasing the delay parameter s, Re k may at the
crossing, changes its sign from − to +, i.e. from a stable focus E(*) to an unstable
the partial derivatives with respect to s, ^1 ðsÞ ¼ ð@ Re k
@s Þk¼ix

1 @Rek
^ ðsÞ ¼ ;
@s k¼ix

@Rek
^1 ðsÞ ¼ ; Lcalc ; RDC ; C1 ; L1 ; Cload ; RL ; . . .; etc: ¼ const; x 2 R þ
@s k¼ix
We check the sign of ^1 ðsÞ according the following rule:

sign½Fx sign V þPx2 s U þ x þ xs s sign½^1 ðsÞ
± ± +
± –
RFID burst switch system stability switching analysis is done according the
below flow chart and based on [BK] geometric stability switch criteria in delay
differential systems with delay dependent parameters article [30, 31].
x ¼ Lcalc ; RDC ; C1 ; L1 ; Cload ; RL ; s; . . .; etc:;
Remark: The analysis consists in identifying the roots of circuit characteristic

equation PðkÞ ¼ QðkÞ eks ¼ 0 situated on the imaginary axis of the complex k-
Plane, where, by increasing the RFID burst switch system parameters. Rek may, at
the crossing, change its sign from “−” to “+”, i.e. from a stable focus E* to an
unstable one, or vice versa. This feature may be further assessed by examining the
sign of the partial derivatives with respect to system parameters. Other sub cases
sestability behavior (s2 ¼ 0; D [ 0 & s2 ¼ D ¼sD ) is not discussed and can be
good reader exercises [12].
Exercises
1. Active RFID system has two sources S1(t), S2(t) and two antennas L1 and L2
(rectangular antennas) as appear in the equivalent circuit. L1 and L2 configu-
ration structure can be represented as L2 inductor antenna which is connected in
the middle of L1 antenna. The overall parameters of two antennas are the same.
P
2
L1 ¼ Lcalc1 ¼ ½lp0 ð½ Xi X3 þ X4 Þ Ncp . Rectangular antennas. L2 ¼
i¼1
P
2 pffiffi
ð1 þ pÞ
Lcalc2 ¼ ½lp0 ð½ Xi X3 þ X4 Þ Nc . X1, X2, X3, and X4 global antenna
i¼1
parameters are the same for inductor antenna L1 and L2.
You can neglect the mutual inductance between inductor antennas L1 and L2. C
(Gama) is the two antennas trim parameter (0 < C < 1).
Exercises 145
1:1 Write RFID system, differential equations as a matrix representation.

1:2 Write the RFID system analog Van der pol’s equations. Find all transfor-
mations between Van der pol system parameters and variables to antenna
system’s variables and parameters.
1:3 Discuss stability, How C trim parameter influences our system stability
switching?
1:4 How RFID system dynamically changes for (A) S1(t) – ON, S2(t) – OFF
(B) S1(t) – OFF, S2(t) – ON (C) S1(t) – ON, S2(t) – ON.
1:5 Find ni (i = 1, 2, 3,…) functions of our RFID system by using regular
perturbation or averaging methods.
1:6 RFID TAG IC capacitance C1 multiple his value C1 ! 2C1, How our
RFID system stability change?
2. Active RFID system includes forcing sources Si(t); i = 1,2,… and antennas
inductors (L1, L2,…). All antennas are rectangular. The following differential
equation describes our RFID system (forced Van der pol equation). R1 and C1
are RFID IC parameters.
Xk
1 X k
€ þð 1 þ
V
1
Þ
1 _ 1
V þ pffiffiffiffiffi P V ¼ ½
1 dVSi

m
R1 R C C R dt
i¼1 Si 1
½ L1 þ Li C1 1 i¼1 Si
i¼2
2:1 Express our RFID system as a matrix differential equation system.

2:2 Find fixed point and discuss stability of our system.
2:3 How our Active RFID system behavior is dependent on k and m
parameters?
2:4 Write the equivalent Van der pol system parameters U(x), a, b when only
one forcing source is active. VSi OFF 8 i 2 ½1::k & i 6¼ n Except
VSn ON; n 62 ½1::k.
2:5 How the dynamic of our system change for the transformation
Pm P
m
Li ! ½L1 þ Li . Find fixed points and discuss the stability issue.
i¼2 i¼2
3. Our passive RFID TAG contains one RFID IC and two rectangular antennas in
the series. The two rectangular antenna parameters are not the same and the
definition is related to global parameters: Xi1, Xi2, Xi3, Xi4. i = 1for the first
antenna and i = 2 for the second antenna. The matrix formulation for RFID
differential equation:
2 3
0 1 0
6 f " 1 #g f C11R1 g 07
2 3 6 l X4 7
6 7
dV1 6 C1 f 0
ðX13 þ X1k Þ Nc p
7
6 dV 6 p 7
2 7
dt
6
4 dt 5 ¼ 6 k¼1;k6 ¼ 3 7
" # 7
dV3 6 l X4 7
6 2 7
dt
6 þ 0 ðX23 þ X2k Þ Ncp g 7
4 p k¼1;k6¼3
5
2 3 2 3 0 0 0
V1 0
4 V2 5 þ 4 0 5
V3 1
R1 and C1 are parameters for RFID TAG IC. V1, V2, V3 are system variables.
All other antenna parameters are the same as discuss in the chapter.
3:1 Find RFID TAG system fixed points and discuss the stability.
3:2 Discuss the system Eigen direction, Eigen solutions, Eigen vectors, and
Eigenvalues behavior for t ! ∞.
3:3 How our system stability is affected by different values of “p” parameter?
Draw Stable/Unstable diagram.
3:4 Analyze RFID TAG system dynamical behavior for X2k ¼ X1k 2
C;
k = 1,…, 4 RFID TAG antennas global parameter index. C is a shifting
parameter between the square of first antenna global parameters (X1k 2
) and
second antenna global parameters (X2k ).
3:5 How our RFID TAG system behavior changes for multiple values of
rectangular antenna’s number of turns (NC) ; NC ! 2 NC . NC is the same
for the first and second RFID TAG antenna.
4. We have delayed in time passive RFID TAG system. Due to electromagnetic
interferences, we have RFID TAG’s voltage and voltage derivative with
delays s þ 1 and s2 1 respectively in time. V1 ðtÞ ! V1 ðt ½s þ 1Þ; V2 ðtÞ !
V2 ðt ½s2 1Þ. We consider no delay effect on dVdt1 ðtÞ and dVdt2 ðtÞ . The RFID
TAG antenna is rectangular. Xi; i = 1, 2, 3, 4 are RFID TAG antenna global
parameters as discuss in the chapter. R1 and C1 are RFID TAG IC parameters.
dV1
¼ V2 ðt ½s2 1Þ
dt
dV2 1
¼ f g V1 ðt ½s þ 1Þ
dt C1 lp0 ½X1 þ X2 X3 þ X4 Ncp
1
V2 ðt ½s2 1Þ
C1 R1
4:1 Find system fixed points and discuss stability for s ¼ 0.

Exercises 147
4:2 Find the system characteristic equation (Dðk; sÞ), s is our delay parameter.
Dðk; sÞ ¼ Pn ðk; sÞ þ Qm ðk; sÞ eks .
4:3 Find polynomial in x representation Fðx; sÞ and sketch 3D function. Find
sin hðsÞ and cos hðsÞ expressions.
4:4 Find U; V; xs expressions and define maps Sn ðsÞ ¼ s sn ðsÞ; s 2 I; n 2
N0 .
4:5 Find K1 ðC1 Þ; K1 ðR1 Þ; K1 ðsÞ; signðK1 ðsÞÞ expressions and discuss
stability switching for different values of s parameters.
5. We have a RFID system with two rectangular antennas (L1, L2) in parallel and
one RFID IC (R1 and C1 parameters). There are parasitic resistances of our
RFID system, rp1 ; rp2 ; rp1 6¼ rp2 . The following figure is equivalent circuit of our
RFID system.
Parameters are the same for the first and second antennas. L1 ðX11 ; X12 ; X13 ;
X14 ; . . .Þ; L2 ðX21 ; X22 ; X23 ; X24 ; . . .Þ; X11 ¼ X21 ; X12 ¼ X22 X13 ¼ X23 ; X14 ¼
X24 . We define four variables for our RFID system. V1(t)—voltage on the first
antenna, V2(t) = dV1(t)/dt—voltage derivative on the first antenna, V3(t)—
voltage on the second antenna, V4(t) = dV3(t)/dt—voltage derivative on the
second antenna.
Remark Voltages on RFID TAG antennas are only on equivalent circuit
inductors L1, L2 without parasitic resistances. Due to electromagnetic inter-
ferences, we get RFID TAG’s antenna voltages and voltages derivatives with
delays in time:
D1 6¼ D2 ; V1 ðtÞ ! V1 ðt sÞ
pffiffiffiffiffiffiffiffiffiffi
V2 ðtÞ ! V2 ðt s D1 Þ; V3 ðtÞ ! V3 ðt s þ 1Þ;
pffiffiffi
V4 ðtÞ ! V4 ðt jD1 D2 j sÞ:
D1 [ 0; D2 [ 0:
5:1 Find RFID system, differential equations, fixed points and discuss stability
for s ¼ 0; Di [ 0 i = 1, 2; D1 6¼ D2 .
5:2 Find the system characteristic equation Dðk; s; D1 ; D2 Þ, s is our delay
parameter and D1, D2 are parameters.
nðs; D1 ; D2 Þ
Dðk; s; D1 ; D2 Þ ¼ Pn ðk; s; D1 ; D2 Þ þ Qm ðk; s; D1 ; D2 Þ eknðs;D1 ;D2 Þ :
5:3 Find polynomial in x representation Fðx; sÞjD1 ;D2 parameters ¼ 0 and sketch
3D function. Find sin hðsÞ and cos hðsÞ expressions.
5:4 Find U, V, xs expressions and define maps Sn ðsÞ ¼ s sn ðsÞ; s 2 I; n 2
N0 for the cases: (A) D1 = D, D2 = 0 ; (B) D1 = 0, D2 = D ;
(C) D1 = D2 = D
5:5 Find
K1 ðsÞ; K1 ðD1 Þ; K1 ðD2 Þ; signðK1 ðsÞÞ; signðK1 ðD1 ÞÞ; signðK1 ðD2 ÞÞ
expressions and discuss stability switching for different values of s; D1 ; D2 .
6. We have triple loop antennas arranged as a shifted gate in X direction.
The RFID TAG is semi passive and contains a battery that enables long reading
distance and also enables the tag to operate independently of the reader. The
double antenna gate is employed due to the fact that this antenna consists of
three parallel loops (primary, secondary, and third loop). Due to electromag-
netic interferences there are differences in time delays with respect to gate
antenna first, second and third loop voltages and voltages derivatives. The delay
voltages are Vi1 ðt s1 Þ; Vi2 ðt s2 Þ; Vi3 ðt s3 Þ respectively (s1 6¼ s2 6¼ s3 ) and
pffiffiffi
ðtDÞ dVi2 ðt½D þ DÞ
and dVi3 ðt½D þ 1Þ
2
delayed voltage derivative dVi1 dt ; dt dt ; s1 0;
s2 0; s3 0; D 0. Each triple loop gate antenna is defined as a three
inductors in series Li1, Li2, Li3 with series parasitic resistors rp1 ; rp2 ; i—index of
the first and second gate. First gate: L11, L12, L13 is mostly formed by traces on
the planar PCB. 2 Lm;12 ; 2 Lm;13 ; 2 Lm;23 , elements represent the mutual
inductances between each two antenna inductors in the gate. The second loop is
within the first loop and third loop is within the second loop. We consider that
the triple loop antennas parameter values are the same in the first and second
gate ðLa1 ; La2 ; La3 ; Lb1 ; Lb2 ; Lb3 ; a1 ; a2 ; a3 Þ.
Exercises 149
2 Aj 2 Aj
L1j ¼ 4 fLbj ln½ þ Laj ln½ þ 2 ½aj þ lcj ðLaj þ Lbj Þg
aj ðLb1 þ lcj Þ aj ðLbj þ lcj Þ
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
lcj ¼ L2aj þ L2bj ; Aj ¼ Laj Lbj
j = 1 first loop, j = 2 s loop, j = 3 third loop. Since three inductors (L11, L12,
L13) are in series and there are mutual inductances, the total antenna inductance
for the first gate:
X
3 X2 pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
LT jfirst gate ¼ L1k þ 2 ½ Lm;iði þ 1Þ þ Lm;13 ; Lm;12 ¼ K1 L11 L12
k¼1 i¼1
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
Lm;13 ¼ K2 L11 L13 ; Lm;23 ¼ K3 L12 L13 :
Lm,i-j is the mutual inductance between inductors i and j. K1, K2, and K3 are the
coupling coefficients of two inductors. 0 K1 1; 0 K2 1; 0 K3 1. We
consider the case our RFID shifted gate system is passive (power source is
disconnected). Remark: no delay effects on RFID system variables derivatives.
6:1 Find RFID double gate differential equations and fixed points (only one
gate).
6:2 Find Jacobian of our RFID shifted gate system and characteristic equation:
Dðk; s1 ; s2 ; s3 ; DÞ.
6:3 Find F(x, s) = 0 and its roots sinh(s) and cosh(s) expressions.
6:4 Find K1 ðs1 Þ; K1 ðs2 Þ; K1 ðs3 Þ; K1 ðDÞ expressions.
6:5 Discuss system stability switching for various values of s1 ; s2 ; s3 ; D.
7. We have RFID detector system which is represented by the following set of
differential equations matrix representation. Nkl (k = 1,…5; l = 1,…,5) global
parameter expressions are the same as describe in Sect. (1.5). Additional ele-
ments are X1, X2 and V(t) second order derivative column matrix element w. V
(t) represents the RFID tag antenna voltage in time, incoming RF small signal
from RFID reader.
0 1
dX 0 1
dt
B dY C 0 1 X
B dt C N11 ... N1n þ X2 B C
B C BY C
B dIL1 C B .. .. .. C B C
B dt C¼B C B C
B C @ . . . A B IL1 C
B dIRj C B C
B dt C Nm1 þ X1 Nmn @ IR j A
@ A n¼m¼5
dIRS IR S
dt
0 R 1 011 0 1
L Lin L
0
B 1 PC B PC B C
B 0 C B0C BwC
B C B C dVðtÞ B C d 2 VðtÞ
B C B
þ B L C VðtÞ þ B 0 C C þB C
B 0 C dt2
1
B 1 C B C dt B C
@ 0 A @0A @0A
0 0 0
7:1 Draw RFID TAG detector circuit which characterizes by our above dif-
ferential equations, matrix representation. What are the additional circuit
components and their location which represents by matrix’s parameters X1,
X2 and w? Remark: probably they are additional Schottky diode’s parasitic
elements.
7:2 Find system fixed points and discuss stability in the case of no parasitic
delay effects si ¼ 0; i = 1,2,….
7:3 Consider that the Schottky detector diode has a package parasitic induc-
tance Lp delay element in time s1 and package parasitic capacitance Cp
delay element in time s2 . Find fixed points coordinate expressions, con-
sider RF in signal VðtÞ ¼ A0 þ B0 f 2 ðtÞ.
VðtÞ ¼ A0 þ B0 f 2 ðtÞ ; jf ðtÞj\1 &A0 jf ðtÞj ; 0\B0 \1:
Find Jacobian of our RFID Schotky detector system.

7:4 Find our system characteristic equation Dðk; s1 ; s2 Þ for three cases:
(A) s1 ¼ s; s2 ¼ 0 (B)s1 ¼ 0; s2 ¼ s (C) s1 ¼ s2 ¼ s.
Exercises 151
7:5 Find ^1 ðsÞ ¼ ð@ Re k

@s Þk¼ix ; LP ; L1 ; Cf ; Rin ; Rs ; CP ; Rj ; . . . ¼ const; x 2 R þ
and discuss stability switching for different values of delay parameter s.
8. We have RFID detector system which is represented by the following set of
differential equations matrix representation. Nkl (k = 1,…5; l = 1,…,5) global
parameter expressions are the same as described in Sect. (1.5). Additional
elements are X1, X2 and V(t) second order derivative column matrix element w1
and w2. V(t) represents the RFID tag antenna voltage in time, the incoming RF
small signal from the RFID reader.
0 1
dX 0 1
dt
B dY C 0 1 X
B dt C N11 þ X2 ... N1n B C
B C BY C
B dIL1 C B .. .. C
.. C B C
B dt C¼B B C
B C @ . . . A B I L1 C
B dIRj C B C
B dt C Nm1 þ X1 Nmn n¼m¼5 @ IRj A
@ A
dIRS IRS
dt
0 1 011 01
LR1 Lin P LP
0
B C B C B C
B 0 C B0C Bw C
B C B C dVðtÞ B 1 C d 2 VðtÞ
þB 1
B L1 C
C VðtÞ þ B 0 C
B C dt þ B 0 C
B C dt2
B C B C B C
@ 0 A @0A @ 0 A
0 0 w2
8:1 Draw RFID TAG detector circuit which characterizes by our above dif-
ferential equations, matrix representation. What are the additional circuit
components and their location which represents by matrix’s parameters X1,
X2 and w1, w2? Remark: probably they are additional Schottky diode’s
parasitic elements.
8:2 Find system fixed points and discuss stability in the case of no parasitic
delay effects si ¼ 0; i = 1,2,….
8:3 Consider that the Schottky detector diode has a package parasitic inductance
Lp delay element in time s and package parasitic capacitance Cp delay
element in time s2 þ 1 . Find fixed points coordinate expressions, consider
RFin signal VðtÞ ¼ A0 þ B0 f 3 ðtÞ VðtÞ ¼ A0 þ B0 f 3 ðtÞ; jf ðtÞj\
1 & A0 jf ðtÞj; 0\B0 \1. Find Jacobian of our RFID Schottky detector
system.
8:4 Find our system characteristic equation Dðk; sÞ.
8:5 Find ^1 ðsÞ ¼ ð@ Re k
@s Þk¼ix ; LP ; L1 ; Cf ; Rin ; Rs ; CP ; Rj ; . . . ¼ const; x 2 R þ
and discuss stability switching for different values of delay parameter s.
9. Active RFID system includes forcing sources Si(t) ; i = 1, 2,… and antennas
inductors (L1, L2,…). All antennas are rectangular. The following differential
equation describes our RFID system (forced Van der pol equation). R1 and C1
are RFID IC parameters. Additional parameters are X1 and X2.
Xk
€ þ ð 1 þ X1
V
1
Þ
1 _
V þ pffiffiffiffiffi
1
V
R1 R C1 P
m
i¼1 Si ½ L1 þ X2 Li C1
i¼2
1 X
k
1 dVSi
¼ ½
C1 i¼1 RSi dt
9:1 Express our RFID system as a matrix differential equation system.

9:2 Find fixed point and discuss stability of our system for different values of
X1 and X2 parameters.
9:3 How our Active RFID system behavior is dependent on k, m, X1 and X2
parameters?
9:4 Write the equivalent Van der pol system parameters U(x), a, b when only
one forcing source is active. VSi OFF8i 2 ½1. . .k & i 6¼ n Except
VSn ON; n 62 ½1::k.
9:5 How the dynamic of our system change for the transformation
Pm Pm
Li ! ½L1 þ L2i . Find fixed points and discuss stability, issue for
i¼2 i¼2
pffiffiffiffi
X1 ¼ X; X2 ¼ 1 þ X.
10. We have delayed in time passive RFID TAG system. Due to electromagnetic
interferences, we have RFID TAG’s voltage and voltage derivative with delays
pffiffiffi
sþ1 and s3 1 respectively in time.
pffiffiffi
V1 ðtÞ ! V1 ðt ½ s þ 1Þ; V2 ðtÞ ! V2 ðt ½s3 1Þ. We consider no delay
effect on dVdt1 ðtÞ and dVdt2 ðtÞ . The RFID TAG antenna is rectangular. Xi; i = 1, 2, 3,
4 are RFID TAG antenna global parameters as discuss in the chapter. R1 and C1
are RFID TAG IC parameters.
dV1
¼ V2 ðt ½s3 1Þ;
dt
dV2 1 pffiffiffi
¼ f l g V1 ðt ½ s þ 1Þ
dt C1 p ½X1 þ X2 X3 þ X4 Nc
0 p
1
V2 ðt ½s3 1Þ
C1 R1
10:1 Find system fixed points and discuss stability for s ¼ 0.

10:2 Find the system characteristic equation (Dðk; sÞ), s is our delay parameter.
Dðk; sÞ ¼ Pn ðk; sÞ þ Qm ðk; sÞ eks .
Exercises 153
10:3 Find polynomial in x representation Fðx; sÞ and sketch 3D function. Find

sin hðsÞ and cos hðsÞ expressions.
10:4 Find U; V; xs expressions and define maps Sn ðsÞ ¼ s sn ðsÞ
Sn ðsÞ ¼ s sn ðsÞ; s 2 I; n 2 N0 :
10:5 Find K1 ðC1 Þ; K1 ðR1 Þ; K1 ðsÞ; signðK1 ðsÞÞ expressions and discuss
stability switching for different value of s parameters.
Chapter 2
Microwave Elements Description
and Stability Analysis
There are three types of microwave circuits which include microwave elements.
The first is a discrete circuit; packaged diodes/transistors mounted in coax and
waveguide assemblies. Second Hybrid MIC (Microwave Integrated Circuit);
diodes/transistors and microstrip fabricated separately and then assembled. The
third is MMIC (Monolithic Microwave Integrated Circuit); diodes, transistors and
microstrip fabricated simultaneously. The monolithic microwave integrated circuit
(MMIC) consists of diodes, transistor, microstrip transmission lines, microstrip
circuits, and other circuit elements, such as lumped capacitors, resistors, etc., which
have parasitic effects influence on overall system stability behavior. The discrete
microwave circuit can be PIN diodes mounted in a coaxial transmission line which
characterize by parasitic effects and delay variables in time. Hybrid microwave
integrated circuit’s wire bonds cause reliability problems and parasitic effects;
stability issue can affect every hybrid microwave integrated circuits. Many receivers
are often at risk of having their front end burned out by high power RF. Receivers
are traditionally protected by a power limiter circuit. The limiter diode is a special
type of the PIN diode. Due to the parasitic effects of microstrip transmission lines
there is a delay in time for input RF signal result in the end. Power limiters use with
transmission line face stability behavior for different delay time values. Reflection
Type Phase Shifter (RTPS), employing a circulator. Micro strip transmission lines
with three port active circulator, stability analysis under time delayed. Many RF
systems are use Active circulator as a passive non-reciprocal three- or four-port
device, in which microwave or radio frequency power entering any port is trans-
mitted to the next port in rotation (only). Micro strip transmission lines fid those
active circulator ports and face a delay parasitic effect of transferring signals in time.
These circulator’s micro strip transmission lines, delays cause to system instability.
Resonant RF network antennas are important to plasma sources with many appli-
cations. The cylindrical resonant RF network antennas run as large volume plasma

DOI 10.1007/978-3-319-45427-6_2
156 2 Microwave Elements Description and Stability Analysis
sources and have stability switching due to system’s copper legs parasitic effects.
The cylindrical RF network antennas structure is 16-leg cylindrical (birdcage) RF
antenna which has electrical circuit and opposite points of RF feeding and
grounding. Due to cylindrical antenna parasitic delayed in time, there is a stability
issue by analyzing its operation. Tunnel diode is the p-n junction device that
exhibits negative resistance. That means when the voltage increases the current
through it decreases. Typical Tunnel Diode (TD) I-V characteristic has two distinct
features: (1) it is STRONGLY non-linear (compare to the resistor I-V). Current-
Voltage relationships for TDs cannot be described using the Ohm’s law (2) it has a
negative differential resistance (NDR) region. Tunnel diode can be a microwave
oscillator. Transient is in the resonant cavity after turning the bias voltage ON. The
resonant circuit with NDR can oscillate. The TD microwave oscillator has parasitic
effects in time and delay variables. Stability is a very crucial issue when designing
microwave oscillator by using Tunnel Diode (TD) [14, 15].
2.1 Microstrip Transmission Lines Delayed in Time

Power Limiters Stability Analysis
Microwave and RF receivers, as well as many instruments, are susceptible to

damage from input signals having amplitudes which exceed some danger level. The
front end of some receivers can be destroyed by power levels. Avoiding such
damage is by using a power limiter which designed around a special type of PIN
diode called a limiter diode. A thin epitaxial I-layer is formed on a heavily N+
doped substrate, after which P+ top contacts are added by diffusion. Typical limiter
diodes have I-layer thickness between 2 and 7 lm, with corresponding values of
breakdown voltage. The diode is mounted in shunt across the microstrip trans-
mission line which leads to the receiver front end, and is provided with a DC bias
return. For incoming signals which are below the threshold level in amplitude, the
diode acts as an ordinary unbiased PIN diode, which is to say that it appears to be a
capacitor of relatively small value. When the incident signal exceeds the threshold
power level, the diode’s I-layer is flooded with carriers during the positive half
cycle of the incoming RF signal. Most of these carriers persist through the negative
half cycle, DC current begins to flow in the loop formed by the diode and bias
return choke, and the diode biases itself to a low value of resistance in a matter of
nanoseconds. Under the influence of this self generated bias current, the diode’s
junction resistance falls to a very low value, shorting out the transmission line. The
limiter circuit, then acts as a reflective switch, reflecting the large signal back to its
source and protecting the circuitry which is “downstream” from the limiter. Our
Microstrip transmission lines with power limiters system delay differential and
delay different model can be analytically by using delay differential equations in
2.1 Microstrip Transmission Lines Delayed in Time Power … 157
dynamically. The need of the incorporation of a time delay is often of the existence
of any stage structure. It is often difficult to analytically study models with delay
dependent parameters, even if only a single discrete delay is present. There is a
practical guidelines that combine graphical information with analytical work to
some function of microstrip delays s1, s2 which can be expressed, explicitly and
thus can be easily depicted by Matlab and other popular software. We need only
look at one such function and locate the zeros. This function often has only two
zeros, providing thresholds for stability switches. As time delay increases, stability
changes from stable to unstable to stable. We emphasize the local stability aspects
of some models with delay dependent parameters. Additionally, there is a general
geometric criterion that, theoretically speaking, can be applied to models with many
delays, or even distributed delays. The simplest case of a first order characteristic
equation, providing more user friendly geometric and analytic criteria for stability
switches. The analytical criteria provided for the first and second order cases can be
used to obtain some insightful analytical statements and can be helpful for con-
ducting simulations. The most obvious way in which to amount limiter diode in
shunt across a microstrip line. Two leads of limiter diode are mounted in parallel to
the transmission line and the third lead is soldered to the ground pad as shown. D1
is a limiter diode [24–26, 33–35] (Fig. 2.1).
The shunt mounted limiter diode equivalent circuit with microstrip lines delayed
in time. The time delay for the first line segment is s1 and the second line segment s2.
See Fig. 2.2.
It is possible to locate several limiter PIN diode on microstrip line, but in the
current chapter we focus on one limiter diode with the specific connection structure.
We consider for simplicity that the microstrip segments resistances are neglected
and either related voltages Vs1 ! e, Vs2 ! e. Then we can define
Va(t) = Vi(t − s1); Vo(t) = Va(t − s2) = Vi(t − (s1 + s2)). We do our stability
analysis of three different cases: s1 ¼ s; s2 ! e; s1 ! e; s2 ¼ s; s1 ¼ s2 ¼ s. We
defined Is1 ; Is2 as the current through first and second delay lines respectively.
Vs1 ; Vs2 ! e ) V0 ðtÞ ¼ Vi ðt ðs1 þ s2 ÞÞ (Vs1 ; Vs2 are voltages of the first and
second delay lines). Is1 ¼ ViRV
s
a
; Is2 ¼ VaRV
s
0
; Is1 ¼ Ia þ Is2 ; Ia ¼ ILi ; i ¼ 0; 1; 2
1 2
Fig. 2.1 Shunt mounted

limiter diode
Fig. 2.2 Shunt mounted

limiter diode equivalent
circuit with microstrip lines
delayed in time
dIa V dV
V ¼ VR1 ¼ VC1 ; Va ¼ ðL0 þ L1 þ L2 Þ þ V; Ia ¼ þ C1
dt R1 dt
dIa 1 dV 2
d V dIa X
2
¼ þ C1 2 ; Va ¼ Li þ V; V ¼ VðtÞ
dt R1 dt dt dt i¼0
1 dV d2V X 2
Va ¼ ½ þ C1 2 Li þ V
R1 dt dt i¼0
2
We consider Vs1 ; Vs2 ! e then Vi ðt s1 Þ ¼ Va ¼ ½R11 dV dt þ C1 dt2
d V
P2
i¼0 Li þ V. Shifting equation Vi ðt s1 Þ ¼ Va ¼ . . . in time by s1 gives equation:
1 dVðt þ s1 Þ d 2 Vðt þ s1 Þ X2
Vi ðtÞ ¼ ½ þ C1 L þ Vðt þ s1 Þ:
i¼0 i
R1 dt dt2
Finally, we get two Power limiter equations (with delays s1, s2)
1 dVðt s2 Þ d 2 Vðt s2 Þ X 2
V0 ðtÞ ¼ ½ þ C1 Li þ Vðt s2 Þ; Vi ðt s1 Þ
R1 dt dt2 i¼0
1 dV d2V X 2
¼½ þ C1 2 Li þ V
R1 dt dt i¼0
We get two Power limiter equations, one of them is a differential equation which
involves input coming signal.
1 dV d2 V X 2
Vi ðt s1 Þ ¼ ½ þ C1 2 Li þ V; xðtÞ ¼ Vi ðt s1 Þ;
R1 dt dt i¼0
X
2
1 X 2
n1 ¼ C 1 Li ; n2 ¼ Li ; n3 ¼ 1; V ¼ f ðn1 ; n2 ; n3 ; xðtÞÞ
i¼0
R1 i¼0
€ n1 þ V_ n2 þ V n3 ¼ xðtÞ ) V
V € þ V_ n2 þ V n3 ¼ 1 xðtÞ: This differential
n1 n1 n1
equation can recognize as forced Van der Pol equation. The basic Van der Pol
equation can be written in the form: V € þ a /ðVÞ V_ þ V ¼ b pðtÞ
n2 1 n 1 n3 1 1
¼ ; 3¼ ; ! ð1 eÞ; b pðtÞ ¼ xðtÞ ¼ xðtÞ
n1 R1 C1 n1 P2 n1 n1 P2
C1 Li C1 Li
i¼0 i¼0
1
¼ Vi ðt s1 Þ
P2
C1 Li
i¼0
1 1 1
a /ðVÞ ¼ ;a¼ ; /ðVÞ ¼ 1; b ¼ ! ð1 eÞ; pðtÞ
R1 C1 R1 C1 P2
C1 Li
i¼0
¼ Vi ðt s1 Þ
We can consider the incoming signal after delay s1 is T periodic and a, b are
non-negative parameters. It is convenient to write our power limiter Van der Pol
equation as autonomous system when ðV; W; hÞ 2 R2 x S2 .
V_ ¼ W R11C1 ; W_ ¼ V þ P1 2 Vi ðh s1 Þ; h_ ¼ 1: We suppose
C1 i¼0
Li
a; b 1, since we are interested in the periodic coming signal response we use 2p
x
periodic transformation.
! ! !
u1 cos x t x1 sin x t V
¼
u2 sin x t x1 cos x t W
du1 1 x2 1
¼ /ðVÞ cos x t ð Þ V sin x t
dt R1 C1 x
1
P sin x t Vi ðt s1 Þ
x C1 2i¼0 Li
du2 1 x2 1
¼ /ðVÞ sin x t ð Þ V cos x t
dt R1 C1 x
1
P cos x t Vi ðt s1 Þ
x C1 2i¼0 Li
First equation solution: V ¼ u1 cos x t u2 sin x t assuming that we are

near resonance, so that jx2 1j; a; b is all small ðr ¼ ð1 x2 Þ R1 C1 =xÞ, we get
du1 1 u1
¼ ½u1 r u2 ðu21 þ u22 Þ;
dt 2 R1 C1 4
du2 1 u2 1
¼ ½u2 þ r u1 ðu21 þ u22 Þ P
dt 2 R1 C1 4 2 x C1 2i¼0 Li
There is no significant difference if we set one of delays s1, s2 to zero for that
Van der Pol equation. The second Power limiter differential equation involves
output voltage.
1 dVðt s2 Þ d 2 Vðt s2 Þ X 2
V0 ðtÞ ¼ ½ þ C1 Li þ Vðt s2 Þ
R1 dt dt2 i¼0
We consider coming signal cause at t = 0 voltage V(t = 0), first power limiter
equation (Van der Pol). Additionally V0(t = 0) = 0. Then we get our second delay
2Þ
2
2Þ
P
differential equation: ½R11 dVðts
dt þ C1 d Vðts
dt2 2i¼0 Li þ Vðt s2 Þ ¼ 0.
V(t = 0) is calculated from the first power limiter Van der Pol equation,
V(t = 0) = U1). We can represent our second power limiter delay differential
equation as a general linear real scalar neutral differential equation with single delay
s2 (s2 > 0).
X
n
dk Xn
dk d0
ak VðtÞ þ bk Vðt sÞ ¼ 0; 0 VðtÞ , VðtÞ
k¼0
dtk k¼0
dtk dt
d0 X2
dk
0
Vðt s2 Þ , Vðt s2 Þ; n ¼ 2; a0 ¼ a1 ¼ a2 ; bk k Vðt s2 Þ ¼ 0
dt k¼0
dt
1 X 2
b0 ¼ 1; b1 ¼ Li ; b2
R1 i¼0
X2
d0 d d2
¼ C1 Li ; b0 0 Vðt s2 Þ þ b1 Vðt s2 Þ þ b2 2 Vðt s2 Þ ¼ 0
i¼0
dt dt dt
d0 d
Since 0
Vðt s2 Þ , Vðt s2 Þ then b0 Vðt s2 Þ þ b1 Vðt s2 Þ þ b2
dt dt
d2
Vðt s2 Þ ¼ 0.
dt2
It is well known that if the characteristic equation associated with a linear neutral
equation has roots only with negative real parts, and if all the roots are uniformly
bounded away from the imaginary axis, then the trivial solution of the linear neutral
equation is uniformly asymptotically stable. Thus the stability analysis of power
limiter second neural differential equation with single delay s2 is very much
equivalent to the problem of determining the conditions under which all roots of its
characteristic equation lie in the half of the complex plane and are uniformly
bounded away from the imaginary axis. In our case ak ¼ 0 then
P P
ð 2k¼0 bk kk Þ eks2 ¼ 0; PðkÞ ¼ 0; QðkÞ ¼ 2k¼0 bk kk .
Theorem 1.0 if |b2| > 1, then for all s2, there is an infinite number of roots of
P
QðkÞ eks2 ¼ 0 whose real parts are positive. b2 ¼ C1 2i¼0 Li )
P2
jC1 i¼0 Li j [ 1.
Theorem 2.0 if |b2| > 1, then the trivial solution of power limiter DDE (Delay
Differential Equation) is unstable for all s2 > 0.
Theorem 3.0 Let f ðk; s2 Þ ¼ k2 þ a k2 eks2 þ gðk; s2 Þ where gðk; s2 Þ is an
analytic function. Assume |a| > 1 and lim Rek [ 0 k12 gðk; s2 Þ ¼ 0 then, for, all s2 [ 0,
jkj ! 1
there is an infinite number of roots of f ðk; s2 Þ ¼ 0 whose real parts are positive. In
fact, there is a sequence {ki} of the roots of f ðk; s2 Þ ¼ 0 such jki j ! 1, and
limi!1 Re ki ¼ s12 ln jaj [ 0 when s2 [ 0.
Theorem 4.0 Let f ðk; s2 Þ ¼ k2 þ gðk; s2 Þ where gðk; s2 Þ is an analytic function.

Assume a ¼ lim Rek [ 0 jk2 gðk; s2 Þj\1 when, as s2 varied, the sum of multiplicities
jkj ! 1
of the roots of f ðk; s2 Þ ¼ 0 in the open right half plane can change only if a root
appears on or crosses the imaginary axis. Back to our second power limiter DDE
which can be considered as the following second order real scalar linear neutral
2Þ 2Þ
2
delay equation: a d Vðts
dt2 þ b dVðts
dt þ c Vðt s2 Þ ¼ 0 where s2 ; a; b; c are
real constant. To find the equilibrium points (fixed points) of delayed power limiter
circuit is by limt!1 Vðt s2 Þ ¼ limt!1 VðtÞ 8 t s2 9 ðt - s2 Þ t, t ! 1t!1 .
2Þ 2Þ
2
dVðtÞ
dt ¼ 0; dVðts
dt ¼ 0; d Vðts
dt2 ¼ 0 and we get one fixed point c V ði¼0Þ ¼ 0;
c V ði¼0Þ ¼ 0 ) c 6¼ 0 ) V ði¼0Þ ¼ 0.
equilibrium points of delayed circuit consists in adding to coordinate V arbitrarily
small increments of exponential form v ekt . This leads to a polynomial charac-
teristic equation in the eigenvalues k. The polynomial characteristic equation
accepts by set the below voltage and voltage derivative respect to time in delayed
power limiter differential equation. The delayed circuit fixed values with an arbi-
trarily small increment of exponential form v ekt is i = 0 (first fixed point), i = 1
(second fixed point), etc., VðtÞ ¼ V ðiÞ þ v ekt
dVðtÞ dVðt s2 Þ
Vðt s2 Þ ¼ V ðiÞ þ v ekðts2 Þ ;¼ v k ekt ;
dt dt
ks2 d Vðt s2 Þ
2
kt
¼vke e ; ¼ v k e eks2
2 kt
dt2
We choose the above expression for our V(t) as small displacement v from our
circuit fixed points at time t = 0. Vðt ¼ 0Þ ¼ V ðiÞ þ v. We get the following char-
acteristic equation: a k2 eks2 þ b k eks2 þ c eks2 ¼ 0. ða k2 þ b k þ cÞ
eks2 ¼ 0. Suppose k ¼ i x; x [ 0 is the root of ða k2 þ b k þ cÞ eks2 ¼ 0
for some s2 . Euler’s formulas: eixs2 ¼ cosðx s2 Þ i sinðx s2 Þ. We get
½ðc a x2 Þ þ i b x eixs2 ¼ 0 and ðc a x2 Þ cosðx s2 Þ þ b x
sinðx s2 Þ þ i ðb x cosðx s2 Þ ðc a x Þ sinðx s2 ÞÞ ¼ 0
2
P P
a ¼ C1 2i¼0 Li ; b ¼ R11 2i¼0 Li ; c ¼ 1; x 6¼ 0; x [ 0. We get two equations:
ðc a x2 Þ cosðx s2 Þ þ b x sinðx s2 Þ
¼ 0; ðb x cosðx s2 Þ ðc a x2 Þ sinðx s2 ÞÞ ¼ 0
Thus ðc a x2 Þ2 þ ðb xÞ2 ¼ 0; Hence a2 x4 þ x2 ðb2 2 c aÞ þ c2 ¼ 0.

Its roots are x2 ¼ 2a1 2 ½2 c a b2 ððb2 2 c aÞ2 4 c2 a2 Þ1=2 . We have
two cases for the above equation ðx2 ¼ . . .Þ.
Case (I): ðb2 2 c aÞ2 ¼ 4 c2 a2 ) b2 ðb2 4 c aÞ ¼ 0
1 X 2 X2
b ! e; Li ! e; R1 ! 1 & Li ! e; b2 4 c a ¼ 0 ) b2
R1 i¼0 i¼0
X2
¼4ca) Li ¼ 4 C1 R21
i¼0
Then x2 ¼ 2a1 2 ½2 c a b2 ¼ ac 12 ðbaÞ2

Case (II):
ðb2 2 c aÞ2 [ 4 c2 a2 ) b2 ðb2 4 c aÞ [ 0; b2 [ 0 ;

b2 4 c a [ 0 ) b2 [ 4 c a
1 X 2
2
X2 X2
1 X 2
ð Li Þ [ 4 C1 Li ; ð Li Þ ½ Li 4 C1 [ 0;
R21 i¼0 i¼0 i¼0
R21 i¼0
X
2
1 X 2
1 X
2
ð Li Þ [ 0; Li 4 C1 [ 0 ) 2 Li [ 4
i¼0
2
R1 i¼0 R1 C1 i¼0
Thus we have two imaginary solutions k ¼ i x with x þ [ x . We need to

determine the sign of the derivative of Re kðs2 Þ at the points where kðs2 Þ is purely
2Þ
imaginary. dsd 2 ekðs2 Þs2 ¼ ðdkðs
ds2 s2 þ kÞ e
kðs2 Þs2
d
ds2 ½a k2 þ b k þ c ¼ ð2 a k þ bÞ ds
dk
2
dk
. To get the expression for ds 2
we need
to calculate d
ds2 f½a k2 ðs2 Þ þ b kðs2 Þ þ c ekðs2 Þs2 g ¼ 0
dk k ða k2 þ b k þ cÞ
¼ ;
ds2 a s2 k2 þ k ð2 a b s2 Þ þ ðb c s2 Þ
dk 1 a s2 k2 þ k ð2 a b s2 Þ þ ðb c s2 Þ
ð Þ ¼
ds2 k ða k2 þ b k þ cÞ
2Þ
If kðs2 Þ ¼ i x is not simple, then dkðs s2 . Since x 6¼ 0; eixs2 6¼ 0
ds2 ¼ 0 at s2 ¼
hence a ði xÞ2 þ i x b þ c ¼ 0 Which implies ðc a x2 Þ þ i x b ¼ 0 and
then
X
2
1 X 2
b ¼ 0 & c a x2 ¼ 0 ) x2 C 1 Li ¼ 1; Li ! e
i¼0
R1 i¼0
dk a s2 k2 þ k ð2 a b s2 Þ þ ðb c s2 Þ
signfReð Þ1 gjk¼ix ¼ signfRe½ gjk¼ix
ds2 k ða k2 þ b k þ cÞ
dk 1 a s2 k þ ð2 a b s2 Þ b c s2
signReð Þ jk¼ix ¼ signfRe½ þ gjk¼ix
ds2 ða k þ b k þ cÞ
2
k ða k2 þ b k þ cÞ
dk 1
Finally, we get the expression for signfReðds 2
Þ gjk¼ix .
dk 1 a s2 x2 b þ ð2 a b s2 Þ ðc a x2 Þ
signfReð Þ gjk¼ix ¼ signf
ds2 ðc a x2 Þ2 þ ðx bÞ2
x2 b ðb c s2 Þ
g
ðx2 bÞ2 þ x2 ðc a x2 Þ2
Since ðc a x2 Þ2 þ ðx bÞ2 [ 0 for any a; b; c; x values we get the

expression:
dk 1
signfReð Þ gjk¼ix ¼ signfa s2 x2 b þ ð2 a b s2 Þ ðc a x2 Þ
ds2
x2 b ðb c s2 Þg
dk
signfReð Þ1 gjk¼ix ¼ fx2 ½2 a2 b2 þ b c s2 þ 2 a c b s2 cg
ds2
dk 1
By inserting the expression for x2 , we can check the sign of fReðds 2
Þ gjk¼ix .
There are two sets of values s2 for which there are imaginary roots:
0 h1 \2p; 0 h2 \2p. sn;1 ¼ xh1þ þ 2pn h2
x þ ; sn;2 ¼ x þ x .
2pn
We choose power limiter critical parameters: Li, R1, C1 and s2 delay parameter
and examine the possibility of stability transitions (bifurcation) due to parameter
values change or s2 delay change. The analysis consists in identifying the root of
power limiter second DDEs characteristic equation situated on the imaginary axis of
the complex k—plane, where by changing circuits parameters or s2 delay,
Rek may, at crossing, changes its sign from “−” to “+” i.e. from stable focus E* to
an unstable one, or vice versa. This feature may be further assessed by examining
the sign of the partial derivatives with respect to circuit parameters or delay s2 .
dðRekÞ dðRekÞ
signf gjk¼ix ; signf gjk¼ix xi ¼ R1 ; C1 ; Li
ds2 dxi

in a power limiter system, about the endemic equilibrium point V(i=0) as a result of
the variation of parameters R1, C1, L1 or L2 in shunt mounted limiter diode
equivalent circuit. Our analysis closely follows the procedure described in details in
reference [BK] for the time delay variation parameter s2 . We keep s2 fixed and
inspect stability switching for variation of parameters R1, C1, L1 or L2 respectively.
We already got the expression for the second power limiter differential equation
which lead the characteristic equation for the eigenvalue k having the form
P P
ð 2k¼0 bk kk Þ eks2 ¼ 0 ; PðkÞ ¼ 0 ; QðkÞ ¼ 2k¼0 bk kk where b0 ¼ 1; b1 =
P2 P2
R1 i¼0 Li ; b2 ¼ C1 i¼0 Li ; ak ¼ 0. We do a little parameters terminology
1
change: k ! j; bk ! cj and we get the following characteristic equation for the

P P
eigenvalue k : ð 2j¼0 cj k j Þ eks2 ¼ 0; PðkÞ ¼ 0; QðkÞ ¼ 2j¼0 cj k j where
P P
c0 ¼ 1; c1 ¼ R11 2i¼0 Li ; c2 ¼ C1 2i¼0 Li ; aj ¼ 0.
Remark: Do not confuse between ci parameters and C1 capacitor element in a
power limiter equivalent circuit and the coefficients faj ðR1 ; C1 ; Li Þ;
cj ðR1 ; C1 ; Li Þg 2 R.
Depend on R1, C1, and Li, but not on s2 . Unless strictly necessary, the desig-
nation of the variable arguments (R1, C1, Li) will subsequently be omitted from P,
Q, aj, cj.
The coefficients aj ; cj are continuous, and differentiable functions of their
arguments, and direct substitution shows that a0 þ c0 6¼ 0 8 R1 ; C1 ; Li 2 < þ ; in our
case: a0 ¼ 0; c0 ¼ 1 ) a0 þ c0 ¼ 1 6¼ 0. i.e., k ¼ 0 is not a root of a power limiter
characteristic equation. Furthermore, PðkÞ; QðkÞ are analytic functions of k, for
present case:
(a) If k ¼ i x; x 2 R, then Pði xÞ þ Qði xÞ 6¼ 0 i.e. P and Q have no com-
mon imaginary roots. This condition was verified numerically in the entire
(R1, C1, Li) domain of interest.
(b) jQðkÞ=PðkÞj is bounded for jkj ! 1, Re k
0. No roots bifurcation from ∞.
Indeed, in the limit jQðkÞ=PðkÞj ¼ Oðjc2 =a3 kjÞ ! 0:
(c) FðxÞ ¼ jPði xÞj2 jQði xÞj2 has at most a finite number of zeros. Indeed,
this is a bi-cubic polynomial in x.
(d) Each positive root xðR1 ; C1 ; Li Þ of FðxÞ ¼ 0 being continuous and differen-
tiable with respect to R1, C1, Li. This condition can only be assessed
numerically.

Pði xÞ ¼ Pði xÞ, and Qði xÞ ¼ Qði xÞ; thus k ¼ i x may be an eigen-
value of the characteristic equation. The analysis consists in identifying the roots of
the characteristic equation situated on the imaginary axis of the complex k plane,
where, by increasing the parameter R1, and/or C1 and/or Li, Re k may, at the
crossing, changes its sign from “−” to “+”, i.e. from a stable focus V(*) to an
unstable one, or vice versa [5, 6]. This feature may be further assessed by exam-
ining the sign of the partial derivatives with respect to R1, C1, and Li.
Reminder: We keep s2 fixed.
@Re k @Re k
K1 ðR1 Þ ¼ ð Þ ; C1 ; Li ¼ const:; ; K1 ðC1 Þ ¼ ð Þ ; R1 ; Li
@R1 k¼ix @C1 k¼ix
¼ const:;
K1 ðLi Þ ¼ ð@ Re k
@Li Þk¼ix ; R1 ; C1 ¼ const:; The subscripts k; x; R1 ; C1 ; Li indicate
the corresponding partial derivatives. Let us first concentrate on KðR1 Þ, remem-
bering that kðR1 ; C1 ; Li Þ; xððR1 ; C1 ; Li Þ, and keeping C1, Li and s2 fixed. The
derivation closely follows that in reference [BK]. Differentiating characteristic
equation with respect to R1, and inverting the derivative, for convenience, one
calculates:
@k 1 Pk ðk; R1 Þ Qðk; R1 Þ þ Qk ðk; R1 Þ Pðk; R1 Þ s2 Pðk; R1 Þ Qðk; R1 Þ

ð Þ ¼ð Þ
@R1 PR1 ðk; R1 Þ Qðk; R1 Þ QR1 ðk; R1 Þ Pðk; R1 Þ
where Pk ¼ @P=@k; . . .; etc. Substituting k ¼ i x, and bear in mind that

Pði xÞ ¼ Pði xÞ, and Qði xÞ ¼ Qði xÞ, i Pk ði xÞ ¼ Px ði xÞ and
i Qk ði xÞ ¼ Qx ði xÞ, and that on the surface exist jPði xÞj2 ¼ jQði xÞj2 , one
obtains:
@k 1 i Px ði x; R1 Þ Pði x; R1 Þ i Qk ði x; R1 Þ Qðk; R1 Þ s2 jPði x; R1 Þj2

@R1 PR1 ði x; R1 Þ Pði x; R1 Þ QR1 ði x; R1 Þ Qði x; R1 Þ
Upon separating into real and imaginary parts, with P ¼ PR þ i PI ; Q ¼

QR þ i QI ; Px ¼ PRx þ i PIx ; Qx ¼ QRx þ i QIx ; PR1 ¼ PRR1 þ i PIR1 ; QR1 ¼
QRR1 þ i QIR1 ; P2 ¼ P2I þ P2R , retaining the real part, and noting that the operators
∂ and Re commute, one come up, after some straightforward algebraic manipula-
tions, with the following result:
2 P2 U xR1 þ V
KðR1 Þ ¼ Fx ðs2 xR1 þ Þ;
FR1 þ 4 V
2 2 P2
PðkÞ ¼ 0 ) Pði xÞ ¼ 0 ) PR ¼ PI ¼ 0
PRx ¼ PIx ¼ PRR1 ¼ PIR1 ¼ 0;
X
2
1 X 2 X2
QðkÞ ¼ cj k j ¼ 1 þ ð Li Þ k þ C1 ð L i Þ k2
j¼0
R1 i¼0 i¼0
X
2
1 X 2 X2
Qðk ¼ i xÞ ¼ cj k j ¼ 1 þ i ð Li Þ x C1 ð Li Þ x2 ;
j¼0
R1 i¼0 i¼0
X
2 X
2
1 X 2
Qði xÞ ¼ cj k j ¼ f1 C1 ð Li Þ x2 g þ i ð Li Þ x
j¼0 i¼0
R1 i¼0
X
2
1 X 2
@QR
QR ¼ 1 C1 ð Li Þ x2 ; QI ¼ ð Li Þ x; QRx ¼
i¼0
R1 i¼0 @x
X
2
@QI 1 X 2
¼ 2 x C1 ð Li Þ; QIx ¼ ¼ ð Li Þ
i¼0
@x R1 i¼0
@QR 1 X 2
U xR1 þ V
QRR1 ¼ ¼ 0; QIR1 ¼ 2 ð Li Þ x; P2 ! 0; s2 xR1
@R1 R1 i¼0 P2
2
KðR1 Þ ! Fx ðU xR1 þ VÞ
FR21 þ 4 V 2
Where for convenience, we have dropped the arguments (i x, R1), and where
FR1 ¼ 2 ½ðPRR1 PR þ PIR1 PI Þ ðQRR1 QR þ QIR1 QI Þ; xR1 ¼ FR1 =Fx
and we get the expressions based on power limiter equivalent parameters:
X
2 X
2
1
Fx ¼ 2 x ð Li Þ ½2 C1 ð Li Þ f2 x2 C12 þ g;
i¼0 i¼0
R21
2 x2 X 2
FR1 ¼ ð Li Þ2
R31 i¼0
P
FR1 x ð 2i¼0 Li Þ
xR1 ¼ ¼
Fx P
2
R31 ½2 C1 ð Li Þ f2 x2 C12 þ 1
R21
g
i¼0
V ¼ ðPR PIR1 PI PRR1 Þ ðQR QIR1 QI QRR1 Þ
1 X 2 X 2
U¼ ð Li Þ f1 þ C1 x2 ð Li Þg;
R1 i¼0 i¼0
1 X 2 X2
V¼ ð L i Þ x ½1 C 1 ð Li Þ x2
R21 i¼0 i¼0
where x 2 R þ . If K(R1) > 0, K(C1) > 0, K(Li) > 0 (or <0), then the crossing
proceeds from “−” to “+” (or from “+” to “−”), respectively. Without loss of
generality, one may consider only k ¼ ix, x > 0 and ignore its complex conjugate.
Writing PðkÞ ¼ PR ðkÞ þ i PI ðkÞ; QðkÞ ¼ QR ðkÞ þ i QI ðkÞ and inserting k ¼ i x
into characteristic equation, x must satisfy the following:
sinðx s2 Þ ¼ gðxÞ ¼
jQði xÞj2
cosðx s2 Þ ¼ hðxÞ ¼
jQði xÞj2
where jQði xÞj2 6¼ 0 in view of the above requirement, and ðg; hÞ 2 R.

Furthermore, it follows from the above equations that, by squaring and adding
sides, x must be a positive root of FðxÞ ¼ jPði xÞj2 jQði xÞj2 ¼ 0.
And the above sin/cos equations, of course, are identical to those in reference
[BK], except that the variable arguments are R1, C1, and Li instead of s2 . Note that F
(x) is independent of s2 ; Thus equation (sin/cos) implies F(x), but not the other way
around. The real and imaginary parts of P and Q are discussed, while F and some of
its elementary properties are presented. One first solves the polynomial F(x),
retaining only the real positive roots x, and discarding the others. The result is a 2D
manifold (surface) x = x(R1, C1) in a three dimensional (3D) space (R1, C1, x),
where x is a continuous and differentiable with respect to its arguments, with the
possible exception of infinite derivatives on 1D continuous lines. Next, one checks
which x’s on the surface also satisfy both (sin/cos) equations for some fixed value of
s2 . This operation results in one, or several continuous lines on the surface. The
projection of these lines on the (R1, C1) plane gives the loci of possible stability
transitions of the dynamical system.
Remark: We can exchange R1 or C1 by Li in the 2D manifold (surface) or in a
three dimensional (3D) space.
We can give the sign of K(R1), without the leading positive factor by:
@Re k 1
signð Þ jk¼ix ¼ signKðR1 Þ; signKðR1 Þ
@R1
U xR1 þ V
¼ signðFx Þ signðs2 xR1 þ Þ
P2
We need now to do the same procedure for C1 parameter and get the expression:
@Rek
K1 ðC1 Þ ¼ ð Þ ; R1 ; Li ¼ const:; ;
@C1 k¼ix
PðkÞ ¼ 0 ) Pði xÞ ¼ 0 ) PR ¼ PI ¼ 0
PRx ¼ PIx ¼ PRC1 ¼ PIC1 ¼ 0; PC1 ¼ PRC1 þ i PIC1 ;
X
2
1 X 2
QR ¼ 1 C1 ð Li Þ x2 ; QI ¼ ð Li Þ x
i¼0
R1 i¼0
@QR X2
@QI
QRC1 ¼ ¼ ð Li Þ x2 ; QIC1 ¼ ¼ 0;
@C1 i¼0
@C 1
@Q
P ¼ 0 ) P2 ¼ 0; QC1 ¼ ¼ QRC1 þ i QIC1
@C1
Fx expression is the same. FC1 ¼ 2 ½ðPRC1 PR þ PIC1 PI Þ ðQRC1 QR þ QIC1 QI Þ

P
xC ¼ FC1 =Fx ; The expression for FC1 is FC1 ¼ 2 ð 2i¼0 Li Þ x2 ½1 C1
P2 1
ð i¼0 Li Þ x2
P
2
x ½1 C1 ð Li Þ x2
FC i¼0
x C1 ¼ 1 ¼
Fx P
2
2 C1 ð Li Þ f2 x2 C12 þ 1
R21
g
i¼0
U expression is the same like our previous calculation.
V ¼ ðPR PIC1 PI PRC1 Þ ðQR QIC1 QI QRC1 Þ ; V

X2
1 X 2
x3 X 2
¼ ð Li Þ x2 ð Li Þ x ¼ ð Li Þ2
i¼0
R1 i¼0
R 1 i¼0
@Rek 1
signð Þ jk¼ix ¼ signKðC1 Þ ; sign KðC1 Þ
@C1
U xC1 þ V
¼ signðFx Þ signðs2 xC1 þ Þ
P2
U is always less than zero (U < 0). V for parameter R1: if x > 0 then V > 0 for
P
C1 ð 2i¼0 Li Þ x2 \1 otherwise V < 0. V for parameter C1: If x > 0 then V < 0.
Now we choose our parameter L1.
X
3
PI ¼ PR ¼ 0 ) P2 ¼ 0; PIL1 ¼ PRL1 ¼ 0 Li ¼ L1 þ L2 þ L3 ;
i¼1
P3 P3 P3
@ Li @ Li @ @Q Li
i¼1
¼ i¼1
¼ i¼1
¼ 1; QL1 ¼
¼ QRL1 þ i QIL1
@L1 @L2 @L3 @L1
P
3 P3
@ Li @ Li
@QR @QI x x
QRL1 ¼ ¼ C1 x2 i¼1 ¼ C1 x2 ; QIL1 ¼ ¼ i¼1 ¼
@L1 @L1 @L1 R1 @L1 R1
FL1 ¼ 2 ½ðPRL1 PR þ PIL1 PI Þ ðQRL1 QR þ QIL1 QI Þ;
X
3
1
FL1 ¼ 2 x2 ½C1 ð Li Þ fC12 x2 þ g
i¼1
R21
P
3
x ½C1 ð
Li Þ fC12 x2 þ R12 g
FL1 i¼1 1
xL1 ¼ ¼ 3
Fx P P3
ð Li Þ ½2 C1 ð Li Þ f2 C12 x2 þ 1
R21
g
i¼1 i¼1
U expression is the same like our previous calculation.

V ¼ ðPR PIL1 PI PRL1 Þ ðQR QIL1 QI QRL1 Þ; V ¼ Rx1 ; If x > 0 then
V < 0 always. Retaining the real part, and noting that the operators ∂ and Re
commute, one come up, after some straightforward algebraic manipulations, with
the following result:
2 P2 U x L1 þ V @Re k 1
KðL1 Þ ¼ Fx ðs2 xL1 þ Þ; signð Þ jk¼ix
þ4 V
FL21 2 P 2 @L1
¼ sign KðL1 Þ
U xL1 þ V
sign KðL1 Þ ¼ signðFx Þ signðs2 xL1 þ Þ
P2
Our switching analysis results are the same if we move from L1 parameter to L2
or to L3 since the partial derivatives are the same.
Summary: We take the assumption that Vi(t) is an incoming signal width
Dt\s1 ; s2 ; Dt ! e. There are three time intervals which we analyze our power
limiter microstrip line system. The first time interval is s1 [ t [ 0, the coming
signal not yet pass the first delay line ðs1 Þ and VA, V, Vout respectively equal to
zero. The second time interval is s1 þ s2 [ t
s1 , the signal has not yet passed the
second delay line ðs2 Þ and mutual interaction between the signal and power limiter
equivalent circuit gives V(t) which is the voltage on resistor R1 and capacitor C1.
The dynamical analysis is done by using forced Van der Pol equation. The forcing
signal X(t) is the coming RF signal. The third time interval is t
s1 þ s2 , the
incoming signal passes both the first and second delay lines and the dynamical
behavior analysis is done by using Delay Differential Equation (DDE) and stability
switching analysis. For simplicity, we consider Vout ! e after s1 + s2 second.
In our analysis we consider the incoming RF signal to power limiter input
circuit, has time interval Dt. We choose Dt, Dt\s1 ; s2 . The incoming RF signal
time interval is less than delay lines times s1 ; s2 . First the incoming RF signal cause
to the voltage V (on shunt mounted limiter equivalent circuit’s R1, C1 elements).
The analysis is based on the Van der Pol’s equation. Second, we analyze output
voltage according to Delay Differential Equation (DDE), V is the main equation
variable in time [39–41] (Fig. 2.3).
Power limiter with microstrip transmission line system stability switching
analysis is done according geometric stability switch criteria [BK] in delay differ-
ential system with delay dependent variables. The first analysis is a power limiter
microstrip system with incoming RF signal. By using Van der Pol topology we find
V(t) voltage after s1 . The second analysis is power limiter microstrip system output
Vin
Incoming Incoming RF
RF signal Signal after first delay line
V t
Vo t [sec]
V t
DD
Van Der
Pol
Fig. 2.3 Shunt mounted limiter diode equivalent circuit with microstrip lines delayed time
diagram
differential equation with delay variable in time Vðt s2 Þ. V is the power limiter
equivalent circuit’s voltage on R1 and C1 . For the second analysis, we find out
system general characteristic equation Dðk; s2 Þ. Find Fðx; s2 Þ for each s2 has at
most a finite number of real zeros. Find x; s2 values which fulfill Fðx; s2 Þ ¼ 0;
xðs2 Þ, only for those values can be stability switching (first condition). Next is to
find those x; s2 values which fulfill the expressions, it is the second condition for
stability switching sinðx s2 Þ ¼ . . . cosðx s2 Þ ¼ . . .. If K(R1) > 0, K(C1) > 0,
K(Li) > 0 then the crossing proceeds from “−” to “+” respectively (stable to
unstable). If K(R1) < 0, K(C1) < 0, K(Li) < 0 then the crossing proceeds from “+”
to “−” respectively (unstable to stable). The analysis consists in identifying the
roots of microstrip power limiter circuit characteristic equation P(k) + Q(k) exp
(−k s2) = 0 situated on the imaginary axis of the complex k—plane, where, by
increasing the circuit parameters R1, C1, Li. Rek may, at the crossing, change its
sign from “−” to “+”, i.e. from a stable focus V* to an unstable one, or vice versa.
This feature may be further assessed by examining the sign of the partial derivatives
with respect to microstrip power limiter circuit parameters R1, C1, Li.
2.2 Three Ports Active Circulator’s Reflection Type Phase

Shifter (RTPS) Circuit Transmission Lines Delayed
in Time System Stability Analysis
Active circulator consisting of three ports, namely P1, P2 and P3. Active circulator is
a three terminal device in which input from one port is transmitted to the next port
in rotation. The active circulator acts as an isolator between the input and the output
signal so that phase shift is well observed. The RF input signal is given at P1 of the
circulator from the left side. This signal from P1 is transmitted to P2. We can
connect LC (L1, C1) components in series to P2 port which results in phase shift and
helps to reflect the signal to P3 at the right. At P3 we get an output RF signal. Each
active circulator terminal faces a delay parasitic effect of signal transferring in time
[25, 26, 35] (Fig. 2.4).
Our circuit is a Reflection Type Phase Shifter (RTPS), employing a circulator. In
the past was little interest in actively circulators since its narrow bandwidth and
problems associated with a hybrid realization. We use active circulators since their
bandwidths have increased considerably as a result of the advances in transistor
technology. Active circulators are ideally suited for realization using MMIC tech-
nology. The circuit employs decade bandwidth active circulator which shows very
low phase error characteristic. Additionally the phase shifter exhibits an excellent
input return loss performance across this decade bandwidth. The circuit configu-
ration of the active circulator used three MESFETs which are the GEC-Marconi
standard library cell F20-FET-4x75. As with all the standard library cells, a very
accurate, ultra-wideband small signal models of the device. MESFET stands for
metal semiconductor field effect transistor. It is similar to a JFET in construction
Fig. 2.4 Three ports decade

bandwidth active circulator
with micro strip delay lines
and LC phase shifter in port
P2
and terminology. The difference is that instead of using a p-n junction for the gate, a
Schottky (metal semiconductor) junction is used. A typical three ports decade
bandwidth active circulator has three MESFETs transistors interconnected with
each other. RF, CF, LF, CC, Rsb plays a major role in the working of the circuit. The
three feedback branches (RF, CF, LF) are used to link all the three transistors in an
end to end fashion. The source resistor (Rsb) is shared among all the three
MESFETs transistors and one transistor is source coupled with the other two
transistors using this source resistor. The circuit works in a symmetric fashion. We
consider MESFET high frequency model taking node capacitors into account.
Figure 2.5 describes the circuit configuration of the active circulator [35, 36].
In Fig. 2.5 we use N-type MESFET but usually the recommended is a sym-
metrical bilateral MESFET. All Cc and Cf capacitors are un-polarized. Once we
inject RF signal to port P1, it passes to port P2 through a feedback branch (RF, CF,
LF). The same is between ports P2 and P3, ports P3 and P1. In case we inject RF signal
to port P2, it reaches the Q1 gate and shorten the Q1’s drain and source. Then Port 2’s
RF signal is shortened to ground through resistor Rsb and didn’t reach port P1.
Fig. 2.5 Circuit

configuration of the active
circulator
2.2 Three Ports Active Circulator’s Reflection Type Phase Shifter … 173
Fig. 2.6 Active circulator system path equivalent circuit
The same is between P1 to P3 and P3 to P2. We consider a varactor which is realized

by connecting together the drain and source terminations of a standard MESFET—
resulting in a Schottky junction. The bias potential is then applied across the
drain/source and gate terminations. Our three ports decade bandwidth active cir-
culator with micro strip delay lines and LC phase shifter in port P2 gets his input RF
signal from the antenna (port P1) and feeds receiver unit by active circulator output
RF signal (Port P3) [1, 2]. The active circulator system can be described by the
system path from RFin port (P1) to RFout port (P3). For simplicity, we ignore
MESFET high frequency equivalent model and took it as a cutoff element in our
system. Figure 2.6 describes our system path from Antenna RF coming signal to the
receiver unit (Rx). Active circulator system path equivalent circuit fulfills current
equation: I1 ¼ I2 þ I3 .
We have three main variables in our active circulator system V1(t), V2(t), V3(t).
I1, I2, I3 are the currents through related branches. We describe system, differential
dVcf
equations. First branch: I1 ¼ Cf dt ; VLf ¼ Lf dIdt1 ; I1 ¼ Cc dVdtcc
d 1 1
½Vcf þ Vcc ¼ I1 ½ þ ; V1 V2 ¼ Vcf þ VLf þ VRf þ VCC ;
dt Cf Cc
dI1
Vcf þ Vcc ¼ V1 V2 Lf I1 Rf
dt
d dV1 dV2 d 2 I1 dI1
½Vcf þ Vcc ¼ Lf 2 Rf ;
dt dt dt dt dt
1 1 dV1 dV2 d 2 I1 dI1
I1 ½ þ ¼ Lf 2 Rf
Cf Cc dt dt dt dt
dVc1 dVc1
Second branch: I2 ¼ C1 dt ; VL1 ¼ L1 dIdt2 ; Vc1 ¼ V2 L1 dIdt2 ; I2 ¼ C1 dt
2
¼ C1 ½dV
dt L1 dt2 .
2 d I2
dVcf
Third branch: I3 ¼ Cf dt ; VLf ¼ Lf dIdt3 ; I3 ¼ Cc dVdtcc
d 1 1
½Vc þ Vcc ¼ I3 ½ þ ; V2 V3 ¼ Vcf þ VLf þ VRf þ VCC ;
dt f Cf Cc
dI3
Vcf þ Vcc ¼ V2 V3 Lf I3 Rf
dt
d dV2 dV3 d 2 I3 dI3
½Vcf þ Vcc ¼ Lf 2 Rf ;
dt dt dt dt dt
1 1 dV2 dV3 d 2 I3 dI3
I3 ½ þ ¼ Lf 2 Rf
Cf Cc dt dt dt dt
We can summarize our system, differential equations:
1 1 dV1 dV2 d 2 I1 dI1 dV2 d 2 I2

I1 ½ þ ¼ Lf 2 Rf ; I2 ¼ C1 ½ L1 2
Cf Cc dt dt dt dt dt dt
1 1 dV2 dV3 d 2 I3 dI3

I3 ½ þ ¼ Lf 2 Rf
Cf Cc dt dt dt dt
We implement Rx (receiver) unit with an equivalent circuit of the input section

of the receiver. The receiver’s amplifier is modeled as a noiseless amplifier pre-
ceded by noise voltage and noise current generators representing amplifier noise
referred to the input. The active circulator’s RFout port is connected to the amplifier
by a transformer with turns ratio m. We shall assume that this is an ideal trans-
former. Figure 2.7 describes the receiver input equivalent circuit.
We can consider the above equivalent circuit as resistor Ra, La, and Lt in the
series. V3 ¼ I3 Ra þ ðLa þ Lt Þ dIdt3 . After we integrated Rx unit differential equation
into our system, differential equations we get the following new system differential
equations:
Fig. 2.7 Receiver input equivalent circuit


1 1 dV1 dV2 d 2 I1 dI1 dV2 d 2 I2
I1 þ ¼ Lf 2 Rf ; I2 ¼ C1 L1 2
Cf Cc dt dt dt dt dt dt

1 1 dV2 d 2 I3 dI3
I3 þ ¼ ðLf þ La þ Lt Þ 2 ðRf þ Ra Þ
Cf Cc dt dt dt

dV2 d 2 I2 dV2 1 d 2 I2
I2 ¼ C1 L1 2 ) ¼ I 2 þ L1 2
dt dt dt C1 dt
We get two main systems, differential equations:
1 1 dV1 1 d 2 I2 d 2 I1 dI1
I1 ½ þ ¼ I2 L1 2 Lf 2 Rf
Cf Cc dt C1 dt dt dt
1 1 1 d 2 I2 d 2 I3 dI3
I3 ½ þ ¼ I2 þ L1 2 ðLf þ La þ Lt Þ 2 ðRf þ Ra Þ
Cf Cc C1 dt dt dt
Since I1 ¼ I2 þ I3 ) I2 ¼ I1 I3 we get the following system, differential

equations:
1 1 dV1 1 1 d 2 I1 d 2 I3
I1 ½ þ ¼ I1 þ I 3 L1 2 þ L 1 2
Cf Cc dt C1 C1 dt dt
d 2 I1 dI1
Lf Rf
dt2 dt
1 1 1 1 d 2 I1
I3 ½ þ ¼ I1 I 3 þ L1 2
Cf Cc C1 C1 dt
d 2 I3 dI3
ðLf þ La þ Lt þ L1 Þ ðRf þ Ra Þ
dt2 dt
dI1 dI3 d 2 I1 dI 0 d 2 I3 dI 0
¼ I10 ; ¼ I30 ; 2 ¼ 1 ; 2 ¼ 3
dt dt dt dt dt dt
We get a new set of system, differential equations:
1 1 1 dV1 1 dI 0 dI 0
I1 ½ þ þ ¼ þ I3 þ L1 3 ðLf þ L1 Þ 1 I10 Rf
Cf Cc C1 dt C1 dt dt
1 1 1 1 dI 0 dI 0
I3 ½ þ þ ¼ I1 þ L1 1 ðLf þ La þ Lt þ L1 Þ 3 I30 ðRf þ Ra Þ;
Cf Cc C1 C1 dt dt
dI1 dI
¼ I10 ; ¼ I30
3
dt dt
For simplicity we define: CR ¼ C1f þ C1c þ C11 and LR ¼ Lf þ La þ Lt þ L1 and

get the following system, differential equations:
dV1 1 dI 0 dI 0
I1 C R ¼ þ I3 þ L1 3 ðLf þ L1 Þ 1 I10 Rf ; I3 CR
dt C1 dt dt
1 dI10 dI30 dI1 dI3
¼ I 1 þ L1 LR I30 ðRf þ Ra Þ; ¼ I10 ; ¼ I30
C1 dt dt dt dt
dI10 dI30
From the above differential equation, we get the expressions for dt and dt :
dI10 dV1 1 ðLRLC

1
CR Þ ðC11 L1LCR R Þ
¼ þ 1
I 1 þ
dt dt fLf þ L1 ð1 LL1 Þg fLf þ L1 ð1 LL1 Þg fLf þ L1 ð1 LLR1 Þg
R R
Rf L1 ðRf þ Ra Þ
I3 I0 I0
fLf þ L1 ð1 LLR1 Þg 1 fLf þ L1 ð1 LLR1 Þg 3
For simplicity we define the following parameters:
1 ðLRLC
1
CR Þ ðC11 L1LCR R Þ
C0 ¼ ; C1 ¼ 1
; C2 ¼ ;
fLf þ L1 ð1 LLR1 Þg fLf þ L1 ð1 LLR1 Þg fLf þ L1 ð1 LLR1 Þg
Rf L1 ðRf þ Ra Þ
C3 ¼ ; C4 ¼
fLf þ L1 ð1 LLR1 Þg fLf þ L1 ð1 LLR1 Þg
dI10 dV1
¼ C0 þ C1 I1 þ C2 I3 þ C3 I10 þ C4 I30
dt dt
dI30 dV1 L1 C0 1 L1 C1 L1 C2 CR
¼ þ½ þ I1 þ ½ I3
dt dt LR LR C1 LR LR LR
L1 C3 0 L1 C4 ðRf þ Ra Þ 0
þ I1 þ ½ I3
LR LR LR
dI30 dV1 L1 C0 1 1 1 L1 C3
¼ þ ½ þ L1 C1 I1 þ ½L1 C2 CR I3 þ
dt dt LR LR C1 LR LR
1
I10 þ ½L1 C4 ðRf þ Ra Þ I30
LR
For simplicity we define the following parameters:
L1 C0 1 1 1
C5 ¼ ; C6 ¼ ½ þ L1 C1 ; C7 ¼ ½L1 C2 CR ;
LR LR C1 LR
L1 C3 1
C8 ¼ ; C9 ¼ ½L1 C4 ðRf þ Ra Þ
LR LR
dI30 dV1
¼ C5 þ C6 I1 þ C7 I3 þ C8 I10 þ C9 I30
dt dt
We can summarize our active circulator system, differential equations for

coming antenna signal and Rx unit. dIdt1 ¼ I10 ; dIdt3 ¼ I30
dI10 dV1 dI 0
¼ C0 þ C1 I1 þ C2 I3 þ C3 I10 þ C4 I30 ; 3
dt dt dt
dV1 0 0
¼ C5 þ C6 I1 þ C7 I3 þ C8 I1 þ C9 I3
dt
We have four system variables: I1 ; I3 ; I10 ; I30 and coming RFin signal V1(t). We
can write the above system, differential equations in a form of matrix representa-
tion: C5 ¼ LLR1 C0
0 1
dI1
B dt C
B C
B C 0 1 0 I1 1 0 0 1
B dI3 C !11 . . . !14
B C BI C B0 C
B dt C B . .. C dV1 ðtÞ
B C B .. C B 3C B C
B dI 0 C ¼@ .. . . AB C þ B C C0
@ I10 A @ 1 A
; !11 ¼ 0;
B 1C dt
B C !41 !44
B dt C I0 L1
B C 3 LR
@ dI A
0
3
dt
!12 ¼ 0; !13 ¼ 1; !14 ¼ 0; !21 ¼ 0; !22 ¼ 0
!23 ¼ 0; !24 ¼ 1; !31 ¼ C1 ; !32 ¼ C2 ; !33 ¼ C3 ; !34 ¼ C4 ; !41 ¼ C6 ; !42

¼ C7 ; !43 ¼ C8 ; !44 ¼ C9
We consider RFin signal V1(t) = A0 + f(t); |f(t)| <= 1 and A0 |f(t)| then
V1 ðtÞ ¼ A0 þ f ðtÞ A0 ) dVdt1 ðtÞ ! e. We can write our matrix representation:
e!0
0 1
dI1
B dt C 0 1
B C
B dI3 C 0 1 I1
B C !11 . . . !14 BI C
B dt C B . .. .. C B 3 C
B C
B dI 0 C ¼ @ .. . . AB Cþe
@ I10 A
B 1C
B dt C !41 !44
B C
@ dI 0 A I30
3
dt
Due to active circulator’s micro strip transmission line delays s1 in the first
port’s current and s3 for the third port current. Accordingly, active circulator’s
micro strip transmission lines, delays D1 for the first port current derivative and D3
for the third port current derivative. I1 ðtÞ ! I1 ðt s1 Þ; I3 ðtÞ ! I3 ðt s3 Þ
I10 ðtÞ ! I10 ðt D1 Þ; I30 ðtÞ ! I30 ðt D3 Þ. We consider no delay effect on
0 0
dI1 dI3 dI1 dI3
dt ; dt ; dt ; dt .
Active circulator’s micro strip transmission lines parasitic effects
influence only on P1 and P3 current and current derivatives I1 ; I3 ; I10 ; I30 (I2 ; I20 which
are a hidden variable in our analysis).
0 1
dI1
B dt C 0 1
B C
B dI3 C 0 1 I1 ðt s1 Þ
B C !11 . . . !14 B C
B dt C B . .. .. C B I3 ðt s3 Þ C
B C¼@ .
B dI 0 C . . . AB Cþe
@ I10 ðt D1 Þ A
B 1C
B dt C ! 41 !44
B C
@ dI 0 A I30 ðt D3 Þ
3
dt
To find equilibrium points (fixed points) of the active circulator system is by
lim I1 ðt s1 Þ ¼ I1 ðtÞ; lim I3 ðt s3 Þ ¼ I3 ðtÞ; lim I10 ðt s1 Þ ¼ I10 ðtÞ;

t!1 t!1 t!1
dI1 dI3
lim I 0 ðt s3 Þ ¼ I30 ðtÞ; ¼ ¼0
t!1 3 dt dt
dI10 dI30
¼ ¼ 0 8 t s1 ; t s3 ; t D1 ; t D3 9 ðt s1 Þ t; ðt s3 Þ t;
dt dt
ðt D1 Þ t; ðt D3 Þ t; t ! 1
We get four equations and the only one fixed point is C7 C6CC1 2 6¼ 0
C6 C2 ð0Þ ð0Þ 0 ð0Þ 0 ð0Þ

C7 6¼ 0 ) E ð0Þ ðI1 ; I3 ; I1 ; I3 Þ ¼ ð0; 0; 0; 0Þ
C1
equilibrium points of the active circulator system consists in adding to coordinate
½I1 ; I3 ; I10 ; I30 arbitrarily small increments of exponential form ½i1 ; i3 ; i01 ; i03 ekt , and
retaining the first order terms in ½I1 ; I3 ; I10 ; I30 . The system of four homogeneous
equations leads to a polynomial characteristic equation in the eigenvalues k. The
polynomial characteristics equations accept by set the below currents and currents
derivative respect to time into active circulator system equations. Active circulator
system fixed values with arbitrarily small increments of exponential form
½i1 ; i3 ; i01 ; i03 ekt are: j = 0 (first fixed point), j = 1 (second fixed point), j = 2 (third
fixed point), etc.,
ðjÞ ðjÞ 0 ðjÞ

I1 ðtÞ ¼ I1 þ i1 ekt ; I3 ðtÞ ¼ I3 þ i3 ekt ; I10 ðtÞ ¼ I1 þ i01 ekt ; I30 ðtÞ
0 ðjÞ
¼ I3 þ i03 ekt
We choose the above expressions for our I1 ðtÞ; I3 ðtÞ and I10 ðtÞ; I30 ðtÞ as a small
displacement ½i1 ; i3 ; i01 ; i03 from the active circulator system fixed points at time
ðjÞ ðjÞ 0 ðjÞ
t = 0. I1 ðt ¼ 0Þ ¼ I1 þ i1 ; I3 ðt ¼ 0Þ ¼ I3 þ i3 ; I10 ðt ¼ 0Þ ¼ I1 þ i01 ; I30 ðt ¼ 0Þ
0 ðjÞ
¼ I3 þ i03 .
For k [ 0; t [ 0 the selected fixed point is stable otherwise k [ 0; t [ 0 is

k [ 0; t [ 0 otherwise go away from the selected fixed point exponentially. k is
the eigenvalue parameter which establishes if the fixed point is stable or Unstable,
additionally his absolute value ðjkjÞ establishes the speed of flow toward or away
from the selected fixed point [3, 4] (Table 2.1).
The speeds of flow toward or away from the selected fixed point for active
circulator system currents and currents derivatives respect to time are
ðiÞ ðiÞ
dI1 ðtÞ I1 ðt þ DtÞ I1 ðtÞ I þ i1 ekðt þ DtÞ ½I1 þ i1 ekt
¼ lim ¼ lim 1
dt Dt!0 Dt Dt!0 Dt
i1 ekt ½ekDt 1 ekDt 1 þ kDt
¼ lim ! k i1 ekt
Dt!0 Dt
dI3 ðtÞ dI 0 ðtÞ dI 0 ðtÞ
¼ k i3 ekt ; 1 ¼ k i01 ekt ; 3 ¼ k i03 ekt
dt dt dt
dI1 ðt s1 Þ dI ðt s Þ
¼ k i1 ekt eks1 ; ¼ k i3 ekt eks3
3 3
dt dt
dI10 ðt D1 Þ dI 0 ðt D1 Þ
¼ k i01 ekt ekD1 ; 3 ¼ k i03 ekt ekD3
dt dt
First, we take the active circulator’s currents I1, I3 differential equations: dIdt1 ¼
I10 ; dIdt3 ¼ I30 and adding coordinates ½I1 ; I3 ; I10 ; I30 arbitrarily small increments of
exponential terms ½i1 ; i3 ; i01 ; i03 ekt and retaining the first order terms in i1 ; i3 ; i01 ; i03 .
Table 2.1 Active circulator system eigenvalues options

k<0 k>0
t=0 ðjÞ ðjÞ
I1 ðt ¼ 0Þ ¼ I1 þ i1 I1 ðt ¼ 0Þ ¼ I1 þ i1
ðjÞ ðjÞ
I3 ðt ¼ 0Þ ¼ I3 þ i3 I3 ðt ¼ 0Þ ¼ I3 þ i3
0ðjÞ 0ðjÞ
I10 ðt ¼ 0Þ ¼ I1 þ i01 I10 ðt ¼ 0Þ ¼ I1 þ i01
0ðjÞ 0ðjÞ
I30 ðt ¼ 0Þ ¼ I3 þ i03 I30 ðt ¼ 0Þ ¼ I3 þ i03
ðjÞ ðjÞ
t>0 I1 ðtÞ ¼ I1 þ i1 ejkjt I1 ðtÞ ¼ I1 þ i1 ejkjt
ðjÞ ðjÞ
I3 ðtÞ ¼ I3 þ i3 ejkjt I3 ðtÞ ¼ I3 þ i3 ejkjt
0ðjÞ 0ðjÞ
0ðjÞ 0ðjÞ
ðiÞ
t>0 I1 ðt ! 1Þ ¼ I1 I1 ðt ! 1; k [ 0Þ i1 ejkjt
t!∞
I3 ðt ! 1Þ ¼
ðiÞ
I3 I3 ðt ! 1; k [ 0Þ i3 ejkjt
I10 ðt ! 1Þ ¼
0ðiÞ
I1 I10 ðt ! 1; k [ 0Þ i01 ejkjt
I30 ðt ! 1Þ ¼
0ðiÞ
I3 I30 ðt ! 1; k [ 0Þ i3 ejkjt
0 ðiÞ 0 ði¼0Þ
k i1 ekt ¼ I1 þ i01 ekt ; I1 ¼ 0 ) k i1 þ i01 ¼ 0
0 ðiÞ 0 ði¼0Þ
k i3 ekt ¼ I3 þ i03 ekt ; I3 ¼ 0 ) k i3 þ i03 ¼ 0
Second, we take the active circulator’s currents derivative I10 ; I30 differential
equations:
dI10 dV1 dI 0
¼ C0 þ C1 I1 þ C2 I3 þ C3 I10 þ C4 I30 ; 3
dt dt dt
dV1 0 0
¼ C5 þ C6 I1 þ C7 I3 þ C8 I1 þ C9 I3
dt
0 0
dV1
dt C0 ! e; dV dt C5 ! e and adding coordinates ½I1 ; I3 ; I1 ; I3 arbitrarily small
1
increments of exponential terms ½i1 ; i3 ; i01 ; i03 ekt and retaining the first order terms
in i1 ; i3 ; i01 ; i03 .
ðjÞ ðjÞ 0ðjÞ 0ðjÞ
k i01 ekt ¼ C1 ðI1 þ i1 ekt Þ þ C2 ðI3 þ i3 ekt Þ þ C3 ðI1 þ i01 ekt Þ þ C4 ðI3 þ i03 ekt Þ
ðjÞ ðjÞ 0ðjÞ 0ðjÞ
k i03 ekt ¼ C6 ðI1 þ i1 ekt Þ þ C7 ðI3 þ i3 ekt Þ þ C8 ðI1 þ i01 ekt Þ þ C9 ðI3 þ i03 ekt Þ
ðj¼0Þ ðj¼0Þ 0ðj¼0Þ 0ðj¼0Þ
I1 ¼ 0 ; I3 ¼ 0 ; I1 ¼ 0 ; I3 ¼0
ðk þ C3 Þ i01 þ C1 i1 þ C2 i3 þ C4 i03 ¼ 0 ; ðk þ C9 Þ i03 þ C6 i1 þ C7 i3 þ C8 i01 ¼ 0
Remark: It is reader exercise to find system Jacobian matrix and to investigate

stability based on system eigenvalues. The system has four eigenvalues:
k1 ; k2 ; k3 ; k4 :
We define I1 ðt s1 Þ ¼ I1ðiÞ þ i1 ekðts1 Þ ; I3 ðt s3 Þ ¼ I3ðiÞ þ i3 ekðts3 Þ I10 ðt D1 Þ ¼
I1 ðiÞ þ i01 ekðtD1 Þ ; I30 ðt D3 Þ ¼ I30 ðiÞ þ i03 ekðtD3 Þ then we get four delayed dif-
0
ferential equations with respect to coordinates ½I1 ; I3 ; I10 ; I30 arbitrarily small incre-
dI10 dI30
ments of exponential ½i1 ; i3 ; i01 ; i03 ekt . We consider no delay effect on dIdt1 ; dIdt3 ; dt ; dt
and get the following equations:
k i1 ekt ¼ i01 ekðtD1 Þ ; k i3 ekt ¼ i03 ekðtD3 Þ

k i01 ekt ¼ C1 i1 ekðts1 Þ þ C2 i3 ekðts3 Þ þ C3 i01 ekðtD1 Þ þ C4 i03 ekðtD3 Þ
k i03 ekt ¼ C6 i1 ekðts1 Þ þ C7 i3 ekðts3 Þ þ C8 i01 ekðtD1 Þ þ C9 i03 ekðtD3 Þ
k i1 þ i01 ekD1 ¼ 0; k i3 þ i03 ekD3 ¼ 0
C1 i1 eks1 þ C2 i3 eks3 þ ðC3 ekD1 kÞ i01 þ C4 i03 ekD3 ¼ 0
C6 i1 eks1 þ C7 i3 eks3 þ C8 i01 ekD1 þ ðC9 ekD3 kÞ i03 ¼ 0
ðj¼0Þ ðj¼0Þ 0 ðj¼0Þ 0 ðj¼0Þ

In the equilibrium fixed point I1 ¼ 0; I3 ¼ 0; I1 ¼ 0; I3 ¼ 0:
The small increments Jacobian of our active circulator system is as bellow:

0
1
0 1i1
N11 . . . N14 B C
B .. .. .. C B i3 C
@ . . . AB C ¼ 0; N11 ¼ k; N12 ¼ 0; N13 ¼ ekD1 ; N14
@ i01 A
N41 N44
i03
¼ 0; N21 ¼ 0; N22 ¼ k; N23 ¼ 0; N24 ¼ ekD3
N31 ¼ C1 eks1 ; N32 ¼ C2 eks3 ; N33 ¼ C3 ekD1 k; N34 ¼ C4 ekD3 ;

N41 ¼ C6 eks1 ; N42 ¼ C7 eks3
N43 ¼ C8 ekD1 ; N44 ¼ C9 ekD3 k
0 1
N11 . . . N14
B .. .. .. C
AkI ¼@ . . . A; detjA k I j ¼ 0
N41 N44
Dðs1 ; s3 ; D1 ; D3 Þ ¼ k4 k3 ðC3 ekD1 þ C9 ekD3 Þ þ k2 fðC3 C9 C8 C4 Þ

ekðD1 þ D3 Þ C7 ekðD3 þ s3 Þ C1 ekðs1 þ D1 Þ g
þ k fðC1 C9 C6 C4 Þ ekðs1 þ D1 þ D3 Þ ðC2 C8 C7 C3 Þ ekðs3 þ D1 þ D3 Þ g
þ ðC1 C7 C6 C2 Þ ekðs1 þ s3 þ D1 þ D3 Þ
We have three stability cases: s1 ¼ s3 ¼ s & D1 ¼ D3 ¼ 0 or s1 ¼ s3 ¼

they are positive parameters. There are other possible simple stability cases: s1 ¼
s; s3 ¼ 0; D1 ¼ D3 ¼ 0 or s1 ¼ 0; s3 ¼ s ; D1 ¼ D3 ¼ 0 ; s1 ¼ s3 ¼ 0 ; D1 ¼
D ; D3 ¼ 0 or s1 ¼ s3 ¼ 0; D1 ¼ 0 ; D3 ¼ D. We need to get characteristics
equations for all above stability analysis cases. We study the occurrence of any
possible stability switching, resulting from the increase of the value of the time
delays s; D; sD for the general characteristic equation Dðk; s=D=sD Þ. If we choose s
parameter, then Dðk; sÞ ¼ Pn ðk; sÞ þ Qm ðk; sÞ eks . The expression for Pn ðk; sÞ is
P
Pn ðk; sÞ ¼ nk¼0 Pk ðsÞ kk ¼ P0 ðsÞ þ P1 ðsÞ k þ P2 ðsÞ k2 þ P3 ðsÞ k3 þ . . .
P
The expression for Qm ðk; sÞ is Qm ðk; sÞ ¼ m k¼0 qk ðsÞ k ¼ q0 ðsÞ þ q1 ðsÞ
k
k þ q2 ðsÞ k þ . . ..
2
The case we analyze is when there is a delay in I1 and I3 currents only. The delay
is the same for I1 and I3 and equal to s (s1 ¼ s; s3 ¼ s) which describe most of
active circulator parasitic effects. The general characteristic equation D(k, s) is as
follow: Dðs1 ¼ s3 ¼ s; D1 ¼ D3 ¼ 0; kÞ
Dðs1 ¼ s3 ¼ s; D1 ¼ D3 ¼ 0; kÞ ¼ Dðs; kÞ
¼ k4 k3 ðC3 þ C9 Þ þ k2 ðC3 C9 C8 C4 Þ þ fk2 ðC7 þ C1 Þ
þ k ðC1 C9 C6 C4 C2 C8 þ C7 C3 Þ
þ ðC1 C7 C6 C2 Þ eks g eks
Under Taylor series approximation: eks 1 k s the Maclaurin series is a

Taylor series expansion of a eks function about zero (0). We get the following
general characteristic equation D(k, s) under Taylor series approximation:
eks 1 k s.
Dðs1 ¼ s3 ¼ s; D1 ¼ D3 ¼ 0; kÞ ¼ Dðs; kÞ ¼ k4 k3 ðC3 þ C9 Þ

þ k2 ðC3 C9 C8 C4 Þ þ fk2 ðC7 þ C1 Þ
þ k ½C1 C9 C6 C4 C2 C8 þ C7 C3 s ðC1 C7 C6 C2 Þ
þ C1 C7 C6 C2 g eks
Pn
The expression for Pn ðk; sÞ is Pn ðk; sÞ ¼ k¼0 Pk ðsÞ kk
X
n
Pn ðk; sÞ ¼ Pk ðsÞ kk ¼ P0 ðsÞ þ P1 ðsÞ k þ P2 ðsÞ k2 þ P3 ðsÞ k3 þ P4 ðsÞ k4
k¼0
¼ k2 ðC3 C9 C8 C4 Þ k3 ðC3 þ C9 Þ þ k4
P0 ðsÞ ¼ 0; P1 ðsÞ ¼ 0; P2 ðsÞ ¼ C3 C9 C8 C4 ; P3 ðsÞ ¼ ðC3 þ C9 Þ;
P4 ðsÞ ¼ 1
Pm
The expression for Qm ðk; sÞ is Qm ðk; sÞ ¼ k¼0 qk ðsÞ kk ¼ q0 ðsÞ þ q1 ðsÞ k
þ q2 ðsÞ k2 :
X
m
Qm ðk; sÞ ¼ qk ðsÞ kk ¼ k2 ðC7 þ C1 Þ
k¼0
þ k ½C1 C9 C6 C4 C2 C8 þ C7 C3 s ðC1 C7 C6 C2 Þ
þ C1 C7 C6 C2
q0 ðsÞ ¼ C1 C7 C6 C2 ;
q1 ðsÞ ¼ C1 C9 C6 C4 C2 C8 þ C7 C3 s ðC1 C7 C6 C2 Þ;
q2 ðsÞ ¼ ðC7 þ C1 Þ
The homogeneous system for I1 ; I3 ; I10 ; I30 leads to a characteristic equation for the
P
eigenvalue k having the form Pðk; sÞ þ Qðk; sÞ eks ¼ 0; PðkÞ ¼ 4j¼0 aj k j ; QðkÞ ¼
P2
j¼0 cj k and the coefficients faj ðqi ; qk ; sÞ; cj ðqi ; qk ; sÞg 2 R depend on qi ; qk and
j
delay s. qi ; qk are any active circulator’s parameters, other parameters keep as a

constant. a0 ¼ 0; a1 ¼ 0; a2 ¼ C3 C9 C8 C4 ; a3 ¼ ðC3 þ C9 Þ ; a4 ¼ 1 ;
c0 ¼ C1 C7 C6 C2 ; c2 ¼ ðC7 þ C1 Þ c1 ¼ C1 C9 C6 C4 C2 C8 þ C7
C3 s ðC1 C7 C6 C2 Þ.
a0 + c0 6¼ 0 for 8 qi ; qk 2 R þ , i.e. k = 0 is not a of PðkÞ þ QðkÞ eks ¼ 0.
requirements of the analysis [BK] can also be verified in the present case.
If k ¼ i x; x 2 R, then Pði xÞ þ Qði xÞ 6¼ 0. jQðkÞ=PðkÞj is bounded for
jkj ! 1, Re k
0. No roots bifurcation from ∞. FðxÞ ¼ jPði xÞj2 jQði xÞj2
has a finite number of zeros. Indeed, this is a polynomial in x. Each positive root
xðqi ; qk Þ of F(x) = 0 is continuous and differentiable respect to qi ; qk . We assume
that Pn ðk; sÞ and Qm ðk; sÞ can’t have common imaginary roots. That is for any real
number x.
pn ðk ¼ i x; sÞ þ Qm ðk ¼ i x; sÞ 6¼ 0 ;
pn ðk ¼ i x; sÞ ¼ x4 x2 ðC3 C9 C8 C4 Þ i x3 ðC3 þ C9 Þ
Qm ðk ¼ i x; sÞ ¼ x2 ðC7 þ C1 Þ þ i x ½C1 C9 C6 C4 C2 C8 þ C7 C3
s ðC1 C7 C6 C2 Þ þ C1 C7 C6 C2
pn ðk ¼ i x; sÞ þ Qm ðk ¼ i x; sÞ ¼ x4 þ x2 fC7 þ C1 C3 C9 þ C8 C4 g
þ C1 C7 C6 C2 þ i x fC1 C9 C6 C4 C2 C8
þ C7 C3 s ðC1 C7 C6 C2 Þ x2 ðC3 þ C9 Þg 6¼ 0
jPði x; sÞj2 ¼ x8 þ x6 fðC3 þ C9 Þ2 2 ðC3 C9 C8 C4 Þg
þ x4 ðC3 C9 C8 C4 Þ4
For simplicity we define a function: XðCj ; s ; 1 j 9Þ

X ¼ XðCj ; s ; 1 j 9Þ ¼ C1 C9 C6 C4 C2 C8 þ C7 C3
s ðC1 C7 C6 C2 Þ
Qm ðk ¼ i x; sÞ ¼ x2 ðC7 þ C1 Þ þ ðC1 C7 C6 C2 Þ
þ i x XðCj ; s ; 1 j 9Þ
Qm ðk ¼ i x; sÞ ¼ x2 ðC7 þ C1 Þ þ C1 C7 C6 C2 þ i x X
jQði x; sÞj2 ¼ x4 ðC7 þ C1 Þ2 þ x2 fX2 þ 2 ðC7 þ C1 Þ ðC1 C7 C6 C2 Þg
þ ðC1 C7 C6 C2 Þ2
Fðx; sÞ ¼ jPði x; sÞj2 jQði x; sÞj2 ¼ x8 þ x6 fðC3 þ C9 Þ2

2 ðC3 C9 C8 C4 Þg þ x4 fðC3 C9 C8 C4 Þ4 ðC7 þ C1 Þ2 g
x2 fX2 þ 2 ðC7 þ C1 Þ ðC1 C7 C6 C2 Þg ðC1 C7 C6 C2 Þ2
U0 ¼ ðC1 C7 C6 C2 Þ2 ; U2 ¼ fX2 þ 2 ðC7 þ C1 Þ ðC1 C7 C6 C2 Þg

U4 ¼ ðC3 C9 C8 C4 Þ4 ðC7 þ C1 Þ2 ;
U6 ¼ ðC3 þ C9 Þ2 2 ðC3 C9 C8 C4 Þ ; U8 ¼ 1
P4
Hence Fðx; sÞ ¼ 0 implies k¼0 U2k x
2k
¼ 0 and its roots are given
by solving the above polynomial. Furthermore PR ði x; sÞ ¼ x4 x2
ðC3 C9 C8 C4 Þ
PR ði x; sÞ ¼ x2 ðx2 C3 C9 þ C8 C4 Þ ; PI ði x; sÞ ¼ x3 ðC3 þ C9 Þ

QR ði x; sÞ ¼ x2 ðC7 þ C1 Þ þ ðC1 C7 C6 C2 Þ ;
QI ði x; sÞ ¼ x X ðCj ; s ; 1 j 9Þ ¼ x X
Hence
sin hðsÞ ¼
jQði x; sÞj2
cos hðsÞ ¼
jQði x; sÞj2
x3 ½ðx2 C3 C9 þ C8 C4 Þ X þ ðC3 þ C9 Þ fx2 ðC7 þ C1 Þ þ ðC1 C7 C6 C2 Þg

sin hðsÞ ¼
x4 ðC7 þ C1 Þ2 þ x2 fX2 þ 2 ðC7 þ C1 Þ ðC1 C7 C6 C2 Þg þ ðC1 C7 C6 C2 Þ2
x2 ðx2 C3 C9 þ C8 C4 Þ fx2 ðC7 þ C1 Þ þ ðC1 C7 C6 C2 Þg x4 ðC3 þ C9 Þ X

cos hðsÞ ¼
x4 ðC7 þ C1 Þ2 þ x2 fX2 þ 2 ðC7 þ C1 Þ ðC1 C7 C6 C2 Þg þ ðC1 C7 C6 C2 Þ2
P
Which jointly with Fðx; sÞ ¼ 0 ) 4k¼0 U2k x2k ¼ 0 that is a continuous and
differentiable in s based on Lemma 1.1. Hence we use Theorem 1.2. This proves the
Theorem 1.3.
Our active circulator homogeneous system for i1 ; i3 ; i01 ; i03 leads to a characteristic
equation for the eigenvalue k having the form P ðkÞ þ Q ðkÞ eks = 0; First case
s1 ¼ s ; s3 ¼ s ; D1 ¼ D3 ¼ 0. The general characteristic equation D(k, s) is as
follow: Dðs1 ¼ s3 ¼ s; D1 ¼ D3 ¼ 0; kÞ ¼ Dðs; kÞ

þ k2 ðC3 C9 C8 C4 Þ þ fk2 ðC7 þ C1 Þ
þ k ðC1 C9 C6 C4 C2 C8 þ C7 C3 Þ þ ðC1 C7 C6 C2 Þ eks g eks
Under Taylor series approximation: eks 1 k s. The Maclaurin series is a

eks 1 k s.

þ k2 ðC3 C9 C8 C4 Þ þ fk2 ðC7 þ C1 Þ þ k ½C1 C9 C6 C4
C2 C8 þ C7 C3 s ðC1 C7 C6 C2 Þ þ C1 C7 C6 C2 g eks

definition: k ! j ; pk ðsÞ ! aj ; qk ðsÞ ! cj ; n ¼ 4 ; m ¼ 2 ; n [ m.
P
Additionally Pn ðk; sÞ ! PðkÞ ; Qm ðk; sÞ ! QðkÞ then PðkÞ ¼ 4j¼0 aj k j ;
P
QðkÞ ¼ 2j¼0 cj k j
Pk ¼ k2 ðC3 C9 C8 C4 Þ k3 ðC3 þ C9 Þ þ k4
Qk ¼ k2 ðC7 þ C1 Þ þ k ½C1 C9 C6 C4 C2 C8 þ C7 C3
s ðC1 C7 C6 C2 Þ þ C1 C7 C6 C2

tion of s such that a0 þ c0 6¼ 0. In the following “−” denotes complex and conju-
gate. PðkÞ; QðkÞ are analytic functions in k and differentiable in s. The coefficients
faj ðRf ; Lf ; Cf ; Cc ; Rsb ; s; . . .Þ and cj ðRf ; Lf ; Cf ; Cc ; Rsb ; s; . . .Þg 2 R depend on
active circulator system’s Rf ; Lf ; Cf ; Cc ; Rsb ; s; . . . values.
a0 ¼ 0; a1 ¼ 0; a2 ¼ C3 C9 C8 C4 ; a3 ¼ (C3 þ C9 ); a4 ¼ 1
c0 ¼ C1 C7 C6 C2 ; c1 ¼ ½C1 C9 C6 C4 C2 C8 þ C7 C3
s ðC1 C7 C6 C2 Þ; c2 ¼ ðC7 þ C1 Þ

ðRf ; Lf ; Cf ; Cc ; Rsb ; s; . . .Þ will subsequently be omitted from P, Q, aj, cj. The
direct substitution shows that a0 þ c0 6¼ 0; C1 C7 C6 C2 6¼ 0
ðLRLC
1
CR Þ L1R ½L1 C2 CR 1
½C11 þ L1 C1 ðC11 L1LCR R Þ
1
LR 6¼ 0
fLf þ L1 ð1 LLR1 Þg fLf þ L1 ð1 LLR1 Þg
ðLRLC
1
CR Þ L1R ½L1 C2 CR L1R ½C11 þ L1 C1 ðC11 L1LCR R Þ
1
6¼ 0
fLf þ L1 ð1 LLR1 Þg
8 Rf ; Lf ; Cf ; Cc ; Rsb ; s; . . . 2 R þ i.e. k ¼ 0 is not a root of the characteristic

equation. Furthermore PðkÞ; QðkÞ are analytic function of k for which the following
requirements of the analysis (see Kuang [5], Sect. 3.4) can also be verified in the
present case [6, 7].
ðRf ; Lf ; Cf ; Cc ; Rsb ; s; . . .Þ domain of interest.
(b) jQ(kÞ=PðkÞj is bounded for jkj ! 1, Re k
0. No roots bifurcation from 1.
Indeed, in the limit
fk2 ðC7 þ C1 Þ þ k ½C1 C9 C6 C4 C2 C8 þ C7 C3

Q(kÞ s ðC1 C7 C6 C2 Þ þ C1 C7 C6 C2 g
j j¼j j
PðkÞ k ðC3 C9 C8 C4 Þ k3 ðC3 þ C9 Þ þ k4
2
Fðx; sÞ ¼ jPði x; sÞj2 jQði x; sÞj2 ¼ x8 þ x6 fðC3 þ C9 Þ2

2 ðC3 C9 C8 C4 Þg þ x4 fðC3 C9 C8 C4 Þ4
ðC7 þ C1 Þ2 g x2 fX2 þ 2 ðC7 þ C1 Þ ðC1 C7 C6 C2 Þg
ðC1 C7 C6 C2 Þ2
Has at most a finite number of zeros. Indeed, this is a polynomial in x (degree

in x8 ).
(d) Each positive root xðRf ; Lf ; Cf ; Cc ; Rsb ; s; . . .Þ of F(xÞ¼ 0 is continuous and

differentiable with Respect to Rf ; Lf ; Cf ; Cc ; Rsb ; s; . . .. This condition can only
be assessed numerically.
Pði xÞ ¼ Pði xÞ, and Qði xÞ ¼ Qði xÞ thus, k ¼ i x, x [ 0 maybe on
eigenvalue of characteristic equation. The analysis consists in identifying the roots
of the characteristic equation situated on the imaginary axis of the complex k—
plane, whereby increasing the parameters Rf ; Lf ; Cf ; Cc ; Rsb ; s; . . ., Rek may, at
the crossing, change its sign from (−) to (+), i.e. from a stable focus
ð0Þ ð0Þ 0 ð0Þ 0 ð0Þ
E ð0Þ ðI1 ; I3 ; I1 ; I3 Þ ¼ ð0; 0; 0; 0Þ to an unstable one, or vice versa. This feature
may be further assessed by examining the sign of the partial derivatives with respect
Re kÞ
to Rf ; Lf ; Cf ; Cc ; Rsb ; s; . . . and gate antenna parameters. ^1 ðRf Þ ¼ ð@ @Rf k¼ix ;
Lf ; Cf ; Cc ; Rsb ; s; . . . ¼ const
@ Re k
^1 ðLf Þ ¼ ð Þ ; Rf ; Cf ; Cc ; Rsb ; s; . . . ¼ const
@Lf k¼ix
@Re k
^1 ðCf Þ ¼ ð Þ ; Rf ; Lf ; Cc ; Rsb ; s; . . . ¼ const
@Cf k¼ix
@ Re k
^1 ðCc Þ ¼ ð Þ ; Rf ; Lf ; Cf ; Rsb ; s; . . . ¼ const
@Cc k¼ix
@ Re k
^1 ðRsb Þ ¼ ð Þ ; Rf ; Lf ; Cf ; Cc ; s; . . . ¼ const
@Rsb k¼ix
@ Re k
^1 ðsÞ ¼ ð Þ ; Rf ; Lf ; Cf ; Cc ; Rsb ; . . . ¼ const ; x 2 R þ :
@s k¼ix
In the case s1 ¼ s3 ¼ s & D1 ¼ D3 ¼ 0 we get the following results: for sim-

plicity we define a function: XðCj ; s; 1 j 9Þ
X ¼ XðCj ; s; 1 j 9Þ ¼ C1 C9 C6 C4 C2 C8 þ C7 C3
s ðC1 C7 C6 C2 Þ
PR ði x; sÞ ¼ x2 ðx2 C3 C9 þ C8 C4 Þ; PI ði x; sÞ ¼ x3 ðC3 þ C9 Þ
QR ði x; sÞ ¼ x2 ðC7 þ C1 Þ þ ðC1 C7 C6 C2 Þ; QI ði x; sÞ
¼ x X ðCj ; s; 1 j 9Þ ¼ x X
U0 ¼ ðC1 C7 C6 C2 Þ2 ; U2 ¼ fX2 þ 2 ðC7 þ C1 Þ ðC1 C7 C6 C2 Þg
U4 ¼ ðC3 C9 C8 C4 Þ4 ðC7 þ C1 Þ2 ; U6 ¼ ðC3 þ C9 Þ2
2 ðC3 C9 C8 C4 Þ ; U8 ¼ 1
X
4
¼ U0 þ U2 x2 þ U4 x4 þ U6 x6 þ U8 x8 ¼ U2k x2k
k¼0
P4
Hence Fðx; sÞ ¼ 0 implies k¼0 U2k x
2k
¼ 0. When writing PðkÞ ¼
PR ðkÞ þ i PI ðkÞ and QðkÞ ¼ QR ðkÞ þ i QI ðkÞ, and inserting k ¼ i x into active
circulator system’s characteristic equation, x must satisfy the following:
jQði xÞj2
jQði xÞj2
where jQði xÞj2 6¼ 0 in view of requirement (a) above, and ðg; hÞ 2 R.

that I R þ 0 is the set where xðsÞ is a positive root of FðxÞ and for s 62 I , xðsÞ is
hðsÞ 2 ½0; 2 p as the solution of sin hðsÞ ¼ . . .; cos hðsÞ ¼ . . .
sin hðsÞ ¼
jQði xÞj2
cos hðsÞ ¼
jQði xÞj2
And the relation between the argument hðsÞ and xðsÞ s for xðsÞ s must be
given by sn ðsÞ ¼ hðsÞxðsÞ
þ n2p
Sn ðsÞ ¼ s sn ðsÞ; s 2 I; n 2 N0 that is a continuous and differentiable in s. In the
following, the subscripts k; x; Rf ; Lf ; Cf ; Cc ; Rsb ; . . . indicate the corresponding
partial derivatives. Let us first concentrate on ^ðxÞ, remember in kðRf ; Lf ; Cf ;
Cc ; Rsb ; . . .Þ and xðRf ; Lf ; Cf ; Cc ; Rsb ; . . .Þ, and keeping all parameters except one
(x) and s. The derivation closely follows that in reference [BK]. Differentiating
active circulator characteristic equation PðkÞ þ QðkÞ eks = 0 with respect to
specific parameter (x), and inverting the derivative, for convenience, one calculates.
Remark: x ¼ Rf ; Lf ; Cf ; Cc ; Rsb ; . . .; etc:;

ð Þ ¼
where Pk ¼ @P
@k ; . . . etc., Substituting k ¼ i x, and bearing i Pði xÞ ¼ Pði xÞ,
Qði xÞ ¼ Qði xÞ then i Pk ði xÞ ¼ Px ði xÞ; i Qk ði xÞ ¼ Qx ði xÞ and
that on the surface jPði xÞj2 ¼ jQði xÞj2 , one obtains


Qx ¼ QRx þ i QIx
P2 ¼ P2R þ P2I . When (x) can be any active circulator parameters R1, C1, And
time delay s etc. Where for convenience, we have dropped the arguments ði; x; xÞ,
and where Fx ¼ 2 ½ðPRx PR þ PIx PI Þ ðQRx QR þ QIx QI Þ.
Fx ¼ 2 ½ðPRx PR þ PIx PI Þ ðQRx QR þ QIx QI Þ; xx ¼ Fx =Fx :
We define U and V:

We choose our specific parameter as time delay x = s
PR ¼ x2 ðx2 C3 C9 þ C8 C4 Þ ; PI ¼ x3 ðC3 þ C9 Þ ;

QR ¼ x2 ðC7 þ C1 Þ þ ðC1 C7 C6 C2 Þ
QI ¼ x X ðCj ; s; 1 j 9Þ ¼ x X ðsÞ ; PRx ¼ 2 x ½2 x2 C3 C9 þ C8 C4
PIx ¼ 3 x2 ðC3 þ C9 Þ ; QRx ¼ 2 x ðC7 þ C1 Þ ; QIx ¼ X ; PRs ¼ 0 ;
PIs ¼ 0 ; QRs ¼ 0; xs ¼ Fs =Fx
@X
QIs ¼ x ¼ x ðC6 C2 C1 C7 Þ
@s
PRx PR ¼ 2 x3 ð2 x2 C3 C9 þ C8 C4 Þ ðx2 C3 C9 þ C8 C4 Þ ;
PIx PI ¼ 3 x5 ðC3 þ C9 Þ2
QRx QR ¼ 2 x ðC7 þ C1 Þ fx2 ðC7 þ C1 Þ þ ðC1 C7 C6 C2 Þg ;
QIx QI ¼ x X2 ðsÞ
Fs ¼ 2 ½ðPRs PR þ PIs PI Þ ðQRs QR þ QIs QI Þ ;
Fs ¼ 2 QIs QI ¼ 2 x2 ðC6 C2 C1 C7 Þ XðsÞ
PR PIx ¼ 3 x4 ðx2 C3 C9 þ C8 C4 Þ ðC3 þ C9 Þ ;
PI PRx ¼ 2 x4 ðC3 þ C9 Þ ½2 x2 C3 C9 þ C8 C4
QR QIx ¼ fx2 ðC7 þ C1 Þ þ ðC1 C7 C6 C2 Þg XðsÞ ;
QI QRx ¼ 2 x2 XðsÞ ðC7 þ C1 Þ
V ¼ ðPR PIs PI PRs Þ ðQR QIs QI QRs Þ
V ¼ QR QIs ¼ fx2 ðC7 þ C1 Þ þ ðC1 C7 C6 C2 Þg
x ðC6 C2 C1 C7 Þ ; Fðx; sÞ ¼ 0:
Differentiating with respect to s and we get Fx @x @x

@s þ Fs ¼ 0 ; s 2 I ) @s ¼
FFxs
@Re k @x Fs
^1 ðsÞ ¼ ð Þ ; ¼ xs ¼ ;
@s k¼ix @s Fx
2 ½U þ s jPj2 þ i Fx
^1 ðsÞ ¼ Ref g
Fs þ i 2 ½V þ x jPj2
@Rek
@s k¼ix
@x U @x
@s þ V
@s jPj2

ð0Þ ð0Þ 0 ð0Þ 0 ð0Þ
active circulator system, about the equilibrium point Eð0Þ ðI0 ; I3 ; I0 ; I3 Þ ¼
ð0; 0; 0; 0Þ as a result of a variation of delay parameter s. The analysis consists in
identifying the roots of our system characteristic equation situated on the imaginary
axis of the complex k-plane where by increasing the delay parameter s, Re k may at
the crossing, changes its sign from − to +, i.e. from a stable focus E(*) to an unstable
the partial derivatives with respect to s,
@ Re k
^1 ðsÞ ¼ ð Þ ; Rf ; Lf ; Cf ; Cc ; Rsb ; . . . ¼ const ; x 2 R þ :
@s k¼ix
For our stability switching analysis, we choose typical active circulator param-
eter values: Rf = 110 Ohm, Lf = 1.4 nH, Cf = 5 pF, Cc = 10 pF, Rsb = 115 Ohm,
MESFET (F20-FET-4x75), L1 = 5 nH, C1 = 5 pF, La = 1.6 nH, Ra = 500 Ohm,
Lt = 7 nH, LR = 15 nH. LR ¼ Lf þ La þ Lt þ L1 ¼ 1:4 nH þ 1:6 nH þ 7 nH þ
5 nH ¼ 15 nH:
1 1 1 1 1 1
CR ¼ þ þ ¼ þ þ ¼ 5 1011 ;
Cf Cc C1 5 1012 10 1012 5 1012
L1
C0 ¼ 2:11 108 ; C1 ¼ C0 ð CR Þ ¼ 9:14 1019
LR C1
1 L1 CR
C2 ¼ C0 ð Þ ¼ 0:703 1019 ; C3 ¼ C0 Rf ¼ 232:1 108 ;
C1 LR
L1 C0
C4 ¼ C0 L1 ðRf þ Ra Þ ¼ 643:55; C5 ¼ ¼ 0:7 108
LR
1 1 1
C6 ¼ ½ þ L1 C1 ¼ 0:171 1020 ; C7 ¼ ½L1 C2 CR ¼ 3 1019 ;
LR C1 LR
L1 C3 1
C8 ¼ ¼ 77:36 108 ; C9 ¼ ½L1 C4 ðRf þ Ra Þ ’ 40:6 109
LR LR
Then we get the expression for Fðx; sÞ for an active circulator parameter’s
value. We find those x; s values which fulfill Fðx; sÞ ¼ 0. We ignore negative,
complex, and imaginary values of x for specific s values. s 2 ½0:001::10. And we
can be express by 3D function Fðx; sÞ ¼ 0. We plot the stability switch diagram
based on different delay values of our active circulator system.
@ Re k 2 ½U þ s jPj2 þ i Fx
@ Re k 2 fFx ðV þ x P2 Þ Fs ðU þ s P2 Þg
s ¼ xh þþ ðsÞ
x þ ðsÞ if only x þ is feasible. Additionally When all active circulator’s parameters
are known and the stability switch due to various time delay values s is described in
the following expression:
signf^1 ðsÞg ¼ signfFx ðxðsÞ; sÞg

signfs xs ðxðsÞÞ þ xðsÞ þ g
jPðxðsÞÞj2
Remark: We know Fðx; sÞ ¼ 0 implies its roots xi ðsÞ and finding those delays
values s which xi is feasible. There are s values, which xi are complex or imag-
inary numbered, then unable to analyze stability [6, 7].
complex, and imaginary values of x for specific s values. s 2 ½0:001::10 and we
can be express by 3D function Fðx; sÞ ¼ 0. We define new MATLAB script
parameters: s!Tau, Ci!Gi (i=0..9), X!Omega, Uj!Phij. Running MATLAB
script for s values (s 2 ½0:001::10) gives the following results.
MATLAB script: Tau=0.1;G0=2.11e8;G1=-9.14e19; G2=0.703e19;
G3=-232.1e8;G4=-643.55;G5=0.7e8;G6=-0.171e20;G7=-3e19;G8=-77.36e8;G9=-
40.6e9; Omega=G1*G9-G6*G4-G2*G8+G7*G3-Tau*(G1*G7-G6*G2);Phi0=-
(G1*G7-G6*G2)^2; Phi2=-(Omega^2+2*(G7+G1)*(G1*G7 G6*G2)); Phi4=
(G3*G9-G8*G4)^4-(G7+G1)^2; Phi6=(G3+G9)^2-2*(G3*G9-G8*G4);Phi8=1; p=
[Phi8 0 Phi6 0 Phi4 0 Phi2 0 Phi0];r=roots(p).
Results: (Table 2.2).
We plot 3D function Fðx; sÞ ¼ 0. s:0!10; x:0!1e20. We define additional
MATLAB script parameters x!w, s!t (Fig. 2.8).
We get two possible real values for x which fulfil Fðx; sÞ ¼ 0 ; Fðx ¼ 0 or x ¼
1:0e þ 020; sÞ ¼ 0 s 2 ½0:001::10. Next is to find those x, s values which fulfil
sin hðsÞ ¼ . . .
Table 2.2 Active circulator s s = 1; s = 10 s = [0…0.1]

roots xi ðsÞ
x1 1.0e+020 1.0e+020
x2 −6.6468 + 6.6468i −6.6468 + 6.6468i
x3 −6.6468 − 6.6468i −6.6468 − 6.6468i
x4 6.6468 + 6.6468i 6.6468 + 6.6468i
x5 6.6468 − 6.6468i 6.6468 − 6.6468i
x6 −0.0000 0
x7 0.0000 0
x8 0.0000 + 0.0000i 0
x9 0.0000 − 0.0000i 0
Fig. 2.8 Active circulator F

(x, s) function
PR QI þ PI QR
sinðx sÞ ¼ and cos hðsÞ ¼ . . .
jQj2
ðPR QR þ PI QI Þ
cosðx sÞ ¼ ; jQj2 ¼ Q2R þ Q2I
jQj2
Case I: x ¼ 0 then PR ¼ 0 ; PI ¼ 0 ; QR ¼ C1 C7 C6 C2 ; QI ¼ 0; sinðx sÞ

¼ . . . fulfil and cosðx sÞ ¼ . . . Can’t fulfil since cosðx sÞjx¼0 6¼ 0.
Case II: x = 1.0e+020 which can fulfil expressions sinðx sÞ ¼ . . . and

cosðx sÞ ¼ . . .. Finally, we plot the stability switch diagram based on different
delay values of our Active circulator system (x = 1.0e20). PR = x2
(x2−9.4233e20), PI = x36.381e10, QR = −x21.214e20+2.8622e39, QI = x
Omega. QIs ¼ x 2:86e39 ; V ¼ QR QIs ; V ¼ QR QIs ¼ x3 3:473e59
PIx ¼ x2 19:143e10 ; QIx ¼ Omega ; PRx ¼ 2 x ½2 x2 9:4233e20 ;

QRx ¼ x 2:428e20
PRx PR ¼ 2 x3 ð2 x2 9:423e20Þ ðx2 9:423e20Þ ;
U ¼ x4 ðx2 6:377e10 þ 6:0119e31Þ x2 Omega 1:214e20
PIx PI ¼ x5 122:15e20 ; QRx QR ¼ x3 2:947e40 ;
QIx QI ¼ x ½Omega2 ; Fs ¼ 2 QIs QI ¼ 2 QIs QI
Fs ¼ 2 QIs QI ¼ 5:72e39 x2 Omega: We plot the function:

gðsÞ ¼ ^1 ðsÞ ¼ ð@ Re k
@s Þk¼ix
@ Re k 2 fFx ðV þ x P2 Þ Fs ðU þ s P2 Þg
gðsÞ ¼ ^1 ðsÞ ¼ ð Þk¼ix ¼
@ Re k
sign ½gðsÞ ¼ sign½^1 ðsÞ ¼ sign½ð Þ
@s k¼ix
2 fFx ðV þ x P Þ Fs ðU þ s P2 Þg
2
¼ sign½
Since Fs2 þ 4 ðV þ x P2 Þ2 [ 0 then sign½^1 ðsÞ ¼ signfFx ðV þ x P2 Þ

Fs ðU þ s P2 Þg
Fs
sign½^1 ðsÞ ¼ signf½Fx ½ðV þ x P2 Þ ðU þ s P2 Þg ;
Fx
Fs @x @F=@x
xs ¼ ; xs ¼ ð Þ1 ¼
Fx @s @F=@s
sign ½^1 ðsÞ ¼ signf½Fx ½V þ xs U þ x P2 þ xs s P2 g ;
1 V þ xs U
sign ½^1 ðsÞ ¼ signf½Fx ½ 2 ½ þ x þ xs sg
P P2
1 V þ xs U
sign ½ 2 [ 0 ) sign ½^1 ðsÞ ¼ signf½Fx ½ þ x þ xs sg
P P2
V þ xs U
sign ½^1 ðsÞ ¼ sign½Fx sign ½ þ x þ xs s;
P2
We check the sign of ^1 ðsÞ according the following rule (Table 2.3).
Table 2.3 Active circulator sign ½Fx sign½V þPx2 s U þ x þ xs s sign ½^1 ðsÞ
stability switching criteria
± ± +
± –
x = 1.0e + 020 and s 2 ½0:001::10. Since it is a very complex function, we rec-
ommend to solve it numerically rather than analytic. We plot the stability switch
diagram based on different delay values of our active circulator system.
We consider Active circulator which connects in a configuration of Reflection
Type Phase Shifter (RTPS) circuit. Due to the parasitic effect, there is a delay in
time for current which flow in and out Active circulator ports. This delay causes to
stability switching for our Active circulator system. We draw our Active circulator
(RTPS) equivalent circuit and get system differential equations. Our variables are
first and third ports currents and currents derivative. Our system dynamic behavior
is dependent on circuit overall parameters and parasitic delay in time. We keep all
circuit parameters fix and change, parasitic delay over various values
s 2 ½0:001::10. Our analysis results extend that of in the way that it deals with
stability switching for different delay values. This implies that our system behavior
of the circuit cannot inspect by short analysis and we must study the full system.
Several very important issues and possibilities were left out of the present analysis.
One possibility is the stability switching by circuit parameters. Every circuit’s
parameter variation can change our system dynamic and stability behavior. This
case can be solved by the same methods combined with alternative and more
technical hypotheses. Moreover, numerical simulations for the active circulator
model studied in references suggest that this result can be extended to enhance
models with more general functions. Still another extension of our results would be
to treat the case of delayed Active circulator’s port currents derivative in time
dI1 ðtD1 Þ dI3 ðtD3 Þ
dt ; dt ; D1 [ 0 ; D3 [ 0 [5, 6]. It would be extremely desirable to con-
firm these cases by mathematical proofs. Active circulator transmission lines are
characterized by parasitic effects which can influence active circulator system sta-
bility in time. There are four main active circulator variables which are affected by
transmission lines parasitic effects, first and third branch currents and currents
derivatives respectively. Each active circulator currents variable under transmission
line parasitic effects is characterized by time delay respectively. The two time
delays are not the same, but can be categorized to some sub cases due to inter-
ferences behavior. The first case we analyze is when there is delay in active cir-
culator first and third branches current and no delay in active circulator first and
third branches current derivative. The second case we analyze is when there is delay
in active circulator first and third branches current derivative and no delay in active
circulator first and third branches current [6, 7]. The third case we analyze is when
there is delay in active circulator first and third branches current and also delay in
active circulator first and third branches current derivative.
(s1 ¼ s3 ¼ D1 ¼ D3 ¼ sD ) [6, 7]. For simplicity of our analysis we consider in the

third case all delays are the same (there is a difference but it is neglected in our
analysis). In each case we derive the related characteristic equation. The charac-
teristic equation is dependent on active circulator overall parameters and interfer-
ences time delay. Upon mathematics manipulation and [BK] theorems and
definitions we derive the expression which gives us a clear picture on active cir-
culator stability map. The stability map gives all possible options for stability
segments, each segment belongs to different time delay value segment. Active
circulator’s stability analysis can be influenced either by system overall parameter
values. We left this analysis and do not discuss it in the current chapter [12].
Lemma 1.1 Assume that xðsÞ is a positive and real root of Fðx; sÞ ¼ 0.
Defined for s 2 I, which is continuous and differentiable. Assume further that if
s 2 I; I R þ 0 , and at some s 2 I, Sn ðs Þ ¼ 0. For some n 2 N0 then a pair of
d Re k dSn ðsÞ
dðs Þ ¼ signf j
g ¼ signfFx ðxðs Þ; s Þg signf j g
ds k¼ixðs Þ ds s¼s
Theorem 1.3 The characteristic equation has a pair of simple and conjugate pure
imaginary roots k ¼ xðs Þ; xðs Þ real at s 2 I if Sn ðs Þ ¼ s sn ðs Þ ¼ 0 for
some n 2 N0 . If xðs Þ ¼ x þ ðs Þ, this pair of simple conjugate pure imaginary
roots crosses the imaginary axis from left to right if d þ ðs Þ [ 0 and crosses the
imaginary axis from right to left if d þ ðs Þ\ 0 where d þ ðs Þ ¼ sign
dSn ðsÞ
fd Re k
ds jk¼ix þ ðs Þ g ¼ signf ds js¼s g. If xðs Þ ¼ x ðs Þ, this pair of simple con-
jugate pure imaginary roots crosses the imaginary axis from left to right if
d ðs Þ [ 0 and crosses the imaginary axis from right to left If d ðs Þ\0 where
dSn ðsÞ
d ðs Þ ¼ signfd Re k
ds jk¼ix ðs Þ g ¼ signf ds js¼s g.
If x ðs Þ ¼ x ðs Þ then Dðs Þ ¼ 0 and signfdRek j
þ ds g ¼ 0, the same is
k¼ixðs Þ
true when S0n ðs Þ ¼ 0. The following result can be useful in identifying values of s
Where stability switches happened.
2.3 Cylindrical RF Network Antennas for Coupled

Plasma Sources Copper Legs Delayed in Time System
Stability Analysis
In this subchapter, Very Critical and useful subject is discussed: cylindrical (closed)
RF network antennas for coupled plasma sources copper legs delayed in time. The
resonant RF networks can be arranged to form large-area or large-volume plasma
sources with properties similar to Inductive Coupled Plasma (ICP) devices. There
are medical applications of Birdcage coils and closed and open configurations of the
antenna for plasma production are possible and can be analyzed by using mathe-
matical formulation. There are systems of an open network antenna as a large-area
planar plasma source and of a closed network antenna as a cylindrical plasma
source. Both are composed of similar electrical meshes. Operation at different
normal modes shows the capability of this antenna type of large-volume plasma
applications [86].
An important issue of proper antenna operation is the location of the RF feeding
and grounding connections on the antenna. There are a large number of different RF
antenna arrangements possible in view of the geometry and RF operation and of
plasma obtained. In our analysis, we investigated only cylindrical RF antenna
which built following a high-pass Birdcage coil. The antenna is mounted outside a
glass tube. The RF antenna consists of 16 copper legs (Fig. 2.9), equally spaced
interconnected with capacitors, each copper leg current has parasitic time delayed
(s1–1 … s1–16). We consider for simplicity that all copper legs parasitic time delayed
are equal (s1–1 = s1–2 = = s1–16) and the voltages on delay units (Ve ) are
neglected Ve ! e. There is a delay in each Copper leg current
I1 ðt s11 Þ; . . .; I16 ðt s116 Þ. We consider all interconnected capacitor values are
the same (C) and all antenna elements inductance values are the same (L). CA1 ¼
CA2 ¼ ¼ CA16 ¼ C ; CB1 ¼ CB2 ¼ ¼ CB16 ¼ CL1 ¼ L2 ¼ ¼ L16 ¼ L ;
Fig. 2.9 Schematic of the

16-leg cylindrical (Birdcage)
RF network antenna (closed) Copper
legs
delays τk
(k=1..16)
τk
2.3 Cylindrical RF Network Antennas for Coupled Plasma Sources … 197
Fig. 2.10 Upper view of

16-leg cylindrical RF antenna
IL1 ¼ I1 ; IL2 ¼ I2 ; . . .; IL16 ¼ I16 : We choose first case: antenna network is fed by
the transmitter unit (S1 = OFF, no direct RF feeding). The upper view of 16-leg
cylindrical RF antenna network described in Fig. 2.10.
The lower view of 16-leg cylindrical RF antenna network described in Fig. 2.11.
Cylindrical RF network antenna system can represent as round strip of capacitors
and inductors (Figs. 2.12 and 2.13). The schematic contains RF feeding signal, S1
switch (S1 = ON for direct RF signal feeding, S1 = OFF for RF signal transmitter
feeding). The upper network connecting nodes are A1, A2,…,A16 and the lower
network connecting nodes are B1, B2,…,B16. Antenna copper leg current parasitic
delays are represented by delay units Tau1–1…Tau1–16 (s11 ; . . .; s116 ). Rp is the
parasitic resistance of RF feeding point (A1). The upper system spaced capacitors
are CA1,…,CA16 and the lower system spaced capacitors are CB1,…,CB16.
Fig. 2.11 Lower view of

16-leg cylindrical RF antenna
Fig. 2.12 16-leg cylindrical RF antenna strip (feeding side)
Fig. 2.13 16-leg cylindrical RF antenna strip (ground side)
d d
ICA1 ¼ CA1 ðVA1 VA2 Þ; ICA2 ¼ CA2 ðVA2 VA3 Þ ;
dt dt
d d
ICA3 ¼ CA3 ðVA3 VA4 Þ; . . .; ICA7 ¼ CA7 ðVA7 VA8 Þ
dt dt
dVA8 d
ICA8 ¼ CA8 ; ICAk ¼ CAk ðVAk VAðk þ 1Þ Þ ; k ¼ 1; . . .; 7 ;
dt dt
d d
ICA16 ¼ CA16 ðVA1 VA16 Þ; ICA15 ¼ CA15 ðVA16 VA15 Þ
dt dt
d d
ICA14 ¼ CA14 ðVA15 VA14 Þ; . . .; ICA10 ¼ CA10 ðVA11 VA10 Þ ;
dt dt
dVA10 d
ICA9 ¼ CA9 ; ICAl ¼ CAl ðVAðl þ 1Þ VAl Þ ; l ¼ 10; . . .; 15
dt dt
d d
ICB1 ¼ CB1 ðVB1 VB2 Þ; ICB2 ¼ CB2 ðVB2 VB3 Þ ; k ¼ 1; . . .; 8 ; ; . . .;
dt dt
d d
ICB8 ¼ CB8 ðVB2 VB3 Þ ; ICBk ¼ CBk ðVBk VBðk þ 1Þ Þ
dt dt
d d
ICB16 ¼ CB16 ðVB1 VB16 Þ; ICB15 ¼ CB15 ðVB16 VB15 Þ ; ; . . .;
dt dt
d
ICB9 ¼ CB9 ðVB10 VB9 Þ
dt
d
ICBl ¼ CBl ðVBðl þ 1Þ VBl Þ ; l ¼ 15; . . .; 9
dt
dIL1 dIL2 dIL3
VA1 VB1 ¼ L1 ; VA2 VB2 ¼ L2 ; VA3 VB3 ¼ L3 ; ; . . .;
dt dt dt
dIL8 dIL9
VA8 VB8 ¼ L8 ; VB9 ¼ L9
dt dt
dIL10 dIL16
VA10 VB10 ¼ L10 ; . . .; VA16 VB16 ¼ L16 ;
dt dt
dILm
VAm VBm ¼ Lm ; m ¼ 1; . . .; 16 ; m 6¼ 9
dt
dIL9
VB9 ¼ L9 ; VA9 ¼ 0 ; A9 ground
dt
IRP ¼ ICA16 þ ICA1 þ IL1 ; ICA1 ¼ ICA2 þ IL2 ;

ICA2 ¼ ICA3 þ IL3 ; . . .; ICA7 ¼ ICA8 þ IL8
ICAl ¼ ICAðl þ 1Þ þ ILðl þ 1Þ ; l ¼ 1; . . .; 7
ICA16 ¼ ICA15 þ IL16 ; ICA15 ¼ ICA14 þ IL15 ;
ICA14 ¼ ICA13 þ IL14 ; . . .; ICA10 ¼ ICA9 þ IL10
ICAk ¼ ICAðk1Þ þ ILk ; k ¼ 16; . . .; 10
IL1 ¼ ICB1 þ ICB16 ; ICB2 ¼ ICB1 þ IL2 ; ICB3 ¼ ICB2 þ IL3 ; ICB4 ¼ ICB3 þ IL4 ; . . .;
ICB8 ¼ ICB7 þ IL8 ; IL9 ¼ ICB8 þ ICB9 ; ICBm ¼ ICBðm1Þ þ ILm ; m ¼ 2; . . .; 8
ICB15 ¼ ICB16 þ IL16 ; ICB14 ¼ ICB15 þ IL15 ; ICB13 ¼ ICB14 þ IL14 ;
ICB12 ¼ ICB13 þ IL13 ; . . .; ICB9 ¼ ICB10 þ IL10
ICBn ¼ ICBðn þ 1Þ þ ILðn þ 1Þ ; n ¼ 15; . . .; 9:
Upon mathematic manipulation we get the following expressions:

ICA1 ICB1 d 2 IL1 d 2 IL2

¼ L1 2 L2 2 ; CA1 ¼ CB1 ¼ C ; L1 ¼ L2 ¼ L ;
CA1 CB1 dt dt
1 d 2 IL1 d 2 IL2
ðICA1 ICB1 Þ ¼ 2
LC dt2 dt
2
ICA2 ICB2 d IL2 d 2 IL3
¼ L2 2 L3 2 ; CA2 ¼ CB2 ¼ C ; L2 ¼ L3 ¼ L ;
CA2 CB2 dt dt
1 d 2 IL2 d 2 IL3
ðICA2 ICB2 Þ ¼ 2 . . .;
LC dt2 dt
ICA7 ICB7 d 2 IL7 d 2 IL8
¼ L7 2 L8 2 ; L7 ¼ L8 ¼ L ;
CA7 CB7 dt dt
2 2
1 d IL7 d IL8
ðICA7 ICB7 Þ ¼ 2
LC dt2 dt
CA7 ¼ CB7 ¼ C ; L1 ¼ L2 ¼ ¼ L16 ¼ L ; k ¼ 1; . . .; 7 ;
1 d 2 ILk d 2 ILðk þ 1Þ
ðICAk ICBk Þ ¼ 2 ; k ¼ 1; . . .; 7
LC dt dt2
1 d 2 IL11 d 2 IL10 1
ðICA10 ICB10 Þ ¼ ; CA10 ¼ CB10 ¼ C ; ðICA15 ICB15 Þ
LC dt2 dt2 LC
d 2 IL16 d 2 IL15
¼ ; CA15 ¼ CB15 ¼ C
dt2 dt2
1 d 2 ILðm þ 1Þ d 2 ILm
ðICAm ICBm Þ ¼ ; m ¼ 10; . . .; 15 ; CA8 ¼ CB8 ¼ C ;
LC dt2 dt2
1 d 2 IL8 d 2 IL9
VA9 ¼ 0 ; ðICA8 ICB8 Þ ¼ þ
LC dt2 dt2
1 d 2 IL9 d 2 IL10
CA9 ¼ CB9 ¼ C ; VA9 ¼ 0 ; ðICA9 ICB9 Þ ¼ þ ;
LC dt2 dt2
1 d 2 IL1 d 2 IL16
CA16 ¼ CB16 ¼ C; VA9 ¼ 0 ; ðICA16 ICB16 Þ ¼
LC dt2 dt2
IR P ¼ ICA16 þ ICA1 þ IL1 ; ICA1 ¼ ICA2 þ IL2 ; ICA2 ¼ ICA3 þ IL3 ;
ICA3 ¼ ICA4 þ IL4 ; ICA4 ¼ ICA5 þ IL5 ; ICA5 ¼ ICA6 þ IL6
ICA6 ¼ ICA7 þ IL7 ; ICA7 ¼ ICA8 þ IL8 ; ICA10 ¼ ICA9 þ IL10 ; ICA11 ¼ ICA10 þ IL11 ;
ICA12 ¼ ICA11 þ IL12 ; ICA13 ¼ ICA12 þ IL13
ICA14 ¼ ICA13 þ IL14 ; ICA15 ¼ ICA14 þ IL15 ; ICA16 ¼ ICA15 þ IL16 ; IL1 ¼ ICB1 þ ICB16 ;
IL9 ¼ ICB8 þ ICB9 ; ICB2 ¼ ICB1 þ IL2
ICB3 ¼ ICB2 þ IL3 ; ICB4 ¼ ICB3 þ IL4 ; ICB5 ¼ ICB4 þ IL5 ; ICB6 ¼ ICB5 þ IL6 ;
ICB7 ¼ ICB6 þ IL7 ; ICB8 ¼ ICB7 þ IL8
ICB9 ¼ ICB10 þ IL10 ; ICB10 ¼ ICB11 þ IL11 ; ICB11 ¼ ICB12 þ IL12 ; ICB12 ¼ ICB13 þ IL13 ;
ICB13 ¼ ICB14 þ IL14 ; ICB14 ¼ ICB15 þ IL15
ICB15 ¼ ICB16 þ IL16 : S1 is OFF for RF signal transmitter feeding:
IRP ¼ 0 ) ICA16 þ ICA1 þ IL1 ¼ 0
X
8 X
8 X
8
ICA1 ¼ ICA8 þ ILk ; ICA2 ¼ ICA8 þ ILk ; ICA3 ¼ ICA8 þ ILk ;
k¼2 k¼3 k¼4
X
8 X
8 X
8
ICA4 ¼ ICA8 þ ILk ; ICA5 ¼ ICA8 þ ILk ; ICA6 ¼ ICA8 þ ILk
k¼5 k¼6 k¼7
X
16 X
15
ICA7 ¼ ICA8 þ IL8 ; ICA16 ¼ ICA9 þ ILk ; ICA15 ¼ ICA9 þ ILk ;
k¼10 k¼10
X
14 X
13 X
12
ICA14 ¼ ICA9 þ ILk ; ICA13 ¼ ICA9 þ ILk ; ICA12 ¼ ICA9 þ ILk
k¼10 k¼10 k¼10
X
11
ICA11 ¼ ICA9 þ ILk ; ICA10 ¼ ICA9 þ IL10 ;
k¼10
X
16 X
16
ICB1 ¼ IL9 ICB16 ILk ; ICB2 ¼ IL9 ICB16 ILk
k¼2;k6¼9 k¼3;k6¼9
X
16 X
16
ICB3 ¼ IL9 ICB16 ILk ; ICB4 ¼ IL9 ICB16 ILk ;
k¼4;k6¼9 k¼5;k6¼9
X
16 X
16
ICB5 ¼ IL9 ICB16 ILk ; ICB6 ¼ IL9 ICB16 ILk
k¼6;k6¼9 k¼7;k6¼9
X
16 X
16
ICB7 ¼ IL9 ICB16 ILk ; ICB8 ¼ IL9 ICB16 ILk ;
k¼8;k6¼9 k¼10
X
16 X
16
ICB9 ¼ ICB16 þ ILk ; ICB10 ¼ ICB16 þ ILk
k¼10 k¼11
X
16 X
16
ICB11 ¼ ICB16 þ ILk ; ICB12 ¼ ICB16 þ ILk ;
k¼12 k¼13
X
16 X
16
ICB13 ¼ ICB16 þ ILk ; ICB14 ¼ ICB16 þ ILk
k¼14 k¼15
ICB15 ¼ ICB16 þ IL16 ; IL1 ¼ ICB1 þ ICB16 : We get the following additional
expressions:
2 d 2 IL1 d 2 IL5 d 2 IL2 d 2 IL4 2

ðIL2 IL4 Þ ¼ 2 ½ 2 ; ðIL6 IL8 Þ
LC dt 2 dt 2 dt 2 dt LC
2 2 2 2
d IL5 d IL9 d IL6 d IL8
¼ þ 2½ 2 2
dt2 dt2 dt dt
2 d 2 IL9 d 2 IL13 d 2 IL10 d 2 IL12 2

ðIL10 IL12 Þ ¼ þ þ2 ½ 2 ; ðIL14 IL16 Þ
LC dt 2 dt 2 dt dt2 LC
2 2 2 2
d IL1 d IL13 d IL14 d IL16
¼ 2
2
þ2 ½ 2
dt dt dt dt2
We add the first and second above equations:
2 d 2 IL1 d 2 IL9
½* fðIL2 IL4 Þ þ ðIL6 IL8 Þg ¼ þ 2
LC dt2 dt2
d IL2 d IL4 d 2 IL6 d 2 IL8
2 2
½ 2 2 þ 2
dt dt dt2 dt
We add the third and forth above equations:
2
½** fðIL10 IL12 Þ þ ðIL14 IL16 Þg
LC
d 2 IL1 d 2 IL9 d 2 IL10 d 2 IL12 d 2 IL14 d 2 IL16
¼ 2
þ 2
þ2 ½ 2 þ
dt dt dt dt2 dt2 dt2
Integrating the last two results ([**]–[*]) gives the following:
2
fIL10 IL12 þ IL14 IL16 þ IL2 IL4 þ IL6 IL8 g
LC
d 2 IL10 d 2 IL12 d 2 IL14 d 2 IL16
¼2½ 2 þ
dt dt2 dt2 dt2
d 2 IL2 d 2 IL4 d 2 IL6 d 2 IL8
þ 2 þ 2
dt2 dt dt2 dt
We define new global variables for our Cylindrical RF network antennas system.
Y ¼ IL10 IL12 þ IL14 IL16 þ IL2 IL4 þ IL6 IL8

dIL10 dIL12 dIL14 dIL16 dIL2 dIL4 dIL6 dIL8
X¼ þ þ þ ;
dt dt dt dt dt dt dt dt
dY dX 1
¼X; ¼ Y
dt dt L C
Due to RF antenna copper leg parasitic effect, we get copper leg’s current and
current derivative with delay s1–k and s2–k (k is leg number index, k = 1,…,16). We
consider for simplicity s1–1 = s1–2 = = s1–16; s2–1 = s2–2 = = s2–16.
0
ILk ðtÞ ! ILk ðt s1k Þ ; ILk ðtÞ ¼ dILkdtðtÞ ; ILk
0 0
ðtÞ ! ILk ðt s2k Þ. We consider no
dI 0 ðtÞ
delay effect on Lkdt . YðtÞ ! Yðt s1 Þ ; XðtÞ ! Xðt s2 Þ: s1 ¼ s11 ¼ s12 ¼
1
¼ s116 s2 ¼ s21 ¼ s22 ¼ ¼ s216 : dYdt ¼ Xðt s2 Þ; dt ¼ LC Yðt s1 Þ.
dX
To find the Equilibrium points (fixed points) of the Cylindrical RF network

antennas system is by limt!1 Yðt s1 Þ ¼ YðtÞ and limt!1 Xðt s2 Þ ¼ XðtÞ.
dt ¼ 0 ; dt ¼ 0; 8 t s1 ; t s2 9ðt s1 Þ t; ðt s2 Þ t , t ! 1.
dY dX
We get two equations and the only fixed point is E ð0Þ ðY ð0Þ ; X ð0Þ Þ ¼ ð0; 0Þ.
ð0Þ ð0Þ ð0Þ ð0Þ ð0Þ ð0Þ ð0Þ ð0Þ

Y ð0Þ ¼ IL10 IL12 þ IL14 IL16 þ IL2 IL4 þ IL6 IL8 ¼ 0
0 ð0Þ 0 ð0Þ 0 ð0Þ 0 ð0Þ 0 ð0Þ 0 ð0Þ 0 ð0Þ 0 ð0Þ
X ð0Þ ¼ IL10 IL12 þ IL14 IL16 þ IL2 IL4 þ IL6 IL8 ¼ 0
equilibrium points of Cylindrical RF network antennas system consists in adding to
coordinates [Y X] arbitrarily small increments of exponential form ½y x ekt , and
retaining the first order terms in y, x. The system of two homogeneous equations
leads to a polynomial characteristics equation in the eigenvalues k. The polynomial
characteristics equations accept by set the below current and current derivative
respect to time into two Cylindrical RF network antennas system equations.
Cylindrical RF network antennas system fixed values with arbitrarily small incre-
ments of exponential form ½y x ekt are: i = 0 (first fixed point), i = 1 (second fixed
point), i = 2 (third fixed point).
YðtÞ ¼ Y ðiÞ þ y ekt ; X ¼ X ðiÞ þ x ekt ; Yðt s1 Þ ¼ Y ðiÞ þ y ekðts1 Þ ; Xðt s2 Þ

¼ X ðiÞ þ x ekðts2 Þ 8 i ¼ 0; 1; 2
We choose the above expressions for our YðtÞ; XðtÞ as small displacement [y x]
from the system fixed points at time t = 0. Yðt ¼ 0Þ ¼ Y ðiÞ þ y ; Xðt ¼ 0Þ ¼
X ðiÞ þ x for k\0; t [ 0 the selected fixed point is stable otherwise k [ 0; t [ 0 is
Unstable. Our Cylindrical RF network antennas system tend to the selected fixed
point exponentially for k\0; t [ 0 otherwise go away from the selected fixed
point exponentially. k is the eigenvalue parameter which establish if the fixed point
is stable or unstable, additionally his absolute value (jkj) establish the speed of flow
toward or away from the selected fixed point [2–6] (Table 2.4).
Table 2.4 Cylindrical RF network antennas system eigenvalues options

k<0 k>0
t=0 Y(t = 0) = Y(i) + y Y(t = 0) = Y(i) + y
X(t = 0) = X(i) + x X(t = 0) = X(i) + x
t>0 Y(t) ¼ YðiÞ þ y ejkjt Y(t) ¼ YðiÞ þ y ejkjt
X(t) ¼ XðiÞ þ x ejkjt X(t) ¼ XðiÞ þ x ejkjt
t!∞ Y(t ! ∞) = Y(i) Y(t ! 1; k [ 0) yejkjt
X(t ! ∞) = X(i) X(t ! 1; k [ 0) xejkjt
The speeds of flow toward or away from the selected fixed point for
Cylindrical RF network antennas system currents and currents derivative respect to
time are
dYðtÞ Yðt þ DtÞ YðtÞ Y ðiÞ þ y ekðt þ DtÞ ½Y ðiÞ þ y ekt

¼ lim ¼ lim
dt Dt!0 Dt Dt!0 Dt
kt kDt kt
y e ½e 1 e 1 þ kDt
kDt y e ½1 þ k Dt 1
¼ lim ! lim ¼ k y ekt
Dt!0 Dt Dt!0 Dt
dXðtÞ Xðt þ DtÞ XðtÞ X ðiÞ þ x ekðt þ DtÞ ½X ðiÞ þ x ekt
¼ lim ¼ lim
dt Dt!0 Dt Dt!0 Dt
x ekt ½ekDt 1 ekDt 1 þ kDt x ekt ½1 þ k Dt 1
¼ lim ! lim ¼ k x ekt
Dt!0 Dt Dt!0 Dt
and the time derivative of the above equations:
dYðtÞ dXðtÞ
¼ y k ekt ; ¼ x k ekt ;
dt dt
dYðt s1 Þ
¼ y k ekðts1 Þ ¼ y k ekt es1 k
dt
dXðt s2 Þ
¼ x k ekðts2 Þ ¼ x k ekt es2 k
dt
First we take the Cylindrical RF network antennas (Y) differential equation:

dY
dt ¼ X and adding to it coordinates [Y X] arbitrarily small increments of expo-
nential form ½y x ekt and retaining the first order terms in y, x.
k y ekt ¼ X ðiÞ þ x ekt ; X ði¼0Þ ¼ 0 ; k y þ x ¼ 0: Second we take the
1
Cylindrical RF network antennas (X) differential equation: dX dt ¼ LC Y and adding
to it coordinates [Y X] arbitrarily small increments of exponential form ½y x ekt
and retaining the first order terms in y, x.
1
k x ekt ¼ LC ½Y ðiÞ þ y ekt ; Y ði¼0Þ ¼ 0 ; k x ¼ LC
1
y ¼ 0. We define
Yðt s1 Þ ¼ Y ðiÞ þ y ekðts1 Þ ; Xðt s2 Þ ¼ X ðiÞ þ x ekðts2 Þ then we get two
delayed differential equations respect to adding to it coordinates [Y X] arbitrarily
small increments of exponential form ½y x ekt . In the equilibrium points:
Y ð0Þ ¼ 0 ; X ð0Þ ¼ 0
k y ekt ¼ X ð0Þ þ x ekðts2 Þ ; X ð0Þ ¼ 0 ) k y

1
¼ x eks2 ; k x ekt ¼ ½Y ð0Þ þ y ekðts1 Þ
LC
Y ð0Þ ¼ 0 ) k x¼ LC
1
y eks1 . We get the following set of eigenvalues
ks2 1
equations: k y þ x e ¼ 0 ; LC y eks1 k x ¼ 0
The small increments Jacobian of our Cylindrical RF network antennas.

! !
k eks2 y 0
1 ks1 ¼

LC e k x 0

k eks2
A k I ¼ 1 ks1 ; detjA k I j ¼ 0 ; Dðk; s1 ; s2 Þ
LC e k
1
¼ k2 þ eks1 eks2
LC
We have three stability analysis cases: s1 = s ; s2 ¼ 0 or s2 = s ; s1 ¼ 0 or s1 ¼

s2 ¼ s otherwise s1 6¼ s2 . We need to get characteristics equations as all above
stability analysis cases. We study the occurrence of any possible stability switching
resulting from the increase of value of the time delay s for the general characteristic
equation Dðk; sÞ. Dðk; sÞ ¼ Pn ðk; sÞ þ Qm ðk; sÞ eks . The expression for Pn ðk; sÞ
P
is Pn ðk; sÞ ¼ nk¼0 Pk ðsÞ kk ¼ P0 ðsÞ þ P1 ðsÞ k þ P2 ðsÞ k2 þ P3 ðsÞ k3 þ . . .
P
The expression for Qm ðk; sÞ is Qm ðk; sÞ ¼ m k¼0 qk ðsÞ k ¼ q0 ðsÞ þ q1 ðsÞ
k
k þ q2 ðsÞ k þ . . .
2
The first case we analyze is when there is delay in Cylindrical RF network

antennas leg’s current and no delay in antennas leg’s current derivative or opposite
s1 = s ; s2 ¼ 0 & s1 = 0 ; s2 = s [4, 5].
1 1
Dðk; s1 ¼ 0; s2 Þ ¼ k2 þ eks2 ¼ k2 þ eks ; Dðk; s1 ; s2 ¼ 0Þ
L C s2 ¼s LC
1 1
¼ k2 þ eks1 js1 ¼s ¼ k2 þ eks
LC LC
Dðk; sÞ ¼ Pn ðk; sÞ þ Qm ðk; sÞ eks : The expression for Pn ðk; sÞ is
X
n
Pn ðk; sÞ ¼ Pk ðsÞ kk ¼ P0 ðsÞ þ P1 ðsÞ k þ P2 ðsÞ k2 ¼ k2 ; P2 ðsÞ ¼ 1 ;
k¼0
P1 ðsÞ ¼ 0 ; P0 ðsÞ ¼ 0
P
The expression for Qm ðk; sÞ is Qm ðk; sÞ ¼ m k¼0 qk ðsÞ k ¼ q0 ðsÞ ¼ LC.
k 1
Our Cylindrical RF network antennas system second order characteristic equa-

tion: Dðk; sÞ ¼ k2 þ aðsÞ k þ bðsÞ k eks þ cðsÞ þ dðsÞ eks .
Then aðsÞ ¼ 0 ; b(sÞ = 0 ; c(sÞ = 0; d(sÞ¼ LC 1
; s 2 R þ 0 and aðsÞ; b(sÞ; c(sÞ;
d(sÞ : R þ 0 ! R are differentiable functions of class C 1 ðR þ 0 Þ such that
c(sÞ þ d(sÞ ¼ LC
1
6¼ 0 for all s 2 R þ 0 and for any s; b(sÞ; d(sÞ are not simultane-
ously zero. We have
Pðk; sÞ ¼ Pn ðk; sÞ ¼ k2 þ aðsÞ k þ cðkÞ ¼ k2 ; Qðk; sÞ ¼ Qm ðk; sÞ

1
¼ bðsÞ k þ dðsÞ ¼
LC
imaginary roots. That is for any real number x; pn ðk ¼ i x; sÞ þ Qm ðk ¼ i x; sÞ 6¼ 0
x2 þ LC1
6¼ 0; Fðx; sÞ ¼ jPði x; sÞj2 jQði x; sÞj2 ¼ ðc x2 Þ2 þ x2 a2
ðx2 b2 þ d 2 Þ
Fðx; sÞ ¼ x4 ðLCÞ
1
2 ; Hence Fðx; sÞ ¼ 0 implies x
4 1
ðLCÞ2
¼ 0 and its roots
pffiffiffiffi pffiffiffi pffiffiffiffi
are given by x2þ ¼ 12 fðb2 þ 2 c a2 Þ þ Dg ¼ 2D ; x2 ¼ 12 fðb2 þ 2 c a2 Þ Dg
pffiffiffi
¼ 2D
D ¼ ðb2 þ 2 c a2 Þ 4 ðc2 d 2 Þ ¼ L24C2 . Therefore the following holds:
pffiffiffiffi pffiffiffiffi
2 x2 ðb2 þ 2 c a2 Þ ¼ D; 2 x2 ¼ D. Furthermore
PR ði x; sÞ ¼ cðsÞ x2 ðsÞ ¼ x2 ðsÞ; PI ði x; sÞ ¼ xðsÞ aðsÞ ¼ 0; QR ði x; sÞ

1
¼ dðsÞ ¼
LC
QI ði x; sÞ ¼ xðsÞ bðsÞ ¼ 0 hence sin hðsÞ ¼ PR ðix;sÞQI ðix;sÞ þ PI ðix;sÞQR ðix;sÞ
jQðix;sÞj2
cos hðsÞ ¼ ;
jQði x; sÞj2
ðc x Þ x b þ x a d
2
sin hðsÞ ¼ ¼0
x 2 b2 þ d 2
cos hðsÞ ¼ ðcxx2Þd þ x ab

2 2
b2 þ d 2 ¼ x2 L C which jointly with x4 ðLCÞ

1
2 ¼ 0.
Defines the maps Sn ðsÞ ¼ s sn ðsÞ; s 2 I , n 2 N0 that are continuous and

differentiable in s based on Lemma 1.1. Hence we use Theorem 1.2. This prove the
Theorem 1.3 and Theorem 1.4.
Remark: a, b, c, d parameters are independent of delay parameter s even we use
aðsÞ; b(sÞ; c(sÞ; d(sÞ.
The second case we analyze is when there is delay both in Cylindrical RF
network antennas leg’s current and current time derivative s1 ¼s; s2 ¼ s [4, 5].
1
Dðk; s1 ¼ s; s2 ¼ sÞ ¼ k2 þ eks eks ; Dðk; sÞ
LC
¼ Pn ðk; sÞ þ Qm ðk; sÞ eks
Pn
The expression for Pn ðk; sÞ is Pn ðk; sÞ ¼ k¼0 Pk ðsÞ kk ¼ P0 ðsÞ þ P1 ðsÞ
k þ P2 ðsÞ k2 ¼ k2
P ðsÞ ¼ 1 ; P1 ðsÞ ¼ 0 ; P0 ðsÞ ¼ 0: The expression for Qm ðk; sÞ ; Qm ðk; sÞ ¼

Pm 2 ks
k¼0 qk ðsÞ k ¼ LC e
k 1
.
Taylor expansion: eks 1 k s þ k 2s since we need n > m [BK] analysis
2 2
P
we choose eks 1 k s then we get Qm ðk; sÞ ¼ m k¼0 qk ðsÞ k ¼ LC
k 1
ð1 k sÞ ¼ LC1
LC
1
k s.
q0 ðs; kÞ ¼ LC ; q1 ðsÞ ¼ LC
1 1
s ; q2 ðsÞ ¼ 0: Our Cylindrical RF network
antennas system second order characteristic equation: Dðk; sÞ ¼ k2 þ aðsÞ
1
k þ bðsÞ k eks þ cðsÞ þ dðsÞ eks then aðsÞ ¼ 0 ; bðsÞ ¼ LC s cðsÞ ¼ 0 ;
dðsÞ ¼ LC and in the same manner like our previous case analysis: Pðk; sÞ ¼
1
Pn ðk; sÞ ¼ k2 ; Qðk; sÞ ¼ Qm ðk; sÞ ¼ LC 1

LC
1
k s. We assume that Pn ðk; sÞ ¼
Pn ðkÞ and Qm ðk; sÞ can’t have common imaginary roots. That is for any real
number x; pn ðk ¼ i x; sÞ þ Qm ðk ¼ i x; sÞ 6¼ 0; x2 i x LC 1
s þ LC
1
6¼ 0
Fðx; sÞ ¼ jPði x; sÞj2 jQði x; sÞj2 ; Pði x; sÞ ¼ x2 ;

PR ði x; sÞ ¼ x2 ; PI ði x; sÞ ¼ 0
1 1 1
Qðk ¼ i x; sÞ ¼ i x sþ ; QI ðk ¼ i x; sÞ ¼ x s;
LC LC LC
1
QR ðk ¼ i x; sÞ ¼
LC
jPði x; sÞj2 ¼ P2I þ P2R ; jQði x; sÞj2 ¼ Q2I þ Q2R ; jPði x; sÞj2 ¼ P2I þ P2R ¼ x4
s2 1 s2 1
jQði x; sÞj2 ¼ x2 2
þ 2
; Fðx; sÞ ¼ x4 x2 2

ðL CÞ ðL CÞ ðL CÞ ðL CÞ2
s 2
Hence Fðx; sÞ ¼ 0 implies x4 x2 ðLCÞ 2
1
ðLCÞ2
¼ 0; Fx ¼ 4 x3 2 x
s2 s 2
ðLCÞ2
¼ 2 x ½2 x2 ðLCÞ2
x2 2 s s
Fs ¼ ; PIx ¼ 0 ; PRx ¼ 2 x ; QIx ¼ ; QRx ¼ 0 ;
ðL CÞ 2 LC
x
PIs ¼ 0 ; PRs ¼ 0 ; QRs ¼ 0 ; QIs ¼
LC
The expressions for U, V can be derive easily [BK]: x ¼ s
U ¼ ðPR PIx PI PRx Þ ðQR QIx QI QRx Þ ;

V ¼ L2xC2 ; U ¼ L2 sC2 ; xs ¼ FFxs and we get the expression:

x2 2s xs

ðLCÞ2 ðLCÞ2
xs ¼ 2 ¼ s2
: Defines the maps Sn ðsÞ ¼ s sn ðsÞ;
2x½2x2 s 2 ½2x2
ðLCÞ ðLCÞ2
s 2 I , n 2 N0 .
Defines the maps Sn ðsÞ ¼ s sn ðsÞ; s 2 I , n 2 N0 that are continuous and
differentiable in s based on Lemma 1.1. Hence we use Theorem 1.2. This prove the
Theorem 1.3 and Theorem 1.4.
Remark: Taylor approximation for eks 1 k s gives us good stability
analysis approximation only for restricted delay time interval.
Our Cylindrical RF network antennas homogeneous system for y, x leads to a
characteristic equation for the eigenvalue k having the for PðkÞ þ QðkÞ eks ¼ 0;
second case s1 ¼s; s2 ¼ s; Dðk; s1 ¼ s; s2 ¼ sÞ ¼ k2 þ LC 1
eks eks : We esti-
mate eks 1 k s. Dðk; s1 ¼ s; s2 ¼ sÞ ¼ k2 þ LC 1
ð1 k sÞ eks
ks
Dðk; s1 ¼ s; s2 ¼ sÞ ¼ k þ ðk LC s þ LCÞ e : We use different parameters
2 1 1
terminology from our last characteristics parameters definition: k ! j; pk ðsÞ !

aj ; qk ðsÞ ! cj ; n ¼ 2; m ¼ 1: Additionally Pn ðk; sÞ ! PðkÞ; Qm ðk; sÞ ! QðkÞ
P P
then PðkÞ ¼ 2j¼0 aj k j and QðkÞ ¼ 1j¼0 cj k j ; PðkÞ ¼ k2 ; Qðk; sÞ ¼ k LC 1

s þ LC.1
n; m 2 N0 ; n [ m and aj ; cj 6¼ R þ 0 ! R are continuous and differentiable

function of s such that a0 þ c0 6¼ 0. In the following “−” denotes complex and
conjugate. PðkÞ; QðkÞ are analytic functions in k and differentiable in s. And the
coefficients : faj ðC; LÞ; cj ðC; L; sÞg 2 R depend on Cylindrical RF network
antennas C, L, s values. a0 ¼ 0; a1 ¼ 0; a2 ¼ 1 ; c0 ¼ LC 1
; c1 ¼ LC
1
s unless
strictly necessary, the designation of the variation arguments ðC; L; sÞ will subse-
quently be omitted from P, Q, aj, cj. The coefficients aj, cj are continuous, and
a0 þ c0 ¼ LC1
6¼ 0; LC
1
6¼ 0 8 C,L,s 2 R þ i.e. k ¼ 0 is not a root of characteristic
equation. Furthermore PðkÞ; QðkÞ are analytic function of k for which the following
requirements of the analysis (see kuang [5], Sect. 3.4) can also be verified in the
present case [4, 5].
(a) If k = i x; x 2 R then P(i x) + Q(i x) 6¼ 0, i.e. P and Q have no common
imaginary roots. This condition was verified numerically in the entire ðC; L; sÞ
domain of interest.
(b) jQðkÞ=PðkÞj is bounded for jkj ! 1, Rek
kLC
1
s þ LC
1
Indeed, in the limit jQ(kÞ=PðkÞj ¼ j k 2 j
s
(c) FðxÞ ¼ jPði xÞj2 jQði xÞj2 ; Fðx; sÞ ¼ x x2 ðLCÞ
2
2
1 4
ðLCÞ2
has at
most a finite number of zeros. Indeed, this is a bi-cubic polynomial in x
(second degree in x2 ).
(d) Each positive root xðC; L; sÞ of F(xÞ¼ 0 is continuous and differentiable with
respect to C; L; s. This condition can only be assessed numerically.
In addition, since the coefficients in P and Q are real, we have Pði xÞ ¼

Pði xÞ and Qði xÞ ¼ Qði xÞ thus k ¼ i x, x [ 0 may be on eigenvalue of
characteristic equation. The analysis consists in identifying the roots of character-
istic equation situated on the imaginary axis of the complex k—plane, where by
increasing the parameters C; L and delay s, Rek may, at the crossing, change its
sign from (−) to (+), i.e. from stable focus E ð0Þ ðY ð0Þ ; X ð0Þ Þ ¼ ð0; 0Þ to an unstable
the partial derivatives with respect to C; L and antenna parameters.
^1 ðCÞ ¼ ð@ Re k 1 @ Re k
@C Þk¼ix ; L; s ¼ const; ^ ðLÞ ¼ ð @L Þk¼ix ; C; s ¼ const
^1 ðsÞ ¼ ð@ RekÞ@s ; C; L; where x 2 R : For the first case s ¼s; s ¼ s we
k¼ix þ 1 2
get the following results PR ði xÞ ¼ x2 ; PI ði xÞ ¼ 0; QR ði xÞ ¼ LC
1
;
QI ði xÞ ¼ xs
LC .
s 2
s 2
FðxÞ ¼ 0 yield to x4 x2 ðLCÞ2
1
ðLCÞ2
¼ 0; v2 ¼ x4 ; v ¼ x2 ; v2 v ðLCÞ2
ðLCÞ
1
2 ¼ 0
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
s2 1 s4 1
v¼ 2
4
þ4 ; v ¼ x2 ) x
2 ðL CÞ 2 ðL CÞ ðL CÞ2
vffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
u
u s 2 1 s4 1
¼ t 2
4
þ4
2 ðL CÞ 2 ðL CÞ ðL CÞ2
s4 s
2
ðLCÞ4
þ 4 ðLCÞ 1
2 [0 always and additional for x 2 R; x2 ¼ 2ðLCÞ
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
s4 s2
ðLCÞ þ 4 ðLCÞ and there are two options: first always exist 2ðLCÞ2 þ 2
1 1 1
2
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
ffi
s4
ðLCÞ4
þ 4 ðLCÞ 1
2 [ 0.
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
ffi qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
s2 s4 s2 s4
2 2 þ 4 ðLCÞ 2 \0; x ¼ 2 LC fLC þ 4g.
1 1 2 1 1
Second 2ðLCÞ ðLCÞ4 ðLCÞ2
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
s4 s2
ðLCÞ2
þ 4 [ LC ; Not exist and always negative for any Cylindrical RF net-
work antennas overall parameters values. We choose only the (+) option (first).
Writing PðkÞ ¼ PR ðkÞ þ i PI ðkÞ and QðkÞ ¼ QR ðkÞ þ i QI ðkÞ, and inserting k ¼
i x into Cylindrical RF network antennas characteristic equation, x must satisfy
the following: sin x s ¼ gðxÞ ¼ PR ðixÞQI ðixÞ þ PI ðixÞQR ðixÞ
jQðixÞj2
.
cos x s ¼ hðxÞ ¼ PR ðixÞQR ðixÞ þ PI ðixÞQI ðixÞ
jQðixÞj2
; where jQði xÞj2 6¼ 0 in view
of requirement (a) above, and ðg; hÞ 2 R. Furthermore, it follows above sin x s
and cos x s equations that, by squaring and adding the sides, x must be a positive
root of FðxÞ ¼ jPði xÞj2 jQði xÞj2 ¼ 0.
Note that FðxÞ is dependent of s. Now it is important to notice that if s 62 I
(assume that I R þ 0 is the set where xðsÞ is a positive root of FðxÞ and for s 62 I ,
xðsÞ is not define. Then for all s in I xðsÞ is satisfies that Fðx; sÞ ¼ 0.
Then there are positive xðsÞ solutions of Fðx; sÞ ¼ 0, and we analyze stability
switches. For any s 2 I where xðsÞ is a positive solution of Fðx; sÞ ¼ 0, we can
define the angle hðsÞ 2 ½0; 2 p as the solution of sin hðsÞ ¼ . . .; cos hðsÞ ¼ . . .
sin hðsÞ ¼
jQði xÞj2
cos hðsÞ ¼
jQði xÞj2
and the relation between the argument hðsÞ and xðsÞ s for s 2 I must be
þ n2p
Sn ðsÞ ¼ s sn ðsÞ; s 2 I; n 2 N0 that are continuous and differentiable in s. In the
following, the subscripts k; x; C; L and Cylindrical RF network antennas param-
eters ðL; C; s etc:; Þ indicate the corresponding partial derivatives. Let us first con-
centrate on ^ðxÞ, remember in kðL; C; s; etc:; Þ and xðL; C; s; etc:; Þ, and keeping
[BK]. Differentiating Cylindrical RF network antennas characteristic equation
PðkÞ þ QðkÞ eks ¼ 0 with respect to specific parameter (x), and inverting the
derivative, for convenience, one calculates.
Remark: x ¼ L; C; s; etc:;

ð Þ ¼
where Pk ¼ @P
@k ; . . . etc., Substituting k ¼ i x, and bearing i Pði xÞ ¼
Pði xÞ; Qði xÞ ¼ Qði xÞ,
then i Pk ði xÞ ¼ Px ði xÞ and i Qk ði xÞ ¼ Qx ði xÞ and that on the sur-
face jPði xÞj2 ¼ jQði xÞj2 , one obtains
@k i Px ði x; xÞ Pði x; xÞ þ i Qk ði x; xÞ Qðk; xÞ s jPði x; xÞj2

ð Þ1 jk¼ix ¼ ð Þ:
Upon separating into real and imaginary parts, with P ¼ PR þ i PI ; Q ¼ QR þ i QI

Px ¼ PRx þ i PIx ; Qx ¼ QRx þ i QIx ; Px ¼ PRx þ i PIx ; Qx ¼ QRx þ i QIx ;
P2 ¼ P2R þ P2I :
When (x) can be any Cylindrical RF network antennas parameters L, C, And
time delay s etc. Where for convenience, we have dropped the arguments ði x; xÞ,
and where Fx ¼ 2 ½ðPRx PR þ PIx PI Þ ðQRx QR þ QIx QI Þ ; xx ¼ Fx =Fx .
Fx ¼ 2 ½ðPRx PR þ PIx PI Þ ðQRx QR þ QIx QI Þ. We define U and V:

We choose our specific parameter as time delay x = s. V ¼ L2xC2 ; U ¼ L2 sC2
x2 2 s
P2 ¼ x4 ; Fs ¼ ; PR ðx; sÞ ¼ x2 ; PI ðx; sÞ ¼ 0 ;
ðL CÞ2
xs 1
QI ðx; sÞ ¼ ; QR ðx; sÞ ¼
LC LC
x
PIs ¼ 0; PRs ¼ 0; QRs ¼ 0; QIs ¼ ) V 6¼ 0;
LC
@F s2
¼ Fx ¼ 4 x3 2 x
@x ðL CÞ2
@F s 2
@x ¼ 2 x ½2 x2 ðLCÞ 2 ; Fðx; sÞ ¼ 0 and differentiating with respect to s
xs
and we get Fx @x @x Fs @x
@s þ Fs ¼ 0; s 2 I ) xs ¼ @s ¼ Fx ; @s ¼
ðLCÞ2
s2
½2x2
ðLCÞ2
@Rek @x xs
^1 ðsÞ ¼ ð Þ ; xs ¼ ¼
@s k¼ix @s ½2 x2 ðL CÞ2 s2
^1 ðsÞ ¼ Ref g
s 2
s ½L21C2 þ x4 þ i x ½2 x2 ðLCÞ 2
¼ Ref x s
g
þ i x ½L21C2 þ x4
2
ðLCÞ2
@Rek
@s k¼ix
@x U @x
@s þ V
@s jPj2
s2
signf^1 ðsÞg ¼ signf2 x ½2 x2 g
ðL CÞ2
xs
s ðLCÞ2 x
xs L2 C 2 ½ s2
þ L2 C 2
½2x2
ðLCÞ2 ðLCÞ2
signfs ½ þxþ g
s
½2 x2 ðLCÞ
2
2
x4
s 2
We define new variables: w1 ; w2 ; w3 : w1 ðx; s; L; CÞ ¼ 2 x ½2 x2 ðLCÞ 2
xs
s ðLCÞ2 x
xs L2 C2 ½ s2
þ L2 C 2
ðLCÞ2 ½2x2
ðLCÞ2
w2 ðx; s; L; CÞ ¼ s ½ ; w3 ðx; s; L; CÞ ¼
s2
½2 x2 ðLCÞ 2 x4
sign f^1 ðsÞg ¼ sign ½w1 sign ½w2 þ x þ w3 . We check the sign of ^1 ðsÞ
according the following rule.
specific x, s. Since it is a very complex function, we recommend to solve it
numerically rather than analytic. We plot the stability switch diagram based on
different delay values of our Cylindrical RF network antennas system. Dðk; s1 ¼
s
eks k LC eks : Taylor expansion: eks 1 k s þ k 2s
2 2
s2 ¼ sÞ ¼ k2 þ LC
1
since we need n > m [BK] analysis we choose eks 1 k s then we get our
Cylindrical RF network antennas system second order characteristic equation:
Dðk; sÞ ¼ k2 þ aðsÞ k þ bðsÞ k eks þ cðsÞ þ dðsÞ eks (Table 2.5).
s 1
aðsÞ ¼ 0; bðsÞ ¼ ; cðsÞ ¼ 0 ; dðsÞ ¼ ; Fðx; sÞ
LC LC
2 2 2 2
¼ jPði x; sÞj jQði x; sÞj ¼ ðc x Þ þ x2 a2 ðx2 b2 þ d 2 Þ
s 2
s 2
Fðx; sÞ ¼ x4 x2 ðLCÞ 2
1
ðLCÞ2
hence Fðx; sÞ ¼ 0 implies x4 x2 ðLCÞ 2
p ffiffiffiffi
1
ðLCÞ2
¼ 0 and its roots are given by x þ ¼ 2 fðb þ 2 c a Þ þ Dg ¼
2 1 2 2
p ffiffiffiffi s2
2 f D þ ðLCÞ2 g
1
1 pffiffiffiffi 1 pffiffiffiffi s2
x2 ¼ fðb2 þ 2 c a2 Þ Dg ¼ f D þ g;
2 2 ðL CÞ2
s2 þ 4
D ¼ ðb2 þ 2 c a2 Þ 4 ðc2 d 2 Þ ¼
ðL CÞ2
s þ4 2
D ¼ ðb2 þ 2 c a2 Þ 4 ðc2 d 2 Þ ¼ ðLCÞ 2 therefore the following holds:
pffiffiffiffi PR ði x; sÞ QI ði x; sÞ þ PI ði x; sÞ QR ði x; sÞ
2 x2 ðb2 þ 2 c a2 Þ ¼ D; sin hðsÞ ¼
jQði x; sÞj2
Table 2.5 Cylindrical RF sign½Fx sign½V þPx2 s U þ x þ xs s sign½^1 ðsÞ

network antennas system
stability switching criteria ± ± +
± –
cos hðsÞ ¼ ;
jQði x; sÞj2
ðc x2 Þ x b þ x a d x3 s L C
sin hðsÞ ¼ ¼
x 2 b2 þ d 2 ðx2 s2 þ 1Þ
cos hðsÞ ¼ ðcxx2Þd þ x ab

2 2
¼ ðxx2 sLC
2
b2 þ d 2 2 þ 1Þ : We consider Cylindrical RF antenna
which mounted outside a Pyrex glass tube of diameter 32 cm and length 50 cm.
The RF antenna consists of 16 copper (Cu) legs equally spaced by 6.7 cm inter-
connected with capacitors of 2.47nF. Copper leg diameter is equal to 1 mm and
length 30 cm = 300 mm (<Pyrex glass tube length, 50 cm). We consider for
Copper (Cu), relative permeability is one. f = 10 MHz is the typical testing fre-
quency for cylindrical (birdcage) antenna. L—inductance (nH), l—length of copper
leg (mm), d—diameter of copper leg, f—testing frequency. l > 100 d
(300 > 100 1 mm), d2 f > 1 mm2 MHz (1 mm2 10 MHz > 1 mm2 MHz).
L = 365.4 nH. L ¼ 15 l ½lnð4ld 1 ¼ 365:4 nH. For our stability switching anal-
ysis we choose typical Cylindrical RF network antennas parameters values (as
calculated): C ¼ 2:47 nF; L ¼ 365:4 nH; Rp ¼ 100Ohm then
LC ¼ 0:00110798 10 . We find those x; s values which fulfill Fðx; sÞ ¼ 0. We
1 18
ignore negative, complex, and imaginary values of x for specific s values. The
below table gives the list.
Remark: We know Fðx; sÞ ¼ 0 implies it roots xi ðsÞ and finding those delays
values s which xi is feasible. There are s values, which xi are complex or imag-
inary numbered, then unable to analyze stability [6, 7]. We find those x; s values
which fulfill Fðx; sÞ ¼ 0. We ignore negative, complex, and imaginary values of x
for specific s values. s 2 ½0:001::10 and we can be express by 3D function
s2
Fðx; sÞ ¼ 0. Fðx; sÞ ¼ x4 x2 ðLCÞ 2
1
ðLCÞ2
X
2
Fðx; sÞ ¼ jPði x; sÞj2 jQði x; sÞj2 ¼ U0 þ U2 x2 þ U4 x4 ¼ U2k x2k
k¼0
s 2 P4
U0 ¼ ðLCÞ
1
2 ; U2 ¼
ðLCÞ2
; U4 ¼ 1 hence Fðx; sÞ ¼ 0 implies k¼0 U2k
sx ¼ 0.
2k
Uj ! Phij. Running MATLAB script for s values (s 2 ½0:001::10) gives the

following results.
MATLAB script: Tau=0.001;C=2.47*1e-9;L=365.4*1e-9;Phi0=-1/(C*L*C*L);
Phi2=-(Tau*Tau)/(C*L*C*L); Phi4=1;p=[Phi4 0 Phi2 0 Phi0];r=roots(p)
(Tables 2.6, 2.7, 2.8, 2.9, 2.10, 2.11 and 2.12).
We can summery our xi ðsÞ results for xi ðsÞ [ 0 and real number (ignore
complex, negative, and imaginary values). We exclude from our table the high and
real xi ðsÞ values (1.0e+007*, 1.0e+012*,…,1.0e+016*) and add results for s = 15
and s = 20 s (Figs. 2.14, 2.15 and Table 2.13).
Table 2.6 Cylindrical RF s s = 0.01 s s = 0.001 s

roots xi ðsÞ x1 1.0e+013* 1.0e+012*
x2 −1.1080 −1.1080
x3 1.1080 1.1080
x4 0.0000 + 0.0000i −0.0000 + 0.0000i
x5 0.0000 − 0.0000i −0.0000 − 0.0000i
Table 2.7 Cylindrical RF s s=1s s = 0.1 s

roots xi ðsÞ x1 1.0e+015* 1.0e+014*
x2 −1.1080 −1.1080
x3 1.1080 1.1080
x4 −0.0000 + 0.0000i 0.0000 + 0.0000i
x5 −0.0000 − 0.0000i 0.0000 − 0.0000i
Table 2.8 Cylindrical RF s s=3s s=2s

roots xi ðsÞ x1 1.0e+015* 1.0e+015*
x2 3.3240 −2.2160
x3 −3.3240 2.2160
x4 0 + 0.0000i −0.0000 + 0.0000i
x5 0 − 0.0000i −0.0000 − 0.0000i

roots xi ðsÞ x1 1.0e+015* 1.0e+015*
x2 −5.5399 4.4319
x3 5.5399 −4.4319
x4 0.0000 + 0.0000i 0 + 0.0000i
x5 0.0000 − 0.0000i 0 − 0.0000i

roots xi ðsÞ x1 1.0e+015* 1.0e+015*
x2 −7.7559 6.6479
x3 7.7559 −6.6479
x4 0.0000 + 0.0000i 0 + 0.0000i
x5 0.0000 − 0.0000i 0 − 0.0000i
Matlab: plot([0 0.001 0.01 0.1 1 2 3 4 5 6 7 8 9 10 15 20], [3.3286 1.1080

1.1080 1.1080 1.1080 2.2160 3.3240 4.4319 5.5399 6.6479 7.7559 8.8639 9.9719
1.1080 1.6620 2.2160],‘-or’). We plot 3D function Fðx; sÞ ¼ 0. s:0!10;
x:0!100. We define additional MATLAB script parameters x!w, s!t.

roots xi ðsÞ x1 1.0e+015* 1.0e+015*
x2 9.9719 8.8639
x3 −9.9719 −8.8639
x4 0 + 0.0000i 0 + 0.0000i
x5 0 − 0.0000i 0 − 0.0000i
Table 2.12 Cylindrical RF s s=0s s = 10 s

roots xi ðsÞ x1 1.0e+007* 1.0e+016*
x2 −3.3286 −1.1080
x3 −0.0000 + 3.3286i 1.1080
x4 −0.0000 − 3.3286i −0.0000 + 0.0000i
x5 3.3286 −0.0000 − 0.0000i
Fig. 2.14 Cylindrical RF

network F(x,s) function for
s1 = s2 = s
Matlab: [w,t]=meshgrid(1:1:100,0:0.01:10);C=2.47*1e-9; L=365.4*1e-9;

f=w.*w.*w.*w-w.*w.*(t.*t)/(C*L*C*L)-1/(C*L*C*L);meshc(f); % x ! w; s ! t.
We get two possible real values for x which fulfil Fðx; sÞ ¼ 0
Fðx ¼ 3:3286 or x ¼ 1:1080 . . . or x¼ 2:2160; sÞ ¼ 0; s 2 ½0:001::10. Next is
to find those x, s values which fulfil sin hðsÞ ¼ . . .; sinðx sÞ ¼ PR QjQj
I þ PI QR
2 and
cos hðsÞ ¼ . . .

network F(x,s) function for
s1 = s2 = s
Table 2.13 Cylindrical RF network antennas system positive and real root xi ðsÞ values and
sinðx sÞ; cosðx sÞ values
x sLC
cosðx sÞ ¼ ðxx2 sLC
3 2
s[s] x sin ðx sÞ ¼ ðx2 s2 þ 1Þ 2 þ 1Þ
0 3.3286 0=0 1 6¼ 9.9e−15

0.001..1 1.1080 −1.22e−18 … −5.51e−16 1.108e−15 … 4.973e−16
2 2.2160 −9.5e−16 2.1e−16
3 3.3240 −9.9e−16 9.9e−17
4 4.4319 −9.9e−16 5.62e−17
5 5.5399 −9.9e−16 3.6e−17
6 6.6479 −9.99−16 2.5055e−17
7 7.7559 −9.9966e−16 1.8413e−17
8 8.8639 −9.9980e−16 1.4099e−17
9 9.9719 −9.9988e−16 1.1141e−17
10 1.1080 −9.9193e−17 8.9525e−18
15 1.6620 −9.9841e−17 4.0048e−18
20 2.2160 −9.9950e−17 2.2552e−18
ðPR QR þ PI QI Þ
cosðx sÞ ¼ ; jQj2 ¼ Q2R þ Q2I ;
jQj2
x3 s L C x2 L C
sinðx sÞ ¼ ; cosðx sÞ ¼
ðx2 s2 þ 1Þ ðx2 s2 þ 1Þ
x3 sLC x2 LC

ðx2 s2 þ 1Þ \0 & ðx2 s2 þ 1Þ [0 then sinðx sÞ\0 and cosðx sÞ [ 0;
2 p [ x s [ p2 3.

network g1 ðs; xÞ F(x,s)
function for s1 = s2 = s
We plot the stability switch diagram based on different delay values of our
Cylindrical RF network antennas system. ^1 ðsÞ ¼ ð@ Re k
@s Þk¼ix ¼
2
Ref2½U þ sjPj þ iFx
F þ i2½V þ xjPj2
g
s
signf^1 ðsÞg ¼ sign½w1 sign½w2 þ x þ w3 : We define the following new

functions (Figs. 2.16 and 2.17):
g1 ¼ w1 ; g2 ¼ w2 þ x þ w3 ; signf^1 ðsÞg ¼ sign½g1 sign½g2
Matlab: [w,t]=meshgrid(1:.01:100,0:0.01:10);C=2.47*1e-9; L=365.4*1e-9;

f=2*w.*(2*w.*w-(t.*t./(C*L*C*L)));meshc(f) % x ! w; s ! t.
Matlab: [w,t]=meshgrid(1:.1:10,0:0.1:10);C=2.47*1e-9; L=365.4*1e-9;m=w.
*t./(2*w.*w.*(L*C*L*C)-t.*t); f=t.*m+w+(m.*t./(L*C*L*C)+w./(L*C*L*C)) ./
(w.*w.*w.*w);meshc(f) % x ! w; s ! t.
gðTauÞ ¼ g1 ðTauÞ g2 ðTauÞ ¼ ^1 ðsÞ ¼ ð@ Re k
@s Þk¼ix : The stability switch occur
only on those delay values (s) which fit the equation: s ¼ xh þþ ðsÞ
ðsÞ and h þ ðsÞ is the
x sLC
3
x LC
2
solution of sin hðsÞ ¼ ðx 2 s2 þ 1Þ ; cos hðsÞ ¼ ðx2 s2 þ 1Þ when x ¼ x þ ðsÞ if only x þ
is feasible. Additionally When all Cylindrical RF network antennas parameters are

known and the stability switch due to various time delay values s is describe in the
below expression (Theorem 1.5):
Fig. 2.17 Cylindrical RF network g2 ðs; xÞ function for s1 = s2 = s
signf^1 ðsÞg ¼ signfFx ðxðsÞ; sÞg signfs

xs ðxðsÞÞ þ xðsÞ þ g
jPðxðsÞÞj2
Remark: We know Fðx; sÞ ¼ 0 implies it roots xi ðsÞ and finding those delays
values s which xi is feasible. There are s values which xi is complex or imaginary
number, then unable to analyse stability [4, 5].
Discussion: We consider Cylindrical RF network antennas system. Due to RF
antenna copper leg parasitic effect we get copper leg’s current and current
derivative with delay s1–k and s2–k (k is leg number index, k = 1,…,16). Those
delays causes to stability switching for our Cylindrical RF network antennas. We
draw our Cylindrical RF network antennas equivalent circuit and get system dif-
ferential equations. Our variables are Y, X which are function of RF antenna copper
leg’s current and current derivative. Our system dynamic behavior is dependent on
circuit overall parameters and parasitic delay in time. We keep all circuit parameters
fix and change, parasitic delay over various values s 2 ½0:001::10. Our analysis
results extend that of in the way that it deals with stability switching for different
delay values. This implies that our system behavior of the circuit cannot inspect by
short analysis and we must study the full system. Several very important issues and
possibilities were left out of the present discussion. One possibility is the stability
switching by circuit parameters. Every circuit’s parameter variation can change our
system dynamic and stability behavior. This case can be solved by the same
methods combined with alternative and more technical hypotheses. Moreover,
numerical simulations for the Cylindrical RF network antennas model studied in
references suggest that this result can be extended to enhance models with more
general functions. Still another extension of our results would be to treat the case of
delayed Cylindrical RF network antennas leg’s higher derivative degree of current.
It would be extremely desirable to confirm these cases by mathematical proofs.
Conclusion: Cylindrical RF network antennas system is characterized by para-
sitic effects which can influence Cylindrical RF network antennas system stability
in time. There are two main Cylindrical RF network antennas variables which are
affected by antenna legs parasitic effects, Y and X functions of antenna leg’s
currents and currents derivatives respectively. Each Cylindrical RF network
antennas system variable under parasitic effects is characterized by time delay
respectively. The two time delays are not the same, but can be categorized to some
sub cases due to antenna leg parasitic behavior. The first case we analyze is when
there is delay in Cylindrical RF network antennas leg’s current and no delay in
antennas leg’s current derivative or opposite. The second case we analyze is when
there is delay both in Cylindrical RF network antennas leg’s current and current
time derivative [4, 5]. For simplicity of our analysis we consider in the second case
all delays are the same (there is a difference but it is neglected in our analysis). In
each case we derive the related characteristic equation. The characteristic equation
is dependent on Cylindrical RF network antennas system overall parameters and
parasitic time delay. Upon mathematics manipulation and [BK] theorems and
definitions we derive the expression which gives us a clear picture on
Cylindrical RF network antennas map. The stability map gives all possible options
for stability segments, each segment belongs to different time delay value segment.
Cylindrical RF network antennas system’s stability analysis can be influenced either
by system overall parameter values. We left this analysis and do not discuss it in the
current subchapter.
Lemma 1.1 Assume that xðsÞ is a positive and real root of Fðx; sÞ ¼ 0.
Defined for s 2 I, which is continuous and differentiable. Assume further that if
s 2 I; I R þ 0 , and at some s 2 I, Sn ðs Þ ¼ 0 for some n 2 N0 then a pair of
dRek dSn ðsÞ

dðs Þ ¼ signf jk¼ixðs Þ g ¼ signfFx ðxðs Þ; s Þg signf j g
ds ds s¼s
n ðsÞ
The theorem becomes signfd Re k
ds jk¼ix g ¼ signfD
1=2
g signfdSds js¼s g.
Theorem 1.3 The characteristic equation: s1 ¼ s; s2 ¼ 0; s1 ¼ 0; s2 ¼ s
Dðk; sÞ ¼ k2 þ aðsÞ k þ bðsÞ k eks þ cðsÞ þ dðsÞ eks ; Dðk; s1 ; s2 Þ

1 1
¼ k2 þ k eks2 þ ekðs1 þ s2 Þ
C1 R1 C1 f#
Has a pair of simple and conjugate pure imaginary roots k ¼ xðs Þ; xðs Þ
real at s 2 I if Sn ðs Þ ¼ s sn ðs Þ ¼ 0 for some n 2 N0 . If xðs Þ ¼ x þ ðs Þ, this
pair of simple conjugate pure imaginary roots crosses the imaginary axis from left
to right if d þ ðs Þ [ 0 and crosses the imaginary axis from right to left if
dSn ðsÞ
d þ ðs Þ\0 where d þ ðs Þ ¼ signfd Re k
ds jk¼ix þ ðs Þ g ¼ signf ds js¼s g: If
xðs Þ ¼ x ðs Þ, this pair of simple conjugate pure imaginary roots cross the
imaginary axis from left to right if d ðs Þ [ 0 and crosses the imaginary axis from
right to left If d ðs Þ\0 where d ðs Þ ¼ signfd Re k
ds jk¼ix ðs Þ g ¼
n ðsÞ
signfdSds js¼s g: If x þ ðs Þ ¼ x ðs Þ ¼ xðs Þ then Dðs Þ ¼ 0 and
signfdRe k 0
ds jk¼ixðs Þ g ¼ 0, the same is true when Sn ðs Þ ¼ 0. The following result can
be useful in identifying values of s where stability switches happened.
Theorem 1.4 Assume that for all s 2 I, xðsÞ is defined as a solution of Fðx; sÞ ¼
0 then d ðsÞ ¼ signfD1=2 ðsÞg signD ðsÞ.
D ðsÞ ¼ x2 ½ðx2 b2 þ d 2 Þ þ a0 ðc x2 Þ þ b d 0 b0 d a c0

þ x x0 ½s ðx2 b2 þ d 2 Þ
daðsÞ 0 dbðsÞ
b d þ a ðc x2 Þ þ 2 x2 a; a0 ¼ ;b ¼ ;
ds ds
dcðsÞ 0 ddðsÞ
c0 ¼ ;d ¼
ds ds
Theorem 1.5 We need to approve the following expression:
signf^1 ðsÞg ¼ signfFx ðxðsÞ; sÞg signfs

xs ðxðsÞÞ þ xðsÞ þ g
jPðxðsÞÞj2
The basic assumption: ^1 ðsÞ ¼ ð@ Re k

@s Þk¼ix
@Rek @Rek
^1 ðsÞ ¼ ð Þk¼ix ; ^1 ðsÞ ¼ ð Þ
@s @s k¼ix
2 fFx ðV þ x P2 Þ Fs ðU þ s P2 Þg
¼
signfFs2 þ 4 ðV þ x P2 Þ2 g [ 0 and xs ¼ FFxs then
@Rek
signf^1 ðsÞg ¼ signfð Þ g ¼ signfFx ðV þ x P2 Þ Fs ðU þ s P2 Þg
@s k¼ix
Fs
signf^1 ðsÞg ¼ signfFx fðV þ x P2 Þ ðU þ s P2 Þgg;
Fx
signf^1 ðsÞg ¼ signfFx fðV þ x P2 Þ þ xs ðU þ s P2 Þgg
signf^1 ðsÞg ¼ signfFx fV þ xs U þ x P2 þ xs s P2 gg;
V þ xs U
signf^1 ðsÞg ¼ signfP2 Fx f þ x þ xs sgg
P2
V þ xs U
signf^1 ðsÞg ¼ signfP2 g signfFx g signf þ x þ xs sg; signfP2 g [ 0
P2
V þ xs U
signf^1 ðsÞg ¼ signfFx g signf þ x þ xs sg:
P2
2.4 Tunnel Diode (TD) as a Microwave Oscillator System

Cavity Parasitic Elements Stability Analysis
Tunnel diode is used in many engineering applications and specialy as a microwave

oscillator. Tunnel diode is the p-n junction device that exhibits negative resistance.
That means when the voltage is increased the current through it decreases. We can
consider the tunnel diode as an oscillator and high-frequency threshold (trigger)
device since it operated at frequencies far greater than the tetrode could, well into
the microwave bands. Applications for tunnel diodes included local oscillators for
UHF television tuners, trigger circuits in oscilloscopes, high-speed counter circuits,
and very fast-rise time pulse generator circuits. The tunnel diode can also be used as
low-noise microwave amplifier. The total current that flows through Tunnel Diode
(TD) is a summation of three elements, Itun, Idiode, Iexcess. Idiode is the P-N junction
VD
½ð Þ1
current Idiode Is e gVth ; Is—saturation current, η—ideality factor,
Vth = k ∙ T/q. VD—tunnel diode voltage. Itun is the tunnel current
VD m
½ð Þ
Itun ¼ VRD0 e V0 ; Typically m = 1…3, Vo = 0.1…0.5v, Ro is the tunnel diode
resistance in the ohmic region, VD is the tunnel diode voltage. Iexcess is an additional
VD VV
tunneling current related to parasitic tunneling via impurities Iexcess ¼ VRDV e½ Vex .
This current usually determines the minimum (valley) current Iv, RV and Vex are the
empirical parameters; in high quality diodes, RV RO, Vex = 1…5v. ID is the
V V
½ðgVD Þ1 ½ð VD Þm
tunnel diode current, ID ¼ Idiode þ Itun þ Iexcess ¼ Is e th þ VD
R0 e 0 þ VD
RV
V V
½ DVex V
e : If we build a circuit with the Tunnel Diode (TD) and resistor.
Case I: the circuit has three possible operating points. The middle point (inter-
section between TD’s characteristic NDR segment and work line) is typically
unstable, depending on parasitic L and C components. The circuit will operate at
one of the other two points (intersection between TD’s characteristic positive dif-
ferential resistance segment and work line).
Case II: The circuit has only one operating point (intersection between TD’s
characteristic NDR segment and work line). The total differential resistance is
negative because R < |Rd|. Depending on the L and C components, the circuit can
be stable (amplifier) or unstable (oscillator). Rd is a TD’s differential resistance
Rd ¼ @V @ID DID . R is a TD’s static resistance R ¼ ID . Typically for linear (Ohmic)
D DVD VD
components R = Rd and for semiconductor devices R 6¼ Rd [47–50]. The tunnel

diode has a region in its voltage current characteristic where the current decreases
with increased forward voltage, known as its negative resistance region. This
characteristic makes the tunnel diode useful in oscillators and as a microwave
amplifier (case II). The basic circuit structure of Tunnel Diode (TD) as a microwave
oscillator includes TD biased voltage and connection to microwave cavity. TD’s
biasing circuit is constructed from DC voltage Vb, inductor Lb which block
oscillation from Vb source and capacitor Cb which shorts to ground Vb bias volt-
age’s oscillations. The DC voltage Vb biases the tunnel diode TD into its negative
resistance region and also supplies the power used in amplifying the input signal Vi.
We have two topological circuit structures: first, switch S1 is ON and switch and S2
is OFF (growing or decreasing oscillations) and second, switch S1 is ON and S2 in
ON (bypass our signal source Vi) and our circuit functions as a oscillator.
A microwave cavity or radio frequency (RF) cavity is a special type of resonator,
consisting of a closed (or largely closed) metal structure that confines electro-
magnetic fields in the microwave region of the spectrum. The structure is either
hollow or filled with dielectric material. A microwave cavity acts similarly to a
resonant circuit with extremely low loss at its frequency of operation. Microwave
resonant cavities can be represented and thought of as simple LC circuits. For a
microwave cavity, the stored electric energy is equal to the stored magnetic energy
at resonance as is the case for a resonant LC circuit. We can represent our
microwave cavity as resonant LC circuit. Due to cavity parasitic effects there are
delays in the currents which flow through equivalent resonant L and C elements and
TD’s voltage derivative in time. s1 is the time delay for the TD’s voltage derivative
in time. s2 is the time delay for the current flows through C element. RL is the load
resistance. Load resistance (RL) is chosen so that RL < |Rd| in the TD’s charac-
teristic NDR region. At the TD operating point, the total circuit differential resis-
tance is negative. We have transient in resonant cavity after turning the bias voltage
to ON state (switch S1 moves to ON state, S2 is in OFF state) (Fig. 2.18).
2.4 Tunnel Diode (TD) as a Microwave Oscillator System Cavity … 223
Fig. 2.18 Tunnel diode as a microwave oscillator
If Rd > 0 or Rd < 0 and RL > |Rd| then oscillation goes down. If Rd < 0 and
RL < |Rd| then the oscillations goes up. The resonator circuit with NDR can
oscillate. Maximum frequency of TD-oscillator is limited by the characteristic
tunneling time fMAX ð2p
1
Þ ðstun
1
Þ, tunneling time in TDs is extremely small: 1ps,
fMAX > 100 GHz. We represent our Microwave resonant cavities thought of as
simple LC circuits. For a microwave cavity, the stored electric energy is equal to the
stored magnetic energy at resonance as is the case for a resonant LC circuit. In
terms of inductance and capacitance, the resonant frequency for a given mnl mode
can be written as Lmnl ¼ l kmnl2
Vcavity ; Cmnl ¼ k4 Ve cavity :
mnl
Vcavity is the cavity volume, kmnl is the mode wavenumber and e, l are per-
mittivity and permeability respectively. The resonant frequency (f) for a given mnl
mode can be written as fmnl ¼ 2ppffiffiffiffiffiffiffiffiffiffiffiffiffi
1
L C
¼ q
1 ffiffiffiffiffi
le
:
mnl mnl 2p
k2
mnl
We consider for our stability analysis the case of switch S1 is ON and S2 in ON

(bypass our signal source Vi) and our TD is functions as a microwave oscillator.
First we need to locate our TD’s voltages segment into negative resistance region
Rd ¼ @V
@ID \0. The expression for tunnel diode current is
D
X
3 V
½ðgVD Þ1 VD ½ðVVD Þm VD ½VDVVV
ID ¼ nðVD Þ ¼ gi ðVD Þ ¼ Is e th þ e 0 þ e ex ;
i¼1
R0 RV
V
½ðgVD Þ1
g1 ðVD Þ ¼ Is e th
VD ½ðVVD Þm VD ½VDVVV
g2 ðVD Þ ¼ e 0 ; g3 ðVD Þ ¼ e ex ;
R0 RV
@g1 ðVD Þ 1 @VD ½ðgVVD Þ1
¼ Is e th
@ID g Vth @ID
@g2 ðVD Þ 1 @VD ½ðVVD Þm VD VD 1 @VD V
½ð D Þm
¼ e 0 þ fm ð Þm1 g e V0
@ID R0 @ID R0 V0 V0 @ID
@g2 ðVD Þ VD m 1 @VD ½ðVVD Þm
¼ f1 m ð Þ g e 0 ;
@ID V0 R0 @ID
@g3 ðVD Þ VD 1 @VD ½VDVVV
¼ ð1 þ Þ e ex
@ID Vex RV @ID
X 3
@ID X 3
@gi X 3
@gi @g1 ðVD Þ @g2 ðVD Þ @g3 ðVD Þ
ID ¼ gi ðVD Þ ) ¼ ) ¼ þ þ ¼1
i¼1
@I D i¼1
@I D i¼1
@I D @ID @ID @ID
1 @VD ½ðgVVD Þ1 VD 1 @VD ½ðVVD Þm
Is e th þ f1 m ð Þm g e 0
g Vth @ID V0 R0 @ID
VD 1 @VD ½VDVVV
þ ð1 þ Þ e ex ¼ 1
Vex RV @ID
@VD 1 V
½ð D Þ1 VD 1 ½ðVD Þm VD 1 ½VDVVV
½Is e gVth þ f1 m ð Þm g e V0 þ ð1 þ Þ e ex ¼ 1
@ID g Vth V0 R0 Vex RV
@VD 1
¼
@ID Is gVth e th
V
½ðgVD Þ1 m
V
½ð D Þm
þ f1 m ð V0 Þ g R10 e V0 þ ð1 þ R1V e½
VD VV
Vex
Vex Þ
1 V D VD
@VD 1 V
½ð D Þ1 VD 1 ½ðVD Þm VD
\0 ) fIs e gVth þ f1 m ð Þm g e V0 þ ð1 þ Þ
@ID g Vth V0 R0 Vex
1 ½VDVVV
e ex g\0
RV
VV is TD’s characteristic valley voltage, VP is TD’s characteristic peak voltage.

It is numerical analysis to find our TD’s voltages segment into negative resis-
tance region Rd ¼ @V@ID \0 by specific TD’s parameters (VP \VD \VV ).
D
VB = VB1 = VB2 (S2 ON). VD ¼ VB VA ; VA ¼ VRL ¼ VC ¼ VL ; IRL ¼ VRAL
dVA dVC dIL VE VB3

IC ¼ C ¼C ; VA ¼ VL ¼ L ; I D ¼ I Lb ¼ I R b ; I R b ¼ ; IC b
dt dt dt Rb
dVE
¼ Cb jdVE ¼ 0; VE ¼ Vb
dt dt ¼0
dI
VLb ¼ VB3 VB ¼ Lb dtLb ¼ Lb dIdtD ; ID ¼ nðVD Þ; ID ¼ IRL þ IC þ IL ;
Vb—Constant supple voltage, VD—Tunnel diode voltage, VC—Capacitor voltage,
VL—Inductor voltage. Vi—Input voltage (is bypass in our case, S2 ON state).
Z
VA dVA 1
nðVD Þ ¼ þC þ VA dt;
RL dt L
Z
dIL VA 1
VA ¼ L ) dt ¼ dIL ) IL ¼ VA dt
dt L L
dnðVD Þ 1 dVA d 2 VA 1
¼ þ C 2 þ VA ;
dt RL dt dt L
dID
Vb ¼ VRb þ VLb þ VD þ VA ¼ ID Rb þ Lb þ VD þ VA
dt
dID dVA dID d 2 ID dVD
VA ¼ Vb ID Rb Lb VD ; jdVb ¼0 ¼ Rb Lb 2
dt dt dt dt dt dt
dVA dID d 2 ID dVD d 2 VA d 2 ID d 3 ID d 2 V D
jdVb ¼ Rb Lb 2 ) 2 ¼ Rb 2 Lb 3 2
dt dt ¼0 dt dt dt dt dt dt dt
We get the following expression:
dnðVD Þ 1 dID d 2 ID dVD d 2 ID d 3 ID d 2 VD

¼ ½Rb Lb 2 þ C ½Rb 2 Lb 3 2
dt RL dt dt dt dt dt dt
1 dID
þ ½Vb ID Rb Lb VD
L dt
dnðVD Þ 1 dID 1 d 2 ID 1 dVD d 2 ID d 3 ID

¼ Rb Lb 2 C Rb 2 C Lb 3
dt RL dt RL dt RL dt dt dt
2
d VD 1 1 1 dID 1
C 2 þ Vb ID Rb Lb VD
dt L L L dt L
dnðVD Þ 1 1 dID Rb Lb d 2 ID 1
¼ Vb ID Rb ½ þ 2 ½Lb þ C Rb
dt L L dt RL L dt RL
d 3 ID 1 1 dVD d 2 VD
C Lb 3 VD C 2
dt L RL dt dt
V V VD VV
½ðgVD Þ1 ½ð VD Þm
ID ¼ nðVD Þ ¼ Is e th þ VD
R0 e 0 þ VD
RV e½ Vex and we need to find
dnðVD Þ
dt :
dnðVD Þ 1 ½ð D Þ1 dVD

V 1 dVD ½ðVVD Þm
¼ Is e gVth þ e 0
dt g Vth dt R0 dt
VD VD 1 dVD V
½ð D Þm
þ ðm ½ m1 Þ e V0
R0 V0 V0 dt
1 dVD ½VDVVV VD 1 dVD ½VDVVV
þ e ex þ e ex
RV dt RV Vex dt

V 1 ½ðVVD Þm dVD
¼ Is e gVth þ e 0
dt g Vth dt R0 dt
VD VD m1 m ½ðVVD Þm dVD
½ e 0
R0 V0 V0 dt
1 ½VDVVV dVD VD VD VV dVD
þ e ex þ e½ Vex
RV dt RV Vex dt

V VD 1 ½ðVD Þm
¼ Is e gVth þ ð1 m ½ m Þ e V0
dt g Vth dt V0 R0
dVD VD 1 ½VDVVV dVD
þ ð1 þ Þ e ex
dt Vex RV dt
dnðVD Þ 1 V
½ð D Þ1 VD 1 ½ðVD Þm VD 1
¼ fIs e gVth þ ð1 m ½ m Þ e V0 þ ð1 þ Þ
dt g Vth V0 R0 Vex RV
VD VV dVD
e½ Vex g
dt
1 V
wðVD Þ ¼ Is e gVth þ ð1 m ½ m Þ e V0 þ ð1 þ Þ
g Vth V0 R0 Vex RV
VD VV
e½ Vex
DÞ dwðVD Þ
dID
dt ¼ dnðV
dt ¼ wðVD Þ dVdtD : We need to find the expression: dt .
dwðVD Þ 1 dVD ½ðgVVD Þ1 2 VD m1 1 dVD 1

V
½ð VD Þm
¼ Is e th m ½ e 0
dt ðg Vth Þ2 dt V0 V0 dt R0
VD 1 VD 1 dVD ½ðVVD Þm
ð1 m ½ m Þ m ð Þm1 e 0
V0 R0 V0 V0 dt
1 dVD 1 ½VDVVV
þ e ex
Vex dt RV
VD 1 1 dVD ½VDVVV
þ ð1 þ Þ e ex
Vex RV Vex dt
dwðVD Þ 1 ½ð D Þ1 dVD

V VD VD
¼ Is 2
e gVth fm þ ð1 m ½ m Þg m ð Þm1
dt ðg Vth Þ dt V0 V0
1 ½ð D Þ dVD
V m VD 1 dVD ½VDVVV
e V0 þ f1 þ ð1 þ Þg e ex
R0 V0 dt Vex Vex RV dt
dwðVD Þ 1 V
½ð D Þ1 VD
¼ fIs 2
e gVth f1 þ m ð1 ½ m Þg
dt ðg Vth Þ V0
VD m1 m VD m
½ð Þ
ð Þ e V0
V0 R0 V0
VD 1 VD VV dVD
þ f2 þ g e½ Vex g
Vex Vex RV dt
1 V
½ðgVD Þ1 VD m
w1 ðVD Þ ¼ Is 2
e th f1 þ m ð1 ½ Þg
ðg Vth Þ V0
VD m1 m V
½ð D Þm
ð Þ e V0
V0 R0 V0
VD 1 VD VV dwðVD Þ dVD
þ f2 þ g e½ Vex ; ¼ w1 ðVD Þ
Vex Vex RV dt dt
dID dnðVD Þ dVD d 2 ID d 2 nðVD Þ

¼ ¼ wðVD Þ ; 2 ¼
dt dt dt dt dt2
dwðVD Þ dVD d 2 VD
¼ ½ þ wðVD Þ ½ 2
dt dt dt
d 2 ID d 2 nðVD Þ dVD dVD d 2 VD

¼ ¼ w 1 ðVD Þ ½ þ wðV D Þ ½
dt2 dt2 dt dt dt2
2
dVD 2 d VD
¼ w1 ðVD Þ ½ þ wðVD Þ ½ 2
dt dt
We define the following new variables: X(t), Y(t). X ¼ dVdtD ; Y ¼ dVdtA

dnðVD Þ 1 dVA d 2 VA 1 dVD 1 dY 1

¼ þ C 2 þ VA ) wðVD Þ ¼ Y þC þ VA
dt RL dt dt L dt RL dt L
dVA dID d 2 ID dVD
jdVb ¼0 ¼ Rb Lb 2 )
dt dt dt dt dt
dVD dVD 2 d 2 VD
Y ¼ Rb wðVD Þ Lb fw1 ðVD Þ ½ þ wðVD Þ ½ 2 g X
dt dt dt
dVD dVD 2 d 2 VD
Y ¼ Rb wðVD Þ Lb w1 ðVD Þ ½ Lb wðVD Þ ½ 2 X
dt dt dt
dX
Y ¼ Rb wðVD Þ X Lb w1 ðVD Þ X Lb wðVD Þ
2
X
dt
dVD 1 dY 1 1 dY 1
wðVD Þ ¼ Y þC þ VA ) wðVD Þ X ¼ Y þC þ VA
dt RL dt L RL dt L
1 dY 1 dY wðVD Þ 1 1
wðVD Þ X ¼ Y þC þ VA ) ¼ X Y VA
RL dt L dt C RL C LC
dX
Y ¼ Rb wðVD Þ X Lb w1 ðVD Þ X 2 Lb wðVD Þ X )
dt
dX Rb Lb w1 ðVD Þ 2 1 1
¼ X X X Y
dt Lb Lb wðVD Þ Lb wðVD Þ Lb wðVD Þ
dX 1 1 w ðVD Þ 2 1
¼ ðRb þ Þ X 1 X Y
dt wðVD Þ Lb wðVD Þ Lb wðVD Þ
We can summery our Tunnel diode as a microwave oscillator system’s differ-

ential equations:
dX 1 1 w ðVD Þ 2 1
¼ ðRb þ Þ X 1 X Y
dY wðVD Þ 1 1 dVD dVA
¼ X Y VA ; ¼X; ¼Y
dt C RL C LC dt dt
To find our system equilibrium points (fixed points), we set dX dt ¼ 0; dt ¼ 0

dY

dt ¼ 0; dt ¼ 0. Then our fixed points are E ðX ; Y ; VD ; VA Þ ¼ ð0; 0; VD ; 0Þ:
dVD dVA
At equilibrium points: limt!1 Xðt s1 Þ ¼ limt!1 XðtÞ 8 t s1 ; limt!1 Yðt s2 Þ

¼ limt!1 YðtÞ 8 t s2 .
Under delay parameters we get the following Tunnel diode system’s Delay
Differential Equation (DDEs).
dt ; dt ;...
dX dY
Remark: Delay parameters don’t effect variables derivative in time
dX 1 1 w ðVD Þ 2 1
¼ ðRb þ Þ Xðt s1 Þ 1 X ðt s1 Þ Yðt s2 Þ
dY wðVD Þ 1 1
¼ Xðt s1 Þ Yðt s2 Þ VA ;
dt C RL C LC
dVD dVA
¼ Xðt s1 Þ ; ¼ Yðt s2 Þ
dt dt
equilibrium points of tunnel diode system consists in adding to its coordinates
½X Y VD VA arbitrarily small increments of exponential terms x y vD vA , and retaining
the first order terms in ½x y vD vA ekt . The system of four homogeneous equations
leads to a polynomial characteristics equation in the eigenvalue k. The polynomial
characteristics equations accept by set the below TD’s variables equations (delayed
and undelayed) into four tunnel diode equations. Tunnel diode’s fixed values with
arbitrarily small increments of exponential form ½x y vD vA ekt are: i = 0 (first
fixed point), i = 1 (second fixed point), i = 2 (third fixed point), etc. Under TD’s
variables X(t) and Y(t), delays in time: XðtÞ ! Xðt s1 Þ; YðtÞ ! Yðt s2 Þ
jIC ðtÞ¼CdVA ¼CYðtÞ ) IC ðtÞ ! IC ðt s2 Þ:
dt
ðiÞ
XðtÞ ¼ X ðiÞ þ x ekt ; YðtÞ ¼ Y ðiÞ þ y ekt ; VD ðtÞ ¼ VD þ vD ekt ;
ðiÞ
VA ðtÞ ¼ VA þ vA ekt
Xðt s1 Þ ¼ X ðiÞ þ x ekðts1 Þ ; Yðt s2 Þ ¼ Y ðiÞ þ y ekðts2 Þ for i ¼ 0; 1; 2; . . .
ðiÞ
Xðt ¼ 0Þ ¼ X ðiÞ þ x; Yðt ¼ 0Þ ¼ Y ðiÞ þ y; VD ðt ¼ 0Þ ¼ VD þ vD ;
ðiÞ
VA ðt ¼ 0Þ ¼ VA þ vA
Xðt s1 Þjt¼0 ¼ X ðiÞ þ x eks1 ; Yðt s2 Þjt¼0 ¼ Y ðiÞ þ y eks2 for i ¼ 0; 1; 2; . . .
for k < 0, t > 0 the selected fixed point is stable otherwise k > 0, t > 0 is Unstable.
Our system tends to the selected fixed point exponentially for k < 0, t > 0 other-
wise go away from the selected fixed point exponentially. k is the eigenvalue
parameter which establish if the fixed point is stable or Unstable, additionally his
absolute value (|k|) establish the speed of flow toward or away from the selected
fixed point [2–5] (Table 2.14).
The speeds of flow toward or away from the selected fixed point for TD system’s
variables are (k = 1, 2):
Table 2.14 Tunnel Diode (TD) as a microwave oscillator system eigenvalues options
k<0 k>0
t=0 Xðt ¼ 0Þ ¼ X ðiÞ þ x Xðt ¼ 0Þ ¼ X ðiÞ þ x
Yðt ¼ 0Þ ¼ Y ðiÞ þ y Yðt ¼ 0Þ ¼ Y ðiÞ þ y
ðiÞ ðiÞ
VD ðt ¼ 0Þ ¼ VD þ vD VD ðt ¼ 0Þ ¼ VD þ vD
ðiÞ ðiÞ
VA ðt ¼ 0Þ ¼ VA þ vA VA ðt ¼ 0Þ ¼ VA þ vA
ðiÞ jkjt ðiÞ
t>0 XðtÞ ¼ X þx e XðtÞ ¼ X þ x ejkjt
YðtÞ ¼ Y ðiÞ þ y ejkjt YðtÞ ¼ Y ðiÞ þ y ejkjt
ðiÞ ðiÞ
VD ðtÞ ¼ VD þ vD ejkjt VD ðtÞ ¼ VD þ vD ejkjt
ðiÞ ðiÞ
VA ðtÞ ¼ VA þ vA ejkjt VA ðtÞ ¼ VA þ vA ejkjt
for i ¼ 0; 1; 2; . . . for i ¼ 0; 1; 2; . . .
t!∞ Xðt ! 1Þ ¼ X ðiÞ Xðt ! 1Þ x ejkjt
Yðt ! 1Þ ¼ Y ðiÞ Yðt ! 1Þ y ejkjt
ðiÞ
VD ðt ! 1Þ ¼ VD VD ðt ! 1Þ vD ejkjt
VA ðt
ðiÞ
! 1Þ ¼ VA VA ðt ! 1Þ vA ejkjt
dXðtÞ Xðt þ DtÞ XðtÞ X ðiÞ þ x ekðt þ DtÞ ½XðtÞ þ x ekt ekDt 1 þ Dtk
¼ lim ¼ lim !
dt Dt!1 Dt Dt!1 Dt
x ekt ðekDt 1Þ x ekt ð1 þ k Dt 1Þ dYðtÞ
lim ¼ ¼ x k ekt ; ¼ y k ekt
Dt!1 Dt Dt dt
dVD ðtÞ dVA ðtÞ dXðt s1 Þ
¼ vD k ekt ; ¼ vA k ekt ; ¼ x k ekt eks1 ;
dt dt dt
dYðt s2 Þ
¼ y k ekt eks2
dt
First we take the Tunnel diode’s voltage derivative differential equations

(dXðtÞ
dt ¼ . . .). dt ¼ ðRb þ wðVD ÞÞ Lb X
First TD’s system differential equation: dX 1 1
w1 ðVD Þ
wðVD Þ X Y Lb wðV and adding to it’s coordinates ½X Y VD VA arbitrarily small
2 1
DÞ
increments of exponential form ½x y vD vA ekt and retaining the first order terms in
x y vD vA .
dX 1 1 w ðVD Þ 2 1
¼ ðRb þ Þ Xðt s1 Þ 1 X ðt s1 Þ Yðt s2 Þ
1 1
x k ekt ¼ ðRb þ ðiÞ
Þ ½X ðiÞ þ x ekðts1 Þ
wðV þ vD e Þ kt L b
D
ðiÞ
w1 ðVD þ vD ekt Þ 1
ðiÞ
½X ðiÞ þ x ekðts1 Þ 2 ½Y ðiÞ þ y ekðts2 Þ ðiÞ
wðVD þ vD ekt Þ Lb wðVD þ vD ekt Þ
1 V
wðVD Þ ¼ Is e gVth þ ð1 m ½ m Þ e V0 þ ð1 þ Þ
g Vth V0 R0 Vex RV
VD VV
e½ Vex
ðiÞ ðiÞ
VD ! VD þ vD ekt ) wðVD Þ ! wðVD þ vD ekt Þ
1 1
ðiÞ
¼ ðiÞ ðiÞ
½V þ vD ekt
wðVD þ vD ekt Þ 1 ½ð D gV Þ1 ½V þ vD ekt m
fIs e th þ ð1 m ½ D Þ
g Vth V0
ðiÞ
þ vD ekt m ðiÞ
1 ½ð½VD Þ ½VD þ vD ekt
e V0
ð1 þ Þ
R0 Vex
ðiÞ
þ vD ekt VV
1 ½½VD
e Vex g
RV
ðiÞ ðiÞ
ðiÞ
1
¼ 1
ðiÞ þ C1 ðVD ; vD ; . . .Þ: We need to find C1 ðVD ; vD ; . . .Þ
wðVD þ vD ekt Þ wðVD Þ
function.
1 1 ðiÞ ðiÞ
ðiÞ
¼ ðiÞ
þ C1 ðVD ; vD ; . . .Þ ) C1 ðVD ; vD ; . . .Þ
1 1
¼ ðiÞ
ðiÞ
1 1
ðiÞ
¼ ðiÞ ðiÞ ðiÞ
wðVD Þ V
½ðgVD Þ1
ðiÞ V
½ð VD Þm
ðiÞ V VV
R1V e½ Vex
VD VD D
Is gV1 th e th þ ð1 m ½ V0 m Þ R10 e 0 þ ð1 þ Vex Þ
ðiÞ ðiÞ ðiÞ

VD þ vD ekt m vD ekt m kt
; B ¼ vDVe0
V VD
Mathematical assumptions: ½ V0 ¼ ½ VD0 þ V0 ; A¼ V0
P
n
xi
m
e e e ¼ e i¼1 ;ðA þ
x1 x2 xn
BÞ
m m m
¼ Am þ Aðm1Þ B þ Aðm2Þ B2 þ þ
1 2 m1
A Bðm1Þ þ Bm
ðiÞ m ðiÞ ðiÞ ðiÞ

VD þ vD ekt VD vD ekt m VD m m V vD ekt
½ ¼½ þ ¼½ þ ½ D ðm1Þ ½
V0 V0 V0 V0 1 V0 V0
ðiÞ
m V vD ekt 2
þ ½ D ðm2Þ ½
2 V0 V0
ðiÞ
m V vD ekt ðm1Þ vD ekt m
þ þ ½ D ½ þ½
m1 V0 V0 V0
ðiÞ m ðiÞ ðiÞ ðiÞ

V þ vD ekt V vD ekt m V vD ekt m V
½ D ¼½ D þ ¼½ D þ ¼ ½ D m
V0 V0 V V0 V0 V0
! 0 !
ðiÞ kt ðiÞ
m V vD e m V v2 e2kt
þ ½ D ðm1Þ ½ þ ½ D ðm2Þ D 2
1 V0 V0 2 V0 V0
ðiÞ ðm1Þ
m V v eðm1Þkt De
vm mkt
þ þ ½ D D ðm1Þ
þ
m1 V0 V0 V0m
D 0 8 m
2 then
We consider vm
ðiÞ m ðiÞ ðiÞ

VD þ vD ekt V vD ekt m V
½ ¼½ D þ ¼ ½ D m
V0 V0 V0 V0
ðiÞ
m V vD ekt
þ ½ D ðm1Þ ½ þ ðe ! 0Þ
1 V0 V0

V
ðiÞ m V ðiÞ ðm1Þ v ekt
ðiÞ m
½f½ VD þ ½ VD ½ þ ðe!0Þg
½V þ vD ekt m
D
V0
½ð D Þ 0
1 0
e V0
¼e

m V
ðiÞ
vD ekt
V
ðiÞ f ½ VD ðm1Þ ½ V0 g
½ VD m 1 0
¼e 0 e
ðiÞ ðiÞ
½V þ vD ekt ½V ðiÞ ðiÞ
½ð D Þ1 ½ D 1 vD ekt ½V þ vD ekt VV V VV vD ekt
gVth gVth ½ gVth
; e½
¼ e½ Vex e½ Vex
D D
e ¼e e Vex
1 1
ðiÞ
¼ ðiÞ ðiÞ
½V þ vD ekt
wðVD þ vD ekt Þ 1 ½ð D gV Þ1 ½V þ vD ekt m
fIs e th þ ð1 m ½ D Þ
g Vth V0
ðiÞ
þ vD ekt m ðiÞ
1 ½ð½VD Þ ½VD þ vD ekt
e V0
þ ð1 þ Þ
R0 Vex
ðiÞ
þ vD ekt VV
1 ½½VD
e Vex g
RV
1 1
ðiÞ
¼ ðiÞ
wðVD þ vD ekt Þ 1
½V
v ekt
½ D 1

V
ðiÞ
½D
gVth
fIs e e gVth þ ð1 m f½ D m
g Vth V0
ðiÞ kt
m V vD e
þ ½ D ðm1Þ ½ gÞ
1 V0 V0

m V ðiÞ ðm1Þ v ekt
ðiÞ
V f ½ VD ½ DV g
1 ½ VD m 1 0 0
e 0 e
R0
ðiÞ ðiÞ
VD vD ekt 1 ½VDVVV ½vDVekt
þ fð1 þ Þþ g e ex e ex g
Vex Vex RV
1 1
ðiÞ
¼ ðiÞ
½V
½ D 1 v ekt V
ðiÞ
gVth ½D
fIs e e gVth þ ð1 m ½ D m Þ
g Vth V0

m V ðm1Þ v ekt
ðiÞ
ðiÞ
f ½ VD ½ DV g
1 ½VVD m 1 0 0
e 0 e
R0
ðiÞ
m V vD ekt
m ½ D ðm1Þ ½
1 V0 V0

ðiÞ
1 ½VVD m f 1 ½ V0 ½ V0 g
D D
e 0 e
R0
ðiÞ ðiÞ
VD 1 ½VDVVV ½vDVekt vD ekt
þ ð1 þ Þ e ex e ex þ
Vex RV Vex
ðiÞ
1 ½VDVVV ½vDVekt
e ex e ex g
RV
ðiÞ
We define the following global parameters: Bi ¼ Bi ðVD Þ 8 i ¼ 1; 2; 3;
ðiÞ
B4 ¼ B4 ðVD ; vD Þ
ðiÞ
½V ðiÞ ðiÞ
1 ½ D 1 V 1 ½VD m
gV
B1 ¼ Is e th ; B2 ¼ ð1 m ½ D m Þ e V0 ; B3
g Vth V0 R0
ðiÞ ðiÞ
VD 1 ½VDVVV
¼ ð1 þ Þ e ex
Vex RV
1 1
ðiÞ
¼
wðVD þ vD ekt Þ m V
ðiÞ
vD ekt
vD ekt f ½ VD ðm1Þ ½ V0 g vD ekt
½ gVth
þ B3 e½ Vex
0
fB1 e þ B2 e 1
ðiÞ
vD ekt 1 ½VDVVV ½vDVekt
þ e ex e ex
Vex RV
ðiÞ
m V vD ekt
m ½ D ðm1Þ ½
1 V0 V0

ðiÞ f ½ VD ½ DV g
V
1 ½ VD m 1 0 0
e 0 e g
R0
We define the following parameter:

ðiÞ ðiÞ ðiÞ
vD ekt 1 ½VDVVV ½vDVekt m VD ðm1Þ vD ekt 1 ½VVD m
B4 ¼ e ex e ex m ½ ½ e 0
Vex RV 1 V0 V0 R0
m VDðiÞ ðm1Þ vD ekt
f ½ V ½ V g
e 1 0 0
1 1
ðiÞ
¼ ;
wðVD þ vD ekt Þ m V
ðiÞ
v ekt
vD ekt f ½ VD ðm1Þ ½ DV g v ekt
½ gV 1 0 0
½ DVex
fB1 e th þ B2 e þ B3 e þ B4 g
1 1
¼ 3
ðiÞ
wðVD Þ P
Bk
k¼1
1 1 ðiÞ ðiÞ
ðiÞ
¼ ðiÞ
1 1
¼ ðiÞ
ðiÞ
ðiÞ 1
C1 ðVD ; vD ; . . .Þ ¼
m ðiÞ
V v ekt
vD ekt f ½ VD ðm1Þ ½ DV g vD ekt
½ 1 þ B3 e½ Vex
0 0
fB1 e gVth þ B2 e þ B4 g
1
3
P
Bk
k¼1

m ðiÞ
V vD ekt
P
3 v ekt f ½ VD ðm1Þ ½ V0 g vD ekt
½D
gVth 1
B3 e½ Vex B4
0
½ Bk B1 e
B2 e
ðiÞ
C1 ðVD ; vD ; . . .Þ ¼ k¼1

vD ekt f ½ V ½ V g vD ekt P
3
½ gV 1 þ B3 e½ Vex þ B4 Þ ½ Bk
0 0
ðB1 e th þ B2 e
k¼1
Binomial coefficients:
Y
n k n iþ1 n ðn 1Þ ðn k þ 1Þ
¼ ¼ j n¼m
k i¼1 k! k! k¼1
Y1
¼ i¼1
ðn i þ 1Þj n¼m ¼ m
k¼1

n n! m
¼ 8 0 k n; n ¼ m & k ¼ 1 ) ¼ m;
k k!ðn kÞ! 1
Y
n k
n iþ1
¼
k i¼1
k!

m ðiÞ
V vD ekt
P
3 vD ekt f ½ VD ðm1Þ ½ V0 g vD ekt
½ gV 1
B3 e½ Vex B4
0
½ Bk B1 e
B2 e th
ðiÞ
C1 ðVD ; vD ; . . .Þ ¼
k¼1

vD e kt f ½ V ½ V g vD ekt P
3
½ 1 þ B3 e½ Vex þ B4 Þ ½ Bk
0 0
ðB1 e gVth þ B2 e
k¼1
½
vD ekt
gVth
vD ekt 1 v2 e2kt vD ekt
e ¼ 1þ þ D þ jvk 08 k
2 ¼ 1 þ þ ðe ! 0Þ
g Vth 2 ½g Vth 2 D g Vth
vD ekt
1þ
g Vth

m VDðiÞ ðm1Þ vD ekt ðiÞ
f ½ V ½ V g
m V vD ekt 1
1 ½ D ðm1Þ ½
0 0
e ¼1 þ
1 V0 V0 2
2 ðiÞ
m V v 2
e 2kt
½ D 2ðm1Þ ½ D 2 þ
1 V0 V0

ðiÞ
f
m V
½ VD ðm1Þ ½
vD ekt ðiÞ
V0 g VD ðm1Þ vD ekt
m
e 1 0
¼1 ½ ½ þ ðe ! 0Þ
V0 1 V0
ðiÞ
m V vD ekt
1 ½ D ðm1Þ ½
1 V0 V0
vD ekt vD ekt 1 v2D e2kt vD ekt vD ekt

e½ Vex ¼ 1þ þ 2
þ ¼ 1þ þ ðe ! 0Þ 1 þ
Vex 2 Vex Vex Vex
X ðiÞ
3
vD ekt m VD ðm1Þ vD ekt
½ Bk B1 ½1 þ B2 f1 ½ ½ g
k¼1
g Vth 1 V0 V0
vD ekt
B3 ½1 þ B4
ðiÞ Vex
C1 ðVD ; vD ; ::::Þ ¼
vD ekt f ½ V ½ V g vD ekt P
3
½ gV 1 þ B3 e½ Vex þ B4 Þ ½ Bk
0 0
ðB1 e th þ B2 e
k¼1
ðiÞ
kt m
½ VD0 ðm1Þ ½vDVe0 B3 vDVeex B4
kt kt
D e V
B1 vgV þ B 2
ðiÞ
th 1
C1 ðVD ; vD ; . . .Þ ¼
v e kt f ½ V ½ V g vD ekt P 3
½D
gVth 1 þ B3 e½ Vex þ B4 Þ ½ Bk
0 0
ðB1 e þ B2 e
k¼1
ðiÞ
B1 þ B2 ½ ½
g Vth 1 V0 V0
vD ekt
B3 B4
ðiÞ Vex
C1 ðVD ; vD ; ::::Þ ¼
m V
ðiÞ
vD ekt
½ gV
þ B3 e½ Vex
0
ðB1 e th þ B2 e 1 Þ
X
3 X
3
½ Bk þ B4 ½ Bk
k¼1 k¼1

m ðiÞ
V vD ekt
½ gVth 1 þ B3 e½ Vex
0
We define: X4 ðvD Þ ¼ B1 e þ B2 e
ðiÞ
m
½ VD0 ðm1Þ ½vDVe0 B3 vDVeex B4
kt kt kt
D e V
B1 vgV þ B2
ðiÞ
th 1
C1 ðVD ; vD ; . . .Þ ¼
P
3 P
3
X4 ðvD Þ ½ Bk þ B4 ½ Bk
k¼1 k¼1
ðiÞ ðiÞ ðiÞ

vD ekt 1 ½VDVVV ½vDVekt m VD ðm1Þ vD ekt 1 ½VVD m
B4 ¼ e ex e ex m ½ ½ e 0
Vex RV 1 V0 V0 R0
f ½ V ½ V g
e 1 0 0
We define the following variables: X1 ðvD Þjv2 0 ¼ X2 ðvD Þjv2 0 ¼ vD

D D
vD ekt vD ekt v2 e2kt

X1 ðvD Þ ¼ vD e½ Vex
’; vD ð1 þ Þ ¼ vD þ D 2 jv2 0 ¼ vD
Vex Vex D

f ½ VD ½ DV g ðiÞ
m V vD ekt
½ D ðm1Þ ½
0 0
X2 ðvD Þ ¼ vD e 1 ¼ vD ð1 Þ
1 V0 V0
ðiÞ
m V v2 ekt
¼ vD ½ D ðm1Þ ½ D
1 V0 V0

f ½ VD ½ D
V0 g
0
X2 ðvD Þ ¼ vD e 1 jv2 0 ¼ vD
D
ðiÞ ðiÞ ðiÞ

ekt 1 ½VDVVV VD ðm1Þ ekt 1 ½VVD m
m
B4 ¼ e ex X1 ðvD Þ m ½ e 0 X2 ðvD Þ
½
Vex RV 1 V0 V0 R0
ðiÞ
V VV

1 m
e½ Vex m
D
B4 jX1 ðvD Þj 2 ¼X2 ðvD Þjv2 0 ¼vD ¼f
v 0
D D Vex RV 1
ðiÞ V
ðiÞ
VD ðm1Þ 1 ½ D m
½ e V0 g vD ekt
V0 V0 R0
V VV
ðiÞ ðiÞ
ðiÞ 1 m V 1
e½ Vex m ½ D ðm1Þ
D
X3 ðVD Þ ¼
Vex RV 1 V0 V0 R0
ðiÞ
V
½ VD m
e 0 ; B4 jX1 ðvD Þj 2 ¼X2 ðvD Þjv2 0 ¼vD
v 0
D D
ðiÞ kt
¼ X3 ðVD Þ vD e

m ðiÞ
V vD ekt
v ekt f ½ VD ðm1Þ ½ V0 g vD ekt
½D
gVth 1 þ B 3 e½Vex
0
X4 ðvD Þ ¼ B1 e þ B2 e
ðiÞ
X4 ðvD Þ ¼ B1 ½1 þ þ B2 f1 ½ ½ g þ B3
g Vth 1 V0 V0
vD ekt
½1 þ
Vex
X ðiÞ
3
vD ekt m V vD ekt vD ekt
X4 ðvD Þ ¼ ½ Bi þ B1 B2 ½ D ðm1Þ ½ þ B3
i¼1
g Vth 1 V0 V0 Vex
X 3 ðiÞ
B1 m V 1 B3
X4 ðvD Þ ¼ ½ Bi þ f B2 ½ D ðm1Þ ½ þ g vD ekt
i¼1
g V th 1 V 0 V 0 V ex
ðiÞ
ðiÞ m
½ VD0 ðm1Þ ½V10 þ
V
We define: X5 ðVD Þ ¼ B1
gVth B2 B3
Vex
1
X
3
ðiÞ
X4 ðvD Þ ¼ ½ Bi þ X5 ðVD Þ vD ekt
i¼1

m ðiÞ
ðiÞ
½ VD0 ðm1Þ ½vDVe0 B3 vDVeex X3 ðVD Þ vD ekt
kt kt kt
D e V
B1 vgV þ B2
ðiÞ
th
1
C1 ðVD ; vD ; . . .Þ ¼
P
3
ðiÞ P
3
ðiÞ P
3
f½ Bi þ X5 ðVD Þ vD ekt g ½ Bk þ X3 ðVD Þ vD ekt ½ Bk
i¼1 k¼1 k¼1

ðiÞ
m ðiÞ
½ VD0 ðm1Þ V10 B3 V1ex X3 ðVD Þg vD ekt
V
fB1 gV1 th þ B2
ðiÞ 1
C1 ðVD ; vD ; . . .Þ ¼
P
3 P3
ðiÞ ðiÞ
½ Bi 2 þ ½ Bk fX5 ðVD Þ þ X3 ðVD Þg vD ekt
i¼1 k¼1
For simplicity we define the following global parameters:

ðiÞ
ðiÞ 1 m V 1 1 ðiÞ
X6 ðVD Þ ¼ B1 þ B2 ½ D ðm1Þ B3 X3 ðVD Þ
g Vth 1 V0 V0 Vex
ðiÞ ðiÞ ðiÞ ðiÞ
X7 ðVD Þ ¼ X5 ðVD Þ þ X3 ðVD Þ; C1 ðVD ; vD ; . . .Þ
ðiÞ
X6 ðVD Þ vD ekt
¼ 3
P P3
ðiÞ
½ Bi 2 þ ½ Bk X7 ðVD Þ vD ekt
i¼1 k¼1
ðiÞ
ðiÞ X6 ðVD Þ vD ekt
C1 ðVD ; vD ; . . .Þ ¼
P
3 P3
ðiÞ
½ Bk f½ Bk þ X7 ðVD Þ vD ekt g
k¼1 k¼1
P
3
ðiÞ
f½ Bk X7 ðVD Þ vD ekt g
k¼1

P3
ðiÞ
f½ Bk X7 ðVD Þ vD ekt g
k¼1
P
3
ðiÞ Q
7
ðiÞ
½ Bk X6 ðVD Þ vD ekt ½ Xi ðVD Þ v2D e2kt
ðiÞ k¼1 i¼6
C1 ðVD ; vD ; . . .Þ ¼
P
3 P
3
ðiÞ
½ Bk f½ Bk 2 X27 ðVD Þ v2D e2kt g
k¼1 k¼1
We consider v2D ! e 0
ðiÞ
ðiÞ X6 ðVD Þ 1
C1 ðVD ; vD ; . . .Þjv2 !e0 ¼ vD ekt ;
D P
3 ðiÞ
wðVD þ vD ekt Þ
½ Bk 2
k¼1
ðiÞ
1 X6 ðVD Þ
¼ þ vD ekt
wðVD Þ
ðiÞ P
3
½ Bk 2
k¼1
We can summery our system global parameters in the below Table 2.15:
Table 2.15 Tunnel Diode (TD) as a microwave oscillator system global parameters
Global parameter Expression
ðiÞ ðiÞ
B1 ðVD Þ ½V
½ D 1
gVth
Is 1
gVth e
ðiÞ ðiÞ
B2 ðVD Þ V
ðiÞ V
½ VD m
ð1 m ½ VD0 m Þ R10 e 0
ðiÞ ðiÞ
B3 ðVD Þ VD
ðiÞ V VV
½ DVex
ð1 þ Vex Þ R1V e
B4 jX1 ðvD Þj 2 ðiÞ
V VV
ðiÞ V
ðiÞ
¼X2 ðvD Þjv2 0 ¼vD m ½ VD m
fVex1RV e½ Vex ½ VD0 ðm1Þ V01R0 e
V
g vD ekt
D
v 0
D D m 0
1
ðiÞ
B4 jX1 ðvD Þj 2 ¼X2 ðvD Þjv2 0 ¼vD ¼ X3 ðVD Þ vD ekt
v 0
D D
X1 ðvD Þjv2 0 ¼ X2 ðvD Þjv2 0 vD

D D
X3 ðVD Þ V VV
m V
½ D m
e½ Vex ½ VD0 ðm1Þ V01R0 e V0
D V
1
Vex RV m
1
ðiÞ
X4 ðvD Þ P3 m
½ VD0 ðm1Þ ½V10 þ VBex3 g vD ekt
V
½ Bi þ fgV B1
B2
i¼1
th 1
ðiÞ
ðiÞ
X5 ðVD Þ B1
B
m
½ VD0 ðm1Þ ½V10 þ VBex3
V
gVth 2
1
ðiÞ
ðiÞ
X6 ðVD Þ B1 gV1 th þ B2
m
½ VD0 ðm1Þ V10 B3 V1ex X3 ðVD Þ
V ðiÞ
1
ðiÞ ðiÞ ðiÞ
X7 ðVD Þ X5 ðVD Þ þ X3 ðVD Þ
ðiÞ ðiÞ
w1 ðVD þ vD ekt Þ w1 ðVD Þ ðiÞ ðiÞ
ðiÞ ¼ ðiÞ þ C2 ðVD ; vD ; . . .Þ. We need to find C2 ðVD ; vD ; . . .Þ
function
ðiÞ
þ vD ekt ðiÞ
1 ½ð
V
D Þ1 VD þ vD ekt m
Is 2
e gVth
f1 þ m ð1 ½ Þg
ðg Vth Þ V0
ðiÞ ðiÞ
þ vD ekt m
VD þ vD ekt m1 m V
½ð D Þ
ð Þ e V0
V0 R0 V0
ðiÞ ðiÞ
VD þ vD ekt 1 V þ vD ekt VV
e½
D
ðiÞ þ f2 þ g Vex
w1 ðVD þ vD ekt Þ Vex V R
ðiÞ
¼ ex V
wðVD þ vD ekt Þ f
½ V ½ V g
vD ekt vD ekt
½ gV 1 þ B3 e½ Vex þ B4 g
0 0
fB1 e th þ B2 e
Under the Taylor series high order elements tend to zero (. . . vkD ! e 8 k
2)
ðiÞ
assumption for exponent functions: B4 jX1 ðvD Þj 2 ¼X2 ðvD Þjv2 0 ¼vD ¼ X3 ðVD Þ vD ekt
v 0
D D
ðiÞ
1 V vD ekt
½gVD 1 ½ gVth
Is 2
e th e
ðg Vth Þ
ðiÞ
VD þ vD ekt m
f1 þ m ð1 ½ Þg
V0
ðiÞ ðiÞ
þ vD ekt m
V þ vD ekt m1 m V
½ð D Þ
ð D Þ e V0
V0 R0 V0
ðiÞ ðiÞ
VD vD ekt 1 V VV vD ekt
e½ Vex e½ Vex
D
ðiÞ þ fð2 þ Þþ g
w1 ðVD þ vD ekt Þ Vex Vex V R
ðiÞ
¼ ex V
wðVD þ vD ekt Þ m V ðiÞ ðm1Þ v ekt
v ekt f ½ VD ½ DV g
½D 1 0 0
fB1 e gVth
þ B2 e
vD ekt
ðiÞ
þ B3 e½ Vex þ X3 ðVD Þ vD ekt g
ðiÞ m ðiÞ ðiÞ

V þ vD ekt V VD ðm1Þ vD ekt
m
½ D ½ D m þ ½ ½
V0 V0 V0 1 V0
ðiÞ ðiÞ ðiÞ
VD þ vD ekt m1 VD m1 m1 V vD ekt
ð Þ ½ þ ½ D ðm2Þ ½
V0 V0 1 V0 V0

V
ðiÞ m V ðiÞ ðm1Þ v ekt m V ðiÞ ðm1Þ vD ekt
ðiÞ m
½½ VD ½ VD ½ ðiÞ ½ VD ½
V þ vD ekt m V0
D
V0 V
½ð D Þ 0
1 0 ½ VD m 1 0
e V0
e ¼e 0 e
V
ðiÞ
vD ekt
ðiÞ
1 ½ D 1 ½ gVth
VD m
Is 2
e gVth e f1 þ m ð1 ½
ðg Vth Þ V0
ðiÞ
m V vD ekt
½ D ðm1Þ ½ Þg
1 V0 V0
ðiÞ ðiÞ
V m1 V
f½ D m1 þ ½ D ðm2Þ
V0 1 V0

ðiÞ ½ VD ½ DV
kt
vD e m V
½
D m
1 0 0
½ g e V0 e
V0 R0 V0
ðiÞ ðiÞ
V 1 V VV vD ekt
e½ Vex e½ Vex
D
þ ð2 þ D Þ
Vex Vex RV
ðiÞ
vD ekt 1 V VV vD ekt
e½ Vex e½ Vex
D
ðiÞ þ
w1 ðVD þ vD ekt Þ Vex Vex RV
ðiÞ
¼
wðVD þ vD ekt Þ m V ðiÞ ðm1Þ vD ekt
vD ekt f ½ VD ½ V0 g
½ gVth 1 0
fB1 e þ B2 e
v ekt
½ DVex ðiÞ
þ B3 e þ X3 ðVD Þ vD ekt g

m ðiÞ
V vD ekt
vD ekt f ½ VD ðm1Þ ½ V0 g
½ gVth 1 0
We already define: X4 ðvD Þ ¼ B1 e þ B2 e þ
vD ekt
½ Vex
B3 e
P ðiÞ
And X4 ðvD Þ ¼ ½ 3i¼1 Bi þ X5 ðVD Þ vD ekt

m ðiÞ
V vD ekt
v ekt f ½ VD ðm1Þ ½ V0 g vD ekt
½D
gVth 1 ðiÞ
þ B 3 e½ Vex þ X3 ðVD Þ vD ekt
0
B1 e þ B2 e
ðiÞ
¼ X4 ðvD Þ þ X3 ðVD Þ vD ekt
ðiÞ
X3
ðiÞ ðiÞ
X4 ðvD Þ þ X3 ðVD Þ vD ekt ¼ ½ Bi þ X5 ðVD Þ vD ekt þ X3 ðVD Þ vD ekt
i¼1
X3
ðiÞ ðiÞ
¼½ Bi þ fX5 ðVD Þ þ X3 ðVD Þg vD ekt
i¼1
V
ðiÞ
vD ekt
ðiÞ
1 ½ D 1 ½ gVth
VD m
Is 2
ðg Vth Þ V0
!
ðiÞ
m V vD ekt
½ D ðm1Þ ½ Þg
1 V0 V0
ðiÞ ðiÞ
VD m1 m1 V
f½ þ ½ D ðm2Þ
V0 1 V0

ðiÞ
kt ½ VD ½ DV
vD e m V
½
D m
1 0 0
½ g e V0 e
V0 R0 V0
ðiÞ ðiÞ
VD 1 V VV vD ekt
e½ Vex e½ Vex
D
þ ð2 þ Þ
Vex Vex RV
ðiÞ
vD ekt 1 V VV vD ekt
e½ Vex e½ Vex
D
ðiÞ þ
w1 ðVD þ vD ekt Þ Vex Vex RV
¼
ðiÞ
wðVD þ vD ekt Þ P3
ðiÞ ðiÞ
½ Bi þ fX5 ðVD Þ þ X3 ðVD Þg vD ekt
i¼1
We need to get an expression of the above equation’s numerator.

V
ðiÞ
vD ekt
ðiÞ
ðiÞ 1 ½ D 1 ½ gVth
VD m
w1 ðVD þ vD ekt Þ ¼ Is 2
ðg Vth Þ V0
!
ðiÞ
m V vD ekt
½ D ðm1Þ ½ Þg
1 V0 V0
ðiÞ ðiÞ
VD m1 m1 V vD ekt
f½ þ ½ D ðm2Þ ½ g
V0 1 V0 V0

V
ðiÞ ½ VD ½ DV ðiÞ
m ½ VD m 1 0 0 V
e 0 e þ ð2 þ D Þ
R0 V0 Vex
ðiÞ ðiÞ
1 V VV vD ekt vD ekt 1 V VV vD ekt
e½ Vex e½ Vex þ e½ Vex e½ Vex
D D

Vex RV Vex Vex RV
V
ðiÞ
vD ekt
ðiÞ
ðiÞ 1 ½ D 1 ½ VD m
gVth
w1 ðVD þ vD ekt Þ ¼ Is 2
e gVth e Þ f1 þ m ð1 ½
ðg Vth Þ V0
!
ðiÞ ðiÞ
m V vD ekt V m1
m ½ D ðm1Þ ½ g f½ D m1 þ
1 V0 V0 V0 1
ðiÞ
VD ðm2Þ vD ekt
½ ½ g
V0 V0

V
ðiÞ ½ VD ½ DV ðiÞ
m ½ VD m 1 0 0 V
e 0 e þ ð2 þ D Þ
R0 V0 Vex
ðiÞ ðiÞ
1 V VV vD ekt vD ekt 1 V VV vD ekt
e½ Vex e½ Vex þ e½ Vex e½ Vex
D D

Vex RV Vex Vex RV
First we take the internal expression’s multiplication:
ðiÞ ðiÞ ðiÞ

VD m m V vD ekt V
f½1 þ m ð1 ½ Þ m ½ D ðm1Þ ½ g f½ D m1
V0 1 V0 V0 V0
ðiÞ kt ðiÞ ðiÞ
m1 V vD e V V
þ ½ D ðm2Þ ½ g ¼ ½1 þ m ð1 ½ D m Þ ½ D m1
1 V0 V0 V0 V0
ðiÞ ðiÞ
V m1 V vD ekt
þ ½1 þ m ð1 ½ D m Þ ½ D ðm2Þ ½
V0 1 V0 V0
ðiÞ ðiÞ
V m V vD ekt m m1
½ D m1 m ½ D ðm1Þ ½ m
V0 1 V0 V0 1 1
ðiÞ ðiÞ
VD ðm1Þ VD ðm2Þ v2D e2kt
½ ½ ½
V0 V0 V02
We consider . . .v2D ! e ¼ 0 then

ðiÞ ðiÞ
m m1 V V v2 e2kt
m ½ D ðm1Þ ½ D ðm2Þ ½ D 2 ! e
1 1 V0 V0 V0
ðiÞ ðiÞ ðiÞ

VD m m V vD ekt V
f½1 þ m ð1 ½ Þ m ½ D ðm1Þ ½ g f½ D m1
V0 1 V0 V0 V0
ðiÞ ðiÞ ðiÞ
m1 V vD ekt V V
þ ½ D ðm2Þ ½ g ¼ ½1 þ m ð1 ½ D m Þ ½ D m1
1 V0 V0 V0 V0
ðiÞ ðiÞ
V m1 V vD ekt
þ ½1 þ m ð1 ½ D m Þ ½ D ðm2Þ ½
V0 1 V0 V0
ðiÞ ðiÞ
VD m1 m V vD ekt
½ m ½ D ðm1Þ ½ ðe ! 0Þ
V0 1 V0 V0

V V V m1 VD ðm2Þ
½1 þ m ð1 ½ D m Þ ½ D m1 þ f½1 þ m ð1 ½ D m Þ ½
V0 V0 V0 1 V0
ðiÞ ðiÞ
V m V vD ekt
½ D m1 m ½ D ðm1Þ g ½
V0 1 V0 V0
ðiÞ ðiÞ
ðiÞ V V
N1 ðVD Þ ¼ ½1 þ m ð1 ½ D m Þ ½ D m1 ;
V0 V0
ðiÞ ðiÞ
ðiÞ VD m m1 V
N2 ðVD Þ ¼ ½1 þ m ð1 ½ Þ ½ D ðm2Þ
V0 1 V0
ðiÞ ðiÞ
V m V ðiÞ ðiÞ vD ekt
½ D m1 m ½ D ðm1Þ ) N1 ðVD Þ þ N2 ðVD Þ ½ ;
V0 1 V0 V0
ðiÞ ðiÞ
N1 ¼ N1 ðVD Þ; N2 ¼ N2 ðVD Þ
ðiÞ
ðiÞ kt 1 V
½gVD 1 ½
vD ekt
gVth
vD ekt
w1 ðVD þ vD e Þ ¼ Is 2
e th e fN1 þ N2 ½ g
ðg Vth Þ V0

m V
ðiÞ
vD ekt
V
ðiÞ ½ VD ðm1Þ ½ V0
ðiÞ
m ½ VD m 1 0 VD
e 0 e þ ð2 þ Þ
R0 V0 Vex
ðiÞ ðiÞ
1 V VV vD ekt 1 V VV vD ekt
e½ Vex e½ Vex þ 2 e½ Vex e½ Vex vD ekt
D D

Vex RV Vex RV
ðiÞ
V
½ D 1
We define for simplicity the following parameters: N3 ¼ Is ðgV1 Þ2
e gVth
th
V
ðiÞ
V VV ðiÞ V VV
ðiÞ ðiÞ
m ½ D m V 1 1
e½ Vex ; N6 ¼ 2 e½ Vex
D D
N4 ¼ e V0 ; N5 ¼ ð2 þ D Þ
R0 V0 Vex Vex RV Vex RV

m V
ðiÞ
vD ekt
½ VD ðm1Þ ½ V0
ðiÞ kt
v ekt
½D
gVth
vD ekt 1 0
w1 ðVD þ vD e Þ ¼ N3 e fN1 þ N2 ½ g N4 e
V0
vD ekt vD ekt
ðiÞ ðiÞ
þ N5 e ½ Vex þ N6 e ½ Vex vD ekt ; N3 ¼ N3 ðVD Þ; N4 ¼ N4 ðVD Þ;
ðiÞ ðiÞ
N5 ¼ N5 ðVD Þ; N6 ¼ N6 ðVD Þ
Under the Taylor series high order elements tend to zero (. . . vkD ! e 8 k
2)
vD ekt vD ekt
½ gVth vD ekt vD ekt
assumption for exponent functions: e 1þ gVth ; e½Vex ¼ 1þ Vex

m V
ðiÞ
vD ekt
!
½ VD ðm1Þ ½ ðiÞ
V0 m VD ðm1Þ vD ekt
e 1 0
1 ½ ½
1 V0 V0

ðiÞ vD ekt vD ekt m
w1 ðVD þ vD ekt Þ ¼ N3 ½1 þ fN1 þ N2 ½ g N4 f1
g Vth V0 1
ðiÞ
VD ðm1Þ vD ekt vD ekt vD ekt
½ ½ g þ N5 ½1 þ þ N6 ½1 þ vD ekt
V0 V0 Vex Vex

ðiÞ kt vD ekt m
w1 ðVD þ vD e Þ ¼ N 3 þ N 3 fN4 N1 þ N4 ðN2 N1
g Vth 1
ðiÞ kt
ðiÞ
V vD e m V v2 e2kt
½ D ðm1Þ Þ N4 N2 ½ D ðm1Þ ½ D 2 g
V0 V0 1 V0 V0
vD ekt v2 e2kt
þ N5 þ N5 þ N6 vD ekt þ N6 D
Vex Vex
All expressions which includes v2D tend to zero (v2D ! e).

ðiÞ vD ekt m
w1 ðVD þ vD ekt Þ ¼ N3 þ N3 N4 N1 N4 ðN2 N1
g Vth 1
ðiÞ
V vD ekt vD ekt
½ D ðm1Þ Þ þ N5 þ N5 þ N6 vD ekt
V0 V0 Vex

ðiÞ kt 1 m
w1 ðVD þ vD e Þ ¼ fN5 þ N3 N4 N1 g þ fN3 ðN2 N1
g Vth 1
ðiÞ
VD ðm1Þ N4 N5
½ Þ þ þ N6 g vD ekt ;
V0 V0 Vex
ðiÞ
w1 ðVD þ vD ekt Þ ¼ U1 þ U2 vD ekt
ðiÞ ðiÞ
We define: U1 ¼ N5 þ N3 N4 N1 ; U1 ¼ U1 ðVD Þ; U2 ¼ U2 ðVD Þ
!
ðiÞ
1 m VD ðm1Þ N4 N5
U2 ¼ N3 ðN2 N1 ½ Þ þ þ N6
g Vth 1 V0 V0 Vex
ðiÞ
w1 ðVD þ vD ekt Þ U1 þ U2 vD ekt
¼
ðiÞ
wðVD þ vD ekt Þ P
3
ðiÞ ðiÞ
i¼1
ðiÞ ðiÞ
w1 ðVD þ vD ekt Þ w1 ðVD Þ ðiÞ ðiÞ
ðiÞ
¼ ðiÞ
ðiÞ ðiÞ
w1 ðVD þ vD ekt Þ w1 ðVD Þ
¼ ðiÞ
ðiÞ
ðiÞ
ðiÞ U1 þ U2 vD ekt w1 ðVD Þ
C2 ðVD ; vD ; ::::Þ ¼
P
3
ðiÞ ðiÞ wðVD Þ
ðiÞ
i¼1
ðiÞ ðiÞ ðiÞ
U1 wðVD Þ þ U2 wðVD Þ vD ekt w1 ðVD Þ
X3
ðiÞ ðiÞ
ð½ Bi þ fX5 ðVD Þ þ X3 ðVD ÞgÞ vD ekt
i¼1
¼
X
3
ðiÞ ðiÞ ðiÞ
½ Bi wðVD Þ þ fX5 ðVD Þ þ X3 ðVD Þg
i¼1
ðiÞ
wðVD Þ vD ekt
ðiÞ ðiÞ ðiÞ
U1 wðVD Þ þ fU2 wðVD Þ w1 ðVD Þ
X
3
ðiÞ ðiÞ
ð½ Bi þ fX5 ðVD Þ þ X3 ðVD ÞgÞg vD ekt
ðiÞ i¼1
C2 ðVD ; vD ; ::::Þ ¼
P
3
½ Bi wðVD Þ þ fX5 ðVD Þ þ X3 ðVD Þg wðVD Þ vD ekt
i¼1
ðiÞ
We define for simplicity the following global parameters: U3 ¼ U1 wðVD Þ
ðiÞ ðiÞ
X3
ðiÞ ðiÞ
U4 ¼ U2 wðVD Þ w1 ðVD Þ ð½ Bi þ fX5 ðVD Þ þ X3 ðVD ÞgÞ;
i¼1
X
3
ðiÞ
U5 ¼ ½ Bi wðVD Þ
i¼1
ðiÞ ðiÞ ðiÞ ðiÞ U3 þ U4 vD ekt

U6 ¼ fX5 ðVD Þ þ X3 ðVD Þg wðVD Þ; C2 ðVD ; vD ; . . .Þ ¼
U5 þ U6 vD ekt
ðiÞ U3 þ U4 vD ekt U5 U6 vD ekt

C2 ðVD ; vD ; . . .Þ ¼
U5 þ U6 vD ekt U5 U6 vD ekt
U3 U5 U3 U6 vD ekt þ U5 U4 vD ekt U4 U6 v2D e2kt
¼
U25 U26 v2D e2kt
All expressions which includes v2D tend to zero (v2D ! e).
ðiÞ U3 U5 þ ðU5 U4 U3 U6 Þ vD ekt

C2 ðVD ; vD ; . . .Þ ¼
U25
U3 ðU5 U4 U3 U6 Þ
¼ þ vD ekt
U5 U25

U3 ¼ U3 ðVD Þ; U4 ¼ U4 ðVD Þ; U5 ¼ U5 ðVD Þ; U6 ¼ U6 ðVD Þ
ðiÞ ðiÞ
w1 ðVD þ vD ekt Þ w1 ðVD Þ U3 ðU5 U4 U3 U6 Þ
ðiÞ
¼ ðiÞ
þ þ vD ekt
wðVD þ vD ekt Þ wðVD Þ U5 U25
We can summery our last analysis in the next Table 2.16):
Table 2.16 Tunnel Diode (TD) as a microwave oscillator system expressions and equivalent
expressions
Expression Equivalent expression
1 ðiÞ ðiÞ
X ðVD Þ ðiÞ X ðVD Þ
ðiÞ
wðVD þ vD ekt Þ
1
ðiÞ þ P
6
3 vD ekt ; C1 ðVD ; vD ; . . .Þjv2 !e0 ¼ P
6
3 vD ekt
wðVD Þ 2 D
½ Bk ½ Bk 2
k¼1 k¼1
ðiÞ
w1 ðVD þ vD ekt Þ ðiÞ ðiÞ
w1 ðVD Þ U3 ðVD Þ
ðiÞ
wðVD þ vD ekt Þ ðiÞ
þ ðiÞ
wðVD Þ U5 ðVD Þ
ðU5 ðVD Þ U4 ðVD Þ U3 ðVD Þ U6 ðVD ÞÞ
þ ðiÞ
U25 ðVD Þ
vD ekt
ðiÞ
ðiÞ U3 ðVD Þ
C2 ðVD ; vD ; . . .Þjv2 !e0 ¼ ðiÞ
D
U5 ðVD Þ
þ ðiÞ
U25 ðVD Þ
kt
vD e
1 1
x k ekt ¼ ðRb þ ðiÞ
Þ ½X ðiÞ þ x ekðts1 Þ
wðVD þ vD ekt Þ Lb
ðiÞ
w1 ðVD þ vD ekt Þ
ðiÞ
½X ðiÞ þ x ekðts1 Þ 2 ½Y ðiÞ þ y ekðts2 Þ
wðVD þ vD ekt Þ
1
ðiÞ
Lb wðVD þ vD ekt Þ
ðiÞ
1 X6 ðVD Þ 1
x k ekt ¼ ðRb þ f þ vD ekt gÞ ½X ðiÞ þ x ekðts1 Þ
wðVD Þ
ðiÞ P
3 Lb
½ Bk 2
k¼1
ðiÞ ðiÞ
w ðVD Þ U3 ðVD Þ
f 1 ðiÞ þ ðiÞ
wðVD Þ U5 ðVD Þ
þ ðiÞ
vD ekt g
U25 ðVD Þ
ðiÞ kðts1 Þ 2 ðiÞ kðts2 Þ
½X þx e ½Y þy e
ðiÞ
1 1 X6 ðV Þ
f þ 3 D vD ekt g
Lb wðV ðiÞ Þ P
D ½ Bk 2
k¼1
½X ðiÞ þ x ekðts1 Þ 2 ¼ ½X ðiÞ 2 þ 2 X ðiÞ x ekðts1 Þ þ x2 e2kðts1 Þ jx2 0

¼ ½X ðiÞ 2 þ 2 X ðiÞ x ekðts1 Þ
ðiÞ
1 X6 ðVD Þ 1
x k ekt ¼ ðRb þ f þ vD ekt gÞ ½X ðiÞ þ x ekðts1 Þ
wðVD Þ
ðiÞ P
3 Lb
½ Bk 2
k¼1
ðiÞ ðiÞ
f 1 ðiÞ þ ðiÞ
wðVD Þ U5 ðVD Þ
þ ðiÞ
vD ekt g
U25 ðVD Þ
1
f½X ðiÞ 2 þ 2 X ðiÞ x ekðts1 Þ g ½Y ðiÞ þ y ekðts2 Þ
Lb
ðiÞ
1 X6 ðVD Þ
f þ vD ekt g
wðVD Þ
ðiÞ P
3
2
½ Bk
k¼1
1 1 1 1
x k ekt ¼ ½Rb þ ðiÞ
X ðiÞ ½Rb þ ðiÞ
x ekðts1 Þ
wðVD Þ Lb wðVD Þ Lb
ðiÞ ðiÞ
X ðiÞ X6 ðVD Þ X6 ðVD Þ
3 vD ekt 3 vD x ekðts1 Þ ekt
Lb P 2 P 2
½ Bk ½ Bk Lb
k¼1 k¼1
w ðVD Þ U3 ðVD Þ w ðVD Þ U3 ðVD Þ
½ 1 ðiÞ þ ðiÞ
½X ðiÞ 2 ½ 1 ðiÞ þ ðiÞ
2 X ðiÞ x ekðts1 Þ
wðVD Þ U5 ðVD Þ wðVD Þ U5 ðVD Þ
ðU5 ðVD Þ U4 ðVD Þ U3 ðVD Þ U6 ðVD ÞÞ ½X ðiÞ 2
ðiÞ
vD ekt
U25 ðVD Þ
ðU5 ðVD Þ U4 ðVD Þ U3 ðVD Þ U6 ðVD ÞÞ 2 X ðiÞ
ðiÞ
U25 ðVD Þ
ðiÞ
Y ðiÞ 1 Y ðiÞ X6 ðVD Þ
vD x ekðts1 Þ ekt vD ekt
Lb wðV ðiÞ Þ Lb P 3
2
D ½ Bk
k¼1
ðiÞ
1 1 X6 ðV Þ
y ekðts2 Þ 3 D vD y ekðts2 Þ ekt
Lb wðVD Þ
ðiÞ L b P
½ Bk 2
k¼1
We consider vD x 0; vD y 0
1 1 1 1
X ðiÞ ½Rb þ ðiÞ
x ekðts1 Þ
wðVD Þ Lb wðVD Þ Lb
ðiÞ ðiÞ ðiÞ
X ðiÞ X6 ðVD Þ w ðVD Þ U3 ðVD Þ
3 vD ekt ½ 1 ðiÞ þ ½X ðiÞ 2
Lb P 2 wðV Þ U ðV
ðiÞ
Þ
½ Bk D 5 D
k¼1
ðiÞ ðiÞ
½ 1 ðiÞ þ ðiÞ
wðVD Þ U5 ðVD Þ
ðiÞ
U25 ðVD Þ
ðiÞ
Y ðiÞ 1 Y ðiÞ X6 ðVD Þ 1
vD ekt vD ekt y ekðts2 Þ
Lb wðV ðiÞ Þ Lb P 3
2 Lb wðVD Þ
ðiÞ
D ½ Bk
k¼1
ðiÞ
11 w ðVD Þ Y ðiÞ 1
X ðiÞ 1 ðiÞ
½X ðiÞ 2
wðVD Þ Lb wðVD Þ Lb wðV ðiÞ Þ
D
ðiÞ ðiÞ
1 1 X ðiÞ X6 ðVD Þ U3 ðVD Þ
½Rb þ x ekðts1 Þ 3 vD ekt ½X ðiÞ 2
ðiÞ
wðVD Þ Lb Lb P ðiÞ
U5 ðVD Þ
½ B k 2
k¼1
ðiÞ ðiÞ
w1 ðVD Þ U3 ðVD Þ
½ ðiÞ
þ ðiÞ
wðVD Þ U5 ðVD Þ
ðiÞ
U25 ðVD Þ
ðiÞ
Y ðiÞ X6 ðVD Þ 1
vD ekt 3 vD ekt y ekðts2 Þ
Lb P 2 L wðV
ðiÞ
Þ
½ Bk b D
k¼1
ðiÞ
w1 ðVD Þ
At fixed point: ðRb þ 1
ðiÞ Þ L1b X ðiÞ ðiÞ ½X ðiÞ 2 Y ðiÞ 1
ðiÞ ¼0
wðVD Þ wðVD Þ Lb wðVD Þ
ðiÞ
1 1 X ðiÞ X6 ðVD Þ
x k ekt ¼ ½Rb þ x ekðts1 Þ
3 vD ekt
ðiÞ
wðVD Þ Lb Lb P 2
½ Bk
k¼1
ðiÞ
U3 ðVD Þ 1
ðiÞ
½X ðiÞ 2 ðiÞ
y ekðts2 Þ
U5 ðVD Þ Lb wðVD Þ
ðiÞ ðiÞ
½ 1 ðiÞ þ ðiÞ
wðVD Þ U5 ðVD Þ
ðiÞ
U25 ðVD Þ
ðiÞ
Y ðiÞ X6 ðVD Þ
vD ekt 3 vD ekt
Lb P
½ Bk 2
k¼1
ðiÞ
U3 ðVD Þ
We need to choose the right parameters which give ðiÞ ½X ðiÞ 2 ¼ 0 since
U5 ðVD Þ
there is no ekt multiplication term. We already approve our fixed points are
ðiÞ
U3 ðVD Þ
E ðX ; Y ; VD ; VA Þ ¼ ð0; 0; VD ; 0Þ. X ðiÞ ¼ 0 ) ðiÞ ½X ðiÞ 2 ¼ 0
U5 ðVD Þ
ðiÞ
1 1 X ðiÞ X6 ðVD Þ
x k ekt ¼ ½Rb þ x ekt eks1
3 vD ekt
ðiÞ
wðVD Þ Lb Lb P 2
½ B k
k¼1
ðiÞ ðiÞ
1 w1 ðVD Þ U3 ðVD Þ
ðiÞ
y ekt eks2 ½ ðiÞ
þ ðiÞ
2 X ðiÞ x ekt eks1
Lb wðVD Þ wðVD Þ U5 ðVD Þ
ðiÞ
vD ekt
U25 ðVD Þ
ðiÞ
Y ðiÞ X6 ðVD Þ
3 vD ekt
Lb P 2
½ Bk
k¼1
Dividing two side of above equation by ekt gives:

ðiÞ
1
1 X ðiÞ X6 ðVD Þ 1
x k ¼ ½Rb þ x eks1
3 vD y eks2
ðiÞ
wðVD Þ Lb Lb P 2 L wðV
ðiÞ
Þ
½ B b D
k
k¼1
ðiÞ ðiÞ
½ 1 ðiÞ þ ðiÞ
2 X ðiÞ x eks1
wðVD Þ U5 ðVD Þ
ðiÞ
U25 ðVD Þ
ðiÞ
Y ðiÞ X6 ðVD Þ
vD 3 vD
Lb P
½ Bk 2
k¼1
ðiÞ ðiÞ
1
1 w ðVD Þ U3 ðVD Þ
fk ð½Rb þ ðiÞ
þ ½ 1 ðiÞ þ ðiÞ
2 X ðiÞ Þ eks1 g x
wðVD Þ Lb wðVD Þ U5 ðVD Þ
ðiÞ
1 X ðiÞ X6 ðVD Þ
eks2 y f 3
Lb wðVD Þ
ðiÞ Lb P
½ B k 2
k¼1
þ ðiÞ
U25 ðVD Þ
ðiÞ
Y ðiÞ X6 ðVD Þ
þ 3 g vD ¼ 0
Lb P
½ B k 2
k¼1

ðiÞ ðiÞ
P1 ¼ P1 ðVD ; X ; . . .Þ
ðiÞ ðiÞ
1 1 w ðVD Þ U3 ðVD Þ
P1 ¼ ½Rb þ ðiÞ
þ ½ 1 ðiÞ
þ ðiÞ
2 X ðiÞ ; P2
wðVD Þ Lb wðVD Þ U5 ðVD Þ
ðiÞ
¼ P2 ðVD ; X ðiÞ ; Y ðiÞ ; . . .Þ
ðiÞ
X ðiÞ X6 ðVD Þ
P2 ¼ 3
Lb P
½ B k 2
k¼1
þ ðiÞ
U25 ðVD Þ
ðiÞ
Y ðiÞ X6 ðVD Þ
þ 3
Lb P
½ B k 2
k¼1
1
fk P1 eks1 g x ðiÞ
eks2 y P2 vD ¼ 0
Lb wðVD Þ
wðVD Þ
Second TD’s system differential equation: dY dt ¼ C X RL C Y LC VA and
1 1
adding to it’s coordinates ½X Y VD VA arbitrarily small increments of exponential

form ½x y vD vA ekt and retaining the first order terms in x y vD vA .
ðiÞ
y k ekt ¼ ½X ðiÞ þ x ekðts1 Þ ½Y ðiÞ þ y ekðts2 Þ
C RL C
1 ðiÞ
½VA þ vA ekt
LC
ðiÞ ðiÞ ðiÞ
wðVD þ vD ekt Þ ¼ wðVD Þ þ C3 ðVD ; vD ; . . .Þ
ðiÞ ðiÞ ðiÞ
) C3 ðVD ; vD ; . . .Þ ¼ wðVD þ vD ekt Þ wðVD Þ
ðiÞ
We already approve B4 jX1 ðvD Þj 2 ¼X2 ðvD Þjv2 0 ¼vD ¼ X3 ðVD Þ vD ekt
v 0
D D

m V
ðiÞ
vD ekt
ðiÞ ½ gVth
þ vD ekt Þ ¼ B1 e þ B 3 e½ Vex
0
wðVD þ B2 e 1 þ B4

m V
ðiÞ
v ekt
vD ekt f ½ VD ðm1Þ ½ DV g vD ekt
ðiÞ ½ gVth
wðVD þ vD ekt Þ ¼ B1 e þ B 3 e½ Vex
0 0
þ B2 e 1
ðiÞ
þ X3 ðVD Þ vD ekt
ðiÞ
X3
ðiÞ
wðVD þ vD ekt Þ ¼ ½ Bk þ X7 ðVD Þ vD ekt
k¼1
ðiÞ
X3
ðiÞ ðiÞ
C3 ðVD ; vD ; . . .Þ ¼ ½ Bk þ X7 ðVD Þ vD ekt wðVD Þ
k¼1
X3
ðiÞ ðiÞ
¼ f½ Bk wðVD Þg þ X7 ðVD Þ vD ekt
k¼1
ðiÞ
X
3
ðiÞ 1 ½ D 1
gV
½V
V

1 ½VD m
ðiÞ ðiÞ
½ Bk wðVD Þ ¼ Is e th þ ð1 m ½ D m Þ e V0
k¼1
g Vth V0 R0
ðiÞ ðiÞ ðiÞ
VD 1 ½VDVVV 1 V
½ð D Þ1
þ ð1 þ Þ e ex fIs e gVth
Vex RV g Vth
VD m 1 ½ðVVD Þm V 1 ½VDVVV
þ ð1 m ½ Þ e 0 þ ð1 þ D Þ e ex g ¼ 0
V0 R0 Vex RV
ðiÞ ðiÞ ðiÞ

C3 ðVD ; vD ; . . .Þ ¼ X7 ðVD Þ vD ekt ) wðVD þ vD ekt Þ
ðiÞ ðiÞ
¼ wðVD Þ þ X7 ðVD Þ vD ekt
ðiÞ ðiÞ
½wðVD Þ þ X7 ðVD Þ vD ekt 1
y k ekt ¼ ½X ðiÞ þ x ekðts1 Þ ½Y ðiÞ þ y
C RL C
1 ðiÞ
ekðts2 Þ ½VA þ vA ekt
LC
ðiÞ ðiÞ
wðVD Þ X7 ðVD Þ vD ekt 1
y k ekt ¼ ½ þ ½X ðiÞ þ x ekðts1 Þ Y ðiÞ
C C RL C
1 1 ðiÞ 1
y ekðts2 Þ V vA ekt
RL C LC A LC
ðiÞ ðiÞ ðiÞ

wðVD Þ ðiÞ wðVD Þ X ðiÞ X7 ðVD Þ vD ekt
y k ekt ¼ X þ x ekðts1 Þ þ
C C C
ðiÞ kðts1 Þ kt
X7 ðVD Þ vD x e e 1
þ Y ðiÞ
C RL C
1 1 ðiÞ 1
y ekðts2 Þ VA vA ekt
RL C LC LC
We consider vD x 0
ðiÞ ðiÞ ðiÞ

wðVD Þ ðiÞ wðVD Þ X ðiÞ X7 ðVD Þ vD ekt
y k ekt ¼ X þ x ekðts1 Þ þ
C C C
1 ðiÞ 1 kðts2 Þ 1 ðiÞ 1
Y ye VA vA ekt
RL C RL C LC LC
ðiÞ ðiÞ
wðVD Þ ðiÞ 1 1 ðiÞ wðVD Þ
y k ekt ¼ X Y ðiÞ V þ x ekðts1 Þ
C RL C LC A C
ðiÞ
X ðiÞ X7 ðVD Þ vD ekt 1 1
þ y ekðts2 Þ vA ekt
C RL C LC
ðiÞ
wðVD Þ ðiÞ
At fixed point: C X ðiÞ RL1C Y ðiÞ LC
1
VA ¼ 0
ðiÞ ðiÞ
wðVD Þ X ðiÞ X7 ðVD Þ vD ekt 1
y k ekt ¼ x ekðts1 Þ þ y ekðts2 Þ
C C RL C
1
vA ekt
LC
ðiÞ ðiÞ
wðVD Þ X ðiÞ X7 ðVD Þ 1 1
yk¼ x eks1 þ vD y eks2 vA
C C RL C LC
ðiÞ ðiÞ
wðVD Þ X ðiÞ X7 ðVD Þ 1 1
x eks1 y k þ vD y eks2 vA ¼ 0
C C RL C LC
ðiÞ ðiÞ
wðVD Þ 1 X ðiÞ X7 ðVD Þ 1
x eks1 y k y eks2 þ vD vA ¼ 0
C RL C C LC
Third TD’s system differential equation: dVdtD ¼ Xðt s1 Þ and adding to it’s
coordinates ½X Y VD VA arbitrarily small increments of exponential form ½x y vD vA
ekt and retaining the first order terms in x y vD vA [9, 10].
vD k ekt ¼ X ðiÞ þ x ekðts1 Þ . At fixed point: Xðt s1 Þjts1 ¼ XðtÞ )

ðiÞ
X ¼ 0.
vD k ekt ¼ x ekt eks1 ) x eks1 vD k ¼ 0. Fourth TD’s system dif-
ferential equation: dVdtA ¼ Yðt s2 Þ and adding to it’s coordinates ½X Y VD VA
arbitrarily small increments of exponential form ½x y vD vA ekt and retaining the
first order terms in x y vD vA . vA k ekt ¼ Y ðiÞ þ y ekðts2 Þ . At fixed point:
Yðt s2 Þjt [ [ s2 ¼ YðtÞ ) Y ðiÞ ¼ 0; vA k ekt ¼ Y ðiÞ þ y ekðts2 Þ ) vA k ekt
¼ y ekðts2 Þ vA k ekt ¼ y ekðts2 Þ ) y eks2 vA k ¼ 0.
We summery our TD system’s four characteristic equations in the eigenvalue k
with delays:
ðiÞ 1 ðiÞ
fk P1 ðVD Þ eks1 g x ðiÞ
eks2 y P2 ðVD Þ vD ¼ 0
Lb wðVD Þ
ðiÞ ðiÞ
wðVD Þ 1 X ðiÞ X7 ðVD Þ 1
eks1 x k y eks2 y þ vD vA ¼ 0
C RL C C LC
x eks1 k vD ¼ 0; y eks2 k vA ¼ 0
The small increments Jacobian of our Gradostat system is as bellow:

0 1 0 x 1
!11 . . . !14
B . B y C
B . .. .. C
C B C ðiÞ
@ . . . AB C ¼ 0; !11 ¼ k P1 ðVD Þ eks1 ;
@ vD A
!41 !44 v A
1 ks2
!12 ¼ ðiÞ
e
Lb wðVD Þ
ðiÞ
ðiÞ wðVD Þ ks1 1
!13 ¼ P2 ðVD Þ; !14 ¼ 0; !21 ¼ e ; !22 ¼ k eks2
C RL C
ðiÞ
X ðiÞ X7 ðVD Þ 1
!23 ¼ ; !24 ¼ ; !31 ¼ eks1 ; !32 ¼ 0; !33 ¼ k; !34 ¼ 0
C LC
!41 ¼ 0; !42 ¼ eks2 ; !43 ¼ 0; !44 ¼ k
0 1
!11 ... !14
B .. .. .. C
AkI ¼@ . . . A; det jA k Ij ¼ 0
!41 !44
0 ðiÞ
1
ks2 X ðiÞ X7 ðVD Þ
B k RL C e LC
1 1
ðiÞ
C C
det jA k Ij ¼ ½k þ P1 ðVD Þ eks1
detB
@ 0 k 0 A
C
eks2 0 k
0 ðiÞ ðiÞ ðiÞ
1
wðVD Þ X X7 ðVD Þ
B C eks1 C LC
1
C
1
þ ðiÞ
e ks2
detB
@ eks1 k 0 A
C
Lb wðVD Þ
0 0 k
0 ðiÞ
1
wðVD Þ
eks1 k RL1C eks2 LC
1
ðiÞ B C C
P2 ðVD Þ det B ks1 C
@ e 0 0 A
0 eks2 k
ðiÞ 1 1
det jA k Ij ¼ ½k þ P1 ðVD Þ eks1 fðk þ eks2 Þ k2 eks2 kg
RL C LC
ðiÞ ðiÞ
1 wðVD Þ ks1 2 X ðiÞ X7 ðVD Þ ks1
þ ðiÞ
eks2 f e k þ e kg
Lb wðVD Þ C C
ðiÞ 1 1
P2 ðVD Þ fðk þ eks2 Þ eks1 k ekðs1 þ s2 Þ g
RL C LC
ðiÞ ðiÞ
Dðk; s1 ; s2 Þ ¼ k4 þ fk3 P1 ðVD Þ þ k2 P2 ðVD Þg eks1
ðiÞ
1 1 1 P1 ðVD Þ 1
þ ðk3 þ k2 Þ eks2 þ fk2 ½ þ
RL C LC C RL Lb
P
2
1 P1 ðVD Þ X ðiÞ X7 ðVD Þ P2 ðVD Þ P2 ðVD Þ k si

þk ½ þ þ þ g e i¼1
C L ðiÞ
Lb wðVD Þ RL LC
We have three stability cases: (1) s1 ¼ s; s2 ¼ 0 (2) s1 ¼ 0; s2 ¼ s

(3) s1 ¼ s2 ¼ s.
We need to get characteristics equations for all above stability analysis cases. We
study the occurrence of any possible stability switching resulting from the increase
of value of the time delay s for the general characteristic equation Dðk; sÞ. If we
choose parameter then Dðk; sÞ ¼ Pn ðk; sÞ þ Qm ðk; sÞ eks ; n [ m:
X
n
Pn ðk; sÞ ¼ pk ðsÞ kk ¼ p0 ðsÞ þ p1 ðsÞ k þ p2 ðsÞ k2 þ p3 ðsÞ k3 þ . . .
k¼0
Xm
Qm ðk; sÞ ¼ qk ðsÞ kk ¼ q0 ðsÞ þ q1 ðsÞ k þ q2 ðsÞ k2 þ q3 ðsÞ k3 þ . . .
k¼0
1 1 ðiÞ ðiÞ
Dðk; s1 ¼ s; s2 ¼ 0Þ ¼ k4 þ k3 þ k2 þ fk3 P1 ðVD Þ þ k2 P2 ðVD Þ
RL C LC
ðiÞ ðiÞ
1 P1 ðVD Þ 1 1 P1 ðVD Þ
þ k2 ½ þ þk ½
C RL Lb C L
ðiÞ ðiÞ ðiÞ
X ðiÞ X7 ðVD Þ P2 ðVD Þ P2 ðVD Þ
þ þ þ g eks
Lb
ðiÞ
wðVD Þ RL LC
ðiÞ ðiÞ 1 1
Dðk; s1 ¼ 0; s2 ¼ sÞ ¼ k4 þ k3 P1 ðVD Þ þ k2 P2 ðVD Þ þ fk3 þ k2
RL C LC
ðiÞ ðiÞ
þ k2 ½ þ þk ½
C RL Lb C L
ðiÞ ðiÞ ðiÞ
þ þ þ g eks
Lb
ðiÞ
wðVD Þ RL LC
ðiÞ 1 ðiÞ 1
Dðk; s1 ¼ s2 ¼ sÞ ¼ k4 þ fk3 ½P1 ðVD Þ þ þ k2 ½P2 ðVD Þ þ g eks
RL C LC
ðiÞ ðiÞ
þ fk2 ½ þ þk ½
C RL Lb C L
ðiÞ ðiÞ ðiÞ
þ þ þ g e2ks
Lb
ðiÞ
wðVD Þ RL LC
Under Taylor series approximation: eks 1 k s. The Maclaurin series is a

eks 1 k s. e2ks eks ð1 k sÞ
ðiÞ 1 ðiÞ 1
Dðk; s1 ¼ s2 ¼ sÞ ¼ k4 þ fk3 ½P1 ðVD Þ þ þ k2 ½P2 ðVD Þ þ g eks
RL C LC
ðiÞ ðiÞ
þ fk2 ½ þ þk ½
C RL Lb C L
ðiÞ ðiÞ ðiÞ
þ þ þ g ð1 k sÞ eks
Lb wðVD Þ
ðiÞ RL LC
ðiÞ ðiÞ 1 ðiÞ ðiÞ 1

B1 ðVD Þ ¼ P1 ðVD Þ þ ; B2 ðVD Þ ¼ P2 ðVD Þ þ ;
RL C LC
ðiÞ
ðiÞ 1 P1 ðVD Þ 1
B3 ðVD Þ ¼ ½ þ
C RL Lb
ðiÞ 1 P1 ðVD Þ X ðiÞ X7 ðVD Þ P2 ðVD Þ ðiÞ P2 ðVD Þ
B4 ðVD Þ ¼ ½ þ þ ; B5 ðVD Þ ¼
C L Lb wðVD Þ
ðiÞ RL LC
ðiÞ ðiÞ ðiÞ

Dðk; s1 ¼ s2 ¼ sÞ ¼ k4 þ fk3 B1 ðVD Þ þ k2 B2 ðVD Þg eks þ fk2 B3 ðVD Þ
ðiÞ ðiÞ
þ k B4 ðVD Þ þ B5 ðVD Þg ð1 k sÞ eks
ðiÞ ðiÞ ðiÞ

Dðk; s1 ¼ s2 ¼ sÞ ¼ k4 þ fk3 B1 ðVD Þ þ k2 B2 ðVD Þg eks þ fk2 B3 ðVD Þ
ðiÞ ðiÞ ðiÞ
þ k B4 ðVD Þ þ B5 ðVD Þg eks k3 B3 ðVD Þ s eks
ðiÞ ðiÞ
k2 B4 ðVD Þ s eks B5 ðVD Þ k s eks
ðiÞ ðiÞ
X3
ðiÞ
Dðk; s1 ¼ s2 ¼ sÞ ¼ k4 þ fk3 ½B1 ðVD Þ B3 ðVD Þ s þ k2 ½ Bk ðVD Þ
k¼2
B4 ðVD Þ s þ k ½B4 ðVD Þ B5 ðVD Þ s þ B5 ðVD Þg eks
(Table 2.17)
Table 2.17 Tunnel Diode (TD) as a microwave oscillator system Pn ðk; sÞ and Qm ðk; sÞ functions
vs s1 and s2 options
s1 ¼ s; s2 ¼ 0 s1 ¼ 0; s2 ¼ s
Pn ðk; sÞ k þk
4 3 1
RL C þk
2 1
LC
ðiÞ
k4 þ k3 P1 ðVD Þ þ k2 P2 ðVD Þ
ðiÞ
Qm ðk; sÞ k3 P1 ðVD Þ
ðiÞ 1 1
k3 þ k2 f
ðiÞ RL C LC
ðiÞ 1 P1 ðVD Þ 1 ðiÞ
þ k2 fP2 ðVD Þ þ ½ þ g 1 P1 ðVD Þ 1
C RL Lb þ ½ þ g
ðiÞ ðiÞ C RL Lb
1 P1 ðVD Þ X ðiÞ X7 ðVD Þ ðiÞ
þk ½ þ ðiÞ 1 P1 ðVD Þ
C L Lb wðVD Þ þk ½
C L
ðiÞ ðiÞ
P2 ðVD Þ P2 ðVD Þ ðiÞ
X ðiÞ X7 ðVD Þ
þ þ þ
RL LC ðiÞ
Lb wðVD Þ
ðiÞ ðiÞ
P2 ðVD Þ P2 ðVD Þ
þ þ
RL LC
n 4 4
m 3 3
Status n>m n>m
We analyze the TD’s system stability for the third case s1 ¼ s2 ¼ s.
ðiÞ ðiÞ
X3
ðiÞ
k¼2
ðiÞ ðiÞ ðiÞ
B4 ðVD Þ s þ k ½B4 ðVD Þ B5 ðVD Þ s
ðiÞ
þ B5 ðVD Þg eks ; n ¼ 4; m ¼ 3; n [ m
ðiÞ ðiÞ
X3
ðiÞ
Pn ðk; sÞ ¼ k4 ; Qm ðk; sÞ ¼ k3 ½B1 ðVD Þ B3 ðVD Þ s þ k2 ½ Bk ðVD Þ
k¼2
B4 ðVD Þ s þ k ½B4 ðVD Þ B5 ðVD Þ s þ B5 ðVD Þ
ðiÞ
p0 ðsÞ ¼ p1 ðsÞ ¼ p2 ðsÞ ¼ p3 ðsÞ ¼ 0; p4 ðsÞ ¼ 1; q0 ðsÞ ¼ B5 ðVD Þ;
ðiÞ ðiÞ
q1 ðsÞ ¼ B4 ðVD Þ B5 ðVD Þ s
X
3
q2 ðsÞ ¼ Bk ðVD Þ B4 ðVD Þ s; q3 ðsÞ ¼ B1 ðVD Þ B3 ðVD Þ s
k¼2
The homogeneous system for X Y VD VA leads to a characteristic equation for the

P
4
eigenvalue k having the form Pðk; sÞ þ Qðk; sÞ eks ¼ 0; Pðk; sÞ ¼ aj k j
j¼0
P
3
Qðk; sÞ ¼ cj k j And the coefficients faj ðqi ; qk ; sÞ; cj ðqi ; qk ; sÞg 2 R depend on
j¼0
qi ; qk and delay s. qi ; qk are any TD’s parameters, other parameters kept as a
ðiÞ
constant a0 ðsÞ ¼ a1 ðsÞ ¼ a2 ðsÞ ¼ a3 ðsÞ ¼ 0; a4 ðsÞ ¼ 1; c0 ðsÞ ¼ B5 ðVD Þ
ðiÞ ðiÞ
X
3
ðiÞ ðiÞ
c1 ðsÞ ¼ B4 ðVD Þ B5 ðVD Þ s; c2 ðsÞ ¼ Bk ðVD Þ B4 ðVD Þ s; c3 ðsÞ
k¼2
ðiÞ ðiÞ
¼ B1 ðVD Þ B3 ðVD Þ s
a0 + c0 6¼ 0 for 8 qi ; qk 2 R þ , i.e. k = 0 is not a of PðkÞ þ QðkÞ eks ¼ 0.
requirements of the analysis [BK] can also be verified in the present case:
polynomial in x.
(d) Each positive root xðqi ; qk Þ of F(x) = 0 is continuous and differentiable
We assume that Pn ðk; sÞ and Qm ðk; sÞ can’t have common imaginary roots. That
is for any real number x; pn ðk ¼ i x; sÞ þ Qm ðk ¼ i x; sÞ 6¼ 0.
pn ðk ¼ i x; sÞ ¼ x4
X3
ðiÞ ðiÞ ðiÞ
Qm ðk ¼ i x; sÞ ¼ x2 ½ Bk ðVD Þ B4 ðVD Þ s þ B5 ðVD Þ
k¼2
þ i fx ½B4 ðVD Þ B5 ðVD Þ s x3 ½B1 ðVD Þ B3 ðVD Þ sg
X3
ðiÞ ðiÞ ðiÞ
pn ðk ¼ i x; sÞ þ Qm ðk ¼ i x; sÞ ¼ x4 x2 ½ Bk ðVD Þ B4 ðVD Þ s þ B5 ðVD Þ
k¼2
þ i fx ½B4 ðVD Þ B5 ðVD Þ s x
3
½B1 ðVD Þ B3 ðVD Þ sg 6¼ 0
X
3
ðiÞ ðiÞ ðiÞ
jPði x; sÞj2 ¼ x8 ; jQði x; sÞj2 ¼ fx2 ½ Bk ðVD Þ B4 ðVD Þ s þ B5 ðVD Þg2
k¼2
þ fx ½B4 ðVD Þ B5 ðVD Þ s x3 ½B1 ðVD Þ B3 ðVD Þ sg2
X3
ðiÞ ðiÞ ðiÞ
X3
ðiÞ ðiÞ
fx2 ½ Bk ðVD Þ B4 ðVD Þ s þ B5 ðVD Þg2 ¼ x4 ½ Bk ðVD Þ B4 ðVD Þ s2
k¼2 k¼2
ðiÞ
X3
ðiÞ ðiÞ ðiÞ
þ ½B5 ðVD Þ2 2 x2 ½ Bk ðVD Þ B4 ðVD Þ s B5 ðVD Þ
k¼2
ðiÞ ðiÞ ðiÞ ðiÞ ðiÞ

fx ½B4 ðVD Þ B5 ðVD Þ s x3 ½B1 ðVD Þ B3 ðVD Þ sg2 ¼ x2 ½B4 ðVD Þ
B5 ðVD Þ s2 þ x6 ½B1 ðVD Þ B3 ðVD Þ s2 2 x4 ½B4 ðVD Þ B5 ðVD Þ s
ðiÞ ðiÞ
½B1 ðVD Þ B3 ðVD Þ s
ðiÞ ðiÞ
X3
ðiÞ ðiÞ
jQði x; sÞj2 ¼ x6 ½B1 ðVD Þ B3 ðVD Þ s2 þ x4 f½ Bk ðVD Þ B4 ðVD Þ s2
k¼2
2 ½B4 ðVD Þ
B5 ðVD Þ
s ½B1 ðVD Þ B3 ðVD Þ sg
ðiÞ ðiÞ
X 3
ðiÞ ðiÞ
þ x2 f½B4 ðVD Þ B5 ðVD Þ s2 2 ½ Bk ðVD Þ B4 ðVD Þ s
k¼2
ðiÞ ðiÞ 2
B5 ðVD Þg þ ½B5 ðVD Þ
ðiÞ ðiÞ
Fðx; sÞ ¼ jPði x; sÞj2 jQði x; sÞj2 ¼ x8 x6 ½B1 ðVD Þ B3 ðVD Þ s2
X
3
x4 f½ Bk ðVD Þ B4 ðVD Þ s2 2 ½B4 ðVD Þ B5 ðVD Þ s
k¼2
½B1 ðVD Þ B3 ðVD Þ sg x2 f½B4 ðVD Þ B5 ðVD Þ s2
X
3
2½ Bk ðVD Þ B4 ðVD Þ s B5 ðVD Þg ½B5 ðVD Þ2
k¼2
ðiÞ
H0 ¼ ½B5 ðVD Þ2 ;
ðiÞ ðiÞ
X
3
ðiÞ ðiÞ ðiÞ
H2 ¼ f½B4 ðVD Þ B5 ðVD Þ s2 2 ½ Bk ðVD Þ B4 ðVD Þ s B5 ðVD Þg
k¼2
X
3
H4 ¼ f½ Bk ðVD Þ B4 ðVD Þ s2 2 ½B4 ðVD Þ B5 ðVD Þ s
k¼2
ðiÞ ðiÞ
½B1 ðVD Þ B3 ðVD Þ sg
ðiÞ ðiÞ P
4
H6 ¼ ½B1 ðVD Þ B3 ðVD Þ s2 ; H8 ¼ 1. Hence Fðx; sÞ ¼ 0 implies H2k
k¼0
x2k ¼ 0.
And its roots are given by solving the above polynomial. Furthermore
X
3
ðiÞ ðiÞ ðiÞ
QR ði x; sÞ ¼ x2 ½ Bk ðVD Þ B4 ðVD Þ s þ B5 ðVD Þ
k¼2
QI ði x; sÞ ¼ x ½B4 ðVD Þ B5 ðVD Þ s x3 ½B1 ðVD Þ B3 ðVD Þ s
Hence sin hðsÞ ¼ PR ðix;sÞQI ðix;sÞ þ PI ðix;sÞQR ðix;sÞ

jQðix;sÞj2
and cos hðsÞ ¼
PR ðix;sÞQR ðix;sÞ þ PI ðix;sÞQI ðix;sÞ
jQðix;sÞj2
. We already approve
ðiÞ ðiÞ
X3
ðiÞ
k¼2
ðiÞ ðiÞ ðiÞ
B4 ðVD Þ s þ k ½B4 ðVD Þ B5 ðVD Þ s
ðiÞ
þ B5 ðVD Þg eks ; n ¼ 4; m ¼ 3; n [ m

definition: k ! j ; pk ðsÞ ! aj ; qk ðsÞ ! cj ; n ¼ 4 ; m ¼ 3 ; n [ m.
P
Additionally Pn ðk; sÞ ! PðkÞ; Qm ðk; sÞ ! QðkÞ then PðkÞ ¼ 4j¼0 aj k j ;
P
QðkÞ ¼ 2j¼0 cj k j
ðiÞ ðiÞ
X
3
ðiÞ ðiÞ
Pk ¼ k4 ; Qk ¼ k3 ½B1 ðVD Þ B3 ðVD Þ s þ k2 ½ Bk ðVD Þ B4 ðVD Þ s
k¼2
ðiÞ ðiÞ ðiÞ
þk ½B4 ðVD Þ B5 ðVD Þ s þ B5 ðVD Þ

tion of s such that a0 þ c0 6¼ 0. In the following “−” denotes complex and conju-
gate. PðkÞ; QðkÞ are analytic functions in k and differentiable in s. The coefficients
faj ðL; C; Rb ; Lb ; RL ; VV ; Vex ; m; s; . . .Þ and cj ðL; C; Rb ; Lb ; RL ; VV ; Vex ; m; s; . . .Þg 2
R depend on TD system’s L; C; Rb ; Lb ; RL ; VV ; Vex ; m; s; . . . values. Unless strictly
necessary, the designation of the varied arguments ðL; C; Rb ; Lb ; RL ;
VV ; Vex ; m; s; . . .Þ will subsequently be omitted from P, Q, aj, cj. The coefficients aj,
cj are continuous, and differentiable functions of their arguments, and direct sub-
ðiÞ
stitution shows that a0 þ c0 6¼ 0; B5 ðVD Þ 6¼ 0. 8 L; C; Rb ; Lb ; RL ; VV ;
Vex ; m; s; . . . 2 R þ i.e. k ¼ 0 is not a root of the characteristic equation.
Furthermore PðkÞ; QðkÞ are analytic function of k for which the following
requirements of the analysis (see Kuang [5], Sect. 3.4) can also be verified in the
present case [4–6].
ðL; C; Rb ; Lb ; RL ; VV ; Vex ; m; s; . . .Þ domain of interest.
(b) jQ(kÞ=PðkÞj is bounded for jkj ! 1, Re k
ðiÞ ðiÞ P
3
ðiÞ ðiÞ
k3 ½B1 ðVD Þ B3 ðVD Þ s þ k2 ½ Bk ðVD Þ B4 ðVD Þ s
k¼2
ðiÞ ðiÞ ðiÞ
Q(kÞ þ k ½B4 ðVD Þ B5 ðVD Þ s þ B5 ðVD Þ
j j¼j j
PðkÞ k4

ðiÞ ðiÞ
Fðx; sÞ ¼ jPði x; sÞj2 jQði x; sÞj2 ¼ x8 þ x6 ½B1 ðVD Þ B3 ðVD Þ s2
X
3
þ x4 f½ Bk ðVD Þ B4 ðVD Þ s2 2 ½B4 ðVD Þ B5 ðVD Þ s
k¼2
½B1 ðVD Þ B3 ðVD Þ sg þ x2 f½B4 ðVD Þ B5 ðVD Þ s2
X3
2½ Bk ðVD Þ B4 ðVD Þ s B5 ðVD Þg þ ½B5 ðVD Þ2
k¼2
Has at most a finite number of zeros. Indeed, this is a polynomial in x .

(d) Each positive root xðL; C; Rb ; Lb ; RL ; VV ; Vex ; m; s; . . .Þ of F(xÞ¼ 0 is con-
tinuous and differentiable with respect to L; C; Rb ; Lb ; RL ; VV ; Vex ; m; s; . . ..

Pði xÞ and Qði xÞ ¼ Qði xÞ thus, k ¼ i x, x [ 0 maybe on eigenvalue of
characteristic equation. The analysis consists in identifying the roots of the char-
acteristic equation situated on the imaginary axis of the complex k—plane, whereby
increasing the parameters L; C; Rb ; Lb ; RL ; VV ; Vex ; m; s; . . ., Rek may, at the
crossing, change its sign from (−) to (+), i.e. from a stable focus
E ðX ; Y ; VD ; VA Þ ¼ ð0; 0; VD ; 0Þ to an unstable one, or vice versa. This feature
to L; C; Rb ; Lb ; RL ; VV ; Vex ; m; s; . . . and TD’s system parameters. x 2 R þ :
@Rek
^1 ðRf Þ ¼ ð Þ ; C; Rb ; Lb ; RL ; VV ; Vex ; m; s; . . . ¼ const
@L k¼ix
@Rek
^1 ðLf Þ ¼ ð Þ ; L; Rb ; Lb ; RL ; VV ; Vex ; m; s; . . . ¼ const
@C k¼ix
@Rek
^1 ðCf Þ ¼ ð Þ ; L; C; Lb ; RL ; VV ; Vex ; m; s; . . . ¼ const
@Rb k¼ix
@Rek
^1 ðCc Þ ¼ ð Þ ; L; C; Rb ; RL ; VV ; Vex ; m; s; . . . ¼ const
@Lb k¼ix
@Rek
^1 ðRsb Þ ¼ ð Þ ; L; C; Rb ; Lb ; VV ; Vex ; m; s; . . . ¼ const
@RL k¼ix
@Rek
^1 ðsÞ ¼ ð Þ ; L; C; Rb ; Lb ; RL ; VV ; Vex ; m; . . . ¼ const
@s k¼ix
inserting k ¼ i x.
Into TD system’s characteristic equation, x must satisfy the following:
sin x s ¼ gðxÞ ¼ ; cos x s ¼ hðxÞ
jQði xÞj2
¼
jQði xÞj2
where jQði xÞj2 6¼ 0 in view of requirement (a) above, and ðg; hÞ 2 R.

that I R þ 0 is the set where xðsÞ is a positive root of FðxÞ and for s 62 I , xðsÞ is
hðsÞ 2 ½0; 2 p as the solution of sin hðsÞ ¼ PR ðixÞQI ðixÞ þ PI ðixÞQR ðixÞ
jQðixÞj2
and
cos hðsÞ ¼ PR ðixÞQR ðixÞ þ PI ðixÞQI ðixÞ
jQðixÞj2
:
xðsÞ s ¼ hðsÞ þ n 2 p 8 n 2 N0 . Hence we can define the maps sn : I !
R þ 0 given by sn ðsÞ ¼ hðsÞxðsÞ þ n2p
; n 2 N0 ; s 2 I. Let us introduce the functions
I ! R ; Sn ðsÞ ¼ s sn ðsÞ; s 2 I; n 2 N0 that is a continuous and differentiable in s.
In the following, the subscripts k; x; L; C; Rb ; Lb ; RL ; VV ; Vex ; m; s; . . . indicate the
corresponding partial derivatives. Let us first concentrate on ^ðxÞ, remember in
kðL; C; Rb ; Lb ; RL ; VV ; Vex ; m; s; . . .Þ and xðL; C; Rb ; Lb ; RL ; VV ; Vex ; m; s; . . .Þ, and
keeping all parameters except one (x) and s. The derivation closely follows that in
reference [BK]. Differentiating TD’s system characteristic equation PðkÞ þ QðkÞ
eks ¼ 0 with respect to specific parameter (x), and inverting the derivative, for
convenience, one calculates:
Remark: x ¼ L; C; Rb ; Lb ; RL ; VV ; Vex ; m; . . .

ð Þ ¼
where Pk ¼ @P
@k ; . . . etc., Substituting k ¼ i x, and bearing i Pði xÞ ¼ Pði xÞ,
Qði xÞ ¼ Qði xÞ then i Pk ði xÞ ¼ Px ði xÞ; i Qk ði xÞ ¼ Qx ði xÞ and
that on the surface jPði xÞj2 ¼ jQði xÞj2 , one obtains:


Qx ¼ QRx þ i QIx ; P2 ¼ P2R þ P2I . When (x) can be any TD’s system parameters
L; C; Rb ; Lb ; RL ; VV ; Vex ; m; . . ., and time delay s etc. Where for convenience, we
have dropped the arguments ði x; xÞ, and where Fx ¼ 2 ½ðPRx PR þ PIx PI Þ
ðQRx QR þ QIx QI Þ; xx ¼ Fx =Fx
Fx ¼ 2 ½ðPRx PR þ PIx PI Þ ðQRx QR þ QIx QI Þ We define U and V:

We choose our specific parameter as time delay x = s. We already find
X3
ðiÞ ðiÞ ðiÞ
QR ði x; sÞ ¼ x2 ½ Bk ðVD Þ B4 ðVD Þ s þ B5 ðVD Þ
k¼2
QI ði x; sÞ ¼ x ½B4 ðVD Þ B5 ðVD Þ s x3 ½B1 ðVD Þ B3 ðVD Þ s
X
3
ðiÞ ðiÞ
PRx ¼ 4 x3 ; PIx ¼ 0; QRx ¼ 2 x ½ Bk ðVD Þ B4 ðVD Þ s
k¼2
QIx ¼ ½B4 ðVD Þ B5 ðVD Þ s 3 x ½B1 ðVD Þ B3 ðVD Þ s; PRs ¼ PIs ¼ 0
2
ðiÞ ðiÞ ðiÞ

QRs ¼ x2 B4 ðVD Þ; QIs ¼ x B5 ðVD Þ þ x3 B3 ðVD Þ
PRx PR ¼ 4 x7 ; PIx PI ¼ 0 ;
X
3
ðiÞ ðiÞ
QRx QR ¼ 2 x3 ½ Bk ðVD Þ B4 ðVD Þ s2
k¼2
X
3
ðiÞ ðiÞ ðiÞ
2x½ Bk ðVD ÞB4 ðVD Þ s B5 ðVD Þ ;
k¼2
QIx QI ¼ f½B4 ðVD Þ B5 ðVD Þ s 3 x2 ½B1 ðVD Þ B3 ðVD Þ sg
fx ½B4 ðVD Þ B5 ðVD Þ s x3 ½B1 ðVD Þ B3 ðVD Þ sg
¼ x ½B4 ðVD Þ B5 ðVD Þ s2 4 x3 ½B1 ðVD Þ B3 ðVD Þ s
½B4 ðVD Þ B5 ðVD Þ s þ 3 x5 ½B1 ðVD Þ B3 ðVD Þ s2
PRs PR ¼ 0; PIs PI ¼ 0;
ðiÞ
X
3
ðiÞ ðiÞ ðiÞ
QRs QR ¼ x2 B4 ðVD Þ fx2 ½ Bk ðVD Þ B4 ðVD Þ s þ B5 ðVD Þg
k¼2
QIs QI ¼ fx B5 ðVD Þ þ x B3 ðVD Þg
3
fx ½B4 ðVD Þ B5 ðVD Þ s
ðiÞ ðiÞ
x3 ½B1 ðVD Þ B3 ðVD Þ sg
X
3
ðiÞ ðiÞ ðiÞ
PR PIx ¼ 0; PI PRx ¼ 0; QR QIx ¼ fx2 ½ Bk ðVD Þ B4 ðVD Þ s þ B5 ðVD Þg
k¼2
f½B4 ðVD Þ B5 ðVD Þ s 3 x2 ½B1 ðVD Þ B3 ðVD Þ sg

QI QRx ¼ fx ½B4 ðVD Þ B5 ðVD Þ s x3 ½B1 ðVD Þ B3 ðVD Þ sg
X
3
ðiÞ ðiÞ
f2 x ½ Bk ðVD Þ B4 ðVD Þ sg
k¼2
PR PIs ¼ 0; PI PRs ¼ 0;
X3
ðiÞ ðiÞ ðiÞ
QR QIs ¼ fx2 ½ Bk ðVD Þ B4 ðVD Þ s þ B5 ðVD Þg
k¼2
ðiÞ ðiÞ
fx B5 ðVD Þ þ x3 B3 ðVD Þg;
QI QRs ¼ fx ½B4 ðVD Þ B5 ðVD Þ s x3 ½B1 ðVD Þ B3 ðVD Þ sg
ðiÞ
x2 B4 ðVD Þ
U ¼ ðPR PIx PI PRx Þ ðQR QIx QI QRx Þj PR PIx ¼0 ¼ QR QIx þ QI QRx
PI PRx ¼0
Vjx¼s ¼ ðPR PIs PI PRs Þ ðQR QIs QI QRs Þj PR PIs ¼0 ¼ QR QIs þ QI QRs
PI PRs ¼0
Fx¼s ¼ 2 ½ðPRs PR þ PIs PI Þ ðQRs QR þ QIs QI Þj PRs PR ¼0 ¼ 2 ðQRs QR þ QIs QI Þ
PIs P¼0
Fx ¼ 2 ½ðPRx PR þ PIx PI Þ ðQRx QR þ QIx QI Þj PRx PR ¼4x7 ¼ 2 ½4 x7 ðQRx QR þ QIx QI Þ
PIx PI ¼0
Fðx; sÞ ¼ 0: Differentiating with respect to s and we get Fx @x

@s þ Fs ¼ 0;
s 2 I ) @x
@s ¼ Fs
Fx
@Rek @x Fs 2 ½U þ s jPj2 þ i Fx
^1 ðsÞ ¼ ð Þk¼ix ; ¼ xs ¼ ; ^1 ðsÞ ¼ Ref g
@s @s Fx Fs þ i 2 ½V þ x jPj2
@Rek
@s k¼ix
@x U @x
@s þ V
@s jPj2

TD’s system, about the equilibrium point E ðX ; Y ; VD ; VA Þ ¼ ð0; 0; VD ; 0Þ as a
result of a variation of delay parameter s. The analysis consists in identifying the
roots of our system characteristic equation situated on the imaginary axis of the
complex k-plane. Where by increasing the delay parameter s, Re k may at the
crossing, changes its sign from − to +, i.e. from a stable focus E(*) to an unstable
the partial derivatives with respect to s, ^1 ðsÞ ¼ ð@ Re k 1
@s Þk¼ix ^ ðsÞ ¼
ð@ RekÞ
@s k¼ix; L; C; R ; L ; R ; V ; V ; m; . . . ¼ const ; x 2 R : [12].
b b L V ex þ
Exercises
1. A two-stage limiter circuit is shown in below figure. The limiter PIN diode at the
output (D2), commonly referred to as the “clean-up stage,” is the diode with
thinner I layer, selected so that the threshold level of the circuit is low enough to
protect the remainder of the receiver components. The limiter diode at the input
(D1), often called the “coarse limiter,” has a thicker I layer for several reasons.
The P layer diameter can be larger for a diode with a thicker I layer while
maintaining a capacitance value that produces low insertion loss under small
input signal conditions. The circuit components are connected by microstrip
segments. We consider for simplicity that the microstrip segments resistances
are neglected and either related voltages Vsk ! e, k = 1,…,5. Two limiter
diode’s equivalent circuit parameters are not the same. We consider coming
signal cause at t = 0 voltage V(t = 0).
Va ðtÞ ¼ Vain ðt s1 Þ;
X
2 X
3
Vb ðtÞ ¼ Va ðt s2 Þ ¼ Vain ðt si Þ; Vc ðtÞ ¼ Vb ðt s3 Þ ¼ Vain ðt si Þ
i¼1 i¼1
X
4
Vcout ðtÞ ¼ Vb ðt s4 Þ ¼ Vain ðt si Þ; IL ðtÞ ! IL ðt s5 Þ
i¼1
1:1 Write two stage limiter system differential equations.

1:2 Try to recognize our system differential equations as forced Van der Pol
equations. Write our two-stage limiter circuit as autonomous system.
1:3 Discuss system stability for the following cases:
P2 P
3 P4 pffiffiffi
(a) s1 ¼ s; i¼1 si ¼ s2 ; si ¼ s3 ; i¼1 si ¼ s3 þ s; s5 ¼ s
i¼1
pffiffiffi P2 pffiffiffi P3 pffiffiffi pffiffiffi P4
(b) s1 ¼ s; si ¼ s s; si ¼ s s þ 3 s; si ¼
i¼1 i¼1 i¼1
pffiffiffi pffiffiffi pffiffiffi
s s þ ð 3 sÞ 2 ; s5 ¼ s
How s value variations influence our system stability?
1:4 How our system dynamic and stability behavior changes when “Clean up”
limiter PIN diode is disconnected?
1:5 How our system dynamic and stability behavior changes when “Coarst”
limiter PIN diode is disconnected?
1:6 Microstrip segment No. 5 delay (s5 ) is dependent on other segment’s delay
P pffiffiffiffi
summation s5 ¼ f ð 4i¼1 si Þ; s1 ¼ s. si þ 1 ¼ si 8 i ¼ 1; ::; 3. Try to find f()
function which our system is stable for any value of s parameter. Which
possible f() functions there is a stability switching under variation of s
parameter. Investigate stability behavior.
2. We have limiter circuit system which characterize by two differential equations

(with delays s1 ; s2 ; s3 ).
pffiffiffiffiffi X X
1 dVðt s2 Þ d 2 Vðt s2 Þ 3 3
V0 ðtÞ ¼ ½ þ C1 Li þ Vðt sk Þ
R1 dt dt2 i¼0 k¼2
1 dV X
2
d V 2 X
3
Vi ðt s1 Þ ¼ ½ þ½ Ci 2 Li þ V
P
3 dt i¼1
dt i¼0
Ri
i¼1
2:1 Draw possible limiter circuits which can fulfil above system differential
equations. Vi—incoming RF signal voltage, Vo—out-going RF signal,
Exercises 269
V—circuit internal voltage variable. si ði ¼ 1; 2; 3Þ represent microstrip

segment delay parameters.
2:2 Find equivalent Van der Pol equation. Represent our system as an auton-
omous system.
2:3 Discuss stability and stability switching for different values of s1 ; s2 ; s3
parameters.
P3
2:4 i¼1 Ri ! 1, Draw our limiter circuit. Find equivalent Van der Pol
equation. Discuss stability and stability switching for different values of Ci,
i = 1,2.
2:5 Discuss system stability behavior for different values of R1 and L2.
3. We have active circulator of four ports, namely P1, P2, P3 and P4. Active cir-
culator is a four terminal device in which input from one port is transmitted to the
next port in rotation. The RF input signal is given at P1 of the circulator from the
left side. This signal from P1 is transmitted to P2. We can connect LC (L1, C1)
components in series to P2 port which results in phase shift and helps to reflect
the signal to P3 at the right. We can connect LC (L2, C2) components in series to
P3 port which results in phase shift and helps to reflect the signal to P4 at the
right. At P4 we get an output RF signal. Each active circulator terminal faces a
delay parasitic effect of signal transferring in time. Our circuit is a Reflection
Type Phase Shifter (RTPS), employing a circulator. The circuit configuration of
the active circulator used four MESFETs which are the GEC-Marconi standard
library cell F20-FET-4x75. A typical four ports decade bandwidth active cir-
culator has four MESFETs transistors interconnected with each other. RF, CF, LF,
CC, Rsb plays a major role in the working of the circuit. The four feedback
branches (RF, CF, LF) are used to link all the four transistors in an end to end
fashion. The source resistor (Rsb) is shared among all the three MESFETs
transistors and one transistor is source coupled with the other two transistors
using this source resistor.
3:1 Draw four ports active circulator configuration and equivalent circuits.
Write system differential equations.
3:2 Find system equilibrium points (fixed points), consider RFin signal
V1 ðtÞ ¼ A0 þ f ðtÞ gðtÞ ; jf ðtÞj ( 1; jgðtÞj ( 1 ; A0 jf ðtÞj; A0 jgðtÞj
Find small increment Jacobian and characteristic equations.

3:3 Discuss stability behavior and stability switching for different values of s
parameter for the following cases:
P P pffiffiffiffiffi P pffiffiffi
(a) s1 ¼ s; 2i¼1 si ¼ s2 ; 3i¼1 si ¼ s3 ; 4i¼1 si ¼ s3 þ s s.
pffiffiffi P2 P p ffiffiffiffi
ffi P pffiffiffi
(b) s1 ¼ s; i¼1 si ¼ s3 þ 1; 3i¼1 si ¼ s5 ; 4i¼1 si ¼ s2 þ s s.
3:4 How our system’s dynamical behavior and stability are influenced if port 3
is disconnected? Write system differential equations and discuss stability
switching for different values of s parameter (s1 ¼ s2 ¼ s3 ¼ s).
3:5 We move port 3 termination components (C2, L2) to port 4 and take our RF
out signal from port 3. How our system’s behavior changes? Write system
differential equations and discuss stability behavior.
4. Consider RF system which characterize by Van der Pol equation as autonomous

system, when ðV; W; hÞ 2 R2 x S2 . We suppose a; b 1, since we are interested
in the periodic coming signal response we use 2p
x periodic transformation. The
following equations describe our system:
du1 1 x2 1
¼ 3 /ðVÞ cos x t ð Þ V sin x t
dt P x
½ Rk C1
k¼1
1 X
2
sin x t Vi ðt sk Þ
P
3 P
2
x½ Ck ½ Li k¼0
k¼1 i¼0
du2 1 x2 1
¼ 3 /ðVÞ sin x t ð Þ V cos x t
dt P x
½ Rk C1
k¼1
1 X
2
cos x t Vi ðt sk Þ
P
3 P
2
x½ Ck Li k¼0
k¼1 i¼0
Exercises 271
4:1 Find the analog basic Van der Pol equation and implement the system by
power limiter diodes and discrete components.
4:2 Discuss stability behavior and stability switching for variation of sk delay
parameter values (k = 0,1,2).
P
4:3 How the system dynamic and stability are changed for ½ 3k¼1 Rk ! 1?
P2
4:4 How the system dynamic and stability are changed for i¼0 Li ! e ?
P2
k¼0 sk ¼ s . How N parameter
N
4:5 Discuss system stability switching for
values influence our system stability switching.
5. We have cylindrical (closed) RF ladder network structure antennas for coupled

plasma sources copper legs which delayed in time by parasitic effects. The
antenna is mounted outside a glass tube. The RF ladder network antenna con-
sists of 16 copper legs (inductors) equally spaced interconnected with capaci-
tors, each copper leg has parasitic time delay (si þ 1 ¼ si þ si1 ; s1 = s2 =
s 8 i ¼ 2; ::; 15). We consider for simplicity that all copper legs voltages on
delay units (Ve) are neglected Ve ! e. There is a delay in each copper leg
current I1 ðt s1 Þ; . . .; I16 ðt s16 Þ. We consider all ladder capacitors are the
same (C) and all ending capacitors are the same (Cend). We consider all antenna
elements inductance values are the same (L). C1 ¼ C2 ¼ ¼ C15 ¼ C16 ¼ C
and CA1 ¼ CA2 ¼ ¼ CA16 ¼ Cend ; L1 ¼ L2 ¼ ¼ L16 ¼ L ; IL1 ¼
I1 ; ; IL16 ¼ I16 : The antenna ladder network structure is fed by the transmitter
unit (S1 = OFF no direct RF feeding).
RF ladder network
Structure antenna
RF Feeding
Rp
(no direct)
Cylindrical RF ladder network structure can represent as round ladder strip of

capacitors and inductors. The schematic contains RF feeding signal, S1 switch
(S1 = ON for RF signal feeding, S1 = OFF for RF signal transmitter feeding).
Rp is the parasitic resistance of RF feeding direct source.
Remark: Only one ground point exists in our Cylindrical RF ladder network
structure, it is connected to point B8.
5:1 Write system differential equations and find fixed points.

5:2 Discuss system stability; find Jacobian of our cylindrical RF ladder network
structure antenna system. Find system general characteristic equations.
5:3 Find K−1(C), K−1(Cend), K−1(s) and K−1(L) functions and discuss stability
switching.
5:4 We disconnect odd copper legs (L2∙k+1; k = 0, 1, 2…6, 7) in our cylindrical
RF ladder network structure antenna. How our system stability switching is
effected? Find system differential equations and fixed points, find Jacobian
and discuss stability switching. Find K−1(s) and K−1(L) functions.
5:5 We disconnect even CA capacitors, disconnect CA2∙k; k = 1, 2…7, 8) in our
cylindrical RF ladder network structure antenna. How our system stability
switching is effected? Find system differential equations and fixed points, find
Jacobian and discuss stability switching. Find K−1() and K−1(L) functions.
5:6 We move our system ground to point B12, How our system behavior
changes? Discuss stability switching, K−1() and K−1(L) functions.
6. We have the following TD’s microwave oscillator which two Tunnel Diodes
(TDs) are connected in series. TD’s biasing circuit is constructed from DC
voltage Vb, inductor Lb which block oscillation from Vb source and capacitor Cb
Exercises 273
which shorts to ground Vb bias voltage’s oscillations. The DC voltage Vb biases

the tunnel diode TD into its negative resistance region and also supplies the
power used in amplifying the input signal Vi. Switch S1 is ON and S2 in ON
(bypass our signal source Vi) and our circuit functions as a oscillator. S3 can be
in OFF state or ON state. Microwave resonant cavities is represented and
thought of as simple LC circuits. We represent our microwave cavity as resonant
LC circuit. Due to cavity parasitic effects there are delays in the currents which
flow through equivalent resonant L and C elements and TD’s voltage derivative
in time. s1 ; s2 are the time delays for the TD’s voltage derivative in time
respectively (D1 & D2). s3 is the time delay for the current flows through C
element. RL is the load resistance. Load resistance (RL) is chosen so that RL < |
min (Rd1, Rd2)| in the TD’s characteristic NDR region. At the Tunnel diodes
operating points, the total circuit differential resistance is negative. We consider
for simplicity, two Tunnel diodes parameters are the same.
6:1 Find tunnel diodes system fixed points for the cases S3 = OFF/ON.
6:2 Find Tunnel diodes system differential equations for the cases
S3 = OFF/ON. How the system dynamic changes for two positions of
switch S3.
6:3 Find tunnel diodes system characteristic equations for two cases of S3
switch.
6:4 Discuss stability switching for variation of si ði ¼ 1; 2; 3Þ delay parameter
values.
6:5 How the system dynamical behavior changes if resistor RL is disconnected?
6:6 What happened if we bypass D2 tunnel diode (short) and S3 = ON con-
stantly? Discuss stability and stability switching for different values of s3
delay parameter.
7. We have planar RTD (Resonant Tunneling Diode) oscillator which eliminate

parasitic bias oscillations in an oscillator circuitry by employing a shunt resistor
to the NDR device. A non-linear (diode) resistor, Schottky diode. Sd and Re are
a Schottky diode and a resistance, respectively, and form the stabilizing
resistance. Ce is a decoupling capacitor (is an RF short circuit), while “TML” is
said to be quarter wave transmission line with signal time delay s. At millimeter
wave and low terahertz frequencies (<1 THz), planar RTD oscillator have been
integrated at the center of the slot antenna, a location at which the antenna input
impedance is infinity.
The above schematic is Reddy’s bias stabilization scheme for sum mm- wave
RTD oscillators. Rb is the resistance of the bias line. Sd, Re and Ce are the
Schottky diode, external resistor and external capacitor, respectively, which
form stabilizing circuit. TML is the quarter wave length transmission line at the
oscillation frequency. RF load is pure resistive Rload and resonant circuit is
parallel LC circuit. We consider that Schottky diode current function is
qðVID RS Þ
ID ¼ IS ½e nKT 1 þ Gx V. n—Ideality factor, IS—Reverse saturation
current, k—Boltzman constant, T—Absolute temperature, V—Bias voltage,
GX ∙ V—Edge leakage current, ID RS —Voltage drop due to ohmic contact
and bulk resistance of semiconductor channel (Ohmic contact is a metal
semiconductor contact that has a linear I-V and non-rectifying characteristics).
Remark: You should consider RTD equation as in subchapter TD’s equation.
The input impedance of a transmission line of length L with characteristic
impedance Z0 and connected to load with impedance ZA (in our case ZA is RTD
element in parallel to RF load and resonant circuit).
Zin ðLÞ ¼ Z0 ½ZZA0 þþ jZ
jZ0 tgðbLÞ
A tgðbLÞ
; Z A ¼ ZAR þ j ZAI
Exercises 275
k
k ZA þ j Z0 tgð2p
k 4Þ Z02 k
Zin ðL ¼ Þ ¼ Z0 ½ k
¼ : Zin ðL ¼ Þ ¼ 50 Ohm
4 Z0 þ j ZA tgð k 4Þ
2p ZA 4
We know the values for ZAR and ZAI . Need to find Z0 expression.
7:1 Find Reddy’s bias stabilization Scheme fixed point and differential
equations.
7:2 Discuss stability switching for different values of s parameter (quarter wave
transmission line delay time).
7:3 What happened if Ce is disconnected? How the system dynamic change?
7:4 Find the following functions: K(Z0), K(ZA), K(Gx), K(VV), K(m).
Remark: VV and m parameters are RTD’s Internal parameters.
7:5 Discuss stability and dynamical behavior if Schottky diode Sd is shorted.
8. We have tunnel diodes system as a microwave oscillator. D1 and D2 are tunnel
diodes which are connected back to back configuration. There are two biasing
circuits (Vb1, Rb1 and Vb2, Rb2). Additionally there is a smart function element
which his voltage (VA − VB) is a complicated function of tunnel diodes cur-
rents respectively. TD’s parasitic effects cause to current time delays s1 and s2
pffiffiffi
respectively ID1 ðtÞ ! ID1 ðt s1 Þ ID2 ðtÞ ! ID2 ðt s2 Þ; s1 ¼ s; s2 ¼ s þ s.
At time t = 0 switch S1, S2 and S3 move to ON state. We consider for sim-
plicity, two tunnel diodes are identical (internal parameters are the same) and
TD’s current function is as discuss in the chapter.
VAB ¼ gðID1 ; ID2 Þ ¼ fmaxðID1 ; ID2 Þ þ ID1 ID2 g C; maxðID1 ; ID2 Þ
1 1
¼ ðID1 þ ID2 Þ þ jID1 ID2 j
2 2
C—Resistive parameter. L, C—Parallel resonant circuit.

8:1 Find system fixed points and differential equations.

8:2 Discuss system stability switching for different values of s time delay
parameter.
8:3 How the system dynamic changed if at time t = 0, switch S1 and S2 move
to ON state and S3 is constantly in OFF state? Find fixed points and discuss
stability switching for different values of s parameter.
8:4 How the system dynamic changed if at time t = 0, switch S1 and S3 move
to ON state and S2 is constantly in OFF state? Find fixed points and discuss
8:5 Find the following functions: K(C), C(R), K(L), and K(C).
8:6 How the system dynamic changed if at time t = 0, switch S1 move to ON
state and S2, S3 is constantly in OFF state? Find fixed points and discuss
9. We have tunnel diode’s bridge structure system as a microwave oscillator. D1,
D2, D3 and D4 are tunnel diodes in our bridge structure. For simplicity all tunnel
diodes are identical (they have the same internal parameters values). There are
two biasing circuits (Vb1, Rb1 and Vb2, Rb2). Switches S1 and S2 connect biasing
circuits to tunnel diode bridge structure. TD’s bridge structure contacts parasitic
effects cause to current time delays s1 and s2 respectively. We have three pos-
sible cases: (1) S1 = ON, S2 = OFF, (2) S1 = OFF, S2 = ON, (3) S1 = ON,
S2 = ON. TD’s current function is as discuss in the chapter. Resonant circuit
No. 1 includes Inductor and capacitor in parallel (L1, C1). Resonant circuit
No. 2 includes inductor and capacitor in parallel (L2, C2). Initially all switched
are in OFF state, at time t = 0 we move to one of the above cases.
Exercises 277
9:1 Find system fixed points and differential equations for the three cases. Find
the related characteristic equations.
9:2 Discuss stability switching for different values of s1 and s2 time delay
pffiffiffiffiffiffiffiffiffiffiffiffi
parameters. How system dynamic changes for s1 ¼ s2 þ 1?
9:3 Find the following functions: K(s1 ), C(s2 ), K(L1), and K(C2).
9:4 D2 tunnel diode is disconnected, how system dynamical behavior changes?
Discuss stability and stability switching.
9:5 D1 tunnel diode is disconnected, how system dynamical behavior changes?
10. We have the following custom TD’s current expression as a function of volt-
ages VDi. VDi represents specific TD’s voltage, the current which flow through
all TDs is the same (TDs are in series).
X
3
ID ¼ nðVD Þ ¼ gk ðVDi Þ
k¼1
P
2
Y
3 V
VDi Y
2 VDi þ 1 m
½ðgVDi Þ1 i¼1 ½ð V0 Þ
¼ Is f e th gþ f e g
i¼1
R0 i¼1
P
2
VDi Y
4 VDi VV
þ i¼1
f e½ Vex g
RV i¼2
Y
3 V
½ðgVDi Þ1
g1 ðVDi Þ ¼ Is f e th g;
i¼1
P
2 P
2
VDi Y
2 V þ1
VDi Y 4 VDi VV
½ð DiV Þm
g2 ðVDi Þ ¼ i¼1 f e 0 g; g3 ðVDi Þ ¼ i¼1 f e½ Vex g
R0 i¼1
RV i¼2
We connect our TDs in series to parallel LC resonant circuit and load Rload.
Additionally, we add biasing circuit (Vb, Rb).
10:1 Find system fixed points and differential equations.
10:2 Discuss system stability switching for different values of L, C resonant
circuit values.
10:3 One of the TD in series is shorted, How it influence our system dynamic?
Discuss fixed points and stability switching for different values of m
parameter.
(Hint: Discuss all possible cases for short TDs).
10:4 We define TD’s contact delays as sl ðl ¼ 1; 2; 3Þ respectively, Find K(sl ),
K(VV), K(Vex) functions.
Chapter 3
Microwave Semiconductor Amplifiers
Analysis
Microwave semiconductor amplifiers are widely used in many engineering appli-

cations. Among microwave semiconductors there are bipolar transistors which
operate at microwave frequencies, microwave Field Effect Transistors (FET)
minimize the adverse effects of transit time and internal capacitance and resistance,
IMPATT (IMPact ionization Avalanche Transit-Time) amplifiers which widely use
at the high end of the microwave band because microwave transistor do not work
well above 30 GHz due to transit time limitations. Stability of these microwave
amplifiers is affected by internal parameters variation and circuit microstrip parasitic
effects. In any microwave system there are parasitic elements with delays in time
which affect the performance and stability of the system. There are also IMPATT
diodes which are a form of high power diode used in high frequencies electronic
and microwave devices. The main advantage of IMPATT diode is the high power
capability and frequency band of 3 to 100 GHz and more. FET combined biasing
and matching circuit has many stability issues which must be taken for every RF
design [14, 15].
3.1 Bipolar Transistor at Microwave Frequencies

Description and Stability Analysis
Bipolar transistors are widely in use for many engineering applications. The bipolar
transistor consists of an N-doped semiconductors the emitter, a P-doped semicon-
ductor called the base, and an N – doped semiconductor called the collector. The
transistor consists of two PN junctions, one between the emitter and the base, the
other between the base and the collector. The base collector junction is reverse
biased by a rather large voltage. The emitter base junction is forward biased. When
the junction is reversed biased no current flows. Current begins to flow when the
junction is forward biased about specific voltage (*0.6 V). No current flows across

DOI 10.1007/978-3-319-45427-6_3
280 3 Microwave Semiconductor Amplifiers Analysis
the reverse biased base collector junction. The amplification of AC signal by a

bipolar transistor is widely used in every circuit. Initial state is when the emitter
base junction is DC biased and there is no AC signal applied. When the input AC
signal to be amplified is added to the DC bias, the voltage between the emitter and
the base increase (AC signal is positive). The current flows from the emitter into the
base. When the input AC signal is negative, it subtracts from the DC bias. There is
also an AC current gain. There is a difference in the operation between a microwave
bipolar transistor and a low frequency bipolar transistor. The operation is the same
except for the degrading effects of transit time, internal capacitance and resistance,
and external lead inductance that occur at microwave frequencies. Because the
electrons take a finite time to travel across the emitter base junction, through the
base, and across the base collector junction, the AC current is reduced at microwave
frequencies, compared with low frequencies. The velocity of microwaves, which is
the same as the velocity of light 3 108 sec m
. Electrons in semiconductor travel
approximately 1/3000 slower, or 10^8 mm per second. The distance that a
microwave or an electron travels in one microwave cycle is the velocity divided by
the frequency (Lcycle ¼ vf). Next is the bipolar transistor at microwave frequencies.
As the microwave signal enters the transistor, it must flow through the capacitance
of the forward biased emitter base junction (Ci). This capacitance exists in low
frequency transistor, but its reactance at low frequencies is negligible. Feedback
capacitance between the collector and base is important at microwave frequencies,
namely the capacitance of the reverse biased base collector junction (Cf). An output
Fig. 3.1 Bipolar transistor at microwave frequencies

3.1 Bipolar Transistor at Microwave Frequencies … 281
collector capacitance between the collector and emitter exists (Co). In addition, the
bonding wire connecting the transistor chip to the input and output microwave
transmission lines have inductance (Lin, Lo). The lower microwave frequency, the less
severe are the transit time effects, because the microwave cycle is longer and the
electrons have time to move through the transistor. At low microwave frequencies, the
base collector junction thickness can be made large and a higher voltage can be applied
and more power can be obtained at low microwave frequencies than at high micro-
wave frequencies. By using Ebers Moll transistor model (see S.M. Sze, Physics of
Semiconductor Devices), we get the following bipolar transistor at microwave fre-
quency equivalent circuit. X(t) is the microwave source input signal [24, 25, 33, 34].
IC ¼ Ic ; IE ¼ Ie .
IRC ¼ ICf þ IC þ IC0 þ ILout ; ILout ¼ IRL ; ILin þ ICf þ IRb ¼ Ib ; Alfa f ¼ af
Alfa R ¼ ar ; IC þ IDC af IDE ¼ 0; Ib ¼ IDC þ IDE ar IDC af IDE
Ib ¼ IDC ð1 ar Þ þ IDE ð1 af Þ; IDE ar IDC ¼ IE ; IC ¼ IDC þ af IDE
VCC VA1
ILin ¼ ICin ; VA2 ¼ VBE ; VA2 VA1 ¼ VBC ; ¼ IRC ;
RC
dIL
VA1 ¼ VLout þ VRL ¼ Lout out þ ILout RL
dt
dVA1 dVCf Vbb VA2
I C0 ¼ C 0 ; VA4 ¼ Vbb ; VA1 VA2 ¼ VCf ; ICf ¼ Cf ; ¼ IR b
dt dt Rb
dILin dVCin
XðtÞ VA2 ¼ VLin þ VCin ; VLin ¼ Lin ; ICin ¼ Cin ; XðtÞ VA3 ¼ VLin
dt dt
Vbe Vbc
VA2 VA3 ¼ VCin ; VA4 ¼ Vbb ; IDE ¼ ISE ½eð VT Þ 1; IDC ¼ ISC ½eð VT Þ 1
Fig. 3.2 Bipolar transistor at microwave frequencies equivalent circuit

af ISE ¼ ar ISC ¼ IS (See S.M. Sze, Physics of Semiconductor Devices).

af ar
bf ¼ ;b ¼ ; IE ¼ IB þ IC ) IB ¼ I E IC ;
1 af r 1 ar
1 1 af 1 1 1 ar 1
1¼ ¼ ; 1¼ ¼
af af bf ar ar br
Vbe Vbc
IE ¼ IDE ar IDC ¼ ISE ½eð VT Þ 1 ar ISC ½eð VT Þ 1
IS Vbe Vbc
¼ ½eð VT Þ 1 IS ½eð VT Þ 1
af
Vbc Vbe
IC ¼ IDC þ af IDE ¼ ISC ½eð VT Þ 1 þ af ISE ½eð VT Þ 1
IS Vbc Vbe
¼ ½eð VT Þ 1 þ IS ½eð VT Þ 1
ar
IS Vbe Vbc IS Vbc Vbe
IB ¼ IE IC ¼ ½eð VT Þ 1 IS ½eð VT Þ 1 f ½eð VT Þ 1 þ IS ½eð VT Þ 1g
af ar
Vbe 1 Vbc 1 IS Vbe IS Vbc
IB ¼ IS ½eð VT Þ 1 ð 1Þ þ IS ½eð VT Þ 1 ð 1Þ ¼ ½eð VT 1 þ ½eð VT Þ 1
af ar bf br
We can define the following terminology variables and parameters:

IC —Transistor collector current, IB —Transistor base current, IE —Transistor
emitter current. bf —Transistor forward common emitter current gain (20 → 500),
br - Transistor reverse common emitter current gain (0 → 20), af —Transistor
common base forward short circuit current gain (0.98 → 0.998). VT—Transistor
thermal voltage VT = k T/q (approximately 26 mV at 300 kelvin ≈ room
temperature).
IC af IE IE þ ar IC Vbe IE þ ar IC

IDC ¼ ; IDE ¼ ) ISE ½eð VT Þ 1 ¼
af ar 1 af ar 1 af ar 1
V
ð Vbe Þ IE þ ar IC V
ð Vbe Þ IE þ ar IC
ISE ½e T 1 ¼ )e T ¼f þ 1g
af ar 1 ISE ðaf ar 1Þ
IE þ ar IC
) Vbe ¼ VT lnf þ 1g
ISE ðaf ar 1Þ
IC af IE Vbc IC a f IE
IDC ¼ ) ISC ½eð VT Þ 1 ¼
af ar 1 af ar 1
IC af IE
) Vbc ¼ VT lnf þ 1g
ISC ðaf ar 1Þ
Vce ¼ Vcb þ Vbe ; Vcb ¼ Vbc ) Vce ¼ Vbc þ Vbe
IE þ ar IC IC a f IE
Vce ¼ VT lnf þ 1g VT lnf þ 1g
ISE ðaf ar 1Þ ISC ðaf ar 1Þ
VCC VA1 dVCf
IRC ¼ ICf þ IC þ IC0 þ ILout ) ¼ Cf þ IC
RC dt
dVA1
þ C0 þ ILout
dt
dILout
VCC ½Lout þ ILout RL dVCf
dt
¼ Cf þ IC
RC dt
d dIL
þ C0 ½Lout out þ ILout RL þ ILout
dt dt
dILout
VCC Lout dt ILout RL dVCf
¼ Cf þ IC
RC dt
d 2 ILout dILout
þ C0 ½Lout þ RL þ ILout
dt2 dt
ILin þ ICf þ IRb ¼ Ib ; ILin ¼ ICin ) ICin þ ICf þ IRb ¼ Ib
dVCin dVCf Vbb VA2
) Cin þ Cf þ ¼ Ib
dt dt Rb
dIL
VA1 VA2 ¼ VCf ) VA2 ¼ VA1 VCf ¼ Lout out þ ILout RL VCf ;
dt
then we get
dI
dVCin dVCf Vbb fLout dtLout þ ILout RL VCf g
ðÞ Cin þ Cf þ ¼ Ib
dt dt Rb
XðtÞ VA2 ¼ VLin þ VCin ) VCin ¼ XðtÞ VA2 VLin
dIL dILin
¼ XðtÞ fLout out þ ILout RL VCf g Lin
dt dt
dILout dILin
VCin ¼ XðtÞ Lout ILout RL þ VCf Lin
dt dt
dVCin dXðtÞ d 2 ILout dILout dVCf d 2 ILin
ðÞ ¼ Lout 2
RL þ Lin
dt dt dt dt dt dt2
d 2 ILout dILout dVCf 2 dVCf
Cin fdXðtÞ
dt Lout dt2 dt RL þ dt Lin d dtILin
2 g þ Cf dt
ðÞ ! ðÞ dIL
Vbb fLout outþ ILout RL VCf g
þ dt
Rb ¼ Ib
We have two global system differential equations:

dILout
VCC Lout ILout RL dVCf d 2 ILout dILout
dt
¼ Cf þ IC þ C0 ½Lout þ RL þ ILout
RC dt dt2 dt
dXðtÞ d 2 ILout dILout dVCf d 2 ILin dVCf
Cin Cin Lout 2
Cin RL þ Cin Cin Lin þ Cf
dt dt dt dt dt2 dt
dI
Vbb fLout dtLout þ ILout RL VCf g
þ ¼ Ib
Rb
dILout
Vbe ¼ VA2 ¼ Lout þ ILout RL VCf ; Vbc ¼ VA2 VA1 ¼ VCf
dt
IS V
ð bc Þ
V
ð be Þ
IC ¼ IDC þ af IDE ¼ ½e VT 1 þ IS ½e VT 1
ar
dILout
VC Lout þ ILout RL VC
IS f dt f
¼ ½eð VT Þ 1 þ IS ½eð VT Þ
1
ar
dILout
Lout þ ILout RL VC VC
IS Vbe IS Vbc IS dt f IS f
IB ¼ ½eð VT Þ 1 þ ½eð VT Þ 1 ¼ ½eð VT Þ
1 þ ½eð VT Þ 1
bf br bf br
dILout dVCf
We define new variables: Y1 ¼ dt ; Y2 ¼ dt
VCC Lout Y1 ILout RL dY1

¼ Cf Y2 þ IC þ C0 ½Lout þ Y1 RL þ ILout
RC dt
dY1 VCC Lout Y1 ILout RL
C0 Lout ¼ Cf Y2 Y1 C0 RL ILout IC
dt RC
dY1 1 1 1 Cf
¼ ½VCC Y1 RL
dt RC C0 Lout C0 Lout
RL 1 1
Y2 Y1 ILout IC
Lout C0 Lout C0 Lout
IC function: IC ¼ n1 ¼ n1 ðVCf ; Y1 ; ILout ; IS ; ar ; . . .Þ
dY1 1 1 1 Cf RL 1
¼ ½VCC Y1 R L Y2 Y1 IL
dt RC C0 Lout C0 Lout Lout C0 Lout out
1
n ðVCf ; Y1 ; ILout ; IS ; ar ; . . .Þ
C0 Lout 1
dXðtÞ d 2 ILout dILout dVCf d 2 ILin

Cin Cin Lout Cin RL þ Cin Cin L in
dt dt2 dt dt dt2
dILout
dVCf Vbb fLout dt þ ILout RL VCf g
þ Cf þ ¼ Ib
dt Rb
Ib function: Ib ¼ n2 ¼ n2 ðVCf ; Y1 ; ILout ; IS ; bf ; . . .Þ
dXðtÞ d 2 ILout d 2 ILin dIL dVCf

Cin Cin ½Lout 2
þ L in 2
out Cin RL þ ½Cin þ Cf
dt dt dt dt dt
dI
Vbb Lout dtLout ILout RL þ VCf
þ ¼ Ib
Rb
We define new variable:
dY3 d 2 ILout d 2 ILin dIL dILin

¼ Lout 2
þ Lin ; Y3 ¼ Lout out þ Lin
dt dt dt2 dt dt
dILout dILin dILin dILin 1 Lout
Y3 ¼ Lout þ Lin ) Y3 ¼ Lout Y1 þ Lin ) ¼ Y3 Y1
dt dt dt dt Lin Lin
dY3 dXðtÞ
Cin ¼ Cin Y1 Cin RL þ ½Cin þ Cf
dt dt
Vbb Lout Y1 ILout RL þ VCf
Y2 þ Ib
Rb
dY3 dXðtÞ Cf Vbb Lout Y1 ILout RL þ VCf 1
¼ Y1 RL þ ½1 þ Y2 þ Ib
dt dt Cin Rb Cin Cin
dY3 Cf Vbb Lout Y1 ILout RL þ VCf
¼ Y1 RL þ ½1 þ Y2 þ
dt Cin Rb Cin
1 dXðtÞ
n ðVCf ; Y1 ; ILout ; IS ; bf ; . . .Þ þ
Cin 2 dt
dY3 Lout Cf RL 1
¼ ½RL þ Y1 þ ½1 þ Y2 ILout þ VCf
dt Rb Cin Cin Rb Cin Rb Cin
1 Vbb dXðtÞ
n ðVCf ; Y1 ; ILout ; IS ; bf ; . . .Þ þ þ
Cin 2 Rb Cin dt
We can summery our circuit differential equations:
dY1 1 1 1 Cf RL 1
¼ ½VCC Y1 RL Y2 Y1 IL
dt RC C0 Lout C0 Lout Lout C0 Lout out
1
n ðVCf ; Y1 ; ILout ; IS ; ar ; . . .Þ
C0 Lout 1
dY3 Lout Cf RL 1
¼ ½RL þ Y1 þ ½1 þ Y2 ILout þ VCf
dt Rb Cin Cin Rb Cin Rb Cin
1 Vbb dXðtÞ
n ðVCf ; Y1 ; ILout ; IS ; bf ; . . .Þ þ þ
Cin 2 Rb Cin dt
dILin 1 Lout dIL dVCf
¼ Y3 Y1 ; out ¼Y1 ; ¼ Y2
dt Lin Lin dt dt
We find our system equilibrium points by setting dY1

dt ¼ 0; dYdt3 ¼ 0; dIdtLin ¼ 0
dILout dVCf dIL dVCf dILin

¼0; ¼ 0: out ¼ 0 ) Y1 ¼ 0; ¼ 0 ) Y2 ¼ 0; ¼ 0 ) Y3 ¼ 0
dt dt dt dt dt
dY1 1 1 1 1 1
¼0) ½VCC RL I
dt RC C0 Lout C0 Lout Lout C0 Lout
n1 ðVCf ; Y1 ¼ 0; ILout ; IS ; ar ; . . .Þ ¼ 0
dY3 RL 1 1
¼ 0 ) ILout þ VCf
dt Rb Cin Rb Cin Cin
Vbb dXðtÞ
n2 ðVCf ; Y1 ¼ 0; ILout ; IS ; bf ; . . .Þ þ þ ¼0
Rb Cin dt
IC function: IC ¼ n1 ¼ n1 ðVCf ; Y1 ; ILout ; IS ; ar ; . . .Þ and Ib function:

Ib ¼ n2 ¼ n2 ðVCf ; Y1 ; ILout ; IS ; bf ; . . .Þ
V I RL V
IS Cf Lout Cf
n1 ðVCf ; Y1 ¼ 0; ILout ; IS ; ar ; . . .Þ ¼ ½eð VT Þ
1 þ IS ½e ð VT Þ
1
ar
I R V V
Lout L
IS Cf
IS Cf
n2 ðVCf ; Y1 ¼ 0; ILout ; IS ; bf ; . . .Þ ¼ ½eð VT Þ
1 þ ½eð VT Þ
1
bf br
We consider Microwave RFin signal X(t) = A0 + fX(t); |fX(t)| 1 and A0 |

dfX ðtÞ
fX(t)| then XðtÞ ¼ A0 þ fX ðtÞ A0 ) dXðtÞ
dt ¼ dt ! e . We get the following two
equations for our ILout ; VCf fixed points: dXðtÞ
dt !e
1 1 1 1 1
½VCC RL ILout
RC C0 Lout C0 Lout C0 Lout
V I RL V
IS Cf Lout Cf
f ½eð VT Þ
1 þ IS ½eð VT Þ
1g ¼ 0
ar
I R V
Lout L
RL 1 1 IS ð
Cf
Þ
ILout þ VCf f ½e VT
1
Rb Cin Rb Cin Cin bf
V
IS Cf
Vbb
þ ½eð VT Þ
1g þ ¼0
br Rb Cin
Our system fixed points are E ðÞ ðY1 ; Y2 ; Y3 ; ILout VCf Þ ¼ ð0; 0; 0; ILout ; VCf Þ . We
V
Cf
need to find the expressions for ILout ; VCf . First exponent is eð VT Þ
. We define new
VC
exponent argument VT
f
¼ C1 ) eC1 1 C1 by using Taylor series approxima-
tion. We define our Taylor series approximation in the interval 0\C1 \1 )
V
VC ð
Cf
Þ VC
0\ VT
f
\1 ) 0\VCf \VT )e VT
1 VT
f
. Second exponent is
I RL V
ð
Lout Cf
Þ ILout RL VC
e VT
. We define new exponent argument VT
f
¼ C2 and there are two
ILout RL VC ILout RL VC ILout RL VC
cases: VT
f
[ 0 and VT
f
\0. First case VT
f
[ 0 ) VT ¼
VC
26 mV ) ILout RL VCf [ 0 ) RL [ f
ILout and by using Taylor series approxi-
I RL V
ð
Lout Cf
Þ ILout RL VC ILout RL VC
mation: eC2 ¼ e VT
1þ VT
f
. Second case VT
f
\0 ) VT ¼
VC
26 mV ) ILout RL VCf \0 ) 0\RL \ f
ILout and by using Taylor series
approximation:
I RL V
Lout Cf
C2 jC2 j ð Þ
e j C \0
2
¼e ¼1 jC2 j ¼e VT
I RL V
ð
Lout Cf
Þ
ILout RL VCf
eC2 jC2 \0 ¼ ejC2 j ¼ 1 jC2 j ¼ e VT
1j jC2 \0
VT
ILout RL VCf
¼ 1þ
VT
We implement our Taylor series approximation expressions into below

equations:

1 1 1 1 1 IS V C
ð# Þ ½VCC RL ILout f f þ IS
RC C0 Lout C0 Lout C0 Lout ar VT
ILout RL VCf
½ g
VT
¼0

RL 1 1 IS IL RL VCf IS V C Vbb
ð##Þ ILout þ VCf f ½ out fgþ ¼0
Rb Cin Rb Cin Cin bf VT br VT Rb Cin
1 1 IS IL IS RL IS VCf
ð#Þ ! Lout ½VCC RL ILout f VCf þ out g¼0
RC Lout ar V T VT VT
1 VC IS IS RL
½VCC Lout RL þ f ½IS ILout ½ þ 1 ¼ 0 ) ILout
RC VT ar VT
VC
1
RC ½VCC Lout RL þ VT
f
½IS aISr
¼
½ISVRT L þ 1

IS ILout RL VCf IS VC
ð##Þ ! ILout RL þ VCf Rb f ½ f g þ Vbb ¼ 0
bf VT br VT

IS ILout RL VCf IS
ILout RL þ VCf Rb ½ þ Rb V þ Vbb ¼ 0
bf VT br VT Cf
IS 1 IS IS
VCf ½1 þ Rb þ Rb ILout RL ½Rb þ 1 þ Vbb ¼ 0
bf VT br VT bf VT
IS IS
VC ½1 þ Rb b þ Rb b þ Vbb
f VT r VT
ILout ¼ f
IS . We have two expressions for ILout then
RL ½Rb b þ 1
f VT
V
Cf IS IS
RC ½VCC Lout RL þ VT
1 I
½IS aSr VC ½1 þ Rb b þ Rb b þ Vbb
f VT r VT
IS RL ¼ f
IS
½ V þ 1 RL ½Rb b V þ 1
T f T
1 IS VC 1 IS
½VCC Lout RL RL ½Rb þ 1 þ f IS ½1 RL ½Rb þ 1
RC bf VT VT ar b f VT
IS IS I S RL IS R L
¼ VCf ½1 þ Rb þ Rb ½ þ 1 þ ½ þ 1 Vbb
b f VT br VT VT VT
1 1 IS Rb I S 1 1 I S RL
VCf f IS ½1 RL ½Rb þ 1 ½1 þ ð þ Þ ½ þ 1g
VT ar bf VT VT bf br VT
I S RL 1 IS
¼½ þ 1 Vbb ½VCC Lout RL RL ½Rb þ 1
VT RC bf VT
½ISVRT L þ 1 Vbb R1C ½VCC Lout RL RL ½Rb b IVS T þ 1
VCf ’ Rb IS
f
1
VT IS ½1 a1r RL ½Rb b IVS T þ 1 ½1 þ VT ðb1 þ b1 Þ ½ISVRT L þ 1
f f r
VCf ½1 þ Rb b IVS T þ Rb b IVS T þ Vbb VCf ½1 þ Rb b IVS T þ Rb b IVS T

ILout ’ f r
¼ f r
RL ½Rb b IVS T þ 1 RL ½Rb b IVS T þ 1

f f
Vbb
þ
RL ½Rb b IVS T þ 1
f
½ISVRT L þ 1 Vbb R1C ½VCC Lout RL RL ½Rb b IVS T þ 1

ILout ’ f 1 Rb IS
f
g
VT IS ½1 a1r RL ½Rb b IVS T þ 1 ½1 þ VT ðb1 þ b1 Þ ½ISVRT L þ 1
f f r
½1 þ Rb b IVS T þ Rb b IVS T Vbb

f f r
gþ
RL ½Rb b IVS T þ 1 RL ½Rb b IVS T þ 1
f f
There is other way to find our system fixed points. First we write our system
differential equations:
dILout dIL 1 RL
VA1 ¼ Lout þ ILout RL ) out ¼ VA1 ILout
dt dt Lout Lout
dVA1 dVA1 1 1
IC0 ¼ C0 ) ¼ I C0 ¼ ðIRC ICf IC ILout Þ; ICf ¼ Ib IRb ILin
dt dt C0 C0
1 1
I C0 ¼ ðIRC ½Ib IRb ILin IC ILout Þ
C0 C0
1
¼ ðIRC ½Ib þ IC þ IRb þ ILin ILout Þ
C0
VCC VA1 1 1 VCC VA1
¼ IRC ) I C0 ¼ ð ½Ib þ IC þ IRb þ ILin ILout Þ
RC C0 C0 RC
Vbb VA2 1 1 VCC VA1 Vbb VA2
¼ IR b ) I C0 ¼ ð ½Ib þ IC þ ½ þ ILin ILout Þ
Rb C0 C0 RC Rb
dILout dIL
IC ¼ n1 ¼ n1 ðVCf ; ; ILout ; IS ; ar ; . . .Þ; Ib ¼ n2 ¼ n2 ðVCf ; out ; ILout ; IS ; bf ; . . .Þ
dt dt
dVA1 1 VCC VA1 dILout
¼ ð ½n2 ðVCf ; ; ILout ; IS ; bf ; . . .Þ
dt C0 RC dt
dIL Vbb VA2
þ n1 ðVCf ; out ; ILout ; IS ; ar ; . . .Þ þ ½ þ ILin ILout Þ;
dt Rb
VA1 VA2 ¼ VCf ) VA2 ¼ VA1 VCf
dVA1 1 VCC VA1 dIL
¼ ð ½n2 ðVCf ; out ; ILout ; IS ; bf ; . . .Þ
dt C0 RC dt
dIL Vbb ðVA1 VCf Þ
þ n1 ðVCf ; out ; ILout ; IS ; ar ; . . .Þ þ ½ þ ILin ILout Þ
dt Rb
dVCf dVCf 1 1 dIL

I Cf ¼ C f ) ¼ I Cf ¼ ðn2 ðVCf ; out ; ILout ; IS ; bf ; . . .Þ IRb ILin Þ
dt dt Cf Cf dt
dVCf 1 dILout Vbb VA2
¼ ðn2 ðVCf ; ; ILout ; IS ; bf ; . . .Þ ½ ILin Þ
dt Cf dt Rb
dVCf 1 dIL Vbb ðVA1 VCf Þ
¼ ðn2 ðVCf ; out ; ILout ; IS ; bf ; . . .Þ ½ ILin Þ
dt Cf dt Rb
dILin dILin 1
VLin ¼ Lin ) ¼ ðXðtÞ ½VA1 VCf VCin Þ;
dt dt Lin
dVCin dVCin 1 1
ICin ¼ Cin ) ¼ ICin ¼ ILin
dt dt Cin Cin
We can summery our system differential equations:
dILout 1 RL
¼ VA1 ILout
dt Lout Lout
dVA1 1 VCC VA1 dIL
¼ ð ½n2 ðVCf ; out ; ILout ; IS ; bf ; . . .Þ
dt C0 RC dt
þ n1 ðVCf ; out ; ILout ; IS ; ar ; . . .Þ þ ½ þ ILin ILout Þ
dt Rb
dVCf 1 dIL Vbb ðVA1 VCf Þ
¼ ðn2 ðVCf ; out ; ILout ; IS ; bf ; . . .Þ ½ ILin Þ
dt Cf dt Rb
dILin 1 dVCin 1
¼ ðXðtÞ ½VA1 VCf VCin Þ; ¼ ILin
dt Lin dt Cin
At equilibrium points (fixed points):
dILout dVA1 dVCf dILin

¼ 0; ¼ 0; ¼ 0; ¼0
dt dt dt dt
dVCin 1 1 RL
¼ 0: ILin ¼ 0; ðXðtÞ ½VA1 VCf VCin Þ ¼ 0; V ILout
¼0
dt Lin Lout A1 Lout

n2 ðVCf ; out ¼ 0; ILout ; IS ; bf ; . . .Þ ½ ¼0
dt Rb

VCC VA1 dIL dIL
½n2 ðVCf ; out ¼ 0; ILout ; IS ; bf ; . . .Þ þ n1 ðVCf ; out ¼ 0; ILout ; IS ; ar ; . . .Þ
RC dt dt

Vbb ðVA1 VCf Þ
þ½ ILout ¼ 0
Rb
We consider Microwave RFin signal X(t) = A0 + fX(t); |fX(t)| <=1 and A0 > > |
fX(t)| then XðtÞ ¼ A0 þ fX ðtÞ A0 .

A0 ½VA1 VCf VCin ¼ 0; VA1

ILout RL ¼ 0;

dILout Vbb ðVA1 VCf Þ
n2 ðVCf ; ¼ 0; ILout ; IS ; bf ; . . .Þ ½ ¼0
dt Rb

VCC VA1 dIL dIL
½n2 ðVCf ; out ¼ 0; ILout ; IS ; bf ; . . .Þ þ n1 ðVCf ; out ¼ 0; ILout ; IS ; ar ; . . .Þ
RC dt dt

Vbb ðVA1 VCf Þ
þ½ ILout ¼ 0; VA1
ILout RL ¼ 0 ) VA1
¼ ILout RL
Rb
Then we get the following equations:
dILout
A0 ½ILout RL VCf VCin ¼ 0; n2 ðVCf ; ¼ 0; ILout ; IS ; bf ; . . .Þ
dt

Vbb ðILout RL VCf Þ
½ ¼0
Rb

VCC ILout RL dIL
½n2 ðVCf ; out ¼ 0; ILout ; IS ; bf ; . . .Þ
RC dt

dIL Vbb ðILout RL VCf Þ
þ n1 ðVCf ; out ¼ 0; ILout ; IS ; ar ; . . .Þ þ ½ ILout ¼ 0
dt Rb
And we can find analytically or numerically our fixed points (equilibrium points)
value: VCf ; ILout . Additionally ILin

¼ 0; VA1
¼ ILout RL .
Stability analysis: We define the following functions:
1 RL
f1 ðVA1 ; ILout ; VCf ; ILin ; VCin Þ ¼ VA1 ILout
Lout Lout
1 VCC VA1 dIL
f2 ðVA1 ; ILout ; VCf ; ILin ; VCin Þ ¼ ð ½n2 ðVCf ; out ; ILout ; IS ; bf ; . . .Þ
C0 RC dt
dILout Vbb ðVA1 VCf Þ
þ n1 ðVCf ; ; ILout ; IS ; ar ; . . .Þ þ ½ þ ILin ILout Þ
dt Rb
1 dIL
f3 ðVA1 ; ILout ; VCf ; ILin ; VCin Þ ¼ ðn2 ðVCf ; out ; ILout ; IS ; bf ; . . .Þ
Cf dt
Vbb ðVA1 VCf Þ
½ ILin Þ
Rb
1
f4 ðVA1 ; ILout ; VCf ; ILin ; VCin Þ ¼ ðXðtÞ ½VA1 VCf VCin Þ;
Lin
1
f5 ðVA1 ; ILout ; VCf ; ILin ; VCin Þ ¼ ILin
Cin
VA1 ¼ VA1 ðtÞ; ILout ¼ ILout ðtÞ; VCf ¼ VCf ðtÞ; ILin ¼ ILin ðtÞ; VCin ¼ VCin ðtÞ
dIL dIL
n1 ¼ n1 ðVCf ; out ; ILout ; IS ; ar ; . . .Þ; n2 ¼n2 ðVCf ; out ; ILout ; IS ; bf ; . . .Þ
dt dt
To classify our bipolar transistor microwave system fixed points, we need to

compute the Jacobian (linearized system) [2–4].
0 @f1 @f1
1
@VA1 . . . @VCin
B . .. .. C @f1 @f1 @f1 @f1
A¼B @ .. .
C
. A; N11 ¼ @VA1 ; N12 ¼ @IL ; N13 ¼ @VC ; N14 ¼ @ILin ; N15
@f5 @f5 out f
@VA1 @VCin
@f1
¼
@VCin
@f2 @f2 @f2 @f2

N21 ¼ ; N22 ¼ ; N23 ¼ ; N24 ¼ ;
@VA1 @ILout @VCf @ILin
@f2 @f3 @f3
N25 ¼ ; N31 ¼ ; N32 ¼
@VCin @VA1 @ILout
@f3 @f3 @f3 @f4
N33 ¼ ; N34 ¼ ; N35 ¼ ; N41 ¼ ;
@VCf @ILin @VCin @VA1
@f4 @f4 @f4
N42 ¼ ; N43 ¼ ; N44 ¼
@ILout @VCf @ILin
@f4 @f5 @f5 @f5 @f5 @f5
N45 ¼ ; N51 ¼ ; N52 ¼ ; N53 ¼ ; N54 ¼ ; N55 ¼
@VCin @VA1 @ILout @VCf @ILin @VCin
@fk @fk @fk @fk @fk
Nk1 ¼ ; Nk2 ¼ ; Nk3 ¼ ; Nk4 ¼ ; Nk5 ¼ ; k ¼ 1; . . .; 5
@VA1 @ILout @VCf @ILin @VCin
0 1
N11 . . . N15
B . .. .. C @f1 1
A¼B @ .. .
C
. A; N11 ¼ @V ¼ L ;
A1 out
N51 N55
@f1 RL @f1
N12 ¼ ¼ ; N13 ¼ ¼0
@ILout Lout @VCf
@f1 @f1 @f2 1 1 @ X 2
1
N14 ¼ ¼ 0; N15 ¼ ¼ 0; N21 ¼ ¼ ð þ ½ n þ Þ
@ILin @VCin @VA1 C0 RC @VA1 k¼1 k Rb
@f2 1 1 1 IS 1 ð
VA1 VC
f
Þ
N21 ¼ ¼ ð þ þ ½1 þ e VT Þ
@VA1 C0 RC Rb VT bf
1 VCC VA1 Vbb ðVA1 VCf Þ

f2 ¼ ð ½n2 þ n1 þ ½ þ ILin ILout Þ;
C0 RC Rb
@ X 2
IS 1 VA1 VC
f
½ nk ¼ ½1 þ eð VT Þ
@VA1 k¼1 VT bf
dIL
VC Lout þ IL RL VC out
IS f dt out f
n1 ¼ IC ¼ ½eð VT Þ 1 þ IS ½eð VT Þ
1
ar
VC VA1 VC
IS f f
¼ ½eð VT Þ 1 þ IS ½eð VT Þ 1
ar
dIL
Lout out þ I
Lout RL VCf VC
IS ð
dt
Þ IS ð fÞ
n2 ¼ IB ¼ ½e VT
1 þ ½e VT 1
bf br
VA1 VC VC
IS f IS f
¼ ½eð VT Þ 1 þ ½eð VT Þ 1
bf br
Exponent function rules:
d f ðxÞ df ðxÞ f ðxÞ A þ B

½e ¼ e ;e ¼ eA eB
dx dx
dILout 1 @n IS ðVV Cf Þ 1 VA1 @n1 IS ðVA1VVCf Þ
¼ ½VA1 ILout RL ; 1 ¼ e T ½ eð VT Þ ; ¼ e T
dt Lout @VCf VT ar @VA1 VT
@n2 IS VC
f 1 VA1 1 @n2 IS VA1 VC
f
¼ eð VT Þ ½ e VT þ ; ¼ eð VT Þ
@VCf VT bf br @VA1 bf VT
@f2 1 @f2 1
N22 ¼ ¼ ; N23 ¼ ¼
@ILout C0 @VCf C0
@n2 @n1 1 1 1 @ X 2
ð½ þ þ Þ ¼ ½ ð n Þ
@VCf @VCf Rb C0 Rb @VCf k¼1 k
@ X 2
IS ðVV Cf Þ 1 VA1 1 VA1 1
ð nk Þ ¼ e T f½ e VT ½ e VT þ g
@VCf k¼1 VT ar bf br
IS ðVV Cf Þ 1 1 VA1 1
¼ e T ½ e VT ð1 þ Þ
VT a r br bf
@f2 1 1 @ X 2
N23 ¼ ¼ ½ ð n Þ
@VCf C0 Rb @VCf k¼1 k
1 1 IS ðVV Cf Þ 1 1 VA1 1
¼ f e T ½ e VT ð1 þ Þg
C0 Rb VT a r br bf
@f2 1 @f2 @f3
N24 ¼ ¼ ; N25 ¼ ¼ 0; N31 ¼
@ILin C0 @VCin @VA1
1 @n2 1 1 IS VA1 VC
f 1
¼ ð þ Þ¼ ð eð VT Þ þ Þ
Cf @VA1 Rb Cf bf VT Rb
@f3 @f3 1 @n 1
N32 ¼ ¼ 0; N33 ¼ ¼ ð 2 Þ
@ILout @VCf Cf @VCf Rb
1 IS ðVCf Þ 1 VA1 1 1
¼ ð e VT ½ e VT þ þ Þ
Cf VT bf br Rb
@f3 1 @f3 1 Vbb ðVA1 VCf Þ
N34 ¼ ¼ ; N35 ¼ ¼ 0; f3 ¼ ðn2 ½ ILin Þ
@ILin Cf @VCin Cf Rb
We consider Microwave RFin signal X(t) = A0 + fX(t); |fX(t)| 1 and A0 |

fX(t)| then XðtÞ ¼ A0 þ fX ðtÞ A0 .
@f4 1 @f4 @f4 1

N41 ¼ ¼ ; N42 ¼ ¼ 0; N43 ¼ ¼ ;
@VA1 Lin @ILout @VCf Lin
@f4 @f4 1
N44 ¼ ¼ 0; N45 ¼ ¼
@ILin @VCin Lin
1 @f5
f4 ðA0 ½VA1 VCf VCin Þ; N51 ¼ ¼ 0;
Lin @VA1
@f5 @f5
N52 ¼ ¼ 0; N53 ¼ ¼0
@ILout @VCf
@f5 1 @f5 1
N54 ¼ ¼ ; N55 ¼ ¼ 0; f5 ¼ ILin
@ILin Cin @VCin Cin

We already found our system fixed points: ILin ¼ 0; VA1 ¼ ILout RL
E ðVA1

; ILout
; VCf
; ILin
; VCin
Þ ¼ ðILout
RL ; ILout
; VCf
; 0; VCin Þ
Our system Jacobian elements for our fixed points coordinates are:
1 RL 1 1
N11 ¼ ; N12 ¼ ; N13 ¼ 0; N14 ¼ 0; N15 ¼ 0; N22 ¼ ; N24 ¼
Lout Lout C0 C0
I RL V
1 1 1 IS 1 Lout Cf
N21 ¼ ð þ þ ½1 þ eð VT Þ Þ; N25 ¼ 0; N32 ¼ 0
C0 RC Rb VT bf

1 1 IS ðVV Cf Þ 1 1 I R
Lout L
N23 ¼ f e T ½ e VT
C0 Rb VT a r br
I R V
1 1 IS Lout L Cf 1
ð1 þ Þg; N31 ¼ ð eð VT Þ þ Þ
bf Cf bf VT Rb
V
1 IS Cf 1 ILout RL 1 1
N33 ¼ ð eð VT Þ ½ e VT þ þ Þ;
Cf V T bf br Rb
1 1
N34 ¼ ; N35 ¼ 0; N54 ¼ ; N55 ¼ 0
Cf Cin
1 1 1
N41 ¼ ; N42 ¼ 0; N43 ¼ ; N44 ¼ 0; N45 ¼ ; N51 ¼ 0; N52 ¼ 0; N53 ¼ 0
Lin Lin Lin
0 1 0 @f1 @f1 1
N11 ... N15 @VA1 ... @VCin
B . C B C
.. ... C B . .. .. C
A ¼ B
@ .. . A j @E ðVA1
¼ B .. . . C j @E ðVA1

;ILout ;VCf ;ILin ;VCin Þ @ A ;ILout ;VCf ;ILin ;VCin Þ
N51 N55 @f5 @f5
¼ ðILout RL ; ILout ; VCf ; 0; VCin Þ @VA1 @VCin
¼ ðILout RL ; ILout ; VCf ; 0; VCin Þ
ðA k IÞ ¼ ðA k IÞ ) det jA k Ij ¼ 0 To classify our system fixed

points.
We define our system Jacobian elements at fixed points:

Nkl @ðILout RL ; ILout ; VCf ; 0; VCin Þ ! Nkl 8 k ¼ 1; . . .; 5; l ¼ 1; . . .; 5
0 1
ðN33 kÞ C1f 0
1 B C
det jA k Ij ¼ ð kÞ ðN22 kÞ detB
@
1
Lin k L1in C
A
Lout
0 1
Cin k
0 1
ðN33 kÞ C1f 0
RL B C
þ fN21 detB
@
1
Lin k L1in C
A
Lout
0 1
Cin k
0 1 0 1
N31 C1f 0 N31 ðN33 kÞ 0
B 1 C 1 B 1 C
N23 detB
@ Lin k L1in C
A þ C0 det@ Lin
1
Lin L1in Ag
0 1
k 0 0 k
Cin
0 1
ðN33 kÞ C1f 0 !
B C k L1in 1 1
L1in
detB
@
1
k L1in C ¼ ðN kÞ det
A 33 þ det Lin
Lin 1
Cin k Cf 0 k
0 1
Cin k
1 1 1
¼ ðN33 kÞ ðk2 þ Þ k
Cin Lin Cf Lin
1 1 1 N33
¼ k3 þ N33 k2 k ð þ Þþ
Lin Cin Cf Cin Lin
0 1
N31 C1f 0 !
B 1 C k L1in 1 L1in L1in
detB
@ Lin k L1in C
A ¼ N
31 det þ
1
Cin k Cf 0 k
0 1
Cin k
1 1 1 N31
¼ N31 ðk2 þ Þþ k ¼ N31 k2 þ kþ
Cin Lin Cf Lin Cf Lin Cin Lin
0 1
N31 ðN33 kÞ 0 1 1
B 1 C L1in Lin L1in
det@ L1in 1
Lin A ¼ N 31 Lin
ðN 33 kÞ
Lin
0 k 0 k
0 0 k
N 1 1 1
¼ 31 k ðN33 kÞ k¼ k2 k ðN31 þ N33 Þ
Lin Lin Lin Lin
We can summery our last results:
1 1
det jA k Ij ¼ ð kÞ ðN22 kÞ fk3 þ N33 k2 k
Lout Lin
1 1 N33
ð þ Þþ g
Cin Cf Cin Lin
RL 1 1 1 N33
þ fN21 ½k3 þ N33 k2 k ð þ Þþ
Lout Lin Cin Cf Cin Lin
1
N23 ðN31 k2 þ k
Cf Lin
N31 1 1 1
þ Þþ ½ k2 k ðN31 þ N33 Þg
Cin Lin C0 Lin Lin
1 N
det jA k Ij ¼ ½k2 k ð þ N22 Þ þ 22
Lout Lout
1 1 1 N33
½k3 þ N33 k2 k ð þ Þþ
Lin Cin Cf Cin Lin
RL 1 1 1
þ fk3 N21 þ N21 N33 k2 k N21 ð þ Þ
Lout Lin Cin Cf
N N N23
þ 21 33 N23 N31 k2 k
Cin Lin Cf Lin
N N 1 1
31 23 þ k2 k ðN31 þ N33 Þg
Cin Lin Lin C0 Lin C0
1 N
det jA k Ij ¼ ½k2 k ð þ N22 Þ þ 22
Lout Lout
1 1 1 N33
½k þ N33 k k
3 2
ð þ Þþ
Lin Cin Cf Cin Lin
RL 1
þ fk3 N21 þ k2 ½N21 N33 N23 N31 þ
Lout Lin C0
1 1 1 N 1
k ½N21 ð þ Þ þ 23 þ ðN31 þ N33 Þ
Lin Cin Cf Cf C0
1
þ ½N21 N33 N31 N23 g
Cin Lin
1 1 1 N33
det jA k Ij ¼ k5 þ N33 k4 k3 ð þ Þ þ k2
Lin Cin Cf Cin Lin
1 1 1 1 1 1
þ k4 ð þ N22 Þ k3 ð þ N22 Þ N33 þ k2 ð þ Þð þ N22 Þ
Lout Lout Lin Cin Cf Lout
1 N33 N N
kð þ N22 Þ k3 22 þ 22 N33 k2 k
Lout Cin Lin Lout Lout
1 N22 1 1 N33 N22
ð þ Þþ
Lin Lout Cin Cf Cin Lin Lout
RL RL 1
þ k3 N þ k2 ½N21 N33 N23 N31 þ
Lout 21 Lout Lin C0
RL 1 1 1 N 1
k ½N21 ð þ Þ þ 23 þ ðN31 þ N33 Þ
Lout Lin Cin Cf Cf C0
1 RL
þ ½N21 N33 N31 N23
Cin Lin Lout
1
det jA k Ij ¼ k5 þ k4 ½N33 þ þ N22 k3
Lout
1 1 1 1 N RL
½ ð þ Þþð þ N22 Þ N33 þ 22 þ N
Lin Cin Cf Lout Lout Lout 21
N33 1 1 1 1 N
þ k2 f þ ð þ Þð þ N22 Þ þ 22
Cin Lin Lin Cin Cf Lout Lout
R 1
N33 þ ½N21 N33 N23 N31 þ
L
g
Lout Lin C0
1 N33 1 N22 1 1
k fð þ N22 Þ þ ð þ Þ
Lout Cin Lin Lin Lout Cin Cf
RL 1 1 1 N 1
þ ½N21 ð þ Þ þ 23 þ ðN31 þ N33 Þg
N33 N 1 RL
þ 22 þ ½N21 N33 N31 N23
Cin Lin Lout Cin Lin Lout
We get fifth degree polynomial in λ (eigenvalue) and define the following

equilibrium parameters:
1 1 1 1
X5 ¼ 1; X4 ¼ N33 þ þ N22 ; X3 ¼ ½ ð þ Þ
Lout Lin Cin Cf
1 N RL
þð þ N22 Þ N33 þ 22 þ N
Lout Lout Lout 21
N33 1 1 1 1
X2 ¼ þ ð þ Þð þ N22 Þ
Cin Lin Lin Cin Cf Lout
N RL 1
þ 22 N33 þ ½N21 N33 N23 N31 þ
Lout Lout Lin C0
1 N33 1 N22 1 1
X1 ¼ fð þ N22 Þ þ ð þ Þ
Lout Cin Lin Lin Lout Cin Cf
RL 1 1 1 N 1
þ ½N21 ð þ Þ þ 23 þ ðN31 þ N33 Þg;
N33 N 1 RL
X0 ¼ 22 þ ½N21 N33 N31 N23
Cin Lin Lout Cin Lin Lout
X
5 X
5
det jA k Ij ¼ ki Xi ; det jA k Ij ¼ 0 ) ki Xi ¼ 0 ) k1 ; k2 ; . . .
i¼0 i¼0
We need to classify our system stability fixed points according to eigenvalues:
Table 3.1 Bipolar transistor at microwave frequencies system stability fixed points and
eigenvalues
System Eigen values System fixed point classification
P
5
( ki Xi ¼ 0 ) k1 ; k2 ; . . .; kn )
i¼0
Number of eigenvalues is n
1 λk > 0 and real 8 k 2 ½1; . . .; n Unstable node
n, k are integers
2 λk < 0 and real 8 k 2 ½1; . . .; n Stable node
n, k are integers
3 At least one Eigen value is negative real Saddle point
number (λl < 0) and all other Eigenvalues
are positive real number λk > 0
8 k 2 ½1; . . .; n; n, k, l are integers, λl < 0;
0 l n; n; l 2 ½0; . . .; n
4 λk < 0 and real 8k 2 ½1; . . .; n Stable spiral (decay oscillation spiral). If at
n, k are integers except λl, λm list one of our Eigenvalues is positive then
kl ¼ c1 þ j c2 ; km ¼ c1 j c2 we have Saddle point spiral
0 l n; 0 m n; l; m 2 ½1; . . .; n
n, m are integer numbers
c1 \0; c2 [ 0 and real number
l 6¼ m & c1 ¼ Reðkm;n Þ\0
5 λk < 0 and real 8k 2 ½1; . . .; n Unstable spiral (growing oscillation spiral)
n, k are integers except λl, λm
kl ¼ c1 þ j c2 ; km ¼ c1 j c2
0 l n; 0 m n; l; m 2 ½1; . . .; n
(continued)
3.2 Field Effect Transistor (FETs) at Microwave … 299
Table 3.1 (continued)

System Eigen values System fixed point classification
P
5
( ki Xi ¼ 0 ) k1 ; k2 ; . . .; kn )
i¼0
Number of eigenvalues is n
n, m are integer numbers
c1 [ 0; c2 [ 0 and real number
l 6¼ m & c1 ¼ Reðkm;n Þ [ 0
6 kl ¼ c1 þ j c2 ; km ¼ c1 j c2 Solutions are periodic with period T ¼ 2p
c
0 l n ; 0 m n ; l; m 2 ½1; . . .; n
2
l, n, m are integer numbers

c1 ¼ 0; c2 [ 0 and real number
l 6¼ m & c1 ¼ Reðkm;n Þ ¼ 0, km;n are pure
imaginary
3.2 Field Effect Transistor (FETs) at Microwave

Frequencies Description
There are three major types of FETs. First type is Junction FET (JFET). The second
type is the metal oxide semiconductor FET (MOSFET). MOSFETs transistor is
widely use as discrete devices in UHF band communications. The third type is
Schottky barrier type FETs made of gallium arsenide. It is known as the Gallium
Arsenide Metal semiconductor FET (GaAs MESFET) which showed performances
better than bipolar transistor. GaAs MESFET provide lower noise and higher gain for
solid state applications, frequency characteristics previously unavailable from bipolar
transistor. It is made by using gallium arsenide. The electron mobility of gallium
arsenide is five to seven times that of silicon. The GaAs FET is different from the
MOSFET by the use of schottky barrier at the gate instead of an oxide layer. GaAs FET
are called “Normally ON” type device, the maximum gate voltage must be zero. The
design of microwave circuits includes active components GaAs MESFET’s and GaAs
or InP based MOD-FET’s. Small Signal Model (SSM’s) are used as a building block
for large signal modeling nonlinear circuits like power amplifiers, mixers, oscillators,
etc., There are equivalent circuit elements which required if small signal broadband
behavior of microwave and millimeter wave FET’s has to be modeled. Optimization is
done by using nonlinear dynamic. The FET is fabricated on a semi insulating sub-
strate, which serves as the transistor support. An epitaxial layer of N-doped semi-
conductor material is deposited on top of the substrate, and practically the FET is built
into epitaxial layer. The FET’s ports are source, gate and drain. The source is at one
end of the transistor, and the drain is at other end. We connect positive voltage to the
drain, and electrons are drawn from the source to the drain. The gate is between the
source and drain on the surface of the epitaxial layer. Microwave FET is constructed
from a metal to semiconductor junction (Schottky junction) at the gate. Another name
to microwave FET is MESFET (gate is a metal to semiconductor junction). FET
semiconductor material is GaAs and silicon not recommended since electrons travel
twice as fast in GaAs as in silicon. Better high frequency performance is obtained in
GaAs FET. Microwave FETs are made with GaAs and called GaAs FETs. When the
source to gate voltage is zero, the electrons move through the entire thickness of the
epitaxial layer and the FET draws the maximum current, saturated drain to source
current IDSS. A microwave FET operated with its gate voltage negative with respect to
its source voltage. A reversed biased Schottky junction is formed around the gate.
Increasing the negative voltage on the gate causes the size of the insulating barrier
region increases, and reducing the current flow from the source to the drain. If the gate
voltage is negative enough, the insulating region around the gate extend across the
entire epitaxial layer and cut off the current flow. FET amplification is performing
because a small voltage applied to the gate controls a large amount of current which
flowing through the transistor. This current used to generate a large voltage in the
output circuit. We interested on the properties of a GaAs MESFET. A linear amplifier
circuit biases the GaAs MESFET. Considering gate bias alone, the range must be from
IDSS, VG = 0 to IDS = 0 at pinch off, VG = VP. FET square law characteristic is done
according to the formula ID ¼ IDSS ð1 VVGSP Þ2 . In this range, the voltage VDS between
the drain and the source has little effect on the current IDS flowing through the channel.
By changing the gate voltage, VG, the drain to source current can be controlled. There
is a transfer characteristic of a GaAs FET with n channels. This FET transfer char-
acteristic is an important basic parameter in circuit design because it sets the bias
conditions and operating point. The operating point line is directly related to the
mutual conductance gm. Mutual conductance is defined as the ratio of the change in
direct current to the minor change in voltage between gate sources. We differentiate ID
@ID
expression with respect to VGS: gm ¼ @V GS
¼ 2 IDSS ð1 VVGSP Þ ð V1P Þ ¼
2IDSS
VP ð1 VVGSP Þ. The most important characteristic when designing a bias circuit for
small signal GaAs FETs is transfer characteristic. There are two methods to bias a
GaAs FET: dual power source method and self-bias method (Auto-bias) [37, 62].
Dual power source: VP \VGS \0 must always apply to a GaAs FET, and we get
qffiffiffiffiffiffi
the expression for VGS, VGS ¼ VP ð1 IDSS ID
Þ.
Fig. 3.3 FET dual source

bias and FET self (auto) bias
method
Self-bias method (Auto-Bias): This is the most universal method for reducing
electrical potential between a gate and the source when there is only one power
source. If the source resistance is RS, and the operating current is ID, then the drop
in electric potential caused by RS is ID RS and the actual electrical potential
between the gate and the source is VGS ¼ ID RS . VGS is negative and the FET can
be turned on. We can get the value for RS:
qffiffiffiffiffiffi
RS ¼ VIDGS ¼ I1D VP ð1 IDSS
ID
Þ. We implement a system of RF FET typical
Band 2 amplifier. We have two bias voltages: VG < 0, VD > 0, input and output
capacitors Cin, Cout, Microwave RF source X(t) and series resistance Rin. The dual
power source bias method (VG < 0, VD > 0) is appropriate for use in higher fre-
quencies. When we connect the source to the ground terminal, source inductance
can be made relatively small. By using this method, higher gain can be obtained and
a lower noise factor anticipated in the higher frequencies. A large DC voltage is
applied between the source and the drain, and the drain is positive with respect to
the source. The gate is biased at a DC negative voltage, which shows IDS as a
function of the gate voltage, current flows through the FET and through RD drain
resistor. In the FET none of the transistor current flows into the gate circuit, because
the gate junction is reverse biased. In a bipolar transistor, a small of the emitter
current flows into the base. The current flowing from the drain to the source, which
is opposite to electron current flow. The gate length determines the transit time of
the FET. The increased power is obtained by using multiple sources, gates, and
drains. We consider Microwave RFin signal X(t) = A0 + fX(t); |fX(t)| 1 and
A0 |fX(t)| then XðtÞ ¼ A0 þ fX ðtÞ A0 .
Remark: The Microwave applications implementation for GaAs FET is rec-
ommended since it has a greatest advantage in the higher frequency band. GaAs
FETs are far better in term of noise, gain and output power saturation characteristics
compare to silicon bipolar transistor and tunnel diodes. Most small signal FETs are
in low noise amplifiers. They are used over the horizon microwave
Fig. 3.4 FET amplifier with microwave source X(t)

communications, and in earth stations communicating with satellites. A low noise

amplifier is designed by minimizing the noise measure (M) which can show by the
expression M ¼ NF1
1 1
. NF is the amplifier noise factor and G is the amplifier gain.
G
The input output impedances of a GaAs FET, there is a difference in impedance
between maximum gain and minimum NF (Noise Figure). This difference is par-
ticularly apparent at lower frequencies. As the frequencies go higher, the difference
seems to decrease. The NF (Noise Factor) will be low when the gain is maximized
at high frequencies. The Noise Factor (NF) is a function of device gain. There are
two possible equivalent circuit topologies for FET small signal modeling analysis.
The first equivalent circuit topology is present in the below figure. For FET pad
capacitance we define Cpgs, Cpds. Rg, Rs, and Rd are FET ports resistances
respectively. Lg, Ls, and Ld are FET ports inductances respectively. Cgd is the
capacitance between the gate and the drain. Cds is the capacitance between the drain
and the source. Cgs is the capacitance between the gate and the source. Ri is the FET
input resistance. The second equivalent circuit topology 19 parameters small signal
equivalent circuit topology that is implemented in our analysis. The broad band
modeling of millimeter wave FET’s requires that the parasitic elements such as pad
capacitances Cpgs, Cpds, and Cpgd are taken into account. Additionally high per-
formance devices such as InP-based MODFET’s often have rather leaky gates, the
reverse currents of which must be modeled by the resistances Rgs, and Rgd. All
other FET parameters are the same like in the first equivalent circuit.
We define circuit node in below schematic as A1, A2 …
Fig. 3.5 Small signal equivalent circuit of FET

Fig. 3.6 Small signal enhance equivalent circuit for FET (version 1)
The above FET equivalent circuits are for high frequency model and operation,
taking the node capacitors and other elements into account. If we switch to low
frequency small signal FET model, all capacitors in the above model disconnected
and all inductors are short. Capacitor impedance is Zc ¼ xC
1
; lim Zc ) lim xC
1
!
x!e x!e
1 and for inductance impedance is Zl ¼ x L Zl ¼ x L ) lim x L ! e. We get
x!e
the low frequency small signal FET model.
Fig. 3.7 Small signal enhance equivalent circuit for FET (version 2)
Table 3.2 Small signal FET model parameters (JFET, MOSFET)

Parameter JFET MOSFET
gm 0.1–10 mA/V 0.1–20 mA/V or more
Rds 0.1–1 MΩ 1–50 KΩ
Cds 0.1–1 pF 0.1–1 pF
Cgs, Cgd 1–10 pF 1–10 pF
Rgs >108 Ω >1010 Ω
Rgd >108 Ω >1010 Ω
Fig. 3.8 Small signal equivalent circuit for FET (Rgs and Rgd disconnected)
More restricted low frequency small signal FET model is taking Rs, Rg, and Rd
small and we neglect them. The range of parameter values for an FET is present in
the below table:
Rgs and Rgd are high resistance elements which can be taken as disconnected in
our low frequency small signal model.
A High-electron-mobility transistor (HEMT), also known as hetero-structure
FET (HFET) or modulation-doped FET (MODFET), is a field-effect transistor
incorporating a junction between two materials with different band gaps (i.e. a
hetero-junction) as the channel instead of a doped region (as is generally the case
for MOSFET). A commonly used material combination is GaAs with AlGaAs,
though there is wide variation, dependent on the application of the device. Devices
incorporating more indium generally show better high-frequency performance,
while in recent years, gallium nitride HEMTs have attracted attention due to their
high-power performance. HEMT transistors are able to operate at higher frequen-
cies than ordinary transistors, up to millimeter wave frequencies, and are used in
high-frequency products. The minimum, maximum and optimal values for each
parameter of the model used for PHEMT SSM (Small Signal Model) Optimization
([62], Fundamental theory and applications, Vol. 43, No. 10, October 1996) is
describe in the below table. Femtofarads (1 fF = 0.001 pF = 10−15 F). Ω—Ohm.
The unit for the conductance (g) is defined by S ¼ X1 ¼ A=V (A—ampere, V—
voltage). Mho is an alternative name of the same unit, the reciprocal of one ohm.
Mho is derived from spelling ohm backwards and is written with an upside-down
capital Greek letter Omega [62].
We do our circuit analysis only for second equivalent circuit topology (full
version) which is high frequency model and good for Input microwave source X(t).
id ¼ gm v ejxTau ; lim id ¼ gm v; lim ejxTau ¼ 1. We describe our circuit
Tau!e Tau!e
nodes Kirchhoff’s current law and elements equations in the below tables.
P
n
Ik ¼ 0; n is the total number of branches with currents flowing towards or
k¼1
away from the node.
Table 3.3 PHEMT SSM No. Parameter Minimum Maximum Optimal

(Small Signal Model)
parameters, minimum, 1 Lg[nH] 0.005 0.5 0.0479
maximum and optimal values 2 Rg[Ω] 0.5 50 7.97
3 Ls[nH] 0.005 0.5 0.011
4 Rs[Ω] 0.5 50 2.68
5 Ld[nH] 0.005 0.5 0.0257
6 Rd[Ω] 0.5 50 4.14
7 Cpgs[fF] 1 100 24.5
8 Cpds[fF] 1 100 18.2
9 Cpgd[fF] 1 100 3.01
10 Rgs[MΩ] 5 500 394
11 Cgs[fF] 10 1000 86.4
12 Ri[Ω] 0.5 50 1.55
13 Rgd[MΩ] 5 500 485
14 Cgd[fF] 1 100 18.8
15 Rdi[Ω] 0.5 50 10.4
16 gm[mS] 10 1000 65.8
17 s [ps] 0.001 0.1 0.0977
18 Rds[Ω] 10 1000 227
19 Cds[fF] 1 100 30.8
Table 3.4 PHEMT SSM (Small Signal Model) circuit nodes and Kirchhoff’s Current Law (KCL)
Circuit node P
n
Kirchhoff’s current law (KCL) - Ik ¼ 0
k¼1
A1 IRin ¼ ICin
A2 VA2 ¼ Vg
A3 ILg ¼ ICpgd þ IRg þ ICpgs
A4 IRg ¼ IRgs þ ICgs þ IRgd þ ICgd
A5 ICgs ¼ IRi
A6 ICgd ¼ IRdi
A7 IRgd þ IRdi ¼ IRd þ IRds þ ICds þ gm v
A8 IRd þ ICpgd ¼ ILd þ ICpds
A9 IRgs þ ICgs þ gm v þ ICds þ IRds ¼ IRs
A10 IRs ¼ ILs
A11 ILd þ IRd1 ¼ ICout ¼ IRL
A12 VA12 ¼ Vd
Below table describes circuit element and related equation.
Table 3.5 PHEMT SSM (Small Signal Model) circuit elements and elements related equations
Circuit Element related equation Circuit Element related equation
element element
Rin XðtÞVA1
Rin ¼ IRin Ri IRi ¼ VA5RV
i
A9
Cin ICin ¼ Cin dtd ðVA1 þ Vg Þ Cds ICds ¼ Cds dtd ðVA7 VA9 Þ
Lg Vg VA3 ¼ Lg
dILg
dt
Rds IRds ¼ VA7RV
ds
A9
Cpgs dVCpgs
ICpgs ¼ Cpgs dt ¼ Cpgs dVdtA3 Rd IRd ¼ VA7RV
d
A8
Cpgd ICpgd ¼ Cpgd dtd ðVA3 VA8 Þ Cpds ICpds ¼ Cpds dVdtA8
Rgd IRgd ¼ VA4RV
gd
A7 RS IRs ¼ VA9 V
RS
A10
Cgd ICgd ¼ Cgd dtd ðVA4 VA6 Þ LS VA10 ¼ LS dIdtLs

Rdi IRdi ¼ VA6RV
di
A7 Ld VA8 VA11 ¼ Ld dIdtLd
Rg IRg ¼ VA3RV
g
A4 Rd1 IRd1 ¼ Vd V
Rd1
A11
Rgs IRgs ¼ VA4RV

gs
A9 Cout ICout ¼ Cout dtd ðVA11 Vout Þ
Cgs ICgs ¼ Cgs dv
dt ¼ Cgs dt ðVA4 VA5 Þ
d RL IRL ¼ VRoutL
We have two methods to investigate our system fixed points.

Method A:
Vg and Vd are constant circuit biasing voltages Vg [ 0 ; Vd [ 0 ;
dVg
dt ¼ 0 ; dt ¼ 0
dVd
XðtÞ VA1
¼ IRin ) XðtÞ VA1 ¼ Rin IRin ) VA1 ¼ XðtÞ Rin IRin
Rin
d dVA1 dVA1 1
ICin ¼ Cin ðVA1 þ Vg Þ ) ICin ¼ Cin ) ¼ ICin
dt dt dt Cin
dILg dILg 1
Vg VA3 ¼ Lg ) ¼ ðVg þ VA3 Þ;
dt dt Lg
dVA3 dVA3 1
ICpgs ¼ Cpgs ) ¼ ICpgs
dt dt Cpgs
d d 1
ICpgd ¼ Cpgd ðVA3 VA8 Þ ) ðVA3 VA8 Þ ¼ ICpgd ;
dt dt Cpgd
VA4 VA7
IRgd ¼ ) VA4 VA7 ¼ IRgd Rgd
Rgd
d d 1
ICgd ¼ Cgd ðVA4 VA6 Þ ) ðVA4 VA6 Þ ¼ ICgd ;
dt dt Cgd
VA6 VA7
IRdi ¼ ) VA6 VA7 ¼ IRdi Rdi
Rdi
VA3 VA4 VA4 VA9
IRg ¼ ) VA3 VA4 ¼ IRg Rg ; IRgs ¼ ) VA4 VA9 ¼ IRgs Rgs ;
Rg Rgs
dv dv 1
ICgs ¼ Cgs ) ¼ ICgs
dt dt Cgs
VA5 VA9
IRi ¼ ) VA5 VA9 ¼ IRi Ri ;
Ri
d d 1
ICds ¼ Cds ðVA7 VA9 Þ ) ðVA7 VA9 Þ ¼ ICds
dt dt Cds
VA7 VA9 VA7 VA8
IRds ¼ ) VA7 VA9 ¼ IRds Rds ; IRd ¼ ) VA7 VA8 ¼ IRd Rd
Rds Rd
dVA8 dVA8 1 VA9 VA10
ICpds ¼ Cpds ) ¼ ICpds ; IRs ¼ ) VA9 VA10 ¼ IRs RS
dt dt Cpds RS
dILs dILs 1 dILd dILd 1
VA10 ¼ LS ) ¼ VA10 ; VA8 VA11 ¼ Ld ) ¼ ðVA8 VA11 Þ
dt dt LS dt dt Ld
Vd VA11
IRd1 ¼ ) Vd VA11 ¼ IRd1 Rd1 ) VA11 ¼ Vd IRd1 Rd1
Rd1
d d 1
ICout ¼ Cout ðVA11 Vout Þ ) ðVA11 Vout Þ ¼ ICout ;
dt dt Cout
Vout
IRL ¼ ) Vout ¼ IRL RL
RL
We do some mathematical manipulations:
dVA1 1 d½XðtÞ Rin IRin 1 dXðtÞ dIRin 1

¼ ICin ) ¼ ICin ) Rin ¼ ICin
dt Cin dt Cin dt dt Cin
dXðtÞ dA0 dIRin 1
XðtÞ ¼ A0 þ fX ðtÞ A0 ) ! e; ¼ 0; ¼ ICin ; IRin ¼ ICin
dt dt dt Cin Rin
dIRin 1 dILg 1 dVA3 1
¼ IRin ; ¼ ðVg þ VA3 Þ; ¼ ICpgs
dt Cin Rin dt Lg dt Cpgs
d 1 dVA3 dVA8 1
ðVA3 VA8 Þ ¼ ICpgd ) ¼ ICpgd
dt Cpgd dt dt Cpgd
dVA8 1 1
) ¼ ICpgs ICpgd
dt Cpgs Cpgd
VA4 VA7 ¼ IRgd Rgd & VA6 VA7 ¼ IRdi Rdi ) VA4 VA6 ¼ IRgd Rgd IRdi Rdi
d 1 d 1
ðVA4 VA6 Þ ¼ ICgd ) ðIRgd Rgd IRdi Rdi Þ ¼ ICgd
dt Cgd dt Cgd
d 1 dIRgd dIRdi 1
ðIRgd Rgd IRdi Rdi Þ ¼ ICgd ) Rgd Rdi ¼ ICgd
dt Cgd dt dt Cgd
VA3 VA4 ¼ IRg Rg & VA4 VA9 ¼ IRgs Rgs ) VA3 VA9 ¼ IRg Rg þ IRgs Rgs
dv 1 d
¼ ICgs ; VA5 VA9 ¼ IRi Ri ; VA7 VA9 ¼ IRds Rds & ðVA7 VA9 Þ
dt Cgs dt
1 dIRds 1
¼ ICds ) ¼ ICds
Cds dt Cds Rds
dVA8 1 dILs 1
VA7 VA8 ¼ IRd Rd ; VA9 VA10 ¼ IRs RS ; ¼ ICpds ; ¼ VA10
dt Cpds dt LS
d 1 dVA3 dVA8 1
ðVA3 VA8 Þ ¼ ICpgd ) ¼
dt Cpgd dt dt Cpgd
dVA8 1 1
ICpgd ) ¼ ICpgs ICpgd
dt Cpgs Cpgd
VA4 VA7 ¼ IRgd Rgd & VA6 VA7 ¼ IRdi Rdi ) VA4 VA6 ¼ IRgd Rgd IRdi Rdi
dILd 1
¼ ðVA8 VA11 Þ & VA11 ¼ Vd IRd1
dt Ld
dILd 1
Rd1 ) ¼ ðVA8 Vd þ IRd1 Rd1 Þ
dt Ld
d 1 d 1
ðVA11 Vout Þ ¼ ICout & Vout ¼ IRL RL ) ðVA11 IRL RL Þ ¼ ICout
dt Cout dt Cout
d 1 d 1
ðVA11 IRL RL Þ ¼ ICout ) ðVd IRd1 Rd1 IRL RL Þ ¼ ICout
dt Cout dt Cout
d 1
ðVd IRd1 Rd1 IRL RL Þ ¼ ICout ;
dt Cout
dVd dIRd1 dIR 1
ICout ¼ IRL ) Rd1 L RL ¼ IR L
dt dt dt Cout
dVd dIRd1 dIRL 1
¼0) Rd1 þ RL ¼
dt dt dt Cout
IRL ; ILd þ IRd1 ¼ IRL ) IRd1 ¼ IRL ILd
dðIRL ILd Þ dIRL 1
Rd1 þ RL ¼ IRL
dt dt Cout
dIRL dILd dIRL 1
) Rd1 Rd1 þ RL ¼ IR L
dt dt dt Cout
dIRL dILd dIRL 1
Rd1 Rd1 þ RL ¼ IRL
dt dt dt Cout
dIRL dILd 1
) ðRd1 þ RL Þ Rd1 ¼ IR L
dt dt Cout
VA7 VA9 ¼ IRds Rds ) VA9 ¼ VA7 IRds Rds ;

VA9 VA10 ¼ IRs RS ) VA10 ¼ VA9 IRs RS
VA10 ¼ VA7 IRds Rds IRs RS ; VA7 ¼ IRd Rd þ VA8 ;
VA10 ¼ IRd Rd þ VA8 IRds Rds IRs RS
dILs 1 1
¼ VA10 ¼ ðIRd Rd þ VA8 IRds Rds IRs RS Þ
dt LS LS
dILs 1
¼ ðIRd Rd þ VA8 IRds Rds ILs RS Þ
dt LS
dIRL dILd 1
ðRd1 þ RL Þ Rd1 ¼ IRL
dt dt Cout
dIRL 1 dILd 1
) ¼ ½ Rd1 IRL
dt ðRd1 þ RL Þ dt Cout
dIRL 1 1 1
¼ ½ ðVA8 Vd þ IRd1 Rd1 Þ Rd1 IRL
dt ðRd1 þ RL Þ Ld Cout
The condition to find our system fixed points is no variation with time for our
00 00
variables. d Variable
dt ¼ 0.
Table 3.6 Small signal equivalent circuit of FET system differential equations, fixed points and
outcome
System differential equation @ Fixed points Outcome

dt ¼ Cin Rin IRin
dIRin 1
dt ¼ 0
dIRin IRin ¼0

dILg dILg ¼ Vg
dt ¼ Lg ðVg þ VA3 Þ dt ¼ 0
1 VA3

dVA3
dt ¼ C1pgs ICpgs dVA3
dt ¼0 ICpgs ¼0

dVA8
dt ¼ C1pgs ICpgs Cpgd
1
ICpgd dVA8
dt ¼0 ICpgd
¼
Cpgd
ICpgs Cpgs
(continued)

System differential equation @ Fixed points Outcome

Rgd
dIRgd
Rdi dIRdi
¼ 1
ICgd
dIRgd
¼ 0; dIdtRdi ¼0 ICgd ¼0
dt dt Cgd dt

dv
dt ¼ C1gs ICgs dv
dt ¼0 ICgs ¼0

dIRds
dt ¼ Cds1Rds ICds dIRds
dt ¼0 ICds ¼0

dVA8
dt ¼ C1pds ICpds dVA8
dt ¼0 ICpds ¼0

dILs
dt ¼ L1S VA10 dILs
dt ¼0 VA10 ¼0

dILd
dt ¼ L1d ðVA8 Vd þ IRd1 Rd1 Þ dILd
dt ¼0 VA8 þ IRd1 Rd1 ¼ Vd
Rd1 þ
dIRL
RL ¼ C1out IRL
dIRL IRL ¼ 0
dt ¼ 0; dt ¼0
dIRd1 dIRd1
dt dt

dILs 1
dt ¼ 0
dILs
IRd Rd þ VA8 IRds Rds
¼ ðIRd Rd þ VA8 IRds Rds
dt LS
ILs RS ¼ 0
ILs RS Þ

dIRL
¼
1 1
½ ðVA8 Vd
dIRL
¼0
Rd1
Ld ðVA8 Vd þ IRd1 Rd1 Þ ¼ 0
dt
dt ðRd1 þ RL Þ Ld
1
þ IRd1 Rd1 Þ Rd1 IRL
Cout
Method B:
dV
Vg and Vd are constant circuit biasing voltages Vg [ 0; Vd [ 0; dtg ¼ 0; dVdtd ¼ 0
XðtÞVA1
(1) KCL @ circuit node A1: Rin ¼ Cin dtd ðVA1 þ Vg Þ.
(2) KCL @ circuit node A3: ILg ¼ Cpgd dtd ðVA3 VA8 Þ þ VA3RV g
A4
þ Cpgs dVdtA3 .
(3) KCL @ circuit node A4:
VA3 VA4 VA4 VA9 VA4 VA7
Rg ¼ Rgs þ Cgs dv
dt þ Rgd þ C gd d
dt ðV A4 VA6 Þ.
VA5 VA9
(4) KCL @ circuit node A5: Cgs dv
dt ¼ Ri .
(5) KCL @ circuit node A6: Cgd dtd ðVA4 VA6 Þ ¼ VA6RV
di
A7
VA4 VA7 VA6 VA7 VA7 VA8 VA7 VA9 d

þ ¼ þ þ Cds ðVA7
Rgd Rdi Rd Rds dt
VA9 Þ þ gm v
VA7 VA8
(7) KCL @ circuit node A8: Rd þ Cpgd dtd ðVA3 VA8 Þ ¼ ILd þ Cpds dVdtA8
VA4 VA9 dv d VA7 VA9

þ Cgs þ gm v þ Cds ðVA7 VA9 Þ þ
Rgs dt dt Rds
VA9 VA10
¼
Rs
VA9 VA10
(9) KCL @ circuit node A10: Rs ¼ ILs .
Vd VA11 d Vout
ILd þ ¼ Cout ðVA11 Vout Þ ¼
Rd1 dt RL
dILg dILs dILd
Vg VA3 ¼ Lg ; VA10 ¼ LS ; VA8 VA11 ¼ Ld
dt dt dt
ILg ¼ Cpgd dtd ðVA3 VA8 Þ þ VA3RV A4
þ Cpgs dVA3
dt )
ð2Þ
g
ILg ¼ ½Cpgs þ Cpgd dt Cpgd dt þ Rg Rg

dVA3 dVA8 VA3 VA4
dVA8 dVA3 VA3 VA4

Cpgd ¼ ½Cpgs þ Cpgd ILg þ )
dt dt Rg Rg
dVA8 Cpgs dVA3 1 VA3 VA4
¼½ þ 1 ILg þ
dt Cpgd dt Cpgd Cpgd Rg Cpgd Rg
VA7 VA8 d dVA8

ð7Þ þ Cpgd ðVA3 VA8 Þ ¼ ILd þ Cpds
Rd dt dt
VA7 VA8 dVA3 dVA8 dVA8
þ Cpgd ILd ¼ Cpds þ Cpgd
Rd dt dt dt
dVA8 VA7 VA8 Cpgd dVA3 1
¼ þ ILd
dt Rd ½Cpds þ Cpgd ½Cpds þ Cpgd dt ½Cpds þ Cpgd
C
½Cpgd
pgs
þ 1 dVdtA3 C1pgd ILg þ Cpgd Rg Cpgd Rg
VA3 VA4
ð2Þ ¼ ð7Þ ! VA7 VA8 Cpgd
¼ Rd ½C pds þ Cpgd
þ ½Cpds þ Cpgd dt ½Cpds þ Cpgd
dVA3 1
ILd
Cpgs Cpgd dVA3

f½ þ 1 g
Cpgd ½Cpds þ Cpgd dt
VA7 VA8 1 1 VA3 VA4
¼ ILd þ ILg þ
Rd ½Cpds þ Cpgd ½Cpds þ Cpgd Cpgd Cpgd Rg Cpgd Rg
C C
We define for simplicity: CT1 ¼ Cpgd
pgs
½Cpds þpgdCpgd þ 1
Cpgs Cpgd Cpgs ½Cpds þ Cpgd þ ½Cpds þ Cpgd Cpgd ½Cpgd 2

CT1 ¼ þ1 ¼
Cpgd ½Cpds þ Cpgd ½Cpds þ Cpgd Cpgd
Cpgs Cpds þ Cpgs Cpgd þ Cpds Cpgd
CT1 ¼
½Cpds þ Cpgd Cpgd
Cpgs Cpds þ Cpgs Cpgd þ Cpds Cpgd dVA3

f g
½Cpds þ Cpgd Cpgd dt
VA7 VA8 1 1 VA3 VA4
¼ ILd þ ILg þ
Rd ½Cpds þ Cpgd ½Cpds þ Cpgd Cpgd Cpgd Rg Cpgd Rg
dVA3 VA7 VA8 1 1

¼ ILd þ ILg
dt Rd ½Cpds þ Cpgd CT1 ½Cpds þ Cpgd CT1 Cpgd CT1
VA3 VA4
þ
Cpgd Rg CT1 Cpgd Rg CT1
XðtÞ VA1 d XðtÞ VA1

¼ Cin ðVA1 þ Vg Þ )
Rin dt Rin
dVA1 dVg dVg
ð1Þ ¼ Cin Cin ; ¼0
dt dt dt
dVA1 XðtÞ VA1
¼ :
dt Rin Cin
VA3 VA4 VA4 VA9 dv VA4 VA7

¼ þ Cgs þ
Rg Rgs dt Rgd
d dv VA5 VA9
ð3Þ þ Cgd ðVA4 VA6 Þ; Cgs ¼
dt dt Ri
¼ þ þ þ Cgd ðVA4 VA6 Þ
Rg Rgs Ri Rgd dt
d VA3 VA4 ðVA4 VA9 Þ ðVA5 VA9 Þ ðVA4 VA7 Þ

Cgd ðVA4 VA6 Þ ¼
dt Rg Rgs Ri Rgd
dVA4 dVA6 VA3 VA4 ðVA4 VA9 Þ ðVA5 VA9 Þ ðVA4 VA7 Þ
¼ þ
dt dt Cgd Rg Cgd Rgs Cgd Ri Cgd Rgd
d VA6 VA7 dVA4 dVA6 VA6 VA7

ð5Þ Cgd ðVA4 VA6 Þ ¼ ) ¼ þ
dt Rdi dt dt Cgd Rdi
(3) = (5) →
VA3 VA4 ðVA4 VA9 Þ ðVA5 VA9 Þ ðVA4 VA7 Þ VA6 VA7
¼
Cgd Rg Cgd Rgs Cgd Ri Cgd Rgd Cgd Rdi
dv VA5 VA9 dv VA5 VA9

ð4Þ Cgs ¼ ) ¼
dt Ri dt Cgs Ri

ð6Þ þ ¼ þ þ Cds ðVA7
Rgd Rdi Rd Rds dt
VA9 Þ þ gm v
d VA4 VA7 VA6 VA7 ðVA7 VA8 Þ ðVA7 VA9 Þ

Cds ðVA7 VA9 Þ ¼ þ gm
dt Rgd Rdi Rd Rds
v
VA4 VA9 dv d VA7 VA9

ð8Þ þ Cgs þ gm v þ Cds ðVA7 VA9 Þ þ
Rgs dt dt Rds
VA9 VA10
¼
Rs
dv VA5 VA9 VA4 VA9 VA5 VA9 d

Cgs ¼ ; þ þ gm v þ Cds ðVA7
dt Ri Rgs Ri dt
VA7 VA9
VA9 Þ þ
Rds
VA9 VA10
¼
Rs
d VA9 VA10 ðVA4 VA9 Þ ðVA5 VA9 Þ ðVA7 VA9 Þ

Cds ðVA7 VA9 Þ ¼
dt Rs Rgs Ri Rds
gm v
(6) = (8) →
VA4 VA7 VA6 VA7 ðVA7 VA8 Þ
þ
Rgd Rdi Rd
VA9 VA10 ðVA4 VA9 Þ ðVA5 VA9 Þ
¼
Rs Rgs Ri
We can summery our last results (circuit node voltages equations) in the below
table:
Table 3.7 Small signal equivalent circuit of FET nodes KCL and circuit nodes voltages equations
Nodes Circuit node voltages equation
KCL
(3) = (5) VA3 VA4
Cgd Rg ðVCA4gdVA9 Þ ðVA5 VA9 Þ ðVA4 VA7 Þ VA6 VA7
Rgs Cgd Ri Cgd Rgd ¼ Cgd Rdi
(*)
(6) = (8) VA4 VA7
Rgd þ VA6 VA7
Rdi ðVA7RV
d
A8 Þ
¼ VA9 V
Rs
A10
ðVA4RV
gs
A9 Þ
ðVA5RV
i
A9 Þ
(**)
dILs VA9 VA10 LS d

VA10 ¼ LS ; IRs ¼ ILs ; IRs ¼ ; VA10 ¼ ðVA9 VA10 Þ
dt RS RS dt
dILg dILg 1
Vg VA3 ¼ Lg ) ¼ ðVg þ VA3 Þ ;
dt dt Lg
dILd dILd 1
VA8 VA11 ¼ Ld ) ¼ ðVA8 VA11 Þ
dt dt Ld
Vd VA11 Vout ðVd VA11 Þ RL
ILd þ ¼ ) Vout ¼ ILd RL þ ;
Rd1 RL Rd1
Vd VA11 d
ILd þ ¼ Cout ðVA11 Vout Þ
Rd1 dt
ðVd VA11 Þ RL
Vout ¼ ILd RL þ ) Vout VA11
Rd1
ðVd VA11 Þ RL
¼ ILd RL þ VA11
Rd1
ðVd þ VA11 Þ RL
VA11 Vout ¼ ILd RL þ þ VA11 ;
Rd1
d dVA11 dVd
ðVA11 Vd Þ ¼ ; ¼0
dt dt dt
dVA11 dVout dILd RL d dVA11
¼ RL þ ðVA11 Vd Þ þ
dt dt dt Rd1 dt dt
d dILd RL dVA11 dILd 1

ðVA11 Vout Þ ¼ RL þ ½ þ 1 ; ¼ ½VA8 VA11
dt dt Rd1 dt dt Ld
d RL RL dVA11
ðVA11 Vout Þ ¼ ½VA8 VA11 þ ½ þ 1
dt Ld Rd1 dt
Vd VA11 d Vd VA11
ILd þ ¼ Cout ðVA11 Vout Þ ) ILd þ
Rd1 dt Rd1
RL dVA11 RL
¼ Cout f½ þ 1 ½VA8 VA11 g
Rd1 dt Ld
RL dVA11 Vd VA11 RL
Cout ½ þ 1 ¼ ILd þ þ Cout ½VA8 VA11
Rd1 dt Rd1 Ld
dVA11 1 Vd
¼ ILd þ
dt Cout ½RRd1L þ 1 Cout ½RRd1L þ 1 Rd1
VA11 RL
þ RL ½VA8 VA11
Cout ½Rd1 þ 1 Rd1 ½Rd1 þ 1 Ld
RL
dVA11 1 Vd RL
¼ ILd þ þ VA8
dt Cout ½RRd1L þ 1 Cout ½RRd1L þ 1 Rd1 ½RRd1L þ 1 Ld
RL 1 1
½ þ RL VA11
Ld Cout Rd1 ½R þ 1
d1
We can summery our system new differential equations representation:
dVA3 VA7 VA8 1 1

¼ ILd þ ILg
VA3 VA4
þ
d VA6 VA7 dv 1 ðVA5 VA9 Þ

ðVA4 VA6 Þ ¼ ; ¼
dt Cgd Rdi dt Cgs Ri
d VA4 VA7 VA6 VA7 ðVA7 VA8 Þ ðVA7 VA9 Þ gm
ðVA7 VA9 Þ ¼ þ v
dt Cds Rgd Cds Rdi Cds Rd Cds Rds Cds
dILs VA9 VA10 d

VA10 ¼ LS ; ¼ ILs ) ðVA9 VA10 Þ
dt Rs dt
Rs dILg 1
¼ VA10 ; ¼ ðVg þ VA3 Þ
LS dt Lg
dILd 1
¼ ðVA8 VA11 Þ
dt Ld
dVA11 1 Vd RL
¼ ILd þ þ RL VA8
dt Cout ½Rd1 þ 1
R L
Cout ½Rd1 þ 1 Rd1 ½Rd1 þ 1 Ld
RL
RL 1 1
½ þ VA11
Ld Cout Rd1 ½RRL þ 1
d1
We need to find our system fixed points. First we analyze our circuit node
voltages equations.
VA3 VA4 ðVA4 VA9 Þ ðVA5 VA9 Þ ðVA4 VA7 Þ VA6 VA7
ð Þ ¼
Cgd Rg Cgd Rgs Cgd Ri Cgd Rgd Cgd Rdi
VA3 VA4 VA4 VA9 VA5 VA9 VA4
þ þ
Cgd Rg Cgd Rg Cgd Rgs Cgd Rgs Cgd Ri Cgd Ri Cgd Rgd
VA7 VA6 VA7
þ ¼
Cgd Rgd Cgd Rdi Cgd Rdi
VA3 1 1 1 VA5 VA6

VA4 f þ þ g
Cgd Rg Cgd Rg Cgd Rgs Cgd Rgd Cgd Ri Cgd Rdi
1 1 1 1
þ VA7 f þ g þ VA9 f þ g¼0
Cgd Rdi Cgd Rgd Cgd Rgs Cgd Ri
VA3 1 1 1 VA5 VA6

VA4 f þ þ g
Cgd Rg Cgd Rg Cgd Rgs Cgd Rgd Cgd Ri Cgd Rdi
1 1
þ VA7 f þ g þ VA8 0 þ VA9
Cgd Rdi Cgd Rgd
1 1
f þ g þ VA10 0 ¼ 0
Cgd Rgs Cgd Ri
1 1 1 1
C11 ¼ ; C12 ¼ f þ þ g; C13
Cgd Rg Cgd Rg Cgd Rgs Cgd Rgd
1 1
¼ ; C14 ¼
Cgd Ri Cgd Rdi
1 1 1 1
C15 ¼ þ ; C16 ¼ 0; C17 ¼ þ ; C18
Cgd Rdi Cgd Rgd Cgd Rgs Cgd Ri
X
8
¼ 0; C1k VAk þ 2 ¼ 0
k¼1
VA4 VA7 VA6 VA7 ðVA7 VA8 Þ

ðÞ þ
Rgd Rdi Rd
VA9 VA10 ðVA4 VA9 Þ ðVA5 VA9 Þ
¼
Rs Rgs Ri
VA4 VA7 VA6 VA7 VA7 VA8 VA9 VA10 VA4 VA9 VA5 VA9
þ þ ¼ þ þ
Rgd Rgd Rdi Rdi Rd Rd Rs Rs Rgs Rgs Ri Ri
1 1 VA5 VA6 1 1 1 VA8

VA4 ð þ Þþ þ VA7 f þ þ gþ VA9
Rgd Rgs Ri Rdi Rgd Rdi Rd Rd
1 1 1 VA10
f þ þ gþ
Rs Rgs Ri Rs
¼0
1 1 VA5 VA6 1 1 1
VA3 0 þ VA4 ð þ Þþ þ VA7 f þ þ g
Rgd Rgs Ri Rdi Rgd Rdi Rd
VA8 1 1 1 VA10
þ VA9 f þ þ gþ ¼0
Rd Rs Rgs Ri Rs
1 1 1 1
C21 ¼ 0; C22 ¼ þ ; C23 ¼ ; C24 ¼ ;
Rgd Rgs Ri Rdi
1 1 1 1
C25 ¼ f þ þ g; C26 ¼
Rgd Rdi Rd Rd
1 1 1 1 X 8
C27 ¼ f þ þ g; C28 ¼ ; C2k VAk þ 2 ¼ 0
Rs Rgs Ri Rs k¼1
ðÞðÞ
X
8 X
8 X
8
C1k VAk þ 2 C2k VAk þ 2 ¼ 0; C1k VAk þ 2
k¼1 k¼1 k¼1
X
8 X
8
C2k VAk þ 2 ¼ ðC1k C2k Þ VAk þ 2
k¼1 k¼1
VA3 C11 þ VA4 ðC12 C22 Þ þ VA5 ðC13 C23 Þ þ VA6 ðC14 C24 Þ þ VA7 ðC15 C25 Þ
VA8 C26 þ VA9 ðC17 C27 Þ VA10 C28 ¼ 0
At fixed points:
dVA3 d dv d
¼ 0 ; ðVA4 VA6 Þ ¼ 0 ; ¼ 0 ; ðVA7 VA9 Þ ¼ 0
dt dt dt dt
d dILg dILd dVA11
ðVA9 VA10 Þ ¼ 0 ; ¼ 0; ¼ 0; ¼0
dt dt dt dt
Fixed points: E ¼ ðVA3

; . . .; VA11
; ILd
; ILg ; v Þ

VA7 VA8 1 1
I þ I
Rd ½Cpds þ Cpgd CT1 ½Cpds þ Cpgd CT1 Ld Cpgd CT1 Lg

VA3 VA4
þ
¼0

VA6 VA7 ¼ 0; VA5 VA9 ¼ 0; VA10 ¼ 0; VA3 ¼ Vg ; VA8 VA11 ¼0

VA4 VA6 1 1 1 1
þ VA7 ½
Cds Rgd Cds Rdi Cds Rgd Cds Rdi Cds Rd Cds Rds

VA8 VA9 gm
þ þ v ¼0
Cds Rd Cds Rds Cds
1 Vd RL
I þ þ
Cout ½RRd1L þ 1 Ld Cout ½RRd1L þ 1 Rd1 ½RRd1L þ 1 Ld
RL 1 1
VA8 ½ þ RL V ¼ 0
Ld Cout Rd1 ½R þ 1 A11
d1
We can minimize the above fixed points equations:

VA7 VA8 1 1
ILd þ
Rd ½Cpds þ Cpgd CT1 ½Cpds þ Cpgd CT1 Cpgd CT1

Vg VA4
ILg þ þ
¼0

VA4 1 1 1 1
VA7 ½ þ þ þ
Cds Rgd Cds Rgd Cds Rdi Cds Rd Cds Rds

1 VA8 VA9 gm
þ þ v
Cds Rdi Cds Rd Cds Rds Cds
¼0
1 Vd RL
ILd þ þ VA8 f RL
Cout ½Rd1 þ 1
RL
Cout ½Rd1 þ 1 Rd1
RL
½Rd1 þ 1 Ld
RL 1 1
½ þ g
d1
¼0
X
8

VA6 ¼ VA7 ; VA5 ¼ VA9 ; VA10 ¼ 0; VA3 ¼ Vg ; VA8 ¼ VA11 ; C1k VAk þ 2
k¼1
X
8
¼ 0; C2k VAk þ 2 ¼ 0
k¼1
3.3 Field Effect Transistor (FETs) at Microwave

Frequencies Stability Analysis
In Sect. 3.2, we present FET system’s differential equations representation:
dVA3 VA7 VA8 1 1

¼ ILd þ ILg
VA3 VA4
þ
d VA6 VA7 dv 1 ðVA5 VA9 Þ

ðVA4 VA6 Þ ¼ ; ¼
dt Cgd Rdi dt Cgs Ri
dILs VA9 VA10 d

VA10 ¼ LS ; ¼ ILs ) ðVA9 VA10 Þ
dt Rs dt
Rs dILg 1
¼ VA10 ; ¼ ðVg þ VA3 Þ
LS dt Lg
dILd 1
¼ ðVA8 VA11 Þ
dt Ld
dVA11 1 Vd RL
¼ ILd þ þ VA8
RL 1 1
½ þ RL VA11
d1
We need to discuss its stability analysis under parameter variation.

dI
We derivate the first equation dVdtA3 ¼ ; dIdtLd ¼ L1d ðVA8 VA11 Þ; dtLg ¼
L1g ðVg þ VA3 Þ
d 2 VA3 d VA7 VA8 1 1

¼ f ILd þ ILg
dt2 dt Rd ½Cpds þ Cpgd CT1 ½Cpds þ Cpgd CT1 Cpgd CT1
VA3 VA4
þ g
d 2 VA3 1 dðVA7 VA8 Þ

¼
dt 2 Rd ½Cpds þ Cpgd CT1 dt
1 dILd 1 dILg
þ
½Cpds þ Cpgd CT1 dt Cpgd CT1 dt
1 dVA3 1 dVA4
þ
Cpgd Rg CT1 dt Cpgd Rg CT1 dt
dILg
Inserting expressions: dILd
dt ¼ ; dt ¼
d 2 VA3 1 dðVA7 VA8 Þ 1

¼
dt 2 Rd ½Cpds þ Cpgd CT1 dt ½Cpds þ Cpgd CT1
1 1 1
ðVA8 VA11 Þ ðVg þ VA3 Þ
Ld Cpgd CT1 Lg
1 dVA3 1 dVA4
þ
d 2 VA3 1 dðVA7 VA8 Þ 1

þ
dt2 Rd ½Cpds þ Cpgd CT1 dt ½Cpds þ Cpgd CT1
1 1 1
ðVA8 VA11 Þ þ ðVg þ VA3 Þ
Ld Cpgd CT1 Lg
1 dVA3 1 dVA4
þ ¼0
d 2 VA3 1 dVA7 1
þ
dt2 Rd ½Cpds þ Cpgd CT1 dt Rd ½Cpds þ Cpgd CT1
dVA8 1 1
þ VA8
dt ½Cpds þ Cpgd CT1 Ld
1 1 1 1
VA11 þ
½Cpds þ Cpgd CT1 Ld Cpgd CT1 Lg
1 1 1 dVA3
Vg þ VA3 þ
Cpgd CT1 Lg Cpgd Rg CT1 dt
1 dVA4
¼0
Cpgd Rg CT1 dt
d 2 VA3
We define the following new variables: dVA3
dt ¼ Y1 ; dt2 ¼ dYdt1 ; dVA4
dt ¼ Y2
dVA7 dVA8
¼ Y3 ; ¼ Y4
dt dt
dY1 1 1
Y3 þ
dt Rd ½Cpds þ Cpgd CT1 Rd ½Cpds þ Cpgd CT1
1 1
Y4 þ VA8
½Cpds þ Cpgd CT1 Ld
1 1 1 1 1
VA11 þ Vg þ
½Cpds þ Cpgd CT1 Ld Cpgd CT1 Lg Cpgd CT1
1 1 1
VA3 þ Y1 Y2 ¼ 0
Lg Cpgd Rg CT1 Cpgd Rg CT1
dY1 1 1
¼ Y3
1 1
Y4 VA8
1 1 1 1
þ VA11 Vg
1 1 1
VA3 Y1
Cpgd CT1 Lg Cpgd Rg CT1
1
þ Y2
Cpgd Rg CT1
We need to derivative equation: dVA11

dt ¼ ; dIdtLd ¼ L1d ðVA8 VA11 Þ ; dVd
dt ¼0
dVA11 1 Vd RL
¼ ILd þ þ
RL 1 1
VA8 ½ þ VA11
d1
d 2 VA11 1 dILd RL
¼ þ RL
dt2 Cout ½RRd1L þ 1 dt ½Rd1 þ 1 Ld
dVA8 RL 1 1 dVA11
½ þ
dt Ld Cout Rd1 ½RRL þ 1 dt
d1
d 2 VA11 1 1 RL
¼ ðVA8 VA11 Þ þ RL
dt2 Cout ½RRd1L þ 1 Ld ½Rd1 þ 1 Ld
dVA8 RL 1 1 dVA11
½ þ RL
dt Ld Cout Rd1 ½R þ 1 dt
d1
d 2 VA11
dt ¼ Y4 ; dVA11
dt ¼ Y5 ; dt2 ¼ dYdt5
dY5 1 1 RL RL 1
¼ ðVA8 VA11 Þ þ RL Y4 ½ þ
dt Cout ½Rd1 þ 1 Ld
R L
½Rd1 þ 1 Ld Ld Cout Rd1
1
RL Y5
½Rd1 þ 1
dY5 1 1 1 1 RL
¼ VA8 VA11 þ RL Y4
R L
Cout ½Rd1 þ 1 Ld
R L
½Rd1 þ 1 Ld
RL 1 1
½ þ Y5
d1
We need to derivative equation: dt ðVA7

d
VA9 Þ ¼ ; dv
dt ¼ C1gs ðVA5RV
i
A9 Þ

d2 d VA4 VA7 VA6 VA7 ðVA7 VA8 Þ ðVA7 VA9 Þ gm

ðVA7 VA9 Þ ¼ f þ
dt2 dt Cds Rgd Cds Rdi Cds Rd Cds Rds Cds
vg
d2 1 dVA4 1 1 dVA7
ðVA7 VA9 Þ ¼
dt 2 Cds Rgd dt Cds Rgd dt dt
1 dVA6 1 dVA7
þ
Cds Rdi dt Cds Rdi dt
1 dVA7 1 dVA8 1 dVA7
þ
Cds Rd dt Cds Rd dt Cds Rds dt
1 dVA9 gm dv
þ
Cds Rds dt Cds dt
d2 1 dVA4 1 1 dVA7 1 dVA6 1 dVA7

ðVA7 VA9 Þ ¼ þ
dt2 Cds Rgd dt Cds Rgd dt dt Cds Rdi dt Cds Rdi dt
1 dVA7 1 dVA8 1 dVA7 1 dVA9 gm 1 ðVA5 VA9 Þ
þ þ
Cds Rd dt Cds Rd dt Cds Rds dt Cds Rds dt Cds Cgs Ri
d2 1 dVA9 dVA7 1 dVA4 1 dVA6

ðVA7 VA9 Þ ¼ ½ þ þ
dt2 Cds Rds dt dt Cds Rgd dt Cds Rdi dt
dVA7 1 1 1 1 dVA8 gm 1 VA5
½ þ þ þ
dt Cds Rgd Cds Rdi Cds Rd Cds Rd dt Cds Cgs Ri
gm 1 VA9
þ
Cds Cgs Ri

dt ¼ Y6 ; Y7 ¼ dVdtA9 dVdtA7 ; dVA7
dt ¼ Y3
dVA9 dVA9
Y7 ¼ Y3 ) ¼ Y7 þ Y 3
dt dt
dY7 1 1 1
¼ Y7 þ Y2 þ Y 6 Y3
dt Cds Rds Cds Rgd Cds Rdi
1 1 1
½ þ þ
Cds Rgd Cds Rdi Cds Rd
1 gm 1 VA5 gm 1 VA9
þ Y4 þ
Cds Rd Cds Cgs Ri Cds Cgs Ri
We can summery our system new differential equations representation:
dVA3 dVA4 dVA7 dVA8

¼ Y1 ¼ f1 ðY1 Þ; ¼ Y2 ¼ f3 ðY2 Þ; ¼ f4 ðY3 Þ ¼ Y3 ; ¼ f5 ðY4 Þ ¼ Y4
dt dt dt dt
dY1 1
¼ f2 ðY1 ; Y2 ; Y3 ; Y4 ; VA3 ; VA8 ; VA11 Þ ¼
dt Rd ½Cpds þ Cpgd CT1
1
Y3 Y4
Rd ½Cpds þ Cpgd CT1
1 1 1 1
VA8 þ VA11
½Cpds þ Cpgd CT1 Ld ½Cpds þ Cpgd CT1 Ld
1 1 1 1
Vg VA3
Cpgd CT1 Lg Cpgd CT1 Lg
1 1
Y1 þ Y2
dVA11 dY5 dVA6 dVA9

¼ f6 ðY5 Þ ¼ Y5 ; ¼ f7 ðVA8 ; VA11 ; Y4 ; Y5 Þ; ¼ f8 ðY6 Þ ¼ Y6 ;
dt dt dt dt
¼ f9 ðY7 ; Y3 Þ ¼ Y7 þ Y3
dY5 1 1 1 1 RL
R L
Cout ½Rd1 þ 1 Ld
R L
½Rd1 þ 1 Ld
RL 1 1
½ þ Y5
d1
dY7 1 1 1
¼ f10 ðVA5 ; VA9 ; Y2 ; Y6 ; Y7 ; Y3 ; Y4 Þ ¼ Y7 þ Y2 þ Y6
1 1 1 1
Y3 ½ þ þ þ
Cds Rgd Cds Rdi Cds Rd Cds Rd
gm 1 VA5 gm 1 VA9
Y4 þ
Cds Cgs Ri Cds Cgs Ri
We have ten differential equations which represent our system.
dVA3 dY1 dVA4 dVA7

¼ f1 ðY1 Þ ; ¼ f2 ðY1 ; Y2 ; Y3 ; Y4 ; VA3 ; VA8 ; VA11 Þ ; ¼ f3 ðY2 Þ; ¼ f4 ðY3 Þ
dt dt dt dt
dVA8 dVA11 dY5 dVA6
¼ f5 ðY4 Þ ; ¼ f6 ðY5 Þ; ¼ f7 ðVA8 ; VA11 ; Y4 ; Y5 Þ ; ¼ f8 ðY6 Þ
dt dt dt dt
dVA9 dY7
¼ f9 ðY7 ; Y3 Þ ; ¼ f10 ðVA5 ; VA9 ; Y2 ; Y6 ; Y7 ; Y3 ; Y4 Þ
dt dt
@ Fixed points:
dVA3 dY1 dVA4 dVA7 dVA8
¼ 0; ¼ 0; ¼ 0; ¼ 0; ¼0
dt dt dt dt dt
dVA11 dY5 dVA6 dVA9 dY7
¼ 0; ¼ 0; ¼ 0; ¼ 0; ¼0
dt dt dt dt dt
Y1 ¼ 0 ; Y2 ¼ 0 ; Y3 ¼ 0 ; Y4 ¼ 0 ; Y5 ¼ 0 ; Y6 ¼ 0 ; Y7 ¼ 0
1 1 1 1
V þ
½Cpds þ Cpgd CT1 Ld A8 ½Cpds þ Cpgd CT1 Ld
1 1 1 1
VA11 Vg V ¼ 0
Cpgd CT1 Lg Cpgd CT1 Lg A3
1 1 1 1
VA8 VA11 ¼0
Cout ½R þ 1 Ld
RL
d1
Cout ½R þ 1 Ld
R L
d1

gm 1 VA5 gm 1 VA9
þ ¼0
equilibrium points of our system consists in adding to coordinates
½VA3 ; VA4 ; VA7 ; VA8 ; Y1 ; VA11 ; Y5 ; VA6 ; VA9 ; . . . arbitrarily small increments of expo-
nential form ½vA3 ; vA4 ; vA7 ; vA8 ; y1 ; vA11 ; y5 ; vA6 ; vA9 ; . . . ekt , and retaining the first
order terms in VA3 ; VA4 ; VA7 ; VA8 ; Y1 ; VA11 ; Y5 ; VA6 ; VA9 ; . . .. The system of homo-
geneous equations leads to a polynomial characteristics equation in the eigenvalues
λ . Our system fixed values with arbitrarily small increments of exponential form
½vA3 ; vA4 ; vA7 ; vA8 ; y1 ; vA11 ; y5 ; vA6 ; vA9 ; . . . ekt are: j = 0 (first fixed point), j = 1
ðjÞ
(second fixed point), j = 2 (third fixed point), etc. VA3 ðtÞ ¼ VA3 þ vA3 ekt ;
ðjÞ
VA4 ðtÞ ¼ VA4 þ vA4 ekt
ðjÞ ðjÞ
VA7 ðtÞ ¼ VA7 þ vA7 ekt ; VA8 ðtÞ ¼ VA8 þ vA8 ekt ;
ðjÞ ðjÞ
Y1 ðtÞ ¼ Y1 þ y1 ekt ; VA11 ðtÞ ¼ VA11 þ vA11 ekt
ðjÞ ðjÞ ðjÞ
Y5 ðtÞ ¼ Y5 þ y5 ekt ; VA6 ðtÞ ¼ VA6 þ vA6 ekt ; VA9 ðtÞ ¼ VA9 þ vA9 ekt . . .:
We choose the above expressions for our VA3 ðtÞ; VA4 ðtÞ; VA7 ðtÞ; VA8 ðtÞ;
Y1 ðtÞ; VA11 ðtÞ; Y5 ðtÞ; VA6 ðtÞ; VA9 ðtÞ; . . . as small displacement ½vA3 ; vA4 ; vA7 ; vA8 ; y1 ;
vA11 ; y5 ; vA6 ; vA9 ; . . .
ðjÞ
from our system fixed points at time t = 0. VA3 ðt ¼ 0Þ ¼ VA3 þ vA3 ;
ðjÞ
VA4 ðt ¼ 0Þ ¼ VA4 þ vA4
ðjÞ ðjÞ
VA7 ðt ¼ 0Þ ¼ VA7 þ vA7 ; VA8 ðt ¼ 0Þ ¼ VA8 þ vA8 ;
ðjÞ ðjÞ
Y1 ðt ¼ 0Þ ¼ Y1 þ y1 ; VA11 ðt ¼ 0Þ ¼ VA11 þ vA11
ðjÞ ðjÞ ðjÞ
Y5 ðt ¼ 0Þ ¼ Y5 þ y5 ; VA6 ðt ¼ 0Þ ¼ VA6 þ vA6 ; VA9 ðt ¼ 0Þ ¼ VA9 þ vA9 . . .. . .
For λ < 0, t > 0 the selected fixed point is stable otherwise λ > 0, t > 0 is
unstable. Our system tends to the selected fixed point exponentially for λ < 0, t > 0
otherwise go away from the selected fixed point exponentially. λ is the eigenvalue
his absolute value |λ| establishes the speed of flow toward or away from the selected
fixed point (Yuri 1995; Jack and Huseyin 1991) [2–4]. The speeds of flow toward or
away from the selected fixed point for system variables derivatives with respect to
time are:
dVA3 ðtÞ dVA4 ðtÞ

¼ vA3 k ekt ; ¼ vA4 k ekt
dt dt
dVA7 ðtÞ dVA8 ðtÞ
¼ vA7 k ekt ; ¼ vA8 k ekt ;
dt dt
dY1 ðtÞ dVA11 ðtÞ
¼ y1 k ekt ; ¼ vA11 k ekt
dt dt
dY5 ðtÞ dVA6 ðtÞ dVA9 ðtÞ
¼ y5 k ekt ; ¼ vA6 k ekt ; ¼ vA9 k ekt . . .
dt dt dt
dVA3 ðjÞ ðjÞ
¼ Y1 ) vA3 k ekt ¼ Y1 þ y1 ekt ; @ fixed point Y1 ¼ 0
dt
ðjÞ dVA4 ðjÞ
Y1 ¼ 0 ; y1 vA3 k ¼ 0; ¼ Y2 ) vA4 k ekt ¼ Y2 þ y2 ekt
dt
ðjÞ
@ fixed point Y2 ¼ 0; y2 vA4 k ¼ 0
dVA7 ðjÞ ðjÞ
¼ Y3 ) vA7 k ekt ¼ Y3 þ y3 ekt @ fixed point Y3 ¼ 0
dt
dVA8 ðjÞ
y3 vA7 k ¼ 0 ; ¼ Y4 ) vA8 k ekt ¼ Y4 þ y4 ekt
dt
ðjÞ
@ fixed point Y4 ¼ 0 ) y4 vA8 k ¼ 0
dY1 1 1
¼ Y3 Y4
1 1 1 1
VA8 þ
½Cpds þ Cpgd CT1 Ld ½Cpds þ Cpgd CT1 Ld
1 1 1 1
VA11 Vg VA3
Cpgd CT1 Lg Cpgd CT1 Lg
1 1
Y1 þ Y2
1 ðjÞ
y1 k ekt ¼ ½Y3 þ y3 ekt
1 ðjÞ
½Y4 þ y4 ekt
1 1 ðjÞ
½VA8 þ vA8 ekt
1 1 ðjÞ 1 1
þ ½VA11 þ vA11 ekt Vg
1 1 ðjÞ 1 ðjÞ
½VA3 þ vA3 ekt ½Y1 þ y1 ekt
Cpgd CT1 Lg Cpgd Rg CT1
1 ðjÞ
þ ½Y2 þ y2 ekt
Cpgd Rg CT1
1 ðjÞ 1 ðjÞ
y1 k ekt ¼ f Y Y
Rd ½Cpds þ Cpgd CT1 3 Rd ½Cpds þ Cpgd CT1 4
1 1 ðjÞ 1 1 ðjÞ 1 1
V þ V Vg
½Cpds þ Cpgd CT1 Ld A8 ½Cpds þ Cpgd CT1 Ld A11 Cpgd CT1 Lg
1 1 ðjÞ 1 ðjÞ 1 ðjÞ
V Y þ Y g
Cpgd CT1 Lg A3 Cpgd Rg CT1 1 Cpgd Rg CT1 2
1 1
y3 ekt y4 ekt
Rd ½Cpds þ Cpgd CT1 Rd ½Cpds þ Cpgd CT1
1 1
vA8 ekt
1 1 1 1
vA11 ekt vA3 ekt
1 1
y1 ekt þ y2 ekt
@ fixed point
1 ðjÞ 1 ðjÞ
f Y Y
Rd ½Cpds þ Cpgd CT1 3 Rd ½Cpds þ Cpgd CT1 4
1 1 ðjÞ 1 1 ðjÞ 1 1
VA8 þ VA11 Vg
½Cpds þ Cpgd CT1 Ld ½Cpds þ Cpgd CT1 Ld Cpgd CT1 Lg
1 1 ðjÞ 1 ðjÞ 1 ðjÞ
V Y þ Y g¼0
Cpgd CT1 Lg A3 Cpgd Rg CT1 1 Cpgd Rg CT1 2
1 1
y1 k ekt ¼ y3 ekt
1 1
y4 ekt vA8 ekt
1 1 1 1
þ vA11 ekt vA3 ekt
1 1
y1 ekt þ y2 ekt
1 1
y1 k ¼ y3
1 1
y4 vA8
1 1 1 1
þ vA11
1 1
vA3 y1 þ y2
1 1
y3
1 1
y4 vA8
1 1 1 1
þ vA11
1 1
vA3 y1 þ y2 y 1 k ¼ 0
ðjÞ ðjÞ
dVA11
dt ¼ Y5 ) vA11 k ekt ¼ Y5 þ y5 ekt ; @ fixed point Y5 ¼ 0 ) vA11 k ekt ¼
y5 ekt
dVA6
vA11 k ekt ¼ y5 ekt ) y5 vA11 k ¼ 0; ¼ Y6 ) vA6 k ekt
dt
ðjÞ
¼ Y6 þ y6 ekt
ðjÞ
@ fixed point Y6 ¼ 0 ) vA6 k ekt ¼ y6 ekt ) y6 vA6 k ¼ 0
ðjÞ ðjÞ
dt ¼ Y7 þ Y3
dVA9
) vA9 k ekt ¼ Y7 þ y7 ekt þ Y3 þ y3 ekt @ fixed point
ðjÞ ðjÞ
Y7 þ Y3 ¼ 0
vA9 k ekt ¼ y7 ekt þ y3 ekt ) vA9 k ¼ y7 þ y3 ) y7 þ y3 vA9 k ¼ 0

dY5 1 1 1 1 RL
dt Cout ½RRd1L þ 1 Ld Cout ½RRd1L þ 1 Ld ½Rd1 þ 1 Ld
RL 1 1
½ þ RL Y5
d1
y5 k ekt
RL ðjÞ RL 1 1 ðjÞ
þ ½Y4 þ y4 ekt ½ þ RL ½Y5 þ y5 ekt
½RRd1L þ 1 Ld Ld Cout Rd1 ½R þ 1
d1
1 1 ðjÞ 1 1 ðjÞ
y5 k ekt ¼ f V V
Cout ½RRd1L þ 1 Ld A8 Cout ½RRd1L þ 1 Ld A11
þ Y ½ þ Y g
½RRd1L þ 1 Ld 4 Ld Cout Rd1 ½RRL þ 1 5
d1
1 1 1 1
þ vA8 ekt
Cout ½Rd1 þ 1 Ld
RL
Cout ½Rd1 þ 1 Ld
R L
RL
vA11 ekt þ RL y4 ekt
½Rd1 þ 1 Ld
RL 1 1
½ þ y5 ekt
d1
1 1 ðjÞ 1 1 ðjÞ
f VA8 VA11
Cout ½Rd1 þ 1 Ld
RL
Cout ½Rd1 þ 1 Ld
R L
@ fixed point
þ Y ½ þ Y g¼0
½RRd1L þ 1 Ld 4 Ld Cout Rd1 ½RRL þ 1 5
d1
1 1 1 1 RL
y5 k ¼ vA8 vA11 þ RL y4
Cout ½Rd1 þ 1 d
RL L Cout ½Rd1 þ 1 d
RL L ½Rd1 þ 1 Ld
RL 1 1
½ þ RL y5
d1
1 1 1 1 RL
vA8 vA11 þ RL y4
Cout ½RRd1L þ 1 Ld Cout ½RRd1L þ 1 Ld ½Rd1 þ 1 Ld
RL 1 1
½ þ y5 y5 k ¼ 0
d1
dY7 1 1 1
¼ Y7 þ Y2 þ Y6
1 1 1 1
Y3 ½ þ þ þ
Cds Rgd Cds Rdi Cds Rd Cds Rd
gm 1 VA5 gm 1 VA9
Y4 þ
1 ðjÞ 1 ðjÞ
y7 k ekt ¼ ½Y7 þ y7 ekt þ ½Y2 þ y2 ekt
Cds Rds Cds Rgd
1 ðjÞ
þ ½Y6 þ y6 ekt
Cds Rdi
ðjÞ 1 1 1
½Y3 þ y3 ekt ½ þ þ
1 ðjÞ
þ ½Y4 þ y4 ekt
Cds Rd
ðjÞ ðjÞ
gm 1 ½VA5 þ vA5 ekt gm 1 ½VA9 þ vA9 ekt
þ
1 ðjÞ 1 ðjÞ 1 ðjÞ

y7 k ekt ¼ f Y7 þ Y2 þ Y
Cds Rds Cds Rgd Cds Rdi 6
ðjÞ 1 1 1 1 ðjÞ
Y3 ½ þ þ þ Y
Cds Rgd Cds Rdi Cds Rd Cds Rd 4
ðjÞ ðjÞ
gm 1 VA5 gm 1 VA9 1
þ gþ y7 ekt
Cds Cgs Ri Cds Cgs Ri Cds Rds
1 1
þ y2 ekt þ y6 ekt
Cds Rgd Cds Rdi
1 1 1
½ þ þ y3 ekt
1 gm 1 gm 1
þ y4 ekt vA5 ekt þ vA9 ekt
@ fixed point
1 ðjÞ 1 ðjÞ 1 ðjÞ ðjÞ

f Y þ Y þ Y Y3
Cds Rds 7 Cds Rgd 2 Cds Rdi 6
1 1 1
½ þ þ
ðjÞ ðjÞ
1 ðjÞ gm 1 VA5 gm 1 VA9
þ Y4 þ g¼0
1 1 1
y7 k ¼ y7 þ y2 þ y6
Cds Rds Cds Rgd Cds Rdi
1 1 1
½ þ þ y3
1 gm 1 gm 1
þ y4 vA5 þ vA9
1 1 1 1 1 1
y7 þ y2 þ y6 ½ þ þ y3
Cds Rds Cds Rgd Cds Rdi Cds Rgd Cds Rdi Cds Rd
1 gm 1 gm 1
þ y4 vA5 þ vA9 y7 k ¼ 0
Summary of our results, we get arbitrarily small increments

(vA3 ; vA4 ; vA7 ; vA8 ; y1 ; vA11 ; y5 ; vA6 ; vA9 ; . . .) ten equations:
y1 vA3 k ¼ 0 ; y2 vA4 k ¼ 0 ; y3 vA7 k ¼ 0 ; y4 vA8 k ¼ 0
1 1 1 1
y3 y4 vA8
Rd ½Cpds þ Cpgd CT1 Rd ½Cpds þ Cpgd CT1 ½Cpds þ Cpgd CT1 Ld
1 1 1 1
þ vA11 vA3
1 1
y1 þ y2 y1 k ¼ 0
y5 vA11 k ¼ 0 ; y6 vA6 k ¼ 0 ; y7 þ y3 vA9 k ¼ 0
1 1 1 1 RL
vA8 vA11 þ RL y4
Cout ½RRd1L þ 1 Ld Cout ½RRd1L þ 1 Ld ½Rd1 þ 1 Ld
RL 1 1
½ þ RL y5 y5 k ¼ 0
d1
1 1 1 1 1 1
y7 þ y2 þ y6 ½ þ þ y3
Cds Rds Cds Rgd Cds Rdi Cds Rgd Cds Rdi Cds Rd
1 gm 1 gm 1
þ y4 vA5 þ vA9 y7 k ¼ 0
0 1
vA 3
B vA 4 C
B C
B vA 7 C 0 1
0 1 Bv C
B
C 0 1 y2
i1 1 ... i1 10 B A8 C v11 . . . v15 B y3 C
B .. .. .. C B y1 C
B
CþB .. .. .. C B

C
B y4 C ¼ 0; i1
@ . . . A B vA11 C @ . . . A B C 1
B C @ y6 A
i10 1 i10 10 B vA C v10 1 v10 5
B 6 C vA 5
B vA C
B 9 C
@ y5 A
y7
¼ k; i1 2 ¼ i1 3 ¼ i1 4 ¼ 0
i1 6 ¼ ¼ i1 ¼ 0 ; i2 1 ¼ 0 ; i2 2 ¼ k ;
10
i2 3 ¼ . . . ¼ i2 10 ; i3 1 ¼ i3 2 ¼ 0 ; i3 3 ¼ k
i3 4 ¼ . . . ¼ i3 10 ¼ 0 ; i4 1 ¼ i4 2 ¼ i4 3 ¼ 0 ; i4 4 ¼ k ;
1 1
i4 5 ¼ . . . ¼ i4 10 ¼ 0 ; i5 1 ¼
Cpgd CT1 Lg
1 1
i5 2 ¼ 0 ; i5 3 ¼ 0 ; i5 4 ¼ ;
1
i5 5 ¼ k
Cpgd Rg CT1
1 1
i5 6 ¼ ; i5 7 ¼ i5 8 ¼ i5 9 ¼ i5 10 ¼ 0
i6 1 ¼ . . . ¼ i6 5 ¼ 0 ; i6 6 ¼ k
i6 7 ¼ 0 ; i6 8 ¼ 0 ; i6 9 ¼ 1 ; i6 10 ¼ 0 ; i7 1 ¼ . . . ¼ i7 6 ¼ 0;
i7 7 ¼ k ; i7 8 ¼ i7 9 ¼ i7 10 ¼ 0
i8 1 ¼ . . . ¼ i8 7 ¼ 0 ; i8 8 ¼ k ; i8 9 ¼ 0 ; i8 10 ¼ 1 ;
1 1
i9 1 ¼ i9 2 ¼ i9 3 ¼ 0 i9 4 ¼
Cout ½Rd1 þ 1 Ld
RL
1 1
i9 5 ¼ 0 ; i9 6 ¼ ; i9 7 ¼ i9 8 ¼0
Cout ½Rd1 þ 1 Ld
RL
RL 1 1
i9 9 ¼ k ½ þ
d1
i9 10 ¼ 0 ; i10 1 ¼ i10
¼ i10 3 ¼ i10 4 ¼ i10 5 ¼ i10 6
2
gm 1
¼ i10 7 ¼ 0; i10 8 ¼
Cds Cgs Ri
1
i10 9 ¼ 0 ; i10 10 ¼ k þ ; v11 ¼ . . . ¼ v15 ¼ 0 ;
Cds Rds
v21 ¼ 1 ; v22 ¼ . . . ¼ v25 ¼ 0
v31 ¼ 0 ; v32 ¼ 1 ; v33 ¼ v34 ¼ v35 ¼ 0 ;
v41 ¼ v42 ¼ 0 ; v43 ¼ 1 ; v44 ¼ 0 ; v45 ¼ 0
1 1
v51 ¼ ; v52 ¼ ;
Cpgd Rg CT1 Rd ½Cpds þ Cpgd CT1
1
v53 ¼
v54 ¼ v55 ¼ 0 ; v61 ¼ . . . ¼ v65 ¼ 0 ; v71 ¼ v72 ¼ v73 ¼ 0;
v74 ¼ 1 ; v75 ¼ 0 ; v81 ¼ 0 ; v82 ¼ 1
v83 ¼ v84 ¼ v85 ¼ 0 ; v91 ¼ v92 ¼ 0;
RL
v93 ¼ RL ; v94 ¼ v95 ¼ 0
½Rd1 þ 1 Ld
1 1 1 1
v10 1 ¼ ; v10 2 ¼ ½ þ þ
Cds Rgd Cds Rgd Cds Rdi Cds Rd
1 1 gm 1
v10 3 ¼ v10 4 ¼ ; v10 5 ¼
Cds Rd Cds Rdi Cds Cgs Ri
Assumption:
0 1 0 1
v11 . . . v15 i1 1 . . . i1 10
B .. .. .. C B .. .. .. C
@ . . . A ! e; detðA k IÞ ¼ 0 ; A k I ¼ @ . . . A
v10 1 v10 5 i10 1 i10 10
0 1
vA3
B C
B vA4 C
B C
B vA C
B 7C
B C
0 1 B vA8 C 0 1
i1 ... i1 10 B C i1 ... i1
1 B y C 1 10
B .. .. .. C B 1 C 0 ; detðA k IÞ ¼ 0 ) detB .. .. .. C
@ . . . A B C @ . . . A 0
B vA11 C
i10 i10 10 B C i10 i10
1 B v C 1 10
B A6 C
B C
B vA9 C
B C
B y C
@ 5A
y7
To effectively apply the stability criterion of Lipunov to our system, we require a

criterion for when the equation detðA k IÞ ¼ 0 has a zero in the left half plane,
without calculating the eigenvalues explicitly. We use criterion of Routh-Hurwitz
[2–4].
3.4 IMPATT Amplifier Stability Analysis
A wide variety of solid state diodes and transistor have been developed for
microwave use. IMPact ionization Avalanche Transit-Time (IMPATT) diode
functions as microwave oscillator. It used to produce carrier signal for microwave
transmission system. IMPATT can operate from a few GHz to a few hundred GHz.
The diode is operated in reverse bias near breakdown, and both the N and N-
regions are completely depleted. Because of the difference in doping between the
“drift region” and “avalanche region”, the electric field is highly peaked in the
avalanche region and nearly flat in drift region. In operation, avalanche breakdown
occurs at the point of highest electric field, and this generates a large number of
hole-electron pairs by impact ionization. The holes are swept into the cathode, but
the electrons travel across the drift region toward anode. As they drift, they induce
image charges on the anode, giving rise to a displacement current in the external
circuit that is 180° out of phase with the nearly sinusoidal voltage waveform. It’s
buildup of microwave oscillations in the diode current and voltage when the diode
is embedded in a resonant cavity and biased at breakdown. The IMPATT diode has
a negative resistance from DC through microwave frequencies. Consequently, it is
prone to oscillate at low frequencies, with the lead inductance from bias circuit
connections. The voltage due to bias circuit oscillations may be large enough to
burn the device out if adequate precautions are not observed. It is prudent practice
to suppress the bias circuit oscillation. Adequate heat sink must be provided for the
diode to operate properly. These IMPATT diodes have been designed to operate in
the pre-collection mode. As the diode is tuned up from a low operating current from
a constant current source, it will be noticed that at the onset of pre-collection mode,
the diode voltage falls down. The power output will increase by several dBs with a
slight shift in the operating frequency. When the circuit is detuned in such a fashion
that the diode falls out of the pre-collection mode, the diode voltage will increase.
The power dissipation will increase as the power output falls down. If the diode is
not adequately heat sink, the diode may burn out. A main advantage is their high
power capability. These diodes are used in a variety of applications from low power
radar systems to alarms. IMPATT oscillator is for higher-power output, higher
efficiency, and higher frequency range of operation. The effect of negative
resistance of IMPATT diode in the amplification of microwave signals. Nonlinear

effects are dominant considerations in power amplifier design because of efficiency
and economy consideration of the device. IMPATT amplifier at lower frequency
band is a very interesting area. Ka-band reflection type IMPATT amplifier has been
developed using a Ka-band IMPATT diode oscillator as an input signal source.
These amplifiers have small size, simple arrangement, and sufficient power addition
for various applications in the field of high frequency communication and radar.
IMPATT amplifiers are used mostly at the high end of the microwave band because
microwave transistors do not work well above 30 GHz due to transit time limita-
tions. IMPATT uses transit time effects to generate microwaves. IMPATT amplifier
consist circulator (three ports circulator) which is connected to an IMPATT diode,
mounted in a resonant circuit or cavity. A circulator must be used to separate the
input and output powers. Since IMPATT is a single port device, circulator must be
used to separate the input and the output power RF signal. The microwave power to
be amplified enters one port of the circulator and is routed into the IMPATT diode
in cavity. The incoming microwave RF signal is amplified and leaves the cavity
from the same port that it entered. It is then routed by the circulator into the output
transmission line [67–74].
IMPATT diode construction: IMPATT diode consists of a PN junction between
the P+ and the N regions, a drift region of intrinsic (I) material, and an N+ con-
nection. IMPATT diode has negative resistance characteristics. The microwave
negative resistance of an IMPATT diode arises out of a phase difference between
the RF voltage and RF current. This phase difference is produced by the lagging RF
current generated in the space charge layer with respect to the applied RF voltage.
Fig. 3.9 IMPATT amplifier circuit with R1 load

3.4 IMPATT Amplifier Stability Analysis 335
An IMPATT diode can function as an amplifier if the load resistance presented to it

is larger in magnitude than the diode’s negative resistance.
The intrinsic region thickness is controlled and the electron transit time through
it is a half a microwave cycle at the selected operational frequency.
Basically IMPATT diode is a form of high power diode used in high frequency
electronics and microwave devices. They operate at frequencies between 3 GHz to
100GHz or more. The main advantage of IMPATT diode is their high power
capability. These diodes are used in a variety of applications from low power radar
systems to alarms. These diodes make excellent microwave generators for many
applications. An IMPATT diode is mounted in a microwave package. The diode is
mounted with its high field region close to a copper heat sink so that the heat
generated at the diode junction can be readily dissipated.
An IMPATT diode can function as an amplifier if the load resistance presented
to it is larger in magnitude than the diode’s negative resistance. A Bias T is required
for operation of the amplifier cavity. The IMPATT diode requires a DC bias cur-
rent. It is necessary to block the DC bias current from the rest of the circuitry. This
has been accomplished using a microstrip inter-digital DC blocking capacitor cir-
cuit. The DC bias current is applied to the diode through a very high RF impedance
(quarter-wavelength long, very narrow piece of copper) terminated in a low RF
impedance (wide piece of copper). The next figure shows IMPATT reflection
amplifier. A circulator is used to separate input and output signals. The maximum
possible DC bias current above which the diode breaks into oscillations can be
readily observed also. An IMPATT diode exhibited the same characteristics of
tending toward saturation with increasing input power levels and a corresponding
increase in bandwidth.
IMPATT diode negative resistance typically varies as a function of the diode RF
current amplitude. RD is the terminal (negative) resistance of the packaged diode
and RL is the diode’s load resistance. IRDI decreases with signal level. RD also
varies with DC bias current and thus the upper limit of bias current is established at
the value that causes IRDI to equal RL. Exceeding this maximum value of bias
current will cause the diode to act as an oscillator instead of an amplifier because
the diodes load resistance, RL, is no longer greater than the magnitude of the
diode’s negative resistance, IRDI. An IMPATT amplifier requires that RL be larger
than IRDI for all values of the RF current through the diode. Since RD varies not
only with different types of diodes, but also with DC bias current and signal level,
the selection of RL for optimum power gain is of prime importance in the design of
the amplifier.
If RL [ jRD j then IMPATT diode acts as an amplifier. If RL \ jRD j then
IMPATT diode acts as an oscillator (RL = Rload).
three port
circulator
Delay line (Tau1) Delay line (Tau3)
ARB1 ARB2
1 3
N1 OUT N1 OUT
2
A RB3 R2
N1
V2 Delay line (Tau2)
Microwave RF Rload
source X(t)
OUT
C1
L1
D1
IMPATT
R1 diode
V1
V1 – DC voltage source (bias voltage to IMPATT diode).

R1 – parasitic resistance of DC voltage source.
V2 – Microwave RF source X(t).
L1, C1 – inductor and capacitor of Bias-T circuit.
ARB1, ARB2, ARB3 – circuit micro strip delay lines.
R2 – circuit load resistance.
D1 – IMPATT diode.
Fig. 3.10 IMPATT amplifier diode negative resistance circuit
IMPATT diode is current controlled negative resistance (CCNR, open circuit

stable, or “S” type). In this type, the voltage is a single valued function of current,
but the current is a multivalued function of voltage. The graph is a curve shaped like
the letter “S”. Negative differential resistance devices such as IMPATT diode are
used to make amplifiers, particularly at microwave frequencies, but not as com-
monly as oscillators. IMPATT diode (negative resistance device) has only one port
(two terminals), unlike two ports devices such as transistors, the outgoing amplified
signal has to leave the IMPATT diode by the same terminals as the incoming signal
enter it. If we do not use the circulator the IMPATT diode negative resistance
amplifier is bilateral. It amplifiers in both directions then there is high sensitivity to
load impedance and feedback problems. To separate the input and output signals,
IMPATT diode negative resistance amplifier use nonreciprocal device such as
isolator and directional couplers. In our case we use reflection amplifier in which
the separation is accomplished by an active circulator. The IMPATT diode chip RF
equivalent circuit includes the active part of the diode (the chip, excluding the
Fig. 3.11 IMPATT diode

chip RF equivalent circuit
package) as a negative resistance RD, and a reactance XD. Included in RD is the

unavoidable parasitic series resistance, RS, contributed by contacts and the un
depleted portion of the N region.
RD ¼ Rc ; Rc [ 0 ; RD \0 ; jRD j ¼ Rc ; XD ¼ jxC 1
D
. If we connect IMPATT
diode directly to bias voltage Vb which locate our IMPATT diode working point
DC characteristic in the negative resistance region. Vb is IMPATT diode bias
voltage source. Rb is voltage source series resistance. S1 is a bypass RF microwave
source switch. X(t) is RF microwave signal source.
dVCd 1
I ¼ ICd þ IRd ¼ IRb ; ICd ¼ Cd ; VCd ¼
dt Cd
Z
ICd dt ; Vb ¼ I Rb þ VCd þ VRd
Z Z
1 d 1
Vb ¼ I Rb þ ICd dt þ VRd ; fVb ¼ I Rb þ ICd dt þ VRd g
Cd dt Cd
dI 1 dVRd dI 1
0 ¼ Rb þ I Cd þ ) Rb þ
dt Cd dt dt Cd
dI dVRd dVRd
I Cd þ ¼ 0; ¼ Rd ¼ Rc
dt dI dI
Fig. 3.12 IMPATT diode and NDR element equivalent circuit

dI 1 dI dI dI 1
Rb þ IC d R c ¼ 0 ) R c R b ¼ IC d ; R c R b [ 0
dt Cd dt dt dt Cd
dI 1 dI=dt 1 d 1
ðRc Rb Þ ¼ I ) ¼ ) fln IðtÞg ¼
dt Cd I Cd ðRc Rb Þ dt Cd ðRc Rb Þ
Z
d 1 1 1
t
f fln IðtÞg ¼ g ) ln IðtÞ ¼ t ) IðtÞ ¼ eCd ðRc Rb Þ
dt Cd ðRc Rb Þ Cd ðRc Rb Þ
1
is the exponential coefficient.
Cd ðRc Rb Þ
We use active circulator in our circuit. Active circulators are ideally suited for
realization using monolithic microwave integrated circuit (MMIC) technology. The
circuit employs decade bandwidth active circulator which shows very low phase
error characteristic. The circuit configuration of the active circulator used three
metal–semiconductor field effect transistors (MESFETs) which are the GEC-Marconi
standard library cell F20-FET-4 × 75. With all the standard library cells, it is a very
accurate ultra-wideband small signal model for the device. It is similar to a junction
gate field-effect transistor (JFET) in construction and terminology. The difference is
that instead of using a p-n junction for gate, a schottky (metal semiconductor)
junction is used. A typical three ports decade bandwidth active circulator has three
MESFETs transistors interconnected with each other. RF, CF, LF, CC, Rsb play a
major role in the working of the circuit. The three feedback branches (RF, CF, LF)
are used to link all the three transistors in an end to end fashion. The source resistor
(Rsb) is shared among all the three MESFETs transistors and one transistor is source
coupled with the other two transistors using this source resistor. The circuit works in
a symmetric fashion. We consider MESFET high frequency model taking node
capacitors into account. Next figure describes the circuit configuration of the active
circulator [36, 37]. We use N-type MESFET but usually the recommended is a
symmetrical bilateral MESFET. All Cc and Cf capacitors are un-polarized. Once we
inject RF signal to port P1, it passes to port P2 through feedback branch (RF, CF, LF).
The same is between ports P2 and P3, ports P3 and P1. In case we inject RF signal to
port P2, it reaches Q1 gate and shorten Q1’s drain and source. Then Port 2’s RF
signal is shortened to ground through resistor Rsb and did not reach port P1. The
same is between P1 to P3 and P3 to P2. We consider a varactor which is realized by
connecting together the drain and source terminations of a standard MESFET,
resulting in a Schottky junction. The bias potential is then applied across the
drain/source and gate terminations. Our three ports decade bandwidth active circu-
lator with micro strip delay lines and IMPATT diode circuit in port P2 gets his input
RF signal from microwave RF source (port P1) and feeds antenna unit by active
circulator output RF signal (Port P3) [36, 37].
Fig. 3.13 Active circulator circuit system (version 1)
Active circulator system can be described by system path from RFin port (P1) to
RFout port (P3). For simplicity we ignore MESFET high frequency equivalent
model and took it as cutoff element in our system. Next figure describes our
IMPATT amplifier system path from microwave RF source X(t) to antenna unit
(Rload). We ignore the signal path from P3 to P1 since our amplifier RF signal is feed
directly to load antenna. Rload is a pure resistive but can be taken with additional
reactance part. We consider fully matching between antenna load resistance and
three power active circulator IMPATT diode circuit, no signal reflections. ARB1,
ARB2, and ARB3 are circuit micro strip delay line, VARB1 ðtÞ ! e ;
VARB2 ðtÞ ! e ; VARB3 ðtÞ ! e. Due to active circulator’s micro strip transmission
lines delays, s1 for the first port current, s2 for the second port current, and s3 for the
third port current. V1 is IMPATT diode bias voltage.
dVRD dVRD
I1 ðtÞ ! I1 ðt s1 Þ ; I2 ðtÞ ! I2 ðt s2 Þ ; I3 ðtÞ ! I3 ðt s3 Þ ; \0 ;
dIRD dIRD
dVRD
¼ RD ; j j ¼ Rc
dIRD
Fig. 3.14 Active circulator circuit system (version 2)
First branch:
dVcf dI1 dVCC

I1 ¼ Cf ; VLf ¼ Lf ; I1 ¼ C C ;
dt dt dt
d
Ceq ½Vc þ VCC ¼ I1 ; Ceq ¼ Cf jjCC
dt f
1 1 1 d 1 1
¼ þ ; ½Vc þ VCC ¼ I1 ð þ Þ;
Ceq Cf CC dt f Cf CC
VP1 VP2 ¼ Vcf þ VLf þ VRf þ VCC
dI1
VCf þ VCC ¼ VP1 VP2 VRf VLf ; VCf þ VCC ¼ VP1 VP2 I1 Rf Lf
dt
d dVP1 dVP2 dI1 d 2 I1
ðVCf þ VCC Þ ¼ R f Lf ; VRf ¼ I1 Rf ; Rd ¼ RD
dt dt dt dt dt2
1 1 dVP1 dVP2 dI1 d 2 I1
I1 ð þ Þ¼ Rf Lf 2
Cf CC dt dt dt dt
Second branch:
dVC1 dIL dVCd

I2 ¼ C 1 ; VL1 ¼ L1 1 ; ICd ¼ Cd ;
dt dt dt
dVRd
¼Rd ; I2 þ IL1 ¼ ICd ; IR1 ¼ IL1
dIRd
ICd ¼ IRD ; V1 VA ¼ VR1 þ VL1 ; VA ¼ VCd þ VRd ;
dVC1 dVCd
VP2 VA ¼ VC1 ; C1 þ IL 1 ¼ C d
dt dt
Z
dVC1 1 dVCd d 2 VC1
C1 þ VL1 dt ¼ Cd ) C1
dt L1 dt dt2
2
1 d VCd
þ VL1 ¼ Cd ; VR1 ¼ IL1 R1
L1 dt2
dVRd dVCd ICd IRd
V1 VCd VRd ¼ VR1 þ VL1 ; ¼Rd ; ¼ ¼
dIRd dt Cd Cd
dVC1 dVP2 dVCd
VC1 ¼ VP2 VA ¼ VP2 VCd VRd ) ¼
dt dt dt
dVRd dVC1 dVP2 IRd dVRd
; ¼
dt dt dt Cd dt
dVC1 dVP2 IRd dVRd
I2 ¼ C1 ¼ C1 ð Þ
dt dt Cd dt
dVP2 IRd dIRd dVRd dVRd
¼ C1 ð Þ; ¼ Rc
dt Cd dt dIRd dIRd
dVP2 IRd dIRd
I2 ¼ C1 ð þ Rc Þ ; I2 þ IL1 ¼ ICd ¼ IRd ) I2
dt Cd dt
dVP2 I2 þ IL1 dðI2 þ IL1 Þ
¼ C1 ð þ Rc Þ
dt Cd dt
Third branch:
dVCf dI3
I3 ¼ Cf ; VLf ¼ Lf ; VRf ¼ I3 Rf ;
dt dt
dVCC dVCf I3
I3 ¼ CC ; IRload ¼ I3 ðt s3 Þ ; ¼
dt dt Cf
dVCC I3 dVCC dVCf I3 I3 d 1 1
¼ ; þ ¼ þ ) ðVCC þ VCf Þ ¼ I3 ð þ Þ
dt CC dt dt Cf CC dt Cf CC
VP2 VP3 ¼ VCf þ VLf þ VRf þ VCC ; VP3 ¼ VRload ¼ Rload IRload ¼ Rload I3 ðt s3 Þ
dI3
VP2 VP3 VLf VRf ¼ VCf þ VCC ; VCf þ VCC ¼ VP2 VP3 Lf I3 Rf
dt
2
d dVP2 dVP3 d I3 dI3
ðVCf þ VCC Þ ¼ Lf 2 Rf ;
dt dt dt dt dt
1 1 dVP2 dVP3 d 2 I3 dI3
I3 ð þ Þ¼ Lf 2 Rf
Cf CC dt dt dt dt
We can summarize our system differential equation:
1 1 dVP1 dVP2 dI1 d 2 I1

I1 ð þ Þ¼ Rf Lf 2 ;
Cf CC dt dt dt dt
dVP2 ðI2 þ IL1 Þ dðI2 þ IL1 Þ
I2 ¼ C1 ð þ Rc Þ
dt Cd dt
1 1 dVP2 dVP3 d 2 I3 dI3
I3 ð þ Þ¼ Lf 2 Rf
Cf CC dt dt dt dt
dVP2 I2 þ IL1 dðI2 þ IL1 Þ
I2 ¼ C1 ð þ Rc Þ
dt Cd dt
dVP2 I2 ðI2 þ IL1 Þ dðI2 þ IL1 Þ
) ¼ þ Rc
dt C1 Cd dt
We get two main system differential equations:
dVP3 dI3 ðt s3 Þ
¼ Rload
VP3 ¼ Rload I3 ðt s3 Þ )
dt dt
1 1 dVP1 I2 ðI2 þ IL1 Þ dðI2 þ IL1 Þ
I1 ð þ Þ¼ ½ þ Rc
Cf CC dt C1 Cd dt
dI1 d 2 I1
Rf Lf 2
dt dt
1 1 I2 ðI2 þ IL1 Þ dðI2 þ IL1 Þ
I3 ð þ Þ¼ þ Rc Rload
Cf CC C1 Cd dt
dI3 ðt s3 Þ d 2 I3 dI3
Lf 2 Rf
dt dt dt
Since I1 ¼ I2 þ I3 ) I2 ¼ I1 I3 we get the following system differential

equations:
1 1 dVP1 I1 I3 ðI1 I3 þ IL1 Þ

I1 ð þ Þ¼ ½ þ
Cf CC dt C1 Cd
dðI1 I3 þ IL1 Þ dI1 d 2 I1
Rc Rf Lf 2
dt dt dt
1 1 I1 I3 ðI1 I3 þ IL1 Þ dðI1 I3 þ IL1 Þ
I3 ð þ Þ¼ þ Rc
Cf CC C1 Cd dt
dI3 ðt s3 Þ d 2 I3 dI3
Rload Lf 2 Rf
dt dt dt
dIL1
We define the following new variables: I10 ¼ dIdt1 ; I30 ¼ dIdt3 ; IL0 1 ¼ dt
dðI1 I3 þ IL1 Þ d 2 I1 dI 0 d 2 I3 dI 0
¼ I10 I30 þ IL0 1 ; 2 ¼ 1 ; 2 ¼ 3
dt dt dt dt dt
1 1 dVP1 I1 I3 ðI1 I3 þ IL1 Þ
I1 ð þ Þ¼ ½ þ
Cf CC dt C1 Cd
0
dI
ðI10 I30 þ IL0 1 Þ Rc I10 Rf Lf 1
dt
dI10 dVP1 I1 I3 ðI1 I3 þ IL1 Þ
Lf ¼ ½ þ
dt dt C1 Cd
1 1
ðI10 I30 þ IL0 1 Þ Rc I10 Rf I1 ð þ Þ
Cf CC
dI 0 dVP1 I1 I3 I1 I3 IL
Lf 1 ¼ ½ þ þ 1
dt dt C1 C1 Cd Cd Cd
I10 Rc þ I30 Rc IL0 1 Rc
1 1
I10 Rf I1 ð þ Þ
Cf CC
dI10 1 dVP1 1 1 1 1 1 1 1 1 I L1
¼ I1 ½ þ þ þ þ I3 ½ þ
dt Lf dt C1 Cd Cf C C Lf C1 Cd Lf Cd Lf
1 Rc Rc
þ I10 ½Rc Rf I30 þ IL0 1
Lf Lf Lf
1 1 I1 I3 ðI1 I3 þ IL1 Þ
I3 ð þ Þ¼ þ ½I10 I30 þ IL0 1
Cf CC C1 Cd
dI3 ðt s3 Þ dI 0
Rc Rload Lf 3 I30 Rf
dt dt
1 1 I1 I3 I1 I3 IL
I3 ð þ Þ¼ þ þ 1 I10 Rc þ I30 Rc IL0 1
Cf CC C1 C1 Cd Cd Cd
dI3 ðt s3 Þ dI 0
Rc Rload Lf 3 I30 Rf
dt dt
dI30 1 1 1 1 1 1
Lf ¼ I1 ½ þ I3 ½ þ þð þ Þ
dt C1 Cd C1 Cd Cf CC
IL dI3 ðt s3 Þ
þ 1 I10 Rc þ I30 ½Rc Rf IL0 1 Rc Rload
Cd dt
0
dI3 1 1 1 1 1 1 1
¼ I1 ½ þ I3 ½ þ þ þ
dt C1 Cd Lf C1 Cd Cf CC
1 IL1 Rc ½Rc Rf
þ I10 þ I30
Lf Cd Lf Lf Lf
Rc Rload dI3 ðt s3 Þ
IL0 1
Lf Lf dt
dIL1
Finally we get system set of differential equations: I10 ¼ dIdt1 ; I30 ¼ dIdt3 ; IL0 1 ¼ dt
dI10 1 dVP1 1 1 1 1 1 1 1 1 I L1
¼ I1 ½ þ þ þ þ I3 ½ þ
dt Lf dt C1 Cd Cf CC Lf C1 Cd Lf Cd Lf
1 Rc Rc
Lf Lf Lf
0
dI3 1 1 1 1 1 1 1
¼ I1 ½ þ I3 ½ þ þ þ
1 I L1 Rc ½Rc Rf
þ I10 þ I30
Lf Cd Lf Lf Lf
Rc Rload dI3 ðt s3 Þ
IL0 1
Lf Lf dt
dI 0 dI 0
We add the above two differential equations: f dt1 ¼ . . .g þ f dt3 ¼ . . .g
dI10 dI 0 1 dVP1 1 1 1 1 1 1 Rf
þ 3¼ I1 ½ þ I3 ½ þ I10
dt dt Lf dt Cf CC Lf Cf CC Lf Lf
R R dI ðt s Þ
I30
f load 3 3

Lf Lf dt
dI10 dI 0 1 dVP1 1 1 1 Rf Rload dI3 ðt s3 Þ
þ 3¼ ½ þ ðI1 þ I3 Þ ðI10 þ I30 Þ
dt dt Lf dt Cf CC Lf Lf Lf dt
We define for simplicity new variables: X ¼ I1 þ I3 ; Y ¼ I10 þ I30 ; dX

dt ¼ Y.
dY 1 dVP1 1 1 1 Rf Rload dI3 ðt s3 Þ dX

¼ ½ þ X Y ; ¼Y
dt Lf dt Cf C C Lf Lf Lf dt dt
I1 flows through delay line (ARB1) before it enters to port 1 then

I1 ðtÞ ! I1 ðt s1 Þ.
I2 flows through delay line (ARB3) before entering to Bias-T and IMPATT
diode circuit. Then I2 ðtÞ ! I2 ðt s2 Þ; accordingly, active circulator’s microstrip
transmission lines delays Δ1 for the first current derivative and Δ3 for the third port
current derivative I30 ðtÞ ! I30 ðt D3 Þ; I10 ðtÞ ! I10 ðt D1 Þ. I3 is the current which
flows through active circulator port 3 and it flows through delay line (ABR2) then
IRload ¼ I3 ðt s3 Þ. I2 ; I20 are hidden variables in our analysis [6, 12, 13]. There is no
time delay in IL0 1 ; IL1 . To find our system fixed point, dY dt ¼ 0 dt ¼ 0. There is no
dX
effect of variables delay in time since t si ) t si t; i ¼ 1; 2; 3 and t

Di ) t Di t ; i ¼ 1; 2; 3 then dI3 ðts dt
3Þ
! dIdt3 ðtÞ ¼ 0 .We get two main fixed
dVP
dVP1 1
points: X*, Y*. Y ¼ 0 and 1
Lf dt ½C1f þ CC
1
L1f X ¼ 0 ) X ¼ ½ 1 þdt1 1 .
Cf CC Lf
We consider the assumption which the IMPATT/Circulator amplifier input

voltage VP1 ¼ C þ nðtÞ ; C is constant voltage and nðtÞ is RF signal in time.
dVP1 dC dnðtÞ dC dnðtÞ dnðtÞ dnðtÞ

¼ þ ; ! 0; ! e or ! X0 sgn½
dt dt dt dt dt dt dt
dnðtÞ dnðtÞ dnðtÞ dnðtÞ dnðtÞ dnðtÞ dnðtÞ
¼ sgn½ j j ; 8 real 9j j ¼ sgn½
dnðtÞ
dj dt j dnðtÞ dnðtÞ dVP1 dnðtÞ
dnðtÞ
¼ sgn½ 8 6¼ 0 ; ¼ 0 for ¼0
d½ dt
dt dt dt dt
dVP1 dnðtÞ dVP1 dnðtÞ

dt ¼ X0 for dt [ 0 ; dt ¼ X0 for dt \0. Then we get some possibilities
for X . X ¼

0 8 dt ¼ 0. X ¼ ½ 1 þX01 1 8 dnðtÞ
dnðtÞ X0 dnðtÞ
dt [ 0 X ¼ ½C1 þ C1 L1 8 dt \0 .
C C L f C f f C f
We get the following options for system fixed points:
dnðtÞ X0 dnðtÞ
E ð0Þ ðX ð0Þ ; Y ð0Þ Þ ¼ ð0; 0Þ 8 ¼ 0 ; E ð1Þ ðX ð1Þ ; Y ð1Þ Þ ¼ ð 1 ; 0Þ 8 [0
dt ½Cf þ C1C L1f dt
X0 dnðtÞ
E ð2Þ ðX ð2Þ ; Y ð2Þ Þ ¼ ð ; 0Þ 8 \0:
½C1f þ C1C L1f dt
Stability analysis: We got system five differential equations:
dI1 0 dI3 0 dIL1

I10 ¼ ; I3 ¼ ; I L1 ¼
dt dt dt
dI10 1 dVP1 1 1 1 1 1 1 1 1 I L1
¼ I1 ½ þ þ þ þ I3 ½ þ
dt Lf dt C1 Cd Cf C C Lf C1 Cd Lf Cd Lf
1 Rc Rc
Lf Lf Lf
dI30 1 1 1 1 1 1 1
¼ I1 ½ þ I3 ½ þ þ þ
1 I L1 Rc ½Rc Rf
þ I10 þ I30
Lf Cd Lf Lf Lf
R c R dI ðt s Þ
IL0 1
load 3 3

Lf Lf dt
We define for simplicity two global parameters:

1 1 1 1
CP 1 ¼ þ þ þ
C1 Cd Cf CC
1 1 dI1 dI3 0 dIL1 0
CP 2 ¼ þ : ¼ I10 ; ¼I3 ; ¼IL1
C1 Cd dt dt dt
dI10 1 dVP1 1 1 I L1
¼ I1 CP 1 þ I3 CP 2
dt Lf dt Lf Lf Cd Lf
1 R c R c
Lf Lf Lf
0
dI3 1 1 I L1 Rc
¼ I1 CP 2 I3 CP 1 þ I10
dt Lf Lf Cd Lf Lf
½Rc Rf Rc Rload dI3 ðt s3 Þ
þ I30 IL0 1
Lf Lf Lf dt
1 Rc 1
N11 ¼ ½Rc Rf ; N12 ¼ ; N13 ¼ CP 1 ;
Lf Lf Lf
1 1
N14 ¼ CP 2 ; N15 ¼
Lf Cd Lf
Rc ½Rc Rf 1
N21 ¼ ; N22 ¼ ; N23 ¼ CP 2 ;
Lf Lf Lf
1 1
N24 ¼ CP 1 ; N25 ¼
Lf Cd Lf
N31 ¼ 1 ; N32 ¼ 0 ; N33 ¼ 0 ; N34 ¼ 0 ; N35 ¼ 0 ; N41 ¼ 0;
N42 ¼ 1 ; N43 ¼ 0 ; N44 ¼ 0
N45 ¼ 0 ; N51 ¼ N55 ¼ 0:
We can write our system differential equations matrix representation:

0 dI 0 1 0R 1
1 0 0 1 0 1 c
B dIdt30 C 0 1 I1 1
B Lf
B C N11 . . . N15 B I0 C B 0 C R C
B dt C B . . . C B 3 C B C 1 dVP1 B B Lfc C
C
B dI1 C ¼ @ . .. .. A B C B C
B dt C . B I1 C þ B 0 C Lf dt þ B
B 0 C C
B dI3 C @ I3 A @ 0 A @0 A
@ dt A N51 N55
dIL1 I L1 0 0
dt 0 1
0
B Rload C
B Lf C dI3 ðt s3 Þ
IL0 1 þ B
B0
C
C
@0 A dt
0
dI10 dI30 dI1 dI3 dIL1 0

We consider no delay effect on dt ; dt ; dt ; dt ; dt . IL1 ! e; s2 ; D2 ! e.
I1 ðtÞ ! I1 ðt s1 Þ ; I10 ðtÞ ! I10 ðt D1 Þ:

0 dI 0 1
1 0 0 1 0 1
B dt0 C 0 1 I1 ðt D1 Þ 1
B dI3 C N . . . N B 0 C B C
B dt C 11 15 B I ðtÞ C B0C
B C B . . . C B 3 C B C
B 1 C¼B . .. .. C B C B C
A B I1 ðt s1 Þ C þ B 0 C
dI
B dt C @ .
B dI C B C B C
B 3 C N51 N55 @ I3 ðtÞ A @0A
@ dt A
dIL1 IL1 ðtÞ 0
dt
0 1
0
B Rload C
B Lf C
1 dVP1 B C dI3 ðt s3 Þ
þBB 0 C
C þe
Lf dt B C dt
@0 A
0
To find equilibrium points (fixed points) of active circulator IMPATT amplifier

system by the following assumptions:
lim I 0 ðt D1 Þ ¼ I10 ðtÞ ; lim I1 ðt s1 Þ ¼ I1 ðtÞ ;

t!1 1 t!1
dI 0 dI 0
lim I3 ðt s3 Þ ¼ I3 ðtÞ ; 1 ¼ 0 ; 3 ¼ 0
t!1 dt dt
dI1 dI3 dIL1 0ðÞ 0ðÞ
¼ 0; ¼ 0; ¼ 0 ; I3 ¼ 0 ; I1 ¼ 0
dt dt dt
1 dVP1 1 1 I
I1 CP 1 þ I3 CP 2 L1 ¼ 0
Lf dt Lf Lf Cd Lf

1 1 I dVP1
I1 CP 2 I3 CP 1 þ L1 ¼ 0 ; þ ðI1 þ I3 Þ ½CP 2 CP 1 ¼ 0
Lf Lf Cd Lf dt
1
dVP
dVP1
þ ðI1 þ I3 Þ ½CP 2 CP 1 ¼ 0 ) I1 þ I3 ¼ P dt P :
dt C 1C 2
The standard local stability analysis about anyone of the equilibrium points of
active circulator IMPATT amplifier system consist in adding to coordinates
½I10 ; I30 ; I1 ; I3 ; IL1 arbitrarily small increments of exponential form ½i01 ; i03 ; i1 ; i3 ; iL1
ekt and retaining the first order terms in I10 ; I30 ; I1 ; I3 ; IL1 . The system of five ho-
mogeneous equations leads to a polynomial characteristics equation in the eigen-
values-λ. The polynomial characteristics equations accept by set the below currents
and currents derivative with respect to time into active circulator IMPATT diode
system equations. Active circulator IMPATT diode system fixed values with
arbitrarily small increments of exponential form ½i01 ; i03 ; i1 ; i3 ; iL1 ekt are: j = 0
(first fixed point), j = 1 (second fixed point), j = 2 (third fixed point), etc.
0ðjÞ 0ðjÞ ðjÞ

I10 ðtÞ ¼ I1 þ i01 ekt ; I30 ðtÞ ¼ I3 þ i03 ekt ; I1 ðtÞ ¼ I1 þ i1 ekt
ðjÞ ðjÞ
I3 ðtÞ ¼ þ i3 ekt ; IL1 ðtÞ ¼
I3 þ iL1 ekt .
I L1
We choose the expressions for our I10 ðtÞ; I30 ðtÞ; I1 ðtÞ; I3 ðtÞ; IL1 ðtÞ as small dis-
placement ½i01 ; i03 ; i1 ; i3 ; iL1 from the active circulator IMPATT diode system fixed
0ðjÞ
points at time t = 0. I10 ðt ¼ 0Þ ¼ I1 þ i01
0ðjÞ ðjÞ ðjÞ

I30 ðt ¼ 0Þ ¼ I3 þ i03 ; I1 ðt ¼ 0Þ ¼ I1 þ i1 ; I3 ðt ¼ 0Þ ¼ I3 þ i3 ; IL1 ðt ¼ 0Þ
ðjÞ
¼ I L1 þ i L1 :
For λ < 0, t > 0, the selected fixed point is stable otherwise λ > 0, t > 0 is
unstable (Table 1). Our system tends to the selected fixed point exponentially for
λ < 0, t > 0 otherwise go away from the selected fixed point exponentially. λ is the
eigenvalue parameter which is established if the fixed point is stable or unstable;
additionally, his absolute value (|λ|) establishes the speed of flow toward or away
from the selected fixed point [2] (Jack and Huseyin 1991). The speeds of flow
toward or away from the selected fixed point for active circulator IMPATT diode
amplifier system currents and currents derivatives with respect to time are:
0ðjÞ 0ðjÞ
dI10 ðtÞ I 0 ðt þ DtÞ I10 ðtÞ I þ i01 ekðt þ DtÞ ðI1 þ i01 ekt Þ
¼ lim 1 ¼ lim 1
dt Dt!0 Dt Dt!0 Dt
i01 ekt ½ekDt 1 ekDt 1 þ kDt dI 0
ðtÞ
¼ lim ! k i01 ekt ; 3 ¼ k i03 ekt
Dt!0 Dt dt
dI1 ðtÞ dI ðtÞ
¼ k i1 ekt ; ¼ k i3 ekt ;
3
dt dt
dIL1 ðtÞ dI 0 ðt D1 Þ
¼ k iL1 ekt ; 1 ¼ k i01 ekt ekD1
dt dt
dI30 ðt D3 Þ dI 0 ðt D1 Þ
¼ k i03 ekt ekD1 ; 1 ¼ k i01 ekt ekD1 ;
dt dt
dI1 ðt s1 Þ
¼ k i1 ekt eks1
dt
dI3 ðt s3 Þ
¼ k i3 ekt eks3 :
dt
First we take The IMPATT amplifier system’s current differential equations:

dI1
dt ¼ I10 ; dIdt3 ¼I30 and adding coordinates ½I10 ; I30 ; I1 ; I3 ; IL1 arbitrarily small increments
of exponential terms ½i01 ; i03 ; i1 ; i3 ; iL1 ekt and retaining the first order terms in
0ðjÞ 0ðjÞ
i01 ; i03 ; i1 ; i3 ; iL1 : k i1 ekt ¼ I1 þ i01 ekt ; I1 ¼ 0 ) i01 k i1 ¼ 0.
0ðjÞ 0ðjÞ
k i3 ekt ¼ I3 þ i03 ekt ; I3 ¼ 0 ) i03 k i3 ¼ 0. Second we take the
active circulator IMPATT diode’s current derivatives I10 ; I30 differential equations:
dI10 1 dVP1 1 1 IL1

¼ I1 CP 1 þ I3 CP 2
dt Lf dt Lf Lf Cd Lf
1 Rc Rc
Lf Lf Lf
0
dI3 1 1 I L1 Rc
¼ I1 CP 2 I3 CP 1 þ I10
dt Lf Lf Cd Lf Lf
½Rc Rf Rc Rload dI3 ðt s3 Þ
þ I30 IL0 1
Lf Lf Lf dt
dV
We already get dtP1 ¼ X0 sgn½dnðtÞ dt . We add coordinates ½I10 ; I30 ; I1 ; I3 ; IL1
0 0 kt
arbitrarily small increments of exponential terms ½i1 ; i3 ; i1 ; i3 ; iL1 e and retaining
the first order terms in i01 ; i03 ; i1 ; i3 ; iL1 .
1 dnðtÞ
ðI1 þ i1 ekt Þ CP 1
ðjÞ
k i01 ekt ¼ X0 sgn½
Lf dt
1 1
þ ðI3 þ i3 ekt Þ CP 2
ðjÞ
Lf Lf
ðjÞ
ðIL1 þ iL1 ekt Þ 0ðjÞ
þ ðI1 þ i01 ekt Þ
Cd Lf
1 0ðjÞ Rc Rc
½Rc Rf ðI3 þ i03 ekt Þ þ ðIL0 1 ! eÞ
Lf Lf Lf
1
dVP
0ðjÞ 0ðjÞ ðjÞ ðjÞ Rc
I3 ¼ 0 ; I1 ¼ 0 ; I1 þ I3 ¼ P dt P ; ðIL0 1 ! eÞ 0
C 1C 2 Lf
1 dnðtÞ 1
I1 C P 1 þ I3
ðjÞ ðjÞ
ð#Þ k i01 ekt ¼ X0 sgn½
Lf dt Lf
1 1
CP 2 i1 CP 1 ekt
Lf Lf
ðjÞ
1 ðI þ iL1 ekt Þ
þ i3 CP 2 ekt L1 þ i01
Lf Cd Lf
1 Rc
ekt ½Rc Rf i03 ekt
Lf Lf
1
k i03 ekt ¼ ðI1 þ i1 ekt Þ CP 2 ðI3 þ i3 ekt Þ CP 1
ðjÞ ðjÞ
Lf
ðjÞ
1 ðI þ iL1 ekt Þ 0ðjÞ Rc
þ L1 ðI1 þ i01 ekt Þ
Lf Cd Lf Lf
0ðjÞ ½Rc Rf Rc Rload
þ ðI3 þ i03 ekt Þ ðIL0 1 ! eÞ
Lf Lf Lf
k i3 ekt eks3 ; eks3 ! 1
k i3 ekt ¼ I1 CP 2 L1f þ i1 CP 2 L1f

0 ðjÞ
ekt I3 CP 1 L1f i3 CP 1 L1f ekt

ðjÞ
ð##Þ ðjÞ
ðIL þ iL1 ekt Þ 0 0
i1 ekt Lfc þ i3
R
þ 1
Cd Lf
½Rc Rf
Lf RLload f
k i3 ekt
Adding two expressions: (#) + (##)
1 dnðtÞ 1
I 1 CP 1
ðjÞ
k i01 ekt þ k i03 ekt ¼ X0 sgn½
Lf dt Lf
1 1
þ I3 CP 2 i1 CP 1 ekt þ i3 CP 2
ðjÞ
Lf Lf
ðjÞ
1 kt ðIL1 þ iL1 ekt Þ
e þ i01 ekt ½Rc Rf
Lf Cd Lf
1 Rc 1 1
þ fI1 CP 2 þ i1 CP 2 ekt I3
ðjÞ ðjÞ
i03 ekt
Lf Lf Lf Lf
ðjÞ
1 1 ðI þ iL1 ekt Þ
CP 1 i3 CP 1 ekt þ L1
Lf Lf Cd L f
Rc ½R c R R
i01 ekt þ i03 ekt k i3 ekt g
f load

Lf Lf Lf
1 dnðtÞ
þ I1 ½CP 2 CP 1
ðjÞ
k ½i01 þ i03 ekt ¼ X0 sgn½
Lf dt
1 1
þ I3 ½CP 2 CP 1
ðjÞ
Lf Lf
1 1
þ i1 ½CP 2 CP 1 ekt þ i3 ½CP 2 CP 1 Rload k ekt
Lf Lf
1 Rc ½Rc Rf 0 kt Rc
þ i01 ekt ½Rc Rf i01 ekt þ i03 ekt i3 e
Lf Lf Lf Lf
1 dnðtÞ 1
þ ½CP 2 CP 1 ðI1 þ I3 Þ
ðjÞ ðjÞ
k ½i01 þ i03 ekt ¼ X0 sgn½
Lf dt Lf
1
þ i1 ½CP 2 CP 1 ekt þ i3 ½CP 2 CP 1 Rload k
Lf
1 Rf Rf
ekt i01 ekt i03 ekt
Lf Lf Lf
dVP
ðjÞ ðjÞ 1
We already found the system fixed points condition: I1 þ I3 ¼ CP dt
CP
1 2
1
dVP
1 dnðtÞ 1
k ½i01 þ i03 ekt ¼ X0 sgn½ þ ½CP 2 CP 1 P dt P
Lf dt Lf C 1 C 2
1
þ i1 ½CP 2 CP 1 ekt þ i3 ½CP 2 CP 1 Rload k
Lf
1 Rf Rf
ekt i01 ekt i03 ekt
Lf Lf Lf
1 dnðtÞ dVP1 1
k ½i01 þ i03 ekt ¼ ½X0 sgn½ þ i1 ½CP 2 CP 1 ekt
Lf dt dt Lf
1
þ i3 ½CP 2 CP 1 Rload k ekt i01 ekt
Lf
Rf Rf
i03 ekt
Lf Lf
dVP1 dnðtÞ 1
¼ X0 sgn½ ) k ½i01 þ i03 ekt ¼ i1 ½CP 2 CP 1 ekt
dt dt Lf
1 Rf Rf
þ i3 ½CP 2 CP 1 Rload k ekt i01 ekt i03 ekt
Lf Lf Lf
We divide two sides of the above equation by ekt
1 1 Rf
k ½i01 þ i03 ¼ i1 ½CP 2 CP 1 þ i3 ½CP 2 CP 1 Rload k i01
Lf Lf Lf
0 Rf
i3
Lf
1 1 Rf
i1 ½CP 2 CP 1 þ i3 ½CP 2 CP 1 Rload k i01
Lf Lf Lf
Rf
i03 k ½i01 þ i03 ¼ 0
Lf
1 1 Rf Rf
i1 ½CP 2 CP 1 þ i3 ðCP 2 CP 1 Þ i01 i03
Lf Lf Lf Lf
1
ði01 þ i03 þ i3 Rload Þ k ¼ 0
Lf
Three cases: CP 2 CP 1 [ Rf ; CP 2 CP 1 \Rf . CP 2 CP 1 ¼ Rf .

We can summery our equations of arbitrarily small increments:
i01 k i1 ¼ 0; i03 k i3 ¼ 0
1 1
i1 ½CP 2 CP 1 þ i3 ðCP 2 CP 1 Þ i01
Lf Lf
Rf Rf 1
i03 ði01 þ i03 þ i3 Rload Þ k ¼ 0
Lf Lf Lf
The active circulator IMPATT diode amplifier system eigenvalues options are
describe in the below table.
Table 3.8 Active circulator IMPATT diode amplifier system eigenvalues options
λ<0 λ>0
0ðjÞ 0ðjÞ
t=0 I10 ðt¼ 0Þ ¼ I1 þ i01 I10 ðt ¼ 0Þ ¼ I1 þ i01
0ðjÞ 0ðjÞ
I30 ðt¼ 0Þ ¼ I3 þ i03 I30 ðt ¼ 0Þ ¼ I3 þ i03
ðjÞ ðjÞ
I1 ðt ¼ 0Þ ¼ I1 þ i1 I1 ðt ¼ 0Þ ¼ I1 þ i1
ðjÞ ðjÞ
I3 ðt ¼ 0Þ ¼ I3 þ i3 I3 ðt ¼ 0Þ ¼ I3 þ i3
ðjÞ ðjÞ
IL1 ðt ¼ 0Þ ¼ IL1 þ iL1 IL1 ðt ¼ 0Þ ¼ IL1 þ iL1
0ðjÞ 0ðjÞ
t>0 I10 ðtÞ ¼ I1 þ i01 ejkjt I10 ðtÞ ¼ I1 þ i01 ejkjt
0ðjÞ 0ðjÞ
ðjÞ ðjÞ
ðjÞ ðjÞ
ðjÞ ðjÞ
IL1 ðtÞ ¼ IL1 þ iL1 ejkjt IL1 ¼ IL1 þ iL1 ejkjt
t > 0; t → ∞ I10 ðt ! 1Þ ¼ I1
0ðjÞ
I10 ðt ! 1Þ i01 ejkjt
0ðjÞ
I30 ðt ! 1Þ ¼ I3 I30 ðt ! 1Þ i03 ejkjt
I1 ðt ! 1Þ ¼ I1
ðjÞ I1 ðt ! 1Þ i1 ejkjt
ðjÞ I3 ðt ! 1Þ i3 ejkjt
I3 ðt ! 1Þ ¼ I3
ðjÞ IL1 ðt ! 1Þ iL1 ejkjt
IL1 ðt ! 1Þ ¼ IL1
We already define
ðjÞ ðjÞ
I1 ðt s1 Þ ¼ I1 þ i1 ekðts1 Þ ; I3 ðt s3 Þ ¼ I3 þ i3 ekðts3 Þ ;
0ðjÞ
I10 ðt D1 Þ ¼ I1 þ i01 ekðtD1 Þ
0ðjÞ ðjÞ
I30 ðt D3 Þ ¼ I3 þ i03 ekðtD3 Þ ; IL1 ¼ IL1 þ iL1 ekt
Then we get four delayed differential equations with respect to coordinates

½I10 ; I30 ; I1 ; I3 ; IL1
arbitrarily small increments of exponential ½i01 ; i03 ; i1 ; i3 ; iL1 ekt .
dI 0 dI 0 dIL1
We consider no delay effect on dt1 ; dt3 ; dIdt1 ; dIdt3 ; dt . We neglect the time delay s2
which related to active circulator branch 2.
0ðjÞ 0ðjÞ ðjÞ

I10 ðtÞ ¼ I1 þ i01 ekt ; I30 ðtÞ ¼ I3 þ i03 ekt ; I1 ðtÞ ¼ I1 þ i1 ekt
ðjÞ ðjÞ
I3 ðtÞ ¼ I3 þ i3 ekt ; IL1 ðtÞ ¼ IL1 þ iL1 ekt
1 Rc 1
N11 ¼ ½Rc Rf ; N12 ¼ ; N13 ¼ CP 1 ;
Lf Lf Lf
1 1
N14 ¼ CP 2 ; N15 ¼
Lf Cd Lf
Rc ½Rc Rf 1
N21 ¼ ; N22 ¼ ; N23 ¼ CP 2 ;
Lf Lf Lf
1 1
N24 ¼ CP 1 ; N25 ¼
Lf Cd Lf
N31 ¼ 1 ; N32 ¼ 0 ; N33 ¼ 0 ; N34 ¼ 0 ; N35 ¼ 0;
N41 ¼ 0 ; N42 ¼ 1 ; N43 ¼ 0 ; N44 ¼ 0
N45 ¼ 0 ; N51 ¼ N55 ¼ 0:
0 dI 0 1
0 1 0 1 0 1
I10 ðt D1 Þ
1
B dIdt30 C 0 1 1 0
B C N11 . . . N15 B I 0 ðtÞ C B0C B Rload C
B dt C B . .. .. C B 3 C B C 1 dVP1 B Lf C
B dI1 C¼@ . B C B C B C
B dt C . . . A B I1 ðt s1 Þ C þ B 0 C L dt þ B 0 C
B dI3 C @ I3 ðtÞ A @0A f @0 A
@ dt A N51 N55
dIL1 IL1 ðtÞ 0 0
dt
dI3 ðt s3 Þ
þe
dt
0ðjÞ 0ðjÞ ðjÞ

i01 k ekt ¼ N11 ðI1 þ i01 ekðtD1 Þ Þ þ N12 ðI3 þ i03 ekt Þ þ N13 ðI1 þ i1 ekðts1 Þ Þ
ðjÞ ðjÞ 1 dVP1
þ N14 ðI3 þ i3 ekt Þ þ N15 ðIL1 þ iL1 ekt Þ þ
Lf dt
0ðjÞ 0ðjÞ ðjÞ ðjÞ ðjÞ 1 dVP1

i01 k ekt ¼ N11 I1 þ N12 I3 þ N13 I1 þ N14 I3 þ N15 IL1 þ þ N11 i01 ekðtD1 Þ
Lf dt
þ N12 i03 ekt þ N13 i1 ekðts1 Þ þ N14 i3 ekt þ N15 iL1 ekt
dI10
We already know that at fixed point: dt ¼0
dI10 0ðjÞ 0ðjÞ ðjÞ ðjÞ ðjÞ 1 dVP1

¼ 0 ) N11 I1 þ N12 I3 þ N13 I1 þ N14 I3 þ N15 IL1 þ ¼0
dt Lf dt
i01 k ekt ¼ N11 i01 ekðtD1 Þ þ N12 i03 ekt

þ N13 i1 ekðts1 Þ þ N14 i3 ekt þ N15 iL1 ekt
i01 ½N11 ekD1 k ekt þ N12 i03 ekt þ N13 i1 ekðts1 Þ
þ N14 i3 ekt þ N15 iL1 ekt ¼ 0
i01 ½N11 ekD1 k þ N12 i03 þ N13 i1 eks1 þ N14 i3 þ N15 iL1 ¼ 0
0ðjÞ 0ðjÞ
k i03 et ¼ N21 ðI1 þ i01 ekðtD1 Þ Þ þ N22 ðI3 þ i03 ekt Þ
ðjÞ ðjÞ ðjÞ
þ N23 ðI1 þ i1 ekðts1 Þ Þ þ N24 ðI3 þ i3 ekt Þ þ N25 ðIL1 þ iL1 ekt Þ
Rload dI3 ðt s3 Þ

Lf dt
dI3 ðtÞ 0 0 dI3 ðts3 Þ

Assumption: since dt ¼I3 ðtÞ then I3 ðt D3 Þ ¼ dt .
0 0ðjÞ 0 kt 0 0ðjÞ 0 kðtD3 Þ 0ðjÞ
I3 ðtÞ ¼ I3 þ i3 e ) I3 ðt D3 Þ ¼ I3 þ i3 e . At fixed point I3 ¼ 0
dI30 dI3 ðts3 Þ
since dt ¼ 0 then dt j@ fixed point ¼ I3 ðt D3 Þj@fixed point ¼ i03 ekðtD3 Þ .
0
0ðjÞ 0ðjÞ ðjÞ

k i03 et ¼ N21 ðI1 þ i01 ekðtD1 Þ Þ þ N22 ðI3 þ i03 ekt Þ þ N23 ðI1 þ i1 ekðts1 Þ Þ
ðjÞ ðjÞ Rload 0 kðtD3 Þ
þ N24 ðI3 þ i3 ekt Þ þ N25 ðIL1 þ iL1 ekt Þ i3 e
Lf
0ðjÞ 0ðjÞ ðjÞ ðjÞ

k i03 et ¼ N21 I1 þ N22 I3 þ N23 I1 þ N24 I3
ðjÞ
þ N25 IL1 þ i01 N21 ekðtD1 Þ
þ i03 N22 ekt þ i1 N23 ekðts1 Þ þ i3 N24 ekt þ iL1
Rload 0 kðtD3 Þ
N25 ekt i3 e
Lf
dI30
We already know that at fixed point: dt ¼0
dI30 0ðjÞ 0ðjÞ ðjÞ ðjÞ ðjÞ

¼ 0 ) N21 I1 þ N22 I3 þ N23 I1 þ N24 I3 þ N25 IL1 ¼ 0
dt
k i03 et ¼ i01 N21 ekðtD1 Þ þ i03 N22 ekt þ i1 N23 ekðts1 Þ þ i3 N24
Rload
ekt þ iL1 N25 ekt k i03 ekðtD3 Þ
Lf
i01 N21 ekt ekD1 þ i03 N22 ekt k i03 et þ i1 N23 ekt eks1 þ i3
Rload 0 kðtD3 Þ
N24 ekt þ iL1 N25 ekt i3 e ¼0
Lf
Rload kD3
i01 N21 ekD1 þ i03 ½N22 e k þ i1 N23
Lf
eks1 þ i3 N24 þ iL1 N25 ¼ 0
dI1 0ðjÞ
¼ N31 I10 ðt D1 Þ ) k i1 ekt ¼ N31 ðI1 þ i01 ekðtD1 Þ Þ ;
dt
0ðjÞ
I1 ¼ 0 ; N31 i01 ekD1 k i1 ¼ 0
dI3 0ðjÞ
¼ N42 I30 ðtÞ ) k i3 ekt ¼ N42 ðI3 þ i03 ekt Þ ;
dt
0ðjÞ
I3 ¼ 0 ; N42 i03 k i3 ¼ 0
L1 is an element of IMPATT diode Bias-T circuit. It forwards DC current to bias

the IMPATT diode in the negative resistance characteristic region. It blocks any RF
signal which comes from V1 (DC voltage source). The IMPATT diode work point
is stable and the DC current which flows through inductor L1 is fixing.
dIL1
dt¼ 0 ) k iL1 ekt ¼ 0 ) k iL1 ¼ 0. We can summery our small increments
equations of our active circulator IMPATT diode amplifier system.
i01 ½N11 ekD1 k þ N12 i03 þ N13 i1 eks1 þ N14 i3 þ N15 iL1 ¼ 0
Rload kD3
i01 N21 ekD1 þ i03 ½N22 e k þ i1
Lf
N23 eks1 þ i3 N24 þ iL1 N25 ¼ 0
N31 i01 ekD1 k i1 ¼ 0 ; N42 i03 k i3 ¼ 0 ; k iL1 ¼ 0 ) k iL1 ¼ 0
The small increments Jacobian of our active circulator IMPATT diode amplifier
system is as follow:
0 1
0 1 i01
!11 ... !15 B 0 C
B i3 C
B . .. .. C B C
B . C B i1 C ¼ 0 ; !11 ¼ N11 ekD1 k;
@ . . . A B C
B C
!51 !55 @ i3 A
iL1
!12 ¼ N12 ; !13 ¼ N13 eks1 ; !14 ¼ N14
Rload
!15 ¼ N15 ; !21 ¼ N21 ekD1 ; !22 ¼ N22
Lf
ekD3 k ; !23 ¼ N23 eks1 ; !24 ¼ N24
!25 ¼ N25 ; !31 ¼ N31 ekD1 ; !32 ¼ 0 ; !33 ¼ k ;
!34 ¼ !35 ¼ 0 ; !41 ¼ 0 ; !42 ¼ N42
!43 ¼ 0 ; !44 ¼ k ; !45 ¼ 0 ; !51 ¼ !52 ¼ !53 ¼ !54 ¼ 0 ; 55 ¼ k
0 1
!11 . . . !15
B .. .. .. C
AkI ¼@ . . . A ; detjA k Ij¼ 0 ; Dðs1 ; D1 ; D3 Þ ¼ detjA k Ij
!51 !55
We inspect the occurrence of any possible stability switching resulting from the
increase of value of time delays s1 ; D1 ; D3 for the active circulator IMPATT diode
amplifier system general characteristic equation Dðs1 ; D1 ; D3 Þ ¼ 0[6].
Dðs1 ; D1 ; D3 Þ ¼ detjA k Ij ¼ ðN11 ekD1 kÞ

0 1
ðN22 RLload
f
ekD3 kÞ . . . N25
B C
detB .. .. .. C
@ . . . A
0 k
0 kD1
1
N21 e N25 ...
B .. .. C
..
N12 detB
@ .
C
. A þ N13 e
.
ks1
0 k
0 1 0 1
N21 ekD1 . . . N25 N21 ekD1 . . . N25
B .. .. .. C B .. .. .. C
detB
@ . .
C B
. A N14 det@ . .
C
. A
0 k 0 k
0 kD1
1
N21 e . . . N24
B .. .. .. C
þ N15 detB
@ . .
C
. A
0 0
P5
We define gi ; i ¼ 1; 2; . . .; 5 functions. Dðs1 ; D1 ; D3 Þ ¼ detjA k Ij ¼ i¼1 gi
gi ¼ gi ðNlk ; Rload ; Lf ; k; D1 ; D3 ; s1 Þ; l ¼ 1; . . .; 5 ; k ¼ 1; . . .; 5
g1 ¼ g1 ðNlk ; Rload ; Lf ; k; D1 ; D3 ; s1 Þ; g2 ¼ g2 ðNlk ; Rload ; Lf ; k; D1 ; D3 ; s1 Þ
g3 ¼ g3 ðNlk ; Rload ; Lf ; k; D1 ; D3 ; s1 Þ; g4 ¼ g4 ðNlk ; Rload ; Lf ; k; D1 ; D3 ; s1 Þ ; g5 ¼ 0
g1 ¼ g1 ðNlk ; Rload ; Lf ; k; D1 ; D3 ; s1 Þ ¼ ðN11 ekD1 kÞ

0 1
ðN22 RLload f
ekD3 kÞ . . . N25
B C
detB .. .. .. C
@ . . . A
0 k
0 1
N21 ekD1 . . . N25
B .. .. .. C
g2 ¼ g2 ðNlk ; Rload ; Lf ; k; D1 ; D3 ; s1 Þ ¼ N12 detB @ . . . A
C
0 k
0 kD1
1
N21 e . . . N25
B .. .. .. C
g3 ¼ g3 ðNlk ; Rload ; Lf ; k; D1 ; D3 ; s1 Þ ¼ N13 eks1 detB
@ . .
C
. A
0 k
0 kD1
1
N21 e . . . N25
B .. .. .. C
g4 ¼ g4 ðNlk ; Rload ; Lf ; k; D1 ; D3 ; s1 Þ ¼ N14 detB @ . . . A
C
0 k
0 kD1
1
N21 e . . . N24
B .. .. .. C
g5 ¼ N15 detB @ . .
C
. A¼0
0 0
Rload kD3
g1 ¼ ðN11 ekD1 kÞ ðN22 k3 þ k3 e þ k4 k2 N24 N42 Þ
Lf
Rload kðD3 þ D1 Þ
¼ N11 N24 N42 ekD1 k2 þ fN11 e
Lf
Rload kD3
N11 N22 ekD1 g k3 þ ðN11 ekD1 e g
Lf
k4 þ N24 N42 k3 þ N22 k4 k5
g2 ¼ N12 fk3 N21 ekD1 N23 N31 k2 ekðs1 þ D1 Þ g ¼ k3 N12 N21 ekD1
þ N23 N31 N12 k2 ekðs1 þ D1 Þ
g3 ¼ N13 eks1 fN22 N31 ekD1

Rload kðD3 þ D1 Þ
k2 þ e N31 k2 þ k3 N31 ekD1
Lf
N24 N31 N42 k ekD1 g ¼ k3 N31 N13 ekðD1 þ s1 Þ
Rload kðD3 þ D1 þ s1 Þ
þ k2 fN31 N13 e
Lf
N31 N13 N22 ekðD1 þ s1 Þ g N24 N31 N42 N13 ekðD1 þ s1 Þ k
g4 ¼ N14 ðN21 N42 ekD1 k2 N23 N31 N42 ekðs1 þ D1 Þ kÞ

¼ N14 N21 N42 ekD1 k2
þ N14 N23 N31 N42 ekðs1 þ D1 Þ k
We analyze our system stability switching for the following cases:

(I) s1 ¼ s ; D1 ¼ D2 ¼ 0 (II) s1 ¼ 0 ; D1 ¼ D2 ¼ D
(III) s1 ¼ s ; D1 ¼ k s ; D2 ¼ ð1 kÞ s ; 0\k\1
We summery our results in the following table: Dðs1 ; D1 ; D3 Þ ¼
P5
detjA k Ij ¼ gi
i¼1
of value of the time delays D; s and balance parameter k (0 < k < 1) for the general
characteristic equation DðD; s; kÞ. If we choose s parameter then Dðk; s; kÞ ¼
P5
gi ¼ Pn ðk; s; kÞ þ Qm ðk; s; kÞ eks . The expression for Pn ðk; sÞ:
i¼1
Table 3.9 IMPATT amplifier system stability switching cases

Case (I) s1 ¼ s ; D1 ¼ D2 ¼ 0 Case (II) s1 ¼ 0; D1 ¼ D2 ¼ D s1 ¼ s; D1 ¼ k s
Case (III)
D2 ¼ ð1 kÞ s; 0\k\1
g1 g1 ¼ N11 N24 N42 k2 g1 ¼ ½N11 N24 N42 k2 g1 ¼ N11 N24 N42 ekks k2
Rload
þ fN11 N11 N22
þð
Rload kD
e N22 Þ N11 k3 þ ðN24 N42 N11 N22 ekks Þ k3
Lf
Lf
þ N24 N42 g k3 þ ðN11 ekks þ N22 Þ k4 k5
Rload
Rload þ ðN11 Þ k4 ekD Rload 3
þ fN11 þ N22 g k4 k5 L f þ ðN11 k
Lf Lf
þ N24 N42 k3 þ N22 k4 k5 Rload kks 4
e k Þ eks
Lf
g2 g2 ¼ k3 N12 N21 g2 ¼ ðN12 N23 N31 k2 g2 ¼ N12 N21 ekks k3
þ k2 N12 N23 N31 eks þ N12 N21 k3 Þ ekD þ N12 N23 N31 ekks eks k2
g3 g3 ¼ fN24 N31 N42 N13 k g3 ¼ fN24 N31 N42 N13 k g3 ¼ N13 N31 ekks eks k3
Rload Rload kD Rload ks
þ ðN13 N31 þ ðN13 N31 e þ fN13 N31 e
Lf Lf Lf
N13 N22 N31 Þ k2 N13 N22 N31 Þ k2 N13 N22 N31 ekks g k2 eks
þ N13 N31 k3 g eks þ N13 N31 k3 g ekD N24 N31 N42 N13 ekks eks k
g4 g4 ¼ N14 N21 N42 k2 g4 ¼ N14 N21 N42 k2 g4 ¼ N14 N21 N42 ekks k2
þ N14 N23 N31 N42 eks k þ N14 N23 N31 N42 eks k þ N14 N23 N31 N42 eks ekks k
X
n
Pn ðk; sÞ ¼ Pj ðsÞ k j ¼ P0 ðsÞ þ P1 ðsÞ k þ P2 ðsÞ k2 þ P3 ðsÞ k3 þ
j¼0
The expression for Qm ðk; sÞ:
X
m
Qm ðk; sÞ ¼ qj ðsÞ k j ¼ q0 ðsÞ þ q1 ðsÞ k þ q2 ðsÞ k2 þ q3 ðsÞ k3 þ
j¼0
P
5
If we choose D parameter then Dðk; DÞ ¼ gi ¼ Pn ðk; DÞ þ Qm ðk; DÞ ekD .
i¼1
The expression for Pn ðk; DÞ:
X
n
Pn ðk; DÞ ¼ Pj ðDÞ k j ¼ P0 ðDÞ þ P1 ðDÞ k þ P2 ðDÞ k2 þ P3 ðDÞ k3 þ
j¼0
The expression for Qm ðk; DÞ [38, 39]:
X
m
Qm ðk; DÞ ¼ qj ðDÞ k j ¼ q0 ðDÞ þ q1 ðDÞ k þ q2 ðDÞ k2 þ q3 ðDÞ k3 þ
j¼0
Remark: Balance parameter k (0 < k < 1) appears only in the third case. For all
cases which s and/or k or D include in Pn and Qm expressions, we take the fol-
lowing assumptions for power Taylor approximation series only for these
expressions.
X
1
ðk k sÞn k2 k 2 s 2
ekks ¼ 1 k k sþ ;
n¼0
n! 2
X
1
ðk sÞn k2 s 2
eks ¼ 1 k sþ
n¼0
n! 2
X
1
ðk k sÞn k2 k 2 s 2
ekks ¼ 1þk k sþ ;
n¼0
n! 2
X
1
ðk sÞn k2 s 2
eks ¼ 1þk sþ
n¼0
n! 2
X
1
ðk DÞn k2 D2
ekD ¼ 1 k Dþ ;
n¼0
n! 2
X
1
ðk DÞn k2 D 2
ekD ¼ 1þk Dþ
n¼0
n! 2
Balance parameter, k only appears in the third case (Case III). We use the
general geometric criterion [BK] and investigate the occurrence of any possible
stability switching resulting from the increase of value of time delay parameters s; D
for the general characteristic equation DðD; s; kÞ ¼ 0 [6].
Dðk; sÞ ¼ Pn ðk; sÞ þ Qm ðk; sÞ eks ; Dðk; DÞ ¼ Pn ðk; DÞ þ Qm ðk; DÞ ekD
P
n
In the case our time delay parameter is s: Pn ðk; sÞ ¼ Pj ðsÞ k j Qm ðk; sÞ ¼
j¼0
P
m
qj ðsÞ k j . In the case our time delay parameter is D:
j¼0
P P
Pn ðk; DÞ ¼ nj¼0 Pj ðDÞ k j ; Qm ðk; DÞ ¼ m j¼0 qj ðDÞ k
j
. n; m 2 N0 ; n [
m; pj ðÞ; qj ðÞ:R þ 0 ! R The expressions are continuous and differentiable func-
tions of s or D.
We summery our Pn ðk; s; kÞ and Qm ðk; s; kÞ expressions in the below table:
There are three cases to analyze stability switching under delay parameter
variation. We choose to analyze the second case (II). It is reader exercise to do the
same analysis for cases I and III. According stability switch criteria [BK],
n; m 2 N0 ; n [ m.
X
1
ðk DÞn
Pn ðk; DÞ ¼ N24 N42 k3 þ N22 k4 k5 ; ekD 1kD
n¼0
n!
Table 3.10 IMPATT amplifier system Pn ðk; s; kÞ and Qm ðk; s; kÞ expressions

Case (I) s1 ¼ s; D1 ¼ D2 ¼ 0 Case (II) s1 ¼ 0; D1 ¼ D2 ¼ D s1 ¼ s; D1 ¼ k s
Case (III)
n = 5; m = 3; n > m n = 5; m = 4; n > m D2 ¼ ð1 kÞ s; 0\k\1
Pn ðk; sÞ Pn ðk; sÞ ¼ k2 ðN11 N24 N42 Pn ðk; DÞ ¼ N24 N42 k3 Pn ðk; s; kÞ ¼ ½N14 N21
Or þ N14 N21 N42 Þ þ N22 k4 k5 N11 N24 N42 ekks k2
Pn ðk; s; kÞ
þ k3 ðN11
Rload
N11 N22 þ ½N24 N42 N11 N22 ekks
Or Lf
þ N12 N21 ekks k3
Pn ðk; DÞ þ N24 N42 þ N12 N21 Þ
þ ðN11 ekks þ N22 Þ k4 k5
Rload
þ k4 ðN11 þ N22 Þ k5
Lf
Qm ðk; sÞ Qm ðk; sÞ ¼ ðN24 N31 N42 N13 Qm ðk; DÞ ¼ ðN24 N31 N42 N13 Rload kks 4
Qm ðk; s; kÞ ¼ e k
Or þ N14 N23 N31 N42 Þ k þ N14 N23 N31 N42 Þ k Lf
Qm ðk; s; kÞ þ ðN12 N23 N31 þ ðN11 N24 N42 þ N12 N23 N31 þ ½N11
Rload
þ N13 N31 ekks k3
Or Lf
Rload Rload kD
Qm ðk; DÞ þ N13 N31 þ N13 N31 e þ ½N12 N23 N31 ekks
Lf Lf
Rload ks
N13 N22 N31 Þ k2 N13 N22 N31 þ N14 N21 N42 Þ k2 þ N13 N31 e
Lf
þ N13 N31 k3 þ fð
Rload kD
e N22 Þ N11
Lf N13 N22 N31 ekks k2
þ N12 N21 þ N13 N31 g k3 þ ½N14 N23 N42
Rload N24 N42 N13 N31 ekks k
þ ðN11 Þ k4
Lf
Qm ðk; DÞ ¼ ðN24 N31 N42 N13 þ N14 N23 N31 N42 Þ k

Rload
þ ðN11 N24 N42 þ N12 N23 N31 þ N13 N31
Lf
ekD N13 N22 N31 þ N14 N21 N42 Þ k2
Rload kD
þ fð e N22 Þ N11 þ N12 N21
Lf
Rload
þ N13 N31 g k3 þ ðN11 Þ k4
Lf

þ ðN11 N24 N42 þ N12 N23 N31 þ N13 N31
Rload
½1 k D N13 N22 N31 þ N14 N21 N42 Þ k2
Lf
Rload
þ fð ½1 k D N22 Þ N11 þ N12
Lf
Rload
N21 þ N13 N31 g k3 þ ðN11 Þ k4
Lf

þ ðN11 N24 N42 þ N12 N23 N31 þ N13 N31
Rload
N13 N22 N31 þ N14 N21 N42 Þ k2
Lf
Rload Rload
þf N11 N22 N11 þ N12 N21 þ N13 N31 N13 N31 Dg
Lf Lf
Rload Rload
k3 þ fN11 D N11 g k4
Lf Lf
Result: n = 5, m = 4; n > m. n; m 2 N0 ; n [ m
The expression Pn ðk; s; kÞ is as follow:
X
n
Pn ðDÞ ¼ Pj ðDÞ k j ¼ P0 ðDÞ þ P1 ðDÞ k þ P2 ðDÞ
j¼0
k2 þ P3 ðDÞ k3 þ P4 ðDÞ k4 þ P5 ðDÞ k5

P0 ðDÞ ¼ 0; P1 ðDÞ ¼ 0; P2 ðDÞ ¼ 0; P5 ðDÞ ¼ 1; P3 ðDÞ ¼ N24 N42 ; P4 ðDÞ ¼ N22
The expression for Qm ðDÞ is as follow:

X
m
Qm ðDÞ ¼ qj ðDÞ k j ¼ q0 ðDÞ þ q1 ðDÞ k þ q2 ðDÞ k2 þ q3 ðDÞ k3 þ q4 ðDÞ k4
j¼0
q0 ðDÞ ¼ 0; q1 ðDÞ ¼ N24 N31 N42 N13 þ N14 N23 N31 N42
q2 ðDÞ ¼ N11 N24 N42 þ N12 N23 N31 þ N13 N31
Rload
N13 N22 N31 þ N14 N21 N42
Lf
Rload Rload
q3 ðDÞ ¼ N11 N22 N11 þ N12 N21 þ N13 N31 N13 N31 D
Lf Lf
Rload Rload
q4 ðDÞ ¼ N11 D N11
Lf Lf
The homogeneous system for I10 ; I30 ; I1 ; I3 ; IL1 leads to a characteristic equation
P
for the eigenvalue λ having the form PðDÞ þ QðDÞ ekD ¼ 0; PðDÞ ¼ 5j¼0 aj
P
k j ; QðDÞ ¼ 4j¼0 cj k j and the coefficients faj ðqi ; ql ; sÞ; cj ðqi ; ql ; sÞg 2 R depend
on qi ; ql and delay s. qi ; ql are any system’s parameters, other parameters kept as a
constant a0 ¼ 0; a1 ¼ 0; a5 ¼ 1
a2 ¼ 0; a3 ¼ N24 N42 ; a4 ¼ N22 ; c0 ¼ 0;

c1 ¼ N24 N31 N42 N13 þ N14 N23 N31 N42
c2 ¼ N11 N24 N42 þ N12 N23 N31 þ N13 N31
Rload
N13 N22 N31 þ N14 N21 N42
Lf
Rload
c3 ¼ N11 N22 N11 þ N12 N21 þ N13 N31 N13 N31
Lf
Rload Rload Rload
D; c4 ¼ N11 D N11
Lf Lf Lf
Unless strictly necessary, the designation of the variation arguments ðqi ; ql Þ will
a0 þ c0 ¼ 0 8 qi ; ql 2 R þ ; that is λ = 0 is of Pðk; DÞ þ Qðk; DÞ eks ¼ 0.
Furthermore, Pðk; DÞ; Qðk; DÞ are analytic functions of λ, for which the following
requirements of the analysis [5, 41] can also be verified in the present case:
polynomial in ω.
(d) Each positive root xðqi ; ql Þ of FðxÞ ¼ 0 is continuous and differentiable with
respect to qi ; ql .
We assume that Pn ðDÞ and Qm ðDÞ can’t have common imaginary roots. That is
for any real number ω;
Pn ðk ¼ i x; DÞ þ Qm ðk ¼ i x; DÞ 6¼ 0; k2 ¼ x2 ; k3 ¼ i x3
k4 ¼ x4 ; k5 ¼ i x5 :Pn ðk ¼ i x; DÞ ¼ N22 x4 i ½N24 N42 x3 þ x5
Qm ðk ¼ i x; DÞ ¼ ðN24 N31 N42 N13 þ N14 N23 N31 N42 Þ i x

Rload
ðN11 N24 N42 þ N12 N23 N31 þ N13 N31 N13 N22 N31 þ N14 N21 N42 Þ x2
Lf
Rload Rload
f N11 N22 N11 þ N12 N21 þ N13 N31 N13 N31 Dg i x3
Lf Lf
Rload Rload
þ fN11 D N11 g x4
Lf Lf
Qm ðk ¼ i x; DÞ ¼ ðN11 N24 N42 þ N12 N23 N31 þ N13 N31

Rload
N13 N22 N31 þ N14 N21 N42 Þ x2
Lf
Rload Rload
þ fN11 D N11 g x4
Lf Lf
þ fðN24 N31 N42 N13 þ N14 N23 N31 N42 Þ x
Rload
ð N11 N22 N11 þ N12 N21 þ N13
Lf
Rload
N31 N13 N31 DÞ x3 g i
Lf
Pn ðk ¼ i x; DÞ þ Qm ðk ¼ i x; DÞ ¼ ðN11 N24 N42 þ N12

Rload
N23 N31 þ N13 N31
Lf
Rload
N13 N22 N31 þ N14 N21 N42 Þ x2 þ fN11
Lf
Rload
D N11 þ N22 g x4
Lf
þ i fðN24 N31 N42 N13 þ N14 N23 N31 N42 Þ x
Rload
ð N11 N22 N11 þ N12 N21 þ N13 N31 N13 N31
Lf
Rload
D þ N24 N42 Þ x3 x5 g 6¼ 0
Lf
jPði x; DÞj2 ¼ N224 N242 x6 þ ½N222 þ 2 N24 N42 x8 þ x10 . We define for
simplicity the following global parameters:
X1 ¼ N11 N24 N42 þ N12 N23 N31 þ N13 N31

Rload
N13 N22 N31 þ N14 N21 N42
Lf
Rload Rload
X2 ðDÞ ¼ N11 D N11 ;
Lf Lf
X3 ¼ N24 N31 N42 N13 þ N14 N23 N31 N42
Rload Rload
X4 ðDÞ ¼ N11 N22 N11 þ N12 N21 þ N13 N31 N13 N31 D
Lf Lf
Qm ðk ¼ i x; DÞ ¼ X1 x2 þ X2 ðDÞ x4 þ fX3 x X4 ðDÞ x3 g i
jQði x; DÞj2 ¼ fX1 x2 þ X2 ðDÞ x4 g2 þ fX3 x X4 ðDÞ x3 g2

jQði x; DÞj2 ¼ X21 x4 þ X22 ðDÞ x8 2 X1 X2 ðDÞ
x6 þ X23 x2 þ X24 ðDÞ x6 2 X3 X4 ðDÞ x4
jQði x; DÞj2 ¼ X23 x2 þ ½X21 2 X3 X4 ðDÞ x4
þ ½X24 ðDÞ 2 X1 X2 ðDÞ x6 þ X22 ðDÞ x8
Fðx; DÞ ¼ jPði x; DÞj2 jQði x; DÞj2 ¼ N224 N242 x6

þ ½N222 þ 2 N24 N42 x8 þ x10
X23 x2 ½X21 2 X3 X4 ðDÞ x4
½X24 ðDÞ 2 X1 X2 ðDÞ x6 X22 ðDÞ x8
Fðx; DÞ ¼ jPði x; DÞj2 jQði x; DÞj2 ¼ X23 x2 ½X21 2 X3 X4 ðDÞ x4

þ N224 N242 x6 ½X24 ðDÞ 2 X1 X2 ðDÞ x6
þ ½N222 þ 2 N24 N42 x8 X22 ðDÞ x8 þ x10
Fðx; DÞ ¼ jPði x; DÞj2 jQði x; DÞj2 ¼ X23 x2 ½X21 2 X3 X4 ðDÞ x4

þ fN224 N242 X24 ðDÞ þ 2 X1 X2 ðDÞg x6
þ fN222 þ 2 N24 N42 X22 ðDÞg x8 þ x10
We define the following parameters for simplicity: U0 ¼ 0; U2 ¼ X23
U4 ¼ ½X21 2 X3 X4 ðDÞ; U6 ¼ N224 N242 X24 ðDÞ þ 2 X1 X2 ðDÞ
U8 ¼ N222 þ 2 N24 N42 X22 ðDÞ; U10 ¼ 1. Hence Fðx; DÞ ¼ 0 implies

P
5
U2k x2k ¼ 0 and its roots are given by solving the polynomial.
k¼0
Furthermore
PR ði x; DÞ ¼ N22 x4 ; PI ði x; DÞ ¼ ½N24 N42 x3 þ x5

QR ði x; DÞ ¼ X1 x2 þ X2 ðDÞ x4 ; QI ði x; DÞ ¼ X3 x X4 ðDÞ x3
Hence
PR ði x; DÞ QI ði x; DÞ þ PI ði x; DÞ QR ði x; DÞ
sin hðDÞ ¼
jQði x; DÞj2
PR ði x; DÞ QR ði x; DÞ þ PI ði x; DÞ QI ði x; DÞ
cos hðDÞ ¼
jQði x; DÞj2
N22 x ½X3 x X4 ðDÞ x3 ½N24 N42 x3 þ x5 ½X1 x2 þ X2 ðDÞ x4
4
sin hðDÞ ¼ 2 2
X3 x þ ½X21 2 X3 X4 ðDÞ x4 þ ½X24 ðDÞ 2 X1 X2 ðDÞ x6 þ X22 ðDÞ x8
N22 x4 ½X1 x2 þ X2 ðDÞ x4 ½N24 N42 x3 þ x5 ½X3 x X4 ðDÞ x3
cos hðDÞ ¼
X23 x2 þ ½X21 2 X3 X4 ðDÞ x4 þ ½X24 ðDÞ 2 X1 X2 ðDÞ x6 þ X22 ðDÞ x8
P
Which jointly with Fðx; DÞ ¼ 0 ) 5k¼0 U2k x2k ¼ 0 that are continuous and
differentiable in Δ, based on Lema 1and Hence we use theorem 1. This proves the
theorem 2.
Lemma 1 Assume that ω(Δ) is a positive and real root of Fðx; DÞ ¼ 0 defined for
D 2 I, which is continuous and differentiable. Assume further that if k ¼ i x
x 2 R then Pn ðk ¼ i x; DÞ þ Qm ðk ¼ i x; DÞ 6¼ 0; D 2 R hold true. Then the
functions Sn ðDÞ; n 2 N0 are continuous and differentiable on I.
Theorem 1 Assume that ω(Δ) is a positive real root of Fðx; DÞ ¼ 0 defined for
D 2 I; IR þ 0 , and at some D 2 I, Sn ðD Þ ¼ 0. For some n 2 N0 then a pair of
simple conjugate pure imaginary roots k þ ðD Þ ¼ i xðD Þ; k ðD Þ ¼ i xðD Þ.
Dðk; DÞ ¼ 0 exist at D ¼ D which crosses the imaginary axis from left to right if
dðD Þ [ 0 and cross the imaginary axis from right to left if dðD Þ\0 where
dRek dSn ðDÞ

dðD Þ ¼ signf jk¼ixðD Þ g ¼ signfFx ðxðD Þ; D Þg signf j g:
dD dD D¼D
Theorem 2 The characteristic equation has a pair of simple and conjugate pure
imaginary roots k ¼ xðD Þ; xðD Þ real at D 2 I if Sn ðD Þ ¼ D Dn ðD Þ ¼ 0
for some n 2 N0 . If xðD Þ ¼ x þ ðD Þ this pair of simple conjugate pure imaginary
roots crosses the imaginary axis from left to right if d þ ðD Þ [ 0 and crosses
the imaginary axis from right to left if d þ ðD Þ\0 where d þ ðD Þ ¼
signfdRe k
dD jk¼ix þ ðD Þ g

dSn ðDÞ
d ðD Þ ¼ signfd Rek j
þ dD
g ¼ signf
k¼ix þ ðD Þ j g. If xðD Þ ¼ x ðD Þ these
dD D¼D
pair of simple conjugates pure imaginary roots cross the imaginary axis from left to
right if d ðD Þ [ 0 and crosses the imaginary axis from right to left if d ðD Þ\0
dSn ðDÞ
where d ðD Þ ¼ signfd Re k
dD jk¼ix ðD Þ g ¼ signf dD jD¼D g.
If x þ ðD Þ ¼
x ðD Þ ¼ xðD Þ then DðD Þ ¼ 0 and signfdRek j
g ¼ 0, the same is true
dD k¼ixðD Þ
when S0n ðD Þ ¼ 0. The following result can be useful in identifying values of Δ
where stability switches happened.
Our IMPATT amplifier homogenous system for i01 ; i03 ; i1 ; i3 ; iL1 leads to a char-
acteristic equation for the eigenvalue λ having the form (second case)
Pðk; DÞ þ Qðk; DÞ ekD ¼ 0. Dðs1 ¼ 0; D1 ¼ D2 ¼ D; kÞ ¼ DðD; kÞ and we use
P ðkDÞn
Taylor series approximation: ekD 1 n¼0 n! 1 k D the Maclaurin ser-
kD
ies is a Taylor series expansion of a e function about zero (0). We get the
following general characteristic equation DðD; kÞ under Taylor series
approximation:
ekD 1 k D.
Dðs1 ¼ 0; D1 ¼ D2 ¼ D; kÞ ¼ DðD; kÞ ¼ N24 N42 k3 þ N22 k4 k5

þ fðN24 N31 N42 N13 þ N14 N23 N31 N42 Þ
k þ ðN11 N24 N42 þ N12 N23 N31
Rload
þ N13 N31 N13 N22 N31 þ N14 N21 N42 Þ
Lf
Rload
k2 þ f N11 N22 N11 þ N12 N21 þ N13 N31 N13 N31
Lf
Rload Rload Rload
Dg k3 þ fN11 D N11 g k4 g ekD
Lf Lf Lf
We use different parameters terminology from our last characteristics param-

eters definition: pj ðDÞ ! aj ; qj ðDÞ ! cj ; n ¼ 5; m ¼ 4; n [ m.
X
5 X
4
Pn ðk; DÞ ! PðkÞ; Qm ðk; DÞ ! QðkÞ; PðkÞ ¼ aj k j ; QðkÞ ¼ cj k j
j¼0 j¼0
PðkÞ ¼ N24 N42 k3 þ N22 k4 k5

QðkÞ ¼ ðN24 N31 N42 N13 þ N14 N23 N31 N42 Þ k
þ ðN11 N24 N42 þ N12 N23 N31 þ N13 N31
Rload
N13 N22 N31 þ N14 N21 N42 Þ k2
Lf
Rload
þf N11 N22 N11 þ N12 N21 þ N13
Lf
Rload Rload Rload
N31 N13 N31 Dg k3 þ fN11 D N11 g k4
Lf Lf Lf
n; m 2 N0 ; n [ m and aj ; cj : R þ 0 ! R are continuous and differentiable function

of Δ such that a0 þ c0 ¼ 0. In the following “─” denotes complex conjugate. P(λ),
Q(λ) are analytic functions in λ and differentiable in Δ. The coefficients:
faj ðCf ; Lf ; Rf ; Cd ; Rd ; R1 ; L1 ; D; . . .Þ and cj ðCf ; Lf ; Rf ; Cd ; Rd ; R1 ; L1 ; D; . . .Þg 2 R

are dependent on IMPATT amplifier system’s Cf ; Lf ; Rf ; Cd ; Rd ; R1 ; L1 ; D; . . .
values.
We already got the following expressions: a0 ¼ 0; a1 ¼ 0; a5 ¼ 1
a2 ¼ 0; a3 ¼ N24 N42 ; a4 ¼ N22 ; c0 ¼ 0; c1

¼ N24 N31 N42 N13 þ N14 N23 N31 N42
c2 ¼ N11 N24 N42 þ N12 N23 N31 þ N13 N31
Rload
N13 N22 N31 þ N14 N21 N42
Lf
Rload
c3 ¼ N11 N22 N11 þ N12 N21 þ N13 N31 N13 N31
Lf
Rload Rload Rload
D; c4 ¼ N11 D N11
Lf Lf Lf
Unless strictly necessary, the designation of the variation arguments

ðCf ; Lf ; Rf ; Cd ; Rd ; R1 ; L1 ; D; . . .Þ will subsequently be omitted from P, Q, aj, cj . The
coefficients aj, cj are continuous, and differentiable functions of their arguments,
and direct substitution shows that a0 þ c0 ¼ 0. In our case λ = 0 is a root of
characteristic equation. Furthermore, P(λ), Q(λ) are analytic function of λ of the
analysis [5] can also be verified in the present case [6].
Remark: In our case Pn ðk ¼ 0; DÞ þ Qm ðk ¼ 0; DÞ ¼ p0 ðDÞ þ q0 ðDÞ ¼
a0 þ c0 ¼ 08 D 2 R þ 0 and λ = 0 is a characteristic root of Dðk; DÞ ¼ 0. It is
against general geometric criterion [BK]. But we inspect our analysis for λ 6¼ 0.
(a) If k ¼ i x; x 2 R then Pði xÞ þ Qði xÞ 6¼ 0, that is P and Q have no
common imaginary roots. This condition was verified numerically in the entire
ðCf ; Lf ; Rf ; Cd ; Rd ; R1 ; L1 ; D; . . .Þ domain interest.
(b) jQðkÞ=PðkÞj is bounded for jkj ! 1; Rek
0, No roots bifurcation from ∞.
ðN24 N31 N42 N13 þ N14 N23 N31 N42 Þ k

þ ðN11 N24 N42 þ N12 N23 N31 þ N13 N31
Rload
N13 N22 N31 þ N14 N21 N42 Þ k2
Lf
Rload Rload
þf N11 N22 N11 þ N12 N21 þ N13 N31 N13 N31 Dg k3
Lf Lf
Rload Rload
þ fN11 D N11 g k4
QðkÞ Lf Lf
j j¼j j
PðkÞ N24 N42 k3 þ N22 k4 k5
ðN24 N31 N42 N13 þ N14 N23 N31 N42 Þ

þ ðN11 N24 N42 þ N12 N23 N31 þ N13 N31
Rload
N13 N22 N31 þ N14 N21 N42 Þ k
Lf
Rload
þf N11 N22 N11 þ N12 N21 þ N13
Lf
Rload
N31 N13 N31 Dg k2
Lf
Rload Rload
þ fN11 D N11 g k3
QðkÞ Lf Lf
j j¼j j
PðkÞ N24 N42 k2 þ N22 k3 k4
FðxÞ ¼ jPði xÞj2 jQði xÞj2 :

Fðx; DÞ ¼ jPði x; DÞj2 jQði x; DÞj2
ðcÞ ¼ X23 x2 ½X21 2 X3 X4 ðDÞ x4
þ fN224 N242 X24 ðDÞ þ 2 X1 X2 ðDÞg x6
þ fN222 þ 2 N24 N42 X22 ðDÞg x8 þ x10
It has at most a finite number of zeros. Indeed, this is a polynomial in ω
(degree in ω10).
(d) Each positive root xðCf ; Lf ; Rf ; Cd ; Rd ; R1 ; L1 ; D; . . .Þ of FðxÞ ¼ 0 is contin-
uous and differentiable with respect to Cf ; Lf ; Rf ; Cd ; Rd ; R1 ; L1 ; D; . . . This
condition can only be assessed numerically.
Pði xÞ ¼ Pði xÞ:
And Qði xÞ ¼ Qði xÞ thus k ¼ i x; x [ 0 may be on eigenvalue of
characteristic equation. The analysis consists in identifying the roots of character-
istic equation situated on the imaginary axis of the complex λ-plane, whereby
increasing the parameters Cf ; Lf ; Rf ; Cd ; Rd ; R1 ; L1 ; D; . . . Reλ may, at the crossing,
0ðÞ 0ðÞ ðÞ
change its sign from (-) to (+), that is, from a stable focus E ðÞ ðI1 ; I3 ; I1 ;
dVP
ðÞ ðÞ ðÞ 1
ðÞ
I3 ; IL1 Þ ¼ ð0; 0; I1 ; CP dt
CP I1 ; IL1 Þ to an unstable one, or vice versa. This
1 2
respect to Cf ; Lf ; Rf ; Cd ; Rd ; R1 ; L1 ; D; . . . and system parameters.
@Rek
K1 ðRf Þ ¼ ð Þ ; Cf ; Lf ; Cd ; Rd ; R1 ; L1 ; D; . . . ¼ const
@Rf k¼ix
@Rek
K1 ðLf Þ ¼ ð Þ ; Cf ; Rf ; Cd ; Rd ; R1 ; L1 ; D; . . . ¼ const
@Lf k¼ix
@Rek
K1 ðCf Þ ¼ ð Þ ; Lf ; Rf ; Cd ; Rd ; R1 ; L1 ; D; . . . ¼ const
@Cf k¼ix
@Rek
K1 ðCd Þ ¼ ð Þ ; Lf ; Rf ; Cf ; Rd ; R1 ; L1 ; D; . . . ¼ const
@Cd k¼ix
@Rek
K1 ðRd Þ ¼ ð Þ ; Lf ; Rf ; Cf ; Cd ; R1 ; L1 ; D; . . . ¼ const
@Rd k¼ix
@Rek
K1 ðDÞ ¼ ð Þ ; Lf ; Rf ; Cf ; Cd ; R1 ; L1 ; Rd ; . . . ¼ const
@D k¼ix
For the second case s1 ¼ 0; D1 ¼ D2 ¼ D we got the following results:
PR ði x; DÞ ¼ N22 x4 ; PI ði x; DÞ ¼ ½N24 N42 x3 þ x5

QR ði x; DÞ ¼ X1 x2 þ X2 ðDÞ x4 ; QI ði x; DÞ ¼ X3 x X4 ðDÞ x3
U0 ¼ 0; U2 ¼ X23 ; U4 ¼ ½X21 2 X3 X4 ðDÞ;
U6 ¼ N224 N242 X24 ðDÞ þ 2 X1 X2 ðDÞ
U8 ¼ N222 þ 2 N24 N42 X22 ðDÞ; U10 ¼ 1;
X
5
Fðx; DÞ ¼ jPði x; DÞj2 jQði x; DÞj2 ¼ U2k x2k
k¼0
P5
Hence Fðx; DÞ ¼ 0 implies k¼0 U2k x
2k
¼ 0. When writing PðkÞ ¼
PR ðkÞ þ i PI ðkÞ and QðkÞ ¼ QR ðkÞ þ i QI ðkÞ, inserting k ¼ i x into IMPATT
amplifier system’s characteristic equation, ω must satisfy the following:
PR ði x; DÞ QI ði x; DÞ þ PI ði x; DÞ QR ði x; DÞ
sin x D ¼ gðxÞ ¼
jQði x; DÞj2
PR ði x; DÞ QR ði x; DÞ þ PI ði x; DÞ QI ði x; DÞ
cos x D ¼ hðxÞ ¼
jQði x; DÞj2
where jQði x; DÞj2 6¼ 0 in view of requirement (a) above, and ðg; hÞ 2 R.

Furthermore, it follows above sin x D and cos x D equations that, by squaring
and adding the sides, ω must be a positive root of FðxÞ ¼ jPði xÞj2
jQði xÞj2 ¼ 0.
Note that FðxÞ is dependent of Δ. Now it is important to notice that if D 62 I
(assume that IR þ 0 is the set where xðDÞ is a positive root of FðxÞ and for,
D 62 I,xðDÞ is not defined, then for all Δ in I xðDÞ satisfies that FðxÞ ¼ 0) . Then
There are no positive xðDÞ solutions for Fðx; DÞ ¼ 0, and we cannot have
stability switches. For any D 2 I, where xðDÞ is a positive solution of Fðx; DÞ ¼ 0,
we can define the angle hðDÞ 2 ½0; 2 p as the solution of sin hðDÞ ¼
; cos hðDÞ ¼ And the relation between the argument hðDÞ and xðDÞ D for
D 2 I must be xðDÞ D ¼ hðDÞ þ n 2 p 8 n 2 N0 . Hence we can define the maps
Dn : I ! R þ 0 given by Dn ðDÞ ¼ hðDÞxðDÞ þ n2p
; n 2 N0 ; D 2 I. Let as introduce the
functions I ! R; Sn ðDÞ ¼ D Dn ðDÞ; D 2 I; n 2 N0 that are continuous and dif-
ferentiable in Δ. In the following, the subscripts k; x; Lf ; Rf ; Cf ; Rd ; R1 ; L1 ; . . .
indicate the corresponding partial derivatives. Let us first concentrate on KðxÞ,
remember in kðLf ; Rf ; Cf ; Rd ; R1 ; L1 ; . . .Þ and xðLf ; Rf ; Cf ; Rd ; R1 ; L1 ; . . .Þ, and
keeping all parameters except (x) and Δ. The derivation closely follows that in
reference [BK]. Differentiating IMPATT amplifier system characteristic equation
PðkÞ þ QðkÞ ekD ¼ 0 with respect to specific parameter (x), and inverting the
derivative, for convenience, one calculates:
Remark:
x ¼ Lf ; Rf ; Cf ; Rd ; R1 ; L1 ; . . .; etc:;
@k 1 Pk ðk; xÞ Qðk; xÞ þ Qk ðk; xÞ Pðk; xÞ D Pðk; xÞ Qðk; xÞ

ð Þ ¼
where Pk ¼ @P
@k …etc., Substituting k ¼ i x and bearing i Pði xÞ ¼ Pði xÞ
Qði xÞ ¼ Qði xÞ; i Pk ði xÞ ¼ Px ði xÞ; i Qk ði xÞ ¼ Qx ði xÞ and that
on the surface jPði xÞj2 ¼ jQði xÞj2 , one obtain the following expression:
@k 1 i Px ði x; xÞ Pði x; xÞ þ i Qk ði x; xÞ Qðk; xÞ D jPði x; xÞj2

ð Þ jk¼ix ¼

Q ¼ QR þ i QI
Px ¼ PRx þ i PIx ; Qx ¼ QRx þ i QIx ; Px ¼ PRx þ i PIx ; Qx ¼ QRx þ i QIx ; P2

¼ P2R þ P2I
When (x) can be any IMPATT diode active circulator system’s parameter Lf, Rf,
Cf,… and time delay Δ etc.,. Where for convenience, we dropped the arguments
(i∙ω, x), and where Fx ¼ 2 ½ðPRx PR þ PIx PI Þ ðQRx QR þ QIx QI Þ
Fx ¼ 2 ½ðPRx PR þ PIx PI Þ ðQRx QR þ QIx QI Þ; xx ¼ Fx =Fx We define
U and V:

We choose our specific parameter as time delay x = Δ.
PR ¼ PR ði x; DÞ ¼ N22 x4 ; PI ¼PI ði x; DÞ ¼ ½N24 N42 x3 þ x5

QR ¼ QR ði x; DÞ ¼ X1 x2 þ X2 ðDÞ x4 ; QI ¼QI ði x; DÞ ¼ X3 x X4 ðDÞ x3
PRx ¼ 4 N22 x3 ; PIx ¼ ½N24 N42 3 x2 þ 5 x4 ;

QRx ¼ X1 2 x þ X2 ðDÞ 4 x3
QIx ¼ X3 X4 ðDÞ 3 x2 ; PRD ¼ 0; PID ¼ 0;
@X2 ðDÞ 4 @X4 ðDÞ 3
QRD ¼ x ; QID ¼ x
@D @D
@X2 ðDÞ Rload @X4 ðDÞ Rload
xD ¼ FD =Fx ; ¼ N11 ; ¼ N13 N31
@D Lf @D Lf
Rload Rload
QRD ¼ N11 x4 ; QID ¼ N13 N31 x3 ; FD ¼ 2 ðQRD QR þ QID QI Þ
Lf Lf
PRx PR ¼ 4 N222 x7 ; PIx PI ¼ ½N24 N42 3 þ 5 x2 ½N24 N42 þ x2 x5
QRx QR ¼ 2 ½X1 þ X2 ðDÞ 2 x2 ½X1 þ X2 ðDÞ x2 x3
QIx QI ¼ ½X3 X4 ðDÞ 3 x2 ½X3 X4 ðDÞ x2 x; PRD PR þ PID PI ¼ 0
PID PI ¼ 0; PRD PR ¼ 0; FD ¼ 2 ½ðPRD PR þ PID PI Þ ðQRD QR þ QID QI Þ
Rload
QRD QR ¼ N11 ðX1 þ X2 ðDÞ x2 Þ x6 ;
Lf
Rload
QID QI ¼ N13 N31 ½X3 X4 ðDÞ x2 x4
Lf
Rload
FD ¼ 2 f N11 ðX1 þ X2 ðDÞ x2 Þ x2
Lf
Rload
þ N13 N31 ½X3 X4 ðDÞ x2 g x4
Lf
PR PIx ¼ N22 ½N24 N42 3 þ 5 x2 x6 ;
PI PRx ¼ 4 ½N24 N42 þ x2 N22 x6
QR QIx ¼ ½X1 þ X2 ðDÞ x2 ½X3 X4 ðDÞ 3 x2 x2
QI QRx ¼ ½X3 X4 ðDÞ x2 ½X1 þ X2 ðDÞ 2 x2 2 x2
V ¼ ðPR PID PI PRD Þ ðQR QID QI QRD Þ; PR PID PI PRD ¼ 0

Rload
V ¼ ðQR QID QI QRD Þ ¼ ð½X1 x2 þ X2 ðDÞ x4 N13 N31 x3
Lf
Rload
½X3 x X4 ðDÞ x3 ½ N11 x4 Þ
Lf
Rload Rload
V ¼ ð½X1 þ X2 ðDÞ x2 N13 N31 x5 þ ½X3 X4 ðDÞ x2 N11 x5 Þ
Lf Lf
F(ω, Δ) = 0. Differentiating with respect to Δ and we get
@x @x FD
Fx þ FD ¼ 0; D 2 I ) ¼ ;
@D @D Fx
@Rek @x FD
K1 ðDÞ ¼ ð Þ ; ¼ xD ¼
@D k¼ix @D Fx
2 ½U þ D jPj2 þ i Fx @Rek
K1 ðDÞ ¼ Ref ; sign K1 ðDÞg ¼ signfð Þ g
2
FD þ i 2 ½V þ x jPj @D k¼ix
U@x
@D þ V
sign K1 ðDÞg ¼ signfFx g signfD @x
@D þ x þ jPj2
g. We shall presently
examine the possibility of stability transition (bifurcations) of our system, about the
dVP
0ðÞ 0ðÞ ðÞ ðÞ ðÞ ðÞ 1
ðÞ
equilibrium point E ðÞ ðI1 ; I3 ; I1 ; I3 ; IL1 Þ ¼ ð0; 0; I1 ; CP dt
CP I1 ; IL1 Þ as
1 2
a result of a variation of delay parameter Δ. The analysis consists in identifying the

complex λ-plane whereby increasing the delay parameter Δ, Reλ may at the
crossing, change its sign from – to +, that is, from a stable focus E(*) to an unstable
the partial derivatives with respect to Δ.
@Rek @Rek
K1 ðDÞ ¼ ð Þ ; K1 ðDÞ ¼ ð Þ ; Lf ; Rf ; Cf ; Rd ; R1 ; L1 ; . . .
@D k¼ix @D k¼ix
¼ const; x 2 R þ
Numerical analysis: We get the expression for Fðx; DÞ system parameters val-
ues. We find those ω, Δ values which fulfil Fðx; DÞ ¼ 0. We ignore negative,
complex, and imaginary values of ω for specific Δ values. D 2 ½0:001. . .10 and we
can express by 3D function Fðx; DÞ ¼ 0. We plot the stability switch diagram
based on different delay values of our system.
@Rek 2 ½U þ D jPj2 þ i Fx
K1 ðDÞ ¼ ð Þk¼ix ¼ Ref g
@D FD þ i 2 ½V þ x jPj2
@Rek 2 fFx ðV þ x P2 Þ FD ðU þ D P2 Þg
K1 ðDÞ ¼ ð Þk¼ix ¼
@D FD2 þ 4 ðV þ x P2 Þ2
The stability switch occur only on those delay values (Δ) which fit the equation:
D ¼ xh þþ ðDÞ
ðDÞ and h þ ðDÞ is the solution of sin hðDÞ ¼ ; cos hðDÞ ¼ . . . when x ¼
x þ ðDÞ if only ω+ is feasible. Additionally, when system’s parameters are known
and the stability switch due to various time delay values Δ is described in the
following expression: (Steven 1994).
signfK1 ðDÞg ¼ signfFx ðxðDÞ; DÞg

UðxðDÞÞ xD ðxðDÞÞ þ VðxðDÞÞ
signfD xD ðxðDÞÞ þ xðDÞ þ g
jPðxðDÞÞj2
Remark: We know Fðx; DÞ ¼ 0 implies it roots xi ðDÞ and finding those delays
values Δ which ωi is feasible. There are Δ values which ωi is complex or imaginary
number, then unable to analyze stability [6, 19, 32].
3.5 Multistage IMPATT Amplifier System Microstrip

Delayed in Time Stability Switching Analysis
In many applications there is a use of multistage IMPATT amplifier.

Multistage IMPATT amplifier is constructed from many single circulator/IMPATT
diode amplifier which are connected as a chain structure. Each circulator/IMPATT
diode amplifier gets the RF signal from the previous amplifier unit and after
amplification feeds the RF signal to the next amplifier unit. In that multistage
IMPATT amplifier structure we use two modes of operation as an amplifier. The
first is the negative resistance mode, where the input signal entering the IMPATT
through the circulator is amplified due to the negative resistance phenomena in the
IMPATT. The amplified signal passes out of the diode through the same port at
which the input signal entered, and because of the circulator, passes into the output
line. The second is the injection locked mode. The IMPATT is biases so that it is
oscillating all the time, but the frequency is locked to the input frequency and the
power leaving the IMPATT is at the same frequency as the input. If we compare the
two modes bandwidth, the negative resistance mode provides the optimal band-
width and the bandwidth of an injection locked amplifier is only few percent. The
efficiency of the injection locked mode is greater than the efficiency of negative
resistance mode. In multistage IMPATT amplifier, all the amplifier unit’s mode of
operation is negative resistance except the last amplifier unit (output stage) which
the mode of operation is injection locked mode.
Fig. 3.15 Multistage IMPATT amplifier circuit
Fig. 3.16 Multistage IMPATT amplifier circuit with microstrip delay lines
Remark: we neglect the microstrip time delay which is connected between each
stage circulator port 2 to IMPATT diode [70–72].
D1, D2, …, Dn are IMPATT diodes. Each single circulator/IMPATT diode
amplifier is connected to the next amplifier unit by microstrip line. We characterize
each microstrip segment as a delay line. We define our multistage IMPATT
amplifier with additional n delay lines, s1 ; . . .; sn which represent (n) microstrip
segments. We neglect the voltages on delay lines Vsi ! e 8 s1 ; . . .; sn ; 1 i n.
There is a delay in the current which flow through each microstrip delay line
IðtÞ ! Iðt si Þ 8 1 i n.
3.5 Multistage IMPATT Amplifier System … 375
The input current to circuit stage (i) is define as Iin(i)(t) and the output current to
circuit stage (i) is define as Iout(i)(t). The transfer function from circuit stage (i) input
current Iin(i)(t) to output current Iout(i)(t) is IoutðiÞ ðtÞ ¼ fi ðIinðiÞ ðtÞ; Di ; Li ; Ci ; VBi Þ
8 1 i n; IinðiÞ ðtÞ ¼ Ioutði1Þ ðt si1 Þ; IoutðiÞ ðtÞ

¼ fi ðIoutði1Þ ðt si1 Þ; Di ; Li ; Ci ; VBi Þ 8
2 i n:Ioutð1Þ ðtÞ ¼ f1 ðIinð1Þ ðtÞ; D1 ; L1 ; C1 ; VB1 Þ; Iinð1Þ ðtÞ ¼ gðXðtÞÞ:
Iinð2Þ ðtÞ ¼ Ioutð1Þ ðt s1 Þ; Iinð3Þ ðtÞ ¼ Ioutð2Þ ðt s2 Þ; Iinðn1Þ ðtÞ ¼ Ioutðn2Þ ðt sn2 Þ
IinðnÞ ðtÞ ¼ Ioutðn1Þ ðt sn1 Þ; IRL ðtÞ ¼ IoutðnÞ ðt sn Þ;
Ioutð1Þ ðtÞ ¼ f1 ðgðXðtÞÞ; D1 ; L1 ; C1 ; VB1 Þ:
Ioutð1Þ ðtÞ ¼ f1 ðIinð1Þ ðtÞ; D1 ; C1 ; L1 ; VB1 Þ; Iinð2Þ ðtÞ ¼ Ioutð1Þ ðt s1 Þ;
Ioutð2Þ ðtÞ ¼ f2 ðIinð2Þ ðtÞ; D2 ; C2 ; L2 ; VB2 Þ
Ioutð2Þ ðtÞ ¼ f2 ðIoutð1Þ ðt s1 Þ; D2 ; C2 ; L2 ; VB2 Þ; Iinð3Þ ðtÞ ¼ Ioutð2Þ ðt s2 Þ;
Ioutð3Þ ðtÞ ¼ f3 ðIinð3Þ ðtÞ; D3 ; C3 ; L3 ; VB3 Þ
Iinðn1Þ ðtÞ ¼ Ioutðn2Þ ðt sn2 Þ; Ioutðn1Þ ðtÞ ¼ fn1 ðIinðn1Þ ðtÞ; Dn1 ; Ln1 ; Cn1 ; VBn1 Þ
Ioutðn1Þ ðtÞ ¼ fn1 ðIoutðn2Þ ðt sn2 Þ; Dn1 ; Ln1 ; Cn1 ; VBn1 Þ;
IinðnÞ ðtÞ ¼ Ioutðn1Þ ðt sn1 Þ
IoutðnÞ ðtÞ ¼ fn ðIoutðn1Þ ðt sn1 Þ; Dn ; Ln ; Cn ; VBn Þ;

IoutðnÞ ðtÞ ¼ fn ðIinðnÞ ðtÞ; Dn ; Ln ; Cn ; VBn Þ
IRL ðtÞ ¼ IoutðnÞ ðt sn Þ; IRL ðtÞ ¼ fn ðIoutðn1Þ ðt sn1 sn Þ; Dn ; Ln ; Cn ; VBn Þ
IRL ðtÞ ¼ fn ðfn1 ðIoutðn2Þ ðt sn2 sn1 sn Þ;
Dn1 ; Ln1 ; Cn1 ; VBn1 Þ; Dn ; Ln ; Cn ; VBn Þ
Ioutðn1Þ ðt sn1 sn Þ ¼ fn1 ðIinðn1Þ ðt sn1 sn Þ;
Dn1 ; Ln1 ; Cn1 ; VBn1 Þ
Ioutðn1Þ ðt sn1 sn Þ ¼ fn1 ðIoutðn2Þ ðt sn2 sn1 sn Þ;
Ioutðn2Þ ðtÞ ¼ fn2 ðIinðn2Þ ðtÞ; Dn2 ; Ln2 ; Cn2 ; VBn2 Þ;

Iinðn2Þ ðtÞ ¼ Ioutðn3Þ ðt sn3 Þ
Ioutðn2Þ ðtÞ ¼ fn2 ðIoutðn3Þ ðt sn3 Þ; Dn2 ; Ln2 ; Cn2 ; VBn2 Þ
Ioutðn2Þ ðt sn2 sn1 sn Þ ¼ fn2 ðIoutðn3Þ ðt sn3 sn2 sn1 sn Þ;
X
n
Ioutðn2Þ ðt sn2 sn1 sn Þ ¼ fn2 ðIoutðn3Þ ðt sk Þ; Dn2 ; Ln2 ; Cn2 ; VBn2 Þ
k¼n3
IRL ðtÞ ¼ fn ðfn1 ðfn2 ðIoutðn3Þ ðt sn3 sn2 sn1 sn Þ; Dn2 ; Ln2 ; Cn2 ; VBn2 Þ;
X n
IRL ðtÞ ¼ fn ðfn1 ðfn2 ðIoutðn3Þ ðt sk Þ; Dn2 ; Ln2 ; Cn2 ; VBn2 Þ;
k¼n3
IRL ðtÞ ¼ fn ðfn1 ðfn2 ð. . .ðf1 ðIinð1Þ ðt s1 s2 . . . sn Þ; D1 ; L1 ; C1 ; VB1 Þ. . .Þ
X
n
IRL ðtÞ ¼ fn ðfn1 ðfn2 ð. . .ðf1 ðIinð1Þ ðt sk Þ; D1 ; L1 ; C1 ; VB1 Þ. . .ÞDn1 ;
k¼1
Ln1 ; Cn1 ; VBn1 Þ; Dn ; Ln ; Cn ; VBn Þ
IinðnkÞ ðtÞ ¼ Ioutðnðk þ 1ÞÞ ðt snðk þ 1Þ Þ 8 k ¼ 0; 1; 2; . . .; n 2

IinðnÞ ðtÞ ¼ Ioutðn1Þ ðt sn1 Þ; . . .; Iinð2Þ ðtÞ ¼ Ioutð1Þ ðt s1 Þ
Fig. 3.17 Multistage

IMPATT amplifier circuit
follow stages
Fig. 3.18 Multistage IMPATT amplifier circuit – n stages

Fig. 3.19 IMPATT amplifier circuit and equivalent circuit
Multistage IMPATT amplifier one stage current transfer function: The transfer
function from circuit stage (i) input current Iin(i)(t) to output current Iout(i)(t) is
IoutðiÞ ðtÞ ¼ fi ðIinðiÞ ðtÞ; Di ; Li ; Ci ; VBi Þ. Circuit stage active circulator can be described
by system path from RFin port (P1) to RFout port (P3). For simplicity we ignore
MESFET high frequency equivalent model and took it as cutoff element in our
system. The equivalent circuit for one stage current transfer function is present in
the below figure. We ignore the circulator signal path from P3 to P1 since our
amplifier RF signal is feed directly to the next IMPATT amplifier stage. We con-
sider fully matching between two follow amplifier stages (stages i − 1 and i), no
signal reflections.
Branch 1: ICC ¼ ILf ¼ IRf ¼ ICf ¼ IinðiÞ ðtÞ; Branch 3: ICC ¼ ILf ¼ IRf ¼ ICf ¼
IoutðiÞ ðtÞ
dVRD dVRD dVRD

RD ¼ Rc ; Rc [ 0; RD \0; jRD j ¼ Rc ; \0; ¼ RD ¼Rd ; j j ¼ Rc
dIRD dIRD dIRD
dVCf dVCC dIinðiÞ ðtÞ
IinðiÞ ðtÞ ¼ Cf ; IinðiÞ ðtÞ ¼ CC ; VRf ¼ IinðiÞ ðtÞ Rf ; VLf ¼ Lf
dt dt Z dt
dIL1 1
IinðiÞ ðtÞ ¼ IC1 þ IoutðiÞ ðtÞ; ICd ¼ IRd ¼ IIMPATT ; VL1 ¼ L1 ) I L1 ¼ VL1 dt
dt L1
Branch 1: VP1 ¼ VCf þ VLf þ VRf þ VCC þ VA ; Branch 3: VA ¼ VCf þ VLf þ VRf þ
VCC þ VP3
dVC1 dVCd
VA VB ¼ VC1 ; IC1 ¼ C1 ; ICd ¼ Cd ; IL1 þ IC1 ¼ IRd ; IC1 ¼ IRd IL1
dt dt
dVRd dVRd
¼ ¼ Rd ; IoutðiÞ ðtÞ ¼ IinðiÞ ðtÞ IC1 ;
dIRd dICd
IoutðiÞ ðtÞ ¼ IinðiÞ ðtÞ IC1 ¼ IinðiÞ ðtÞ IRd þ IL1
Z
dVCd 1
ICd ¼ Cd ) VCd ¼ ICd dt;
dt Cd
Z
1
IoutðiÞ ðtÞ ¼ IinðiÞ ðtÞ IRd þ ðVBi VB Þ dt
L1
Z
1
IoutðiÞ ðtÞ ¼ IinðiÞ ðtÞ IRd þ VL1 dt; VL1 ¼ VBi VB ;
L1
Z
1
VB ¼ VCd þ VRd ¼ ICd dt þ VRd
Cd
Z
dVC1 dðVA VB Þ 1
VA VB ¼ VC1 ; IC1 ¼ C1 ; IC1 ¼ C1 ) VA VB ¼ IC1 dt
dt dt C1
Z Z
1 1
VB ¼ VA IC1 dt; IoutðiÞ ðtÞ ¼ IinðiÞ ðtÞ IRd þ ðVBi VB Þ dt; VBi const
C1 L1
Z Z
1 1
IoutðiÞ ðtÞ ¼ IinðiÞ ðtÞ IRd þ VBi dt VB dt
L1 L
Z1
1 1
¼ IinðiÞ ðtÞ IRd þ VBi t VB dt
L1 L1
1 1
IoutðiÞ ðtÞ ¼ IinðiÞ ðtÞ IRd þ VBi t
L1 L1
Z Z
1
½VA IC1 dt dt; IC1 ðtÞ ¼ IinðiÞ ðtÞ IoutðiÞ ðtÞ
C1
VA ¼ VP1 ðVCf þ VLf þ VRf þ VCC Þ@Branch1 ;
Z
dVCf 1
IinðiÞ ðtÞ ¼ Cf ) V Cf ¼ IinðiÞ ðtÞ dt
dt Cf
Z Z
1 dIinðiÞ ðtÞ 1
VA ¼ VP1 ð IinðiÞ ðtÞ dt þ Lf þ IinðiÞ ðtÞ Rf þ IinðiÞ ðtÞ dtÞ
Cf dt CC
dVA dVP1 1 d 2 IinðiÞ ðtÞ dIinðiÞ ðtÞ 1
¼ ½ IinðiÞ ðtÞ þ Lf þ Rf þ IinðiÞ ðtÞ
dt dt Cf dt2 dt CC
dVA dVP1 1 1 d 2 IinðiÞ ðtÞ dIinðiÞ ðtÞ
¼ ½ð þ Þ IinðiÞ ðtÞ þ Lf þ Rf
dt dt CC Cf dt2 dt
We are interested to find out our IoutðiÞ ðtÞ expression.

Z ZZ
1 1 1
IoutðiÞ ðtÞ ¼ IinðiÞ ðtÞ IRd þ VBi t VA dt þ IC1 dt dt
L1 L1 L1 C1
Z
dIoutðiÞ ðtÞ dIinðiÞ ðtÞ dIRd 1 1 1
¼ þ VBi VA þ IC1 dt
dt dt dt L1 L1 L1 C1
Z
d dIoutðiÞ ðtÞ dIinðiÞ ðtÞ dIRd 1 1 1
f ¼ þ VBi VA þ IC1 dtg
dt dt dt dt L1 L1 L1 C1
d 2 IoutðiÞ ðtÞ d 2 IinðiÞ ðtÞ d 2 IRd 1 dVBi 1 dVA 1 dVBi

¼ 2 þ þ I C1 ; ¼0
dt 2 dt 2 dt L1 dt L1 dt L1 C1 dt
d 2 IoutðiÞ ðtÞ d 2 IinðiÞ ðtÞ d 2 IRd 1 dVA 1

ð Þ ¼ 2 þ ½IinðiÞ ðtÞ IoutðiÞ ðtÞ
dt 2 dt 2 dt L1 dt L1 C1
dVA dVP1 1 1 d 2 IinðiÞ ðtÞ dIinðiÞ ðtÞ

ðÞ ¼ ½ð þ Þ IinðiÞ ðtÞ þ Lf þ Rf
dt dt CC Cf dt2 dt
We consider the assumption which the IMPATT/Circulator amplifier input

voltage VP1 ¼ C þ nðtÞ; C is constant voltage and nðtÞ is RF signal in time.
dVP1 dC dnðtÞ dC dnðtÞ dnðtÞ dnðtÞ

¼ þ ; ! 0; ! e Or ! X0 sgn½
dVA dnðtÞ 1 1 d 2 IinðiÞ ðtÞ dIinðiÞ ðtÞ

¼ X0 sgn½ ½ð þ Þ IinðiÞ ðtÞ þ Lf þ Rf
dt dt CC Cf dt2 dt
dnðtÞ dnðtÞ dnðtÞ dnðtÞ dnðtÞ dnðtÞ dnðtÞ

¼ sgn½ j j; 8 real 9j j ¼ sgn½
dnðtÞ
dj j
dt ¼ sgn½dnðtÞ 8 dnðtÞ 6¼ 0; dVP1 ¼ 0 for dnðtÞ ¼ 0
dnðtÞ dt dt dt dt
d½
dt
dVP1 dnðtÞ dVP1 dnðtÞ

¼ X0 for [ 0; ¼ X0 for \0
dt dt dt dt
dVP1
Fig. 3.20 IMPATT amplifier circuit VP1 and dt functions of time
We neglect the IMPATT diode reciprocal negative resistance (dtd ½R1d ! e)

characteristic slope changes over time.
d 2 IR d d dIR d dIR dVRd dIRd 1 d 2 IR d 1 dVRd

2
¼ ½ d ¼ ½ d ; ¼ ; 2d ¼ ½
dt dt dt dt dVRd dt dVRd Rd dt dt Rd dt
d 2 IR d d 1 dVRd d 1 dVRd 1 d 2 VRd d 1
¼ ½ ¼ ½ þ ; ½ !e
dt2 dt Rd dt dt Rd dt Rd dt2 dt Rd
d 2 IR d 1 d 2 VRd 1 d dVRd dVRd d Rd d 2 IR
2
¼ 2
¼ ½ ; ¼ Rd ; ! e; 2 d ! e
dt Rd dt Rd dt dt dt dt dt
d 2 IoutðiÞ ðtÞ d 2 IinðiÞ ðtÞ 1 dVA 1

ð Þ ¼ þ ½IinðiÞ ðtÞ IoutðiÞ ðtÞ
dt2 dt2 L1 dt L1 C1
ðÞ ! ðÞ
d 2 IoutðiÞ ðtÞ d 2 IinðiÞ ðtÞ 1 dnðtÞ 1 1 d 2 IinðiÞ ðtÞ
¼ fX 0 sgn½ ½ð þ Þ I inðiÞ ðtÞ þ L f
dt2 dt2 L1 dt CC Cf dt2
dIinðiÞ ðtÞ 1
þ Rf g þ ½IinðiÞ ðtÞ IoutðiÞ ðtÞ
dt L1 C1
d 2 IoutðiÞ ðtÞ d 2 IinðiÞ ðtÞ X0 dnðtÞ 1 1 1 Lf d 2 IinðiÞ ðtÞ
¼ sgn½ þ ð þ Þ I inðiÞ ðtÞ þ
dt2 dt2 L1 dt L1 CC Cf L1 dt2
dIinðiÞ ðtÞ Rf 1 1
þ þ IinðiÞ ðtÞ IoutðiÞ ðtÞ
dt L1 L1 C1 L1 C1
d 2 IoutðiÞ ðtÞ 1 Lf d 2 IinðiÞ ðtÞ Rf dIinðiÞ ðtÞ
þ IoutðiÞ ðtÞ ¼ ð1 þ Þ þ
dt 2 L1 C1 L1 dt2 L1 dt
1 1 1 1 X0 dnðtÞ
þ ½ð þ Þ þ IinðiÞ ðtÞ sgn½
L1 CC Cf C1 L1 dt
dVP1 dnðtÞ
Case I: dt ¼ 0 for dt ¼0
d 2 IoutðiÞ ðtÞ 1 Lf d 2 IinðiÞ ðtÞ Rf dIinðiÞ ðtÞ 1

þ IoutðiÞ ðtÞ ¼ ð1 þ Þ þ þ
dt 2 L1 C1 L1 dt2 L1 dt L1
1 1 1
½ð þ Þ þ IinðiÞ ðtÞ
CC Cf C1
dVP1 dnðtÞ
Case II: dt ¼ X0 for dt [0

þ I outðiÞ ðtÞ ¼ ð1 þ Þ þ
dt2 L1 C1 L1 dt2 L1 dt
1 1 1 1 X0
þ ½ð þ Þ þ IinðiÞ ðtÞ
L1 CC Cf C1 L1
dVP1 dnðtÞ
Case III: dt ¼ X0 for dt \0

þ I outðiÞ ðtÞ ¼ ð1 þ Þ þ
dt2 L1 C1 L1 dt2 L1 dt
1 1 1 1 X0
þ ½ð þ Þ þ IinðiÞ ðtÞ þ
L1 CC Cf C1 L1
d 2 IinðiÞ ðtÞ dIinðiÞ ðtÞ
We define for simplicity a new function: w ¼ wð dt2 ; dt ; IinðiÞ ðtÞ; . . .Þ
Lf d 2 IinðiÞ ðtÞ Rf dIinðiÞ ðtÞ 1 1 1 1 X0

w ¼ ð1 þ Þ þ þ ½ð þ Þ þ IinðiÞ ðtÞ
L1 dt2 L1 dt L1 CC Cf C1 L1
dnðtÞ
sgn½
dt
d 2 IoutðiÞ ðtÞ 1 d 2 IinðiÞ ðtÞ dIinðiÞ ðtÞ

þ I outðiÞ ðtÞ ¼ wð ; ; IinðiÞ ðtÞ; . . .Þ
dt2 L1 C1 dt2 dt
Next it is reader exercise to find IoutðiÞ ðtÞ ¼ fi ðIinðiÞ ðtÞ; Di ; Li ; Ci ; VBi Þ function.
3.6 FET Combined Biasing and Matching Circuit

Stability Analysis
FET RF transistor is biased by using two power supplies, one for VDS and the other
for VGS. Another way to bias microwave FET is to use source resistor. The source
resistor has the advantage of providing feedback to stabilize the FET performance
and requires only one power supply. Once the transistor has been properly biased, it
must be matched to microstrip transmission line. By implementing biasing and
matching elements to RF microwave FET, each electrode must be simultaneously
connected to an RF circuit and a DC circuit, and the two circuits must not interfere.
The required isolation between the biasing and the matching circuit is done with RF
chocks (which pass the DC and block the RF) and coupling capacitors which pass
RF and block DC [33, 34, 62].
Cin—Input coupling capacitor. It allows the input microwave signal X(t) to enter
the transistor gate but prevents the input microstrip line from shorting out the gate
bias voltage.
Lin—We use RF chock because the gate must be connected to DC ground, but
the RF must not leak through this ground.
The RF FET source port is connected to RF ground through coupling capacitor
Cs which allows the FET source to be at RF ground. It allows the biasing source
resistor Rs to be used between the source and DC ground.
The FET drain port is connected to the drain resistor Rd through RF choke Ld,
which is connected to the positive supply voltage Vdd. The RF choke Ld presents
Fig. 3.21 FET combined biasing and matching circuit

3.6 FET Combined Biasing and Matching Circuit Stability Analysis 383
the microwave signal from being shorted out by the drain resistor Rd and the power
supply Vdd. Coupling capacitor Cout allows the RF microwave signal to pass into
the output microstrip line and the load Rload, but prevents the output microstrip line,
which is a DC ground from shorting out the drain voltage. We represent our circuit
microstrip elements as a delay lines Tau1, Tau2, Tau3, Tau4 (s1 ; s2 ; s3 ; s4 ).
We neglect the voltages on delay lines Vsi ! e 8 s1 ; . . .; s4 ; 1 i 4. There
is a delay in the current which flow through each microstrip delay line
IðtÞ ! Iðt si Þ 8 1 i 4. Next is to investigate how these delay line elements
influence our circuit performance and stability. Input and output matching circuits
can be π or T models. We use for our analysis the FET small signal equivalent
circuit (reduced version) and input and output T matching network. We give dif-
ferent name to RF chocke Ld, Ldd (Ld is the FET small signal equivalent circuit
drain inductance) and drain resistor Rd, Rdd (Rd is the FET small signal equivalent
circuit drain resistance). Ld ! Ldd ; Rd ! Rdd . Since we neglect the voltages on the
delay lines Vsi ! e 8 s1 ; . . .; s4 ; 1 i 4 the voltages on each delay line’s ports is
the same (consider it like a short contact). We write the FET small signal equivalent
circuit for our analysis:
dðVA2 VA3 Þ VA VA8

ICin1 ¼ Cin1 ; IRd ¼ 7 ; IRi ¼ ICgs
dt Rd
dðVA1 VA2 Þ
ICin ¼ Cin ; ICin1 ðtÞ ¼ ICin ðt s1 Þ; VA1 ¼ XðtÞ;
dt
dILin1 dðVA3 VA4 Þ dILin
VA3 ¼ Lin1 ; ICin2 ¼ Cin2 ; VA4 ¼ Lin
dt dt dt
dILg
VA4 VA5 ¼ VLg ¼ Lg ; ICin ðt s1 Þ ¼ ICin1 ðtÞ
dt
¼ ILin1 ðtÞ þ ICin2 ðtÞ; ICin2 ðt s2 Þ ¼ ILin ðtÞ þ ILg ðtÞ
dVA5
ICpgs ¼ Cpgs ; ILg ¼ IRg þ ICpgs ; ICgs ¼ IRi ;
dt
VA VA6 dðVA6 VA7 Þ
IRg ¼ 5 ; ICgd ¼ Cgd
Rg dt
dV
IRg ¼ ICgd þ ICgs ; V ¼ VA6 VA9 ; ICgs ¼ Cgs ;
dt
VA VA15
IRi ¼ 9 ; ICgd ¼ Id þ ICds þ IRds þ IRd
Ri
ICds ¼ Cds ; IRds ¼ 7 ;
dt Rds
VA VA16 dIL
IRd ¼ ILd þ ICpds ; IRs ¼ 15 ; VA16 ¼ VLS ¼ LS S
RS dt
dVA8 dILd
IRS ¼ ILS ; ICpds ¼ Cpds ; VLd ¼ VA8 VA10 ¼ Ld ;
dt dt
Vdd VA11
IRdd ¼ ILdd ; ¼ IRdd
Rdd
Fig. 3.22 FET combined biasing and matching equivalent circuit
dILdd
VA11 VA10 ¼ Ldd ; ILd ðt s3 Þ þ ILdd ðt s3 Þ
dt
dðVA10 VA12 Þ
¼ ICout1 ðtÞ ; ICout1 ¼ Cout1
dt
dILout1
VA12 ¼ Lout1 ;
dt
dðVA12 VA13 Þ
ICout1 ¼ ILout1 þ ICout2 ; ICout2 ¼ Cout2 ; ICout ðtÞ ¼ ICout2 ðt s4 Þ
dt
dðVA13 VA14 Þ VA
ICout ¼ IRload ; ICout ¼ Cout ; IRload ¼ 14 ; VO ¼ VA14
dt Rload
We write FET small signal equivalent circuit’s Kirchhoff’s current law (KCL)
for circuit’s node A2, A3,…,A16.
Table 3.11 FET combined biasing and matching equivalent circuit’s Kirchhoff’s Current Law
(KCL) and expressions
KCL @ Ax (x = 2,…,16) Expression of Kirchhoff’s current law
1 A2 ICin1 ðtÞ ¼ ICin ðt s1 Þ
2 A3 ICin ðt s1 Þ ¼ ICin1 ðtÞ ¼ ILin1 ðtÞ þ ICin2 ðtÞ
3 A4 ICin2 ðt s2 Þ ¼ ILin ðtÞ þ ILg ðtÞ
4 A5 ILg ¼ IRg þ ICpgs
5 A6 IRg ¼ ICgd þ ICgs
6 A7 ICgd ¼ Id þ ICds þ IRds þ IRd
7 A8 IRd ¼ ILd þ ICpds
8 A9 ICgs ¼ IRi
9 A10 ILd ðt s3 Þ þ ILdd ðt s3 Þ ¼ ICout1 ðtÞ
10 A11 IRdd ¼ ILdd
(continued)

KCL @ Ax (x = 2,…,16) Expression of Kirchhoff’s current law
11 A12 ICout1 ¼ ILout1 þ ICout2
12 A13 ICout ðtÞ ¼ ICout2 ðt s4 Þ
13 A14 ICout ¼ IRload
14 A15 IRS ¼ IRi þ Id þ ICds þ IRds
15 A16 IRS ¼ ILS
d d
ICin ¼ Cin ðVA1 VA2 Þ ¼ Cin ðXðtÞ VA2 Þ;
dt dt
d d dIL
ICin1 ¼ Cin1 ðVA2 VA3 Þ ¼ Cin1 ðVA2 Lin1 in1 Þ
dt dt dt
dXðtÞ dVA2 dVA2 d 2 ILin1
ICin ¼ Cin ð Þ; ICin1 ¼ Cin1 ð Lin1 Þ;
dt dt dt dt2
1 dXðtÞ dVA2
ICin ¼
Cin dt dt
dVA2 dXðtÞ 1 1 dVA2 d 2 ILin1
¼ ICin ; ICin1 ¼ Lin1
dt dt Cin Cin1 dt dt2
2
dVA2 1 d ILin1
) ¼ ICin1 þ Lin1
dt Cin1 dt2
dXðtÞ 1 1 d 2 ILin1
ICin ¼ ICin1 þ Lin1 ;
dt Cin Cin1 dt2
d d dIL dIL
ICin2 ¼ Cin2 ðVA3 VA4 Þ ¼ Cin2 ðLin1 in1 Lin in Þ
dt dt dt dt
d 2 ILin1 d 2 ILin
ICin2 ¼ Cin2 ðLin1 Lin 2 Þ;
dt2 dt Z
dILg dILin 1 dIL
VA4 VA5 ¼ Lg ) Lin ICpgs dt ¼ Lg g
dt dt Cpgs dt
d 2 ILin 1 d 2 I Lg d
Lin 2
I C pgs
¼ Lg 2
; ICds ¼ Cds ðVA7 VA15 Þ;
dt Cpgs dt dt
VA7 VA15
IRds ¼ ) IRds Rds ¼ VA7 VA15
Rds
Z
dIR 1 1
ICds ¼ Cds Rds ds ; VA5 ¼ ICpgs dt; VA8 ¼
dt Cpgs Cpds
Z Z
1
ICpds dt; VA6 ¼ ICpgs dt IRg Rg
Cpgs
Z
1 d 1
VA 7 ¼ I R d Rd þ ICpds dt; ICgd ¼ Cgd ½
Cpds dt Cpgs
Z Z
1
ICpgs dt IRg Rg IRd Rd ICpds dt
Cpds
1 1 dIR dIR
ICgd ¼ Cgd ½ ICpgs ICpds Rg g Rd d ;
Cpgs Cpds dt dt
dIRS
VA15 ¼ IRs Rs þ LS
Z dt
1 dIR 1 dIR
V¼ ICgs dt; Ri i ¼ ICpgs Rg g
Cgs dt Cpgs dt
dIRS d 2 IR 1
RS LS 2 S ICgs
dt dt Cgs
dICgs 1 dIR dIR d 2 IR 1
IRi ¼ ICgs ; Ri ¼ ICpgs Rg g RS S LS 2 S ICgs
dt Cpgs dt dt dt Cgs
Z Z
1 1 dIL
VA 8 ¼ ICpds dt; VA10 ¼ ICpds dt Ld d ; VA11 ¼ Vdd Rdd IRdd
Cpds Cpds dt
Z
1 dILd dILdd
Vdd Rdd IRdd ICpds dt þ Ld ¼ Ldd
Cpds dt dt
Z
d 1 dILd d dIL
fVdd Rdd IRdd ICpds dt þ Ld g ¼ fLdd dd g
dt Cpds dt dt dt
dVdd dIR 1 d 2 IL d 2 ILdd dVdd
Rdd dd ICpds þ Ld 2 d ¼ Ldd ; ¼0
dt dt Cpds dt dt2 dt
dIRdd 1 d 2 IL d 2 ILdd
Rdd ICpds þ Ld 2 d ¼ Ldd
dt Cpds dt dt2
Z Z
1 1 dIL
VA 8 ¼ ICpds dt; VA10 ¼ ICpds dt Ld d
Cpds Cpds dt
Z
d 1 dILd dILout1
ICout1 ¼ Cout1 ½ ICpds dt Ld Lout1
dt Cpds dt dt
1 d 2 IL d 2 ILout1
¼ Cout1 ½ ICpds Ld 2 d Lout1
Cpds dt dt2
Z
dIL 1
VA12 ¼ Lout1 out1 ; VA14 ¼ IRload Rload ; VA13 ¼ ICout dt þ IRload Rload
dt Cout
Z
d dILout1 1
ICout2 ¼ Cout2 ½Lout1 ICout dt IRload Rload
dt dt Cout
d 2 ILout1 1 dIR
¼ Cout2 ½Lout1 ICout Rload load
dt2 Cout dt
2
d ILout1 1 dIR
ICout2 ¼ ICout ¼ IRload ; ICout2 ¼ Cout2 ½Lout1 ICout Rload load
dt2 Cout dt
d 2 ILout1 1 dIRload
IRload ¼ Cout2 ½Lout1 IRload Rload
dt2 Cout dt
Our system equations with delays:
ICin1 ðtÞ ¼ ICin ðt s1 Þ ; ILd ðt s3 Þ þ ILdd ðt s3 Þ ¼ ICout1 ðtÞ

ICin ðt s1 Þ ¼ ICin1 ðtÞ ¼ ILin1 ðtÞ þ ICin2 ðtÞ;
ICin2 ðt s2 Þ ¼ ILin ðtÞ þ ILg ðtÞ ; ICout ðtÞ ¼ ICout2 ðt s4 Þ
dXðtÞ 1 1 d 2 ILin1 d 2 ILin1 d 2 ILin

ICin ¼ ICin1 þ Lin1 ; I C in2
¼ Cin2 ðL in1 L in Þ
dt Cin Cin1 dt2 dt2 dt2
d 2 IL 1 d 2 IL dIR
Lin 2 in ICpgs ¼ Lg 2 g ; ICds ¼ Cds Rds ds
dt Cpgs dt dt
1 1 dIRg dIRd
ICgd ¼ Cgd ½ ICpgs ICpds Rg Rd
Cpgs Cpds dt dt

Ri ¼ ICpgs Rg g RS S LS 2 S ICgs
Rdd ICpds þ Ld 2 d ¼ Ldd
dt Cpds dt dt2
1 d 2 IL d 2 ILout1
ICout1 ¼ Cout1 ½ ICpds Ld 2 d Lout1 ;
Cpds dt dt2
d 2 ILout1 1 dIR
IRload ¼ Cout2 ½Lout1 2
IRload Rload load
dt Cout dt
We implement delay variables in the above system differential equations:
dXðtÞ 1 1 d 2 ILin1
ICin ¼ ICin ðt s1 Þ þ Lin1 ; ICin2
dt Cin Cin1 dt2
d ½ICin ðt s1 Þ ICin2 ðtÞ
2
d 2 ILin
¼ Cin2 ðLin1 L in Þ
dt2 dt2
d 2 ILin 1 d 2 ½ICin2 ðt s2 Þ ILin ðtÞ dIR

Lin 2
ICpgs ¼ Lg 2
; ICds ¼ Cds Rds ds
dt Cpgs dt dt
1 1 dIR dIR
ICgd ¼ Cgd ½ ICpgs ICpds Rg g Rd d
Cpgs Cpds dt dt

Ri ¼ ICpgs Rg g RS S LS 2 S ICgs
Rdd ICpds þ Ld 2 d ¼ Ldd ; ICout ¼ IRload ; IRload ðtÞ¼ICout ðtÞ
dt Cpds dt dt2
¼ ICout2 ðt s4 Þ
1 d 2 IL d 2 ILout1
ILd ðt s3 Þ þ ILdd ðt s3 Þ ¼ Cout1 ½ ICpds Ld 2 d Lout1
Cpds dt dt2
d 2 ILout1 1 dICout2 ðt s4 Þ
ICout2 ðt s4 Þ ¼ Cout2 ½Lout1 ICout2 ðt s4 Þ Rload
dt2 Cout dt
To find equilibrium points (fixed points) of our system we define
lim ICin ðt s1 Þ ¼ ICin ðtÞ; lim ILd ðt s3 Þ ¼ ILd ðtÞ ; lim ILdd ðt s3 Þ ¼ ILdd ðtÞ
t!1 t!1 t!1
lim ICin2 ðt s2 Þ ¼ ICin2 ðtÞ; lim ICout2 ðt s4 Þ ¼ ICout2 ðtÞ
t!1 t!1
d 2 ILin1 d 2 ½ICin ðt s1 Þ ICin2 ðtÞ

2
¼ 0; ¼ 0;
dt dt2
d 2 ILin d 2 ½ICin2 ðt s2 Þ ILin ðtÞ dIR
2
¼ 0 ; 2
¼ 0; ds ¼ 0
dt dt dt
dIRd dICgs dIRg dIRS d 2 IR
¼ 0; Ri ¼ 0; ¼ 0; ¼ 0; 2 S ¼ 0;
dt dt dt dt dt
dIRdd d 2 I Ld d 2 ILdd
¼ 0; 2 ¼ 0; ¼0
dt dt dt2
d 2 I Ld d 2 ILout1 dIC ðt s4 Þ
2
¼ 0; ¼ 0; out2 ¼0
dt dt2 dt
We get our system fixed points: dXðtÞ

dt ¼ ½C1in1 þ Cin
1
IC in ; IC in2 ¼ 0; IC pgs ¼ 0;
IC ds ¼ 0
IC gd ¼ 0; IC pds ¼ 0; IC gd ¼ 0; IC gs ¼ 0; ILd þ ILdd ¼ 0; IC out2 ¼ 0
We consider microwave RF source XðtÞ ¼ A0 þ f ðtÞ; jf ðtÞj\1 & A0 jf ðtÞj

df ðtÞ
then XðtÞjA0 jf ðtÞj XðtÞjA0 jf ðtÞj ¼ A0 þ f ðtÞ A0 ; dXðtÞ
dt jA0 jf ðtÞj ¼ dt ! e.
Stability analysis:
The standard local stability analysis about any one of the equilibrium points of
the small signal equivalent circuit for FET consists in adding to coordinate
½ICin ; ILin1 ; ICin2 ; ILin ; ICpgs ,ICds ; IRds ; ICgd ; ICpds ; IRg ; IRd ,ICgs ; IRs ; IRdd ; ILd ; ILdd ; ILout1 ; ICout2
arbitrarily small increments of exponential form ½x; y; iL1 ; iRj ; iRs ekt and retaining
the first order terms in ICin ; ILin1 ; ICin2 ; ILin ; ICpgs ; ICds ; IRds ; ICgd ; . . . The system of ho-
values. The polynomial characteristic equations accept by set the below circuit
variables and circuit variables derivative with respect to time into equivalent circuit
for FET equations. FET circuit fixed values with arbitrarily small increments of
exponential form ½iCin ; iLin1 ; iCin2 ; iLin ; iCpgs ; iCds ; iRds ; iCgd ; . . . ekt are: j = 0 (first fixed
point), j = 1 (second fixed point), j = 2 (third fixed point), etc.
ðjÞ ðjÞ
ICin ðtÞ ¼ ICin þ iCin ekt ; ICin ðt s1 Þ ¼ ICin þ iCin ekðts1 Þ ;
ðjÞ dILin1 ðtÞ
ILin1 ðtÞ ¼ ILin1 þ iLin1 ekt ; ¼ iLin1 k ekt
dt
d 2 ILin1 ðtÞ dIC ðt s1 Þ
¼ iLin1 k2 ekt ; in ¼ iCin k ekðts1 Þ ;
dt2 dt
d 2 ICin ðt s1 Þ
¼ iCin k2 ekðts1 Þ
dt2
ðjÞ dIL ðtÞ d 2 ILin ðtÞ
ILin ðtÞ ¼ ILin þ iLin ekt ; in ¼ iLin k ekt ;
dt dt2
kt ðjÞ kt
¼ iLin k e ; ICin2 ðtÞ ¼ ICin2 þ iCin2 e
2
d 2 ICin2 ðtÞ ðjÞ

¼ iCin2 k2 ekt ; ICpgs ðtÞ ¼ ICpgs þ iCpgs ekt ;
dt2
ðjÞ
ICin2 ðt s2 Þ ¼ ICin2 þ iCin2 ekðts2 Þ
d 2 ICin2 ðt s2 Þ ðjÞ
¼ iCin2 k2 ekðts2 Þ ; ICds ¼ ICds þ iCds ekt ;
dt2
ðjÞ dIR
IRds ¼ IRds þ iRds ekt ; ds ¼ iRds k ekt
dt
ðjÞ ðjÞ
ICgd ðtÞ ¼ ICgd þ iCgd ekt ; ICpds ðtÞ ¼ ICpds þ iCpds ekt ;
ðjÞ dIRg ðtÞ
IRg ðtÞ ¼ IRg þ iRg ekt ; ¼ iRg k ekt
dt
ðjÞ dIRd ðtÞ

IRd ðtÞ ¼ IRd þ iRd ekt ; ¼ iRd k ekt ;
dt
ðjÞ dIC ðtÞ
ICgs ðtÞ ¼ ICgs þ iCgs ekt ; gs ¼ iCgs k ekt
dt
ðjÞ kt dIRs ðtÞ
IRs ðtÞ ¼ IRs þ iRs e ; ¼ iRs k ekt ;
dt
d 2 IRs ðtÞ ðjÞ
¼ iRs k2 ekt ; IRdd ðtÞ ¼ IRdd þ iRdd ekt
dt2
dIRdd ðtÞ ðjÞ d 2 ILd ðtÞ
¼ iRdd k ekt ; ILd ðtÞ ¼ ILd þ iLd ekt ; ¼ iLd
dt dt2
ðjÞ
k2 ekt ; ILdd ðtÞ ¼ ILdd þ iLdd ekt
d 2 ILdd ðtÞ ðjÞ
¼ iLdd k2 ekt ; ILd ðt s3 Þ ¼ ILd þ iLd ekðts3 Þ ;
dt2
ðjÞ
ILdd ðt s3 Þ ¼ ILdd þ iLdd ekðts3 Þ
ðjÞ d 2 ILout1 ðtÞ
ILout1 ðtÞ ¼ ILout1 þ iLout1 ekt ; ¼ iLout1 k2 ekt ;
dt2
ðjÞ
ICout2 ðt s4 Þ ¼ ICout2 þ iCout2 ekðts4 Þ
dICout2 ðt s4 Þ
¼ iCout2 k2 ekðts4 Þ
dt
By implementing the above delay equations, we get the following system

eigenvalues equations:
dXðtÞ 1 ðjÞ 1 ðjÞ

½ ! e ðI þ iCin ekt Þ ¼ ðI þ iCin ekðts1 Þ Þ þ Lin1 iLin1 k2 ekt ;
dt Cin Cin Cin1 Cin
ðjÞ
ICin2 þ iCin2 ekt ¼ Cin2 ðLin1 ½iCin eks1 iCin2 k2 ekt Lin iLin k2 ekt Þ
1 ðjÞ
Lin iLin k2 ekt ½I þ iCpgs ekt ¼ Lg ½iCin2 eks2 iLin k2 ekt
Cpgs Cpgs
ðjÞ
ICds þ iCds ekt ¼ Cds Rds iRds k ekt
ðjÞ 1 ðjÞ
ICgd þ iCgd ekt ¼ Cgd f ½I þ iCpgs ekt
Cpgs Cpgs
1 ðjÞ
½I þ iCpds ekt Rg iRg k ekt Rd iRd k ekt g
Cpds Cpds
1 ðjÞ
Ri iCgs k ekt ¼ ½I þ iCpgs ekt Rg iRg k ekt RS iRs k ekt
Cpgs Cpgs
1 ðjÞ
LS iRs k2 ekt ½I þ iCgs ekt
Cgs Cgs
1 ðjÞ
Rdd iRdd k ekt ½I þ iCpds ekt þ Ld iLd k2 ekt ¼ Ldd iLdd k2 ekt
Cpds Cpds
ðjÞ ðjÞ 1 ðjÞ

ILd þ iLd ekðts3 Þ þ ILdd þ iLdd ekðts3 Þ ¼ Cout1 ½ ðI þ iCpds ekt Þ
Cpds Cpds
Ld iLd k2 ekt Lout1 iLout1 k2 ekt
ðjÞ
ICout2 þ iCout2 ekðts4 Þ ¼ Cout2 ½Lout1 iLout1 k2 ekt
1 ðjÞ
ðI þ iCout2 ekðts4 Þ Þ Rload iCout2 k2 ekðts4 Þ
Cout Cout2
We implement our system fixed points values:
dXðtÞ ðjÞ ðjÞ ðjÞ ðjÞ

! e ) ICin ¼ IC in ¼ 0; ICin2 ¼IC in2 ¼ 0; ICpgs ¼IC pgs ¼ 0; ICds ¼IC ds ¼ 0
dt
ICpgs ¼ IC pgs ; ICgd ¼ IC gd ; IC gd ¼ 0; ICgs ¼IC gs ¼ 0; ICpds ¼IC pds ¼ 0
ðjÞ ðjÞ ðjÞ
ILdd ¼ ILdd ; ILd ¼ ILd ; ILd þ ILdd ¼ 0; ICout2 ¼IC out2 ¼ 0
System set of eigenvalues equations (eliminating ekt term):
1 1
iCin ¼ iCin eks1 þ Lin1 iLin1 k2 ; iCin2
Cin Cin1
¼ Cin2 ðLin1 ½iCin eks1 iCin2 Lin iLin Þ k2
1
Lin iLin k2 iCpgs ¼ Lg ½iCin2 eks2 iLin k2 ; iCds ¼ Cds Rds iRds k
Cpgs
1 1
iCgd ¼ Cgd iCpgs Cgd iCpds Cgd ½Rg iRg þ Rd iRd k
Cpgs Cpds
1 1
Ri iCgs k ¼ iCpgs Rg iRg k RS iRs k LS iRs k2 iCgs
Cpgs Cgs
1
Rdd iRdd k iCpds þ Ld iLd k2 ¼ Ldd iLdd k2
Cpds
1
iLd eks3 þ iLdd eks3 ¼ Cout1 ½ iCpds Ld iLd k2 Lout1 iLout1 k2
Cpds
1
iCout2 eks4 ¼ Cout2 ½Lout1 iLout1 k2 iCout2 eks4 Rload iCout2 k2 eks4
Cout
Remark: it a reader task to analyze system stability under delay parameters

variation based on eigenvalues equations.
Exercises
1. We have system of a bipolar transistor that activated at microwave frequency

and include feedback loop (Cfx ; Rfx ; RA ). RA is a variable resistor (3 ports) which
established the strength of our RF circuit feedback. We can divide our variable
resistor RA to two resistive sections: RA d; RA ð1 dÞ. Delta parameter (d) is
the feedback balance resistor parameter 0\d\1. Consider the base diagram of
bipolar transistor at microwave frequency circuit (Sect. 3.1) with additional
feedback loop which describe in the below figure.
1:1 Draw the circuit of bipolar transistor with feedback loop at microwave
frequencies equivalent circuit. Input Microwave RF source X(t). We con-
sider Microwave RFin signal X(t) = A0 + fX(t); |fX(t)| 1 and A0 |
dfX ðtÞ
fX(t)| then XðtÞ ¼ A0 þ fX ðtÞ A0 ) dXðtÞdt ¼ dt ! e . Find system dif-
ferential equations and fixed points.
1:2 Discuss stability analysis, linearization, and find system Jacobian elements
at fixed points. How Delta (d ; 0\d\1) feedback balance resistor param-
eter influences our system stability?
Exercises 393
1:3 Classify system stability fixed points according to eigenvalues variation.

How Delta (d ; 0\d\1) parameter influences our eigenvalues variation?
1:4 What happened if our feedback loop Delta (d ; 0\d\1) parameter is
constantly equal to one d ¼ 1 ? How our system behaviors change?
1:5 What happened if our feedback loop Delta (d; 0\d\1) parameter is con-
stantly equal to zero d ¼ 0 ? How our system behaviors change?
2. We have system of a bipolar transistor that activated at microwave frequency
and includes two feedback loops, first loop CfA ; RfA ; RA and second loop
CfB ; RfB ; RB . RA and RB are variable resistors (3 ports) which establish the
strength of our RF circuit feedback. We can divide our system circuit feedback
variable resistors to two sections respectively. dA RA ; ð1 dA Þ RA and
dB RB ; ð1 dB Þ RB . They dA ; dB are two feedback balance resistors parameters
0\dA \1; 0\dB \1; dA 6¼ dB .
Consider the base diagram of bipolar transistor at microwave frequency circuit
(Sect. 3.1) with two additional feedback loops which describe in the below
figure.
2:1 Draw system circuit bipolar transistor with two feedback loops equivalent
circuit at microwave frequencies. Input Microwave RF source X(t). We
consider Microwave RFin signal X(t) = A0 + fX(t); |fX(t)| 1 and A0 |
dfX ðtÞ
fX(t)| then XðtÞ ¼ A0 þ fX ðtÞ A0 ) dXðtÞ
dt ¼ dt ! e
Find system differential equations and fixed points.
2:2 Discuss stability analysis, linearization, and find Jacobian elements at fixed
points. How DeltaA (dA ; 0\dA \1) and DeltaB (dB ; 0\dB \1) feedback
balance resistor parameters influence our system stability? dA 6¼ dB
2:3 Classify system stability fixed points according to eigenvalues. How dA ; dB

parameters influence our system eigenvalues variation?
2:4 What happened if our feedback loop dA parameter is constantly equal to
zero (dA ¼ 0) or one (dA ¼ 1); dB ; 0\dB \1 ? How our system behavior
changes?
2:5 What happened if our feedback loop dB parameter is constantly equal to zero
(dB ¼ 0) or one (dB ¼ 1);dA ; 0\dA \1 ? How our system behavior change?
3. Consider RF FETs amplifier system which includes two RF FETs with
peripheral components. We consider Microwave RFin signal X(t) = A0 + fX(t); |
fX(t)| <=1 and A0 > > |fX(t)| then XðtÞ ¼ A0 þ fX ðtÞ A0 .
The FET equivalent circuits are for high frequency model and operation, taking
the node capacitors and other elements into account. If we switch to low fre-
quency small signal FET model, all capacitors in the above model disconnected
and all inductors are short.
3:1. Draw full small signal equivalent circuit for FETs system and write FETs
system differential Eqs.
3:2. Find FETs system fixed points, How our system fixed points change if
R3 = 0? How our system fixed points change if capacitor C1 is
disconnected?
3:3. Discuss RF FETs system standard local stability analysis about anyone of
the equilibrium points. How the stability changes if R3 = 0?
3:4. Classify RF FETs system fixed points and discuss bifurcation for different
values of R1, R2, R3 resistors.
Exercises 395
3:5. Resistor R3 is disconnected, How our RF FETs system dynamical behavior

changes?
3:6. Capacitor C1 is disconnected, How our system bifurcation behavior
changes?
4. We have a schematic of an RF amplifier using a JFET as the active element. The
configuration of the JFET amplifier is common gate. The circuit includes two
iron-core inductors L1, L2 (RF chockes). The iron core inductors block the high
frequency AC signals from getting to the DC power supply. +V is positive DC
voltage source and –V is negative DC voltage source. +V and –V are biasing
voltages to our FET circuit.
We consider RFin signal XðtÞ ¼ A0 þ fx ðtÞ; jfx ðtÞj\ ¼ 1 and A0 jfx ðtÞj then
XðtÞ ¼ A0 þ fx ðtÞjA0 jfx ðtÞj A0 ; jfx ðtÞj 1.
4:1 Draw full small signal equivalent circuit for FET RF amplifier and write
system differential equations.
4:2 Find JFET amplifier circuit fixed points, How our system fixed points
change if C4 is disconnected?
4:3 Discuss JFET amplifier standard local stability analysis about anyone of the
equilibrium points. How the stability changes if C3 is disconnected?
4:4 Classify JFET amplifier circuit fixed points and discuss bifurcation for
different values of L3, L4.
4:5 L3 is disconnected, How our JFET amplifier circuit behavior changes?
4:6 Capacitor C5 is disconnected, How our system bifurcation behavior
changes?
5. We have IMPATT amplifier system which is constructed from three ports active
circulator and two IMPATT diodes. The circulator is used to separate input and
output signals. We consider IMPATT diodes chip RF equivalent circuit which
can be represent as a series resistor and capacitor RD1, CD1 and RD2, CD2
respectively for the first IMPATT diode D1 and second IMPATT diode D2.
RD1 < 0, RD2 < 0. RD1 and RD2 are the terminal negative resistances of the
packaged diodes (D1 and D2). IMPATT diodes act as an amplifier in the neg-
ative differential resistance characteristics.
s1 is the time delay for ARB1 microstrip. s2 is the time delay for ARB2
microstrip. s3 is the time delay for ARB3 microstrip. Rload is the circuit load
resistance. V1, V2 are DC voltage sources (bias voltages to IMPATT diodes D1
and D2 respectively). V1 6¼ V2 is different biasing voltages. R1, R2 are parasitic
resistances of DC voltage sources. L1, CA—inductor and capacitor of Bias-T
circuit for D1. L2, CB—inductor and capacitor of Bias-T circuit for D2. ARB1,
ARB2, and ARB3 are circuit microstrip delay lines (s1 ; s2 ; s3 ). D1, D2—first and
second IMPATT diodes.
5:1 Draw system amplifier full equivalent circuit and find differential equations.
Find fixed points and consider amplifier system input voltage
VP1 ¼ C þ nðtÞ, Γ is constant voltage and ξ(t) is RF signal in time.
5:2 Write system differential equations in matrix representation. Discuss system
eigenvalues and related fixed points classification.
5:3 If we short inductor L1 in our amplifier system, How it influences system
stability?
5:4 If we short RA in our amplifier system, How it influences system stability?
Exercises 397
5:5 If we short diode D2, How it influences system differential equations, fixed
points and stability?
5:6 If we disconnect diode D1, How it influences system differential equations,
fixed points and stability?
Hint: The delay parameters s1 ; s2 and s3 are related to currents which flows
through micro strips ARB1, ARB2, and ARB3. We can consider additional delay
parameters D1 ; D2 ; D3 for the current derivatives which flows through micro strips
respectively.
6. We have two sets of multistage IMPATT amplifiers. The output of our system
is a summation of two set’s outputs. The first set of multistage IMPATT
amplifier has additional k1 delay lines s1 ; s3 ; s5 ; . . .; s2n1 for n n ¼ 1; 2; . . .; k1
which represents ð2 k1 1Þ IMPATT diode/active circulator circuit stages.
The delay lines for the first set are exist only on odd stages outputs. The second
set of multistage IMPATT amplifiers has additional k2 delay lines
s2 ; s4 ; s6 ; . . .; s2n ; n ¼ 1; 2; . . .; k2 which represents ð2 k2 Þ IMPATT
diode/active circulator circuit stages. The delay lines for the second set are exist
only on even stages outputs (k1 6¼ k2). The first set of multistage IMPATT
amplifiers is feed by microwave RF signal X1(t) and the second set of multi-
stage IMPATT amplifiers is feed by microwave RF signal X2(t); X1(t) 6¼ X2(t).
X1 ðtÞ ¼ C1 þ n1 ðtÞ X2 ðtÞ ¼ C2 þ n2 ðtÞ. Γ1, Γ2 are constant voltages (Γ1 6¼ Γ2)
and ξ1(t), ξ2(t) are RF signals in time ξ1(t) 6¼ ξ2(t).
Each system IMPATT equivalent circuit is constructed from IMPATT diode

and Bias-T circuit which includes DC voltage source (bias voltage to IMPATT
diode), parasitic resistance of DC voltage source, inductor and capacitor. We
neglect the microstrip time delay which is connected between each stage cir-
culator’s port 2 to IMPATT diode. We neglect the voltages on system delay
lines Vs2n1 ! e8n ¼ 1; 2; . . .; k1 and Vs2n ! e 8 n ¼ 1; 2; . . .; k2 . There is a
delay in the current that flow through each microstrip delay line in the two sets
of multistage IMPATT amplifiers. 1 ð2 n 1Þ ð2 k1 1Þ for the first set
and 1 2 n ð2 k2 Þ for the second set. The input current for the circuit stage
(i) is defined as Iin(i)(t) in the two sets respectively and the output current for
circuit stage (i) is defined as Iout(i)(t) in the two sets respectively. The transfer
define as IoutðiÞ ðtÞ ¼ . . .; IoutðiÞ ðtÞ ¼ g1i ðIinðiÞ ðtÞ; Di ; Li ; Ci ; VBi Þ for the first set
and IoutðiÞ ðtÞ ¼ ;IoutðiÞ ðtÞ ¼ g2i ðIinðiÞ ðtÞ; Di ; Li ; Ci ; VBi Þ for the second set
respectively. g1i ðIinðiÞ ðtÞ; Di ; Li ; Ci ; VBi Þ 6¼ g2i ðIinðiÞ ðtÞ; Di ; Li ; Ci ; VBi Þ
6:1 Find the expression for IRload(t) as a function of the two set’s outputs
functions.
6:2 Find the two functions for the first and second sets: g1i ðIinðiÞ ðtÞ; Di ; Li ;
Ci ; VBi Þ and g2i ðIinðiÞ ðtÞ; Di ; Li ; Ci ; VBi Þ
6:3 Find the expressions for IRload(t) if X1(t) = 0 or X2(t) = 0.
6:4 How our system behavior changes if each IMPATT/active circulator
amplifier stage in the two sets suffers from short Bias-T’s inductor or
shorted Bias-T’s capacitor?
6:5 How our system behavior changes if two set’s input RF signals are as
follow: X1 ðtÞ ¼ C þ nðtÞ; X2 ðtÞ ¼ a1 X1 ðtÞ þ a2 X12 ðtÞ. Γ is a constant
voltage. ξ(t) is a RF signal. a1, a2 are constants.
7. We have system of three sets of multistage IMPATT amplifiers. The RF
microwave signal input to the third set is a summation of two signals: first is the
output signal from the first set and second is the output signal from the second
set. The first set is constructed from multistage IMPATT amplifier which has a
delay line in each output stage. The number of IMPATT/active circulator stages
in the first set is k1 and the number of delay lines is k1 (s1 ; s2 ; . . .; sk1 ). k1 can be
odd or even number. The second set of multistage IMPATT amplifier has
additional k2 delay lines s1 ; s3 ; s5 ; . . .; s2n1 ; n ¼ 1; 2; . . .; k2 which represent
ð2 k2 1Þ IMPATT/ active circulator circuit stages. The delay lines for the
second set are exist only on odd stages outputs. The third set of multistage
IMPATT amplifier has additional k3 delay lines s2 ; s4 ; s6 ; . . .; s2n ; n ¼
1; 2; . . .; k3 which represent 2 k3 IMPATT/active circulator stages. The delay
lines for the third set are exist only on even stages outputs (k1 6¼ k2 6¼ k3 ). The
first set of multistage IMPATT amplifiers is feed by microwave RF signal
X1(t); and the second set of multistage IMPATT amplifier is feed by microwave
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
RF signal X2(t); X1(t) 6¼ X2(t). X1 ðtÞ ¼ C þ nðtÞ; X2 ðtÞ ¼ 1 þ X1 ðtÞ.
Γ is constant voltage and ξ(t) is RF signal in time. Each system IMPATT
equivalent circuit is constructed from IMPATT diode and Bias-T circuit which
includes DC voltage source (bias voltage to IMPATT diode), parasitic resis-
tance of the DC voltage source, inductor and capacitor. We consider IMPATT
diode acts as an amplifier. We neglect the microstrip time delay which is
connected between each stage circulator port 2 to IMPATT diode. We neglect
the voltages on system delay lines Vsn ! e 8 n ¼ 1; 2; . . .; k1 ; Vs2n1 ! e 8 n ¼
1; 2; . . .; k2 and Vs2n ! e 8 n ¼ 1; 2; . . .; k3 .
Exercises 399
There is a delay in the current that flow through each microstrip delay line in
the three sets of multistage IMPATT amplifier in our system.
8 1 n k1 first set ; 8 1 n k2 second set ; 8 1 n k3 third set
The input current for circuit stage (i) is defined as Iin(i)(t) in all sets and the
output current for circuit stage (i) is defined as Iout(t) in all sets. The transfer
function from circuit stage (i) input current Iin(i)(t) to output current Iout(i)(t) in
set m (m = 1,2,3 stand for set number). IoutðiÞ ðtÞ ¼ gmi ðIinðiÞ ðtÞ,Di ;
Li ; Ci ; VBi Þ 8 m ¼ 1; 2; 3 (set number).
gl1 i ðIinðiÞ ðtÞ; . . .Þ 6¼ gl2 i ðIinðiÞ ðtÞ; . . .Þ ; l1 6¼ l2 ; l1 ¼ 1; 2; 3 ; l2 ¼ 1; 2; 3
7:1 Find the expression for IRload(t) as a function of the three set’s output
functions.
7:2 Find the three functions (m = 1,2,3), gmi ðIinðiÞ ðtÞ; Di ; Li ; Ci ; VBi Þ.
7:3 Find the expression for IRload(t) if X1(t) = 0.
7:4 Find the expression for IRload(t) if X2(t) = 0.
7:5 How our system behavior changes if each IMPATT diode/active circulator
amplifier stage in the second set is suffered from short Bias-T’s inductor or
short Bias-T’s capacitor?
7:6 How our system behavior changes if each IMPATT diode/active circulator
amplifier stage in the third set is suffered from short Bias-T’s inductor or
short Bias-T’s capacitor?
8. We have system of two sets of multistage IMPATT amplifiers which are

connected through isolators to loads network (RA, RB, and R − Q). The RF
microwave signal input to the first set is X1(t) and for the second set is X2(t).
The first set of multistage IMPATT amplifier has additional k1 delay lines
s1 ; s3 ; s5 ; . . .; s2n1 for n n ¼ 1; 2; . . .; k1 which represents ð2 k1 1Þ
IMPATT diode/active circulator circuit stages. The delay lines for the first set
are exist only on odd stages outputs. The second set of multistage IMPATT
amplifiers has additional k2 delay lines s2 ; s4 ; s6 ; . . .; s2n ; n ¼ 1; 2; . . .; k2
which represents ð2 k2 Þ IMPATT diode/active circulator circuit stages. The
delay lines for the second set are exist only on even stages outputs (k1 6¼ k2).
X1(t) 6¼ X2(t). X1 ðtÞ ¼ C1 þ n21 ðtÞ ; X2 ðtÞ ¼ n2 ðtÞ C2 þ X1 ðtÞ. Γ1, Γ2 are
constant voltages (Γ1 6¼ Γ2) and ξ1(t), ξ2(t) are RF signals in time ξ1(t) 6¼ ξ2(t).
Consider that isolators A and B are ideal and transparent to RF signal in one
direction.
Each system IMPATT equivalent circuit is constructed from IMPATT diode

and Bias-T circuit which includes DC voltage source (bias voltage to IMPATT
diode), parasitic resistance of DC voltage source, inductor and capacitor. We
neglect the microstrip time delay which is connected between each stage cir-
culator’s port 2 to IMPATT diode. We neglect the voltages on system delay
lines Vs2n1 ! e8 n ¼ 1; 2; . . .; k1 and Vs2n ! e 8 n ¼ 1; 2; . . .; k2 . There is a
delay in the current that flow through each microstrip delay line in the two sets
of multistage IMPATT amplifiers. 1 ð2 n 1Þ ð2 k1 1Þ for the first set
and 1 2 n ð2 k2 Þ for the second set. The input current for the circuit stage
(i) is defined as Iin(i)(t) in the two sets respectively and the output current for
circuit stage (i) is defined as Iout(i)(t) in the two sets respectively. The transfer
define as IoutðiÞ ðtÞ ¼ . . . ; IoutðiÞ ðtÞ ¼ g1i ðIinðiÞ ðtÞ; Di ; Li ; Ci ; VBi Þ for the first set
and IoutðiÞ ðtÞ ¼ . . . ; IoutðiÞ ðtÞ ¼ g2i ðIinðiÞ ðtÞ; Di ; Li ; Ci ; VBi Þ for the second set
respectively. g1i ðIinðiÞ ðtÞ; Di ; Li ; Ci ; VBi Þ 6¼ g2i ðIinðiÞ ðtÞ; Di ; Li ; Ci ; VBi Þ
Exercises 401
8:1 Find the expression for IRA(t), IRB(t), IR-Q(t), as a function of the two set’s
outputs functions.
8:2 Find the two functions for the first and second sets: g1i ðIinðiÞ ðtÞ; Di ; Li ;
Ci ; VBi Þ and g2i ðIinðiÞ ðtÞ; Di ; Li ; Ci ; VBi Þ
8:3 Find the expressions for IR-Q(t)if X1(t) = 0 or X2(t) = 0.
8:4 How our system behavior changes if each IMPATT/active circulator
amplifier stage in the two sets suffers from short Bias-T’s inductor or
shorted Bias-T’s capacitor and additionally RA is disconnected?
8:5 How our system behavior changes if two set’s input RF signals are as
follow: X1 ðtÞ ¼ C þ nðtÞ; X2 ðtÞ ¼ a1 X1 ðtÞ þ a2 X12 ðtÞ. Γ is a constant
voltage. ξ(t) is a RF signal. a1, a2 are constants and RB is disconnected?
9. We have a system of two sets of multistage IMPATT amplifier which feed load
resistance RL. Each circulator/IMPATT diode amplifier stage gets RF signal
from the previous amplifier unit and after amplification feeds the RF signal to
the next amplifier unit. Input RF microwave signal to the first set is defined as
X1(t) and to the second set X2(t). The function expression which characterize
our load resistance current in time is as follow:
X
n
IRL ¼ fn ðfn1 ðfn2 ð. . .ðf1 ðIinð1Þ ðt sk Þ; D1 ; L1 ; C1 ; VB1 Þ. . .Þ; Dn1 ; Ln1 ; Cn1 ; VBn1 Þ;
k¼3
Dn ; Ln ; Cn ; VBn Þ þ g2 ðg1 ðIinð1Þ ðt s1 s2 Þ; D1 ; L1 ; C1 ; VB1 Þ; D2 ; L2 ; C2 ; VB2 Þ
The first set is characterized by the f functions and the second set is charac-
terized by the g function.
First set: sk ¼ jsk1 ð1Þk þ sk2 ð1Þk þ 1 j 8 n
k
3 is recursive function
of micro strip delay line parameter in the first set.
s3 ¼ js1 s2 j ; s4 ¼ js3 s2 j ; s5 ¼ js3 s4 j ; s6 ¼ js5 s4 j. . .
Second set:
0
s1 [ 0; s2 [ 0; Ioutð2Þ ðtÞ ¼ Ioutð2Þ ðt s2 Þ; Iinð1Þ ðtÞ ¼ n2 ðX2 ðtÞÞ
Ioutð2Þ ðtÞ ¼ g2 ðIoutð1Þ ðt s1 Þ; D2 ; L2 ; C2 ; VB2 Þ; Iinð2Þ ðtÞ ¼ Ioutð1Þ ðt s1 Þ
0
Ioutð1Þ ðtÞ ¼ g1 ðIinð1Þ ðtÞ; D1 ; L1 ; C1 ; VB1 Þ; Ioutð2Þ ðtÞ ¼ g2 ðIoutð1Þ ðt s1 s2 Þ; D2 ; L2 ; C2 ; VB2 Þ
0
Ioutð2Þ ðtÞ ¼ g2 ðg1 ðIinð1Þ ðt s1 s2 Þ; D1 ; L1 ; C1 ; VB1 Þ; D2 ; L2 ; C2 ; VB2 Þ:
0
The system’s summation operator inputs are Ioutð2Þ ðtÞ and fn ðfn1 ðfn2 ð. . .Þ
9:1 Draw our two sets amplifier system.

9:2 How our system dynamic change for sk1 [ sk2 and for sk1 \sk2 ?
(First set micro strip delay lines parameters, n ≥ k ≥ 3).
9:3 The system’s first set of IMPATT multistage amplifiers has micro strip
delay lines only on odd places of IMPATT diode/active circulator stages.
Write the related functions for IRL(t) and draw our system. How the
dynamic of the system changes?
9:4 System’s first set stages is suffered from Bias-T’s short capacitor.
How our system functionality changes? Find the related f functions for that
case.
9:5 We change the direction polarity of each IMPATT diode in the first set.
How it influences our system functionality. Find the expression for IRL(t)
and transformation function in each multistage amplifier in the first set.
9:6 The recursive function of micro strip delay lines in the first set is as follow:
sk ¼ jsk1 ð1Þk þ jsk2 ð1Þk þ 1 sk3 ð1Þk jj 8 n

k
4; s3
0
g1 ðs; kÞ ¼ sk2 ð1Þk þ 1 sk3 ð1Þk ;
g2 ðs; kÞ ¼ sk1 ð1Þk ; sk ¼ jg2 ðs; kÞ þ jg1 ðs; kÞjj
How our system dynamic changes? Write all possible options for
s3 ; s4 ; s5 ; s6 ; . . .; sn . Analyze the IRL(t) expression for that case.
10. We have a system of RF FET transistor combined biasing and matching circuit.
The RF FET transistor is N channel JFET and the input and output matching
circuits are Pi type. We represent our circuit microstrip elements as a delay lines τ1,
…,τ6 (τk+1 = τk + Γ τk−1) for C 2 ½1; . . .; 10 ; k ¼ 2; . . .; 5 and τ1 = τ, τ2 = τ2.
We neglect the voltages on the delay lines Vsi ! e ; s1 ; . . .; s6 ; 1 i 6. There
is a delay in the current which flows through each microstrip delay line
IðtÞ ! Iðt si Þ 8 1 i 6. We use for our analysis the FET small signal
equivalent circuit (reduced version). The RF choke Ldd presents the microwave
signal from being shorted out by the drain resistor Rdd through RF choke Ldd,
which is connected to the positive supply voltage Vdd. Input and output capac-
itors are Cin and Cout. X(t) is a input microwave RF source.
Exercises 403
10:1 Draw our system complete circuit which includes small signal equivalent
circuit for FET.
10:2 Write system represented differential equations with delay variables in
time (τ1,…,τ6 delay parameters).
10:3 Find our system small increments Jacobian and characteristic Eq.
10:4 Discuss stability switching under delay parameter (τ) variation and Γ
parameter variation.
10:5 How our system dynamic changes if sk þ 1 ¼ sk C þ Csk sk1 sk þ 1 ¼
sk C þ Csk sk1 8 C 2 ½1; . . .; 10 ; k ¼ 2; . . .; 5. Discuss stability
Switching under delay parameter (τ) variation and Γ parameter variation.
10:6 Resistor Rs is disconnected. How our system dynamic changes? Discuss
stability switching under delay parameter (τ) variation and Γ parameter
variation.
Chapter 4
Small Signal (SS) Amplifiers and Matching
Network Stability Analysis
An amplifier is an active device that has the ability to amplify voltage, amplify
current or amplify both voltage and current. There are some types of amplifiers.
Amplifiers types: zero frequency amplifiers (DC amplifiers), low frequency
amplifiers (audio amplifiers), and high frequency amplifiers (RF amplifiers).
Power is P = VI, when current (I) or voltage (V) is raising then will create power
amplification. Amplifiers come in three basic flavors: Common Base (CB)
amplifiers, Common Collector (CC) amplifiers, and Common Emitters (CE)
amplifiers. It depends whether the base, collector, or emitter is common to both
the input and output of the amplifier. Common Base (CB) amplifier, input signal
inserted at emitter (E) and output signal taken from the collector (C). The CB
amplifier can operate as a voltage amplifier for low input impedance circuits. The
most popular amplifier circuit is Common Emitter (CE). The CE amplifier has a
greater current gain and voltage gain combination than any other type. CE
amplifier makes excellent power amplifiers. Common Emitter (CE) amplifier,
input signal inserted at the base (B) and output signal taken from the collector
(C). The CE amplifier’s output voltage is shifted by *180° in phase compared to
CE amplifier’s input signal. At RF frequencies there is an effect of “positive
feedback”, which creates amplifier instability and oscillation. “Positive feed-
back” is when there is internal feedback capacitance between the transistor’s
collector and its base. It cause to undesired CE oscillations. At a specific fre-
quency this capacitance will send an in phase signal back into the base input from
the collector’s output. This back in phase signal creates an “oscillator”. The
Common Collector (CC) amplifier is emitter follower amplifier. The CC
amplifier has the input signal inserted into the base and output signal from the
emitter. The CC amplifier has current and power gain, voltage gain less than one.
CC amplifier’s common use buffer or active impedance matching circuit. The CC

DOI 10.1007/978-3-319-45427-6_4
406 4 Small Signal (SS) Amplifiers and Matching Network …
amplifier has high input impedance and low output impedance. There is no phase
inversion between CC amplifier’s input and output. When an amplifier’s output
impedance matches the load impedance maximum power is transferred to the
load and all reflections are eliminated. When an amplifier’s output impedance
unmatched the load impedance there are reflections and less than maximum
power is transferred to the load. There are instabilities behaviors in these three
types of amplifiers causes by circuit micro-strip delays in time parasitic effects.
We use RF matching network which able and facilitate impedance matching and
filtering of signal, coupling between RF stages. There are typical amplifiers
matching networks: L matching network, T matching network, and PI matching
network. In design of microwave matching network, device parasitic effects of
length on RF circuit matching and stability. Many RF circuits contain Bias-T
three ports network. The function of the Bias-T is to simultaneously allow a DC
bias voltage and RF test signal to be applied to the port of a transistor during
measurement and operation. Bias-T three ports network suffers from instability
under delayed micro-strip in time. The passive filter with Bias-T suffers from
instability under parameters variation. Many RF circuits include a PIN diode.
A PIN diode is a diode with a wide, lightly doped ‘near’ intrinsic semiconductor
region between a p-type semiconductor and a n-type semiconductor region.
A PIN diode suitable for many applications: Attenuators, fast switches, photo
detectors, and high voltage power electronics applications. A PIN diode operates
under high level injection. The PIN diode suffers from instability under param-
eters variations.
4.1 Small Signal (SS) Amplifiers and Matching Network
Amplifier can be categorized in two manner, first according to signal level (small
signal amplifier, power/large signal amplifier) and second according to DC biasing
scheme of the active component (Classes A, B, AB, C). There are also other classes,
such as class D (D stands for digital), class E and class F. These all uses the
transistor/FET as a switch [24–26]. Input and output voltage relation of the
P3
amplifier can be modeled simply as: Vout ðtÞ ¼ ak vkin þ H:O:T; Vout ðtÞ ¼
k¼1
a1 vi ðtÞ þ a2 v2i ðtÞ þ a3 v3i ðtÞ þ H:O:T. The general block diagram is described
below (Fig. 4.1):
If we compare small signal versus Large-signal operation amplifier output
voltage expression: Large signal Vout ðtÞ ¼ a1 vi ðtÞ þ a2 v2i ðtÞ þ a3 v3i ðtÞ þ
P3
H:O:T;Vout ðtÞ ¼ ak vkin þ H:O:T. We get usually non-sinusoidal waveform
k¼1
4.1 Small Signal (SS) Amplifiers … 407
Fig. 4.1 Small Signal (SS) amplifier and matching network
amplifier output signal. Small signal, output voltage expression is linear Vout ðtÞ ’
a1 vi ðtÞ and the output signal is sinusoidal waveform. All amplifiers are inherently
nonlinear. However when the input signal is small, the input and output relationship
of the amplifier is approximately linear.
Vout ðtÞ ¼ a1 vi ðtÞ þ a2 v2i ðtÞ þ a3 v3i ðtÞ þ H:O:T a1 vi ðtÞ. When vi ðtÞ ! e
ð\2:6 mVÞ Then Vout ðtÞ a1 vi ðtÞ (Fig. 4.2).
The active component can be BJT transistor (Fig. 4.3):
The linear relationship applies also to current and power. An amplifier that
fulfills these conditions: Small signal operation and linearity is called Small Signal
Amplifier (SSA). If a SSA amplifier contains BJT and FET, these components can
Fig. 4.2 Small Signal (SS) amplifier and Zs, Zload

Fig. 4.3 Small Signal (SS) amplifier BJT transistor active component
be replaced by their respective small signal model, for instance the Hybrid Pi model
of BJT. To determine the performance of an amplifier, the following characteristics
are typically observed: power gain, bandwidth (operation frequency range), noise
figure, phase response, gain compression, dynamic range, harmonic distortion,
intermodulation distortion and Third Order Intercept point (TOI). The character-
istics which are important to small signal amplifier are bandwidth, noise figure, and
phase response. The important parameters of Large-signal amplifier (Related to
linearity) are gain compression, dynamic range, harmonic distortion, intermodula-
tion distortion, and Third Order Intercept point (TOI).
Typical RF amplifier schematics (Fig. 4.4):
Under AC and Small Signal (SS) conditions, the BJT can be replaced with linear
Hybrid Pi model (Fig. 4.5).
At low frequencies it is assumed that the transistor responds instantly to charges
of input voltage or current but actually because the mechanism of the transport of
charge carriers from emitter to collector is one of diffusion. The transistor behavior at
high frequencies is inspected by examine this diffusion mechanism in more details.
The Hybrid Pi model gives a reasonable compromise between accuracy and sim-
plicity. Using this model, a detailed analysis of a single stage CE transistor amplifier
is made. The Common Emitter (CE) is the most important practical configuration.
Fig. 4.4 Typical RF amplifier schematics
Fig. 4.5 AC Small Signal (SS) amplifier Hybrid Pi model
The CE model is valid at high frequencies. The circuit called the Hybrid Pi,
Giacoletto or model. The analyses of circuits using this model are not too difficult
and give results which are excellent with experiment at all frequencies for which the
transistor gives reasonable amplification. All parameters (resistances and capaci-

tances, inductances) in the model are assumed to be independent of frequency. They
may vary with the quiescent operation point, but under given bias conditions are
reasonably constant for small signal swings [92,93].
Remark: like a PN diode, we can break the BJT up into a large signal analysis
and small signal analysis and “linearize” the non-linear behavior of the Ebers-Moll
model. Small signal models are only useful for forward active mode and thus, are
derived under this condition. Saturation and cutoff are used for switches which
involve very large voltage/current swings from the on to off states. Small signal
models are used to determine amplifier characteristics. When a diode voltage
exceeds a certain value, the non-linear behavior of the diode leads to distortion of
the current/voltage curves, and if the BJT transistor inputs/outputs exceed certain
limits, the full Ebers-Moll model must be used.
For amplifiers functioning at RF and microwave frequencies, usually of interest
is the input and output power relation. The ratio of output power over input power
is called power gain (G), usually expressed in dB. Power gain:
output power
G ¼ 10 log 10 ð input power Þ dB. There are a number of definitions for power
gain. Furthermore G is a function of frequency and the input signal level. Power
gain is preferred for high frequency amplifiers as the impedance encountered is
usually low (due to presence of parasitic capacitance).
Power = Voltage Current. If the amplifier is required to drive 50 Ohm load
the voltage across the load may be small, although the corresponding current may
be large (there is a current gain). For amplifiers functioning at lower frequency
(such as IF frequency), it is the voltage gain that is of interest, since impedance
encounter is usually higher (less parasitic). If the output of an IF amplifier drives the
modulator circuits, which are usually digital systems, the impedance looking into
the digital system is high and large voltage can be developed across it. Thus
working with voltage gain is more convenient. When the input driving signal is
small, the amplifier is linear and harmonic components are almost non-existent.
When the input driving signal is too large the amplifier becomes nonlinear.
Harmonics are introduced at the output. Harmonic generation reduces the gain of
the amplifier, as some of the output power at the fundamental frequency is shifted to
higher harmonics. This result is in gain compression. The amplifier also introduces
noise into the output in addition to the noise from the environment.
Phase consideration is important for amplifier working with wideband signals.
For signal to be amplified with no distortion, two requirements are needed from
linear systems theory. First the magnitude of the power gain transfer function must
be a constant with respect to frequency (f). Second the phase of the power gain
transfer function must be a linear function of (f). A linear phase produces a constant
time delay for all signal frequencies, and a linear phase shift produces different time
delay for different frequencies. Property means that all frequency components will
be amplified by similar amount and implies that all frequency components will be
delayed by similar amount. Essence of Small Signal Amplifier (SSA) design: In
essence, designing a small signal amplifier with transistor or Monolithic Microwave
Integrated Circuit (MMIC) implies finding the suitable load and source impedance
to be connected to the output and input port, and getting the required transducer
power gain GT, bandwidth and other characteristics. An amplifier is a circuit
designed to enlarged electrical signals. When there is no input, there should be no
output; this condition is known as stable. On the contrary, if the amplifier produces
an output when there is no input, it is unstable. In fact the amplifier becomes an
oscillator. Thus a stability analysis is required to determine whether an amplifier
circuit is stable or not. Stability analysis is also carried out by assuming a
small-signal amplifier, since the initial signal that causes oscillation is always very
small. Stability of an amplifier is affected by the load and source impedance con-
nected to its two ports. An unstable or marginally stable amplifier can be made
more stable. When amplifier is unstable, or stable region is too small there are some
steps which need to be done: use negative feedback to reduce amplifier gain,
redesign DC biasing, finding new operating point (or Q point) that will result in
more stable amplifier, add some resistive loss to the circuit to improve stability, and
use a new component with better stability.
When an amplifier’s output impedance matches the load impedance, maximum
power is transferred to the load. When amplifier’s output impedance matches the
load impedance all reflections are eliminated. When an amplifier’s output impe-
dance unmatched the load impedance there are reflections and less than maximum
power is transferred to the load. In order to develop maximum power the Zout of the
amplifier must be complex conjugate of the Zin of the load. Amplifier matching
allows to amplifier maximum power transfer and attenuation of harmonics to be
achieved between stages. RF matching network able to facilitates impedance
matching, filtering of signal and coupling between RF stages. The amplifier
matching networks types are L matching network, T matching network, and PI
matching network.
L matchingnetwork: LC matching topology is especially for narrowband impe-
dance matching. L network name is due to its L shape. L network can furnish low
pass filtering to decrease harmonic output. Two stages (amplifier, load) are con-
structed with simple low pass L network between stages. Low pass L network can
matches a higher output impedance source (Zhigh) to a lower input impedance load
(Zlow). Low pass L network can matches a lower output impedance source (Zlow) to
a higher input impedance load (Zhigh). We need to analyze the stability of typical
RF amplifier under RF and small signal conditions, the BJT can be replaced with
linear Hybrid Pi model and additional input and output matching circuits. We define
Xs(t) as the RF source and Rs RF source parasitic resistance (Fig. 4.6).
Fig. 4.6 Amplifier Hybrid Pi model and additional input and output matching circuits
Xs ðtÞ VA1 dVA1 dIL d

IRs ¼ ; I C 1 ¼ C1 ; VA1 VA2 ¼ L1 1 ; ICC1 ¼ CC1 ðVA2 VA3 Þ
Rs dt dt dt
dIL3 VA4 dIL
IL1 ¼ ICC1 ; VA3 VA4 ¼ L3 ; IRb2 ¼ ; IL3 ¼ IRb2 ; VA3 VA13 ¼ L2 2 ;
dt Rb2 dt
VA13 VA9 dILbplg VA5 VA6
IRb1 ¼ VA3 VA5 ¼ Lbplg ; IRbbp ¼ ; ILbplg ¼ IRbbp ;
Rb1 dt Rbbp
VA
IRc ¼ 9 ; Vpi ¼ VA6 VA8 ¼ VCe
Rc
dVpi VA VA8 dðVA6 VA7 Þ VA VA7
ICe ¼ Ce ; IRbpe ¼ 6 ; I C c ¼ Cc ; IRbpc ¼ 6 ;
dt Rbpe dt Rbpc
VA VA8
IRce ¼ 7
Rce
dILeplg dIL dLLc
VA8 ¼ Leplg ; VA7 VA10 ¼ Lcplg cplg ; VA9 VA10 ¼ Lc ;
dt dt dt
dðVA10 VA11 Þ
ICc2 ¼ Cc2
dt
dVA11 dIL
I C2 ¼ C 2 ; VA11 VA12 ¼ L4 4 ; VA12 ¼ Vout ; IL4 ¼ IRload ;
dt dt
Vout VA12
IRload ¼ ¼
Rload Rload
Table 4.1 Amplifier Hybrid Pi model and matching circuit Kirchhoff Current Law (KCL) for all
nodes
Node KCL @ Ai Node KCL @ Ai
number number
A1 IRs ¼ IL1 þ IC1 A8 ICe þ IRbpe þ gm Vpi þ IRce ¼ ILeplg
A2 IL1 ¼ ICc1 A9 IRb1 ¼ IRc þ ILc
A3 ICc1 ¼ IL2 þ ILbplg þ IL3 A10 ILc þ ILcplg ¼ ICc2
A4 IL3 ¼ IRb2 A11 ICc2 ¼ IL4 þ IC2
A5 ILbplg ¼ IRbbp A12 IL4 ¼ IRload
A6 IRbbp ¼ IRbpc þ ICc þ IRbpe þ ICe A13 IL2 ¼ IRb1
A7 ICc þ IRbpc ¼ gm Vpi þ IRce þ ILcplg
We can write our circuit Kirchhoff Current Law (KCL) for all nodes (Table 4.1):
Xs ðtÞ VA1
IRs ¼ ) IRs Rs ¼ Xs ðtÞ VA1 ) VA1 ¼ Xs ðtÞ IRs Rs
Rs

dVA1 dXs ðtÞ dIRs dIL
I C1 ¼ C1 ¼ C1 Rs ; VA1 VA2 ¼ L1 1 ) VA2
dt dt dt dt
dIL1
¼ VA1 L1
dt

dIL1 dVA2 dVA3
VA2 ¼ Xs ðtÞ IRs Rs L1 ; ICC1 ¼ CC1
dt dt dt
dXs ðtÞ dIRs d 2 IL1 dVA3
¼ CC1 Rs L1 2
dt dt dt dt
VA4 dIL
IRb2 ¼ ) VA4 ¼ IRb2 Rb2 ; VA3 VA4 ¼ L3 3 ) VA3
Rb2 dt
dIL3 dVA3 dIRb2 d 2 IL
¼ IRb2 Rb2 þ L3 ; ¼ Rb2 þ L3 2 3
dt dt dt dt

dXs ðtÞ d 2 IL 1 dIRb2 d 2 IL 3
System equation No. 1: ICC1 ¼ CC1 dt dIdtRs Rs L1 dt2 dt Rb2 L3 dt2
dIL2 dIL dIL dIL

VA3 VA13 ¼ L2 ) VA13 ¼ VA3 L2 2 ¼ IRb2 Rb2 þ L3 3 L2 2
dt dt dt dt
VA13 VA9
IRb1 ¼ ) IRb1 Rb1 ¼ VA13 VA9 ) VA9 ¼ VA13 IRb1 Rb1
Rb1
dIL3 dIL
VA9 ¼ VA13 IRb1 Rb1 ¼ IRb2 Rb2 þ L3 L2 2 IRb1 Rb1
dt dt
dILbplg dIL dIL

VA3 VA5 ¼ Lbplg ) VA5 ¼ IRb2 Rb2 þ L3 3 Lbplg bplg
dt dt dt
VA5 VA6
IRbbp ¼ ) IRbbp Rbbp ¼ VA5 VA6 ) VA6
Rbbp
dIL dIL
¼ IRb2 Rb2 þ L3 3 Lbplg bplg IRbbp Rbbp
dt dt
VA9 VA VA8
IR c ¼ ) VA9 ¼ IRc Rc ; IRbpe ¼ 6 ) IRbpe Rbpe ¼ VA6 VA8 ) VA8 ¼ VA6 IRbpe Rbpe
Rc Rbpe
dIL dIL
VA8 ¼ VA6 IRbpe Rbpe ¼ IRb2 Rb2 þ L3 3 Lbplg bplg IRbbp Rbbp IRbpe Rbpe
dt dt

I Cc ¼ C c ; IRbpc ¼ 6 ) I Cc
dt Rbpc
dIR
¼ Cc Rbpc bpc ; VA7 VA8 ¼IRce Rce
dt
dIL3 dIL
VA7 ¼VA8 þ IRce Rce ¼ IRb2 Rb2 þ L3 Lbplg bplg IRbbp Rbbp IRbpe
dt dt
Rbpe þ IRce Rce
System equation No. 2:
dILeplg dIL dIL

VA8 ¼ Leplg ) IRb2 Rb2 þ L3 3 Lbplg bplg IRbbp Rbbp IRbpe Rbpe
dt dt dt
dILeplg
¼ Leplg
dt
dILcplg dIL
VA7 VA10 ¼ Lcplg ) VA10 ¼ VA7 Lcplg cplg
dt dt
dIL dIL
VA10 ¼ IRb2 Rb2 þ L3 3 Lbplg bplg IRbbp Rbbp IRbpe Rbpe þ IRce Rce
dt dt
dILcplg
Lcplg
dt
dLLc dLLc dVA10 ICc2 IC
VA9 VA10 ¼ Lc ) VA9 ¼ VA10 þ Lc ; ¼ þ 2
dt dt dt Cc2 C2
dIL dIL
dt dt
dIL dLLc
Lcplg cplg þ Lc
dt dt

dðVA10 VA11 Þ dVA10 dVA11
ICc2 ¼ Cc2 ) ICc2 ¼ Cc2 ) ICc2
dt dt dt
dVA10 IC2
¼ Cc2
dt C2

ICc2 I C2 d dIL dIL
þ ¼ IRb2 Rb2 þ L3 3 Lbplg bplg IRbbp Rbbp
Cc2 C2 dt dt dt

dIL
IRbpe Rbpe þ IRce Rce Lcplg cplg
dt
ICc2 IC dIRb2 d 2 IL d 2 ILbplg dIRbbp dIR
þ 2¼ Rb2 þ L3 2 3 Lbplg 2
Rbbp bpe
Cc2 C2 dt dt dt dt dt
dIRce d 2 ILcplg
Rbpe þ Rce Lcplg
dt dt2
dIL4 dIL VA
VA11 VA12 ¼ L4 ) VA11 ¼ VA12 þ L4 4 ; IRload ¼ 12 ) VA12
dt dt Rload
¼ IRload Rload

dIL4 dIRload d 2 IL
VA11 ¼ IRload Rload þ L4 ; IC2 ¼ C2 Rload þ L4 2 4
dt dt dt
We can summery our system equations:

dXs ðtÞ dIRs d 2 IL dIR d 2 IL
ICC1 ¼ CC1 Rs L1 2 1 b2 Rb2 L3 2 3
dt dt dt dt dt
dIL3 dIL dIL

IRb2 Rb2 þ L3 Lbplg bplg IRbbp Rbbp IRbpe Rbpe ¼ Leplg eplg
dt dt dt

þ 2¼ Rb2 þ L3 2 3 Lbplg Rbbp bpe Rbpe
Cc2 C2 dt dt dt2 dt dt
2
dIRce d ILcplg dIRload d 2 I L4
þ Rce Lcplg I C2 ¼ C 2 Rload þ L4 2
dt dt2 dt dt
Since ICc2 ¼ IL4 þ IC2 ¼ IRload þ IC2 ) IC2 ¼ ICc2 IRload then

dXs ðtÞ dIRs d 2 IL1 dIRb2 d 2 I L3
ICC1 ¼ CC1 Rs L1 2 Rb2 L3 2
dt dt dt dt dt
dIL3 dIL dIL

dt dt dt

þ 2¼ Rb2 þ L3 2 3 Lbplg Rbbp bpe
dIRce d 2 ILcplg
Rbpe þ Rce Lcplg
dt dt2

dI d2 I
ICc2 IRload ¼ C2 Rdtload Rload þ L4 dt2L4 . We can restrict our two differential
equations to one differential equation:

d dIL dIL d 2 ILeplg
IRb2 Rb2 þ L3 3 Lbplg bplg IRbbp Rbbp IRbpe Rbpe ¼ Leplg
dt dt dt dt2
dIRb2 d 2 IL d 2 ILbplg dIRbbp dIR d 2 ILeplg
Rb2 þ L3 2 3 Lbplg 2
Rbbp bpe Rbpe ¼ Leplg
dt dt dt dt dt dt2

þ 2¼ Rb2 þ L3 2 3 Lbplg Rbbp bpe
dIRce d 2 ILcplg
Rbpe þ Rce Lcplg
dt dt2
I I d 2 ILcplg d 2 ILeplg
One differential equation: CCc2c2 þ CC22 dIdtRce Rce þ Lcplg dt2 ¼ Leplg dt2
We get system three differential equations:
IL3 ¼ IRb2 ; IL1 ¼ ICc1 ; ICc2 ¼ IL4 þ IC2 ) IC2 ¼ ICc2 IRload

IL1 ¼ CC1 Rs L1 2 1 b2 Rb2 L3 2 3
dt dt dt dt dt
ICc2 IC dIR d 2 ILcplg d 2 ILeplg

þ 2 ce Rce þ Lcplg ¼ Leplg
Cc2 C2 dt dt2 dt2

dIRload d 2 IRload
ICc2 IRload ¼ C2 Rload þ L4
dt dt2
We need to find our system equilibrium points (fixed points). It is done by setting
dIRs dIRb2 dIL1 d 2 IL1 dIL3 d 2 IL3
dt ¼ 0; dt ¼ 0; dt ¼0) dt2 ¼ 0; dt ¼0) dt2 ¼ 0; dIdtRce ¼ 0
dILcplg d 2 ILcplg dILeplg d 2 ILeplg dIRload d 2 IRload

¼0) ¼ 0; ¼ 0 ) ¼ 0; ¼ 0 ) ¼0
dt dt2 dt dt2 dt dt2
We get the following system fixed points:
dXs ðtÞ 1 1 IR

IL1 ¼ CC1 ; ICc2 ð þ Þ load ¼ 0; IC c2 IRload ¼ 0
dt Cc2 C2 C2
dXs ðtÞ dXs ðtÞ

IC c2 IRload ¼ 0 ) IC 2 ¼ 0; IL1 ¼ CC1 ) IC c1 ¼ CC1
dt dt
dXs ðtÞ dXs ðtÞ dXs ðtÞ
IL1 ¼ CC1 ) CC1 ¼ IRs IC 1 ; CC1 ¼ IL2 þ ILbplg þ IL3
dt dt dt
We consider input RF source Xs ðtÞ ¼ A0 þ nðtÞ; jnðtÞj\1&A0 jnðtÞjThen

dXs ðtÞj
Xs ðtÞjA0 jnðtÞj ¼ A0 þ nðtÞ A0 ; A0 jnðtÞj
dt ¼ dnðtÞ
dt ! e.
Based on the above assumption, we get the following system fixed points:

1 1 IRload
IL1 ¼ 0; IC c2 þ ¼ 0; IC c2 IRload ¼ 0; IRs IC 1 ) IRs ¼ IC 1
Cc2 C2 C2
IC c2 IRload ¼ 0 ) IC 2 ¼ 0; IL1 ¼ 0 ) IC c1 ¼ 0; IL1 ¼ 0 ) IC 1

¼ IRs ; IL2 þ ILbplg þ IL3 ¼ 0
E ðIL1 ; IC c2 ; IRload ; IRce ; IC 2 ; IC 1 ; IRs ; IL2 ; ILbplg ; IL3 Þ
¼ ð0; IRload ; IRload ; IRce ; 0; IRs ; IRs ; ðILbplg þ IL3 Þ; ILbplg ; IL3 Þ
librium points of the Small Signal (SS) amplifier with matching network consists in
adding to coordinate [IL1 ; ICc2 ; IRload ; IRce ; IC2 ; IC1 ; IRs ; IL2 ; ILbplg ; IL3 ] arbitrarily small
increments of exponentially form ½iL1 ; iCc2 ; iRload ; iRce ; iC2 ; iC1 ; iRs ; iL2 ; iLbplg ; iL3 ekt
and retaining the first order terms in
IL1 ; ICc2 ; IRload ; IRce ; IC2 ; IC1 ; IRs ; IL2 ; ILbplg ; IL3 . The system of homogenous equations
leads to a polynomial characteristic equation in the eigenvalues. The polynomial
characteristic equations accept by set of the below circuit variables, circuit variables
derivative and circuit variables second order derivative with respect to time into
Small Signal (SS) amplifier with matching network equivalent circuit [2–4]. Our
Small Signal (SS) amplifier with matching network equivalent circuit fixed values
with arbitrarily small increments of exponential form ½iL1 ; iCc2 ; iRload ; iRce ; iC2 ; iC1 ;
iRs ; iL2 ; iLbplg ; iL3 ekt are: j = 0 (first fixed point), j = 1 (second fixed point), j = 2
(third fixed point), etc.,
dIL1 d 2 IL1 dIL3 d 2 IL3
We define new variables: Y1 ¼ dt ) dYdt1 ¼ dt2 ; Y2 ¼ dt ) dYdt2 ¼ dt2

dIRload dY3 d 2 IRload dY3
Y3 ¼ ) ¼ ; ICc2 IRload ¼ C2 Y3 Rload þ L4
dt dt dt2 dt

dXs ðtÞ dIRs dY1 dIRb2 dY2
I L 1 ¼ C C1 !e Rs L1 Rb2 L3
dt dt dt dt dt
1 1 IR dIR d 2 ILcplg d 2 ILeplg

ICc2 ð þ Þ load ce Rce þ Lcplg ¼ L eplg
Cc2 C2 C2 dt dt2 dt2
2 2
dIL dY4 d ILcplg dIL dY5 d ILeplg
Y4 ¼ cplg ) ¼ 2
; Y5 ¼ eplg ) ¼
dt dt dt dt dt dt2
1 1 IR dIR dY4 dY5
ICc2 ð þ Þ load ce Rce þ Lcplg ¼ Leplg
Cc2 C2 C2 dt dt dt
ðjÞ ðjÞ ðjÞ
Y1 ðtÞ ¼ Y1 þ y1 ekt ; Y2 ðtÞ ¼ Y2 þ y2 ekt ; Y3 ðtÞ ¼ Y3 þ y3 ekt ; IL1 ðtÞ
ðjÞ
¼ IL1 þ iL1 ekt
ðjÞ ðjÞ dY4 ðtÞ dY5 ðtÞ

Y4 ðtÞ ¼ Y4 þ y4 ekt ; Y5 ðtÞ ¼ Y5 þ y5 ekt ; ¼ y4 k ekt ; ¼ y5 k ekt
dt dt
ðjÞ ðjÞ dIL ðtÞ dIL ðtÞ
IL3 ðtÞ ¼ IL3 þ iL3 ekt ; IRload ðtÞ ¼ IRload þ iRload ekt ; 1 ¼ iL1 k ekt ; 3 ¼ iL3 k ekt
dt dt
dIRload ðtÞ dIL ðjÞ

¼ iRload k ekt ; Y1 ¼ 1 ) Y1 þ y1 ekt ¼ iL1 k ekt ; ICc2 ðtÞ
dt dt
ðjÞ
¼ ICc2 þ iCc2 ekt
dIL1 ðjÞ dIL3 ðjÞ dIRload ðjÞ

At fixed point dt ¼ 0 ) Y1 ¼ 0; dt ¼ 0 ) Y2 ¼ 0; dt ¼ 0 ) Y3 ¼ 0
dILcplg ðjÞ dIL ðjÞ dIL

¼ 0 ) Y4 ¼ 0; eplg ¼ 0 ) Y5 ¼ 0; cplg jY ðjÞ ¼0 ) iLcplg k þ y4 ¼ 0
dt dt dt 4
dILeplg
j ðjÞ ) iLeplg k ekt ¼ y5 ekt ; iLeplg k þ y5 ¼ 0
dt Y5 ¼0
ðjÞ ðjÞ ðjÞ

IRs ðtÞ ¼ IRs þ iRs ekt ; IRb2 ðtÞ ¼ IRb2 þ iRb2 ekt ; IRCe ðtÞ ¼ IRCe þ iRCe ekt
ðjÞ ðjÞ dIL1

ILcplg ðtÞ ¼ ILcplg þ iLcplg ekt ; ILeplg ðtÞ ¼ ILeplg þ iLeplg ekt ; Y1 ¼ j ðjÞ ) y1 ekt
dt Y1 ¼0
¼ iL1 k ekt
dIL3
iL1 k þ y1 ¼ 0; Y2 ¼ j ðjÞ ) y2 ekt ¼ iL3 k ekt ; iL3 k þ y2 ¼ 0
dt Y2 ¼0
dIRload
Y3 ¼ j ðjÞ ) y3 ekt ¼ iRload k ekt ; iRload k þ y3 ¼ 0
dt Y3 ¼0
dY3
ICc2 IRload ¼ C2 Y3 Rload þ L4
dt
ðjÞ ðjÞ ðjÞ
ICc2 þ iCc2 ekt IRload iRload ekt ¼ C2 ðY3 þ y3 ekt Þ Rload þ L4 y3 k ekt Þ
ðjÞ ðjÞ ðjÞ

@ fixed point dY3
dt ¼ 0 ) ICc2 IRload C2 Rload Y3 ¼ 0
ðjÞ ðjÞ ðjÞ

ICc2 IRload C2 Y3 Rload þ iCc2 ekt iRload ekt
¼ y3 C2 Rload ekt þ C2 L4 y3 k ekt
iCc2 ekt iRload ekt ¼ y3 C2 Rload ekt þ C2 L4 y3 k ekt
y3 C2 Rload þ C2 L4 y3 k ¼ iCc2 iRload

) y3 C2 L4 k y3 C2 Rload þ iCc2 iRload
¼0

dXs ðtÞ dIRs dY1 dIRb2 dY2
! e; IL1 ¼ CC1 R s L1 Rb2 L3
dt dt dt dt dt
ðjÞ
IL1 þ iL1 ekt ¼ CC1 ðiRs k ekt Rs L1 y1 k ekt iRb2 k ekt Rb2 L3
y2 k ekt Þ
ðjÞ
@ fixed point IL1 ¼ 0; iL1 ¼ CC1 ðiRs Rs L1 y1 iRb2 Rb2 L3 y2 Þ k

1 1 IR dIR d 2 ILcplg d 2 ILeplg
ICc2 þ load ce Rce þ Lcplg ¼ L eplg
Cc2 C2 C2 dt dt2 dt2

1 1 IR dIR dY4 dY5
ICc2 þ load ce Rce þ Lcplg ¼ Leplg
Cc2 C2 C2 dt dt dt

1 1
I
ðjÞ
R þ i R e kt
ðjÞ load
ICc2 þ iCc2 ekt iRCe k ekt Rce
load
þ
Cc2 C2 C2
þ Lcplg y4 k ekt ¼ Leplg y5 k ekt
ðjÞ
ðjÞ IR
@ fixed point ICc2 ðC1c2 þ C2 Þ
1
load
C2 ¼0
ðjÞ
ðjÞ 1 1 1 1 IR iR
ICc2 ð þ Þ þ iCc2 ð þ Þ ekt load load ekt iRCe k ekt Rce
Cc2 C2 Cc2 C2 C2 C2
kt kt
þ Lcplg y4 k e ¼ Leplg y5 k e
ðjÞ
ðjÞ 1 1 IR 1 1 iR
fICc2 ð þ Þ load g þ iCc2 ð þ Þ ekt load ekt iRCe k ekt Rce
Cc2 C2 C2 Cc2 C2 C2
kt kt
þ Lcplg y4 k e ¼ Leplg y5 k e

1 1 iR
iCc2 þ ekt load ekt iRCe k ekt Rce þ Lcplg y4 k ekt
Cc2 C2 C2
¼ Leplg y5 k ekt
1 1 iR
iCc2 ð þ Þ load iRCe k Rce þ Lcplg y4 k ¼ Leplg y5 k
Cc2 C2 C2
Small Signal (SS) amplifier with matching network system matrix

iLcplg ; iLeplg ; iL1 ; iL3 ; iRload ; iCc2 ; iRs ; iRb2 ; iRCe ; y1 ; y2 ; y3 ; y4 ; y5 can be constructed from
the below list of equations:
iLcplg k þ y4 ¼ 0; iLeplg k þ y5 ¼ 0; iL1 k þ y1 ¼ 0; iL3 k þ y2 ¼ 0;

iRload k þ y3 ¼ 0
y3 C2 L4 k y3 C2 Rload þ iCc2 iRload ¼ 0
iL1 ¼ CC1 ðiRs Rs L1 y1 iRb2 Rb2 L3 y2 Þ k

1 1 iR
iCc2 þ load iRCe k Rce þ Lcplg y4 k ¼ Leplg y5 k
Cc2 C2 C2
Assumption: We consider for simplicity that arbitrarily small increments ele-

ments iRs ; y1 ; iRb2 ; y2 are proximally the same iRs y1 iRb2 y2 ! z1 and we
represent them as small increment element z1 . Additionally, arbitrarily small
increments elements iRCe ; y4 ; y5 are proximally the same iRCe y4 y5 ! z2 and we
represent them as small increment element z2 .
iLcplg k þ z2 ¼ 0; iLeplg k þ z2 ¼ 0; iL1 k þ z1 ¼ 0; iL3 k þ z1 ¼ 0;

iRload k þ y3 ¼ 0
Rload 1 1
y3 k y3 þ iCc2 iRload
L4 C2 L4 C2 L4
1
¼ 0; z1 k iL1 ¼0
CC1 ðRs þ L1 þ Rb2 þ L3 Þ

Cc2 þ C2
1 1
iRload
z2 k þ iCc2 ¼0
ðRce Lcplg þ Leplg Þ C2 ðRce Lcplg þ Leplg Þ
Small Signal (SS) amplifier with matching network system matrixes:

0 1
iLcplg
00 1
B C
B iLeplg C
B C B0 C
0 1 B C B
B i L1 C B 0
C
C
i11 . . . i18 B C B C
B .. .. .. C B
B i L C
3 C
B
B
0 C
C iCc2 ¼ 0
@ . . . AB CþB0 C
B iRload C B 1 C
i81 i88 B C B C2 L4 C
B y3 C B C
B C @0 1 1 A
B C ðC þ C Þ
@ z1 A c2 2
ðRce Lcplg þ Leplg Þ
z2
i11 ¼ k; i12 ¼ . . . ¼ i17 ¼ 0; i18 ¼ 1; i21 ¼ 0; i22 ¼ k; i23 ¼ . . . ¼ i27 ¼ 0; i28 ¼ 1
i31 ¼ i32 ¼ 0; i33 ¼ k; i34 ¼ i35 ¼ i36 ¼ 0; i37 ¼ 1; i38 ¼ 0; i41 ¼ i42 ¼ i43 ¼ 0;
i44 ¼ k
i45 ¼ i46 ¼ 0; i47 ¼ 1; i48 ¼ 0; i51 ¼ . . . ¼ i54 ¼ 0; i55 ¼ k; i56 ¼ 1; i57 ¼ i58 ¼ 0
1 Rload
i61 ¼ . . . ¼ i64 ¼ 0; i65 ¼ ; i66 ¼ k ; i67 ¼ i68 ¼ 0
C2 L4 L4
1
i71 ¼ i72 ¼ 0; i73 ¼ ; i74 ¼ i75 ¼ i76 ¼ 0; i77 ¼ k; i78 ¼ 0
CC1 ðRs þ L1 þ Rb2 þ L3 Þ
1
i81 ¼ . . . ¼ i84 ¼ 0; i85 ¼ ; i86 ¼ i87 ¼ 0; i88 ¼ k
C2 ðRce Lcplg þ Leplg Þ
0 1
0 1 iLcplg
0 B C
B0 C B iLeplg C
B C B C
B0 C 0 1 B C
B C B i L1 C
B0 C i11 . . . i18 B C
B C .. C B i L3 C
We consider B C ! eB .
@ ..
.. B C0
B01 C . . AB C
B C B iRload C
B C2 L4 C i81 i88 B C
B0 C B y3 C
B C B C
@ þ C1 A B C
@ z1 A
1
Cc2 2
ðRce Lcplg þ Leplg Þ
z2
0 1 0 1
i11 . . . i18 i11 ... i18
B .. .. .. C B .. .. .. C
AkI ¼@ . . . A ; det(A k IÞ ¼ 0 ) det@ . . . A¼0
i81 i88 i81 i88

[2–4].
4.2 Small Signal (SS) Amplifiers PI & T’s Matching

Network and Transformation
In (4.1) we discuss the stability and fixed points (equilibrium points) analysis of
Small Signal (SS) amplifiers with input and output L-matching network. Another
way is to use Small Signal (SS) amplifier with input and output PI & T’s matching
networks. The PI or T matching networks constructed from three elements impe-
dance matching and are used in many narrow band applications. The narrow band is
due to the higher loaded Q over what the L network possesses. PI and T networks
also permit any Q to be selected and always PI and T Q’s (loaded quality factor) is
bigger or equal to the L network Q’s.
The Qis desired for a particular applications
and is calculated with the formula Q ¼ f2 f
fc
1
when utilizing with high Q inductor.
Q is the loaded quality factor of the matching circuit. fc is the center frequency of
the circuit. f2 is the upper frequency and need to pass with little loss. f1 is the lower
frequency and need to pass with little loss. We consider PI network as a two L
networks which attached back to back. Any PI network can be transformed to an
equivalent T network. This is also known as the Wye-Delta transformation, which
is the terminology used in power distribution and electrical engineering. The PI
matching network is equivalent to the Delta and the T matching network is
equivalent to the Wye (or star) form [25, 26, 33]. The typical RF amplifier sche-
matic with input and output matching circuit is as follow (Fig. 4.7):
4.2 Small Signal (SS) Amplifiers PI & T’s … 423
Fig. 4.7 Small Signal (SS) amplifier with PI matching networks
Under AC and Small Signal (SS) conditions, the BJT can be replaced with linear
Hybrid PI model (Fig. 4.8):
Fig. 4.8 Small Signal (SS) amplifier Hybrid PI model with PI matching networks
Fig. 4.9 Linear Hybrid PI model and additional input and output matching circuits
We need to analyze the stability of typical RF amplifier under RF and small

signal conditions, the BJT can be replaced with linear Hybrid Pi model and addi-
tional input and output matching circuits. We define Xs(t) as the RF source and
Rs RF source parasitic resistance (Fig. 4.9).
Xs ðtÞ VA1 dVA1 dVA2 dIL

IRs ¼ ; ICa1 ¼ Ca1 ; ICa2 ¼ Ca2 ; VLa ¼VA1 VA2 ¼ La a
Rs dt dt dt
d
ICC1 ¼ CC1 ðVA2 VA3 Þ; VA1 ¼ VCa1 ; VA2 ¼ VCa2 ; ILa ¼ ICa2 þ ICC1 ; IRs ¼ ILa þ ICa1
dt
dIL VA dIL VA VA 9
VA3 VA4 ¼ L3 3 ; IRb2 ¼ 4 ; IL3 ¼ IRb2 ; VA3 VA13 ¼ L2 2 ; IRb1 ¼ 13
dt Rb2 dt Rb1
dILbplg VA5 VA6 VA9
VA3 VA5 ¼ Lbplg ; IRbbp ¼ ; ILbplg ¼ IRbbp ; IRc ¼ ; Vpi ¼ VA6 VA8 ¼ VCe
dt Rbbp Rc
dVpi VA VA 8 dðVA6 VA7 Þ VA VA 7 VA VA8
ICe ¼ Ce ; IRbpe ¼ 6 ; ICc ¼ Cc ; IRbpc ¼ 6 ; IRce ¼ 7
dt Rbpe dt Rbpc Rce
dILeplg dILcplg dLLc dðVA10 VA11 Þ
VA8 ¼ Leplg ; VA7 VA10 ¼ Lcplg ; VA9 VA10 ¼ Lc ; ICc2 ¼ Cc2
dt dt dt dt
dVA11 dIL dVA12

ICb1 ¼ Cb1 ; VLb ¼VA11 VA12 ¼ Lb b ; VA12 ¼ Vout ; ICb2 ¼ Cb2 ;
dt dt dt
Vout VA
IRload ¼ ¼ 12
Rload Rload
VA11 ¼ VCb1 ; VA12 ¼ VCb2 ; ICc2 ¼ ILb þ ICb1 ; ILb ¼ ICb2 þ IRload
Table 4.2 Linear Hybrid PI model and additional matching circuits Kirchhoff’s Current Law
(KCL) for all nodes
Node KCL @ Ai Node KCL @ Ai
number number
A1 IRs ¼ ILa þ ICa1 A8 ICe þ IRbpe þ gm Vpi þ IRce ¼ ILeplg
A2 ILa ¼ ICa2 þ ICC1 A9 IRb1 ¼ IRc þ ILc
A4 IL3 ¼ IRb2 A11 ICc2 ¼ ILb þ ICb1
A5 ILbplg ¼ IRbbp A12 ILb ¼ ICb2 þ IRload
We can write our circuit Kirchhoff’s Current Law (KCL) for all nodes:
Xs ðtÞ VA1
Rs
dVA1 d dXs ðtÞ dIR
ICa1 ¼ Ca1 ¼ Ca1 ðXs ðtÞ IRs Rs Þ ¼ Ca1 Ca1 Rs s
dt dt dt dt
dILa dILa dIL
VLa ¼VA1 VA2 ¼ La ) VA2 ¼ VA1 La ¼ Xs ðtÞ IRs Rs La a
dt dt dt
dVA2 d dILa
ICa2 ¼ Ca2 ¼ Ca2 ðXs ðtÞ IRs Rs La Þ
dt dt dt
dXs ðtÞ dIR d 2 IL
¼ Ca2 Ca2 Rs s Ca2 La 2 a
dt dt dt
d dVA2 dVA3
ICC1 ¼ CC1 ðVA2 VA3 Þ ¼ CC1 C C1
dt dt dt
d dILa dVA3
¼ CC1 ðXs ðtÞ IRs Rs La Þ C C1
dt dt dt
d dXs ðtÞ dIRs d 2 IL
ICC1 ¼ CC1 ðVA2 VA3 Þ ¼ CC1 CC1 Rs CC1 La 2 a
dt dt dt dt
dVA3
CC1
dt
VA4 dIL dIL

IRb2 ¼ ) VA4 ¼ IRb2 Rb2 ; VA3 VA4 ¼ L3 3 ) VA3 ¼ VA4 þ L3 3
Rb2 dt dt
dIL3
¼ IRb2 Rb2 þ L3
dt
dVA3 dVA4 d 2 IL dIRb2 d 2 IL

¼ þ L3 2 3 ¼ Rb2 þ L3 2 3
dt dt dt dt dt
d
ICC1 ¼ CC1 ðVA2 VA3 Þ
dt
dXs ðtÞ dIRs d 2 IL dIR
¼ CC1 CC1 Rs CC1 La 2 a CC1 ð b2 Rb2 þ L3
dt dt dt dt
d 2 I L3
2 Þ
dt

ICC1 ¼ CC1 CC1 Rs CC1 La 2 a b2 CC1 Rb2
dt dt dt dt
d 2 I L3
C C1 L3 2
dt
dIL2 dIL dIL dIL
dt dt dt dt
VA13 VA9
Rb1
dIL3 dIL
dt dt
dILbplg dIL dIL
dt dt dt
VA5 VA6
Rbbp
dIL dIL
¼ IRb2 Rb2 þ L3 3 Lbplg bplg IRbbp Rbbp
dt dt
VA9 VA VA8
IRc ¼ ) VA9 ¼ IRc Rc ; IRbpe ¼ 6 ) IRbpe Rbpe
Rc Rbpe
¼ VA6 VA8 ) VA8 ¼ VA6 IRbpe Rbpe
dIL3 dIL
VA8 ¼ VA6 IRbpe Rbpe ¼ IRb2 Rb2 þ L3 Lbplg bplg IRbbp Rbbp
dt dt
IRbpe Rbpe

I Cc ¼ C c ; IRbpc ¼ 6 ) I Cc
dt Rbpc
dIR
¼ Cc Rbpc bpc ; VA7 VA8 ¼IRce Rce
dt
dIL3 dIL
dt dt
Rbpe þ IRce Rce
dILeplg dIL dIL

dt dt dt
dILeplg
¼ Leplg
dt
dILcplg dIL
dt dt
dIL dIL
dt dt
dILcplg
Lcplg
dt
VA9 VA10 ¼ Lc ) VA9 ¼ VA10 þ Lc ; ¼ þ b1
dt dt dt Cc2 Cb1
dIL dIL
dt dt
dILcplg dLLc
Lcplg þ Lc
dt dt

ICc2 ¼ Cc2 ) ICc2 ¼ Cc2 ) ICc2
dt dt dt
dVA10 ICb1
¼ Cc2
dt Cb1

ICc2 IC d dIL dIL
þ b1 ¼ IRb2 Rb2 þ L3 3 Lbplg bplg IRbbp Rbbp
Cc2 Cb1 dt dt dt
dILcplg o
IRbpe Rbpe þ IRce Rce Lcplg
dt

þ b1 ¼ Rb2 þ L3 2 3 Lbplg 2
Rbbp bpe
Cc2 Cb1 dt dt dt dt dt
dIRce d 2 ILcplg
Rbpe þ Rce Lcplg
dt dt2
dILb dIL VA
VA11 VA12 ¼ Lb ) VA11 ¼ VA12 þ Lb b ; IRload ¼ 12 ) VA12
dt dt Rload
¼ IRload Rload
dILb dVA12 dIR

VA11 ¼ IRload Rload þ Lb ; VA12 ¼ VCb2 ; ICb2 ¼ Cb2 ¼ Cb2 Rload load
dt dt dt
dIRload
ILb ¼ ICb2 þ IRload ) ILb ¼ Cb2 Rload þ IRload
dt

ICC1 ¼ CC1 CC1 Rs CC1 La 2 a b2 CC1 Rb2 CC1 L3
dt dt dt dt
d 2 I L3

dt2
dIL3 dIL dIL

dt dt dt

þ b1 ¼ Rb2 þ L3 2 3 Lbplg Rbbp bpe
Cc2 Cb1 dt dt dt2 dt dt
dIRce d 2 ILcplg
Rbpe þ Rce Lcplg
dt dt2
dIRload
ILb ¼ Cb2 Rload þ IRload
dt
Since
ICc2 ¼ ILb þ ICb1 ) ICb1 ¼ ICc2 ILb ; ILb ¼ ICb2 þ IRload ; ICb1 ¼ ICc2 ICb2 IRload

ICC1 ¼ CC1 CC1 Rs CC1 La 2 a b2 CC1 Rb2 CC1 L3
dt dt dt dt
d 2 I L3

dt2
dIL3 dIL dIL

dt dt dt

þ b1 ¼ Rb2 þ L3 2 3 Lbplg 2
Rbbp bpe
Cc2 Cb1 dt dt dt dt dt
dIRce d 2 ILcplg
Rbpe þ Rce Lcplg
dt dt2

d dIL3 dILbplg d 2 ILeplg
IRb2 Rb2 þ L3 Lbplg IRbbp Rbbp IRbpe Rbpe ¼ Leplg
dt dt dt dt2
ICc2 ICb1 d 2 ILeplg dIRce

We get one differential equation: Cc2 þ Cb1 ¼ Leplg dt2 þ dt Rce
2
d ILcplg
Lcplg dt2

1 1 IL
ICb1 ¼ ICc2 ILb ) ICc2 þ b
Cc2 Cb1 Cb1
d 2 ILeplg dIRce d 2 ILcplg
¼ Leplg þ Rce Lcplg
dt2 dt dt2
We get for our system three differential equations:

ICC1 ¼ CC1 CC1 Rs CC1 La 2 a b2 CC1 Rb2 CC1 L3 2 3
dt dt dt dt dt

1 1 IL b d 2 ILeplg dIRce d 2 ILcplg
ICc2 þ ¼ Leplg þ Rce Lcplg ;
Cc2 Cb1 Cb1 dt2 dt dt2
dIR
ILb ¼ Cb2 Rload load þ IRload
dt
We need to find our system equilibrium points (fixed points). It is done by

dIRs 2 dIRb2 dIL3 d 2 IL3
setting dt ¼ 0; dIdtLa ¼ 0 ) ddtI2La ¼ 0; dt ¼ 0; dt ) dt2 ¼0
dILeplg d 2 ILeplg dIRce dILcplg d 2 ILcplg dIR

¼0) ¼ 0; ¼ 0; ) ¼ 0; load ¼ 0
dt dt2 dt dt dt2 dt
We get the system fixed points (equilibrium points):
dXs ðtÞ 1 1 IL

IC C ¼ CC1 ; ICc2 ð þ Þ b ¼ 0 ) ILb
1 dt Cc2 Cb1 Cb1
1 1
¼ IC c2 ð þ
Þ Cb1 ; ILb ¼ IRload
Cc2 Cb1
We consider input RF source Xs ðtÞ ¼ A0 þ nðtÞ;jnðtÞj\1&A0 jnðtÞj

dXs ðtÞjA jnðtÞj
Then. Xs ðtÞjA0 jnðtÞj ¼ A0 þ nðtÞ A0 ; dt
0
¼ dnðtÞ
dt ! e:
Based on the above assumption, we get the following system fixed points:

1 1 ILb 1 1
IC C ¼ 0; IC c2 þ
¼ 0 ) ILb ¼ ICc2 þ Cb1 ; ILb
1 Cc2 Cb1 Cb1 Cc2 Cb1
¼ IRload

E IC C ; IRs ; ILa ; IRb2 ; IL3 ; IC c2 ; ILb ; ILeplg ; IRce ; ILcplg ; IRload

1
1 1
¼ 0; IRs ; ILa ; IRb2 ; IL3 ; IC c2 ; IC c2 þ Cb1 ; ILeplg ; IRce ; ILcplg ; IC c2
C c2 C b1
C1c2 þ 1
Cb1 Cb1
equilibrium points of the Smallh Signal (SS) amplifier with matching networki
consists in adding to coordinate ICC1 ; IRs ; ILa ; IRb2 ; IL3 ; ICc2 ; ILb ; ILeplg ; IRce ; ILcplg ; IRload
arbitrarily small increments of exponentially form ½iCC1 ; iRs ; iLa ; iRb2 ; iL3 ;
iCc2 ; iLb ; iLeplg ; iRce ; iLcplg ; iRload ekt and retaining the first order terms in
ICC1 ; IRs ; ILa ; IRb2 ; IL3 ; ICc2 ; ILb ; ILeplg ; IRce ; ILcplg ; IRload [3, 4]. The system of homogenous
equations leads to a polynomial characteristic equation in the eigenvalues. The
polynomial characteristic equations accept by set of the below circuit variables,
circuit variables derivative and circuit variables second order derivative with respect
to time into Small Signal (SS) amplifier with matching network equivalent circuit.
Our Small Signal (SS) amplifier with matching network equivalent circuit fixed
values with arbitrarily small increments of exponential form ½iCC1 ; iRs ; iLa ; iRb2 ; iL3 ;
iCc2 ; iLb ; iLeplg ; iRce ; iLcplg ; iRload ekt are: j = 0 (first fixed point), j = 1 (second fixed
point), j = 2 (third fixed point), etc.,
2 dIL3 d 2 IL3
We define new variables: Y1 ¼ dIdtLa ) dYdt1 ¼ ddtI2La ; Y2 ¼ dt ) dYdt2 ¼ dt2

dXs ðtÞ dIR dY1 dIRb2 dY2
ICC1 ¼ CC1 ! e s CC1 Rs CC1 La CC1 Rb2 CC1 L3
dt dt dt dt dt
2 2
dIL dY3 d ILeplg dIL dY4 d ILcplg
Y3 ¼ eplg ) ¼ ; Y4 ¼ cplg ) ¼
dt dt dt2 dt dt dt2

1 1 I Lb dY3 dIRce dY4
ICc2 þ ¼ Leplg þ Rce Lcplg
Cc2 Cb1 Cb1 dt dt dt

Y1 ðtÞ ¼ Y1 þ y1 ekt ; Y2 ðtÞ ¼ Y2 þ y2 ekt ; Y3 ðtÞ ¼ Y3 þ y3 ekt ; Y4 ðtÞ ¼ Y4 þ y4 ekt
ICc1 ðtÞ ¼ ICc1 þ iCc1 ekt ; IRs ðtÞ ¼ IRs þ iRs ekt ; IRb2 ðtÞ ¼ IRb2 þ iRb2 ekt ; ICc2 ðtÞ ¼ ICc2 þ iCc2 ekt
ðjÞ ðjÞ dIRs ðtÞ dY1 ðtÞ
ILb ðtÞ ¼ ILb þ iLb ekt ; IRce ðtÞ ¼ IRce þ iRce ekt ; ¼ iRs k ekt ; ¼ y1 k ekt
dt dt
dIRb2 ðtÞ dY2 ðtÞ ðjÞ dIR ðtÞ
¼ iRb2 k ekt ; ¼ y2 k ekt ; IRload ðtÞ ¼ IRload þ iRload ekt ; load ¼ iRload k ekt
dt dt dt
dILa ðjÞ dIL3 ðjÞ dILeplg ðjÞ

At fixed point dt ¼ 0 ) Y1 ¼ 0; dt ¼ 0 ) Y2 ¼ 0; dt ¼ 0 ) Y3 ¼ 0
dILcplg ðjÞ dIL

¼ 0 ) Y4 ¼ 0; Y1 ¼ a jY ðjÞ ¼0 ) y1 ekt ¼ iLa k ekt ; y1 iLa k ¼ 0
dt dt 1
dIL3 dIL
Y2 ¼ jY ðjÞ ¼0 ) y2 ekt ¼ iL3 k ekt ; y2 iL3 k ¼ 0; Y3 ¼ eplg jY ðjÞ ¼0
dt 2 dt 3
) y3 ekt ¼ iLeplg k ekt
dILcplg
y3 iLeplg k ¼ 0; Y4 ¼ j ðjÞ ) y4 ekt ¼ iLcplg k ekt ;
dt Y4 ¼0
dXs ðtÞ
y4 iLcplg k ¼ 0; !e¼0
dt
dIRs dY1 dIRb2 dY2
I CC 1 ¼ CC1 Rs CC1 La CC1 Rb2 CC1 L3
dt dt dt dt
ðjÞ
ICc1 þ iCc1 ekt ¼ iRs k ekt CC1 Rs CC1 La y1 k ekt iRb2 k ekt CC1
Rb2 CC1 L3 y2 k ekt
ðjÞ
@ fixed point ICc1 ¼ 0 ; iCc1 ¼ ðiRs Rs þ La y1 þ iRb2 Rb2 þ L3 y2 Þ CC1 k
We divide the two side of the above equation by iCc1 term
iCc1 ¼ iRs k CC1 Rs CC1 La y1 k iRb2 k CC1 Rb2 CC1 L3 y2 k

1 1 IL dY3 dIRce dY4
ICc2 þ b ¼ Leplg þ Rce Lcplg
Cc2 Cb1 Cb1 dt dt dt

1 1
I
ðjÞ
Lb þ i L e kt
ðjÞ b
ICc2 þ iCc2 ekt þ
Cc2 Cb1 Cb1
¼ Leplg y3 k ekt þ iRce k ekt Rce Lcplg y4 k ekt
I

@ fixed point IC c2 1
Cc2 þ 1
Cb1 CLb1b ¼ 0 ) ILb ¼ IC c2 1
Cc2 þ 1
Cb1 Cb1
( ðjÞ
)
ðjÞ 1 1 IL 1 1 iLb ekt
ICc2 þ b þ iCc2 þ ekt
Cc2 Cb1 Cb1 Cc2 Cb1 Cb1
¼ Leplg y3 k ekt þ iRce k ekt Rce Lcplg y4 k ekt

1 1 iL ekt
iCc2 þ ekt b ¼ Leplg y3 k ekt þ iRce k ekt Rce Lcplg
Cc2 Cb1 Cb1
y4 k ekt
We divide the two side of the above equation by ekt term.

1 1 iLb
iCc2 þ ¼ Leplg y3 k þ iRce k Rce Lcplg y4 k
Cc2 Cb1 Cb1
dIRload ðjÞ
ILb ¼ Cb2 Rload þ IRload ) ILb þ iLb ekt
dt
ðjÞ
¼ Cb2 Rload iRload k ekt þ IRload þ iRload ekt
@ fixed point ILb ¼ IRload ; ILb ¼ IC c2 ðC1c2 þ Cb1 Þ

1
Cb1
n o
ðjÞ ðjÞ
ILb IRload þ iLb ekt ¼ Cb2 Rload iRload k ekt þ iRload
n o
ðjÞ ðjÞ
ekt ; ILb IRload jI ¼I
Lb Rload
¼0
Cb2 Rload iRload k ekt ¼ iLb ekt iRload ekt ) iLb iRload Cb2 Rload iRload k
¼0
We can summery our Small Signal (SS) amplifier with matching network (PI
input and output matching networks) small increments elements:
y1 iLa k ¼ 0; y2 iL3 k ¼ 0; y3 iLeplg k ¼ 0; y4 iLcplg k ¼ 0
iCc1 ¼ iRs k CC1 Rs CC1 La y1 k iRb2 k CC1 Rb2 CC1 L3 y2 k

1 1 iL
iCc2 þ b ¼ Leplg y3 k þ iRce k Rce Lcplg y4 k
Cc2 Cb1 Cb1
iLb iRload Cb2 Rload iRload k ¼ 0
Assumption: We consider for simplicity that arbitrarily small increments

elements iRs ; y1 ; iRb2 ; y2 are proximally the same iRs y1 iRb2 y2 ! z1 and we
represent them as small increment element z1 . Additionally, arbitrarily small
increments elements y3 ; iRCe ; y4 are proximally the same y3 iRCe y4 ! z2 and we
represent them as small increment element z2 .
z1 iLa k ¼ 0; z2 iL3 k ¼ 0; z3 iLeplg k ¼ 0; z4 iLcplg k ¼ 0
1
iCc1 z1 k ¼ 0
ðCC1 Rs þ CC1 La þ CC1 Rb2 þ CC1 L3 Þ

1
Cc2 þ 1
Cb1 i Lb
iCc2 z2 k ¼ 0
ðLeplg þ Rce Lcplg Þ Cb1 ðLeplg þ Rce Lcplg Þ
iLb iRload Cb2 Rload iRload k ¼ 0

1 1
) i Lb iRload iRload k
Cb2 Rload Cb2 Rload
¼0
Small Signal (SS) amplifier with matching network system matrixes:

0 1
i La
B i C
B L3 C
0 1 B C 0 1 0 1
i11 . . . i17 B iLeplg C u11 . . . u13 iCc1
B C
B .. .. .. C B C B .. .. .. C B i C ¼ 0
@ . . . AB iL C þ @ .
B cplg C . . A @ Cc2 A
i71 i77 B C u71 u73 iLb
B z1 C
B C
@ z2 A
iRload
i11 ¼ k; i12 ¼ i13 ¼ i14 ¼ 0; i15 ¼ 1; i16 ¼ i17 ¼ 0; i21 ¼ 0; i22 ¼ k;
i23 ¼ i24 ¼ 0; i25 ¼ 1; i26 ¼ i27 ¼ 0
i31 ¼ i32 ¼ 0; i33 ¼ k; i34 ¼ i35 ¼ 0; i36 ¼ 1; i37 ¼ 0; i41 ¼ i42 ¼ i43 ¼ 0
i44 ¼ k; i45 ¼ 0; i46 ¼ 1; i47 ¼ 0; i51 ¼ . . . ¼ i54 ¼ 0; i55 ¼ k; i56 ¼ i57 ¼ 0
1
i61 ¼ i65 ¼ 0; i66 ¼ k; i67 ¼ 0; i71 ¼ . . . ¼ i76 ¼ 0; i77 ¼ k
Cb2 Rload
u11 ¼ u12 ¼ u13 ¼ 0; u21 ¼ u22 ¼ u23 ¼ 0; u31 ¼ u32 ¼ u33 ¼ 0; u41 ¼ u42
¼ u43 ¼ 0
1
u51 ¼ ; u ¼ u53 ¼ 0
ðCC1 Rs þ CC1 La þ CC1 Rb2 þ CC1 L3 Þ 52
ðC1c2 þ Cb1 Þ
1
1
u61 ¼ 0; u62 ¼ ; u63 ¼
ðLeplg þ Rce Lcplg Þ Cb1 ðLeplg þ Rce Lcplg Þ
1
u71 ¼ u72 ¼ 0; u73 ¼
Cb2 Rload
0 1
i La
B i C
B L3 C
0 1 0 1 B C
u11 ... u13 i11 . . . i17 B iLeplg C
B C
B .. .. .. C ! e; B .. .. . C B C
We consider @ . . . A @ . . .. A B
B
iLcplg C 0
C
u71 u73 i71 i77 B z C
B 1C
B C
@ z2 A
iRload
0 1 0 1
i11 ... i17 i11 . . . i17
B .. .. .. C B .. .. . C
AkI ¼@ . . . A; detðA k IÞ ¼ 0 ) det@ . . .. A ¼ 0
i71 i77 i71 i77

[2–4].
4.3 Small Signal (SS) Amplifiers Matching Network Stability Analysis … 435
4.3 Small Signal (SS) Amplifiers Matching Network

Stability Analysis Under Microstrip Parasitic
Parameters Variation
In our stability analysis of Small Signal (SS) amplifiers which include input and
output matching networks, we need to consider the microstrip lines that connect our
input RF source to input matching network and input matching network to Small
Signal (SS) amplifier. Additionally there is a microstrip lines between Small Signal
(SS) amplifier to the output matching network and between output matching net-
work to the load (Rload). We represent in our stability analysis the microstrip lines
as a parasitic delay lines in time. We define our microstrip line’s delay parameters
as s1, s2, s3, s4 respectively. Under AC and Small Signal (SS) conditions, the BJT
can be replaced with linear Hybrid PI model, input and output matching circuits, RF
input source, load resistance and microstrip delay lines. The amplifier matching
networks types are L matching network in our analysis [25, 26, 33] (Fig. 4.10).
Remark: microstrip lines have many parasitic effects. We neglect all those effects
and concentrate on representation of microstrip line as a delay line. We consider
that the voltage on microstrip delay line is very small and the assumption that the
current flow through each microstrip is delay in time.
If we define the voltage on (i) delay line as Vsi (i = 1, 2, 3…). Vsi ! e . If we
define the current that flow through microstrip as a IðtÞ then the effect of parasitic
delay in time is Iðt si Þ.IðtÞ ! Iðt si Þ. We consider the following:
Tau1 = s1, Tau2 = s2, Tau3 = s3, Tau4 = s4, Vs1 ¼ ðVA11 VA12 Þ ! e;
Vs2 ¼ ðVA21 VA22 Þ ! e
Fig. 4.10 Small Signal (SS) amplifier with matching networks (Hybrid PI equivalent circuit for
BJT transistor)
Vs3 ¼ ðVA111 VA112 Þ ! e; Vs4 ¼ ðVA121 VA122 Þ ! e;

ðVA11 VA12 Þ ! e ) VA1 ¼ VA11 ¼ VA12
ðVA21 VA22 Þ ! e ) VA2 ¼ VA21 ¼ VA22 ;
ðVA111 VA112 Þ ! e ) VA11 ¼ VA111 ¼ VA112
ðVA121 VA122 Þ ! e ) VA12 ¼ VA121 ¼ VA122
Xs ðtÞ VA1 dVA1 dIL d

IR s ¼ ; I C1 ¼ C 1 ; VA1 VA2 ¼ L1 1 ; ICC1 ¼ CC1 ðVA2 VA3 Þ
Rs dt dt dt
dIL VA
IL1 ðt s2 Þ ¼ ICc1 ðtÞ; VA3 VA4 ¼ L3 3 ; IRb2 ¼ 4 ; IL3 ¼ IRb2 ;
dt Rb2
dIL2 VA13 VA9
VA3 VA13 ¼ L2 ; IRb1 ¼
dt Rb1
dILbplg VA VA 6 VA
VA3 VA5 ¼ Lbplg ; IRbbp ¼ 5 ; ILbplg ¼ IRbbp ; IRc ¼ 9 ;
dt Rbbp Rc
Vpi ¼ VA6 VA8 ¼ VCe
dVpi VA VA 8 dðVA6 VA7 Þ VA VA7
I Ce ¼ C e ; IRbpe ¼ 6 ; I Cc ¼ C c ; IRbpc ¼ 6 ;
dt Rbpe dt Rbpc
VA VA 8
IRce ¼ 7
Rce
dIL dIL dLLc
VA8 ¼ Leplg eplg ; VA7 VA10 ¼ Lcplg cplg ; VA9 VA10 ¼ Lc ;
dt dt dt
dðVA10 VA11 Þ
ICc2 ¼ Cc2
dt
dVA11 dIL Vout VA

I C2 ¼ C 2 ; VA11 VA12 ¼ L4 4 ; VA12 ¼ Vout ; IL4 ¼ IRload ; IRload ¼ ¼ 12
dt dt Rload Rload
We can write our circuit Kirchhoff’s Current Law (KCL) for all nodes:
Table 4.3 Small Signal (SS) amplifier with matching networks circuit Kirchhoff’s Current Law
(KCL) for all nodes
Node number KCL @ Ai Node number KCL @ Ai
A1 IRs ðt s1 Þ ¼ IL1 ðtÞ þ IC1 ðtÞ A8 ICe þ IRbpe þ gm Vpi þ IRce ¼ ILeplg
(A1 = A1-1 = A1-2)
A2 IL1 ðt s2 Þ ¼ ICc1 ðtÞ A9 IRb1 ¼ IRc þ ILc
(A2 = A2-1 = A2-2)
A4 IL3 ¼ IRb2 A11 ICc2 ðt s3 Þ ¼ IL4 ðtÞ þ IC2 ðtÞ
(A11 = A11-1 = A11-2)
A5 ILbplg ¼ IRbbp A12 IL4 ðt s4 Þ ¼ IRload ðtÞ
(A12 = A12-1 = A12-2)
Xs ðtÞ VA1
Rs
dVA1 dXs ðtÞ dIRs
IC1 ¼ C1 ¼ C1 ð Rs Þ; VA1 VA2
dt dt dt
dIL dIL
¼ L1 1 ) VA2 ¼ VA1 L1 1
dt dt

dIL1 dVA2 dVA3
VA2 ¼ Xs ðtÞ IRs Rs L1 ; ICC1 ¼ CC1
dt dt dt
dXs ðtÞ dIRs d 2 IL1 dVA3
¼ CC1 Rs L1 2
dt dt dt dt
VA4 dIL
IRb2 ¼ ) VA4 ¼ IRb2 Rb2 ; VA3 VA4 ¼ L3 3 ) VA3
Rb2 dt
dIL3 dVA3 dIRb2 d 2 IL
¼ IRb2 Rb2 þ L3 ; ¼ Rb2 þ L3 2 3
dt dt dt dt
dXs ðtÞ dIRs d 2 IL1 dIRb2 d 2 IL
I CC 1 ¼ CC1 ð Rs L1 2 Rb2 L3 2 3 Þ
dt dt dt dt dt
IL1 ðt s2 Þ ¼ ICc1 ðtÞ; IL1 ðtÞ¼IRs ðt s1 Þ IC1 ðtÞ; ICc1 ðtÞ ¼ IL1 ðt s2 Þ
¼ IRs ðt s1 s2 Þ IC1 ðt s2 Þ
P
2
ICc1 ðtÞ ¼ IL1 ðt s2 Þ ¼ IRs ðt si Þ IC1 ðt s2 Þ System equation No. 1:
i¼1
!
X
2
IRs t si IC1 ðt s2 Þ
i¼1

¼ CC1 Rs L1 2 Rb2 L3 2
dt dt dt dt dt
dIL2 dIL dIL dIL

dt dt dt dt
VA13 VA9
Rb1
dIL3 dIL
dt dt
dILbplg dIL dIL
dt dt dt
VA5 VA6
Rbbp
dIL dIL
¼ IRb 2 Rb2 þ L3 3 Lbplg bplg IRbbp Rbbp
dt dt
VA9 VA VA8
IRc ¼ ) VA9 ¼ IRc Rc ; IRbpe ¼ 6 ) IRbpe Rbpe
Rc Rbpe
¼ VA6 VA8 ) VA8 ¼ VA6 IRbpe Rbpe
dIL3 dIL
VA8 ¼ VA6 IRbpe Rbpe ¼ IRb2 Rb2 þ L3 Lbplg bplg IRbbp Rbbp
dt dt
IRbpe Rbpe
dðVA6 VA7 Þ VA VA7 dIR

I Cc ¼ C c ; IRbpc ¼ 6 ) ICc ¼ Cc Rbpc bpc ;
dt Rbpc dt
VA7 VA8 ¼IRce Rce
dIL3 dIL
dt dt
Rbpe þ IRce Rce
dILeplg dIL dIL

dt dt dt
dILeplg
¼ Leplg
dt
dILcplg dIL
dt dt
dIL dIL
dt dt
dILcplg
Lcplg
dt
VA9 VA10 ¼ Lc ) VA9 ¼ VA10 þ Lc ; ¼ þ 2
dt dt dt Cc2 C2
dIL dIL
dt dt
dILcplg dLLc
Lcplg þ Lc
dt dt

ICc2 ¼ Cc2 ) ICc2 ¼ Cc2 ð Þ ) ICc2
dt dt dt
dVA10 IC2
¼ Cc2 ð Þ
dt C2

ICc2 IC d dIL dIL
þ 2¼ IRb2 Rb2 þ L3 3 Lbplg bplg IRbbp Rbbp IRbpe Rbpe
Cc2 C2 dt dt dt
dILcplg
þ IRce Rce Lcplg
dt
ICc2 IC2 dIRb2 d 2 I L3 d 2 ILbplg dIRbbp dIR
þ ¼ Rb2 þ L3 2 Lbplg 2
Rbbp bpe
Cc2 C2 dt dt dt dt dt
dIRce d 2 ILcplg
Rbpe þ Rce Lcplg
dt dt2
ICc2 ðt s3 Þ ¼ IL4 ðtÞ þ IC2 ðtÞ ) IC2 ðtÞ ¼ ICc2 ðt s3 Þ IL4 ðtÞ
ICc2 IC ðt s3 Þ IL4 ðtÞ dIRb2 d 2 IL d 2 ILbplg

þ c2 ¼ Rb2 þ L3 2 3 Lbplg
Cc2 C2 C2 dt dt dt2
dIRbbp dIR dIRce d 2 ILcplg
Rbbp bpe Rbpe þ Rce Lcplg
dt dt dt dt2
dIL4 dIL VA
VA11 VA12 ¼ L4 ) VA11 ¼ VA12 þ L4 4 ; IRload ¼ 12 ) VA12
dt dt Rload
¼ IRload Rload
System equation No. 4: IRload ðtÞ ¼ IL4 ðt s4 Þ; IC2 ðtÞ ¼ ICc2 ðt s3 Þ IL4 ðtÞ

dIL4 dIRload d 2 IL
VA11 ¼ IRload Rload þ L4 ; IC2 ¼ C2 Rload þ L4 2 4
dt dt dt

dIL4 ðt s4 Þ d 2 I L4
ICc2 ðt s3 Þ IL4 ðtÞ ¼ C2 Rload þ L4 2
dt dt

!
X
2
IR s t si IC1 ðt s2 Þ
i¼1

¼ CC1 Rs L1 2 1 b2 Rb2 L3 2 3
dt dt dt dt dt
dIL dI L dIL
IRb2 Rb2 þ L3 3 Lbplg bplg IRbbp Rbbp IRbpe Rbpe ¼ Leplg eplg
dt dt dt
ICc2 IC ðt s3 Þ IL4 ðtÞ dIRb2 d 2 IL d 2 ILbplg

þ c2 ¼ Rb2 þ L3 2 3 Lbplg
Cc2 C2 C2 dt dt dt2
2
dIR dIR dIRce d ILcplg
bbp Rbbp bpe Rbpe þ Rce Lcplg
dt dt dt dt2

dIL4 ðt s4 Þ d 2 I L4
ICc2 ðt s3 Þ IL4 ðtÞ ¼ C2 Rload þ L4 2
dt dt

d dIL3 dILbplg d 2 ILeplg
IRb2 Rb2 þ L3 Lbplg IRbbp Rbbp IRbpe Rbpe ¼ Leplg
dt dt dt dt2
dIRb2 d 2 IL d 2 ILbplg dIRbbp dIR d 2 ILeplg

Rb2 þ L3 2 3 Lbplg 2
Rbbp bpe Rbpe ¼ Leplg
dt dt dt dt dt dt2
We can merge our system two differential equations to one differential equation:
ICc2 IC ðt s3 Þ IL4 ðtÞ d 2 ILeplg dIRce d 2 ILcplg

þ c2 ¼ Leplg þ R ce L cplg
Cc2 C2 C2 dt2 dt dt2
We can summery our system three differential equations:
X
2
IRs ðt si Þ IC1 ðt s2 Þ
i¼1

¼ CC1 Rs L1 2 Rb2 L3 2
dt dt dt dt dt
ICc2 IC ðt s3 Þ IL4 ðtÞ d 2 ILeplg dIRce d 2 ILcplg

þ c2 ¼ Leplg þ R ce L cplg
Cc2 C2 C2 dt2 dt dt2

dIL4 ðt s4 Þ d 2 IL
ICc2 ðt s3 Þ IL4 ðtÞ ¼ C2 Rload þ L4 2 4
dt dt
To find our equilibrium points (fixed points) of the Small Signal (SS) amplifier

P2
with L—matching networks is by lim IRs t si ¼ IRs ðtÞ;
t!1 i¼1
lim ICc2 ðt s3 Þ ¼ ICc2 ðtÞ
t!1
X
2
lim IC1 ðt s2 Þ ¼ IC1 ðtÞ; lim IL4 ðt s4 Þ ¼ IL4 ðtÞ; t s2 ; t s4 ; t si ; t s 3
t!1 t!1
i¼1
dIRs dIR dIR dIL ðt s4 Þ dIL d 2 IL
¼ 0; b2 ¼ 0; ce ¼ 0; 4 ¼ 0; 1 ¼ 0 ) 2 1 ¼ 0
dt dt dt dt dt dt
dIL3 d 2 I L3 dILeplg d 2 ILeplg dILcplg d 2 ILcplg
¼ 0 ) 2 ¼ 0; ¼0) ¼ 0; ) ¼0
dt dt dt dt2 dt dt2
dIL4 d 2 IL dXs ðtÞ 1 1 I
¼ 0 ) 2 4 ¼ 0; IRs IC 1 ¼ CC1 ; ICc2 ð þ Þ L4 ¼ 0
dt dt dt Cc2 C2 C2
dX ðtÞ
IC c2 IL4 ¼ 0 ) IC c2 ¼ IL4 ;
s
!e
dt
ICc1 ¼ IL2 þ ILbplg þ IL3 ¼ IL2 þ IRbbp þ IL3 ¼ IL2 þ ðIRbpc þ ICc Þ þ IRbpe þ ICe þ IL3
ICc þ IRbpc ¼ gm Vpi þ IRce þ ILcplg ; ICc1 ¼ IL2 þ ðIRbpc þ ICc Þ þ IRbpe þ ICe þ IL3
ICc1 ¼ IL2 þ gm Vpi þ IRce þ ILcplg þ IRbpe þ ICe þ IL3 ; ICe þ IRbpe þ gm Vpi þ IRce ¼ ILeplg
ICe þ IRbpe þ gm Vpi ¼ ILeplg IRce ; ICc1 ¼ IL2 þ ILeplg IRce þ IRce þ ILcplg þ IL3
ICc1 ¼ IL2 þ ILeplg þ ILcplg þ IL3 ; ICc2 ¼ILc þ ILcplg ¼ IRb1 IRc þ ILcplg ;
IRb1 ¼ IRc þ ILc ) ILc ¼ IRb1 IRc
ICc2 ¼ILc þ ILcplg ¼ IRb1 IRc þ ILcplg ¼ IL2 IRc þ ILcplg ; IL2 ¼ICc1 ILbplg IL3
ICc2 ¼ IL2 IRc þ ILcplg ¼ ICc1 ILbplg IL3 IRc þ ILcplg ; ICc1 ðtÞ¼IL1 ðt s2 Þ
ICc2 ¼ IL1 ðt s2 Þ ILbplg IL3 IRc þ ILcplg ; ILbplg ¼IRbbp ¼ IRbpc þ ICc þ IRbpe þ ICe
ILbplg ¼ ðIRbpc þ ICc Þ þ IRbpe þ ICe ; ICc þ IRbpc ¼ gm Vpi þ IRce þ ILcplg
ILbplg ¼ ðgm Vpi þ IRce þ ILcplg Þ þ IRbpe þ ICe ; ICe þ IRbpe þ gm Vpi þ IRce ¼ ILeplg
ILbplg ¼ ðgm Vpi þ IRce þ IRbpe þ ICe Þ þ ILcplg ¼ ILeplg þ ILcplg ; ILbplg ¼ ILeplg þ ILcplg
ICc2 ¼ IL1 ðt s2 Þ ILbplg IL3 IRc þ ILcplg ¼ IL1 ðt s2 Þ ILeplg ILcplg
IL3 IRc þ ILcplg
ICc2 ¼ IL1 ðt s2 Þ ILeplg IL3 IRc ; ICc2 ðt s3 Þ

¼ IL1 ðt s2 s3 Þ ILeplg ðt s3 Þ IL3 ðt s3 Þ IRc ðt s3 Þ
!
X
3
ICc2 ðt s3 Þ ¼ IL1 t si ILeplg ðt s3 Þ IL3 ðt s3 Þ IRc ðt s3 Þ; IRb2 ¼IL3
i¼2
We can write our system set of differential equations:

!
X
2
dXs ðtÞ dIRs d 2 IL dIL d 2 IL
I Rs t si IC1 ðt s2 Þ ¼ CC1 Rs L1 2 1 3 Rb2 L3 2 3
i¼1
dt dt dt dt dt
1 1 X3
½IL1 ðt s2 Þ ILeplg IL3 IRc þ ½IL1 ðt si Þ ILeplg ðt s3 Þ IL3 ðt s3 Þ
Cc2 C2 i¼2
IL4 ðtÞ d 2 ILeplg dIRce d 2 ILcplg
IRc ðt s3 Þ ¼ Leplg þ Rce L cplg
C2 dt2 dt dt2
X
3
IL1 ðt si Þ ILeplg ðt s3 Þ IL3 ðt s3 Þ IRc ðt s3 Þ IL4 ðtÞ

i¼2
¼ C2 Rload þ L4 2 4
dt dt
Remark: Some of our system variables include time argument X(t) and other
without time argument. Always consider these system variables as a function of
time.
We consider L13 ¼ L1 L3 ; Lecplg ¼Leplg Lcplg . We define for simplicity of
dIL1 dIL3 d 2 Y2 d 2 IL1 d 2 IL3 dIL3
our analysis new variables: Y1 ¼ dIdtRs ; Y2 ¼ dt þ dt ; dt2 ¼ dt2 þ dt2 ; Y3 ¼ dt .
dIL dILcplg d 2 ILeplg d 2 ILcplg
Y4 ¼ dIdtRce ; Y5 ¼ dteplg dt ; dY5
dt ¼ dt2 dt2 . The system new set of delay
differential equations:
X
2
dXs ðtÞ dY2
IRs ðt si Þ IC1 ðt s2 Þ ¼ CC1 ð Y1 Rs L13 Y3 Rb2 Þ
i¼1
dt dt
1 1 X3
½IL1 ðt s2 Þ ILeplg IL3 IRc þ ½IL1 ðt si Þ ILeplg ðt s3 Þ IL3 ðt s3 Þ
Cc2 C2 i¼2
IL4 ðtÞ dY5
IRc ðt s3 Þ ¼ Lecplg þ Y4 Rce
C2 dt
X
3
IL1 ðt si Þ ILeplg ðt s3 Þ IL3 ðt s3 Þ IRc ðt s3 Þ IL4 ðtÞ
i¼2
¼ C2 ð Rload þ L4 2 4 Þ
dt dt
We can rewrite our new set of system delay differential equations:
dIL4 dY6 d 2 IL4 dIL ðt s4 Þ

Y6 ¼; ¼ 2 ; Y6 ðt s4 Þ¼ 4
dt dt dt dt
!
dY2 1 X2
1 1 1
¼ IRs t si þ IC1 ðt s2 Þ Y1 Rs
dt CC1 L13 i¼1
C C1 L 13 L 13 L13
1 dXs ðtÞ
Y3 Rb2 þ
L13 dt
dY5 1 1 X3
¼ ½IL1 ðt s2 Þ ILeplg IL3 IRc þ ½IL1 ðt si Þ
dt Cc2 Lecplg C2 Lecplg i¼2
IL4 ðtÞ Rce
ILeplg ðt s3 Þ IL3 ðt s3 Þ IRc ðt s3 Þ Y4
C2 Lecplg Lecplg
dIRs dIL dIL3 dIL dIR dIL dIL dIL

¼ Y1 ; 1 þ ¼ Y2 ; 3 ¼ Y3 ; ce ¼Y4 ; eplg cplg ¼ Y5 ; 4 ¼ Y6
We need to find our system equilibrium points (fixed points) based on the new
set of delay differential equations. It is done by setting dYdt2 ¼ 0; dYdt5 ¼ 0; dYdt6 ¼ 0
dIRs dIL dIL3 dIL dIR dIL dIL dIL

¼ 0; 1 þ ¼ 0; 3 ¼ 0; ce ¼0; eplg cplg ¼ 0; 4 ¼ 0
@ fixed points
P
2 P
2
lim IRs ðt si Þ IRs ðtÞ; lim IL1 ðt s2 Þ ¼ IL1 ðtÞ; t si ; t s2
t!1 i¼1 t!1 i¼1
X
3 X
3
lim IC1 ðt s2 Þ ¼ IC1 ðtÞ; lim IL1 ðt si Þ ¼ IL1 ðtÞ; t si ; lim ILeplg ðt s3 Þ ¼ ILeplg ðtÞ
t!1 t!1 t!1
i¼2 i¼2
X
3 X
3
lim IL3 ðt s3 Þ ¼ IL3 ðtÞ; t s3 ; lim IRc ðt s3 Þ ¼ IRc ðtÞ; lim IL1 ðt si Þ ¼ IL1 ðtÞ; t si
t!1 t!1 t!1
i¼2 i¼2
lim Y6 ðt s4 Þ ¼ Y6 ðtÞ; t s4
t!1
1 1 1 dXs ðtÞ
I þ I þ ¼ 0;
CC1 L13 Rs CC1 L13 C1 L13 dt

1 1 ½IL1 ILeplg IL3 IRc IL4
ð þ Þ ¼0
Cc2 C2 Lecplg C2 Lecplg
1
ðI ILeplg IL3 IRc IL4 Þ ¼ 0; Y1 ¼ 0;
C2 L4 L1
Y2 ¼ 0; Y3 ¼ 0; Y4 ¼ 0; Y5 ¼ 0; Y6 ¼ 0
equilibrium points of the Small Signal
(SS) amplifier with L matching networks
consists in adding to coordinate IRs ; IC1 ; IL1 ; IL3 ; ILeplg ; IRc ; IL4 ; Y1 ; Y2 ; Y3 ; Y4 ; Y5 ; Y6
arbitrarily small increments of exponentially form ½iRs ; iC1 ; iL1 ; iL3 ; iLeplg ; iRc ; iL4 ;
y1 ; y2 ; y3 ; y4 ; y5 ; y6 ekt and retaining the first order terms in IRs ; IC1 ; IL1 ; IL3 ; ILeplg ;
IRc ; IL4 ; Y1 ; Y2 ; Y3 ; Y4 ; Y5 ; Y6 [2, 3]. The system of homogenous equations leads to a
polynomial characteristic equation in the eigenvalues. The polynomial character-
istic equations accept by set of the below circuit variables, circuit variables
Small Signal (SS) amplifier with L matching networks equivalent circuit. Our Small
Signal (SS) amplifier with matching L networks equivalent circuit fixed values with
arbitrarily small increments of exponential form ½iRs ; iC1 ; iL1 ; iL3 ; iLeplg ; iRc ;
iL4 ; y1 ; y2 ; y3 ; y4 ; y5 ; y6 ekt are: j = 0(first fixed point), j = 1(second fixed point),
j = 2(third fixed point), etc.,
P 2
ðjÞ ðjÞ
X2
ðjÞ
kðt si Þ
kt kt
Y5 ðtÞ ¼ Y5 þ y5 e ; Y6 ðtÞ ¼ Y6 þ y6 e ; IRs ðt si Þ ¼ IRs þ iRs e i¼1
i¼1
ðjÞ ðjÞ
IC1 ðt s2 Þ ¼ IC1 þ iC1 ekðts2 Þ ; IL1 ðt s2 Þ ¼ IL1 þ iL1 ekðts2 Þ ;
ðjÞ
ILeplg ðtÞ ¼ ILeplg þ iLeplg ekt
P
3
ðjÞ ðjÞ
X
3
ðjÞ
kðt si Þ
kt kt
IL3 ðtÞ ¼ IL3 þ iL3 e ; IRc ðtÞ ¼ IRc þ iRc e ; IL1 ðt si Þ ¼ IL 1 þ iL1 e i¼2
i¼2
ðjÞ kðts3 Þ ðjÞ
ILeplg ðt s3 Þ ¼ ILeplg þ iLeplg e ; IL3 ðt s3 Þ ¼ IL3 þ iL3 ekðts3 Þ ;
ðjÞ
IRc ðt s3 Þ ¼ IRc þ iRc ekðts3 Þ
IL4 ðtÞ
dIRs dIL dIL3
¼ iRs k ekt ; 1 þ ¼ ½iL1 þ iL1 k ekt
dt dt dt
dIL3 dIR dIL dIL

¼ iL3 k ekt ; ce ¼ iRce k ekt ; eplg cplg ¼ ½iLeplg iLcplg k ekt ;
dt dt dt dt
dIL4 ðjÞ ðjÞ
¼ iL4 k ekt ILeplg ¼ ILeplg þ iLeplg ekt ; ILeplg ðt s3 Þ ¼ ILeplg þ iLeplg ekðts3 Þ
dt
dY2 1 X 2
1 1 1
¼ IRs ðt si Þ þ IC1 ðt s2 Þ Y1 Rs
dt C C1 L13 i¼1
CC1 L13 L13 L13
1 dXs ðtÞ
Y3 Rb2 þ
L13 dt
P2
1 kðt si Þ 1
kt ðjÞ ðjÞ
y2 k e ¼ ðIRs þ iRs e i¼1 Þþ ðI þ iC1 ekðts2 Þ Þ
CC1 L13 CC1 L13 C1
1 ðjÞ 1 ðjÞ 1 dXs ðtÞ
ðY1 þ y1 ekt Þ Rs ðY3 þ y3 ekt Þ Rb2 þ
L13 L13 L13 dt
We consider input RF source Xs ðtÞ ¼ A0 þ nðtÞ; jnðtÞj\1 & A0 jnðtÞj

dXs ðtÞjA jnðtÞj
Then Xs ðtÞjA0 jnðtÞj ¼ A0 þ nðtÞ A0 ; dt
0
¼ dnðtÞ
dt ! e.
0 1
P
2
k t si
1 B ðjÞ C
y2 k ekt ¼ @IRs þ iRs e i¼1
A
CC1 L13
1 1
ðjÞ ðjÞ
þ IC1 þ iC1 ekðts2 Þ ðY1 þ y1 ekt Þ Rs
CC1 L13 L13
1 ðjÞ
Y3 þ y3 ekt Rb2
L13

P 2
k t si
1 ðjÞ 1 1 ðjÞ
y2 k ekt ¼ I iR e i¼1
þ I
CC1 L13 Rs CC1 L13 s CC1 L13 C1
1
þ iC1 ekðts2 Þ
CC1 L13
1 ðjÞ Rs Rb ðjÞ Rb
Y1 Rs y1 ekt 2 Y3 2 y3 ekt
L13 L13 L13 L13
1 ðjÞ 1 ðjÞ 1 ðjÞ Rb ðjÞ
y2 k ekt ¼ I þ I Y Rs 2 Y3
CC1 L13 Rs CC1 L13 C1 L13 1 L13

P2
k t si
1
iRs e i¼1
CC1 L13
1 Rs Rb
þ iC1 ekðts2 Þ y1 ekt 2 y3 ekt
CC1 L13 L13 L13
ðjÞ ðjÞ ðjÞ R ðjÞ

@ fixed point CC 1L13 IRs þ 1
CC1 L13 IC1 L113 Y1 Rs Lb132 Y3 ¼ 0
1
P2
1 kðt si Þ 1 Rs
kt
y2 k e ¼ iRs e i¼1 þ iC1 ekðts2 Þ y1 ekt
CC1 L13 CC1 L13 L13
Rb
2 y3 ekt
L13
Dividing two sides of the above by ekt term
P
2
1 k si 1 Rs Rb
y2 k iRs e i¼1 þ iC1 eks2 y1 2 y3 ¼ 0
CC1 L13 CC1 L13 L13 L13
dY5 1 1 X3
¼ ½IL1 ðt s2 Þ ILeplg IL3 IRc þ ½IL1 ðt si Þ
dt Cc2 Lecplg C2 Lecplg i¼2
IL4 ðtÞ Rce
ILeplg ðt s3 Þ IL3 ðt s3 Þ IRc ðt s3 Þ Y4
C2 Lecplg Lecplg
1 h
ðjÞ ðjÞ
y5 k ekt ¼ IL1 þ iL1 ekðts2 Þ ILep lg þ iLep lg ekt
Cc2 Lecplg
i
ðjÞ ðjÞ
IL3 þ iL3 ekt IRc þ iRc ekt
0 P3 1
1 kðt si Þ
ðjÞ A ILðjÞ þ iLep lg ekðts3 Þ
þ ½@IL1 þ iL1 e i¼2
C2 Lecplg ep lg

ðjÞ
IL3 þ iL3 ekðts3 Þ

ðjÞ
IL4 þ iL4 ekt R
ðjÞ ðjÞ
IRc þ iRc ekðts3 Þ Y4 þ y4 ekt
ce
C2 Lecplg Lecplg

kt 1 ðjÞ 1 kðts2 Þ
y5 k e ¼ I þ iL e
Cc2 Lecplg L1 Cc2 Lecplg 1

1 ðjÞ 1
ILeplg þ iLeplg ekt
Cc2 Lecplg Cc2 Lecplg

1 ðjÞ 1
IL 3 þ iL3 ekt

1 ðjÞ 1 kt
I þ iR e
Cc2 Lecplg Rc Cc2 Lecplg c
0 P3 1
1 1 kðt si Þ
ðjÞ
þ@ I þ iL e i¼2 A
C2 Lecplg L1 C2 Lecplg 1

1 ðjÞ 1
ILeplg þ iLeplg ekðts3 Þ
C2 Lecplg C2 Lecplg

1 ðjÞ 1 kðts3 Þ
I þ iL e
C2 Lecplg L3 C2 Lecplg 3

1 ðjÞ 1
IR c þ iRc ekðts3 Þ
C2 Lecplg C2 Lecplg
ðjÞ
IL 4 iL4 ekt ðjÞ Rce Rce
Y4 þ y4 ekt
C2 Lecplg C2 Lecplg Lecplg Lecplg

y5 k ekt ¼ f I I I
Cc2 Lecplg L1 Cc2 Lecplg Leplg Cc2 Lecplg L3
1 ðjÞ 1 ðjÞ
I þ I
Cc2 Lecplg Rc C2 Lecplg L1
I I I
C2 Lecplg Leplg C2 Lecplg L3 C2 Lecplg Rc
ðjÞ
IL 4 ðjÞ Rce 1
Y4 gþ iL ekðts2 Þ
C2 Lecplg Lecplg Cc2 Lecplg 1
1 1
iL ekt iL ekt
Cc2 Lecplg eplg Cc2 Lecplg 3
P3
1 1 kðt si Þ
iRc ekt þ iL1 e i¼2
Cc2 Lecplg C2 Lecplg
1 1
iL ekðts3 Þ iL ekðts3 Þ
C2 Lecplg eplg C2 Lecplg 3
1
iR ekðts3 Þ
C2 Lecplg c
iL4 ekt Rce
y4 ekt
C2 Lecplg Lecplg
@ fixed point
1 ðjÞ 1 ðjÞ 1 ðjÞ 1 ðjÞ

f I I I I
Cc2 Lecplg L1 Cc2 Lecplg Leplg Cc2 Lecplg L3 Cc2 Lecplg Rc
þ I L1 ILeplg I L3 I
C2 Lecplg C2 Lecplg C2 Lecplg C2 Lecplg Rc
ðjÞ
I L4 ðjÞ Rce
Y4 g¼0
C2 Lecplg Lecplg
1 1
y5 k ekt ¼ iL ekðts2 Þ iL ekt
Cc2 Lecplg 1 Cc2 Lecplg eplg
1 1
iL ekt iR ekt
Cc2 Lecplg 3 Cc2 Lecplg c
P3
1 kðt si Þ 1
þ i L1 e i¼2 iL ekðts3 Þ
C2 Lecplg C2 Lecplg eplg
1 1
iL3 ekðts3 Þ iR ekðts3 Þ
C2 Lecplg C2 Lecplg c
iL4 ekt Rce
y4 ekt
C2 Lecplg Lecplg
1 1 1
y5 k ¼ iL eks2 iL iL
Cc2 Lecplg 1 Cc2 Lecplg eplg Cc2 Lecplg 3
P3
1 1 k si
iRc þ iL1 e i¼2
Cc2 Lecplg C2 Lecplg
1 1
iLeplg eks3 iL eks3
C2 Lecplg C2 Lecplg 3
1 i L4 Rce
iRc eks3 y4
C2 Lecplg C2 Lecplg Lecplg
1 1 1
y5 k þ iL eks2 iL iL
Cc2 Lecplg 1 Cc2 Lecplg eplg Cc2 Lecplg 3
P3
1 1 k si 1
iR þ iL e i¼2 iL eks3
Cc2 Lecplg c C2 Lecplg 1 C2 Lecplg eplg
1 1 iL4 Rce
iL eks3 iR eks3 y4 ¼0
C2 Lecplg 3 C2 Lecplg c C2 Lecplg Lecplg
P3
1 k si 1
y5 k þ ðeks2 þ e i¼2 Þ iL1 ð1 þ eks3 Þ iLeplg
1 1 i L4
ð1 þ eks3 Þ iL3 ð1 þ eks3 Þ iRc
C2 Lecplg C2 Lecplg C2 Lecplg
Rce
y4 ¼0
Lecplg
dY6 1 X3
1 1
¼ IL1 ðt si Þ ILeplg ðt s3 Þ IL ðt s3 Þ
dt C2 L4 i¼2
C2 L4 C2 L4 3
1 1 1
IRc ðt s3 Þ IL ðtÞ Y6 ðt s4 Þ Rload
C2 L4 C2 L4 4 L4
P3
1 kðt si Þ 1
ðjÞ ðjÞ
y6 k ekt ¼ ðIL1 þ iL1 e i¼2 Þ ðI þ iLeplg ekðts3 Þ Þ
C2 L4 C2 L4 Leplg
1 ðjÞ 1 ðjÞ
ðI þ iL3 ekðts3 Þ Þ ðI þ iRc ekðts3 Þ Þ
C2 L4 L3 C2 L4 Rc
1 ðjÞ 1 ðjÞ
ðI þ iL4 ekt Þ ðY6 þ y6 ekðts4 Þ Þ Rload
C2 L4 L4 L4
P3
1 1 kðt si Þ 1
kt ðjÞ ðjÞ
y6 k e ¼ I L1 þ iL1 e i¼2 I
C2 L4 C2 L4 C2 L4 Leplg
1 1 ðjÞ 1
iL ekðts3 Þ I iL ekðts3 Þ
C2 L4 eplg C2 L4 L3 C2 L4 3
1 ðjÞ 1 1 ðjÞ 1
IRc iRc ekðts3 Þ I L4 iL ekt
C2 L4 C2 L4 C2 L4 C2 L4 4
Rload ðjÞ Rload
Y6 y6 ekðts4 Þ
L4 L4

y6 k ekt ¼ I I I I
C2 L4 L1 C2 L4 Leplg C2 L4 L3 C2 L4 Rc

1 ðjÞ Rload ðjÞ
I L4 Y6
C2 L4 L4
P3
1 kðt si Þ 1
þ i L1 e i¼2 iL ekðts3 Þ
C2 L4 C2 L4 eplg
1 1
iL ekðts3 Þ iR ekðts3 Þ
C2 L4 3 C2 L4 c
1 Rload
iL4 ekt y6 ekðts4 Þ
C2 L4 L4
@ fixed point
1 ðjÞ 1 ðjÞ 1 ðjÞ 1 ðjÞ 1 ðjÞ

I I I I I
C2 L4 L1 C2 L4 Leplg C2 L4 L3 C2 L4 Rc C2 L4 L4
Rload ðjÞ
Y6 ¼ 0
L4
P3
1 kðt si Þ 1
kt
y6 k e ¼ i L1 e i¼2 iL ekðts3 Þ
C2 L4 C2 L4 eplg
1 1
iL ekðts3 Þ iR ekðts3 Þ
C2 L4 3 C2 L4 c
1 Rload
iL4 ekt y6 ekðts4 Þ
C2 L4 L4
P3
1 kðt si Þ 1
kt
y6 k e þ i L1 e i¼2 iL ekðts3 Þ
C2 L4 C2 L4 eplg
1 1
iL3 ekðts3 Þ iR ekðts3 Þ
C2 L4 C2 L4 c
1 Rload
iL ekt y6 ekðts4 Þ ¼ 0
C2 L4 4 L4
P3
1 k si 1
y6 k þ iL1 e i¼2 iL eks3
C2 L4 C2 L4 eplg
1 1
iL eks3 iR eks3
C2 L4 3 C2 L4 c
1 Rload
iL y6 eks4 ¼ 0
C2 L4 4 L4
dIRs dIL dIL3 dIL dIR

¼ Y1 ; 1 þ ¼ Y2 ; 3 ¼ Y3 ; ce ¼Y4 ;
dt dt dt dt dt
dILeplg dILcplg dIL4
¼ Y5 ; ¼ Y6
dt dt dt
dIRs ðjÞ
¼ Y1 ) iRs k ekt ¼ Y1 þ y1 ekt )Y ðjÞ ¼0 iRs k þ y1 ¼ 0
dt 1
dIL1 dIL3 ðjÞ

þ ¼ Y2 ) iL1 k ekt þ iL3 k ekt ¼ Y2 þ y2 ekt )Y ðjÞ ¼0
dt dt 2
ðiL1 þ iL3 Þ k þ y2 ¼ 0
dIL3 ðjÞ
¼ Y3 ) iL3 k ekt ¼ Y3 þ y3 ekt )Y ðjÞ ¼0 iL3 k þ y3 ¼ 0
dt 3
dIRce ðjÞ
¼Y4 ) iRce k ekt ¼ Y4 þ y4 ekt )Y ðjÞ ¼0 iRce k þ y4 ¼ 0
dt 4
dILeplg dILcplg ðjÞ

¼ Y5 ) iLeplg k ekt iLcplg k ekt ¼ Y5 þ y5 ekt )Y ðjÞ ¼0
dt dt 5
ðiLeplg iLcplg Þ k þ y5 ¼ 0
dIL4 ðjÞ
¼ Y6 ) iL4 k ekt ¼ Y6 þ y6 ekt )Y ðjÞ ¼0 iL4 k þ y6 ¼ 0
dt 6
We can summery our results:
P
2
1 k si 1 Rs Rb
y2 k iRs e i¼1 þ iC1 eks2 y1 2 y3 ¼ 0
CC1 L13 CC1 L13 L13 L13
P 3
1 k si 1
ks2
y5 k þ ðe þ e i¼2 Þ iL1 ð1 þ eks3 Þ iLeplg
1
ð1 þ eks3 Þ iL3
C2 Lecplg
1 i L4 Rce
ð1 þ eks3 Þ iRc y4 ¼0
C2 Lecplg C2 Lecplg Lecplg
P3
Rload ks4 1 k si 1

ðk þ e Þ y6 þ iL1 e i¼2 iL eks3
L4 C2 L4 C2 L4 eplg
1
iL eks3
C2 L4 3
1 1
iR eks3 iL ¼ 0
C2 L4 c C2 L4 4
iRs k þ y1 ¼ 0; ðiL1 þ iL3 Þ k þ y2 ¼ 0; iL3 k þ y3 ¼ 0; iRce k þ y4 ¼ 0

ðiLeplg iLcplg Þ k þ y5 ¼ 0; iL4 k þ y6 ¼ 0; iL3 k þ y3 ¼ 0 ) y3 ¼ iL3 k
ðiL1 þ iL3 Þ k þ y2 ¼ 0 ) iL1 k iL3 k þ y2 ¼ 0 ) iL1 k y3 þ y2 ¼ 0
The small increments of our Small Signal (SS) amplifier with L matching
networks can be divided to two matrixes. The first matrix is (8x8) and the second
matrix is (6x8).
0 1
y2
B C 0 1
B iRs C i C1
B C B C
0 1 B C
B y5 C 0 1 B y1 C
!11 . . . !18 B C N11 ... N16 B C
B C
B .. .. .. C B i C
B L1 C þ B .. .. .. C B y3 C
@ . . . A Bi C @ . . . A B C ¼ 0; !11
B L3 C B iLeplg C
!81 !88 B C N81 N86 B C
B iL4 C B i C
B C @ Rc A
B y C
@ 6A y4
iRce
P2
1 k si
¼ k; !12 ¼ e i¼1
CC1 L13
!13 ¼ !14 ¼ !15 ¼ !16 ¼ !17 ¼ !18 ¼ 0; !21 ¼ 0; !22 ¼ k; !23 ¼ !24 ¼ !25
¼ !26 ¼ !27 ¼ !28 ¼ 0
P
3
1 k si
ks2
!31 ¼ !32 ¼ 0; !33 ¼ k; !34 ¼ ðe þ e i¼2 Þ;
Cc2 Lecplg
1
!35 ¼ ð1 þ eks3 Þ
C2 Lecplg
1
!36 ¼ ; !37 ¼ !38 ¼ 0; !41 ¼ 1; !42 ¼ !43 ¼ 0;
C2 Lecplg
!44 ¼ k; !45 ¼ !46 ¼ !47 ¼ !48 ¼ 0
!51 ¼ !52 ¼ !53 ¼ !54 ¼ 0; !55 ¼ k; !56 ¼ !57 ¼ !58 ¼ 0;

!61 ¼ !62 ¼ !63 ¼ !64 ¼ !65 ¼ 0
P3
1 k si
!66 ¼ k; !67 ¼ 1; !68 ¼ 0; !71 ¼ !72 ¼ !73 ¼ 0; !74 ¼ e i¼2 ;
C2 L4
1
!75 ¼ eks3
C2 L4
1 Rload ks4
!76 ¼ ; !77 ¼ ðk þ e Þ; !78 ¼ 0; !81 ¼ . . . ¼ !87 ¼ 0;
C2 L4 L4
!88 ¼ k
1 Rs Rb
N11 ¼ eks2 ; N12 ¼ ; N13 ¼ 2 ; N14 ¼ N15 ¼ N16 ¼ 0
CC1 L13 L13 L13
N21 ¼ 0; N22 ¼ 1; N23 ¼ N24 ¼ N25 ¼ N26 ¼ 0; N31 ¼ N32 ¼ N33 ¼ 0;
1
N34 ¼ ð1 þ eks3 Þ
Cc2 Lecplg
1 Rce
N35 ¼ ð1 þ eks3 Þ; N36 ¼ ; N41 ¼ N42 ¼ 0;
C2 Lecplg Lecplg
N43 ¼ 1; N44 ¼ N45 ¼ N46 ¼ 0
N51 ¼ N52 ¼ 0; N53 ¼ 1; N54 ¼ N55 ¼ N56 ¼ 0; N61 ¼ . . . ¼ N66 ¼ 0;
N71 ¼ N72 ¼ N73 ¼ 0
1 1
N74 ¼ eks3 ; N75 ¼ eks3 ; N76 ¼ 0; N81 ¼ . . . ¼ N85 ¼ 0;
C2 L4 C2 L4
N86 ¼ 1
0 1
iC1
B C
0 1 B y1 C
N11 . . . N16 B C
B C
B .. .. .. C B y3 C
We consider in our analysis the following: @ . . . AB C!e
B iLeplg C
N61 N66 B C
B i C
@ Rc A
y4
0 1
N11 . . . N16
B . .. .. C
Since N61 ¼ . . . ¼ N66 ¼ 0 then det@ .. . . A¼0
N61 N66
We consider in our stability analysis small increments Jacobian of our Small
Signal (SS) amplifier with L matching networks, first matrix is (8 8) [5, 6].
0 1
y2
B C
B iRs C
B C
0 1 B
B y5 CC
!11 . . . !18 B C
B .. .. .. C B
B
i L1 C
C ¼ 0; A k I
@ . . . AB i L3 C
B C
!81 !88 B C
B i L4 C
B C
B y6 C
@ A
iRce
0 1
!11 . . . !18
B .. .. .. C
¼@ . . . A; detjA k Ij ¼ 0
!81 !88
1 Rload ks4
Dðk; s1 ; s2 ; s3 ; s4 Þ ¼ k8 þ k6 þ k7 e ; s4 ¼ s
C2 L4 L4
1 Rload ks
Dðk; sÞ ¼ k8 þ k6 þ k7 e
C2 L4 L4
We need to get the characteristic equation for stability analysis. We study the
occurrence of any possible stability switching resulting from the increase of value
of the time delay s parameter then Dðk; sÞ ¼ Pn ðk; sÞ þ Qm ðk; sÞ eks .
Pn
The expression for Pn ðk; sÞ is Pn ðk; sÞ ¼ pk ðsÞ kk . The expression for
k¼0
P
m
Qm ðk; sÞ is Qm ðk; sÞ ¼ qk ðsÞ k . Pn¼8 ðkÞ ¼ k8 þ k6 C21L4 ; Qm¼7 ðkÞ¼k7 RLload
k
4
k¼0
1
n ¼ 8; m ¼ 7; n [ m; p0 ðsÞ ¼ . . . ¼ p5 ðsÞ ¼ 0; p6 ðsÞ ¼ ; p7 ðsÞ ¼ 0; p8 ðsÞ
C2 L4
¼1
q0 ðsÞ ¼ . . . ¼ q6 ðsÞ; q7 ðsÞ ¼ RLload 4

. The homogeneous system for IRs ; IC1 ; IL1 ; IL3 ;
ILeplg ; IRc ; IL4 ; Y1 ; Y2 ; Y3 ; Y4 ; Y5 ; Y6 leads to a characteristic equation for the eigen-
P8
value k having the form Pðk; sÞ þ Qðk; sÞ eks ¼ 0; Pðk; sÞ ¼ aj k j Qðk; sÞ ¼
j¼0
P
7
cj k j and the coefficients faj ðqi ; qk ; sÞ; cj ðqi ; qk ; sÞg 2 R depend on qi ; qk and
j¼0
delay s. qi ; qk are any Small Signal (SS) amplifier with L matching network
parameters, other parameters kept as a constant. Unless strictly necessary, the
designation of variation arguments (qi ; qk ) will subsequently be omitted from P, Q,
aj and cj. The coefficients aj, cj are continuous, and differentiable functions of their
arguments. Furthermore P(k), Q(k) are analytic functions of k, for which the fol-
lowing requirements of the analysis (Kuang and Cong 2005; Kuang 1993) can also
be verified in the present case.
(a) If k ¼ i x; x 2 R then Pði xÞ þ Qði xÞ 6¼ 0.
(b) jQðkÞ=PðkÞj is bounded for jkj ! 1; Rek 0. No roots bifurcation from ∞.
(c) FðxÞ ¼ jPði xÞj2 jQði xÞj2 . It has a finite number of zeros; indeed, this is
a polynomial in x.
(d) Each positive root xðqi ; qk Þ of FðxÞ ¼ 0 is continuous and differentiable with
That is for any real number x; Pn ðk ¼ i x; sÞ þ Qm ðk ¼ i x; sÞ 6¼ 0.
1 Rload
Pn ðk ¼ i x; sÞ ¼ x8 x6 ; Qm ðk ¼ i x; sÞ ¼ i x7
C2 L4 L4
1 R load
Pn ðk ¼ i x; sÞ þ Qm ðk ¼ i x; sÞ ¼ x8 x6 i x7 6¼ 0
C2 L4 L4
1 1 R2
jPði xÞj2 ¼ x16 þ x12 2 2 2 x14 ; jQði xÞj2 ¼ x14 load
C2 L4 C2 L4 L24
2
1 1 14 Rload
FðxÞ ¼ jPði xÞj2 jQði xÞj2 ¼ x16 þ x12 2 x 14
x
C22 L24 C2 L4 L24
1 R2 1
FðxÞ ¼ jPði xÞj2 jQði xÞj2 ¼ x16 ½2 þ load x14 þ x12 2 2
C2 L4 L24 C2 L4
1 R2 1
U16 ¼ 1; U14 ¼ ½2 þ load ; U12 ¼ 2 2 ; U2k ¼ 0 8 k ¼ 0; . . .; 5
C2 L4 L42 C2 L4
P
8
Hence FðxÞ ¼ 0 implies U2k x2k ¼ 0 and its roots are given by solving
k¼0
the polynomial. Furthermore
1
PR ði x; sÞ ¼ x8 x6 ; PI ði x; sÞ ¼ 0; QR ði x; sÞ ¼ 0;
C2 L4
Rload
QI ði x; sÞ ¼ x7
L4
sin hðsÞ ¼
jQði xÞj2
cos hðsÞ ¼
jQði xÞj2

½x8 x6 C21L4 x7 RLload 1 1 L4
sin hðsÞ ¼ 4
¼ x ; cos hðsÞ ¼ 0
x14
R2load x C2 L4 Rload
L24
We can use different parameters terminology from our last characteristics

parameters definition: k ! j; pk ðsÞ ! aj ; qk ðsÞ ! cj ; n ¼ 8; m ¼ 7; n [ m.
Additionally
P8 P
7
Pn ðk; sÞ ! PðkÞ; Qm ðk; sÞ ! QðkÞ; PðkÞ ¼ aj k j ; QðkÞ ¼ cj k j
j¼0 j¼0
1 Rload
PðkÞ þ QðkÞ eks ¼ 0; PðkÞ ¼ k8 þ k6 ; QðkÞ ¼ k7
C2 L4 L4
n; m 2 N0 ; n [ m; aj ; cj : R þ 0 ! R are continuous and differentiable function

of s. In the following “—” denotes complex and conjugate. PðkÞ; QðkÞ are analytic
functions in k and differentiable in s. The coefficients aj ; cj as follow
faj ðC2 ; L4 ; Rload;... Þ&cj ðC2 ; L4 ; Rload;... Þg 2 R depend on Small Signal (SS) amplifier
with L matching network system’s C2, L 4, Rload… values.
a0 ¼ . . . ¼ a5 ¼ 0; a6 ¼ C21L4
a7 ¼ 0; a8 ¼ 1; c0 ¼ c1 ¼ . . . ¼ c6 ¼ 0; c7 ¼ RLload
4
. Unless strictly necessary, the
designation of the variation arguments (C2 ; L4 ; Rload;... ) will subsequently be
omitted from P, Q, aj, cj. The coefficients aj, cj are continuous and differential
functions of their arguments. Furthermore P(k), Q(k) are analytic function of s for
which the following requirements of the analysis (Kuang 1993) can also be verified
in the present case (Beretta and Kuang 2002) [5, 6].
(a) If k ¼ i x; x 2 R then Pði xÞ þ Qði xÞ 6¼ 0 that is P and Q have no
common imaginary roots. This condition can verified numerically in the entire
C2 ; L4 ; Rload;... domain of interest.
(b) j QðkÞ
PðkÞ j is bounded for jkj ! 1; Rek 0. No roots bifurcation from ∞. Indeed
Rload
k7
in the limit j QðkÞ
PðkÞ j ¼ j k8 þ k6
L4
1 j.
C2 L4
R2load
(c) FðxÞ ¼ jPði xÞj2 jQði xÞj2 ¼ x16 ½2 C21L4 þ L24
x14 þ x12 C21L2
2 4
It has at most a finite number of zeros. Indeed, this is a polynomial in x
(degree in x16).
(d) Each positive root x C2 ; L4 ; Rload;... of FðxÞ ¼ 0 is continuous and differ-
entiable with respect to C2 ; L4 ; Rload;... this condition can only be assessed
numerically.
In addition, since the coefficients in P and Q are real we have Pði xÞ ¼

Pði xÞ;Qði xÞ ¼ Qði xÞ. k ¼ i x; x [ 0 may be on eigenvalue of charac-
teristic equation. The analysis consists in identifying the roots of characteristic
equation situated on the imaginary axis of the complex k plane, whereby increasing
the parameters C2 ; L4 ; Rload;... Re k may at the crossing, change its sign from (−) to

(+), that is, from a stable focus E IRs ; IC 1 ; IL1 ; IL3 ; . . . to an unstable one, or vice
versa. This feature may be further assessed by examining the sign of the partial
derivatives with respect to C2 ; L4 ; Rload;... and Small Signal (SS) amplifier with L
matching networks parameters.

@Rek @Rek
K1 ðC2 Þ ¼ ; L4 ; Rload ; . . . ¼ const; K1 ðL4 Þ ¼ ;
@C2 k¼ix @L4 k¼ix
C2 ; Rload ; . . . ¼ const
X
@Rek 8
K1 ðRload Þ ¼ ; C2 ; L4 ; . . . ¼ const: Fðx; sÞ ¼ 0 ) U2k x2k ¼ 0:
@Rload k¼ix k¼0
When writing PðkÞ ¼ PR ðkÞ þ i PI ðkÞ; QðkÞ ¼ QR ðkÞ þ i QI ðkÞ and inserting
k ¼ i x; x 2 R into Small Signal (SS) amplifier with L matching networks sys-
tem’s characteristic equation, x must satisfy the following:
jQði xÞj2
jQði xÞj2

Note that FðxÞ can be dependent of s. If s 62 I (assume I
R þ 0 is the set where
xðsÞ is a positive root of FðxÞ and for s 62 I, xðsÞ is not defined then for all s in I,
xðsÞ satisfies that Fðx; sÞ) then there are no positive xðsÞ solutions for Fðx; sÞ ¼ 0
and we cannot have stability switches. For any s 2 I, where xðsÞ is a positive
solution of Fðx; sÞ ¼ 0, we can define the angle hðsÞ 2 ½0; 2 p as the solution of
sin hðsÞ ¼
jQði xÞj2
cos hðsÞ ¼
jQði xÞj2
þ n2p
; n 2 N0 ; s 2 I. Let us introduce the functions
I ! R; Sn ðsÞ ¼ s sn ðsÞ; s 2 I; n 2 N0 that are continuous and differentiable
in s. In the following, the subscripts k; x; C2 ; L4 ; Rload;... indicate the corresponding
partial derivatives. Let us first concentrate on K(x), remember in kðC2 ; L4 ; Rload;... Þ

and xðC2 ; L4 ; Rload;... Þ, and keeping all parameters except one (x) and s. The
derivation closely follows that in reference [BK]. Differentiation Small Signal
(SS) amplifier with L matching networks characteristic equation: PðkÞ þ QðkÞ
eks ¼ 0 with respect to specific parameter (x), and inverting the derivative, for
convenience one calculates:
Remark: x ¼ L4 ; Rload ; C2 ; . . .etc

ð Þ ¼
where Pk ¼ @P
@k …etc., substituting k ¼ i x and bearing Pði xÞ ¼ Pði xÞ and
Qði xÞ ¼ Qði xÞ. Then i Pk ði xÞ ¼ Px ði xÞ;i Qk ði xÞ ¼ Qx ði xÞ and
that on the surface jPði xÞj2 ¼ jQði xÞj2 , one obtain:

ð Þ1 jk¼ix ¼

Q ¼ QR þ i QI
Px ¼ PRx þ i PIx ; Qx ¼ QRx þ i QIx ; Px ¼ PRx þ i PIx ; Qx ¼ QRx þ i QIx ;

P2 ¼ P2R þ P2I
When (x) can be any Small Signal (SS) amplifier with L matching network
parameters C2 ; L4 ; Rload;... and any time delay s etc. Where for convenience, we
dropped the arguments ði x; xÞ, and where
1 Rload
PR ¼ x8 x6 ; PI ¼ 0; QR ¼ 0; QI ¼ x7 ;
C2 L4 L4
1
PRx ¼ 8 x7 6 x5
C2 L4
Rload
PIx ¼ 0; QRx ¼ 0; QIx ¼ 7 x6 ; PRs ¼ 0; PIs ¼ 0; QRs ¼ 0; QIs ¼ 0
L4
1 2
P2 ¼ P2R þ P2I ¼ x12 ½x2
C2 L4
Fx
Fx ¼ 2 ½ðPRx PR þ PIx PI Þ ðQRx QR þ QIx QI Þ; xx ¼
Fx
We define U and V: U ¼ ðPR PIx PI PRx Þ ðQR QIx QI QRx Þ

V ¼ ðPR PIx PI PRx Þ ðQR QIx QI QRx Þ. First we choose our specific
parameter as time delay s. PRx PR ¼ ½8 x7 6 x5 C21L4 ½x8 x6 C21L4
1 1
PRx PR ¼ 2 x11 ½4 x4 7 x2 þ3 ; QRx QR ¼ 0; PIx PI ¼ 0
C2 L4 ðC2 L4 Þ2
2
Rload
QIx QI ¼ 7 x13 ; PR PIx ¼ 0; PI PRx ¼ 0; QR QIx ¼ 0; QI QRx ¼ 0
L4
( " # )
1 1 Rload 2
Fs ¼ 0; Fx ¼ 2 2 x 4 x 7 x
11 4 2
þ3 7x 13
C2 L 4 ðC2 L4 Þ2 L4
( 2 ! )
2 Rload 6
Fx ¼ 2 x11 8 x4 7 x2 þ þ ; xs ¼ 0; V ¼ 0; U ¼ 0
C2 L4 L4 ðC2 L4 Þ2
@x @x Fs @Rek
Fx þ Fs ¼ 0; s 2 I ) xs ¼ ¼ ; K1 ðsÞ ¼ ð Þ
@s @s Fx @s k¼ix
( )
2 ½U þ s jPj2 þ i Fx @Rek
K1 ðsÞ ¼ Re ; signfK 1
ðsÞg ¼ sign
Fs þ i 2 ½V þ x jPj2 @s k¼ix
( )
@x
@x U @s þ V
signfK1 ðsÞg ¼ signfFx g sign s þxþ
@s jPj2
" ( )#
1 2 Rload 2 6
signfK ðsÞg ¼ sign 2 x 8 x 7 x ð
11 4 2
þ½ Þþ signfxg
C 2 L4 L4 ðC2 L4 Þ2
Remark: Since P and Q are independent on s parameter, signfK1 ðsÞg function

Is independent on s parameter and it is only a function of x.
We shall presently examine the possibility of stability transitions (bifurcation) of
Small Signal (SS) amplifier with L matching networks system, about the equilib-
rium point E ðIRs ; IC 1 ; IL1 ; IL3 ; . . .Þ as a result of a variation of delay parameter s or
any other system parameter (C2 ; L4 ; Rload;... ). The analysis consists in identifying the
complex k-plane whereby increasing the delay parameter s or other system’s
parameter, Re k may at the crossing, change its sign from—to +, that is, from a
assessed by examining the sign of the partial derivatives with respect to s or other
system parameter C2 ; L4 ; Rload;... .
It is a reader exercise to find the expressions for signfK1 ðC2 Þg;
signfK1 ðL4 Þg; . . .
And discuss stability switching for different value of system parameter.
Remark: Our system Jacobian is related to part of small increments, so our
stability analysis is not fully implemented to the actual behavior of our system. Our
stability switching analysis is under these assumptions.
4.4 Bias—T Three Port Network Stability Switching

Under Delayed Micro Strip in Time
The function of the bias T is to simultaneously allow a DC bias voltage and an RF

test signal to be applied to the port of a transistor during measurement. In S—
parameter measurement system, the DC bias is applied at the port labeled “DC”,
and the RF test signal from the Vector Network Analyzer (VNA)) is applied to the
port labeled “RF”. At the RF + DC port, both RF and DC voltages are applied to
the device. Basic Bias T schematics: capacitor transfers only RF signal and block
DC. Inductor transfers DC signal and block RF [91–93] (Fig. 4.11)..
A bias T is a three ports network designed to provide power to remote devices,
such as amplifiers, over the same coaxial cable that RF signals are conveyed.
Commercially available bias Ts are available in both “connectorized” and surface
mount versions. These units are typically expensive and, although designed to be
wideband, often suffer in performance at frequencies below 50 MHz. It is con-
sisting of one inductor and one capacitor the bias T circuit is simple, but particular
consideration must be given to component selection. The basic topology and means
of operation, of a bias T network is described in the below figure (Fig. 4.12).
The shunt capacitor (C2) on the DC port should not be considered optional. It
increases isolation between the RF ports and the DC supply connection by routing
any remaining RF leakage on the supply side of the inductor to ground. The circuit
is evaluated both as a single unit and in the intended configuration with two bias Ts
connected together and transferring power. Bias T design considerations: Finding a
Fig. 4.11 Bias-T three ports schematic

4.4 Bias—T Three Port Network Stability Switching … 461
Fig. 4.12 Basic topology of

Bias-T network
DC/RF isolation inductor is the main challenge. The inductor must provide high
reactance across the bands of interest (XL ! ∞; ZL = XL) and it must carry the
required DC current. Inductor’s Q must be high to minimize inductor power dis-
sipation due to the RF signal
ðQ ! 1 ) PL ðaverageÞ ! e ) R ! eÞ. The real component of an inductor
has loss due to the resistive component. The stored energy in the inductor is marked
as EL. The average power dissipated in an inductor is marked as Pavg.
We analyze the stability of Bias T circuit which feed coax cable. For Bias T’s
microstrip lines circuit connections, we represent microstrip as a delay lines in time.
We neglect the voltage on delay lines Vsi ! e then Vsi ! e 8 i ¼ 1; 2; . . .; 5
(Fig. 4.13).
The delay is in the current that flows through microstrip represented delay line
IðtÞ ! Iðt si Þ (Fig. 4.14).
The purpose of the inductor L1 is to prevent the RF signal from entering DC
path, and the purpose of the capacitor C1 is to keep the DC signal from entering the
RF path. The inductor and capacitor should be designed such that the upper cut-off
frequency of the low pass DC path is lower than the lower cut-off frequency of the
high pass RF path. We define Rload as the total resistance seen at the RF + DC port
(purely resistive). The equivalent circuit for the proposed system is as follow
(Fig. 4.15):
Terminology: Tau1 $ s1 ; Tau2 $ s2 ; Tau3 $ s3 ;Tau4 $ s4 ; Tau5 $ s5
Fig. 4.13 Bias-T microstrip delay line in time

Fig. 4.14 Full Bias-T circuit implementation with delay lines
Fig. 4.15 Full Bias-T circuit implementation with delay lines and nodes index
Vs1 ! e ) A2 ¼ A21 ¼ A22 ; Vs2 ! e ) A3 ¼ A31 ¼ A32

fVs3 ! eg&fVs4 ! eg&fVs5 ! eg ) A5 ¼ A51 ¼ A52 ¼ A53 ¼ A54
Vs ðtÞ VA2
VA1 ¼Vs ðtÞ; IRs ¼ ; IC1 ðtÞ ¼ IRs ðt s1 Þ;
Rs
KCL @ A31 ) IC1 ðt s2 Þ þ IL1 ðt s2 Þ ¼ IRload ðtÞ
KCL @ A52 ) IRb ðt s5 Þ ¼ I52a ðtÞ þ I52b ðtÞ; IC2 ¼ I52a ðt s4 Þ; IR1 ðtÞ
¼ I52b ðt s3 Þ; IL1 ðtÞ ¼ IR1 ðtÞ
d d
IC1 ¼ C1 ðVA22 VA31 Þ; VA2 ¼ VA22 ; VA3 ¼ VA31 ; IC1 ¼ C1 ðVA2 VA3 Þ
dt dt
dVA52 dVA5 Vb VA54
IC2 ¼ C2 dt , since A5 ¼ A51 ¼ A52 , IC2 ¼ C2 dt ; IRb ¼ Rb and
Vb VA5 VA51 VA4 VA5 VA4
since A5 ¼ A54 ) IRb ¼ Rb ; IR1 ¼ R1 Since A5 ¼ A51 ) IR 1 ¼
dIL1
R1 VL1 ¼ L1 dt ; VL1 ¼ VA4 VA31 , since VA3 ¼ VA31 ; VL1 ¼ VA4 VA3 ;
dIL1
VA4 VA3 ¼ L1 dt
VA32 VA3
VA3 ¼ VA31 ¼ VA32 ) IRload ¼ ¼
Rload Rload
Remark: we consider in our analysis VA52 [ VA31 then the current flow is from
node A5-2 to node A3-1 otherwise ðVA52 \VA31 Þ the current flow is from node A3-1
to node A5-2.
Vs ðtÞ VA2
IR s ¼ ) Vs ðtÞ VA2 ¼ IRs Rs ) VA2 ¼ Vs ðtÞ IRs Rs
Rs
IC1 ðtÞ ¼ IRs ðt s1 Þ; IC1 ðt s2 Þ þ IL1 ðt s2 Þ
¼ IRload ðtÞ; IRs ðt s1 s2 Þ þ IL1 ðt s2 Þ ¼ IRload ðtÞ
dVA5 dVA5
IC2 ¼ I52a ðt s4 Þ; IC2 ¼ C2 ; I52a ðt s4 Þ ¼ C2 ) VA5
Z dt dt
1
¼ I52a ðt s4 Þ dt
C2
Vb VA5
IRb ¼ ) Vb VA5 ¼ IRb Rb ) VA5 ¼ Vb IRb Rb ; IR1
Rb
VA VA4
¼ 5 ) VA5 VA4 ¼ IR1 R1
R1
VA3 X 2
IRload ¼ ) VA3 ¼ IRload Rload ; VA3 ¼ Rload ½IRs ðt si Þ þ IL1 ðt s2 Þ
Rload i¼1
Z
d 1
IC1 ¼ C1 ðVA2 VA3 Þ ) VA2 VA3 ¼ IC1 dt; VA2 VA3
dt C1
Z
1
¼ IRs ðt s1 Þ dtIRb ðt s5 Þ ¼ I52a ðtÞ þ I52b ðtÞ ) I52b ðtÞ
C1
¼ IRb ðt s5 Þ I52a ðtÞ; I52b ðt s3 Þ ¼ IRb ðt s5 s3 Þ I52a ðt s3 ÞIR1
VA VA4
¼ 5 ) VA5 VA4 ¼ IR1 R1 ; VA5 VA4 ¼ IR1 R1 ;
R1
IR1 ðtÞ ¼ I52b ðt s3 Þ; IR1 ðtÞ ¼ I52b ðt s3 Þ
¼ IRb ðt s5 s3 Þ I52a ðt s3 Þ;
VA5 VA4 ¼ ½IRb ðt s5 s3 Þ I52a ðt s3 Þ R1
We can summery our intermediate equations:
Z
1
VA2 ¼ Vs ðtÞ IRs Rs ;VA5 ¼ I52a ðt s4 Þ dt; VA5 ¼ Vb IRb Rb
C2
" #
X2
VA5 VA4 ¼ IR1 R1 ¼ I52b ðt s3 Þ R1 ; VA3 ¼ Rload IRs ðt si Þ þ IL1 ðt s2 Þ
i¼1
Z
1
VA2 VA3 ¼ IRs ðt s1 Þ dt; VA5 VA4 ¼ ½IRb ðt s5 s3 Þ I52a ðt s3 Þ R1
C1
2 0 1 3
6 B X
5 C 7
VA5 VA4 ¼ 6 B
4IRb @t si C 7
A I52a ðt s3 Þ5 R1
i¼3
i6¼4
Z Z
1 1
VA5 ¼ I52a ðt s4 Þ dt; VA5 ¼ Vb IRb Rb ) I52a ðt s4 Þ dt ¼ Vb IRb Rb
C C2
2Z
d 1 d 1 dVb dIRb dVb
I52a ðt s4 Þ dt ¼ fVb IRb Rb g ) I52a ðt s4 Þ ¼ Rb ; ¼0
dt C2 dt C2 dt dt dt
1 dIR dIRb 1
I52a ðt s4 Þ ¼ b Rb ) ¼ I52a ðt s4 Þ
C2 dt dt C2 Rb
Z
1
Vs ðtÞ IRs Rs VA3 ¼ IRs ðt s1 Þ dt ) VA2
C1
X
2 Z
1
Rload ½IRs ðt si Þ þ IL1 ðt s2 Þ ¼ IRs ðt s1 Þ dt
i¼1
C1
X
2 Z
1
Vs ðtÞ IRs Rs Rload ½IRs ðt si Þ þ IL1 ðt s2 Þ ¼ IRs ðt s1 Þ dt
i¼1
C1
X2 Z
d d 1
fVs ðtÞ IRs Rs Rload ½IRs ðt si Þ þ IL1 ðt s2 Þg ¼ f IRs ðt s1 Þ dtg
dt i¼1
dt C1
P
2
dIRs ðt si Þ
dVs ðtÞ dIRs i¼1 dIL1 ðt s2 Þ IR ðt s1 Þ
Rs Rload ½ þ ¼ s
dt dt dt dt C1
2 0 1 3
6 B X
5 C 7 dIL1
VA4 ¼ VA5 6I
4 @R b
Bt si C 7
A I52a ðt s3 Þ5 R1 ; VA4 ¼ VA3 þ L1 dt
i¼3
i6¼4
" ! !#
dIL1 X5
VA3 þ L1 ¼ VA5 IRb t si I52a ðt s3 R1
dt i¼3
i6¼4
" ! # Z
X
2
dIL 1
Rload IRs t si þ IL1 ðt s2 Þ þ L1 1 ¼ I52a ðt s4 Þ dt
i¼1
dt C2
" !#
X
5
IRb t si Þ I52a ðt s3 R1
i¼3
i6¼4
Z " ! #
1 X
2
I52a ðt s4 Þ dt ¼ Rload IRs t si þ IL1 ðt s2 Þ
C2 i¼1
" !#
dIL1 X5
þ L1 þ IR b t si Þ I52a ðt s3 R1
dt i¼3
i6¼4
Z " ! #
d 1 d X2
I52a ðt s4 Þ dt ¼ fRload IRs t si þ IL1 ðt s2 Þ
dt C2 dt i¼1
dIL1 X5
þ L1 þ ½IRb ðt si Þ I52a ðt s3 Þ R1 g
dt i¼3
i6¼4
P
2
dIRs ðt si Þ
I52a ðt s4 Þ i¼1 dIL1 ðt s2 Þ d 2 IL
¼ Rload þ Rload þ L1 2 1
C2 dt dt dt
P
5
dIRb ðt si Þ
i¼3
i6¼4 dI52a ðt s3 Þ
þ R1 R1
dt dt
P2
dIRs ðt si Þ
2
d IL I52a ðt s4 Þ i¼1 dIL ðt s2 Þ
L1 2 1 ¼ Rload Rload 1
dt C2 dt dt
P5
dIRb ðt si Þ
i¼3
R1 þ R1
dt dt
dIRb
We can summery our equations as follow: dt ¼ C21Rb I52a ðt s4 Þ
P
2
dIRs ðt si Þ
dVs ðtÞ dIRs i¼1 dIL1 ðt s2 Þ IR ðt s1 Þ
Rs Rload ½ þ ¼ s
dt dt dt dt C1
P
2
dIRs ðt si Þ
d IL2
I52a ðt s4 Þ i¼1 dIL1 ðt s2 Þ
L1 2 1 ¼ Rload Rload
dt C2 dt dt
P5
dIRb ðt si Þ
i¼3
R1 þ R1
dt dt
We can merge two differential equations:

2 3
P2
dIRs t si
6 dIL1 ðt s2 Þ7
Rload 6 i¼1
þ 7 ¼ dVs ðtÞ dIRs Rs IRs ðt s1 Þ
4 dt dt 5 dt dt C1
2 3
P2
dIRs t si
d 2 IL1 I52a ðt s4 Þ 6 dIL1 ðt s2 Þ7
L1 ¼ Rload 6
4
i¼1
þ 7
5
dt2 C2 dt dt
0 1
B P
5
C
dIRb @t si A
i¼3
R1 þ R1
dt dt

d 2 IL1 I52a ðt s4 Þ dVs ðtÞ dIRs IRs ðt s1 Þ
L1 ¼ R s
dt2 C2 dt dt C1
0 1
B P
5
C
dIRb @t si A
i¼3
R1 þ R1
dt dt
We can summery our system two differential equations:
dIRb 1
¼ I52a ðt s4 Þ
dt C2 Rb

d 2 IL I52a ðt s4 Þ dVs ðtÞ dIRs IR ðt s1 Þ
L1 2 1 ¼ Rs s
dt C2 dt dt C1
0 1
P 5
dIRb @t si A
i¼3
R1 þ R1
dt dt
dIRb d 2 IL1
Some definitions for new variables: dt ¼ C21Rb I52a ðt s4 Þ; dYdt1 ¼ dt2 ;
dIL1
dt ¼ Y1

dY1 I52a ðt s4 Þ dVs ðtÞ dIRs IRs ðt s1 Þ
L1 ¼ Rs
dt C2 dt dt C1
0 1
B P
5
C
dIRb @t si A
i¼3
R1 þ R1
dt dt
P
5
dIRb ðt si Þ
i¼3
dIL1 i6¼4 ðts3 Þ
At fixed point: dt ¼ 0; dYdt1 ¼ 0; dIdtRs ¼ 0; dt ¼ 0; dI52adt ¼0
X
5
lim I52a ðt s4 Þ ¼ I52a ðtÞ; lim IRs ðt s1 Þ ¼ IRs ðtÞ; lim IRb ðt si Þ ¼ IRb ðtÞ
t!1 t!1 t!1
i¼3
i6¼4
X
5
lim I52a ðt s3 Þ ¼ I52a ðtÞ; t [ [ s4 ; t [ [ s1 ; t [ [ si ; t [ [ s3
t!1
i¼3
i6¼4
j is the Index of system fixed points, first fixed point j = 0, second fixed point j = 1,
third fixed point j = 2, etc.,
dIRb 1 ðjÞ dIL ðjÞ

¼0) I52a ðt s4 Þ ¼ 0 ) I52a ¼ 0; 1 ¼ 0 ) Y1 ¼ 0
dt C2 Rb dt
" ðjÞ
#
ðjÞ
I52a dVs ðtÞ IRs ðjÞ ðjÞ dVs ðtÞ dIRs dI52a
¼ 0 ) I52a þ IRs ¼ C1 ; Y2 ¼ ; Y3 ¼
C2 dt C1 dt dt dt
dIRs ðjÞ dI52a ðjÞ
¼ 0 ) Y2 ¼ 0; ¼ 0 ) Y3 ¼ 0
dt dt
equilibrium
points of the Bias-T circuit which feed coax cable consists in adding to
coordinate Y1 ; Y2 ; Y3; I52a ; IRs arbitrarily small increments of exponentially form
½y1 ; y2 ; y3 ; i52a ; iRs ekt and retaining the first order terms in Y1 ; Y2 ; Y3; I52a ; IRs . The
system of homogenous equations leads to a polynomial characteristic equation in
the eigenvalues. The polynomial characteristic equations accept by set of the below
circuit variables, circuit variables derivative and circuit variables second order
derivative with respect to time into Bias-T circuit which feed coax cable equivalent
circuit. Our Bias-T circuit which feed coax cable equivalent circuit fixed values
with arbitrarily small increments of exponential form ½y1 ; y2 ; y3 ; i52a ; iRs ekt are:
j = 0 (first fixed point), j = 1 (second fixed point), j = 2 (third fixed point), etc.,
ðjÞ ðjÞ
IL1 ¼ IL1 þ iL1 ekt ; Y1 ¼ Y1 þ y1 ekt ; I52a
ðjÞ ðjÞ ðjÞ
¼ I52a þ i52a ekt ; IRs ¼ IRs þ iRs ekt ; IRb ¼ IRb þ iRb ekt
ðjÞ ðjÞ
I52a ðt s4 Þ ¼ I52a þ i52a ekðts4 Þ ; IRs ðt s1 Þ ¼ IRs þ iRs ekðts1 Þ ;
X
5
ðjÞ
X
5
IRb ðt si Þ ¼ IRb þ iRb exp½k ðt si Þ
i¼3 i¼3
i6¼4 i6¼4
dIL1 dY1 dI52a
¼ iL1 k ekt ; ¼ y1 k ekt ;
dt dt dt
kt dIRs kt dIRb
¼ i52a k e ; ¼ iRs k e ; ¼ iRb k ekt
dt dt
dI52a ðt s4 Þ dIR ðt s1 Þ
¼ i52a k ekt eks4 ; s
dt dt !
P 5
dIRb t si
i¼3
i6¼4
¼ iRs k ekt eks1 ;
dt
X
5
¼ iRb k ekt exp½k ðt si Þ
i¼3
i6¼4
ðjÞ dI52a ðt s3 Þ
I52a ðt s3 Þ ¼ I52a þ i52a ekðts3 Þ ;
dt
kt ks3 ðjÞ
¼ i52a k e e ; Y2 ¼ Y2 þ y2 ekt
!
P 5
dIRb t si
i¼3
i6¼4 dIRb X 5
¼ exp½k ðt si Þ
dt dt i¼3
i6¼4
1 X5
dI52a ðt s3 Þ dI52a ks3
¼ I52a ðt s4 Þ exp½k ðt si Þ; ¼ e
C2 Rb i¼3
dt dt
i6¼4
For k < 0, t > 0, the selected fixed point is stable otherwise k > 0, t > 0 is
unstable. Our system tends to the selected fixed point exponentially for k < 0, t > 0
his absolute value |k| established the speed of flow toward or away from the selected
fixed point [5, 6].
dIRb
We can rewrite our system differential equations: dt ¼ C21Rb I52a ðt s4 Þ;
dIL1
dt ¼ Y1
dY1 I52a ðt s4 Þ dVs ðtÞ dIRs IR ðt s1 Þ 1

L1 ¼ þ Rs þ s þ I52a ðt s4 Þ
dt C2 dt dt C1 C2 Rb
X5
dI52a ks3
exp½k ðt si Þ R1 þ e R1
i¼3
dt
i6¼4
2 3
dY1 1 6 R1 X5 7
L1 ¼ I52a ðt s4 Þ 6
41þ exp½k ðt si Þ7
5
dt C2 Rb i¼3
i6¼4
dVs ðtÞ dIRs IR ðt s1 Þ dI52a ks3

þ Rs þ s þ e R1
dt dt C1 dt
dY1 1 R1 X5
¼ I52a ðt s4 Þ ½1 þ exp½k ðt si Þ
dt L1 C2 Rb i¼3
i6¼4
1 dVs ðtÞ dIRs Rs IR ðt s1 Þ dI52a ks3 R1
þ þ s þ e
L1 dt dt L1 C1 L1 dt L1
We define two new variables: Y2 ¼ dIdtRs ; Y3 ¼ dIdt52a
dY1 1 R1 X5
¼ I52a ðt s4 Þ ½1 þ exp½k ðt si Þ þ Y2
dt L1 C2 Rb i¼3
i6¼4
Rs IR ðt s1 Þ R1 1 dVs ðtÞ
þ s þ Y3 eks3
L1 C1 L1 L1 L1 dt
Expression for differential equation: dY2

dt ¼ . . .. . .:
" ! #
X
2
VA3 ¼ Rload IRs t si þ IL1 ðt s2 Þ ; VA2 VA3
i¼1
Z
1
¼ IC1 dt; IC1 ðtÞ ¼ IRs ðt s1 Þ
C1
Z X
2
1
VA2 ¼ IRs ðt s1 Þ dt þ Rload ½IRs ðt si Þ þ IL1 ðt s2 Þ;
C1 i¼1
Vs ðtÞ VA2 1 1
IRs ¼ ¼ Vs ðtÞ VA2
Rs Rs Rs
( Z " ! #)
1 1 1 X2
IRs ¼ Vs ðtÞ IRs ðt s1 Þ dt þ Rload IRs t si þ IL1 ðt s2 Þ
Rs Rs C1 i¼1
8 2 39
> P
2
>
> dIRs ðt si Þ >
dIRs 1 dVs ðtÞ 1 < 1 6 dIL1 ðt s2 Þ7=
¼ I ðt s1 Þ þ Rload 6
4
i¼1
þ 7
5>
dt Rs dt >C 1 R s
Rs > dt dt >
: ;
P
2
dIRs ðt si Þ X5
i¼1 dIRs ðtÞ dIR ðt s1 Þ dIRs ðtÞ ks1
¼ exp½k ðt si Þ; s ¼ e
dt dt i¼3
dt dt
i6¼4
P
2
dIRs ðt si Þ
dIRs 1 dVs ðtÞ 1 Rload i¼1 Rload dIL1 ðt s2 Þ
¼ IRs ðt s1 Þ
dt Rs dt C 1 Rs Rs dt Rs dt
dIRs 1 dVs ðtÞ 1 Rload dIRs ðtÞ X5

Rload dIL1 ðt s2 Þ
¼ IRs ðt s1 Þ exp½k ðt si Þ
dt Rs dt C1 Rs Rs dt i¼3
Rs dt
i6¼4
dIRs Rload X
5
1 dVs ðtÞ 1 Rload dIL1 ðt s2 Þ
½1 þ exp½k ðt si Þ ¼ IRs ðt s1 Þ
dt Rs i¼3
Rs dt C1 Rs Rs dt
i6¼4
We derive the two side of the above equation and get.
d 2 IRs Rload X5
1 d 2 Vs ðtÞ 1 dIR ðtÞ Rload d 2 IL1 ðt s2 Þ
½1 þ exp½k ðt si Þ ¼ s eks1
dt 2 Rs i¼3
Rs dt 2 C 1 Rs dt Rs dt2
i6¼4
dY2 Rload X5
1 d 2 Vs ðtÞ 1 Rload d 2 IL1 ðt s2 Þ
½1 þ exp½k ðt si Þ ¼ Y2 eks1
dt Rs i¼3
R s dt 2 C 1 R s Rs dt2
i6¼4
d 2 IL1 ðts2 Þ P
5
We consider dt2 ! e ¼ 0 then dY2
dt ½1 þ Rload
Rs exp½k ðt si Þ ¼ R1s
i¼3
i6¼4
d 2 Vs ðtÞ
dt2 C11Rs Y2 eks1
dY2 1 d 2 Vs ðtÞ 1
¼ Y2 eks1
dt P
5 dt2 P
5
Rs ½1 þ Rload
Rs exp½k ðt si Þ C1 Rs ½1 þ Rload
Rs exp½k ðt si Þ
i¼3 i¼3
i6¼4 i6¼4
P
2
Case I: s1 ! eð¼ 0Þ; s2 ! eð¼ 0Þ; si ! e
i¼1
Assumption: Our Bias-T circuit’s first and second microstrips parasitic effect
P
2
(delay parameters in time) is minor, s1 ! eð¼ 0Þ; s2 ! eð¼ 0Þ; si ! e
i¼1
X
5
Rload X5
Rload
lim exp½k ðt si Þ ¼ 1; 1 þ exp½k ðt si Þ ¼ 1 þ ;
si !e Rs Rs
i¼1;2 i¼3 i¼3
i6¼4 i6¼4
dY2 1 d Vs ðtÞ
2
1
¼ Y2 eks1
dt Rs ½1 þ Rs
Rload dt2 C1 Rs ½1 þ Rs
Rload
dY1 1 R1 X5
¼ I52a ðt s4 Þ ½1 þ exp½k ðt si Þ
dt L1 C2 Rb i¼3
i6¼4
Rs IR ðtÞ R1 1 dVs ðtÞ
þ Y2 þ s þ Y3 eks3
L1 C1 L1 L1 L1 dt
Expression for differential equation: dY3

dt ¼ . . .. . .:
dIC2 dI52a ðt s4 Þ dI52a ðt s4 Þ dI52a ðtÞ ks4

IC2 ¼ I52a ðt s4 Þ ) ¼ ; ¼ e
dt dt dt dt
dVA5 Vb VA5 dVA5 dVb dIRb dVb
IC 2 ¼ C 2 ; IR b ¼ ) VA5 ¼ Vb IRb Rb ; ¼ Rb ; ¼0
dt Rb dt dt dt dt
dIR dIC2 d 2 IR d 2 IR dI52a ðtÞ ks4
IC2 ¼ C2 Rb b ) ¼ C2 Rb 2 b ; C2 Rb 2 b ¼ e
dt dt dt dt dt
d 2 IR dI52a ðtÞ ks4 d 3 IR d 2 I52a ðtÞ ks4
C 2 Rb 2 b ¼ e ) C2 Rb 3 b ¼ e ;
dt dt dt dt2
d 2 I52a ðtÞ dY3
¼
dt2 dt
d 3 IRb dY3 ks4 d 3 IRb dY3

C2 Rb ¼ e ; 3 ! ðe ¼ 0Þ ) ¼0
dt3 dt dt dt
We can summery our system five Delay Differential Equations (DDE):
dY1 1 R1 X5
¼ I52a ðt s4 Þ ½1 þ exp½k ðt si Þ þ Y2
dt L1 C2 Rb i¼3
i6¼4
Rs IR ðtÞ R1 1 dVs ðtÞ
þ s þ Y3 eks3
L1 C1 L1 L1 L1 dt
dY2 1 d 2 Vs ðtÞ 1 dY3

¼ Y2 eks1 ;
dt Rs ½1 þ Rs
Rload dt 2
C1 Rs ½1 þ Rs
Rload dt
dIRs dI52a
¼ 0; ¼Y2 ; ¼ Y3
dt dt
Remark: Some of our system variables include time argument X(t) and other
without time argument. Always consider these system variables as a function of
time.
We consider RF signal source Vs ðtÞ ¼ A0 þ nðtÞ; jnðtÞj\1 & A0 jnðtÞj [
dVs ðtÞjA0 jnðtÞj dnðtÞ d 2 Vs ðtÞjA0 jnðtÞj

Vs ðtÞjA0 jnðtÞj ¼ A0 þ nðtÞ A0 ; ¼ ! e;
dt dt dt2
d 2 nðtÞ
¼ ! e:
dt2
Under the above consideration we can rewrite our system five Delay Differential
P2
Equations (DDE): s1 ! eð¼ 0Þ; s2 ! eð¼ 0Þ; si ! e; eks1 ! 1
i¼1
2 3
dY1 1 6 R1 X5 7
¼ I52a ðt s4 Þ 61 þ exp½k ðt si Þ7
5 þ Y2
dt L1 C2 4 Rb i¼3
i6¼4
Rs IR ðtÞ R1
þ s þ Y3 eks3
L1 C1 L1 L1
dY2 1 dY3 dIR dI52a

¼ h i Y2 ; ¼ 0; s ¼Y2 ; ¼ Y3
dt C1 Rs 1 þ Rload dt dt dt
Rs
&&&
2 3
dY1 1 6 R1 X5 7
¼ I52a ðt s4 Þ 61 þ exp½k ðt si Þ7
5 þ Y2
dt L1 C2 4 Rb i¼3
i6¼4
Rs IR ðtÞ R1
þ s þ Y3 eks3
L1 C1 L1 L1
2 3
ðjÞ 1 6 R1 X
5 7
y1 k ekt ¼ ½I52a þ i52a ekðts4 Þ 61þ exp½k ðt si Þ7
L1 C2 4 Rb i¼3
5
i6¼4
ðjÞ Rs 1 ðjÞ
þ ½Y2 þ y2 ekt þ ½I þ iRs ekt
L1 C1 L1 Rs
ðjÞ R1
þ ½Y3 þ y3 ekt eks3
L1
2 3
ðjÞ 1 6 R1 X5 7
y1 k ekt ¼ I52a 6
4 1 þ exp½k ðt si Þ7
5
L1 C2 Rb i¼3
i6¼4
Rs
ðjÞ 1 ðjÞ ðjÞ R1
þ Y2 þ I þ Y3 eks3
L1 C1 L1 Rs L1
1 R1 X5
þ ½1 þ exp½k ðt si Þ i52a ekðts4 Þ
L1 C2 Rb i¼3
i6¼4
Rs 1 R1
þ y2 ekt þ iR ekt þ y3 ekt eks3
L1 C1 L1 s L1
ðjÞ P
5
ðjÞ ðjÞ
At fixed point I52a L11C2 ½1 þ R1
Rb exp½k ðt si Þ þ Y2 LR1s þ 1
C1 L1 IRs þ
i¼3
i6¼4
ðjÞ
Y3 eks3 RL11 ¼ 0
1 R1 X5
Rs
y1 k ekt ¼ ½1 þ exp½k ðt si Þ i52a ekt eks4 þ y2
L1 C2 Rb i¼3
L1
i6¼4
kt 1 R1
e þ iR ekt þ y3 ekt eks3
C1 L1 s L1
Dividing the two side of the above equation by ekt term gives the equation:
2 3
1 6 R1 X5 7 Rs 1
y1 k þ 6
41 þ exp½k ðt si Þ7
5 i52a eks4 þ y2 þ
L1 C2 Rb i¼3
L1 C1 L1
i6¼4
R1
iRs þ y3 eks3
L1
¼0
dY2 1
¼ h i Y2 ; y2 k ekt
dt C1 Rs 1 þ Rload
Rs
1
¼ h i ½Y2ðjÞ þ y2 ekt
C1 Rs 1 þ Rload
Rs

ðjÞ
At fixed point 1
R Y2 ¼ 0; k 1
R y2 ¼ 0
C1 Rs ½1 þ load
Rs C1 Rs ½1 þ load
Rs
dY3 dIR
¼ 0 ) y3 k ekt ¼ 0 ) y3 k ¼ 0; s ¼Y2
dt dt
ðjÞ
) iRs k ekt ¼Y2 þ y2 ekt
ðjÞ
At fixed point Y2 ¼ 0 ) iRs k ekt ¼y2 ekt ) iRs k þ y2 ¼ 0
ðjÞ ðjÞ
¼ Y3 ) i52a k ekt ¼ Y3 þ y3 ekt . At fixed point Y3 ¼ 0 ) i52a
dI52a
dt
k þ y3 ¼ 0
The small increments Jacobian of our Bias-T circuit is as follow:

0 1
y1
0 1 B C
!11 . . . !15 B y2 C
B C
B . .. .. C B C
B . C B y3 C ¼ 0; !11 ¼ k; !12 ¼ Rs ;
@ . . . A B C
B C L1
!51 !55 B iR C
@ sA
i52a
R1 ks3 1
!13 ¼ e ; !14 ¼
L1 C1 L1
2 3
1 6 R1 X5 7 ks
!15 ¼ 6 1 þ exp½k ðt si Þ7
5e
4
; !21 ¼ 0;
L1 C2 4 Rb i¼3
i6¼4
1
!22 ¼ k h i
C1 Rs 1 þ Rload
Rs
!23 ¼ 0; !24 ¼ 0; !25 ¼ 0; !31 ¼ 0; !32 ¼ 0; !33 ¼ k; !34 ¼ 0; !35 ¼ 0

!41 ¼ 0; !42 ¼ 1; !43 ¼ 0; !44 ¼ k; !45 ¼ 0; !51 ¼ 0; !52 ¼ 0; !53 ¼ 1;
!54 ¼ 0; !55 ¼ k
0 1
!11 . . . !15
B .. .. .. C
jA k Ij ¼ @ . . . A ; detjA k Ij ¼ 0. We define for simplicity
!51 !55
the following parameters: r1 ¼ LR1s ; r2 ¼ RL11 ; r3 ¼ C11L1 ; r4 ¼ L11C2 ; r5 ¼ RR1b
1
r6 ¼ h i ; !12 ¼ r1 ; !13 ¼ r2 eks3 ;
C1 Rs 1 þ RRload
s
2 3
6 X
5 7 ks
!14 ¼ r3 ; !15 ¼ r4 6
4 1 þ r 5 exp½k ðt si Þ7
5e
4
i¼3
i6¼4
22 ¼ k þ r6 ; detjA k Ij ¼ k4 ðk þ r6 Þ ¼ 0;
1
k1 ¼ 0; k2 ¼ r6 ¼
C1 Rs ½1 þ RRload
s

We get a stability solution which is independent on Bias-T microstrip delay lines

parameters, k1 = 0, k2 < 0 then our stability map is attracting line [26].
P2
Case II: 1 s1 [ 0; 1 s2 [ 0; 1 si [ 0 then our system delay differ-
i¼1
d 2 Vs ðtÞ d 2 IL1 ðts2 Þ
ential equations are as below. dt2 ! e; dt2 !e
2 3
dY1 1 6 R1 X
5 7
¼ I52a ðt s4 Þ 61þ exp½k ðt si Þ7
5 þ Y2
dt L1 C2 4 Rb i¼3
i6¼4
Rs IR ðt s1 Þ R1 1 dVs ðtÞ
þ s þ Y3 eks3
L1 C1 L1 L1 L1 dt
dY2 1
¼ 2 3 Y2 eks1 ;
dt
6 P
5
7
C1 Rs 41 þ Rload
Rs exp½k ðt si Þ5
i¼3
i6¼4
dY3 dIR dI52a
¼ 0; s ¼Y2 ; ¼ Y3
dt dt dt
Small increment equations: i52a k þ y3 ¼ 0

" #
1 R1 X5
Rs
y1 k þ 1þ exp½k ðt si Þ i52a eks4 þ y2
L1 C2 Rb i¼3
L1
i6¼4
1 R1
þ iRs eks1 þ y3 eks3 ¼0
C1 L1 L1
1
y2 k 2 3 y2 eks1 ¼ 0;
6 P
5
7
C1 Rs 41 þ Rload
Rs exp½k ðt si Þ5
i¼3
i6¼4
y3 k ¼ 0; iRs k þ y2 ¼ 0
The small increments Jacobian of our Bias-T circuit is as follow:

0 1
y1
0 1 B C
!11 ...!15 B y2 C
B C
B . .. .. C B C
B . C B C
@ . . . A B y3 C ¼ 0;
B C
!51 !55 B iR C
@ sA
i52a
Rs R1 ks3 1
!11 ¼ k; !12 ¼ ; !13 ¼ e ; !14 ¼ eks1
L1 L1 C1 L1
2 P5 3
k si
6 i¼3 7
1 6 R1 i6¼4 7 ks4
!15 ¼ 61 þ e 7e ; !21 ¼ 0;
L1 C2 4 Rb 5
1
!22 ¼ k 2 P2 3 eks1
k si
C1 Rs 41 þ Rload
Rs e i¼1 5
!23 ¼ 0; !24 ¼ 0; !25 ¼ 0; !31 ¼ 0;

!32 ¼ 0; !33 ¼ k; !34 ¼ 0; !35 ¼ 0
!41 ¼ 0; !42 ¼ 1; !43 ¼ 0; !44 ¼ k;
!45 ¼ 0; !51 ¼ 0; !52 ¼ 0; !53 ¼ 1; !54 ¼ 0; !55 ¼ k
0 1
!11 . . . !15
B .. .. .. C
jA k I j ¼ @ . . . A; detjA k I j ¼ 0: We define for simplicity
!51 !55
the following parameters: r1 ¼ LR1s ; r2 ¼ RL11 ; r3 ¼ C11L1 ; r4 ¼ L11C2 ; r5 ¼ RR1b
1
r6 ðs1 ; s2 Þ ¼ 2 P2 3 eks1 ; 12 ¼ r1 ; 13 ¼ r2 eks3 ;
k si
C1 Rs 41 þ Rload
Rs e i¼1 5
14 ¼ r3 eks1
2 P
5 3
k si
6 i¼3 7
6 i6¼4 7 ks4
15 ¼ r4 61 þ r5 e 7e ; 22 ¼ k þ r6 ; detjA k Ij ¼ k4 ðk þ r6 Þ ¼ 0
4 5
detjA k I j ¼ k4 ðk þ r6 Þ ¼ k5 þ k4 r6 ðs1 ; s2 Þ; Dðs1 ; s2 Þ ¼ k5 þ k4 r6 ðs1 ; s2 Þ

We need to find Dðs1 ; s2 Þ for the following cases: (A) s1 ¼ s; s2 ¼ 0

(B) s1 ¼ 0; s2 ¼ s (C) s1 ¼ s; s2 ¼ s. We need to get characteristics equations for
all above stability analysis cases. We study the occurrence of any possible stability
switching, resulting from the increase of the value of the time delays s1 ; s2 for the
general characteristic equation Dðs1 ; s2 Þ. If we choose s as a parameter, then the
expression: Dðk; sÞ ¼ Pn ðk; sÞ þ Qm ðk; sÞ eks ; n; m 2 R0 ; n [ m.
We analyze the stability switching for the third case (C) s1 ¼ s; s2 ¼ s.
1
Dðs1 ; s2 Þ ¼ k5 k4 eks ;
C1 Rs ½1 þ RRload
s
e 2ks
Pn ðk; sÞ ¼ k5 ; Qm ðk; sÞ ¼ k4 f ðk; sÞ

1 1
Qm ðk; sÞ ¼ k4 h i ; f ðk; sÞ ¼ h i
C1 Rs 1 þ Rs e2ks
Rload
C1 Rs 1 þ RRload
s
e 2ks
The exponential function e1ðk;sÞ ; 1ðk; sÞ ¼ 2 k s can be characterized in a

variety of equivalent ways. In particular it may be defined by the following power
P
1 n
series: 1 ¼ 1ðk; sÞ; e1ðk;sÞ ¼ 1 1 þ 12! 13! þ . . . ¼ 1
2 3 n
n! ð1Þ .
n¼0
12
1ðk;sÞ
13!
3
We take it as approximation expression: e 1 1þ 2!

4 s3 Rload 2ks
e2ks 1 k 2 s þ k2 2 s2 k3 ; C1 Rs 1 þ e
3 Rs
¼ C1 Rs þ C1 Rload e2ks
C1 Rs þ C1 Rload e2ks ¼ C1 ½Rs þ Rload k 2 C1 Rload s
4 s3
þ k2 2 C1 Rload s2 k3 C1 Rload
3
1
f ðk; sÞ
C1 ½Rs þ Rload k 2 C1 Rload s þ k2 2 C1 Rload s2 k3 C1 Rload 4s3
3
Qm ðk; sÞ ¼ k4 f ðk; sÞ

k4
¼
C1 ½Rs þ Rload k 2 C1 Rload s þ k2 2 C1 Rload s2 k3 C1 Rload 4s3
3
Qm ðk; sÞ C1 ½Rs þ Rload k 2 C1 Rload Qm ðk; sÞ s

þ k2 2 C1 Rload Qm ðk; sÞ s2
4 s3 X4
k3 C1 Rload Qm ðk; sÞ þ k4 ¼ 0; wi ðk; sÞ ki ¼ 0
3 i¼0
w0 ðk; sÞ ¼ Qm ðk; sÞ C1 ½Rs þ Rload ; w1 ðk; sÞ ¼ 2 C1 Rload Qm ðk; sÞ s

w2 ðk; sÞ ¼ 2 C1 Rload Qm ðk; sÞ s2 ; w3 ðk; sÞ
4 s3
¼ C1 Rload Qm ðk; sÞ ; w4 ðk; sÞ ¼ 1
3
We can solve a quartic function by factoring it into a product of two quadratic

P
4
equations: wi ðk; sÞ ki ¼ ðk2 þ k C1 þ C2 Þ ðk2 þ k C3 þ C4 Þ.
i¼0
X
4
wi ðk; sÞ ki ¼ k4 þ k3 ðC3 þ C1 Þ þ k2 ðC4 þ C1 C3 þ C2 Þ
i¼0
þ k ðC1 C4 þ C3 C2 Þ þ C2 C4
w0 ðk; sÞ ¼ C2 C4 ; w1 ðk; sÞ ¼ C1 C4 þ C3 C2 ; w2 ðk; sÞ

¼ C4 þ C1 C3 þ C2
w3 ðk; sÞ ¼ C3 þ C1 ; w4 ðk; sÞ ¼ 1
Remark: it is easier to solve the above equations numerically rather than ana-
lytically. The target is to find the two quadratic equations parameters as a function
of Qm ðk; sÞ.Ck ðQm ðk; sÞ; s; . . .Þ; k ¼ 1; 2; 3; 4.
P
4
We have two possible solutions: wi ðk; sÞ ki ¼ ðk2 þ k C1 þ C2 Þ
i¼0
ðk2 þ k C3 þ C4 Þ ¼ 0
k2 þ k C1 ðQm ðk; sÞ; s; . . .Þ þ C2 ðQm ðk; sÞ; s; . . .Þ ¼ 0 or k2 þ k C3 ðQm ðk; sÞ;
s; . . .Þ þ C4 ðQm ðk; sÞ; s; . . .Þ ¼ 0.
X
m X
n¼5
Qm ðk; sÞ ¼ qk ðsÞ kk ; m\n ¼ 5; Pn ðk; sÞ ¼ k5 ; Pn¼5 ðk; sÞ ¼ pk ðsÞ kk
k¼0 k¼0
p0 ðsÞ ¼ 0; p1 ðsÞ ¼ 0; p2 ðsÞ ¼ 0; p3 ðsÞ ¼ 0; p4 ðsÞ ¼ 0; p5 ðsÞ ¼ 1
The homogenous system for Y1 ; Y2 ; Y3 ; IRs ; I52a leads to a characteristic equation

P5
for the eigenvalue k having the form Pðk; sÞ þ Qðk; sÞ eks ¼ 0; PðkÞ ¼ aj k j
j¼0
P
m\5
QðkÞ ¼ cj k j . The coefficients faj ðqi ; qk ; sÞ; cj ðqi ; qk ; sÞg 2 R depend on
j¼0
qi ; qk and delay parameter s. qi ; qk are any Bias-T circuit’s global parameter, other
parameters kept as a constant. Unless strictly necessary, the designation of the
varied arguments ðqi ; qk Þ will subsequently be omitted from P, Q, aj, and cj. The
direct substitution shows that a0 + c0 6¼ 0 for qi ; qk 2 R þ ; that is k = 0 is not of
Pðk; sÞ þ Qðk; sÞ eks ¼ 0. Furthermore Pðk; sÞ; Qðk; sÞ are analytic functions of
k, for which the following requirements of the analysis (Kuang and Cong 2005;
Kuang 1993) can also be verified in the present case.
(a) If k ¼ i x; x 2 R then Pði xÞ þ Qði xÞ 6¼ 0.
(b) j QðkÞ
PðkÞ j is bounded for jkj ! 1. Rek 0 no roots bifurcation from ∞.
polynomial in x.
(d) Each positive root xðqi ; qk Þ of FðxÞ ¼ 0 is continuous and differentiable with
That is for any real number x. Pn ðk ¼ i x; sÞ þ Qm ðk ¼ i x; sÞ 6¼ 0.
X
m \5
Pn ðk ¼ i x; sÞ ¼ i x5 ; Qm ðk ¼ i x; sÞ ¼ qk ðsÞ ik xk ; 0 k m\n ¼ 5
k¼0
ik ; i0 ¼ 1; i1 ¼ i; i2 ¼ 1; i3 ¼ i; i4 ¼ 1; Pn ðk ¼ i x; sÞ þ Qm ðk ¼ i x; sÞ 6¼ 0
X
m\5
i x5 þ qk ðsÞ ik xk 6¼ 0; jPðix; sÞj2 ¼ x10
k¼0
" #2 " #2
X
2 X
1
2k þ 1
jQðix; sÞj2m¼4 ¼ q2k ðsÞ ð1Þ x k 2k
þ k
q2k þ 1 ðsÞ ð1Þ x
k¼0 k¼0
" #2
2 2
X
2
k
FðxÞ ¼ jPði xÞj jQði xÞj ¼ x 10
q2k ðsÞ ð1Þ x 2k
k¼0
" #2
X
1
k 2k þ 1
k¼0
X
2 X
1
q2k ðsÞ ð1Þk x2k ¼ q0 ðsÞ q2 ðsÞ x2 þ q4 ðsÞ x4 ; q2k þ 1 ðsÞ ð1Þk
k¼0 k¼0
2k þ 1
x ¼ q1 ðsÞ x q3 ðsÞ x 3
" #2
X
2
k
q2k ðsÞ ð1Þ x 2k
¼ q20 ðsÞ 2 q0 ðsÞ q2 ðsÞ x2 þ ½q22 ðsÞ þ 2 q0 ðsÞ q4 ðsÞ x4
k¼0
2 q2 ðsÞ q4 ðsÞ x6 þ q24 ðsÞ x8

" #2
X
1
k 2k þ 1
q2k þ 1 ðsÞ ð1Þ x ¼ q21 ðsÞ x2 2 q1 ðsÞ q3 ðsÞ x4 þ q23 ðsÞ x6
k¼0
" #2 " #2
X2 X1
k k 2k þ 1
q2k ðsÞ ð1Þ x þ
2k
k¼0 k¼0
¼ q20 ðsÞ þ ½q21 ðsÞ 2 q0 ðsÞ q2 ðsÞ x2

þ ½q22 ðsÞ þ 2 q0 ðsÞ q4 ðsÞ 2 q1 ðsÞ q3 ðsÞ x4 þ ½q23 ðsÞ
2 q2 ðsÞ q4 ðsÞ x6 þ q24 ðsÞ x8
FðxÞ ¼ jPði xÞj2 jQði xÞj2 ¼ q20 ðsÞ ½q21 ðsÞ 2 q0 ðsÞ q2 ðsÞ x2
½q22 ðsÞ þ 2 q0 ðsÞ q4 ðsÞ 2 q1 ðsÞ q3 ðsÞ x4
½q23 ðsÞ 2 q2 ðsÞ q4 ðsÞ x6 q24 ðsÞ x8 þ x10
We define the following parameters for simplicity P0 ; P2 ; P4 ; P6 ; P8 ; P10
P0 ðsÞ ¼ q20 ðsÞ; P2 ðsÞ ¼ ½q21 ðsÞ 2 q0 ðsÞ q2 ðsÞ; P4 ðsÞ

¼ ½q22 ðsÞ þ 2 q0 ðsÞ q4 ðsÞ 2 q1 ðsÞ q3 ðsÞ
P6 ðsÞ ¼ ½q23 ðsÞ 2 q2 ðsÞ q4 ðsÞ; P8 ðsÞ ¼ q24 ðsÞ; P10 ðsÞ ¼ 1:
P
5
Hence FðxÞ ¼ 0 implies P2k x2k ¼ 0 and its roots are given by solving
k¼0
the above polynomial. QI ði x; sÞ ¼ q1 ðsÞ x q3 ðsÞ x3
PR ði x; sÞ ¼ 0; PI ði x; sÞ ¼ x5 ; QR ði x; sÞ ¼ q0 ðsÞ q2 ðsÞ x2 þ q4 ðsÞ x4

sin hðsÞ ¼
jQði x; sÞj2
cos hðsÞ ¼
jQði x; sÞj2

definition: k ! j; pk ðsÞ ! aj ; qk ðsÞ ! cj ; n ¼ 5; m\5; n [ m
P5
Additionally Pn ðk; sÞ ! PðkÞ; Qm ðk; sÞ ! Qðk; sÞ then PðkÞ ¼ aj k j
j¼0
X
m\5
Qðk; sÞ ¼ cj k j ; PðkÞ ¼ k5 ; a0 ¼ a1 ¼ a2 ¼ a3 ¼ a4 ¼ 0; a5 ¼ 1
j¼0
X
m\5
Qðk; sÞ ¼ c j k j ¼ c 0 þ c 1 k þ c 2 k2 þ c 3 k3 þ c 4 k4
j¼0
n; m 2 R0 ; n [ m and aj ; cj :R þ 0 ! R are continuous and differentiable function

of s such that a0 þ c0 6¼ 0. In the following “—” denotes complex and conjugate.
PðkÞ; Qðk; sÞ are analytic functions in k and differentiable in s. The coefficients
faj ðRs ; C1 ; L1 ; R1 ; C2 ; Rload ; Rb ; sÞ and cj ðRs ; C1 ; L1 ; R1 ; C2 ; Rload ; Rb ; sÞg 2 R
depend on Bias-T’s Rs ; C1 ; L1 ; R1 ; C2 ; Rload ; Rb ; s values. Unless strictly necessary,
the designation of the varied arguments: ðRs ; C1 ; L1 ; R1 ; C2 ; Rload ; Rb ; sÞ will sub-
sequently be omitted from P; Q; aj ; cj . The coefficients aj, cj are continuous and
a0 þ c0 6¼ 0.a0 ¼ 0 ; c0 ¼ q0 ðsÞ; a0 þ c0 6¼ 0 ) q0 ðsÞ 6¼ 0.
8 Rs ; C1 ; L1 ; R1 ; C2 ; Rload ; Rb ; s 2 R þ i.e. k = 0 is not a root of the characteristic
equation. Furthermore PðkÞ; Qðk; sÞ are analytic functions of k for which the fol-
lowing requirements of the analysis (see Kuang 1993, Sect. 3.4) can also be verified
in the present case.
(a) If k ¼ i x; x 2 R then Pði xÞ þ Qði xÞ 6¼ 0 i.e. P and Q have no common
ðRs ; C1 ; L1 ; R1 ; C2 ; Rload ; Rb ; sÞ domain of interest.
(b) j Qðk;sÞ
PðkÞ j is bounded for jkj ! 1, Rek 0. No roots bifurcation from ∞. Indeed,
Pm\5 j
cj k P
m\5
in the limit: j Qðk;sÞ
PðkÞ j ¼ j j¼0
k5
j¼j cj kj5 j
j¼0
2 2 P5
(c) FðxÞ ¼ jPði xÞj jQði xÞj ; FðxÞ ¼ P2k x2k has at most a finite
k¼0
number of zeros. Indeed, this is a polynomial in x (degree in x10).
(d) Each positive root xðRs ; C1 ; L1 ; R1 ; C2 ; Rload ; Rb ; sÞ of FðxÞ ¼ 0 is continuous
and differentiable with respect to Rs ; C1 ; L1 ; R1 ; C2 ; Rload ; Rb ; s. This condition
can be assessed numerically.

Pði xÞ and Qði xÞ ¼ Qði xÞ thus, x > 0 maybe on eigenvalue of character-
istic equations. The analysis consists in identifying the roots of the characteristic
equation situated on the imaginary axis of the complex k—plane, whereby
increasing the parameters: Rs ; C1 ; L1 ; R1 ; C2 ; Rload ; Rb ; s Rek may, at the crossing,
ðjÞ ðjÞ ðjÞ ðjÞ ðjÞ
change its sign from (−) to (+), i.e. from stable focus E ðjÞ ðY1 ; Y2 ; Y3 ; I52a ; IRs Þ
j ¼ 0; 1; 2; . . . to an unstable one, or vice versa. This feature may be further
assessed by examining the sign of the partial derivatives with respect to
Rs ; C1 ; L1 ; R1 ; C2 ; Rload ; Rb ; s.
@Rek
K1 ðsÞ ¼ ð Þ ; Rs ; C1 ; L1 ; R1 ; C2 ; Rload ; Rb ¼ Const; x 2 R þ
@s k¼ix
When writing PðkÞ ¼ PR ðkÞ þ i PI ðkÞ; Qðk; sÞ ¼ QR ðk; sÞ þ i QI ðk; sÞ and

inserting k ¼ i x into Bias-T circuit’s characteristic equation, x must satisfy the
following:
jQði x; sÞj2
jQði x; sÞj2
where jQði x; sÞj2 6¼ 0 in view of requirement (a) above, and ðg; hÞ 2 R.

Note: F(x) is dependent on s. Now it is important to notice that if s 62 I(assume
that I
R þ 0 is the set where xðsÞ is a positive root of F(x) and for s 62 I,xðsÞ is not
defined. Then for all s in I, xðsÞ is satisfied that FðxÞ ¼ 0. Then there are no
positive xðsÞ solutions for FðxÞ ¼ 0, and we cannot have stability switches. For
s 2 I where xðsÞ is a positive solution of FðxÞ ¼ 0, we can define the angle
hðsÞ 2 ½0; 2 p as the solution of sin hðsÞ ¼ . . .; cos hðsÞ ¼ . . .
xðsÞ s ¼ hðsÞ þ 2 n p 8 n 2 R0 . Hence we can define the maps sn : I ! R þ 0
given by sn ðsÞ ¼ hðsÞxðsÞ þ 2np
; n 2 R0 ; s 2 I. Let us introduce the functions: I !
R; Sn ðsÞ ¼ s sn ðsÞ; s 2 I; n 2 R0 that is continuous and differentiable in s. In the
following the subscripts k; x; Rs ; C1 ; L1 ; R1 ; C2 ; Rload ; Rb ; :: indicate the corre-
sponding partial derivatives. Let us first concentrate on K(x), remember in
kðRs ; C1 ; L1 ; R1 ; C2 ; Rload ; Rb ; ::Þ and xðRs ; C1 ; L1 ; R1 ; C2 ; Rload ; Rb ; ::Þ, and keeping
[BK]. Differentiating Bias-T circuit characteristic equation PðkÞ þ Qðk; sÞ eks ¼
0 with respect to specific parameter (x), and inverting the derivative, for conve-
nience, one calculates:
Remark: x ¼ Rs ; C1 ; L1 ; R1 ; C2 ; Rload ; Rb ; ::

ð Þ ¼
@Q @Q
where Pk ¼ @P @P
@k ; Qk ¼ @k ; Px ¼ @x ; Qx ¼ @x , substituting k ¼ i x and bearing
@Pðk;xÞ
Pði xÞ ¼ Pði xÞ; Qði xÞ ¼ Qði xÞ: Then @k ¼ @Pðk;xÞ @Pðk;xÞ
@½ix ¼ i@x ¼
i @Pðk;xÞ
@x
i @Pðk;xÞ @Pðk;xÞ
@k ¼ @x ; i Pk ði xÞ ¼ Px ði xÞ; i Qk ði xÞ ¼ Qx ði xÞ and that on
the surface jPði xÞj2 ¼ jQði xÞj2 , one obtains:
@k i Px ði x; xÞ Pði x; xÞ þ i Qk ði x; xÞ Qðk; xÞ s jPði xÞj2

ð Þ1 jk¼ix ¼ ð Þ

Q ¼ QR þ i QI

¼ P2R þ P2I ¼ x10
When (x) can be any Bias-T circuit parameter’s Rs ; C1 ; L1 ; R1 ; C2 ; Rload ; Rb and s

etc. Where for convenience, we have dropped the arguments ði x; xÞ, and where
Fx ¼ 2 ½ðPRx PR þ PIx PI Þ ðQRx QR þ QIx QI Þ; xx ¼ F Fx
x
Fx ¼ 2 ½ðPRx PR þ PIx PI Þ ðQRx QR þ QIx QI Þ. We define U and V:

QI ¼ q1 ðsÞ x q3 ðsÞ x3 ; PR ¼ 0; PI ¼ x5 ; QR ¼ q0 ðsÞ q2 ðsÞ x2 þ q4 ðsÞ x4

PRx ¼ 0; PIx ¼ 5 x4 ; QIx ¼ q1 ðsÞ q3 ðsÞ 3 x2 ;
QRx ¼ 2 q2 ðsÞ x þ 4 q4 ðsÞ x3
@q1 ðsÞ @q3 ðsÞ 3
PRs ¼ 0; PIs ¼ 0; QIs ¼ x x ;
@s @s
@q0 ðsÞ @q2 ðsÞ 2 @q4 ðsÞ 4
QRs ¼ x þ x
@s @s @s
Fs
xs ¼ ; PRx PR ¼ 0; QRx QR ¼ ½2 q2 ðsÞ x þ 4 q4 ðsÞ x3
Fx
½q0 ðsÞ q2 ðsÞ x2 þ q4 ðsÞ x4
QRx QR ¼ 2 q0 ðsÞ q2 ðsÞ x þ 2 ½q22 ðsÞ þ 2 q0 ðsÞ q4 ðsÞ x3
6 q2 ðsÞ q4 ðsÞ x5 þ 4 q24 ðsÞ x7
!1 ðsÞ ¼ 2 q0 ðsÞ q2 ðsÞ; !3 ðsÞ ¼ 2 ½q22 ðsÞ þ 2 q0 ðsÞ q4 ðsÞ;
!5 ðsÞ ¼ 6 q2 ðsÞ q4 ðsÞ
X
4
!7 ðsÞ ¼ 4 q24 ðsÞ; QRx QR ¼ !2k1 x2k1 ; PR PIx ¼ 0; PI PRx ¼ 0;
k¼1
PIx PI ¼ 5 x9
QIx QI ¼ ½q1 ðsÞ q3 ðsÞ 3 x2 ½q1 ðsÞ x q3 ðsÞ x3
¼ q21 ðsÞ x 4 q1 ðsÞ q3 ðsÞ x3 þ 3 q23 ðsÞ x5
w1 ðsÞ ¼ q21 ðsÞ; w3 ðsÞ ¼ 4 q1 ðsÞ q3 ðsÞ; w5 ðsÞ ¼ 3 q23 ðsÞ; QIx QI
X
3
¼ w2k1 ðsÞ x2k1
k¼1
QR QIx ¼ ½q0 ðsÞ q2 ðsÞ x2 þ q4 ðsÞ x4 ½q1 ðsÞ q3 ðsÞ 3 x2 ¼ q0 ðsÞ q1 ðsÞ
½3 q0 ðsÞ q3 ðsÞ þ q2 ðsÞ q1 ðsÞ x2 þ ½3 q3 ðsÞ q2 ðsÞ þ q4 ðsÞ q1 ðsÞ x4
3 q4 ðsÞ q3 ðsÞ x6
A0 ðsÞ ¼ q0 ðsÞ q1 ðsÞ; A2 ðsÞ ¼ ½3 q0 ðsÞ q3 ðsÞ þ q2 ðsÞ q1 ðsÞ

A4 ðsÞ ¼ 3 q3 ðsÞ q2 ðsÞ þ q4 ðsÞ q1 ðsÞ; A6 ðsÞ ¼ 3 q4 ðsÞ q3 ðsÞ;
X
3
QR QIx ¼ A2k ðsÞ x2k
k¼0
QI QRx ¼ ½q1 ðsÞ x q3 ðsÞ x3 ½2 q2 ðsÞ x þ 4 q4 ðsÞ x3
¼ 2 q1 ðsÞ q2 ðsÞ x2
þ 2 ½2 q4 ðsÞ q1 ðsÞ þ q2 ðsÞ q3 ðsÞ x4 4 q3 ðsÞ q4 ðsÞ x6
n2 ðsÞ ¼ 2 q1 ðsÞ q2 ðsÞ; n4 ðsÞ ¼ 2 ½2 q4 ðsÞ q1 ðsÞ þ q2 ðsÞ q3 ðsÞ;
n6 ðsÞ ¼ 4 q3 ðsÞ q4 ðsÞ
X
3
QI QRx ¼ n2k ðsÞ x2k ; PR PIs ¼ 0; PI PRs ¼ 0
k¼1
QR QIs ¼ ½q0 ðsÞ q2 ðsÞ x2 þ q4 ðsÞ x4 ½ x x
@s @s
@q1 ðsÞ
¼ q0 ðsÞ x
@s
@q3 ðsÞ @q1 ðsÞ @q3 ðsÞ
½q0 ðsÞ þ q2 ðsÞ x3 þ ½q2 ðsÞ
@s @s @s
þ q4 ðsÞ x q4 ðsÞ
5
x
@s @s
f1 ðsÞ ¼ q0 ðsÞ ; f3 ðsÞ ¼ q0 ðsÞ þ q2 ðsÞ
@s @s @s
f5 ðsÞ ¼ q2 ðsÞ þ q4 ðsÞ ; f7 ðsÞ ¼ q4 ðsÞ ;
@s @s @s
X 3
QR QIs ¼ f2k þ 1 x2k þ 1
K¼0

QI QRs ¼ ½q1 ðsÞ x q3 ðsÞ x3 ½ x þ x
@s @s @s
¼ q1 ðsÞ x ½q1 ðsÞ þ q3 ðsÞ x3
@s @s @s
@q4 ðsÞ @q2 ðsÞ @q4 ðsÞ 7
þ ½q1 ðsÞ þ q3 ðsÞ x5 q3 ðsÞ x
@s @s @s
g1 ðsÞ ¼ q1 ðsÞ ; g3 ðsÞ ¼ ½q1 ðsÞ þ q3 ðsÞ
@s @s @s
g5 ðsÞ ¼ q1 ðsÞ þ q3 ðsÞ ; g7 ðsÞ ¼ q3 ðsÞ ;
@s @s @s
X3
QI QRs ¼ g2k þ 1 ðsÞ x2k þ 1
k¼0
X
3
Fx ¼ 10 x9 2 ½ ½w2k1 ðsÞ þ 2k1 x2k1 þ 7 x7
k¼1
Fs ¼ 2 ðQRs QR þ QIs QI Þ ¼ 2 ð½ x þ x
@s @s @s
½q0 ðsÞ q2 ðsÞ x2 þ q4 ðsÞ x4 þ ½ x x
@s @s
½q1 ðsÞ x q3 ðsÞ x3 Þ
@q0 ðsÞ @q0 ðsÞ

Fs ¼ 2 ðQRs QR þ QIs QI Þ ¼ 2 f q0 ðsÞ ½ q2 ðsÞ
@s @s
@q2 ðsÞ @q1 ðsÞ
þ q0 ðsÞ q1 ðsÞ x2
@s @s
@q0 ðsÞ @q2 ðsÞ @q4 ðsÞ @q1 ðsÞ
þ½ q4 ðsÞ þ q2 ðsÞ þ q0 ðsÞ q3 ðsÞ
@s @s @s @s
q1 ðsÞ x4 ½ q4 ðsÞ þ q2 ðsÞ q3 ðsÞ
@s @s @s @s
@q4 ðsÞ
x6 þ q4 ðsÞ x8 g
@s
@q0 ðsÞ @q0 ðsÞ
Fs ¼ 2 ðQRs QR þ QIs QI Þ ¼ 2 q0 ðsÞ þ 2 ½ q2 ðsÞ
@s @s
þ q0 ðsÞ q1 ðsÞ x2 2 ½ q4 ðsÞ þ q2 ðsÞ
@s @s @s @s
þ q0 ðsÞ q3 ðsÞ q1 ðsÞ x4
@s @s @s
þ2 ½ q4 ðsÞ þ q2 ðsÞ q3 ðsÞ x6 2 q4 ðsÞ x8
@s @s @s @s
We define for simplicity the following functions: B0 ðsÞ ¼ 2 @q@s

0 ðsÞ
q0 ðsÞ

B2 ðsÞ ¼ 2 q2 ðsÞ þ q0 ðsÞ q1 ðsÞ
@s @s @s

B4 ðsÞ ¼ 2 q4 ðsÞ þ q2 ðsÞ þ q0 ðsÞ
@s @s @s

@q1 ðsÞ @q3 ðsÞ
q3 ðsÞ q1 ðsÞ
@s @s

B6 ðsÞ ¼ 2 q4 ðsÞ þ q2 ðsÞ q3 ðsÞ ;
@s @s @s
@q4 ðsÞ
B8 ðsÞ ¼ 2 q4 ðsÞ
@s
X 4
Fs ¼ 2 ðQRs QR þ QIs QI Þ ¼ B2k ðsÞ x2k
k¼0
X
3 X
3
¼ ð A2k ðsÞ x2k n2k ðsÞ x2k Þ
k¼0 k¼1
X
3 X
3 X
3
U¼ n2k ðsÞ x2k A2k ðsÞ x2k ¼ ½n2k ðsÞ A2k ðsÞ x2k A0 ðsÞ
k¼1 k¼0 k¼1
X
3 X
3
Vjx¼s ¼ ðQR QIs QI QRs Þ ¼ ð f2k þ 1 x2k þ 1 g2k þ 1 ðsÞ x2k þ 1 Þ
k¼0 k¼0
X
3
2k þ 1
¼ ½g2k þ 1 ðsÞ f2k þ 1 x
k¼0
Fðx; sÞ ¼ 0. Differentiating with respect to s and we get Fx @x @s þ Fs ¼ 0

@x Fs @ Rek
s2I) ¼ xs ¼ ; K1 ðsÞ ¼ ;
@s Fx @s k¼ix
( )
K1 ðsÞ ¼ Re

@ Rek
signfK1 ðsÞg ¼ sign ;
@s k¼ix
( )
1 @x U @x
@s þ V
signfK ðsÞg ¼ signfFx g sign s þxþ
@s jPj2

ðjÞ ðjÞ ðjÞ ðjÞ ðjÞ
Bias-T circuit, about the equilibrium points EðjÞ ðY1 ; Y2 ; Y3 ; I52a ; IRs Þ, j = 0, 1,…
As a result of a variation of delay parameter s. The analysis consists in identi-

fying the roots of our system characteristic equation situated on the imaginary axis
of the complex k-plane where by increasing the delay parameter s, Rek may at the
crossing, changes its sign from − to +, i.e. from stable focus E(j) to an unstable one,
or vice versa. This feature may be further assessed by examining the sign of the
partial derivatives with respect to s,
@Rek
K1 ðsÞ ¼ ð Þ ; Rs ; C1 ; L1 ; R1 ; C2 ; Rload ; Rb ; . . . ¼ const; x 2 R þ
@s k¼ix
( " #)
X3
signfK1 ðsÞg ¼ sign 10 x9 2 ½w2k1 ðsÞ þ !2k1 x2k1 þ !7 x7
k¼1
0 1
P
4
B B2k ðsÞ x2k C
B k¼0 C
signfs B 3 C þ x
@ P A
10 x9 2 ½w2k1 ðsÞ þ !2k1 x2k1 þ !7 x7
k¼1
0 1
" # P4
X3 B B 2k ðsÞ x 2k
C
B k¼0 C
f ½n2k ðsÞ A2k ðsÞ x2k A0 ðsÞ B 3 C
@ P A
k¼1 10 x9 2 ½w2k1 ðsÞ þ !2k1 x2k1 þ 7 x7
k¼1
X
3
þ ½g2k þ 1 ðsÞ f2k þ 1 x2k þ 1 g
k¼0
þ g
x10
s ¼ xh þþ ðsÞ
x þ ðsÞ if only x þ is feasible. Additionally, when all Bias-T circuit parameters are
known and the stability switch due to various time delay values s is described in the
following expression:
signfK1 ðsÞg ¼ signfF

(x ðxðsÞ; sÞg )
sign s xs ðxðsÞÞ þ xðsÞ þ
jPðxðsÞÞj2
Remark: We know Fx ðxðsÞ; sÞ ¼ 0 implies its roots xi ðsÞ and finding those
delays values s which xi is feasible. There are s values which give complex xi or
imaginary number, then unable to analyze the stability.
4.5 PIN Diode Stability Analysis Under Parameters Variation 489
4.5 PIN Diode Stability Analysis Under Parameters

Variation
A PIN diode is a diode with a wide, lightly doped ‘near’ intrinsic semiconductor
region between a p-type semiconductors an n-type semiconductor region. The
p-type and n-type regions are typically heavily doped because they are used for
ohmic contacts. The wide intrinsic region is in contrast to an ordinary PN diode.
The wide intrinsic region makes the PIN diode an inferior rectifier (one typical
function of a diode). The wide intrinsic region makes the PIN diode suitable for
many applications [33,91,92]. The PIN diode suitable for many applications:
attenuators, fast switches, photo detectors and high voltage power electronics
applications. A PIN diode operates under what is known as high level injection.
PIN’s intrinsic “i” region is flooded with charge carriers from the “p” and “n”
regions. Its function can be likened to filling up a water bucket with a hole on the
side. Once the water reaches the hole’s level it will begin to pour out. A PIN diode
obeys the standard diode equation for low frequency signals. At higher frequencies,
the diode looks like an almost perfect (very linear, even for large signals) resistor.
At low frequencies, the charge can be removed and the diode turns off. At higher
frequencies, there is not enough time to remove the charge, so the diode never turns
off. The high frequency resistance is inversely proportional to the DC bias current
through the diode. A PIN diode, suitably biased, therefore acts as a variable resistor.
The high frequency resistance may vary over a wide range (from 0.1 to 10 k-ohm in
some cases; the useful range is smaller, though). The wide intrinsic region also
means the diode will have a low capacitance when reverse biased.
PIN diode fundamentals: A PIN diode is a semiconductor device that operates as
a variable resistor at RF and microwave frequencies. The resistance value of the
PIN diode is determined only by the forward biased DC current. In switch and
attenuator applications, the PIN diode should ideally control the RF signal level
without introduction distortion which might change the shape of the RF signal. An
important additional feature of the PIN diode is its ability to control large RF
signals while using much smaller level of DC excitation. A model of a PIN diode
chip is presented and the chip is prepared by starting with a wafer of almost
intrinsically pure silicon, having high resistivity and long lifetime. A P-region is
then diffused into one diode surface and an N-region is diffused into the other
surface. The resulting intrinsic or I-region thickness (W) is a function of the
thickness of the original silicon wafer.
The area of the chip (A) depends upon how many small sections are defined
from the original wafer. The performance of the PIN diode primarily depends on
chip geometry and nature of the semiconductor material in the finished diode,
particularly in the I-region. The characteristics of PIN diodes are controlled

thickness I-regions having long carrier lifetimes and very high resistivity. These
characteristics enhance the ability to control RF signals with a minimum of dis-
tortion while requiring low DC supply. When a PIN diode is forward biased, holes
and electrons are injected from the P and N regions into the I-region. These charges
do not recombine immediately. Instead, a finite quantity of charge always remains
stored and results in a lowering of the resistivity of the I-region. The quantity of
stored charge, Q depends on the recombination time (s)—the carrier life time, and
the forward bias current (IF), Q = IFs. The resistance of the I-region under forward
2
bias, Rs is inversely proportional to Q. Rs ¼ ðl þWl ÞQ ðohmÞ, W—I-region width,
N p
lN —electron mobility, lp - hole mobility. We get the expression for Rs as an

W2
inverse function of current Rs ¼ ðl þ lp ÞIF s ðohmÞ. The equation is independent of
N
area (A). Rs is slightly dependent upon area because the effective lifetime varies
with area and thickness sðA; WÞ due to edge recombination effects. Typically, PIN
diodes display a resistance characteristic consistent with this model. Resistance of
the order of 0.1 Ohm at 1A forward bias increasing to about 10000 Ohm (10Kohm)
at 1lA. The forward bias represents a realistic range for a PIN diode. The maximum
forward resistance, Rs(max), of a PIN diode is generally specified at 100 mA
forward bias current. Some PIN diodes suppliers specifies not only the Rs(max) but
also the Rs(min) at a lower forward bias current (*10 mA). It ensures a wide range
of diode resistance which is particularly important in attenuator applications. At the
lower frequencies Rs is not constant but increases as the frequency is lowered. The
normal PIN diodes which are designed to operate in RF/Microwave frequencies
exhibit this increase in Rs in the 1–10 MHz range.
A properly designed PIN will maintain constant Rs well into the 10 kHz region.
The results obtained are valid over an extremely broad frequency range. The
practical low resistance limitations result from package parasitic inductances and
junction contact resistances. Both of which are minimized in the construction of
PIN diodes. The high resistance range of PIN diodes is usually limited by the effect
of the diode capacitance (Ct). The maximum dynamic range of the PIN diode at
high frequencies, this diode reactance may have to be tuned out. The “skin effect” is
much less pronounced in relatively poor conductors such as silicon, than with good
metallic conductors. The “skin depth” is proportional to the square root of the
resistivity of the conducting material. RF signals penetrate deeply into the semi-
conductor and “skin effect” is not a significant factor in PIN diodes below X-band
frequencies. At DC and very low frequencies, the PIN diode is similar to a PN
diode. The diode resistance is described by the dynamic resistance of the I–V
characteristics at any quiescent bias point. The DC dynamic resistance point is not,
however, valid in PIN diodes at frequencies above which the period is shorter than
the transit time of the I-region. The frequency at which this occurs, fT is called
transit time frequency and may be considered the lower frequency limit. The lower
frequency limit is primarily a function of W, the I-region thickness and can be
expressed at fT = 1300/W, where W is the I-region thickness in microns.
The conductance of the diode is proportional to the stored charge and the charge
is in turn related to the diode current by Id ¼ dQ dt þ s where Id is the diode current,
d Qd
Qd is the charge stored in the diode, s is diode recombination lifetime. If the diode is
biased with only a constant current, the stored charge is constant and is equal to
Qd ¼ Id s. The PIN diode store charge equation:
Id ðtÞ ¼ dQdtd ðtÞ þ QdsðtÞ, Id ðtÞ is a function of time. First we consider the simpler
dQd ðtÞ
dQd ðtÞ Qd ðtÞ
equation Id ðtÞ ¼ 0. The equation is dt þ s ¼ 0 or dt
Qd ðtÞ ¼ 1s
Z R1 R1
dQd 1 1
¼ dt; ln Qd ¼ dt þ const; Qd ¼ e sdt þ const ¼ A e sdt ;
Qd s s
A ¼ econst
R
Is ¼ 1s dt ) dIdts ¼ 1s ; Qd ¼ A eIs ; Qd eIs ¼ A. We can see how to solve our
equation. If we differentiate equation Qd eIs ¼ A respect to time t and use
dt ¼ s , we get dt ðQd e Þ ¼ dt e þ Qd e dt ; dt ðQd e Þ ¼ e ½ dt þ Qd
dIs 1 d Is dQd Is Is dIs d Is Is dQd
dt e is the integrating factor. Since dt þ Qd dt jdIs ¼1 ¼ Id then

dIs Is dQd dIs
dt s
dt ðQd eIs Þ ¼ eIs ½dQ

dt þ Qd dt ¼ e Id . Id and e are functions of t only, we
d d dIs Is Is
can now integrate both sides of dt ðQd e Þ ¼ e Id with respect to time t to get
d Is Is
R R R
Qd eIs ¼ eIs Id dt þ const; Qd ¼ eIs eIs Id dt þ const; Is ¼ 1s dt. If we
considerRthat recombination lifetime parameter s is independent on time then
Is ¼ 1s dt ¼ 1s t þ const. If the bias consists of both a constant current and a
low frequency RF or time varying signal, then the DC component of stored charge
will be modulated by the presence of an AC component. The degree of modulation
depends on the relative level the two charge components and the frequency of the
RF signal. At signal frequency below fc ¼ 2ps 1
the RF signal has about the same
effect as the DC bias. Above fc , the modulation effect decreases. The lifetime of PIN
diodes is determined by design and is based on the desired switching speed.
Typically, diode recombination lifetime s can be in the range of 0.005 lsec to over
3 lsec. At frequencies bellow fc , the PIN diode behaves as an ordinary PN junction
diode. The RF signal incident on the diode will be rectified and considerable
distortion of the signal will occur. In the vicinity of fc , the diode begins to behave as
a linear resistor with a small nonlinear component. At frequencies well above fc , the
Fig. 4.16 PIN diode low

frequency equivalent circuit
diode appears essentially as a pure linear resistance whose value can be controlled
by the DC or a low frequency control signal. The equivalent circuit of the PIN diode
also depends on the frequency. At frequencies much less than fc the equivalent
circuit is as shown in the below figure (normal PN junction) [24] (Fig. 4.16).
In this circuit LP is the package inductance, CP is the package
capacitance, Rs is
the series resistance, and Rj is the junction resistance Rj ¼ nkT
qIdc . Typical value for
n is 1.8 then at room temperature and Rj ¼ Idc ðmAÞ
48
. Idc is the forward DC bias current,
Cj(V) is the junction capacitance which is a function of the applied voltage. At
frequencies much higher than fc , we can draw equivalent circuit is as shown in the
below figure. LP, CP, and Rs are the same as in the low frequency equivalent circuit.
The element CI represents the I-layer capacitance which is constant and dependent
only on the geometry of the I-layer (typical values of CI are between 0.02 and 2 pF
and are dependent on diode design). The element RI represents the effective RF
resistance of the I-layer. This resistance is constant with respect to RF signal,
providing the signal frequency much higher than fc . It is variable by the DC or very
low frequency control current (Fig. 4.17).
Fig. 4.17 PIN diode high

frequency equivalent circuit
We analyze the stability of Single Pole Single Throw (SPST)) PIN switches.
The RF switch circuit requires a few components and a very simple to implement.
Nevertheless it is able to act as an RF switch for radio frequency or RF applications
and is adequate for many applications. The RF switch circuit comprises a single
PIN diode (D1), an RF inductor or choke (Lc), a current limiting resistor or RF
choke (Ld) and a DC block capacitor (Cc). In operation, when a positive potential is
applied to the control point current, this forward biases the diode and as a result the
radio frequency signal is able to pass through the circuit. When a negative bias is
applied to the circuit, the diode become reverse biased and is effectively switched
off. Under these conditions the depletion layer in the diode becomes wide and does
not allow signal to pass. PIN diodes have a number of advantages as switches. In
the first place they are more linear than ordinary PN junction diodes. This means
that in their action as a radio frequency switch they do not create as many spurious
products. Secondly when reverse biased and switched off, the depletion layer is
wider than with an ordinary diode and this provides for greater isolation when
switching. By varying the amount of bias on the PIN diode it is possible to vary the
level of attenuation provided. In this way the circuit can be used as a very simple
RF attenuator. Although the circuit can be used as an RF attenuator, more effective
RF attenuator circuits are available for more demanding applications [92, 93]
(Fig. 4.18).
Fig. 4.18 PIN diode attenuator and switch
Applications of PIN diode:

• A variable resistor in a variable attenuator, a function that few other components
can achieve as effectively. The fact that when it is forward biased, the diode is
linear, behaving like a resistor, can be put to good use in a variety of
applications.
• The PIN diode can also be used as an RF switch. In the forward direction it can
be biased sufficiently to ensure it has a low resistance to the RF that needs to be
passed, and when a reverse bias is applied it acts as an open circuit, with only a
relatively small level of capacitance.
• PIN diode is for use in RF protection circuits. When used with RF, the diode
normally behaves like a resistor when a small bias is applied. Hover this is only
true for RF levels below a certain level. Above this the resistance drops con-
siderably. Thus it can be used to protect a sensitive receiver from the effects of a
large transmitter if it is placed across the receiver input.
We consider the RF source as a high frequency signal and use PIN diode high
frequency equivalent circuit.
VA1 VA2 d
VA1 ¼ Vs ðtÞ; IRsource ¼ ; IRsource ¼ ICc ; ICc ¼ Cc ðVA2 VA3 Þ; VLd
Rsource dt
dIL
¼ Vd VA3 ¼ Ld d
dt
(Fig. 4.19).
Fig. 4.19 PIN diode attenuator and switch equivalent circuit
dILp VA VA5 VA VA6

VLp ¼ VA3 VA4 ¼ Lp ; ILp ¼ IRs ; IRs ¼ 4 ; IR I ¼ 5 ; I CI
dt Rs RI
d
¼ CI ðVA5 VA6 Þ
dt
d dIL VA6
I Cp ¼ C p ðVA3 VA6 Þ; VA6 ¼ VLc ¼ Lc c ; IRload ¼ ; VLc ¼ VRload ¼ Vout
dt dt Rload
We write our circuit Kirchhoff’s Current Law (KCL) for all nodes:
Vs ðtÞ VA2
IRsource ¼ ) Vs ðtÞ VA2 ¼ IRsource Rsource ; VA2 ¼ Vs ðtÞ IRsource Rsource
Rsource
Z
d 1 d 1
ICc ¼ Cc ðVA2 VA3 Þ ) ICc ¼ ðVA2 VA3 Þ; VA2 VA3 ¼ ICc dt
dt Cc dt Cc
Z Z
1 1
VA3 ¼ VA2 ICc dt; VA3 ¼ Vs ðtÞ IRsource Rsource ICc dt
Cc Cc
Table 4.4 PIN diode attenuator and switch equivalent circuit Kirchhoff’s Current Law (KCL) for
all nodes
Node number KCL @ Ai
A2 IRsource ¼ ICc
A3 ICc þ ILd ¼ ILp þ ICp
A4 ILp ¼ IRs
A5 IRs ¼ IRI þ ICI
A6 IRI þ ICI þ ICp ¼ ILc þ IRload
Z
dIL 1 dILd
Vd VA3 ¼ Ld d ; Vd ½Vs ðtÞ IRsource Rsource ICc dt ¼ Ld
dt Cc dt
Z
d 1 d 2 I Ld
fVd ½Vs ðtÞ IRsource Rsource ICc dtg ¼ Ld
dt Cc dt2
dVd dVs ðtÞ dIRsource 1 d 2 IL dVd

þ Rsource þ ICc ¼ Ld 2 d ; !e
dt dt dt Cc dt dt
d 2 ILd
First system differential equation: dVdts ðtÞ þ dIRsource
dt Rsource þ 1
Cc ICc ¼ Ld dt2
dILp dIL
VA3 VA4 ¼ Lp ) VA4 ¼ VA3 p ; VA4
dt dt Z
1 dIL
¼ Vs ðtÞ IRsource Rsource ICc dt p
Cc dt
dILp dIL
VA3 VA4 ¼ Lp ) VA4 ¼ VA3 Lp p ; VA4 ¼ Vs ðtÞ IRsource Rsource
Z dt dt
1 dILp
ICc dt Lp
Cc dt
VA4 VA5
IRs ¼ ) VA4 VA5 ¼ IRs Rs ; VA5 ¼ VA4 IRs Rs
Rs
Z
1 dIL
VA5 ¼ Vs ðtÞ IRsource Rsource ICc dt Lp p IRs Rs
Cc dt
VA5 VA6
IRI ¼ ) VA5 VA6 ¼ IRI RI ; VA6 ¼ VA5 IRI RI
RI
Z
1 dIL
VA5 ¼ Vs ðtÞ IRsource Rsource ICc dt Lp p IRs Rs
Cc dt
VA5 VA6
IR I ¼ ) VA5 VA6 ¼ IRI RI ; VA6 ¼ VA5 IRI RI
RI
Z
1 dIL
VA6 ¼ Vs ðtÞ IRsource Rsource ICc dt Lp p IRs Rs IRI RI
Cc dt
d VA VA6
ICI ¼ CI ðVA5 VA6 Þ; IRI ¼ 5 ) VA5 VA6 ¼ IRI RI ; ICI
dt RI
dIR
¼ CI RI I
dt
Z
d IC d 1
I Cp ¼ Cp ðVA3 VA6 Þ ) p ¼ ðVA3 VA6 Þ; VA3 VA6 ¼ ICp dt
dt Cp dt Cp
Z
1
Vs ðtÞ IRsource Rsource ICc dt fVs ðtÞ IRsource Rsource
Cc
Z
1 dI L
ICc dt Lp p
Cc dt
Z
1
IRs Rs IRI RI g ¼ ICp dt
Cp
Z
dILp 1 d dILp
Lp þ IRs Rs þ IRI RI ¼ ICp dt; Lp þ IRs Rs þ IRI RI
dt Cp dt dt
1
¼ ICp
Cp
d 2 I Lp dIRs dIRI 1 1 dIR

Lp 2
þ Rs þ RI ¼ ICp ; ICI ¼ RI I
dt dt dt Cp CI dt
d 2 ILp dIRs
Second system differential equation: Lp dt2 þ dt Rs þ 1
CI ICI ¼ C1p ICp
dILc VA6 dIL

VA6 ¼ Lc ; IRload ¼ ; Lc c ¼ IRload Rload
dt Rload dt
Third system differential equation: Lc dIdtLc ¼ IRload Rload

d 2 ILd dVs ðtÞ 1 dIRsource Rsource 1

¼ þ þ I Cc
dt 2 dt Ld dt Ld Cc Ld
d 2 I Lp 1 1 dIR Rs dIL Rload

¼ I Cp I CI s ; c ¼ IRload
dt 2 Cp Lp CI Lp dt Lp dt Lc
dILp
Since IRs ¼ ILp ; dIdtRs ¼ dt ; IRsource ¼ ICc ; dIRdt
source
¼ dIdtCc
d 2 I Ld dVs ðtÞ 1 dICc Rsource 1

¼ þ þ ICc
dt 2 dt Ld dt Ld Cc Ld
d 2 I Lp 1 1 dIL Rs dIL Rload

¼ I Cp I CI p ; c ¼ IRload
dt 2 Cp Lp CI Lp dt Lp dt Lc
We define new variables:
dICc dIL dY2 d 2 ILd dIL dY3 d 2 ILp

Y1 ¼ ; Y2 ¼ d , ¼ 2 ; Y3 ¼ p ) ¼ 2
dY2 dVs ðtÞ 1 Rsource 1

¼ þ Y1 þ I Cc
dt dt Ld Ld Cc Ld
dY3 1 1 Rs dIL Rload

¼ I Cp I C I Y3 ; c ¼ IRload
dt Cp Lp CI Lp Lp dt Lc
We can summery our new system differential equations:
dICc dIL dIL dY2 dVs ðtÞ 1 Rsource 1

¼ Y1 ; d ¼Y2 ; p ¼Y3 ; ¼ þ Y1 þ ICc
dt dt dt dt dt Ld Ld Cc Ld
dY3 1 1 Rs dIL Rload

¼ I Cp I C I Y3 ; c ¼ IRload
dt Cp Lp CI Lp Lp dt Lc
dICc dILd dILp

At fixed points: dt ¼ 0; dt ¼0; dt ¼0; dYdt2 ¼ 0; dYdt3 ¼ 0; dIdtLc ¼ 0
dVs ðtÞ 1 1 1
Y1 ¼ 0; Y2 ¼ 0; Y3 ¼ 0; þ I ¼ 0; IC P IC I ¼ 0; IRload ¼ 0
dt Cc Cc Cp CI
equilibrium
points of PIN diode attenuator and switch circuit in adding to coor-
dinate Y1 ; Y2 ; Y3; ICc ; ILd ; ILp ; ICp ; ICI ; ILc ; IRload arbitrarily small increments of
exponentially form ½y1 ; y2 ; y3 ; iCc ; iLd ; iLp ; iCp ; iCI ; iLc ; iRload ekt and retaining the
first order terms in Y1 ; Y2 ; Y3; ICc ; ILd ; ILp ; ICp ; ICI ; ILc ; IRload . The system of
homogenous equations leads to a polynomial characteristic equation in the eigen-
values. The polynomial characteristic equations accept by set of the below circuit
variables, circuit variables derivative and circuit variables second order derivative
with respect to time into PIN diode attenuator and switch circuit equivalent circuit.
Our PIN diode attenuator and switch circuit equivalent circuit fixed values with
arbitrarily small increments of exponential form ½y1 ; y2 ; y3 ; iCc ; iLd ; iLp ; iCp ; iCI ;
iLc ; iRload ekt are: j = 0 (first fixed point), j = 1 (second fixed point), j = 2 (third
fixed point), etc.,

Y1 ¼ Y1 þ y1 ekt ; Y2 ¼ Y2 þ y2 ekt ; Y3 ¼ Y3 þ y3 ekt ; ICc ¼ ICc þ iCc ekt
ILd ¼ ILd þ iLd ekt ; ILp ¼ ILp þ iLp ekt ; ICp ¼ ICp þ iCp ekt ; ICI ¼ ICI þ iCI ekt
ðjÞ ðjÞ dICc dIL
ILc ¼ ILc þ iLc ekt ; IRload ¼ IRload þ iRload ekt ; ¼ iCc k ekt ; d ¼ iLd k ekt
dt dt
dILp dY dY dI
¼ iLp k ekt ; ¼ y2 k ekt ; ¼ y3 k ekt ; c ¼ iLc k ekt
2 3 L
dt dt dt dt
For k < 0, t > 0, the selected fixed point is stable otherwise k > 0, t > 0 is
unstable. Our system tends to the selected fixed point exponentially for k < 0, t > 0
his absolute value |k| established the speed of flow toward or away from the selected
fixed point [2–4].
We can rewrite our system differential equations:
ðjÞ ðjÞ
iCc k ekt ¼ Y1 þ y1 ekt At fixed point Y1 ¼ 0 then y1 iCc k ¼ 0
ðjÞ ðjÞ
iLd k ekt ¼Y2 þ y2 ekt At fixed point Y2 ¼ 0 then y2 iLd k ¼ 0
ðjÞ ðjÞ
iLp k ekt ¼Y3 þ y3 ekt At fixed point Y3 ¼ 0 then y3 iLp k ¼ 0
dY2 dVs ðtÞ 1 Rsource 1 dVs ðtÞ

¼ þ Y1 þ I Cc ; !e
dt dt Ld Ld Cc Ld dt
ðjÞ Rsource 1 ðjÞ
y2 k ekt ¼ ½Y1 þ y1 ekt þ ½I þ iCc ekt
Ld Cc Ld Cc
ðjÞ Rsource 1 ðjÞ Rsource kt 1
y2 k ekt ¼ Y1 þ I Cc þ y 1 e þ iCc ekt
Ld Cc Ld Ld Cc Ld
ðjÞ ðjÞ
At fixed point Y1 Rsource
Ld þ
1
Cc Ld ICc ¼ 0 then y1 Rsource
Ld þ iCc Cc Ld y2 k ¼ 0
1
1 ðjÞ 1 ðjÞ ðjÞ Rs

y3 k ekt ¼ ½I þ iCp ekt ½I þ iCI ekt ½Y3 þ y3 ekt
Cp Lp Cp CI Lp CI Lp
1 ðjÞ 1 ðjÞ ðjÞ Rs
y3 k ekt ¼ ½ I Cp I C I Y3
Cp Lp CI Lp Lp
1 1 Rs
þ i Cp ekt iCI ekt y3 ekt
Cp Lp CI Lp Lp
ðjÞ ðjÞ ðjÞ

At fixed point 1
Cp Lp ICp CI1Lp ICI Y3 LRps ¼ 0
1 1 Rs
i Cp iCI y3 y3 k ¼ 0
Cp Lp CI Lp Lp
Rload ðjÞ
iLc k ekt ¼ ½IRload þ iRload ekt ; iLc k ekt
Lc
Rload ðjÞ Rload kt
¼ IRload þ iRload e
Lc Lc
ðjÞ
At fixed point RLload c
IRload ¼ 0 then iRload RLload c
i Lc k ¼ 0
PIN
diode attenuator and switch circuit
system matrix
y1 ; y2 ; y3 ; iCc ; iLd ; iLp ; iCp ; iCI ; iLc ; iRload can be constructed from the below list of
equations:
Rsource 1
y1 iCc k ¼ 0; y2 iLd k ¼ 0; y3 iLp k ¼ 0; y1 þ iCc y2 k
Ld Cc Ld
¼0
1 1 Rs Rload
i Cp iCI y3 y3 k ¼ 0; iRload iLc k ¼ 0
Cp Lp CI Lp Lp Lc
PIN diode attenuator and switch circuit system matrixes:

0 1
iCc
B C 0 1
0 1 B iLd C 0 1 y1
i11 . . . i16 B C u11 . . . u14 B C
B C
B .. .. .. C B iLp C B ..
þ .. .. C B
B
iCp C
C¼0
@ . . . A B C @ . . . A B
B y2 C
B C @ iCI C
A
i61 i66 By C u61 u64
@ 3A iRload
iLc
i11 ¼ k; i12 ¼ . . . ¼ i16 ¼ 0; i21 ¼ 0; i22 ¼ k; i23 ¼ 0; i24 ¼ 1; i25 ¼ i26 ¼ 0
1
i31 ¼ i32 ¼ 0; i33 ¼ k; i34 ¼ 0; i35 ¼ 1; i36 ¼ 0; i41 ¼ ; i42 ¼ i43 ¼ 0; i44
Cc Ld
¼ k; i45 ¼ i46 ¼ 0
Rs
i51 ¼ . . . ¼ i54 ¼ 0; i55 ¼ k ; i56 ¼ 0; i61 ¼ . . . ¼ i65 ¼ 0; i66 ¼ k; u11
Lp
¼ 1; u12 ¼ u13 ¼ u14 ¼ 0
Rsource
u21 ¼ . . . ¼ u24 ¼ 0; u31 ¼ . . . ¼ u34 ¼ 0; u41 ¼ ; u42 ¼ u43 ¼ u44 ¼ 0
Ld
1 1
u51 ¼ 0; u52 ¼ ; u53 ¼ ; u ¼ 0; u61 ¼ u62 ¼ u63 ¼ 0; u64
Cp Lp CI Lp 54
Rload
¼
Lc
We consider
0 1
i Cc
B C
0 1 0 1 B i Ld C
u11 ... u14 i11 . . . i16 B C
B C
B .. .. .. C ! e; B .. .. .. C B iLp C
@ . . . A @ . . . A B C0
B y2 C
u61 u64 i61 i66 B C
By C
@ 3A
iLc
0 1 0 1
i11 . . . i16 i11 ... i16
B .. .. .. C B .. .. .. C
AkI ¼@ . . . A ; det(A k IÞ ¼ 0 ) det@ . . . A¼0
i61 i66 i61 i66

[2–4].
Exercises
1. We have amplifier system which contains BJT transistor amplifier. The amplifier
is operated as a Small Signal Amplifier. We represent our BJT transistor by
Hybrid Pi model of BJT. Our amplifier input and output matching networks are
T-type. Amplifier’s load is represented by parallel resistor (Rload) and capacitor
(Cload). We feed the amplifier by input RF source.
1:1 Draw our amplifier system by using BJT transistor linear Hybrid Pi model,
input and output block capacitors, input and output T-type matching net-
works, amplifier load circuit and input RF source.
1:2 Write our amplifier system differential equations. Find fixed points and
discuss stability. Remark: We consider input RF source Xs ðtÞ ¼
A0 þ nðtÞ; jnðtÞj\1&A0 [ [ jnðtÞj then Xs ðtÞjA0 [ [ jnðtÞj ¼ A0 þ nðtÞ A0 ;
dXs ðtÞjA [ [ jnðtÞj
0
dt¼ dnðtÞ
dt ! e.
1:3 How our amplifier system stability and dynamical behavior change if the
input matching network is T-type and output matching network is Pi-type?
1:4 We change our amplifier load circuit to parallel resistor (Rload) and induc-
tance (Lload). How the dynamical behavior and stability of the circuit
change? Find fixed points.
2. We have amplifier system which contains two BJT transistors amplifiers in
cascade. The amplifiers are operated as a Small Signal Amplifiers. We represent
our BJT transistors by two Hybrid Pi model of BJTs. Our first amplifier input
and output matching networks are Pi-type and second amplifier input and output
matching networks are T-type. Last amplifier’s load is represented by parallel
inductance (Lload) and capacitor (Cload). We feed the two amplifiers by input RF
source.
2:1 Draw our amplifiers system by using BJT transistors linear Hybrid Pi
models, input and output block capacitors, first and second amplifiers
matching networks, last amplifier load circuit and input RF source.
2:2 Write our amplifiers system differential equations. Find fixed points and
discuss stability. Remark: We consider first amplifier circuit input RF source
Xs ðtÞ ¼ A0 þ nðtÞ; ; jnðtÞj\1 & A0 [ [ jnðtÞj then
dXs ðtÞj
Xs ðtÞjA0 [ [ jnðtÞj ¼ A0 þ nðtÞ A0 ; A0 [ [ jnðtÞj
dt ¼ dnðtÞ
dt ! e
2:3 How our amplifier system stability and dynamical behavior change if the
load is pure resistive (Rload)?
2:4 How our amplifier system stability and dynamical behavior change if all
input and output matching networks are L-type?
3. We have Small Signal (SS) amplifier with PI and T’s matching networks
equivalent circuit. Under AC and Small Signal (SS) conditions, we replace the
BJT transistor with linear Hybrid PI model. Input matching network constructed
from two Pi-type matching networks in cascade. Output matching network can
be L-type (switch S1 position A) or PI-type (switch S1 position B).
Exercises 503
3:1 Switch S1 is in (A) position, Find circuit differential equations and fixed
points. Discuss stability of the circuit.
3:2 Switch S2 is in (B) position, Find circuit differential equations and fixed
point s. Discuss stability of the circuit.
3:3 Capacitors Ca11 and Ca12 are disconnected. How the circuit dynamical
behavior and stability is changed?
3:4 Capacitor Cm is a function of Ca11 and Ca12. Cm ¼ Ca11 C1 þ Ca12 C2
C1 ¼ a C2 ; C1 ; C2 ; a 2 R þ . How the dynamical behavior and stability of
the circuit change for different values of a parameter?
Remark: We consider amplifier circuit input RF source Xs ðtÞ ¼
A0 þ nðtÞ; jnðtÞj\1 & A0 jnðtÞj then
dXs ðtÞjA [ [ jnðtÞj
Xs ðtÞjA0 [ [ jnðtÞj ¼ A0 þ nðtÞ A0 ; 0
dt ¼ dnðtÞ
dt ! e
4. We have balanced amplifier which contains two BJT NPN transistors (Q1 and
Q2), input and output matching networks (RM1, Cc1, RM2, Cc2, RM3, Cc3, RM4,
Cc4), feedback resistor (Rf), two input RF sources (Vs1(t) and Vs2(t)), biasing
resistors (Rb1, Rc1, Rb2, Rc2), and output load (Rload)
Use BJT NPN transistor’s Small Signal (SS) Hybrid Pi equivalent circuit in your
analysis.
4:1 Find circuit differential equations and fixed points. Discuss stability of the
circuit.
4:2 Resistor Rm4 is disconnected, how it influences on the circuit dynamical
behavior and stability.
4:3 We short capacitors Cc2 and Cc3. How the circuit dynamical behavior and
stability is changed?
4:4. Capacitors Cc3 and Cc4 are functions of capacitors Cc1 and Cc2.
Cc3 ¼ Cc1 C1 þ Cc2 ð1 C1 Þ; Cc4 ¼ Cc1 C2 þ Cc2 ð1 C2 Þ

C1 6¼ C2 ; 0\C1 ; C2 \1; C1 ; C2 2 R þ
How the dynamical behavior and stability of the circuit change for different
values of C1 and C2 parameters?
4:5 Circuit feedback resistor (Rf) is disconnected, How the dynamical behavior
and stability of the circuit is changed?
Remarks: We consider balance amplifier circuit input RF sources Vs1 ðtÞ ¼
A01 þ n1 ðtÞ; jn1 ðtÞj\1; A01 jn1 ðtÞj then Vs1 ðtÞjA01 jn1 ðtÞj ¼ A01 þ n1 ðtÞ A01 ;
dVs1 ðtÞjA [[ jn ðtÞj
01
dt
1
¼ dndt1 ðtÞ ! e; Vs2 ðtÞ ¼ A02 þ n2 ðtÞ; jn2 ðtÞj\1; A02 jn2 ðtÞj then
dVs2 ðtÞjA [[ jn ðtÞj dn2 ðtÞ
Vs2 ðtÞjA02 [[ jn2 ðtÞj ¼ A02 þ n2 ðtÞ A02 ; 02
dt
2
¼ dt !e
5. We have Common Emitter (CE) and Common Base (CB) BJT transistors
amplifier circuit. Q1 is connected as CE and Q2 is connected as CB. Common
Emitter (CE) amplifier, input signal inserted at base (B) and output signal is
taken from the collector (C). The CE amplifier’s output voltage is shifted
by *180° in phase compared to CE amplifier’s input signal. Common Base
(CB) amplifier, input signal inserted at emitter (E) and output signal taken from
the collector (C). The CB amplifier can operate as a voltage amplifier for low
input impedance.
Exercises 505
We define two circuit outputs. First circuit output (Vout1) is a voltage phase shift
by 180° and second circuit output (Vout2) has no voltage phase shift. The input RF
source is Vs(t) with serial resistor Rs and is injected to Q1 base. There are three
possible loads connections to out circuit, Rload1, Rload2, and Rload. R1, R2, R3, and
R4 are circuit biasing voltage resistors. Under AC and Small Signal (SS) conditions,
we replace the BJT transistors (Q1 and Q2) with linear Hybrid PI model. Switches
S1 and S2 connect the loads to our circuit. In the current circuit there are no
matching networks between the circuit amplifier and loads (Rload1, Rload2, and
Rload).
5:1 Find circuit differential equations and fixed points. Discuss stability of the
circuit for the cases: (1) S1 is in position A and S2 is in position A, (2) S1 is in
position A and S2 is in position B, (3) S1 and S2 are in position B, (4) S1 is in
position B and S2 in position A.
5:2 Resistor R1 is disconnected, how it influences on the circuit dynamical
behavior and stability?
5:3 We add L—matching network between output one and Rload1. Switches S1 and
S2 are in position A. Find circuit differential equations and fixed points.
Discuss stability of the circuit.
5:4 We add PI matching network between output two and Rload2. Switches S1 and
S2 are in position A. Find circuit differential equations and fixed points.
Discuss stability of the circuit.
Remarks: We consider CE and CB transistor amplifier circuit input RF sources

Vs ðtÞ ¼ A0 þ nðtÞ; jnðtÞj\1; A0 jnðtÞj then Vs ðtÞjA0 jnðtÞj ¼ A0 þ nðtÞ A0 ;
dVs ðtÞjA [[ jnðtÞj
0
dt ¼ dnðtÞ
dt ! e
6. We have Bias-T circuit with 4th order filter. Bias-T’s RF choke is L2 and
capacitor is C1. We consider our 4th order filter with Bias-T feed coax cable
with RF + DC signal. The coax and the remote device which is connected to the
coax cable are taken as a pure resistive load (Rload). Due to parasitic effects of
the circuit microstrip lines there are some delay elements in time in our circuit
ðs1 ; s2 ; s3 Þ. Assumptions: The voltages on the microstrip represented delay lines
are neglected Vsi ! eði ¼ 1; 2; 3Þ and the delay is in the current that flows
through each delay line IðtÞ ! Iðt sÞ. V1—DC voltage source, V2—RF
voltage source. Tau1 $ s1 ; Tau2 $ s2 ; Tau3 $ s3
6:1 Write 4th order filter with Bias-T circuit differential Eqs.
6:2 Discuss stability and dynamical behavior of the circuit for
si ! eð¼ 0Þ 8 i ¼ 1; 2; 3. Find circuit fixed points and characteristic Eq.
6:3 Discuss stability and stability switching for si [ 0 8 i ¼ 1; 2; 3 due to dif-
ferent values of s parameter s1 ¼ s; s2 ¼ s2 ; s3 ¼ s þ 1 .
6:4 How the circuit dynamical behavior changes if L3 is disconnected?
6:5 How the circuit dynamical behavior changes if we short inductor L1?
6:6. Discuss stability and stability switching for si [ 0 8(i ¼ 1; 2; 3 due )to!dif-
0 for s\s C
ferent values of s parameter s1 ¼ s; s2 ¼ s2 ; s3 ¼
s3 for s sC
Exercises 507
sC is a critical delay parameter value. s3 ðsÞ ¼ s3 Uðs sC Þ. How the cir-

cuit stability and stability switching is dependent on the critical delay
parameter (sC )?
7. We have Bias-T circuit which is driving VCSEL (Vertical Cavity Surface

Emitting Laser) diode (or other laser diode). It can be done with no IC.
The VCSEL or laser diode is biased with a DC current until it just begins to lase.
An RF sine wave then applied to this laser through a Bias-T network.
The VCSEL or laser diode is biased with a DC current until it just begins to lase.
An RF sine wave is then applied to this laser through a Bias-T network. The
internal capacitance and structure of these small lasers then does something.
Instead of slowly turning ON and OFF with the application of the RF, the diode
is driven for threshold during one part of the sine wave and then begin to store
energy on the opposite swing of the wave. When it has reached a certain level of
stored energy (gain), it “snaps” on and releases all of this energy as laser light.
The laser then turns off because all of the gain was extracted. The phenomenon
is called “gain-switching”. Tuning the laser diode: if DC is too high, the laser
diode may produce light all the time. If the RF is too high, the laser diode may
produce an “after pulse”. If either is too low, the laser diode won’t produce the
desired amplitude. Due to parasitic effects of the circuit microstrip lines there are
some delay elements in time in our circuit ðs1 ; s2 Þ. Assumptions: The voltages
on the microstrip represented delay lines are neglected Vsi ! eði ¼ 1; 2Þ and the
delay is in the current that flows through each delay line IðtÞ ! Iðt sÞ.
7:1 Write VCSEL diode driving Bias-T circuit differential equations. Take into
consideration in your differential equations the full equivalent circuit of
VCSEL diode.
7:2 Discuss stability and dynamical behavior for the circuit si ¼ 0; i ¼ 1; 2.

7:3 Discuss stability and stability switching for si 6¼ 0; si [ 0 8 i ¼ 1; 2 due to
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
different values of s parameter s1 ¼ s2 ; s2 ¼ s þ s2 þ C s; C 2 R . How
the stability changes for different values of C parameter (s is constant)?
7:4 Discuss stability and stability switching
( for si [ 0 8 i ¼ 1; 2 due)to!different
s þ 1 for 0 \s\sC
values of s parameter s1 ¼ s4 ; s2 ¼ . sC is a
s3 for s sC
critical delay parameter value. How the circuit stability and stability
switching is dependent on the critical delay parameter (sC )? s2 ðsÞ ¼
ð s þ 1Þ ½UðsÞ Uðs sC Þ þ s3 Uðs sC Þ
Cp—pad capacitance, Li—inductance of interconnect metal, Rm—resistance

from mirror stack, Ca—aperture capacitance, Ra—aperture resistance.
8. We have SPST (Single Pole Single Throw) PIN diode circuit. The output RF
port is connected to load resistance (Rload). We consider the RF source as a high
frequency signal and use PIN diode high frequency equivalent circuit. Consider
dVs ðtÞ
dt ! e
Exercises 509
8:1 Draw PIN diode SPST circuit equivalent circuit.

8:2 Write circuit differential equations and find fixed points.
8:3 Find circuit eigenvalues expressions and discuss stability switching for
parameter values variation.
8:4 We replace inductor Ld by resistor Rd. How the circuit dynamic changes?
Find circuit differential equations and fixed points. Discuss stability.
8:5 We add PIN diode D2 in parallel to D1. How the circuit dynamical behavior
changes? Find circuit differential equations and fixed point. Discuss
stability.
9. We have SPST with series and shunt PIN didoes (D1 and D2). The output RF
port is connected to load resistance (Rload). We consider the RF source as a high
frequency signal and use PIN diode high frequency equivalent circuit. Consider
dVs ðtÞ
dt ! e.
9:1 Draw PIN diodes (D1 and D2) SPST circuit equivalent circuit.
parameters variation.
9:4 Inductor Ls is disconnected, How the circuit dynamic changes? Find circuit
differential equations and fixed points. Discuss stability.
9:5 We disconnect diode D2, How the circuit dynamic changes? Find circuit
10. We have high isolation generic PIN SPST circuit. PIN diode switches can
handle very large power signals. The important diode parameters for switches
that must handle power levels higher than 1 w include the diode’s voltage
rating and thermal resistance. Other diode parameters, such as series resistance,
capacitance and I layer thickness, are also contributing factors to the determi-
nation of maximum power handling. We inject the RF signal (Vs(t)) through
circuit RF common port and can switch it to RF Out1 or RF Out1. We have two
bias voltage sources to our circuit (Vss1 and Vss2). Circuit RF chokes are Ld1,
Ld2, and Ld3. Three Cblock capacitors (Cb1, Cb2, and Cb3). Two Cfilter capacitors
(Cf1 and Cf2). Two PIN diodes are low capacitance and the other two PIN
diodes are low resistance. Rload1 and Rload2 are our circuit’s load resistances.
We consider for simplicity dVdts ðtÞ ! e. Circuit PIN diodes parameters are not the
Exercises 511
same and there is slightly differences. Use PIN diode high frequency equivalent
circuit in your analysis.
Ld ¼ Ld1 ; Ld2 ¼ C1 Ld1 ; Ld3 ¼ ð1 C1 Þ Ld1 ; 0\C1 \1; Cf

¼ Cf 1 ; Cf 2 ¼ ð1 C2 Þ Cf 1 ; 0\C2 \1
10:1 Draw PIN diodes (D1,…,D4) SPST circuit equivalent circuit.

parameters C1, C2 variation.
10:4 Diode D2 is disconnected, how the dynamical of the circuit is changes?
Find circuit differential equations and discuss stability.
10:5 We short diode D3, how the dynamical of the circuit is changed? Find
circuit differential equations and discuss stability.
10:6 We short Rload2, How the dynamical of the circuit is changed? Discuss
stability switching for parameters C1, C2 variations.
10:7 We disconnect Cfilter capacitors, How the dynamical of the circuit is
changed? Discuss stability switching for circuit parameters variations.
Chapter 5
Power Amplifier (PA) System Stability
Analysis
Generally, large signal or Power Amplifier (PA) are used in the output stages of
audio amplifier systems to derive a loudspeaker load. Power amplifier must be able
to supply the high peak currents required to drive the low impedance speaker. One
method used to distinguish the electrical characteristics of different types of
amplifiers is by “class”, and as such amplifiers are classified according to their
circuit configuration and method of operation. Then Amplifier Classes is the term
used to differentiate between the different amplifier types. Amplifier Classes rep-
resent the amount of the output signal which varies within the amplifier circuit over
one cycle of operation when excited by a sinusoidal input signal. The classification
of amplifiers range from entirely linear operation (for use in high-fidelity signal
amplification) with very low efficiency, to entirely non-linear (where a faithful
signal reproduction is not so important) operation but with a much higher effi-
ciency, while others are a compromise between the two. Amplifier classes are
mainly lumped into two basic groups. The first are the classically controlled con-
duction angle amplifiers forming the more common amplifier classes of A, B, AB
and C, which are defined by the length of their conduction state over some portion
of the output waveform, such that the output stage transistor operation lies some-
where between being “fully-ON” and “fully-OFF”. The second set of amplifiers are
the newer so-called “switching” amplifier classes of D, E, F, G, S, T etc., which use
digital circuits and pulse width modulation (PWM) to constantly switch the signal
between “fully-ON” and “fully-OFF” driving the output hard into the transistors
saturation and cut-off regions. The most commonly constructed amplifier classes are
those that are used as audio amplifiers, mainly class A, B, AB and C. Class A
Amplifiers are the most common type of amplifier class due mainly to their simple
design. Class A, literally means “the best class” of amplifier due mainly to their low
signal distortion levels and are probably the best sounding of all the amplifier
classes mentioned here. The class A amplifier has the highest linearity over the
other amplifier classes and as such operates in the linear portion of the character-
istics curve. Class B amplifiers were invented as a solution to the efficiency and
heating problems associated with the previous class A amplifier. The basic class B
DOI 10.1007/978-3-319-45427-6_5
514 5 Power Amplifier (PA) System Stability Analysis
amplifier uses two complimentary transistors either bipolar of FET for each half of
the waveform with its output stage configured in a “push-pull” type arrangement, so
that each transistor device amplifies only half of the output waveform. Class AB
Amplifier is a combination of the “Class A” and the “Class B” type amplifiers.
The AB classification of amplifier is currently one of the most common used types
of audio power amplifier design. The class AB amplifier is a variation of a class B
amplifier, except that both devices are allowed to conduct at the same time around
the waveforms crossover point eliminating the crossover distortion problems of the
previous class B amplifier. The Class C Amplifier design has the greatest efficiency
but the poorest linearity of the classes of amplifiers mentioned here. The previous
classes, A, B and AB are considered linear amplifiers, as the output signals
amplitude and phase are linearly related to the input signals amplitude and phase.
Class C amplifier is heavily biased so that the output current is zero for more than
one half of an input sinusoidal signal cycle with the transistor idling at its cut-off
point. The conduction angle for the transistor is significantly less than 180°, and is
generally around the 90° area. We analyse the stability of these amplifiers by
inspecting the equivalent circuit differential equations, fixed points, bifurcation and
stability switching for circuit parameters variation. BJT transistor is replaced by
large signal model in our analysis.
We use in our analysis the Bipolar transistor model for large signal circuit
simulation: The BJT model used in circuit simulation can accurately represent the
DC and dynamic currents of the transistor in response to VBE(t) and VCE(t).
A typical circuit simulation model or compact model is made of the Ebers-Moll
model when VBE and VBC are two driving forces for IC and IB, plus additional
enhancements for high level injection, voltage dependent capacitances that accu-
rately represent the charge storage in the transistor, and parasitic resistances.
This BJT model is known as the Gummel-Poon model. The Ebers-Moll BJT model
is a good large signal. if the inputs/outputs exceed certain limits, the full Ebers-Moll
model must be used. When certain parameters are omitted, the Gummel–Poon
model reduces to the simpler Ebers–Moll model. Gummel-Poon nonlinear model is
the “large signal model”. Large signal models is closer to reality but is computa-
tionally complex or even intractable. Additionally we discuss the stability of
wideband LNA with negative feedback.
5.1 Class AB Push-Pull Power Amplifiers Stability

Analysis Under Parameters Variation
Class AB Amplifier is a combination of the “Class A” and the “Class B” type

amplifiers. The AB classification of amplifier is currently one of the most common
used types of audio power amplifier design. Push-pull mechanism is essential for
realizing practical class AB power amplifiers. Circuit diagram of a typical class AB
push-pull amplifier is shown in the next figure. The technical designation is
5.1 Class AB Push-Pull Power … 515
“complementary symmetry class AB power amplifier”. The active elements used in

this circuit (Q1 and Q2) are complementary symmetric transistors and it means the
transistors are similar in all aspects except one is NPN and the other is PNP. The
use of this complementary pair eliminates the bulky transformer for phase splitting
the input signal for driving the individual transistor. The NPN transistor alone will
conduct the positive half cycle and PNP transistor alone will conduct the negative
half cycle. Pre-biasing is given to the transistors using the network comprising of
resistors R1, R2 and biasing diodes D1 and D2. The NPN transistor will start
conducting when its base voltage is above the base emitter voltage (VBE 0.7 v)
and a PNP transistor will start conducting when its base voltage is below the base
emitter voltage (VBE −0.7 v). A forward biased diode will drop approximately
0.7 v across it and the biasing diodes used here will keep the transistor slightly
forward biased even if there is no input signal. Compensating diodes (D1, D2)
characteristics must match as close as possible to the transistors. Resistors R1 and
R2 are used for forward biasing the diodes and they drop 0.7 v across it for biasing
the individual transistors (Q1, Q2). C1 and C2 are input DC decoupling capacitors.
The advantages of class AB power amplifier: no cross over distortion, no need for
the bulky coupling transformers, and no hum in the output. The disadvantages of
class AB power amplifier: efficiency is slightly less when compared to class B
configuration, there will be some DC components in the output as the load is
directly coupled, and capacitive coupling can eliminate DC components but it is not
practical in case of heavy loads. For simplicity in our circuit analysis we reduces
BJT transistors Gummel–Poon model to the simpler Ebers–Moll model [24–26]
(Fig. 5.1).
Fig. 5.1 Class AB push-pull amplifier

VBEQ VBCQ VBEQ VBCQ

Is 1 1 1 Is 1
IEQ1 ¼ ðe VT 1Þ Is ðe VT 1Þ; ICQ1 ¼ Is ðe VT 1Þ ðe VT 1Þ
af ar
VBEQ VBCQ VEBQ VCBQ
Is 1 Is 1 Is 2 2
IBQ1 ¼ ðe VT 1Þ þ ðe VT 1Þ; IEQ2 ¼ ðe VT 1Þ Is ðe VT 1Þ
bf br af
VEBQ VCBQ VEBQ VCBQ
2 Is 2 Is 2 Is 2
ICQ2 ¼ Is ðe VT 1Þ ðe VT 1Þ; IBQ2 ¼ ðe VT 1Þ þ ðe VT 1Þ
ar bf br
ID ID
VD1 ¼ Vt lnð 1 þ 1Þ; VD2 ¼ Vt lnð 2 þ 1Þ; ID ¼ ID1 ¼ ID2 ; IEQ1 ¼ IRload þ IEQ2
I0 I0
It can be shown that aF ISE ¼ aR ISC ¼ IS (see S.M. Sze, Physics of

Semiconductor Devices) (Fig. 5.2).
Fig. 5.2 Class AB push-pull amplifier equivalent circuit

Vcc VB1 VB d
ICQ1 þ IBQ1 ¼ IEQ1 ; IEQ2 ¼ ICQ2 þ IBQ2 ; IR1 ¼ ; IR2 ¼ 2 ; IC1 ¼ C1 ðVA1 VB1 Þ
R1 R2 dt
d Isc
IC2 ¼ C2 ðVA1 VB2 Þ; IRs ¼ IC1 þ IC2 ; IR1 þ IC1 ¼ IBQ1 þ ID1 ¼ IBQ1 þ ID ; lnð Þ 0
dt Ise
V A3 Vs ðtÞ VA1
ID2 þ IC2 þ IBQ2 ¼ IR2 ; IRload ¼ ; VECQ2 ¼ VA3 ; VCEQ1 þ VECQ2 ¼ Vcc ; IRs ¼
Rload Rs
Collector emitter voltage expression for BJT NPN and PNP transistor
½ar IC IE þ ðar af 1Þ Ise

VCENPN VT lnf g;
½IC af IE þ ðaf ar 1Þ Isc
IC af IE þ ðaf ar 1Þ Isc
VCEPNP VT lnf g
ar IC IE þ ðar af 1Þ Ise
we can write the following circuit collector emitter equations for Q1 and Q2:
½ar ICQ1 IEQ1 þ ðar af 1Þ Ise

VCEQ1 VT lnf g;
½ICQ1 af IEQ1 þ ðaf ar 1Þ Isc
ICQ af IEQ2 þ ðaf ar 1Þ Isc
VCEQ2 VT lnf 2 g
ar ICQ2 IEQ2 þ ðar af 1Þ Ise
VECQ2 ¼ VCEQ2 ; VECQ2 VT lnf g;
ICQ2 af IEQ2 þ ðaf ar 1Þ Isc
VCEQ1 þ VECQ2 ¼ Vcc

VT lnf g
þ VT lnf g ¼ Vcc
We can rewrite transistors collector and emitter currents:
ICQ1 ¼ bf IBQ1 þ ðbf þ 1Þ ICBQ1 0 ; IEQ1 ¼ ðbf þ 1Þ IBQ1 þ ðbf þ 1Þ ICBQ1 0

ICQ2 ¼ bf IBQ2 þ ðbf þ 1Þ ICBQ2 0 ; IEQ2 ¼ ðbf þ 1Þ IBQ2 þ ðbf þ 1Þ ICBQ2 0
Remark Q1 and Q2 are complementary symmetric transistors and it means the

transistors are similar in all aspects (parameters are the same, βf, αf,…) except one is
NPN and the other is PNP.
af bf
bf : 20 ! 500; bf ¼ ; af ¼ ; af : 0:95 ! 0:99; bf [ br
1 af 1 þ bf
ar br
br : 0 ! 20; br ¼ ; ar ¼ ; ar : 0 ! 0:95; af [ ar
1 ar 1 þ br
IEQ1
IEQ1 ¼ ðbf þ 1Þ IBQ1 þ ðbf þ 1Þ ICBQ1 0 ; IBQ1 ¼ ICBQ1 0
ðbf þ 1Þ
IEQ2
IEQ2 ¼ ðbf þ 1Þ IBQ2 þ ðbf þ 1Þ ICBQ2 0 ; IBQ2 ¼ ICBQ2 0
ðbf þ 1Þ
bf IEQ1 bf IEQ2
ICQ1 ¼ þ ICBQ1 0 ; ICQ2 ¼ þ ICBQ2 0
ðbf þ 1Þ ðbf þ 1Þ
Vcc VB1 VB
IR1 ¼ ) IR1 R1 ¼ Vcc VB1 ; VB1 ¼ Vcc IR1 R1 ; IR2 ¼ 2 ) VB2
R1 R2
¼ IR2 R2
Z
d 1
IC1 ¼ C1 ðVA1 VB1 Þ ) VA1 VB1 ¼ IC1 dt;
dt C1
Z
1
VA1 ¼ VB1 þ IC1 dt
C1
Z Z
d 1 1
IC1 ¼ C1 ðVA1 VB1 Þ ) VA1 VB1 ¼ IC1 dt; VA1 ¼ VB1 þ IC1 dt
dt C1 C1
Z Z
d 1 1
IC2 ¼ C2 ðVA1 VB2 Þ ) VA1 VB2 ¼ IC2 dt; VA1 ¼ VB2 þ IC2 dt
dt C2 C2
Z Z
1 1
VB 1 þ IC1 dt ¼ VB2 þ IC2 dt;
C1 C2
Z Z
1 1
Vcc IR1 R1 þ IC1 dt ¼ IR2 R2 þ IC2 dt
C1 C2
Z Z
d 1 d 1
fVcc IR1 R1 þ IC1 dtg ¼ fIR R2 þ IC2 dtg
dt C1 dt 2 C2
dVcc dIR1 1 dIR2 1 dVcc

R1 þ I C1 ¼ R2 þ I C2 ; !e
dt dt C1 dt C2 dt
dIR1 dIR2
We get the equation: dt R1 þ 1
C1 I C1 ¼ dt R2 þ 1
C2 I C2
VA3
IRload ¼ ) VA3 ¼ IRload Rload ; VA3 ¼ VECQ2
Rload
¼ VT lnf g
VT ar ICQ2 IEQ2 þ ðar af 1Þ Ise

IRload ¼ lnf g;
Rload ICQ2 af IEQ2 þ ðaf ar 1Þ Isc
ID
Vcc ¼ IR1 R1 þ 2 lnð þ 1Þ þ IR2 R2
I0
VBEQ VBCQ
Is 1 1
IEQ1 ¼ IRload þ IEQ2 ) ðe VT 1Þ Is ðe VT 1Þ
af
VEBQ VCBQ
Is 2 2
¼ IRload þ ðe VT 1Þ Is ðe VT 1Þ
af
IR1 þ IC1 ¼ IBQ1 þ ID ) IR1 ¼ IBQ1 þ ID IC1 ; IR2 ¼ ID þ IC2 þ IBQ2
We add one equation to the other and get the following expression:
ðÞ IR1 ¼ IBQ1 þ ID IC1 ; ðÞ IR2 ¼ ID þ IC2 þ IBQ2 ; ðÞ þ ðÞ ) IR1 þ IR2
¼ IBQ1 þ IBQ2 þ IC2 IC1 þ 2 ID
dIR1 1 dIR2 1 dIR dIR2

R1 þ I C1 ¼ R2 þ IC2 ) ð 1 R1 þ R2 Þ
dt C1 dt C2 dt dt
1 1
¼ I C2 I C1
C2 C1
dIR1 dIR2
If C = C1 = C2 then ð dt R1 þ dt R2 Þ ¼ C1 ðIC2 IC1 Þ
IR1 þ IR2 ¼ IBQ1 þ IBQ2 þ IC2 IC1 þ 2 ID ) IC2 IC1

¼ IR1 þ IR2 ðIBQ1 þ IBQ2 Þ 2 ID
X
2 X
2
dIR1 dIR2
I C2 I C1 ¼ IR k IBQk 2 ID ; ð R1 þ R2 Þ
k¼1 k¼1
dt dt
1 X 2 X2
¼ ð IRk IBQk 2 ID Þ
C k¼1 k¼1
ID
Vcc ¼ IR1 R1 þ 2 lnð þ 1Þ þ IR2 R2 ;
I0
1
ID ¼ I0 fexp½ ðVcc IR1 R1 IR2 R2 Þ 1g
2
dIR1 dIR2 1 X 2 X 2
ð R1 þ R2 Þ ¼ ð IRk IBQk 2 I0
dt dt C k¼1 k¼1
1
fexp½ ðVcc IR1 R1 IR2 R2 Þ 1gÞ
2

IRload ¼ IEQ1 IEQ2 ; IRload ¼ lnf g

IEQ1 IEQ2 ¼ lnf g
b I
VT ar ½ðbf þEQ1Þ2 þ ICBQ2 0 IEQ2 þ ðar af 1Þ Ise
IEQ1 IEQ2 ¼ lnf b IEQ f g;
Rload f 2
þ ICBQ 0 af IEQ þ ðaf ar 1Þ Isc
ðbf þ 1Þ 2 2
VCEQ1 þ VECQ2 ¼ Vcc
b I
VT ar ½ðbf þEQ1Þ2 þ ICBQ2 0 IEQ2 þ ðar af 1Þ Ise
n1 ðIEQ1 ; IEQ2 Þ ¼ lnf b IEQ f g
Rload f 2
þ ICBQ 0 af IEQ þ ðaf ar 1Þ Isc
ðbf þ 1Þ 2 2
IEQ1 þ IEQ2 ¼ 0

VT lnf g
þ VT lnf g ¼ Vcc
b IEQ1
ar ½ðbf þ ICBQ1 0 IEQ1 þ ðar af 1Þ Ise
f þ 1Þ
lnf b I
g
½ðbf þEQ1Þ1 þ ICBQ1 0 af IEQ1 þ ðaf ar 1Þ Isc
f
b IEQ2
ar ½ðbf þ ICBQ2 0 IEQ2 þ ðar af 1Þ Ise Vcc
f þ 1Þ
þ lnf b I
g¼
½ðbf þEQ1Þ2 þ ICBQ2 0 af IEQ2 þ ðaf ar 1Þ Isc VT
f
b IEQ1
f þ 1Þ
ln½f b I
g
f
b IEQ2
ar ½ðbf þ ICBQ2 0 IEQ2 þ ðar af 1Þ Ise Vcc
f þ 1Þ
f b I
g ¼
½ðbf þEQ1Þ2 þ ICBQ2 0 af IEQ2 þ ðaf ar 1Þ Isc VT
f
b IEQ1
½VVcc f þ 1Þ
e T ¼f b I
g
f
b IEQ2
f þ 1Þ
f b I
g
f
b IEQ1
f þ 1Þ
n2 ðIEQ1 ; IEQ2 Þ ¼ f b I
g
f
b IEQ2
f þ 1Þ
f b I
g;
f
n2 ðIEQ1 ; IEQ2 Þ ¼ e½ VT
Vcc
We need to solve the following two equations:
½Vcc
n1 ðIEQ1 ; IEQ2 Þ ¼ 0; n2 ðIEQ1 ; IEQ2 Þ ¼ e VT
Assumptions: ar : 0 ! 0:95; ar ¼ 0:95 1; bf : 20 ! 500; bf 1 ) bf þ 1 bf
bf bf
1; af : 0:95 ! 0:99; af 1; ð af Þ ! e ! 0; X ðar ; bf Þ
ar bf
¼ 1
ðbf þ 1Þ
a b
VT ððb r þ f1Þ 1Þ IEQ2 þ ar ICBQ2 0 þ ðar af 1Þ Ise
n1 ðIEQ1 ; IEQ2 Þ ¼ lnf f b g
Rload ð f af Þ IEQ þ ICBQ 0 þ ðaf ar 1Þ Isc
ðbf þ 1Þ 2 2
IEQ1 þ IEQ2 ¼ 0
a b
VT ððb r þ f1Þ 1Þ IEQ2 þ ar ICBQ2 0 þ ðar af 1Þ Ise
n1 ðIEQ1 ; IEQ2 Þ ¼ lnf f g
Rload ICBQ2 0 þ ðaf ar 1Þ Isc
IEQ1 þ IEQ2 ¼ 0
C1 ðar ; af ; Ise ; ICBQ2 0 Þ ¼ ar ICBQ2 0 þ ðar af 1Þ Ise ; C2 ðar ; af ; Isc ; ICBQ2 0 Þ

¼ ICBQ2 0 þ ðaf ar 1Þ Isc
VT Xðar ; bf Þ IEQ2 þ C1 ðar ; af ; Ise ; ICBQ2 0 Þ

n1 ðIEQ1 ; IEQ2 Þ ¼ lnf g IEQ1 þ IEQ2
Rload C2 ðar ; af ; Isc ; ICBQ2 0 Þ
¼0
Xðar ; bf Þ IEQ2 þ C1 ðar ; af ; Ise ; ICBQ2 0 Þ Rload

lnf g ¼ ðIEQ1 IEQ2 Þ
C2 ðar ; af ; Isc ; ICBQ2 0 Þ VT
Rload Xðar ; bf Þ IEQ2 þ C1 ðar ; af ; Ise ; ICBQ2 0 Þ

expfðIEQ1 IEQ2 Þ g¼
VT C2 ðar ; af ; Isc ; ICBQ2 0 Þ
Rnload
X1 ðIEQ IEQ Þn
Rload 1 2 VTn
expfðIEQ1 IEQ2 Þ g¼
VT n¼0
n!
R2load R3load
Rload ðIEQ1 IEQ2 Þ2 VT2
ðIEQ1 IEQ2 Þ3 VT3
¼ 1 þ ðIEQ1 IEQ2 Þ þ þ þ
VT 2 6
R2load
Rload Rload ðIEQ1 IEQ2 Þ2 VT2
expfðIEQ1 IEQ2 Þ g 1 þ ðIEQ1 IEQ2 Þ þ
VT VT 2
C1 ¼ C1 ðar ; af ; Ise ; ICBQ2 0 Þ; C2 ¼ C2 ðar ; af ; Isc ; ICBQ2 0 Þ; X ¼ Xðar ; bf Þ
R2
Rload ðIEQ1 IEQ2 Þ2 Vload
2
1 þ ðIEQ1 IEQ2 Þ þ T
VT 2
Xðar ; bf Þ IEQ2 þ C1 ðar ; af ; Ise ; ICBQ2 0 Þ

C2 ðar ; af ; Isc ; ICBQ2 0 Þ
R2
ðIEQ1 IEQ2 Þ2 load
V2
ðÞ 1 þ ðIEQ1 IEQ2 Þ Rload
VT þ 2
T
CX2 IEQ2 þ C1
C2
b IEQ1
f þ 1Þ
n2 ðIEQ1 ; IEQ2 Þ ¼ f b IEQ1
g
½ðbf þ ICBQ1 0 af IEQ1 þ ðaf ar 1Þ Isc
f þ 1Þ
b IEQ2
f þ 1Þ
f b IEQ2
g;
½ðbf þ ICBQ2 0 af IEQ2 þ ðaf ar 1Þ Isc
f þ 1Þ
n2 ðIEQ1 ; IEQ2 Þ ¼ e½ VT
Vcc
bf
ðar ðb þ 1Þ 1Þ IEQ1 þ ar ICBQ1 0 þ ðar af 1Þ Ise
f bf
f
g
ððb þ 1Þ af Þ IEQ1 þ ICBQ1 0 þ ðaf ar 1Þ Isc
f
bf
ðar ðb þ 1Þ 1Þ IEQ2 þ ar ICBQ2 0 þ ðar af 1Þ Ise
¼ e½ VT
Vcc
f f
bf
g
ððb þ 1Þ af Þ IEQ2 þ ICBQ2 0 þ ðaf ar 1Þ Isc
f
Assumptions:
bf bf
ððb þ 1Þ af Þ ! e; X ¼ ar ðb þ 1Þ 1; C3 ¼ ar ICBQ1 0 þ ðar af 1Þ Ise
f f
C4 ¼ ICBQ1 0 þ ðaf ar 1Þ Isc ; C1 ¼ ar ICBQ2 0 þ ðar af 1Þ Ise ;

C2 ¼ ICBQ2 0 þ ðaf ar 1Þ Isc
ðÞ ðXIEQ1 þ C3 Þ ðXIEQ2 þ C1 Þ ¼ e½VVccT

C4 C2
We can summary our intermediate results:

R2
ðIEQ1 IEQ2 Þ2 load
V2
ðÞ 1 þ ðIEQ1 IEQ2 Þ RVload þ 2
T
CX2 IEQ2 þ C1
C2
T
ðÞ XIEQ1 þ C3 XIEQ2 þ C1
ð C4 Þ ð C2 Þ ¼ e½ VT
Vcc
1 C4 C2 e½ VT
Vcc
IEQ1 ¼ ð C3 Þ
X X IEQ2 þ C1
1 C4 C2 e½ VT
Vcc
Rload
1þð ð C3 Þ IEQ2 Þ
X X IEQ2 þ C1 VT
V
½ cc
2 e T
R2load
ðX1 ðCXI4 C
V
EQ þ C1
C3 Þ IEQ2 Þ2 VT2 X C1
þ 2
IEQ2 þ
2 C2 C2
C4 C2 Rload e½ VT C4 C2 e½ VT
Vcc Vcc
Rload C3 Rload
IEQ2 þ1 þ ð½ 2 IEQ2 2
X VT IEQ2 þ X C1 VT
2 VT X VT X IEQ2 þ X C1
C4 C2 e½ VT
Vcc
C3 C3 R2 X C1
þ ½ 2 2 ½ 2 IEQ2 Þ load2 IEQ2 þ
X X X IEQ2 þ X C1 2 VT C 2 C2
C4 C2 Rload e½ VT
Vcc
Rload C3 Rload
IEQ2 þ1
X2 VT IEQ2 þ X C1 VT VT X VT
C24 C22 e2½ VT C4 C2 e½ VT
Vcc Vcc
þð þ IEQ
2
2 IEQ2
½X2 IEQ2 þ X C1 2 2
½X2 IEQ2 þ X C1
½Vcc
C3 C3 C4 C2 e VT R2 X C1
þ ½ 2 2 ½ 2 IEQ2 Þ load2 IEQ2 þ
X X X IEQ2 þ X C1 2 VT C2 C2
C4 C2 Rload e½ VT
Vcc
Rload C3 Rload
IEQ2 þ1
2 V T X VT
C24 C22 e2½ VT 2V C4 C2 e½ VT 2V
Vcc
R2
load load
Vcc
R2
2 R2load 2
þ T
þ IEQ
2
2 2 T
IEQ2
½X2 IEQ2 þ X C1 2 2
2 VT2 ½X IEQ2 þ X C1
½Vcc R2
C3 C4 C2 e T 2VT2
load
C3 X C1
V
R2 R2
þ ½ 2 load2 2 ½ 2 IEQ2 load2 IEQ2 þ
X 2 VT X X IEQ2 þ X C1 2 VT C2 C2
C4 C2 Rload e½ VT
Vcc
C3 R2load Rload X R2
þ IEQ2 ½2 þ IEQ
2
load2
2 X 2 VT 2 VT C2 2
2 VT
C24 C22 e2½ VT 2V ½V
Vcc
R2 Vcc
R2
C3 C4 C2 e T 2VT2
load load
2
þ 2 2 ð T
Þ
½X IEQ2 þ X C1 2 X X2 IEQ2 þ X C1
½Vcc R2
C4 C2 e VT 2V
load
2 C1 C3 R2 C3 Rload
2 T
IEQ2 ½ 2 load2 þ 1
½X IEQ2 þ X C1
2 C2 X 2 VT X VT
The above equation can be solve numerically and we get some options for IEQ2
values. We ignore complex and negative values [33].
C1 C3 R2 C3 Rload
w1 ðIEQ2 Þ ½ 2 load2 þ 1;
C2 X 2 VT X VT
IEQ2 ¼ fn ðC1 ; . . .; C4 ; X; Rload ; VT ; . . .Þ; n ¼ 1; 2; . . .
½Vcc
1 C4 C2 e VT
IEQ1 ¼ ð C3 Þ;
X X fn ðC1 ; . . .; C4 ; X; Rload ; VT ; . . .Þ þ C1
IEQ1 ¼ gn ðC1 ; . . .; C4 ; X; Rload ; VT ; . . .Þ
Summary: We get some options for IEQ1 ; IEQ2 values, and ignore negative and
complex results.
IEQ1 ¼ gn ðC1 ; . . .; C4 ; X; Rload ; VT ; . . .Þ;
IEQ2 ¼ fn ðC1 ; . . .; C4 ; X; Rload ; VT ; . . .Þ; n ¼ 1; 2; . . .
Back to our previous differential equation:
dIR1 dIR2 1 X 2 X2
ð R1 þ R2 Þ ¼ ð IRk IBQk 2 I0
dt dt C k¼1 k¼1
1
fexp½ ðVcc IR1 R1 IR2 R2 Þ 1gÞ
2
We consider R1 R2 and R ¼ R2 ; R ¼ R1 then we can write the above circuit

differential equation:
dIR1 dIR2 1 X 2 X2
1 X2
ð þ ÞR¼ ð IRk IBQk 2 I0 fexp½ ðVcc R IRk Þ
dt dt C k¼1 k¼1
2 k¼1
1gÞ
We define new variable
X
2
dX X 2
dIRk
X¼ IRk ; ¼
k¼1
dt k¼1
dt
IEQ1 gn ðC1 ; . . .; C4 ; X; Rload ; VT ; . . .Þ

IBQ1 ¼ ICBQ1 0 ¼ ICBQ1 0
IEQ2 fn ðC1 ; . . .; C4 ; X; Rload ; VT ; . . .Þ

IBQ2 ¼ ICBQ2 0 ¼ ICBQ2 0
X
2
gn ðC1 ; . . .; C4 ; X; Rload ; VT ; . . .Þ fn ðC1 ; . . .; C4 ; X; Rload ; VT ; . . .Þ
IBQk ¼ þ
k¼1
X2
ICBQk 0
k¼1
gn ¼ gn ðC1 ; . . .; C4 ; X; Rload ; VT ; . . .Þ; fn ¼ fn ðC1 ; . . .; C4 ; X; Rload ; VT ; . . .Þ
X
2
gn þ f n X
2
IBQk ¼ ICBQk 0 ; n ¼ 1; 2; . . .
k¼1
ðbf þ 1Þ k¼1
X
2
dIR 1 X 2
gn þ f n X2
R k
¼ ð IR k þ ICBQk 0 2 I0
k¼1
dt C k¼1 ðbf þ 1Þ k¼1
1 X2
fexp½ ðVcc R IRk Þ 1gÞ
2 k¼1
dX 1 gn þ f n X2
1
¼ ðX þ ICBQk 0 2 I0 fexp½ ðVcc R XÞ 1gÞ
dt CR ðbf þ 1Þ k¼1
2
The first stage is to find our circuit fixed point: dX

dt ¼0
gn þ fn X2
1
X ðjÞ þ ICBQk 0 2 I0 fexp½ ðVcc R X ðjÞ Þ 1g ¼ 0;
ðbf þ 1Þ k¼1
2
j ¼ 0; 1; 2; . . .
1 gn þ fn X
2
X ðjÞ 2 I0 fexp½ ðVcc R X ðjÞ Þ 1g ¼ ICBQk 0
2 ðbf þ 1Þ k¼1
1 gn þ fn X
2
X ðjÞ 2 I0 fexp½ ðVcc R X ðjÞ Þ 1g ¼ ICBQk 0
2 ðbf þ 1Þ k¼1
X1 1 ðjÞ n
1 2n ðVcc R X Þ
exp½ ðVcc R X ðjÞ Þ ¼
2 n¼0
n!
1 1
¼ 1þ ðVcc R X ðjÞ Þ þ ðVcc R X ðjÞ Þ2 þ
2 8
1 1 1
exp½ ðVcc R X ðjÞ Þ 1 þ ðVcc R X ðjÞ Þ þ ðVcc R X ðjÞ Þ2
2 2 8
1 gn þ f n X
2
X ðjÞ I0 fðVcc R X ðjÞ Þ þ ðVcc R X ðjÞ Þ2 g ¼ ICBQk 0
4 ðbf þ 1Þ k¼1
1
ð1 þ I0 RÞ X ðjÞ I0 Vcc I0 ½Vcc
2
2 Vcc R X ðjÞ þ R2 ½X ðjÞ 2
4
gn þ fn X2
¼ ICBQk 0
ðbf þ 1Þ k¼1
1 1
I0 R2 ½X ðjÞ 2 ð1 þ I0 R þ I0 Vcc RÞ X ðjÞ
4 2
1 gn þ f n X
2
þ ½I0 Vcc þ I0 Vcc 2
þ ICBQk 0 ¼ 0
4 ðbf þ 1Þ k¼1
vffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
u
u 1 2
u ð1 þ I0 R þ I0 Vcc RÞ
u 2
ð1 þ I0 R þ 12 I0 Vcc RÞ u
u X
t I0 R2 ½I0 Vcc þ 1 I0 V 2 þ gn þ fn
2
ICBQk 0
4 cc
ðbf þ 1Þ k¼1
X ðjÞ ¼ 1
2 I0 R2
We get two groups of fixed points for our circuit: gn ; fn 2 R þ ; n ¼ 1; 2; . . .

u
u 1 2
u 2
u
ð1 þ I0 R þ 2 I0 Vcc RÞ þ u
1
X
2
ICBQk 0
4 cc
ðbf þ 1Þ k¼1
ðjÞ
Xgroup ¼
1 1
2 I0 R2
u
u 1 2
u 2
ð1 þ I0 R þ 12 I0 Vcc RÞ u
u X
2
ICBQk 0
4 cc
ðbf þ 1Þ k¼1
ðjÞ
Xgroup ¼
2 1
2 I0 R2
We ignore in our analysis negative and complex fixed points values. We get a set
of one dimension systems dXdt ¼ nn ðXÞ; n ¼ 1; 2; . . .
1 g n þ fn X2
1
nn ðXÞ ¼ ðX þ ICBQk 0 2 I0 fexp½ ðVcc R XÞ 1gÞ
CR ðbf þ 1Þ k¼1
2
dX 1 gn þ f n X2
1
¼ ðX þ ICBQk 0 2 I0 fexp½ ðVcc R XÞ 1gÞ
dt CR ðbf þ 1Þ k¼1
2
We can draw the graphs nn ðXÞ; n ¼ 1; 2; . . . and then use it to sketch the vector
fields on the real line. A fluid is flowing along the real line with a local velocities
nn ðXÞ. This imaginary fluid is called the phase fluid of our class AB push-pull
amplifier system, and the real line is the phase space. The flow is to the right where
nn ðXÞ [ 0 and to the left where nn ðXÞ\0. To find the solutions to dX dt ¼ nn ðXÞ; n ¼
1; 2; . . . starting from an arbitrary initial condition X0 As time goes, the phase point
moves along the X-axis according to some functions nn ðXÞ; n ¼ 1; 2; . . . This
function is called the trajectory based at X0 , and it represents the solutions of the
differential equation starting from the initial conditions X0 . A picture which shows
all the qualitatively different trajectories of our class AB push-pull amplifier system,
is called a phase portrait. The appearance of the phase portrait is controlled by the
ðjÞ ðjÞ ðjÞ
fixed points Xgroup1 or Xgroup2 ; j ¼ 0; 1; . . ., defined by nn ðXgroup1 Þ ¼ 0;
ðjÞ
nn ðXgroup2 Þ ¼ 0; n ¼ 1; 2; . . .; they correspond to stagnation points of the flow. Our
system fixed points represent equilibrium solutions (steady state, constant, rest
ðjÞ ðjÞ
solutions), since if X ¼ Xgroup1 or X ¼ Xgroup2 ; j ¼ 0; 1; . . . initially, then XðtÞ ¼
ðjÞ ðjÞ
Xgroup1 or XðtÞ ¼ Xgroup2 ; j ¼ 0; 1; . . . for all time. An class AB push-pull amplifier
system equilibrium is defined to be stable if all sufficiently small disturbances away
from it damp out in time. Stable system equilibria are represented geometrically by
stable fixed points. Conversely, unstable equilibria, in which disturbances grow in

time, are represented by unstable fixed points [2–4].
5.2 Class C Power Amplifier (PA) with Parallel Resonance

Circuit Stability Analysis Under Parameters Variation
The Class C Amplifier design has the greatest efficiency but the poorest linearity of
the classes of amplifiers. The class C amplifier is heavily biased so that the output
current is zero for more than one half of an input sinusoidal signal cycle with the
transistor idling at its cut-off point. Due to its heavy audio distortion, class C
amplifiers are commonly used in high frequency sine wave oscillators and certain
types of radio frequency amplifiers. The class C amplifier conduction angle is
slightly less than 180°. The most common application of the Class C amplifier is the
RF (radio frequency) circuits like RF oscillator, RF amplifier etc. where there are
additional tuned circuits for retrieving the original input signal from the pulsed
output of the Class C amplifier and so the distortion caused by the amplifier has
little effect on the final output. Biasing resistor Rb pulls the base of Q1 further
downwards and the Q-point will be set some way below the cut-off point in the DC
load line. As a result the transistor will start conducting only after the input signal
amplitude has risen above the base emitter voltage (Vbe * 0.7 V) plus the
downward bias voltage caused by Rb. That is the reason why the major portion of
the input signal is absent in the output signal. Inductor L1 and capacitor C1 forms a
tank circuit which aids in the extraction of the required signal from the pulsed
output of the transistor. Class C operation means that the collector current flows for
less than 180° of the ac cycle. This implies that the collector current of a class C
amplifier is highly non-sinusoidal because current flows in pulses. To avoid dis-
tortion, class C amplifier makes use of a resonant tank circuit. This results in a
sinusoidal output voltage. Actual job of the active element (transistor Q1) here is to
produce a series of current pulses according to the input and make it flow through
the resonant circuit. Values of L1 and C1 are so selected that the resonant circuit
oscillates in the frequency of the input signal. Since the resonant circuit oscillates in
one frequency (generally the carrier frequency) all other frequencies are attenuated
and the required frequency can be squeezed out using a suitably tuned load.
Harmonics or noise present in the output signal can be eliminated using additional
filters. A coupling transformer can be used for transferring the power to the
load. The Class C amplifier has high efficiency and it is excellent in RF applications
[24–26] (Fig. 5.3).
5.2 Class C Power Amplifier (PA) with Parallel Resonance … 529
Fig. 5.3 Class C power amplifier
Inductor L1 and capacitor C1 forms a tank circuit which aids in the extraction of
the required signal from the pulsed output of the transistor. Class C operation means
that the collector current flows for less than 180° of the ac cycle. This implies that
the collector current of a class C amplifier is highly non-sinusoidal because current
flows in pulses. To avoid distortion, class C amplifier makes use of a resonant tank
circuit. This results in a sinusoidal output voltage. Actual job of the active element
(transistor Q1) here is to produce a series of current pulses according to the input
and make it flow through the resonant circuit. Values of L1 and C1 are so selected
that the resonant circuit oscillates in the frequency of the input signal. Since the
resonant circuit oscillates in one frequency (generally the carrier frequency) all
other frequencies are attenuated and the required frequency can be squeezed out
using a suitably tuned load. Harmonics or noise present in the output signal can be
eliminated using additional filters. A coupling transformer can be used for trans-
ferring the power to the load. The Class C amplifier has high efficiency and it is
excellent in RF applications.
Since the input RF signal is a large signal we use in our analysis the Ebers-Moll
BJT model for Q1. The Ebers-Moll BJT model is a good large signal, steady state
model of the transistor and allows the state of conduction of the device to be easily
determined for different modes of operation of the device. The different modes of
operation are determined by the manner in which the junctions are biased [91–93]
(Fig. 5.4).
Fig. 5.4 Class C power amplifier equivalent circuit
Vs ðtÞ VA1 d
ICQ1 þ IBQ1 ¼ IEQ1 ; IRs ¼ ; ICin ¼ Cin ðVA1 VA2 Þ; IRs ¼ ICin ; ICin
Rs dt
¼ IBQ1 þ IRb
Collector emitter voltage expression for BJT NPN:

VCENPN VT lnf g;
VCEQ1 VT lnf g
VA dIL
IRb ¼ 2 ; IL1 þ IC1 ¼ ICQ1 þ ICout ; ICout ¼ IRload ; Vcc VA3 ¼ L1 1 ;
Rb dt
d
IC1 ¼ C1 ðVcc VA3 Þ
dt
d VA4
VA2 ¼ VBEQ1 ; VCEQ1 ¼ VCout þ VRload ; ICout ¼ Cout ðVA3 VA4 Þ; IRload ¼
dt Rload
Vs ðtÞ VA1 d
IR s ¼ ) VA1 ¼ Vs ðtÞ IRs Rs ; ICin ¼ Cin ðVA1 VA2 Þ
Rs dt
Z
1
) VA1 VA2 ¼ ICin dt
Cin
Z Z
1 1
VA1 ¼ VA2 þ ICin dt; Vs ðtÞ IRs Rs ¼ VA2 þ ICin dt;
Cin Cin
Z
1
VA2 ¼ Vs ðtÞ IRs Rs ICin dt
Cin
VA2
IRb ¼ ) VA2 ¼ IRb Rb ; IRb Rb
Rb
Z
1
¼ Vs ðtÞ IRs Rs ICin dt; VCEQ1 ¼ VA3
Cin
Z
d 1
fIRb Rb ¼ Vs ðtÞ IRs Rs ICin dtg;
dt Cin
dIRb dVs ðtÞ dIRs 1
Rb ¼ Rs ICin
dt dt dt Cin
dIL d d 2 IL
Vcc VA3 ¼ L1 1 ; IC1 ¼ C1 ðVcc VA3 Þ; IC1 ¼ C1 L1 2 1
dt dt dt
Z
d 1
ICout ¼ Cout ðVA3 VA4 Þ ) VA3 VA4 ¼ ICout dt; VA3 IRload Rload
dt
Z Cout
1
¼ ICout dt
Cout

VCEQ1 ¼ VA3 ; VT lnf g
½ICQ1 Zaf IEQ1 þ ðaf ar 1Þ Isc
1
¼ IRload Rload þ ICout dt
Cout
d ½ar ICQ1 IEQ1 þ ðar af 1Þ Ise

ðlnf gÞ
dt ½ICQ1 af IEQ1 þ ðaf ar 1Þ Isc
dICQ1 dIEQ1
ðar Þ ½ICQ1 af IEQ1 þ ðaf ar 1Þ Isc
dt dt
dICQ1 dIEQ1
ð af Þ ½ar ICQ1 IEQ1 þ ðar af 1Þ Ise
¼ dt dt
dICQ1 dIEQ1
dt dt
dICQ1 dIEQ1
VT f dt dt g
dIRload 1
¼ Rload þ ICout
dt Cout
dICQ1 dIEQ1
dt dt
dICQ1 dIEQ1
ICout ¼ IRload ; VT f dt dt g
dICout 1
¼ Rload þ ICout
dt Cout
d 2 IL1
We can summary our equations: IRs ¼ ICin ; IC1 ¼ C1 L1 dt2
dIRb dVs ðtÞ dICin 1

Rb ¼ Rs ICin ;
dt dt dt Cin
dICin dIRb 1 dVs ðtÞ Rs 1
Y1 ¼ ; ¼ Y1 ICin
dt dt Rb dt Rb Cin Rb
d 2 IL
ICQ1 ¼ IL1 þ IC1 ICout ¼ IL1 þ C1 L1 2 1 ICout ; IBQ1 ¼ ICin IRb
dt
d 2 IL
IEQ1 ¼ IBQ1 þ ICQ1 ¼ ICin IRb þ IL1 þ C1 L1 2 1 ICout
dt
dICQ1 dIL1 d 3 IL dICout

¼ þ C1 L1 3 1 ;
dt dt dt dt
dIEQ1 dICin dIRb dIL1 d 3 IL dICout
¼ þ þ C1 L1 3 1
dt dt dt dt dt dt
d 2 I L1
ICQ1 af IEQ1 ¼ ð1 af Þ IL1 þ ð1 af Þ C1 L1 þ ðaf 1Þ ICout af
dt2
ICin þ af IRb
d 2 I L1
ar ICQ1 IEQ1 ¼ ðar 1Þ IL1 þ ðar 1Þ C1 L1 þ ð1 ar Þ ICout
dt2
ICin þ IRb
dICQ1 dtIEQ1 dIL d 3 IL dICout

ar ¼ ðar 1Þ 1 þ ðar 1Þ C1 L1 3 1 þ ð1 ar Þ
dt dt dt dt dt
dICin dIRb
þ
dt dt
dICQ1 dIEQ1 dIL d 3 IL dICout

af ¼ ð1 af Þ 1 þ ð1 af Þ C1 L1 3 1 þ ðaf 1Þ
dt dt dt dt dt
dICin dIR
af þ af b
dt dt
We define new variables in our system:
dICin dICout dIL dY3 d 2 IL1

Y1 ¼ ; Y2 ¼ ; Y3 ¼ 1 ; Y4 ¼ ¼ 2
dt dt dt dt dt
ICQ1 af IEQ1 ¼ ð1 af Þ IL1 þ ð1 af Þ C1 L1 Y4 þ ðaf 1Þ ICout af ICin þ af IRb
ar ICQ1 IEQ1 ¼ ðar 1Þ IL1 þ ðar 1Þ C1 L1 Y4 þ ð1 ar Þ ICout ICin þ IRb
dICQ1 dtIEQ1 dY4

ar ¼ ðar 1Þ Y3 þ ðar 1Þ C1 L1 þ ð1 ar Þ Y2 Y1
dt dt dt
1 dVs ðtÞ Rs 1
þ Y1 ICin
Rb dt Rb Cin Rb
dICQ1 dIEQ1 dY4

af ¼ ð1 af Þ Y3 þ ð1 af Þ C1 L1 þ ðaf 1Þ Y2
dt dt dt
1 dVs ðtÞ Rs 1
af Y1 þ af ð Y1 ICin Þ
Rb dt Rb Cin Rb
We define for simplicity of our analysis four functions:
ICQ1 af IEQ1 ¼ ð1 af Þ IL1 þ ð1 af Þ C1 L1 Y4

þ ðaf 1Þ ICout af ICin þ af IRb
g1 ðIL1 ; Y4 ; ICout ; ICin ; IRb Þ ¼ ð1 af Þ IL1 þ ð1 af Þ C1 L1 Y4
ICQ1 af IEQ1 ¼ g1 ðIL1 ; Y4 ; ICout ; ICin ; IRb Þ; g1 ¼ g1 ðIL1 ; Y4 ; ICout ; ICin ; IRb Þ
ar ICQ1 IEQ1 ¼ ðar 1Þ IL1 þ ðar 1Þ C1 L1 Y4

þ ð1 ar Þ ICout ICin þ IRb
g2 ðIL1 ; Y4 ; ICout ; ICin ; IRb Þ ¼ ðar 1Þ IL1 þ ðar 1Þ C1 L1 Y4
þ ð1 ar Þ ICout ICin þ IRb
ar ICQ1 IEQ1 ¼ g2 ðIL1 ; Y4 ; ICout ; ICin ; IRb Þ; g2 ¼ g2 ðIL1 ; Y4 ; ICout ; ICin ; IRb Þ
dICQ1 dtIEQ1 dY4 dVs ðtÞ

ar ¼ ðar 1Þ C1 L1 þ g3 ðY3 ; Y2 ; Y1 ; ICin ; Þ;
dt dt dt dt
dVs ðtÞ
g3 ¼ g3 ðY3 ; Y2 ; Y1 ; ICin ; Þ
dt
dVs ðtÞ Rs
g3 ðY3 ; Y2 ; Y1 ; ICin ; Þ ¼ ðar 1Þ Y3 þ ð1 ar Þ Y2 Y1 ð1 þ Þ
dt Rb
1 dVs ðtÞ 1
þ ICin
Rb dt Cin Rb
dICQ1 dIEQ1 dY4 dVs ðtÞ

af ¼ ð1 af Þ C1 L1 þ g4 ðY3 ; Y2 ; Y1 ; ICin ; Þ;
dt dt dt dt
dVs ðtÞ
g4 ¼ g4 ðY3 ; Y2 ; Y1 ; ICin ; Þ
dt
dVs ðtÞ Rs
g4 ðY3 ; Y2 ; Y1 ; ICin ; Þ ¼ ð1 af Þ Y3 þ ðaf 1Þ Y2 Y1 af ð1 þ Þ
dt Rb
af dVs ðtÞ af
þ ICin
Rb dt Cin Rb
The main system differential equation can be present as follow:
dY4
½ðar 1Þ C1 L1 þ g3 ½g1 þ ðaf ar 1Þ Isc
dt
dY4
½ð1 af Þ C1 L1 þ g4 ½g2 þ ðar af 1Þ Ise
VT f dt g
½g1 þ ðaf ar 1Þ Isc ½g2 þ ðar af 1Þ Ise
1
¼ Y2 Rload þ ICout
Cout
dY4
½ðar 1Þ C1 L1 þ g3 ½g1 þ ðaf ar 1Þ Isc
dt
dY4
½ð1 af Þ C1 L1 þ g4 ½g2 þ ðar af 1Þ Ise
dt
1 1
¼ ½Y2 Rload þ ICout ½g1 þ ðaf ar 1Þ Isc ½g2 þ ðar af 1Þ Ise
VT Cout
dY4
fðar 1Þ ½g1 þ ðaf ar 1Þ Isc ð1 af Þ ½g2 þ ðar af 1Þ Ise g C1 L1
dt
1 1
¼ ½Y2 Rload þ ICout ½g1 þ ðaf ar 1Þ Isc ½g2 þ ðar af 1Þ Ise
VT Cout
g3 ½g1 þ ðaf ar 1Þ Isc g4 ½g2 þ ðar af 1Þ Ise
1 1
½Y2 Rload þ ICout ½g1 þ ðaf ar 1Þ Isc ½g2 þ ðar af 1Þ Ise
VT Cout
dY4 g3 ½g1 þ ðaf ar 1Þ Isc g4 ½g2 þ ðar af 1Þ Ise
¼
dt fðar 1Þ ½g1 þ ðaf ar 1Þ Isc ð1 af Þ ½g2 þ ðar af 1Þ Ise g C1 L1
We can summary our system differential equations:
dICin dICout dIL1 dY3

¼ Y1 ; ¼ Y2 ; ¼ Y3 ; ¼ Y4
dt dt dt dt
1 1
VT Cout
¼
The first stage is to find our system fixed points: It is done by letting
dICin dICout dIL1 dY3

¼ 0 ) Y1 ¼ 0; ¼ 0 ) Y2 ¼ 0; ¼ 0 ) Y3 ¼ 0; ¼ 0 ) Y4
dt dt dt dt
dY4
¼ 0; ¼0
dt
g1 ðIL1 ; Y4 ; IC out ; IC in ; IRb Þ ¼ ð1 af Þ IL1 þ ð1 af Þ C1 L1 Y4

þ ðaf 1Þ IC out af IC in þ af IRb
g2 ðIL1 ; Y4 ; IC out ; IC in ; IRb Þ ¼ ðar 1Þ IL1 þ ðar 1Þ C1 L1 Y4
þ ð1 ar Þ IC out IC in þ IRb
dVs ðtÞ 1 dVs ðtÞ 1

g3 ¼ g3 ðY3 ; Y2 ; Y1 ; IC in ; Þ¼ I
dt Rb dt Cin Rb Cin
dVs ðtÞ af dVs ðtÞ af
g4 ¼ g4 ðY3 ; Y2 ; Y1 ; IC in ; Þ¼ ICin
dt Rb dt Cin Rb
dY4
¼ 0 ) fðar 1Þ ½g1 þ ðaf ar 1Þ Isc ð1 af Þ
dt
½g2 þ ðar af 1Þ Ise g C1 L1 6¼ 0
C1 L1 6¼ 0 ) ðar 1Þ ½g1 þ ðaf ar 1Þ Isc ð1 af Þ
½g2 þ ðar af 1Þ Ise 6¼ 0
1 1
I ½g1 þ ðaf ar 1Þ Isc ½g2 þ ðar af 1Þ Ise
VT Cout Cout
g3 ½g1 þ ðaf ar 1Þ Isc g4 ½g2 þ ðar af 1Þ Ise ¼ 0
a
Assumption dVdts ðtÞ ! e; g3 ¼ Cin1Rb IC in ; g4 ¼ Cin R f
b
ICin
equilibrium points of the class C power amplifier equivalent circuit consists in
adding to coordinate ½IL1 ; IRb ; IRload ; ICout ; ICin ; Y1 ; Y2 ; Y3 ; Y4 arbitrarily small incre-
ments of exponentially form ½iL1 ; iRb ; iRload ; iCout ; iCin ; y1 ; y2 ; y3 ; y4 ekt and retaining
the first order terms in IL1 ; IRb ; IRload ; ICout ; ICin ; Y1 ; Y2 ; Y3 ; Y4 . The system of
homogenous equations leads to a polynomial characteristic equation in the eigen-
values [4]. The polynomial characteristic equations accept by set of the below
circuit variables, circuit variables derivative and circuit variables second order
derivative with respect to time into class C power amplifier [2–4]. Our class C
power amplifier equivalent circuit fixed values with arbitrarily small increments of
exponential form ½iL1 ; iRb ; iRload ; iCout ; iCin ; y1 ; y2 ; y3 ; y4 ekt are: j = 0 (first fixed
point), j = 1 (second fixed point), j = 2 (third fixed point), etc.,
ðjÞ ðjÞ ðjÞ

Y1 ðtÞ ¼ Y1 þ y1 ekt ; Y2 ðtÞ ¼ Y2 þ y2 ekt ; Y3 ðtÞ ¼ Y3 þ y3 ekt ;
ðjÞ
IL1 ðtÞ ¼ IL1 þ iL1 ekt
ðjÞ ðjÞ ðjÞ

Y4 ðtÞ ¼ Y4 þ y4 ekt ; ICin ðtÞ ¼ ICin þ iCin ekt ; ICout ðtÞ ¼ ICout þ iCout ekt ;
ðjÞ
IRb ðtÞ ¼ IRb þ iRb ekt
dICin dIC dIL dY3

¼ iCin k ekt ; out ¼ iCout k ekt ; 1 ¼ iL1 k ekt ; ¼ y3 k ekt ;
dt dt dt dt
dY4
¼ y4 k ekt
dt
ðjÞ dIRload ðtÞ dIR ðtÞ

IRload ðtÞ ¼ IRload þ iRload ekt ; ¼ iRload k ekt ; b ¼ iRb k ekt ;
dt dt
dVs ðtÞ
!e
dt
ðjÞ Rs 1 ðjÞ
iRb k ekt ¼ ½Y1 þ y1 ekt ½I þ iCin ekt
Rb Cin Rb Cin
ðjÞ Rs 1 ðjÞ Rs 1
iRb k ekt ¼ Y1 ICin y1 ekt iCin ekt
Rb Cin Rb Rb Cin Rb
ðjÞ ðjÞ
At fixed point Y1 RRbs C1in Rb ICin ¼ 0: iRb k y1 RRbs iCin C1in Rb ¼ 0
dICin ðjÞ ðjÞ

¼ Y1 ; iCin k ekt ¼ Y1 þ y1 ekt ; Y1 ¼ 0; iCin k þ y1 ¼ 0
dt
dICout ðjÞ ðjÞ
¼ Y2 ; iCout k ekt ¼ Y2 þ y2 ekt ; Y2 ¼ 0; iCout k þ y2 ¼ 0
dt
dIL1 ðjÞ ðjÞ

¼ Y3 ; iL1 k ekt ¼ Y3 ðtÞ ¼ Y3 þ y3 ekt ; Y3 ¼ 0; iL1 k þ y3 ¼ 0
dt
dY3 ðjÞ ðjÞ
¼ Y4 ; y3 k ekt ¼ Y4 þ y4 ekt ; Y4 ¼ 0; y3 k þ y4 ¼ 0
dt
ðjÞ
g1 ðIL1 ðtÞ; Y4 ðtÞ; ICout ðtÞ; ICin ðtÞ; IRb ðtÞÞ ¼ ð1 af Þ ðIL1 þ iL1 ekt Þ
ðjÞ ðjÞ
þ ð1 af Þ C1 L1 ðY4 þ y4 ekt Þ þ ðaf 1Þ ðICout þ iCout ekt Þ
ðjÞ ðjÞ
af ðICin þ iCin ekt Þ þ af ðIRb þ iRb ekt Þ
ðjÞ ðjÞ
g1 ðIL1 ðtÞ; Y4 ðtÞ; ICout ðtÞ; ICin ðtÞ; IRb ðtÞÞ ¼ ð1 af Þ IL1 þ ð1 af Þ C1 L1 Y4
ðjÞ ðjÞ ðjÞ
þ ðaf 1Þ ICout af ICin þ af IRb þ iL1 ð1 af Þ ekt þ y4 ð1 af Þ C1 L1 ekt
þ iCout ðaf 1Þ ekt iCin af ekt þ iRb af ekt
At fixed point:
ðjÞ ðjÞ
g1 ¼ g1 ðIL1 ; Y4 ; IC out ; IC in ; IRb Þ ¼ ð1 af Þ IL1 þ ð1 af Þ C1 L1 Y4
ðjÞ ðjÞ ðjÞ
g1 ðtÞ ¼ g1 ðIL1 ðtÞ; Y4 ðtÞ; ICout ðtÞ; ICin ðtÞ; IRb ðtÞÞ ¼ g1 ðIL1 ; Y4 ; IC out ; IC in ; IRb Þ þ iL1 ð1 af Þ ekt
þ y4 ð1 af Þ C1 L1 ekt þ iCout ðaf 1Þ ekt iCin af ekt þ iRb af ekt
g1 ðtÞ ¼ g1 þ ½iL1 ð1 af Þ þ y4 ð1 af Þ C1 L1 þ iCout ðaf 1Þ iCin af

þ iRb af ekt
ðjÞ
g2 ðIL1 ðtÞ; Y4 ðtÞ; ICout ðtÞ; ICin ðtÞ; IRb ðtÞÞ ¼ ðar 1Þ IL1 þ iL1 ðar 1Þ ekt
ðjÞ ðjÞ
þ ðar 1Þ C1 L1 Y4 þ y4 ðar 1Þ C1 L1 ekt þ ð1 ar Þ ICout
ðjÞ ðjÞ
þ iCout ð1 ar Þ ekt ICin iCin ekt þ IRb þ iRb ekt
ðjÞ ðjÞ
g2 ðIL1 ðtÞ; Y4 ðtÞ; ICout ðtÞ; ICin ðtÞ; IRb ðtÞÞ ¼ ðar 1Þ IL1 þ ðar 1Þ C1 L1 Y4
ðjÞ ðjÞ ðjÞ
þ ð1 ar Þ ICout ICin þ IRb þ iL1 ðar 1Þ ekt þ y4 ðar 1Þ C1 L1 ekt
þ iCout ð1 ar Þ ekt iCin ekt þ iRb ekt

At fixed point: ðar 1Þ IL1 þ ðar 1Þ C1 L1 Y4 þ ð1 ar Þ ICout ICin þ
ðjÞ
IRb ¼ 0
ðjÞ ðjÞ
g2 ¼ g2 ðIL1 ; Y4 ; IC out ; IC in ; IRb Þ ¼ ðar 1Þ IL1 þ ðar 1Þ C1 L1 Y4
ðjÞ ðjÞ ðjÞ
þ ð1 ar Þ ICout ICin þ IRb ¼ 0
g2 ðtÞ ¼ g2 þ ½iL1 ðar 1Þ þ y4 ðar 1Þ C1 L1 þ iCout ð1 ar Þ iCin þ iRb ekt
dVs ðtÞ ðjÞ

g3 ðY3 ðtÞ; Y2 ðtÞ; Y1 ðtÞ; ICin ðtÞ; Þ ¼ ðar 1Þ ½Y3 þ y3 ekt
dt
ðjÞ
þ ð1 ar Þ ½Y2 þ y2 ekt
ðjÞ Rs 1 dVs ðtÞ
½Y1 þ y1 ekt ð1 þ Þþ
Rb Rb dt
1 ðjÞ
½I þ iCin ekt
Cin Rb Cin
dVs ðtÞ ðjÞ ðjÞ ðjÞ Rs

g3 ðY3 ðtÞ; Y2 ðtÞ; Y1 ðtÞ; ICin ðtÞ; Þ ¼ ðar 1Þ Y3 þ ð1 ar Þ Y2 Y1 ð1 þ Þ
dt Rb
1 ðjÞ
I þ y3 ðar 1Þ ekt þ y2 ð1 ar Þ ekt
Cin Rb Cin
Rs 1 1 dVs ðtÞ
y1 ð1 þ Þ ekt iCin ekt þ
Rb Cin Rb Rb dt
dV ðtÞ ðjÞ ðjÞ
g3 ¼ g3 ðY3 ; Y2 ; Y1 ; IC in ;
s
Þ ¼ ðar 1Þ Y3 þ ð1 ar Þ Y2
dt
ðjÞ Rs 1 ðjÞ
Y1 ð1 þ Þ I
Rb Cin Rb Cin
dVs ðtÞ
g3 ðY3 ðtÞ; Y2 ðtÞ; Y1 ðtÞ; ICin ðtÞ; Þ ¼ g3 þ y3 ðar 1Þ ekt þ y2 ð1 ar Þ ekt
dt
Rs 1 1 dVs ðtÞ dVs ðtÞ
y1 ð1 þ Þ ekt iCin ekt þ ; !e
Rb Cin Rb Rb dt dt
dVs ðtÞ Rs 1
g3 ð ! eÞ ¼ g3 þ ½y3 ðar 1Þ þ y2 ð1 ar Þ y1 ð1 þ Þ iCin ekt
dt Rb Cin Rb
dVs ðtÞ ðjÞ ðjÞ
g4 ðY3 ðtÞ; Y2 ðtÞ; Y1 ðtÞ; ICin ; Þ ¼ ð1 af Þ ½Y3 þ y3 ekt þ ðaf 1Þ ½Y2 þ y2 ekt
dt
ðjÞ Rs af dVs ðtÞ af ðjÞ
½Y1 þ y1 ekt af ð1 þ Þ þ ½I þ iCin ekt
Rb Rb dt Cin Rb Cin
dVs ðtÞ ðjÞ

g4 ðY3 ðtÞ; Y2 ðtÞ; Y1 ðtÞ; ICin ; Þ ¼ ð1 af Þ Y3 þ y3 ð1 af Þ ekt
dt
ðjÞ
þ ðaf 1Þ Y2 þ y2 ðaf 1Þ ekt
ðjÞ Rs Rs
Y1 af ð1 þ Þ y1 af ð1 þ Þ ekt
Rb Rb
af dVs ðtÞ af ðjÞ af
þ ICin iCin ekt
Rb dt Cin Rb Cin Rb
dVs ðtÞ ðjÞ ðjÞ

g4 ðY3 ðtÞ; Y2 ðtÞ; Y1 ðtÞ; ICin ; Þ ¼ ð1 af Þ Y3 þ ðaf 1Þ Y2
dt
ðjÞ Rs af ðjÞ
Y1 af ð1 þ Þ ICin þ y3 ð1 af Þ ekt þ y2 ðaf 1Þ ekt
Rb Cin Rb
Rs kt af af dVs ðtÞ
y1 af ð1 þ Þ e iCin ekt þ
Rb Cin Rb Rb dt
dV ðtÞ ðjÞ ðjÞ
g4 ¼ g4 ðY3 ; Y2 ; Y1 ; IC in ;
s
Þ ¼ ð1 af Þ Y3 þ ðaf 1Þ Y2
dt
ðjÞ Rs af ðjÞ
Y1 af ð1 þ Þ I
Rb Cin Rb Cin
dVs ðtÞ
g4 ðY3 ðtÞ; Y2 ðtÞ; Y1 ðtÞ; ICin ðtÞ; Þ ¼ g4 þ y3 ð1 af Þ ekt þ y2 ðaf 1Þ ekt
dt
Rs af af dVs ðtÞ dVs ðtÞ
y1 af ð1 þ Þ ekt iCin ekt þ ; !e
Rb Cin Rb Rb dt dt
dVs ðtÞ
g4 ð ! eÞ
dt
Rs af
¼ g4 þ ½y3 ð1 af Þ þ y2 ðaf 1Þ y1 af ð1 þ Þ iCin
Rb Cin Rb
ekt
Back to our last differential equation:
1 1
VT Cout
¼
At fixed point: dY4

dt ¼0
1 ðjÞ 1 ðjÞ
½Y2 Rload þ I ½g1 þ ðaf ar 1Þ Isc ½g2 þ ðar af 1Þ Ise
VT Cout Cout

¼0
We define function: dY4

dt ¼ tðY2 ; ICout ; . . .Þ
1 1
VT Cout
tðY2 ; ICout ; . . .Þ ¼
First we extract the above function tðY2 ; ICout ; . . .Þ denominator.

¼ fðar 1Þ g1 ð1 af Þ g2 þ ðar 1Þ ðaf ar 1Þ Isc
ð1 af Þ ðar af 1Þ Ise g C1 L1
We define global parameter:
C1 ¼ C1 ðar ; af ; Isc ; Ise Þ ¼ ðar 1Þ ðaf ar 1Þ Isc ð1 af Þ ðar af 1Þ Ise

¼ fðar 1Þ g1 ð1 af Þ g2 þ C1 ðar ; af ; Isc ; Ise Þg C1 L1

¼ fðar 1Þ ðg1 þ ½iL1 ð1 af Þ þ y4 ð1 af Þ C1 L1 þ iCout ðaf 1Þ
iCin af þ iRb af ekt Þ ð1 af Þ ðg2 þ ½iL1 ðar 1Þ
þ y4 ðar 1Þ C1 L1 þ iCout ð1 ar Þ iCin þ iRb ekt Þ
þ C1 ðar ; af ; Isc ; Ise Þg C1 L1

¼ fðar 1Þ g1 þ ½iL1 ðar 1Þ ð1 af Þ þ y4 ðar 1Þ ð1 af Þ C1 L1
þ iCout ðar 1Þ ðaf 1Þ iCin ðar 1Þ af þ iRb ðar 1Þ af ekt
ð1 af Þ g2 ½iL1 ð1 af Þ ðar 1Þ þ y4 ð1 af Þ ðar 1Þ
C1 L1 þ iCout ð1 af Þ ð1 ar Þ ð1 af Þ iCin þ ð1 af Þ iRb ekt
þ C1 ðar ; af ; Isc ; Ise Þg C1 L1

¼ fðar 1Þ g1 ð1 af Þ g2 þ C1 ðar ; af ; Isc ; Ise Þ þ ½iL1 ðar 1Þ ð1 af Þ
þ y4 ðar 1Þ ð1 af Þ C1 L1 þ iCout ðar 1Þ ðaf 1Þ
iCin ðar 1Þ af þ iRb ðar 1Þ af ekt
½iL1 ð1 af Þ ðar 1Þ þ y4 ð1 af Þ ðar 1Þ C1 L1
þ iCout ð1 af Þ ð1 ar Þ
ð1 af Þ iCin þ ð1 af Þ iRb ekt g C1 L1

¼ fðar 1Þ g1 ð1 af Þ g2 þ C1 ðar ; af ; Isc ; Ise Þ
þ ½ð1 af Þ ðar 1Þ af iCin ekt þ ½ðar 1Þ af ð1 af Þ iRb ekt g C1 L1
We define the following new system parameters for simplicity.
C2 ðg1 ; g2 ; C1 ; ar ; af Þ ¼ ðar 1Þ g1 ð1 af Þ g2 þ C1 ðar ; af ; Isc ; Ise Þ

C3 ðaf ; ar Þ ¼ ð1 af Þ ðar 1Þ af ; C3 ðaf ; ar Þ ¼ ðar 1Þ af ð1 af Þ

¼ fC2 ðg1 ; g2 ; C1 ; ar ; af Þ þ C3 ðaf ; ar Þ ðiCin iRb Þ ekt g C1 L1
Second we extract the above function tðY2 ; ICout ; . . .Þ numerator.

We define for simplicity three functions which there summation gives the
function tðY2 ; ICout ; . . .Þ numerator.
1 1
U1 ðY2 ; g1 ; g2 ; af ; ar ; . . .Þ ¼ ½Y2 Rload þ ICout ½g1 þ ðaf ar 1Þ Isc
VT Cout
½g2 þ ðar af 1Þ Ise
U2 ðg1 ; g3 ; af ; ar ; Isc Þ ¼ g3 ½g1 þ ðaf ar 1Þ Isc ; U3 ðg2 ; g4 ; af ; ar ; Ise Þ
¼ g4 ½g2 þ ðar af 1Þ Ise
P3
Function tðY2 ; ICout ; . . .Þ numerator is define as Uk .U1 ðY2 ; g1 ; g2 ;
P k¼1
af ; ar ; . . .Þ ¼ V1T 8j¼1 U1j .
U1j¼1 ¼ g1 g2 Y2 Rload ¼ fg1 þ ½iL1 ð1 af Þ þ y4 ð1 af Þ C1 L1

þ iCout ðaf 1Þ iCin af þ iRb af ekt g
fg2 þ ½iL1 ðar 1Þ þ y4 ðar 1Þ C1 L1
ðjÞ
þ iCout ð1 ar Þ iCin þ iRb ekt g fY2 þ y2 ekt g Rload
U1j¼1 ¼ fg1 g2 þ g1 ½iL1 ðar 1Þ þ y4 ðar 1Þ C1 L1 þ iCout ð1 ar Þ iCin þ iRb ekt
þ g2 ½iL1 ð1 af Þ þ y4 ð1 af Þ C1 L1 þ iCout ðaf 1Þ iCin af þ iRb af ekt
þ ½iL1 ð1 af Þ þ y4 ð1 af Þ C1 L1 þ iCout ðaf 1Þ iCin af þ iRb af ekt
ðjÞ
½iL1 ðar 1Þ þ y4 ðar 1Þ C1 L1 þ iCout ð1 ar Þ iCin þ iRb ekt g fY2 þ y2 ekt g Rload
dICout ðjÞ
Since iL1 iL1 ! e; y4 iL1 ! e; iCout iL1 ! e. . .; dt ¼ 0 ) Y2 ¼ 0 then
U1j¼1 ¼ fg1 g2 þ g1 ½iL1 ðar 1Þ þ y4 ðar 1Þ C1 L1

þ iCout ð1 ar Þ iCin þ iRb ekt þ g2 ½iL1 ð1 af Þ
þ y4 ð1 af Þ C1 L1 þ iCout ðaf 1Þ iCin af þ iRb af ekt g y2 ekt Rload
U1j¼1 ¼ fg1 g2 y2 ekt þ g1 ½iL1 ðar 1Þ þ y4 ðar 1Þ C1 L1

þ iCout ð1 ar Þ iCin þ iRb y2 ekt ekt þ g2 ½iL1 ð1 af Þ
þ y4 ð1 af Þ C1 L1 þ iCout ðaf 1Þ iCin af þ iRb af y2 ekt ekt g Rload
Since iL1 y2 ! e; y4 y2 ! e; . . .; iL1 y2 ! e; y4 y2 ! e. . .; U1j¼1 ¼ g1 g2

Rload y2 ekt
U1j¼2 ¼ Y2 g1 ðar af 1Þ Ise Rload

ðjÞ
¼ ðY2 þ y2 ekt Þ fg1 þ ½iL1 ð1 af Þ
þ y4 ð1 af Þ C1 L1 þ iCout ðaf 1Þ
iCin af þ iRb af ekt g ðar af 1Þ Ise Rload
dICout ðjÞ
Since dt ¼ 0 ) Y2 ¼ 0; y2 iL1 ! e; y2 y4 ! e. . . then
U1j¼2 ¼ y2 ekt g1 ðar af 1Þ Ise Rload

¼ g1 ðar af 1Þ Ise Rload y2 ekt
U1j¼3 ¼ Y2 g2 ðar af 1Þ Isc Rload

ðjÞ
¼ ðY2 þ y2 ekt Þ fg2 þ ½iL1 ðar 1Þ þ y4 ðar 1Þ
C1 L1 þ iCout ð1 ar Þ iCin þ iRb ekt g
ðar af 1Þ Isc Rload
dICout ðjÞ
Since dt ¼ 0 ) Y2 ¼ 0; y2 iL1 ! e; y2 y4 ! e. . . then
U1j¼3 ¼ Y2 g2 ðar af 1Þ Isc Rload ¼ g2 ðar af 1Þ Isc Rload y2 ekt
dICout ðjÞ
U1j¼4 ¼ Y2 Rload ðar af 1Þ2 Isc Ise ; ¼ 0 ) Y2 ¼ 0;
dt
U1j¼4 ¼ Rload ðar af 1Þ2 Isc Ise y2 ekt
1 ðjÞ
U1j¼5 ¼ ICout g1 g2 ¼ ðICout þ iCout ekt Þ fg1 þ ½iL1 ð1 af Þ þ y4 ð1 af Þ C1 L1
Cout
þ iCout ðaf 1Þ iCin af þ iRb af ekt g fg2 þ ½iL1 ðar 1Þ þ y4 ðar 1Þ C1 L1
1
þ iCout ð1 ar Þ iCin þ iRb ekt g
Cout
ðjÞ
U1j¼5 ¼ ðICout þ iCout ekt Þ fg1 g2 þ g1 ½iL1 ðar 1Þ þ y4 ðar 1Þ C1 L1
þ iCout ð1 ar Þ iCin þ iRb ekt þ g2 ½iL1 ð1 af Þ þ y4 ð1 af Þ C1 L1
þ iCout ðaf 1Þ iCin af þ iRb af ekt þ ½iL1 ð1 af Þ þ y4 ð1 af Þ C1 L1
þ iCout ðaf 1Þ iCin af þ iRb af ekt ½iL1 ðar 1Þ þ y4 ðar 1Þ C1 L1
1
Cout
dICout ðjÞ
Since iL1 iL1 ! e; y4 iL1 ! e; iCout iL1 ! e. . .; dt ¼ 0 ) Y2 ¼ 0 then
ðjÞ
U1j¼5 ¼ ðICout þ iCout ekt Þ fg1 g2 þ g1 ½iL1 ðar 1Þ þ y4 ðar 1Þ C1 L1
þ iCout ð1 ar Þ iCin þ iRb ekt þ g2 ½iL1 ð1 af Þ þ y4 ð1 af Þ C1 L1
1
þ iCout ðaf 1Þ iCin af þ iRb af ekt g
Cout
ðjÞ
U1j¼5 ¼ ðICout þ iCout ekt Þ ðg1 g2 þ fiL1 ½g1 ðar 1Þ þ g2 ð1 af Þ
þ ½g1 ðar 1Þ þ g2 ð1 af Þ C1 L1 y4 þ iCout ½g1 ð1 ar Þ þ g2 ðaf 1Þ
1
iCin ½g1 þ g2 af þ iRb ½g1 þ g2 af g ekt Þ
Cout
ðjÞ ðjÞ
U1j¼5 ¼ ðICout g1 g2 þ ICout fiL1 ½g1 ðar 1Þ þ g2 ð1 af Þ
þ ½g1 ðar 1Þ þ g2 ð1 af Þ C1 L1 y4
þ iCout ½g1 ð1 ar Þ þ g2 ðaf 1Þ iCin ½g1 þ g2 af
þ iRb ½g1 þ g2 af g ekt þ g1 g2 iCout ekt
þ fiL1 ½g1 ðar 1Þ þ g2 ð1 af Þ
þ iCout ½g1 ð1 ar Þ þ g2 ðaf 1Þ iCin ½g1 þ g2 af
1
þ iRb ½g1 þ g2 af g iCout ekt ekt Þ
Cout
Since iL1 iCout ! e; y4 iCout ! e. . .
ðjÞ ðjÞ
U1j¼5 ¼ ðICout g1 g2 þ ICout fiL1 ½g1 ðar 1Þ þ g2 ð1 af Þ
þ ½g1 ðar 1Þ þ g2 ð1 af Þ C1 L1 y4 þ iCout ½g1 ð1 ar Þ þ g2 ðaf 1Þ
1
iCin ½g1 þ g2 af þ iRb ½g1 þ g2 af g ekt þ g1 g2 iCout ekt Þ
Cout
f1 ðiL1 ; y4 ; . . .Þ ¼ iL1 ½g1 ðar 1Þ þ g2 ð1 af Þ

þ iCout ½g1 ð1 ar Þ þ g2 ðaf 1Þ
iCin ½g1 þ g2 af þ iRb ½g1 þ g2 af
ðjÞ ðjÞ 1
U1j¼5 ¼ ðICout g1 g2 þ ICout f1 ðiL1 ; y4 ; . . .Þ ekt þ g1 g2 iCout ekt Þ
Cout
1
U1j¼6 ¼ ICout g1 ðar af 1Þ Ise
Cout
ðjÞ
¼ ðICout þ iCout ekt Þ fg1 þ ½iL1 ð1 af Þ þ y4 ð1 af Þ C1 L1
1
þ iCout ðaf 1Þ iCin af þ iRb af ekt g ðar af 1Þ Ise
Cout
1
Cout
ðjÞ ðjÞ
¼ fICout g1 þ ICout ½iL1 ð1 af Þ þ y4 ð1 af Þ C1 L1
þ iCout ðaf 1Þ iCin af þ iRb af ekt þ iCout ekt g1
þ iCout ekt ½iL1 ð1 af Þ
þ y4 ð1 af Þ C1 L1 þ iCout ðaf 1Þ
1
iCin af þ iRb af ekt g ðar af 1Þ Ise
Cout
Since iL1 iCout ! e; y4 iCout ! e . . .
1
Cout
ðjÞ ðjÞ
¼ fICout g1 þ ICout ½iL1 ð1 af Þ þ y4 ð1 af Þ C1 L1
1
þ iCout ðaf 1Þ iCin af þ iRb af ekt þ iCout ekt g1 g ðar af 1Þ Ise
Cout
f2 ðiL1 ; y4 ; . . .Þ ¼ iL1 ð1 af Þ þ y4 ð1 af Þ C1 L1 þ iCout ðaf 1Þ iCin

af þ iRb af
1 ðjÞ ðjÞ
U1j¼6 ¼ ICout g1 ðar af 1Þ Ise ¼ fICout g1 þ ICout f2 ðiL1 ; y4 ; . . .Þ ekt
Cout
1
þ g1 iCout ekt g ðar af 1Þ Ise
Cout
1
U1j¼7 ¼ ICout g2 ðar af 1Þ Isc
Cout
ðjÞ
¼ ðICout þ iCout ekt Þ fg2 þ ½iL1 ðar 1Þ þ y4 ðar 1Þ C1 L1
1
þ iCout ð1 ar Þ iCin þ iRb ekt g ðar af 1Þ Isc
Cout
1 ðjÞ ðjÞ
U1j¼7 ¼ ICout g2 ðar af 1Þ Isc ¼ fICout g2 þ ICout ½iL1 ðar 1Þ
Cout
þ y4 ðar 1Þ C1 L1 þ iCout ð1 ar Þ iCin þ iRb ekt
þ g2 iCout ekt þ ½iL1 ðar 1Þ
þ y4 ðar 1Þ C1 L1 þ iCout ð1 ar Þ iCin þ iRb
1
iCout ekt ekt g ðar af 1Þ Isc
Cout
Since iL1 iCout ! e; y4 iCout ! e . . .
1
U1j¼7 ¼ ICout g2 ðar af 1Þ Isc
Cout
ðjÞ ðjÞ
¼ fICout g2 þ ICout ½iL1 ðar 1Þ
þ y4 ðar 1Þ C1 L1 þ iCout ð1 ar Þ iCin þ iRb ekt
1
þ g2 iCout ekt g ðar af 1Þ Isc
Cout
f3 ðiL1 ; y4 ; . . .Þ ¼ iL1 ðar 1Þ þ y4 ðar 1Þ C1 L1 þ iCout ð1 ar Þ iCin þ iRb

1 ðjÞ ðjÞ
U1j¼7 ¼ ICout g2 ðar af 1Þ Isc ¼ fICout g2 þ ICout f3 ðiL1 ; y4 ; . . .Þ ekt
Cout
1
Cout
1 ðjÞ 1
U1j¼8 ¼ ICout ðar af 1Þ2 Isc Ise ¼ ðICout þ iCout ekt Þ ðar af 1Þ2 Isc Ise
Cout Cout
1 ðjÞ 1
U1j¼8 ¼ ICout ðar af 1Þ2 Isc Ise ¼ ICout ðar af 1Þ2 Isc Ise
Cout Cout
1
þ ðar af 1Þ2 Isc Ise iCout ekt
Cout
We can summary our last results in the following Table 5.1.
Table 5.1 Class C power amplifier (PA) with parallel resonance, U1j expressions
U1j Expression
j ¼ 1 ðU1j¼1 Þ U1j¼1 ¼ g1 g2 Rload y2 ekt
j ¼ 2 ðU1j¼2 Þ U1j¼2 ¼ g1 ðar af 1Þ Ise Rload y2 ekt
j ¼ 3 ðU1j¼3 Þ U1j¼3 ¼ g2 ðar af 1Þ Isc Rload y2 ekt
j ¼ 4 ðU1j¼4 Þ U1j¼4 ¼ Rload ðar af 1Þ2 Isc Ise y2 ekt
j ¼ 5 ðU1j¼5 Þ ðjÞ ðjÞ
U1j¼5 ¼ ðICout g1 g2 þ ICout f1 ðiL1 ; y4 ; . . .Þ ekt þ g1 g2 iCout ekt Þ C1out
U1j¼6 ¼ fICout g1 þ ICout f2 ðiL1 ; y4 ; . . .Þ ekt
1
þ g1 iCout ekt g ðar af 1Þ Ise
Cout
U1j¼7 ¼ fICout g2 þ ICout f3 ðiL1 ; y4 ; . . .Þ ekt
1
Cout
j ¼ 8 ðU1j¼8 Þ ðjÞ 1
U1j¼8 ¼ ICout ðar af 1Þ2 Isc Ise
Cout
1
þ ðar af 1Þ2 Isc Ise iCout ekt
Cout
U2 ðg1 ; g3 ; af ; ar ; Isc Þ ¼ g3 ½g1 þ ðaf ar 1Þ Isc

dVs ðtÞ
¼ g3 ð ! eÞ ½g1 þ ðaf ar 1Þ Isc
dt
g1 ðtÞ ¼ g1 þ ½iL1 ð1 af Þ þ y4 ð1 af Þ C1 L1 þ iCout ðaf 1Þ iCin af þ iRb af ekt
g2 ðtÞ ¼ g2 þ ½iL1 ðar 1Þ þ y4 ðar 1Þ C1 L1 þ iCout ð1 ar Þ iCin þ iRb ekt
dVs ðtÞ Rs 1
g3 ð ! eÞ ¼ g3 þ ½y3 ðar 1Þ þ y 2 ð1 ar Þ y1 ð1 þ Þ iCin ekt
dt Rb Cin Rb
a
g4 ðdVdts ðtÞ ! eÞ ¼ g4 þ ½y3 ð1 af Þ þ y2 ðaf 1Þ y1 af ð1 þ Rb Þ
Rs
iCin Cin R
f
b
ekt
dVs ðtÞ
U2 ðg1 ; g3 ð ! eÞ; af ; ar ; Isc Þ ¼ fg3 þ ½y3 ðar 1Þ þ y2 ð1 ar Þ
dt
Rs 1
y1 ð1 þ Þ iCin ekt g f½g1 þ ðaf ar 1Þ Isc þ ½iL1 ð1 af Þ
Rb C in Rb
þ y4 ð1 af Þ C1 L1 þ iCout ðaf 1Þ iCin af þ iRb af ekt g
dVs ðtÞ
U2 ðg1 ; g3 ð ! eÞ; af ; ar ; Isc Þ ¼ fg3 ½g1 þ ðaf ar 1Þ Isc
dt
Rs 1
þ ½g1 þ ðaf ar 1Þ Isc ½y3 ðar 1Þ þ y2 ð1 ar Þ y1 ð1 þ Þ iCin ekt
Rb Cin Rb
Rs 1
þ ½y3 ðar 1Þ þ y2 ð1 ar Þ y1 ð1 þ Þ iCin ekt ½iL1 ð1 af Þ
Rb Cin Rb
þ y4 ð1 af Þ C1 L1 þ iCout ðaf 1Þ iCin af þ iRb af ekt g
Since y3 iL1 ! e; y3 y4 ! e . . .
dVs ðtÞ
dt
Rs 1
þ ½g1 þ ðaf ar 1Þ Isc ½y3 ðar 1Þ þ y2 ð1 ar Þ y1 ð1 þ Þ iCin ekt g
Rb Cin Rb
We define two functions:
f4 ðiL1 ; y4 ; . . .Þ ¼ iL1 ð1 af Þ þ y4 ð1 af Þ C1 L1 þ iCout ðaf 1Þ iCin af þ iRb af

Rs 1
f5 ðy3 ; y2 ; . . .Þ ¼ y3 ðar 1Þ þ y2 ð1 ar Þ y1 ð1 þ Þ iCin
Rb Cin Rb
dVs ðtÞ
dt
þ g3 f4 ðiL1 ; y4 ; . . .Þ ekt þ ½g1 þ ðaf ar 1Þ Isc f5 ðy3 ; y2 ; . . .Þ ekt g
dVs ðtÞ dVs ðtÞ

U3 ðg2 ; g4 ð ! eÞ; af ; ar ; Ise Þ ¼ g4 ð ! eÞ ½g2 þ ðar af 1Þ Ise
dt dt
dVs ðtÞ
U3 ðg2 ; g4 ð ! eÞ; af ; ar ; Ise Þ ¼ fg4 þ ½y3 ð1 af Þ þ y2 ðaf 1Þ
dt
Rs af
y1 af ð1 þ Þ iCin ekt g fg2 þ ½iL1 ðar 1Þ þ y4 ðar 1Þ C1 L1
Rb Cin Rb
þ iCout ð1 ar Þ iCin þ iRb ekt þ ðar af 1Þ Ise g
dVs ðtÞ
U3 ðg2 ; g4 ð ! eÞ; af ; ar ; Ise Þ ¼ fg4 ½g2 þ ðar af 1Þ Ise
dt
þ g4 ½iL1 ðar 1Þ þ y4 ðar 1Þ C1 L1 þ iCout ð1 ar Þ iCin þ iRb ekt
Rs
þ ½g2 þ ðar af 1Þ Ise ½y3 ð1 af Þ þ y2 ðaf 1Þ y1 af ð1 þ Þ
Rb
af kt
iCin e þ ½y3 ð1 af Þ þ y2 ðaf 1Þ
Cin Rb
Rs af
y1 af ð1 þ Þ iCin ekt ½iL1 ðar 1Þ þ y4 ðar 1Þ C1 L1
Rb Cin Rb
Since y3 iL1 ! e; y3 y4 ! e . . .
dVs ðtÞ
dt
þ g4 ½iL1 ðar 1Þ þ y4 ðar 1Þ C1 L1
þ iCout ð1 ar Þ iCin þ iRb ekt þ ½g2 þ ðar af 1Þ Ise
Rs
½y3 ð1 af Þ þ y2 ðaf 1Þ y1 af ð1 þ Þ
Rb
af
iCin ekt g
Cin Rb
We define two functions:
f6 ðiL1 ; y4 ; . . .Þ ¼ iL1 ðar 1Þ þ y4 ðar 1Þ C1 L1 þ iCout ð1 ar Þ iCin þ iRb

Rs af
f7 ðy3 ; y2 ; . . .Þ ¼ y3 ð1 af Þ þ y2 ðaf 1Þ y1 af ð1 þ Þ iCin
Rb Cin Rb
dVs ðtÞ
dt
þ g4 f6 ðiL1 ; y4 ; . . .Þ ekt þ ½g2 þ ðar af 1Þ Ise f7 ðy3 ; y2 ; . . .Þ ekt g
Finally we get the enhance expression for tðY2 ; ICout ; . . .Þ

P3
k¼1 Uk
tðY2 ; ICout ; . . .Þ ¼
fC2 ðg1 ; g2 ; C1 ; ar ; af Þ þ C3 ðaf ; ar Þ ðiCin iRb Þ ekt g C1 L1
P8 P3
1
VT j¼1 U1j þ k¼2 Uk
tðY2 ; ICout ; . . .Þ ¼
fC2 ðg1 ; g2 ; C1 ; ar ; af Þ þ C3 ðaf ; ar Þ ðiCin iRb Þ ekt g C1 L1
P8
1 1 j¼1 U1j
tðY2 ; ICout ; . . . ¼ f
C1 L1 VT ½C2 ðg1 ; g2 ; C1 ; ar ; af Þ þ C3 ðaf ; ar Þ ðiCin iRb Þ ekt
P3
k¼2 Uk
þ g
½C2 ðg1 ; g2 ; C1 ; ar ; af Þ þ C3 ðaf ; ar Þ ðiCin iRb Þ ekt
1 1
tðY2 ; ICout ; . . .Þ ¼ f t1 ðY2 ; ICout ; . . .Þ þ t2 ðY2 ; ICout ; . . .Þg
C1 L1 VT
1 1 X2 X2
tðY2 ; ICout ; . . .Þ ¼ f t1k ðY2 ; ICout ; . . .Þ þ t2k ðY2 ; ICout ; . . .Þg
C1 L1 VT k¼1 k¼1
P
8
U1j
j¼1
t1 ðY2 ; ICout ; . . .Þ ¼
P3
k¼2 Uk
t2 ðY2 ; ICout ; . . .Þ ¼
1 1
tðY2 ; ICout ; . . .Þ ¼ f t1 ðY2 ; ICout ; . . .Þ þ t2 ðY2 ; ICout ; . . .Þg
C1 L1 VT
Stage 1:
P8 ½C2 ðg1 ; g2 ; C1 ; ar ; af Þ

j¼1 U1j C3 ðaf ; ar Þ ðiCin iRb Þ ekt
t1 ðY2 ; ICout ; . . .Þ ¼
½C2 ðg1 ; g2 ; C1 ; ar ; af Þ ½C2 ðg1 ; g2 ; C1 ; ar ; af Þ
þ C3 ðaf ; ar Þ ðiCin iRb Þ e C3 ðaf ; ar Þ ðiCin iRb Þ ekt
kt
P8
½C2 ðg1 ; g2 ; C1 ; ar ; af Þ C3 ðaf ; ar Þ ðiCin iRb Þ ekt j¼1 U1j
t1 ðY2 ; ICout ; . . .Þ ¼
½C22 ðg1 ; g2 ; C1 ; ar ; af Þ C23 ðaf ; ar Þ ðiCin iRb Þ e2kt
2
ðiCin iRb Þ2 ¼ i2Cin 2 iCin iRb þ i2Rb ; i2Cin ! e; iCin iRb ! e; i2Rb ! e; ðiCin iRb Þ2
!e
P8
½C2 ðg1 ; g2 ; C1 ; ar ; af Þ C3 ðaf ; ar Þ ðiCin iRb Þ ekt j¼1 U1j
t1 ðY2 ; ICout ; . . .Þ ¼
½C22 ðg1 ; g2 ; C1 ; ar ; af Þ
P8 P8
j¼1 U1j C3 ðaf ; ar Þ ðiCin iRb Þ ekt j¼1 U1j
t1 ðY2 ; ICout ; . . .Þ ¼
C2 ðg1 ; g2 ; C1 ; ar ; af Þ C22 ðg1 ; g2 ; C1 ; ar ; af Þ
P8
j¼1 U1j
t11 ðY2 ; ICout ; . . .Þ ¼ ; t12 ðY2 ; ICout ; . . .Þ
C2 ðg1 ; g2 ; C1 ; ar ; af Þ
P
C3 ðaf ; ar Þ ðiCin iRb Þ ekt 8j¼1 U1j
¼
C22 ðg1 ; g2 ; C1 ; ar ; af Þ
X
2
t1 ðY2 ; ICout ; . . .Þ ¼ t1k ðY2 ; ICout ; . . .Þ ¼ t11 ðY2 ; ICout ; . . .Þ þ t12 ðY2 ; ICout ; . . .Þ
k¼1
½g1 g2 Rload y2 ekt þ g1 ðar af 1Þ Ise Rload y2 ekt

þ g2 ðar af 1Þ Isc Rload y2 ekt þ Rload ðar af 1Þ2 Isc Ise y2 ekt
ðjÞ ðjÞ
þ ICout g1 g2 C1out þ ICout C1out f1 ðiL1 ; y4 ; . . .Þ ekt þ g1 g2 C1out iCout ekt
ðjÞ ðjÞ
þ fICout g1 þ ICout f2 ðiL1 ; y4 ; . . .Þ ekt þ g1 iCout ekt g ðar af 1Þ C1out Ise
ðjÞ ðjÞ
þ fICout g2 þ ICout f3 ðiL1 ; y4 ; . . .Þ ekt þ g2 iCout ekt g ðar af 1Þ C1out Isc
ðjÞ
þ ICout 1
Cout ðar af 1Þ2 Isc Ise þ Cout ðar af
1
1Þ2 Isc Ise iCout ekt
t11 ðY2 ; ICout ; . . .Þ ¼
C2 ðg1 ; g2 ; C1 ; ar ; af Þ
ðjÞ
½ICout C1out ½ðar af 1Þ2 Isc Ise þ g1 g2 þ g1 ðar af 1Þ Ise þ g2 ðar af 1Þ Isc
þ ½g1 g2 Rload y2 þ g1 ðar af 1Þ Ise Rload y2
þ g2 ðar af 1Þ Isc Rload y2 þ Rload ðar af 1Þ2 Isc Ise y2
ðjÞ ðjÞ
þ ICout C1out f1 ðiL1 ; y4 ; . . .Þ þ g1 g2 C1out iCout þ ICout ðar af 1Þ C1out Ise f2 ðiL1 ; y4 ; . . .Þ
ðjÞ
þ g1 ðar af 1Þ C1out Ise iCout þ ICout ðar af 1Þ C1out Isc f3 ðiL1 ; y4 ; . . .Þ
þ g2 ðar af 1Þ C1out Isc iCout þ C1out ðar af 1Þ2 Isc Ise iCout ekt
t11 ðY2 ; ICout ; . . .Þ ¼
C2 ðg1 ; g2 ; C1 ; ar ; af Þ
ðjÞ ðjÞ 1
X1 ðICout ; g1 ; g2 ; . . .Þ ¼ ICout ½ðar af 1Þ2 Isc Ise þ g1 g2
Cout
þ g1 ðar af 1Þ Ise þ g2 ðar af 1Þ Isc
ðjÞ
X1 ðICout ; g1 ; g2 ; . . .Þ
t11 ðY2 ; ICout ; . . .Þ ¼
C2 ðg1 ; g2 ; C1 ; ar ; af Þ
½g1 g2 Rload y2 þ g1 ðar af 1Þ Ise Rload y2
ðjÞ ðjÞ
þ ICout C1out f1 ðiL1 ; y4 ; . . .Þ þ g1 g2 C1out iCout þ ICout ðar af 1Þ C1out Ise f2 ðiL1 ; y4 ; . . .Þ
ðjÞ
þ g1 ðar af 1Þ C1out Ise iCout þ ICout ðar af 1Þ C1out Isc f3 ðiL1 ; y4 ; . . .Þ
þ g2 ðar af 1ÞCout
1
Isc iCout þ Cout1
ðar af 1Þ2 Isc Ise iCout ekt
þ C2 ðg1 ;g2 ;C1 ;ar ;af Þ
ðjÞ
X1 ðI ;g ;g ;...Þ !1 ðy2 ;f1 ðiL1 ;y4 ;...Þ;iCout ;...Þekt
We define t11 ðY2 ; ICout ; . . .Þ ¼ C2 ðgC;gout ;C11 ;a2r ;af Þ þ C2 ðg1 ;g2 ;C1 ;ar ;af Þ
1 2
!1 ðy2 ; f1 ðiL1 ; y4 ; . . .Þ; iCout ; . . .Þ ¼ ½g1 g2 Rload y2 þ g1 ðar af 1Þ Ise Rload y2
ðjÞ 1 1 ðjÞ 1
þ ICout f1 ðiL1 ; y4 ; . . .Þ þ g1 g2 iCout þ ICout ðar af 1Þ Ise f2 ðiL1 ; y4 ; . . .Þ
Cout Cout Cout
1 ðjÞ 1
þ g1 ðar af 1Þ Ise iCout þ ICout ðar af 1Þ Isc f3 ðiL1 ; y4 ; . . .Þ
Cout Cout
1 1
þ g2 ðar af 1Þ Isc iCout þ ðar af 1Þ2 Isc Ise iCout
Cout Cout
P
8
C3 ðaf ; ar Þ ðiCin iRb Þ ekt U1j
j¼1
t12 ðY2 ; ICout ; . . .Þ ¼
C22 ðg1 ; g2 ; C1 ; ar ; af Þ
C3 ðaf ; ar Þ ðiCin iRb Þ ekt ½g1 g2 Rload y2 ekt þ g1 ðar af 1Þ Ise Rload y2 ekt
þ g2 ðar af 1Þ Isc Rload y2 ekt þ Rload ðar af 1Þ2 Isc Ise y2 ekt
ðjÞ 1 ðjÞ 1 1
þ ICout g1 g2 þ ICout f1 ðiL1 ; y4 ; . . .Þ ekt þ g1 g2 iCout ekt
Cout Cout Cout
ðjÞ ðjÞ 1
þ fICout g1 þ ICout f2 ðiL1 ; y4 ; . . .Þ ekt þ g1 iCout ekt g ðar af 1Þ Ise
Cout
ðjÞ ðjÞ 1
þ fICout g2 þ ICout f3 ðiL1 ; y4 ; . . .Þ ekt þ g2 iCout ekt g ðar af 1Þ Isc
Cout
ðjÞ 1 1
þ ICout 2
ðar af 1Þ Isc Ise þ ðar af 1Þ Isc Ise iCout ekt
2
Cout Cout
t12 ðY2 ; ICout ; . . .Þ ¼ 2
C2 ðg1 ; g2 ; C1 ; ar ; af Þ
ðjÞ 1
C3 ðaf ; ar Þ ðiCin iRb Þ ekt ½ICout ½ðar af 1Þ2 Isc Ise þ g1 g2
Cout
þ C3 ðaf ; ar Þ ðiCin iRb Þ ekt ½g1 g2 Rload y2 þ g1 ðar af 1Þ Ise Rload y2
ðjÞ 1 1 ðjÞ 1
þ ICout f1 ðiL1 ; y4 ; :::Þ þ g1 g2 iCout þ ICout ðar af 1Þ Ise f2 ðiL1 ; y4 ; :::Þ
Cout Cout Cout
1 ðjÞ 1
þ g1 ðar af 1Þ Ise iCout þ ICout ðar af 1Þ Isc f3 ðiL1 ; y4 ; :::Þ
Cout Cout
1 1
þ g2 ðar af 1Þ Isc iCout þ ðar af 1Þ2 Isc Ise iCout ekt
Cout Cout
t12 ðY2 ; ICout ; :::Þ =
C22 ðg1 ; g2 ; C1 ; ar ; af Þ
Since ðiCin iRb Þ y2 ! e; ðiCin iRb Þ f1 ðiL1 ; y4 ; . . .Þ ! e. . .
1 ðjÞ
C3 ðaf ; ar Þ ðiCin iRb Þ ekt ½ICout ½ðar af 1Þ2 Isc Ise þ g1 g2
Cout
t12 ðY2 ; ICout ; . . .Þ ¼
C22 ðg1 ; g2 ; C1 ; ar ; af Þ
!2 ðg1 ;g2 ;ar ;af ;...ÞC3 ðaf ;ar ÞðiCin iRb Þekt
We define t12 ðY2 ; ICout ; . . .Þ ¼ C22 ðg1 ;g2 ;C1 ;ar ;af Þ
ðjÞ 1
!2 ðg1 ; g2 ; ar ; af ; . . .Þ ¼ ½ICout ½ðar af 1Þ2 Isc Ise þ g1 g2
Cout
Stage 2:
P3
k¼2 Uk
t2 ðY2 ; ICout ; . . .Þ ¼
dVs ðtÞ
U2 ðg1 ; g3 ð ! eÞ; af ; ar ; Isc Þ ¼ fg3 ½g1 þ ðaf ar 1Þ Isc þ g3 f4 ðiL1 ; y4 ; . . .Þ ekt
dt
þ ½g1 þ ðaf ar 1Þ Isc f5 ðy3 ; y2 ; . . .Þ ekt g
dVs ðtÞ
U3 ðg2 ; g4 ð ! eÞ; af ; ar ; Ise Þ ¼ fg4 ½g2 þ ðar af 1Þ Ise þ g4 f6 ðiL1 ; y4 ; . . .Þ ekt
dt
þ ½g2 þ ðar af 1Þ Ise f7 ðy3 ; y2 ; . . .Þ ekt g
U2 ðg1 ; g3 ðdVdts ðtÞ ! eÞ; af ; ar ; Isc Þ þ U3 ðg2 ; g4 ðdVdts ðtÞ ! eÞ; af ; ar ; Ise Þ
t2 ðY2 ; ICout ; . . .Þ ¼
U2 ðg1 ; g3 ðdVdts ðtÞ ! eÞ; af ; ar ; Isc Þ

t2 ðY2 ; ICout ; . . .Þ ¼
U3 ðg2 ; g4 ðdVdts ðtÞ ! eÞ; af ; ar ; Ise Þ
þ

t21 ðY2 ; ICout ; . . .Þ ¼

t22 ðY2 ; ICout ; . . .Þ ¼
X
2
t2 ðY2 ; ICout ; . . .Þ ¼ t2k ðY2 ; ICout ; . . .Þ ¼ t21 ðY2 ; ICout ; . . .Þ þ t22 ðY2 ; ICout ; . . .Þ
k¼1

t21 ðY2 ; ICout ; . . .Þ ¼ f g
½C2 ðg1 ; g2 ; C1 ; ar ; af Þ
þ C3 ðaf ; ar Þ ðiCin iRb Þ ekt
½C2 ðg1 ; g2 ; C1 ; ar ; af Þ
C3 ðaf ; ar Þ ðiCin iRb Þ ekt
f g
½C2 ðg1 ; g2 ; C1 ; ar ; af Þ
dVs ðtÞ
U2 ðg1 ; g3 ð ! eÞ; af ; ar ; Isc Þ ½C2 ðg1 ; g2 ; C1 ; ar ; af Þ
dt
t21 ðY2 ; ICout ; . . .Þ ¼
C22 ðg1 ; g2 ; C1 ; ar ; af Þ C23 ðaf ; ar Þ ðiCin iRb Þ2 e2kt
Assume ðiCin iRb Þ2 ! e then
dVs ðtÞ
U2 ðg1 ; g3 ð ! eÞ; af ; ar ; Isc Þ ½C2 ðg1 ; g2 ; C1 ; ar ; af Þ
dt
t21 ðY2 ; ICout ; . . .Þ ¼
C22 ðg1 ; g2 ; C1 ; ar ; af Þ

t21 ðY2 ; ICout ; . . .Þ ¼
C2 ðg1 ; g2 ; C1 ; ar ; af Þ
U2 ðg1 ; g3 ðdVdts ðtÞ ! eÞ; af ; ar ; Isc Þ C3 ðaf ; ar Þ ðiCin iRb Þ ekt

C22 ðg1 ; g2 ; C1 ; ar ; af Þ

t21 ðY2 ; ICout ; . . .Þ ¼
C2 ðg1 ; g2 ; C1 ; ar ; af Þ
fg3 ½g1 þ ðaf ar 1Þ Isc þ g3 f4 ðiL1 ; y4 ; . . .Þ ekt
þ ½g1 þ ðaf ar 1Þ Isc f5 ðy3 ; y2 ; . . .Þ ekt g C3 ðaf ; ar Þ ðiCin iRb Þ ekt
þ
C22 ðg1 ; g2 ; C1 ; ar ; af Þ
Since f4 ðiL1 ; y4 ; . . .Þ ðiCin iRb Þ ! e; f5 ðy3 ; y2 ; . . .Þ ðiCin iRb Þ ! e

t21 ðY2 ; ICout ; . . .Þ ¼
C2 ðg1 ; g2 ; C1 ; ar ; af Þ
g3 ½g1 þ ðaf ar 1Þ Isc C3 ðaf ; ar Þ ðiCin iRb Þ ekt
þ
C22 ðg1 ; g2 ; C1 ; ar ; af Þ
fg3 f4 ðiL1 ; y4 ; . . .Þ
g3 ½g1 þ ðaf ar 1Þ Isc þ ½g1 þ ðaf ar 1Þ Isc f5 ðy3 ; y2 ; . . .Þg ekt
t21 ðY2 ; ICout ; . . .Þ ¼
g3 ½g1 þ ðaf ar 1Þ Isc C3 ðaf ; ar Þ ðiCin iRb Þ ekt
þ
C22 ðg1 ; g2 ; C1 ; ar ; af Þ
g3 ½g1 þ ðaf ar 1ÞIsc

We define t21 ðY2 ; ICout ; . . .Þ ¼ C2 ðg1 ;g2 ;C1 ;ar ;af Þ þ !4 ðiCin ; iRb ; f4 ðiL1 ; y4 ; . . .Þ;
kt
. . .Þ e
g3 ½g1 þ ðaf ar 1Þ Isc C3 ðaf ; ar Þ ðiCin iRb Þ

!4 ðiCin ; iRb ; f4 ðiL1 ; y4 ; . . .Þ; . . .Þ ¼
C22 ðg1 ; g2 ; C1 ; ar ; af Þ
fg f4 ðiL1 ; y4 ; . . .Þ þ ½g1 þ ðaf ar 1Þ Isc f5 ðy3 ; y2 ; . . .Þg
3
C2 ðg1 ; g2 ; C1 ; ar ; af Þ

t22 ðY2 ; ICout ; . . .Þ ¼
½C2 ðg1 ; g2 ; C1 ; ar ; af Þ
þ C3 ðaf ; ar Þ ðiCin iRb Þ ekt
½C2 ðg1 ; g2 ; C1 ; ar ; af Þ
f g
½C2 ðg1 ; g2 ; C1 ; ar ; af Þ
dVs ðtÞ
U3 ðg2 ; g4 ð ! eÞ; af ; ar ; Ise Þ ½C2 ðg1 ; g2 ; C1 ; ar ; af Þ
dt
t22 ðY2 ; ICout ; . . .Þ ¼
C22 ðg1 ; g2 ; C1 ; ar ; af Þ C23 ðaf ; ar Þ ðiCin iRb Þ2 e2kt
Assume ðiCin iRb Þ2 ! e then
dVs ðtÞ
U3 ðg2 ; g4 ð ! eÞ; af ; ar ; Ise Þ ½C2 ðg1 ; g2 ; C1 ; ar ; af Þ
dt
t22 ðY2 ; ICout ; . . .Þ ¼
C22 ðg1 ; g2 ; C1 ; ar ; af Þ

t22 ðY2 ; ICout ; . . .Þ ¼
C2 ðg1 ; g2 ; C1 ; ar ; af Þ
U3 ðg2 ; g4 ðdVdts ðtÞ ! eÞ; af ; ar ; Ise Þ C3 ðaf ; ar Þ ðiCin iRb Þ ekt

C22 ðg1 ; g2 ; C1 ; ar ; af Þ

t22 ðY2 ; ICout ; . . .Þ ¼
C2 ðg1 ; g2 ; C1 ; ar ; af Þ
fg4 ½g2 þ ðar af 1Þ Ise þ g4 f6 ðiL1 ; y4 ; . . .Þ ekt
þ ½g2 þ ðar af 1Þ Ise f7 ðy3 ; y2 ; . . .Þ ekt g C3 ðaf ; ar Þ ðiCin iRb Þ ekt
þ
C22 ðg1 ; g2 ; C1 ; ar ; af Þ
Since f6 ðiL1 ; y4 ; . . .Þ ðiCin iRb Þ ! e; f7 ðy3 ; y2 ; . . .Þ ðiCin iRb Þ ! e

t22 ðY2 ; ICout ; . . .Þ ¼
C2 ðg1 ; g2 ; C1 ; ar ; af Þ
g4 ½g2 þ ðar af 1Þ Ise C3 ðaf ; ar Þ ðiCin iRb Þ ekt
þ
C22 ðg1 ; g2 ; C1 ; ar ; af Þ
g4 ½g2 þ ðar af 1Þ Ise

t22 ðY2 ; ICout ; . . .Þ ¼
C2 ðg1 ; g2 ; C1 ; ar ; af Þ
fg4 f6 ðiL1 ; y4 ; . . .Þ
þ ½g2 þ ðar af 1Þ Ise f7 ðy3 ; y2 ; . . .Þg ekt

C2 ðg1 ; g2 ; C1 ; ar ; af Þ
g4 ½g2 þ ðar af 1Þ Ise C3 ðaf ; ar Þ ðiCin iRb Þ ekt
þ
C22 ðg1 ; g2 ; C1 ; ar ; af Þ
g4 ½g2 þ ðar af 1ÞIse

We define t22 ðY2 ; ICout ; . . .Þ ¼ C2 ðg1 ;g2 ;C1 ;ar ;af Þ þ !3 ðiCin ; iRb ; . . .Þ ekt
!3 ðiCin ; iRb ; f6 ðiL1 ; y4 ; . . .Þ; f7 ðy3 ; y2 ; . . .ÞÞ

g4 ½g2 þ ðar af 1Þ Ise C3 ðaf ; ar Þ ðiCin iRb Þ
¼
C22 ðg1 ; g2 ; C1 ; ar ; af Þ
fg f6 ðiL1 ; y4 ; . . .Þ þ ½g2 þ ðar af 1Þ Ise f7 ðy3 ; y2 ; . . .Þg
4
C2 ðg1 ; g2 ; C1 ; ar ; af Þ
Summary: We define function: dY4

dt ¼ tðY2 ; ICout ; . . .Þ
dY4 1 1 X2 X2
¼ f t1k ðY2 ; ICout ; . . .Þ þ t2k ðY2 ; ICout ; . . .Þg
dt C1 L1 VT k¼1 k¼1
dY4 1 X
2
1 X
2
¼ t1k ðY2 ; ICout ; . . .Þ þ t2k ðY2 ; ICout ; . . .Þ
dt C1 L1 VT k¼1 C1 L1 k¼1
ðjÞ
X1 ðICout ; g1 ; g2 ; . . .Þ !1 ðy2 ; f1 ðiL1 ; y4 ; . . .Þ; iCout ; . . .Þ ekt
t11 ðY2 ; ICout ; . . .Þ ¼ þ
!2 ðg1 ; g2 ; ar ; af ; . . .Þ C3 ðaf ; ar Þ ðiCin iRb Þ ekt

t12 ðY2 ; ICout ; . . .Þ ¼
C22 ðg1 ; g2 ; C1 ; ar ; af Þ
g3 ½g1 þ ðaf ar 1Þ Isc

t21 ðY2 ; ICout ; . . .Þ ¼ þ !4 ðiCin ; iRb ; f4 ðiL1 ; y4 ; . . .Þ; . . .
C2 ðg1 ; g2 ; C1 ; ar ; af Þ
ekt
g4 ½g2 þ ðar af 1Þ Ise

t22 ðY2 ; ICout ; . . .Þ ¼ þ !3 ðiCin ; iRb ; . . .Þ ekt
C2 ðg1 ; g2 ; C1 ; ar ; af Þ
ðjÞ
1 X1 ðICout ; g1 ; g2 ; . . .Þ !1 ðy2 ; f1 ðiL1 ; y4 ; . . .Þ; iCout ; . . .Þ ekt
tðY2 ; ICout ; . . .Þ ¼ f þ
C1 L1 VT C2 ðg1 ; g2 ; C1 ; ar ; af Þ C2 ðg1 ; g2 ; C1 ; ar ; af Þ
!2 ðg1 ; g2 ; ar ; af ; . . .Þ C3 ðaf ; ar Þ ðiCin iRb Þ ekt
þ g
C22 ðg1 ; g2 ; C1 ; ar ; af Þ
1 g ½g1 þ ðaf ar 1Þ Isc
þ f 3
C1 L1 C2 ðg1 ; g2 ; C1 ; ar ; af Þ
g ½g2 þ ðar af 1Þ Ise
þ !4 ðiCin ; iRb ; f4 ðiL1 ; y4 ; . . .Þ; . . .Þ ekt þ f 4 g
C2 ðg1 ; g2 ; C1 ; ar ; af Þ
þ !3 ðiCin ; iRb ; . . .Þ ekt g
ðjÞ
1 X1 ðICout ; g1 ; g2 ; . . .Þ
tðY2 ; ICout ; . . .Þ ¼
C1 L1 VT C2 ðg1 ; g2 ; C1 ; ar ; af Þ
1 !1 ðy2 ; f1 ðiL1 ; y4 ; . . .Þ; iCout ; . . .Þ ekt
þ
C1 L1 VT C2 ðg1 ; g2 ; C1 ; ar ; af Þ
1 !2 ðg1 ; g2 ; ar ; af ; . . .Þ C3 ðaf ; ar Þ ðiCin iRb Þ ekt
þ
C1 L1 VT C22 ðg1 ; g2 ; C1 ; ar ; af Þ
1 g ½g1 þ ðaf ar 1Þ Isc 1
3 þ !4 ðiCin ; iRb ; f4 ðiL1 ; y4 ; . . .Þ; . . .Þ ekt
C1 L1 C2 ðg1 ; g2 ; C1 ; ar ; af Þ C1 L1
1 g ½g2 þ ðar af 1Þ Ise 1
4 þ !3 ðiCin ; iRb ; . . .Þ ekt
C1 L1 C2 ðg1 ; g2 ; C1 ; ar ; af Þ C1 L1
ðjÞ
fV1T X1 ðICout ; g1 ; g2 ; . . .Þ g3 ½g1 þ ðaf ar 1Þ Isc g4 ½g2 þ ðar af 1Þ Ise g
tðY2 ; ICout ; . . .Þ ¼
C1 L1 C2 ðg1 ; g2 ; C1 ; ar ; af Þ
1 !1 ðy2 ; f1 ðiL1 ; y4 ; . . .Þ; iCout ; . . .Þ 1
þf þ
C 1 L 1 VT C2 ðg1 ; g2 ; C1 ; ar ; af Þ C1 L1 VT
!2 ðg1 ; g2 ; ar ; af ; . . .Þ C3 ðaf ; ar Þ ðiCin iRb Þ

C22 ðg1 ; g2 ; C1 ; ar ; af Þ
1 1
þ !4 ðiCin ; iRb ; f4 ðiL1 ; y4 ; . . .Þ; . . .Þ þ !3 ðiCin ; iRb ; . . .Þg ekt
C1 L1 C1 L1
At fixed point:
ðjÞ
fV1T X1 ðICout ; g1 ; g2 ; . . .Þ g3 ½g1 þ ðaf ar 1Þ Isc g4 ½g2 þ ðar af 1Þ Ise g
¼0
C1 L1 C2 ðg1 ; g2 ; C1 ; ar ; af Þ
1 !1 ðy2 ; f1 ðiL1 ; y4 ; . . .Þ; iCout ; . . .Þ

tðY2 ; ICout ; . . .Þ ¼ f
C 1 L 1 VT C2 ðg1 ; g2 ; C1 ; ar ; af Þ
1 !2 ðg1 ; g2 ; ar ; af ; . . .Þ C3 ðaf ; ar Þ ðiCin iRb Þ
þ
C1 L1 VT C22 ðg1 ; g2 ; C1 ; ar ; af Þ
1 1
þ !4 ðiCin ; iRb ; f4 ðiL1 ; y4 ; . . .Þ; . . .Þ þ !3 ðiCin ; iRb ; . . .Þg ekt
C1 L1 C1 L1
dY4 dY4
¼ tðY2 ; ICout ; . . .Þ; ¼ y4 k ekt
dt dt
1 !1 ðy2 ; f1 ðiL1 ; y4 ; . . .Þ; iCout ; . . .Þ
y4 k þ f
C1 L1 VT C2 ðg1 ; g2 ; C1 ; ar ; af Þ
þ
C1 L1 VT C22 ðg1 ; g2 ; C1 ; ar ; af Þ
1 1
þ !4 ðiCin ; iRb ; f4 ðiL1 ; y4 ; . . .Þ; . . .Þ þ !3 ðiCin ; iRb ; . . .Þg ¼ 0
C1 L1 C1 L1
We build our system matrices ðiRb ; iCin ; iCout ; iL1 ; y3 ; y4 Þ and ðy1 ; y2 Þ
Rs 1
iRb k y1 iCin ¼ 0; iCin k þ y1 ¼ 0; iCout k þ y2 ¼ 0
Rb Cin Rb
iL1 k þ y3 ¼ 0; y3 k þ y4 ¼ 0
1 !1 ðy2 ; f1 ðiL1 ; y4 ; . . .Þ; iCout ; . . .Þ

y4 k þ f
C1 L1 VT C2 ðg1 ; g2 ; C1 ; ar ; af Þ
þ
C1 L1 VT C22 ðg1 ; g2 ; C1 ; ar ; af Þ
1 1
þ !4 ðiCin ; iRb ; f4 ðiL1 ; y4 ; . . .Þ; . . .Þ þ !3 ðiCin ; iRb ; . . .Þg ¼ 0
C1 L1 C1 L1
0 1 0 1
m11iRb m12
B C B
0 1 B iCin C B m21
B C m22 C
C
i11 . . . i16 B C B C !
B . . C B iCout C B m31 m32 C
B . . . ... C B C B C y1
@ . AB CþB
B
C ¼ 0; i11 ¼ k;
B iL C B m41 m42 C
C y2
B 1C B C
i61 . . . i66 B C
B y3 C B m m52 C
@ A @ 51 A
y4 m61 m62
1
i12 ¼
Cin Rb
i13 ¼ 0; i14 ¼ i15 ¼ i16 ¼ 0; i21 ¼ 0; i22 ¼ k; i23 ¼ i24 ¼ i25 ¼ i26 ¼ 0
i31 ¼ i32 ¼ 0; i33 ¼ k; i34 ¼ i35 ¼ i36 ¼ 0; i41 ¼ i42 ¼ i43 ¼ 0; i44 ¼ k; i45
¼ 1; i46 ¼ 0
i51 ¼ i52 ¼ i53 ¼ i54 ¼ 0; i55 ¼ k; i56 ¼ 1; i66 ¼ k. To find i16 ; . . .; i56 ; m61 ;
m62 , we need to do some analytic work with the expression:
1 !1 ðy2 ; f1 ðiL1 ; y4 ; . . .Þ; iCout ; . . .Þ

C1 L1 VT C2 ðg1 ; g2 ; C1 ; ar ; af Þ
þ
C1 L1 VT C22 ðg1 ; g2 ; C1 ; ar ; af Þ
1 1
þ !4 ðiCin ; iRb ; f4 ðiL1 ; y4 ; . . .Þ; . . .Þ þ !3 ðiCin ; iRb ; . . .Þ
C1 L1 C1 L1
Rs
m11 ¼ ; m12 ¼ 0; m21 ¼ 1; m22 ¼ 0; m31 ¼ 0; m32 ¼ 1; m41 ¼ m42 ¼ 0; m51 ¼ m52
Rb
¼0
Assumption mlk ; l ¼ 1; . . .; 6; k ¼ 1; 2 elements are neglected compare to ilk ele-

ments l ¼ 1; . . .; 6; k ¼ 1; . . .; 6.
0 1 0 1
i11 ... i16 i11 . . . i16
B .. .. .. C B .. .. . C
ðA k IÞ ¼ @ . . . A; det ðA k IÞ ¼ 0 ) det@ . . .. A ¼ 0
i61 ... i66 i61 . . . i66

without calculating the eigenvalues explicit. We use criterion of Routh-Hurwitz
[2–4].
5.3 Single Ended Class B Amplifier … 559
5.3 Single Ended Class B Amplifier Gummel-Poon Model

Analysis Under Parameters Variation
Class B amplifier is a type of power amplifier where the active device (transistor)
conducts only for one half cycle of the input signal. That means the conduction
angle is 180° for a Class B amplifier. Since the active device is switched off for half
the input cycle, the active device dissipates less power and hence the efficiency is
improved. Theoretical maximum efficiency of Class B power amplifier is 78.5 %. it
improves the power efficiency, it creates a lot of distortion. You can find class B
amplifier in the RF power amplifiers where the distortion is not a matter of major
concern. We use in our stability analysis The BJT NPN transistor Gummel-Poon
model since the input signal is large. The small signal S-parameters are not useful
for large-signal or high power circuit design such as power amplifier, mixers,
frequency converters because the active devices (transistor/FET/diode) in these
circuits usually operate in the nonlinear regions. In large signal circuits the voltage
and current variation will be large, for BJT this means the variation of the transistor
terminals voltages will be greater than VT. We must use large signal model of the
transistor, such as Ebers-Molls model, the Gummel-Poon model, the VBIC model.
The most popular large-signal model for BJT is the Spice Gummel Poon
(SGP) model. A more recent alternative to the SGP model is the Vertical Bipolar
Intercompany Model (VBIC) model which offers more accuracies as compared to
SGP model. The Spice Gummel Poon model is based on the device physics of
bipolar junction transistor. The Gummel Poon model is a compact model for bipolar
junction transistor which also takes into account effects of low currents and at high
level injection signal [91–93] (Fig. 5.5).
The base of the transistor Q1 is not biased and the negative half cycle of the
input waveform is missing in the output. Even though it improves the power
Fig. 5.5 Single ended class B amplifier

Fig. 5.6 Single ended class B amplifier equivalent circuit with BJT transistor Gummel-Poon
model
efficiency, it creates a lot of distortion. Only half the information present in the
input will be available in the output and that is a bad thing. Single ended Class B
amplifiers are not used in present day practical audio amplifier application and they
can be found only in some earlier gadgets. Another place where you can find them
is the RF power amplifiers where the distortion is not a matter of major concern.
The single ended class B amplifier equivalent circuit with BJT transistor
Gummel-Poon model is present in the below schematic. (Fig. 5.6).
The Gummel-Poon schematic equivalent circuit shows the large signal schematic
of the Gummel-Poon model. It represents the physical transistor, a current-controlled
output current sink, and two diode structures including their capacitors.
Iss qVBC Iss qVBE

Iss qVBE qVBC
Iec ¼ ðenr kB T 1Þ; Icc ¼ ðenf kB T 1Þ; Ict ¼ Icc Iec ¼ ½enf kB T enr kB T
qb qb qb
Iec Iss qVBC I cc I ss
qVBE
I D1 ¼ ¼ ðenr kB T 1Þ; ID3 ¼ ¼ ðenf kB T 1Þ
br qb br bf qb b f
qVBC qVBE
ID2 ¼ C4 Iss ðencl kB T 1Þ; ID4 ¼ C2 Iss ðenel kB T 1Þ; IRc ¼ Irc ; IRs ¼ ICin ; ICout ¼ IRload
P
KCL @ (C): C ¼ C0 ; Irc ! e; VC0 VC ! e; IRc þ ICjx þ ICbc þ 2k¼1 IDk ¼ ICt þ ICcs
P
KCL @ (B): B ¼ B0 ; Irb ! e; VB0 VB ! e; ICin ¼ ICjx þ ICbc þ ICbe þ 4k¼1 IDk
P4
KCL @ (E): E ¼ E0 ; Ire ! e; VE VE0 ! e; ICbe þ k¼3 IDk þ ICt ¼ IRe þ ICout
d d
ICjx ¼ Cjx ðVB0 VC ÞjVB0 ¼VB ¼ Cjx ðVB VC Þ;
dt dt
d d
ICin ¼ Cin ðVA1 VB0 ÞjVB0 ¼VB ¼ Cin ðVA1 VB Þ
dt dt
d d
ICbc ¼ Cbc ðVB VC Þ; ICbe ¼ Cbe ðVB VE Þ;
dt dt
d d
ICout ¼ Cout ðVE0 VA2 ÞjVE0 ¼VE ¼ Cout ðVE VA2 Þ
dt dt
dVc Vs ðtÞ VA1 VA2 Vout
ICcs ¼ Ccs ; IRs ¼ ; VA2 ¼ Vout ; IRload ¼ ¼ ; IC ¼ IRload
dt Rs Rload Rload out
Vcc Vc0 Vcc Vc VE0 VE

IR c ¼ jVc ¼Vc0 ¼ ; IRe ¼ jVE ¼VE0 ¼
Rc Rc Re Re
Gummel-Poon model parameters:

nf forward current emission coefficient.
nr reverse current emission coefficient.
qb variable model parameter (early and Kirk effects).
nel ; ncl ; C2 ; C4 model parameters responsible for low current effects.
kB T
q
thermal voltage. bf —forward current gain. br — reverse current
gain.
P
KCL @ (C): C ¼ C 0 ; Irc ! e; VC0 VC ! e; IRc þ ICjx þ ICbc þ 2k¼1 IDk ¼
ICt þ ICcs
Vcc Vc d 1 qVBC qVBC

þ ½Cjx þ Cbc ðVB VC Þ þ Iss ½ ðenr kB T 1Þ þ C4 ðencl kB T 1Þ
Rc dt qb br
Iss nqVBE qVBC dV c
¼ ½e f kB T enr kB T Þ þ Ccs ; VBC ¼ VB VC
qb dt
We define for simplicity new functions:
1 qVBC qVBC
n1 ðVBC Þ ¼ Iss ½ ðenr kB T 1Þ þ C4 ðencl kB T 1Þ; n2 ðVBE ; VBC Þ ¼ ICt
qb b r
Iss nqV BE qVBC
¼ ½e f kB T enr kB T
qb
Vcc Vc dVBC dVc

þ ½Cjx þ Cbc þ n1 ðVBC Þ ¼ n2 ðVBE ; VBC Þ þ Ccs
Rc dt dt
P4
KCL @ (B): B ¼ B0 ; Irb ! e; VB0 VB ! e; ICin ¼ ICjx þ ICbc þ ICbe þ k¼1 IDk
d d d
Cin ðVA1 VB Þ ¼ Cjx ðVB VC Þ þ Cbc ðVB VC Þ
dt dt dt
d Iss qVBC
þ Cbe ðVB VE Þ þ ðenr kB T 1Þ
dt qb br
qVBC Iss qVBE
þ C4 Iss ðencl kB T 1Þ þ ðenf kB T 1Þ
qb bf
qVBE
þ C2 Iss ðenel kB T 1Þ
d dVBC dVBE
Cin ðVA1 VB Þ ¼ ½Cjx þ Cbc þ Cbe þ n3 ðVBE ; VBC Þ
dt dt dt
1 qVBC qVBC
n3 ðVBE ; VBC Þ ¼ Iss ½ ðenr kB T 1Þ þ C4 ðencl kB T 1Þ
qb br
1 qVBE qVBE
þ Iss ½ ðenf kB T 1Þ þ C2 ðenel kB T 1Þ
qb b f
1 qVBE qVBE
n4 ðVBE Þ ¼ Iss ½ ðenf kB T 1Þ þ C2 ðenel kB T 1Þ;
qb bf
n3 ðVBE ; VBC Þ ¼ n1 ðVBC Þ þ n4 ðVBE Þ
P4
KCL @ (E): E ¼ E0 ; Ire ! e; VE VE0 ! e; ICbe þ k¼3 IDk þ ICt ¼ IRe þ ICout
d 1 qVBE qVBE
Cbe ðVB VE Þ þ Iss ½ ðenf kB T 1Þ þ C2 ðenel kB T 1Þ
dt qb bf
Iss qVBE qVBC VE VA2
þ ½enf kB T enr kB T Þ ¼ þ
qb Re Rload
dVBE VE VA2
Cbe þ n4 ðVBE Þ þ n2 ðVBE ; VBC Þ ¼ þ ;
dt Re Rload
dVBE VE d
Cbe þ n4 ðVBE Þ þ n2 ðVBE ; VBC Þ ¼ þ Cout ðVE VA2 Þ
dt Re dt
VA2 d d
ICout ¼ IRload ) ¼ Cout ðVE VA2 Þ ) VA2 ¼ Rload Cout ðVE VA2 Þ
Rload dt dt
Vs ðtÞ VA1 d
IRs ¼ ICin ) ¼ Cin ðVA1 VB Þ ) Vs ðtÞ VA1
Rs dt
d
¼ Rs Cin ðVA1 VB Þ
dt
Summary of our system differential equations:
Vcc Vc dVBC dVc

þ ½Cjx þ Cbc þ n1 ðVBC Þ ¼ n2 ðVBE ; VBC Þ þ Ccs
Rc dt dt
d dVBC dVBE
Cin ðVA1 VB Þ ¼ ½Cjx þ Cbc þ Cbe þ n3 ðVBE ; VBC Þ
dt dt dt
dVBE VE VA2
dt Re Rload
dVBE VE d
Cbe þ n4 ðVBE Þ þ n2 ðVBE ; VBC Þ ¼ þ Cout ðVE VA2 Þ
dt Re dt
d d
VA2 ¼ Rload Cout ðVE VA2 Þ; Vs ðtÞ VA1 ¼ Rs Cin ðVA1 VB Þ
dt dt
&&
VB ¼ VBE þ VE ; VC ¼ VCB þ VBE þ VE ; VBC ¼ VCB ; VCB ¼ VBC ; VC

¼ VBE VBC þ VE
Vcc ½VBE VBC þ VE dVBC

þ ½Cjx þ Cbc þ n1 ðVBC Þ
Rc dt
d
¼ n2 ðVBE ; VBC Þ þ Ccs ðVBE VBC þ VE Þ
dt
þ ½Cjx þ Cbc þ Ccs þ n1 ðVBC Þ
Rc dt
dVBE dVE
¼ n2 ðVBE ; VBC Þ þ Ccs þ Ccs
dt dt
d dVBC dVBE
Cin ðVA1 VBE VE Þ ¼ ½Cjx þ Cbc þ Cbe þ n3 ðVBE ; VBC Þ
dt dt dt
dVA1 dVE dVBC dVBE
Cin Cin ¼ ½Cjx þ Cbc þ ½Cbe þ Cin þ n3 ðVBE ; VBC Þ
dt dt dt dt
dVBE VE VA2
dt Re Rload
dVE dVA2
VA2 ¼ Rload Cout Rload Cout
dt dt
d
Vs ðtÞ VA1 ¼ Rs Cin ðVA1 VBE VE Þ ) Vs ðtÞ VA1
dt
dVA1 dVBE dVE
¼ Rs Cin Rs Cin Rs Cin
dt dt dt
dVs ðtÞ dA0 dnðtÞ dnðtÞ

Vs ðtÞ ¼ A0 þ nðtÞjA0 nðtÞ A0 ; ¼ þ j e;
dt dt dt A0 nðtÞ dt
dVs ðtÞ
! e; !e
dt
Assumption IRs Rs ! e ) VA1 Vs ðtÞ; dVdts ðtÞ ! e ) dtA1 ! e; Vs ðtÞ VA1 ! e
dV
and we get the following system differential equations:

þ ½Cjx þ Cbc þ Ccs þ n1 ðVBC Þ
Rc dt
dVBE dVE
¼ n2 ðVBE ; VBC Þ þ Ccs þ Ccs
dt dt
dVE dVBC dVBE
Cin ¼ ½Cjx þ Cbc þ ½Cbe þ Cin þ n3 ðVBE ; VBC Þ; Rs Cin
dt dt dt
dVBE dVE
Rs Cin
dt dt
¼0
dVBE VE VA2
Cbe þ n4 ðVBE Þ þ n2 ðVBE ; VBC Þ ¼ þ ; VA2
dt Re Rload
dVE dVA2
¼ Rload Cout Rload Cout
dt dt
dVA1 dVBE dVE dVBE dVE
Rs Cin ð ! eÞ Rs Cin Rs Cin ¼ 0; ¼
dt dt dt dt dt
We get the following set of system differential equations:

þ ½Cjx þ Cbc þ Ccs þ n1 ðVBC Þ ¼ n2 ðVBE ; VBC Þ
Rc dt
dVBE dVBC dVBE

Cin ¼ ½Cjx þ Cbc þ ½Cbe þ Cin þ n3 ðVBE ; VBC Þ
dt dt dt
dVBE VE VA2
dt Re Rload
dVBE dVA2
VA2 ¼ Rload Cout Rload Cout
dt dt
First differential equation:

þ ½Cjx þ Cbc þ Ccs þ n1 ðVBC Þ ¼ n2 ðVBE ; VBC Þ
Rc dt
VE ¼ Vcc VBE þ VBC Rc n2 ðVBE ; VBC Þ þ Rc n1 ðVBC Þ þ Rc ½Cjx þ Cbc þ Ccs

dVBC

dt
Third differential equation:
dVBE VE VA2
Cbe þ n4 ðVBE Þ þ n2 ðVBE ; VBC Þ ¼ þ
dt Re Rload
dVBE VA2
VE ¼ Cbe Re þ n4 ðVBE Þ Re þ n2 ðVBE ; VBC Þ Re Re
dt Rload
Combine of the first and third differential equations:
dVBC
Vcc VBE þ VBC Rc n2 ðVBE ; VBC Þ þ Rc n1 ðVBC Þ þ Rc ½Cjx þ Cbc þ Ccs
dt
dVBE VA2
¼ Cbe Re þ n4 ðVBE Þ Re þ n2 ðVBE ; VBC Þ Re Re
dt Rload
dVBC
Vcc VBE þ VBC þ Rc n1 ðVBC Þ þ Rc ½Cjx þ Cbc þ Ccs
dt
dVBE VA2
¼ Cbe Re þ n4 ðVBE Þ Re þ ½Re þ Rc n2 ðVBE ; VBC Þ Re
dt Rload
We can summary our system reduced three differential equations:

dVBC
dt
(1)
dVBE VA2
dt Rload
(2) Cin dVdtBE ¼ ½Cjx þ Cbc dVdtBC þ ½Cbe þ Cin dVdtBE þ n3 ðVBE ; VBC Þ
dVA2
(3) VA2 ¼ Rload Cout dVdtBE Rload Cout dt
dVA2 dVA2
(3) VA2 ¼ Rload Cout dVdtBE Rload Cout dt ) dt ¼ dVdtBE Rload1Cout VA2
(2) Cin dVdtBE ¼ ½Cjx þ Cbc dVdtBC þ ½Cbe þ Cin dVdtBE þ n3 ðVBE ; VBC Þ
dVBC dVBE
½Cjx þ Cbc þ Cbe þ n3 ðVBE ; VBC Þ ¼ 0
dt dt
dVBE dVBC
) Cbe ¼ ½Cjx þ Cbc n3 ðVBE ; VBC Þ
dt dt
dVBE Cjx þ Cbc dVBC 1
¼ ½ n ðVBE ; VBC Þ
dt Cbe dt Cbe 3
dVBC
dt
(1)
dVBE VA2
dt Rload
dVBC
dt
Cjx þ Cbc dVBC
¼ Cbe Re f½
Cbe dt
1 VA2
þ n3 ðVBE ; VBC Þg þ n4 ðVBE Þ Re þ ½Re þ Rc n2 ðVBE ; VBC Þ Re
Cbe Rload
dVBC
ðRc ½Cjx þ Cbc þ Ccs þ Re ½Cjx þ Cbc Þ ¼ Re n3 ðVBE ; VBC Þ þ n4 ðVBE Þ Re
dt
VA 2
þ ½Re þ Rc n2 ðVBE ; VBC Þ Rc n1 ðVBC Þ Re Vcc þ VBE VBC
Rload
We define for simplicity global parameter C1 ¼ C1 ðRc ; Re ; . . .Þ
C1 ðRc ; Re ; . . .Þ ¼ ðRc ½Cjx þ Cbc þ Ccs þ Re ½Cjx þ Cbc Þ
dVBC Re Re
¼ n ðVBE ; VBC Þ þ n4 ðVBE Þ
dt C1 ðRc ; Re ; . . .Þ 3 C1 ðRc ; Re ; . . .Þ
Re þ Rc Rc
þ½ n2 ðVBE ; VBC Þ n ðVBC Þ
C1 ðRc ; Re ; . . .Þ C1 ðRc ; Re ; . . .Þ 1
VA2 Re VBE VBC Vcc
þ
Rload C1 ðRc ; Re ; . . .Þ C1 ðRc ; Re ; . . .Þ
Re Re þ Rc
X1 ðRc ; Re ; . . .Þ ¼ ; X2 ðRc ; Re ; . . .Þ ¼ ; X3 ðRc ; Re ; . . .Þ
C1 ðRc ; Re ; . . .Þ C1 ðRc ; Re ; . . .Þ
Rc
¼
C1 ðRc ; Re ; . . .Þ
dVBC
¼ X1 ðRc ; Re ; . . .Þ n3 ðVBE ; VBC Þ þ n4 ðVBE Þ X1 ðRc ; Re ; . . .Þ
dt
þ X2 ðRc ; Re ; . . .Þ n2 ðVBE ; VBC Þ X3 ðRc ; Re ; . . .Þ n1 ðVBC Þ
VA2 VBE VBC Vcc
X1 ðRc ; Re ; . . .Þ þ
Rload C1 ðRc ; Re ; . . .Þ
dVBE Cjx þ Cbc

¼ ½ fX1 ðRc ; Re ; . . .Þ n3 ðVBE ; VBC Þ þ n4 ðVBE Þ X1 ðRc ; Re ; . . .Þ
dt Cbe
þ X2 ðRc ; Re ; . . .Þ n2 ðVBE ; VBC Þ X3 ðRc ; Re ; . . .Þ n1 ðVBC Þ
VA2 VBE VBC Vcc 1
X1 ðRc ; Re ; . . .Þ þ g n ðVBE ; VBC Þ
Rload C1 ðRc ; Re ; . . .Þ Cbe 3
dVA2 Cjx þ Cbc

¼½ fX1 ðRc ; Re ; . . .Þ n3 ðVBE ; VBC Þ þ n4 ðVBE Þ X1 ðRc ; Re ; . . .Þ
dt Cbe
V A2
þ X2 ðRc ; Re ; . . .Þ n2 ðVBE ; VBC Þ X3 ðRc ; Re ; . . .Þ n1 ðVBC Þ X1 ðRc ; Re ; . . .Þ
Rload
VBE VBC Vcc 1 1
þ gþ n ðVBE ; VBC Þ V A2
C1 ðRc ; Re ; . . .Þ Cbe 3 Rload Cout
We define three functions:
w1 ðVBE ; VBC ; VA2 ; . . .Þ ¼ X1 ðRc ; Re ; . . .Þ n3 ðVBE ; VBC Þ

þ n4 ðVBE Þ X1 ðRc ; Re ; . . .Þ þ X2 ðRc ; Re ; . . .Þ n2 ðVBE ; VBC Þ
VA2
X3 ðRc ; Re ; . . .Þ n1 ðVBC Þ X1 ðRc ; Re ; . . .Þ
Rload
VBE VBC Vcc
þ
C1 ðRc ; Re ; . . .Þ
Cjx þ Cbc
w2 ðVBE ; VBC ; VA2 ; . . .Þ ¼ ½ fX1 ðRc ; Re ; . . .Þ n3 ðVBE ; VBC Þ
Cbe
VA2
Rload
VBE VBC Vcc 1
þ g n ðVBE ; VBC Þ
C1 ðRc ; Re ; . . .Þ Cbe 3
Cjx þ Cbc
w3 ðVBE ; VBC ; VA2 ; . . .Þ ¼ ½ fX1 ðRc ; Re ; . . .Þ n3 ðVBE ; VBC Þ
Cbe
VA2
Rload
VBE VBC Vcc 1 1
þ gþ n ðVBE ; VBC Þ VA 2
C1 ðRc ; Re ; . . .Þ Cbe 3 Rload Cout
Our system differential equations: dVBC

dt ¼ w1 ðVBE ; VBC ; VA2 ; . . .Þ
dVBE dVA2
¼ w2 ðVBE ; VBC ; VA2 ; . . .Þ; ¼ w3 ðVBE ; VBC ; VA2 ; . . .Þ
dt dt
dVA2
To find system fixed points: dVBC
dt ¼ 0; dVBE
dt ¼ 0; dt ¼0

w1 ðVBE ; VBC ; VA2 ; . . .Þ ¼ 0; w2 ðVBE

; VBC ; VA2 ; . . .Þ ¼ 0; w3 ðVBE

; VBC ; VA2 ; . . .Þ ¼ 0
Fixed point and linearization: We approximate our system phase portrait near a

fixed point VBE ; VBC ; VA2 by corresponding it as a linear system.
We consider the system dVdtBC ¼ w1 ðVBE ; VBC ; VA2 Þ; dVdtBE ¼ w2 ðVBE ; VBC ; VA2 Þ
dVA2
¼ w3 ðVBE ; VBC ; VA2 Þ and suppose that ðVBE
dt ; VBC ; VA2 Þ is a fixed point, i.e.,

w1 ðVBE ; VBC ; VA2 Þ ¼ 0; w2 ðVBE ; VBC ; VA2 Þ ¼ 0; w3 ðVBE ; VBC ; VA2 Þ ¼ 0. Let u ¼

VBC VBC v ¼ VBE VBE ; w ¼ VA2 VA2 denote the components of a small dis-
turbance from the fixed point. To see whether the disturbance grows or decays, we
dVA2
dt ¼ dt ; dt ¼ dt ; dt ¼ dt
dVBC dv dVBE dw
need to derive differential equations for u, v, and w. du

Since VBE ; VBC ; VA2 are constants. By substitution:
du dVBC
¼ ¼ w1 ðv þ VBE ; u þ VBC ; w þ VA2 Þ ¼ w1 ðVBE

; VBC ; VA2 Þ
dt dt
@w1 @w1 @w1
þu þv þw þ Oðu2 ; v2 ; w2 ; uvw; . . .Þ
@VBC @VBE @VA2
@w1 @w1 @w1

Since w1 ðVBE ; VBC ; VA2 Þ ¼ 0 then du
dt ¼ dVdtBC ¼ u @V BC
þ v @V BE
þ w @V A
þ
2
Oðu2 ; v2 ; w2 ; . . .Þ.

The partial derivatives are to be evaluated at the fixed point ðVBE ; VBC ; VA2 Þ and
they are numbers and not functions. Also the shorthand notation Oðu2 ; v2 ; w2 ; . . .Þ:
Denotes quadratic terms in u, v, and w and it extremely small. Similarly we find
@w2 @w2 @w3
dv
dt ¼ dt ¼ u @VBC þ v @VBE þ w @VA þ Oðu ; v ; w ; . . .Þ and the expression
dVBE 2 2 2
2
dVA2 @w3 @w3 @w3
dw
dt ¼ dt ¼ u @VBC
þ v @V BE
þ w @VA
þ Oðu2 ; v2 ; w2 ; . . .Þ. We denote (u, v, w) a
2
disturbance [2–4].
0 1
du 0 @w 1 0 1
B dt C 1 @w1 @w1
B C @V @VBE @VA2 u
B dv C B @wBC2 @w2 @w2 C B C
B C B C @ v A þ quadratic term
B dt C ¼ @ @VBC @VBE @VA2 A
B C @w3 @w3 @w3
@ dw A @VBC @VBE @VA2
w
dt
0 @w1 @w1 @w1 1

@V @VBE @VA2
B @wBC2 @w2 @w2 C
The matrix A ¼ B C is called the Jacobian matrix at
@ @VBC @VBE @VA2 A
@w3 @w3 @w3
@VBC @VBE @VA2 ;V ;V Þ
ðVBE BC A 2

the fixed point ðVBE ; VBC ; VA2 Þ.
The quadratic terms are tiny and we neglect them
altogether. We obtain the linearized system.
0 1
du 0 @w 1
B dt C 1 @w1 @w1
B C @VBC @VBE @VA2
B dv C B @w2 @w2 @w2 C
B C B C
B dt C ¼ @ @VBC @VBE @VA2 A
B C @w3 @w3 @w3
@ dw A @VBC @VBE @VA2
dt
@w1 @n ðVBE ; VBC Þ @n ðVBE ; VBC Þ

¼ X1 ðRc ; Re ; . . .Þ 3 þ X2 ðRc ; Re ; . . .Þ 2
@VBC @VBC @VBC
@n1 ðVBC Þ 1
X3 ðRc ; Re ; . . .Þ
@VBC C1 ðRc ; Re ; . . .Þ
@n1 ðVBC Þ q 1 qVBC C4 nqVkBCT @n1 ðVBC Þ

¼ Iss ½ enr kB T þ e cl B ;
@VBC kB T qb br nr ncl @VBC
@n3 ðVBE ; VBC Þ
¼
@VBC
@n2 ðVBE ; VBC Þ Iss q qVBC

¼ enr kB T ;
@VBC qb nr k B T
@n3 ðVBE ; VBC Þ q 1 qVBC C4 nqVkBCT
¼ Iss ½ enr kB T þ e cl B
@VBC kB T qb br nr ncl
@w1 @n ðVBE ; VBC Þ @n4 ðVBE Þ

¼ X1 ðRc ; Re ; . . .Þ 3 þ X1 ðRc ; Re ; . . .Þ
@VBE @VBE @VBE
@n ðVBE ; VBC Þ 1
þ X2 ðRc ; Re ; . . . 2 þ
@VBE C1 ðRc ; Re ; . . .Þ
@n3 ðVBE ; VBC Þ q 1 qVBE

C2 nqVkBET
¼ Iss ½ enf kB T þ e el B
@VBE k B T qb b f nf nel
@n4 ðVBE Þ q 1 qVBE

C2 nqVkBET @n3 ðVBE ; VBC Þ
¼ Iss ½ enf kB T þ e el B ;
@VBE kB T qb bf nf nel @VBE
@n4 ðVBE Þ
¼
@VBE
@n2 ðVBE ; VBC Þ Iss q qVBE

@w1 1
¼ enf kB T ; ¼ X1 ðRc ; Re ; . . .Þ
@VBE qb nf k B T @VA2 Rload
@w2 Cjx þ Cbc @n ðVBE ; VBC Þ @n ðVBE ; VBC Þ

¼ ½ fX1 ðRc ; Re ; . . .Þ 3 þ X2 ðRc ; Re ; . . .Þ 2
@VBC Cbe @VBC @VBC
@n1 ðVBC Þ 1 1 @n3 ðVBE ; VBC Þ
X3 ðRc ; Re ; . . .Þ g
@VBC C1 ðRc ; Re ; . . .Þ Cbe @VBC
@w2 Cjx þ Cbc @n ðVBE ; VBC Þ @n4 ðVBE Þ

¼ ½ fX1 ðRc ; Re ; . . .Þ 3 þ X1 ðRc ; Re ; . . .Þ
@VBE Cbe @VBE @VBE
@n ðVBE ; VBC Þ 1 1 @n ðVBE ; VBC Þ
þ X2 ðRc ; Re ; . . .Þ 2 þ g @VBE 3
@VBE C1 ðRc ; Re ; . . .Þ Cbe @VBE
@w2 Cjx þ Cbc 1

¼½ X1 ðRc ; Re ; . . .Þ
@VA2 Cbe Rload
@w3 Cjx þ Cbc @n ðVBE ; VBC Þ @n ðVBE ; VBC Þ

¼½ fX1 ðRc ; Re ; . . .Þ 3 þ X2 ðRc ; Re ; . . .Þ 2
@VBC Cbe @VBC @VBC
@n1 ðVBC Þ 1 1 @n3 ðVBE ; VBC Þ
X3 ðRc ; Re ; . . .Þ gþ
@VBC C1 ðRc ; Re ; . . .Þ Cbe @VBC
@w3 Cjx þ Cbc @n ðVBE ; VBC Þ @n4 ðVBE Þ

¼½ fX1 ðRc ; Re ; . . .Þ 3 þ X1 ðRc ; Re ; . . .Þ
@VBE Cbe @VBE @VBE
@n ðVBE ; VBC Þ 1 1 @n3 ðVBE ; VBC Þ
þ X2 ðRc ; Re ; . . .Þ 2 þ gþ
@VBE C1 ðRc ; Re ; . . .Þ Cbe @VBE
@w3 Cjx þ Cbc 1 1

¼ ½ X1 ðRc ; Re ; . . .Þ
@VA2 Cbe Rload Rload Cout
We already found matrix A. The eigenvalues of a matrix A are given by the

characteristic equation detðA k IÞ ¼ 0, where I is the identity matrix 3 × 3.
0 1
@w1 @w1 @w1 0 1
@V
B @wBC2
@VBE @VA2
C k 0 0
B @w2 @w2 C
AkI ¼ @ @VBC @VBE @VA2 A @ 0 k 0A
@w3 @w3 @w3 0 0 k
@VBC @VBE @VA2 ;V ;V Þ
ðVBE BC A 2
0 @w1 @w1 @w1 1

@VBC k @VBE @VA2
B @w2 @w2 @w2 C
B C
AkI ¼ @ @VBC @VBE k @VA2 A
@w3 @w3 @w3
@VBC @VBE @VA2 k ;V ;V Þ
ðVBE BC A 2
0 @w2 @w2
1
@w1 @VBE k @VA2
detðA k IÞ ¼ ð kÞðV ;V ;V Þ det@ A
@VBC BE BC A2 @w3 @w3
k
@VBE @VA2 ;V ;V Þ
ðVBE BC A
0 @w2 @w2
1 2
@w1 @VBC @VA2

ð Þ det@ A
@VBE ðVBE ;VBC ;VA2 Þ @w3 @w3
k
@VBC @VA2 ;V ;V Þ
ðVBE BC A
!
2
@w2 @w2
@w1 @VBC @VBE k
þð Þ det
@VA2 ðVBE ;VBC ;VA2 Þ @w3 @w3
@VBC @VBE ;V ;V Þ
ðVBE BC A 2
@w1 @w @w @w3 @w2

detðA k IÞ ¼ ð kÞðV ;V ;V Þ ½ð 2 kÞ ð 3 kÞ
@VBC BE BC A2 @VBE @VA2 @VBE @VA2 ðVBE ;VBC ;VA2 Þ
@w @w @w @w3 @w2
ð 1 ÞðV ;V ;V Þ ½ 2 ð 3 kÞ
@VBE BE BC A2 @VBC @VA2 @VBC @VA2 ðVBE ;VBC ;VA2 Þ
@w @w @w3 @w3 @w
þ ð 1 ÞðV ;V ;V Þ ½ 2 ð 2 kÞðV ;V ;V Þ
@VA2 BE BC A2 @VBC @VBE @VBC @VBE BE BC A2
@w1 @w @w3 @w3 @w2

detðA k IÞ ¼ ð kÞðV ;V ;V Þ ½ð 2 Þ
@VBC BE BC A2 @VBE @VA2 @VBE @VA2
@w @w3 @w
ð 2 þ Þ k þ k2 ðV ;V ;V Þ ð 1 ÞðV ;V ;V Þ
@VBE @VA2 BE BC A2 @VBE BE BC A2
@w @w3 @w3 @w2 @w2
½ð 2 Þ kðV ;V ;V Þ
@VBC @VA2 @VBC @VA2 @VBC BE BC A2
@w1 @w2 @w3 @w3 @w2 @w3

þð Þ ½ð Þþ kðV ;V ;V Þ
@VA2 ðVBE ;VBC ;VA2 Þ @VBC @VBE @VBC @VBE @VBC BE BC A2
@w1 @w @w3 @w3 @w2 @w1 @w @w3

detðA k IÞ ¼ ½ ð 2 Þ ð 2 þ Þk
@VBC @VBE @VA2 @VBE @VA2 @VBC @VBE @VA2
@w1 @w @w3 @w3 @w2
þ k2 ð 2 Þk
@VBC @VBE @VA2 @VBE @VA2
@w @w3 @w @w @w3
þð 2 þ Þ k2 k3 ðV ;V ;V Þ ½ð 1 Þ ð 2
@VBE @VA2 BE BC A2 @VBE @VBC @VA2
@w3 @w2 @w @w2
Þ þ ½ð 1 Þ k
@VBC @VA2 ðVBE ;VBC ;VA2 Þ @VBE @VBC ðVBE ;VBC ;VA2 Þ
@w @w @w3 @w3 @w2 @w @w3
þ ½ð 1 Þ ð 2 Þþð 1Þ kðV ;V ;V Þ
@VA2 @VBC @VBE @VBC @VBE @VA2 @VBC BE BC A2
@w1 @w @w3
detðA k IÞ ¼ k3 þ ½ þð 2 þ Þ k2
@VBC @VBE @VA2 ðVBE ;VBC ;VA2 Þ
@w @w3 @w1 @w @w3
þ ½ð 1 Þ ð 2 þ Þ
@VA2 @VBC @VBC @VBE @VA2
@w @w3 @w3 @w2 @w @w2
ð 2 Þþð 1 Þ k
@VBE @VA2 @VBE @VA2 @VBE @VBC ðVBE ;VBC ;VA2 Þ
@w @w @w3 @w3 @w2
þ½ 1 ð 2 Þ
@VBC @VBE @VA2 @VBE @VA2
@w @w @w3 @w3 @w2
ð 1Þð 2 Þ
@VBE @VBC @VA2 @VBC @VA2
@w @w @w3 @w3 @w2
þð 1Þ ð 2 Þ
@VA2 @VBC @VBE @VBC @VBE ðVBE ;VBC ;VA2 Þ
The eigenvalues of a matrix A are given by the characteristic equation

P
detðA k IÞ ¼ 0; detðA k IÞ ¼ 3k¼0 Nk kk ¼ 0.
@w1 @w @w3
N3 ¼ 1; N2 ¼ ½ þð 2 þ Þ
@VBC @VBE @VA2 ðVBE ;VBC ;VA2 Þ
@w @w3 @w1 @w @w3
N1 ¼ ½ð 1 Þ ð 2 þ Þ
@VA2 @VBC @VBC @VBE @VA2
@w @w3 @w3 @w2 @w @w2
ð 2 Þþð 1 Þ
@VBE @VA2 @VBE @VA2 @VBE @VBC ðVBE ;VBC ;VA2 Þ
@w1 @w @w3 @w3 @w2 @w @w @w3 @w3 @w2

N0 ¼ ½ ð 2 Þð 1Þð 2 Þ
@VBC @VBE @VA2 @VBE @VA2 @VBE @VBC @VA2 @VBC @VA2
@w @w @w3 @w3 @w2
þð 1Þ ð 2 Þ
@VA2 @VBC @VBE @VBC @VBE ðVBE ;VBC ;VA2 Þ
We get three eigenvalues (λ1, λ2, λ3) for our system and need to classify them. If

k1 ; k2 ; k3 2 R then our fixed point VBE ; VBC ; VA2 is classify in the below Table 5.2
Table 5.2 Single ended class B amplifier system eigenvalues and stability classification

System eigenvalues Stability classification VBE ; VBC ; VA2
λ1 > 0, λ2 > 0, λ3 > 0 Unstable node
λ1 > 0, λ2 > 0, λ3 < 0 Saddle point
λ1 > 0, λ2 < 0, λ3 < 0 Saddle point
λ1 < 0, λ2 < 0, λ3 < 0 Stable node
If k1 ; k2 2 C; k1 ¼ g1 þ i g2 ; k2 ¼ g1 i g2 ; g1 ; g2 2 R; k3 2 R then our fixed

point VBE ; VBC ; VA2 is classify in the below table

System eigenvalues Stability classification VBE ; VBC ; VA2
g1 [ 0; k3 [ 0 Unstable spiral node
g1 [ 0; k3 \0 Unstable spiral saddle
g1 \0; k3 [ 0 Unstable spiral saddle
g1 \0; k3 \0 Stable spiral node
If k1 ; k2 \0; k3 ¼ 0 or k1 ; k3 \0; k2 ¼ 0 or k2 ; k3 \0; k1 ¼ 0 then we get

attracting line. If k1 ; k2 [ 0; k3 ¼ 0 or k1 ; k3 [ 0; k2 ¼ 0 or k2 ; k3 [ 0; k1 ¼ 0 then
we get repelling line.
5.4 Wideband Low Noise Amplifier (LNA) with Negative

Feedback Circuit Stability Analysis Under Circuit’s
When we want to amplify a very low power signal, we use Low Noise Amplifier
(LNA). It is done without degrading its signal to noise ratio (SNR). LNA device is a
crucial element in every RF receiver system and it amplifies the signal that comes
from the antenna. Regular amplifier will increase the power of both the signal and the
noise which come from the antenna and present at the amplifier’s input. Additionally
amplifiers are not ideal and they add noise to the input signal. Low Noise Amplifiers
(LNAs) are designed to minimize the additional noise. The target is to minimize the
additional noise by considering tradeoffs that include impedance matching. choosing
the amplifier technology, and selecting low-noise biasing conditions. Low-noise
amplifiers are found in many radio communications systems, medical instruments,
and electronic equipments. The conventional LNA operates on a single band, while
wideband LNA operate typically from 100 MHz to 1GHz and it hard to design. It is a
challenge to design broadband amplifier with the best performances. One architecture
is the combination of several narrows band LNA circuit into a single wideband LNA
circuit. LNA feedback technique is proposed to simultaneously achieve improvement
in bandwidth and on its gain, noise figure and return loss. The negative feedback
technique can be used in wideband amplifier to provide a flat gain response and to
reduce the input and output VSWR. It controls the amplifier performance due to
technical specifications variation from transistor to transistor and in band stability is
also improved by employing negative feedback. The LNA is most important block in
any receiving system because the receiving system sensitivity is generally determined
by its gain and noise figure. LNAs figures of merit are reduced Noise Figure (NF),
moderated gain, good input/output impedance matching, low power consumption,
isolation between input and output, acceptable linearity (low distortion), and stability.
There are many ways to design LNAs. It can be single ended or differential, single
stage, multistage, depending on type of engineering application and applications.
Typical single end LNA system is a two stage single ended LNA. The first stage
cascode amplifier is chosen for its simple input matching, its higher gain compared to
an inductively degenerated common emitter amplifier, and its high reverse isolation
and higher stability compared to a common base counterpart. The second stage
consists of a common emitter cascode amplifier without emitter degeneration for
higher gain. Differential LNA is composed of two stages. The first stage is a dif-
ferential cascode amplifier using LC impedance peaking network as load. This load
impedance can be made very large across the desired frequency band in order to force
the output current to flow into the following stage. In practice the series resistance in
the transmission line inductor will limit the impedance peaking effect. The second
stage is a conventional emitter coupled differential amplifier with high common mode
rejection. It is used to amplify the desired signal and compress the common mode
signal. The two stages are connected by a coupling capacitor C. The single ended
architecture has one disadvantage that it is very sensitive to parasitic ground
inductance. A differential LNA can beneficial while the noise figure is higher than
single ended design. We get higher gain by using multi stage LNA but the problem is
that is difficult to maintain stability than single stage LNA. The selection of design
option depends on type of application and specific design targets. The wideband
LNA is required to be in single stage, low power consumption, and minimum
components. The most important design considerations in a LNA design are stability,
noise, power gain, bandwidth, and DC requirements. The DC biasing circuit is used
to bias the selected transistor and the input and output matching network is important
for maximum power transfer in the circuit. LNAs operate in class A mode, charac-
terized by a bias point at the center of maximum voltage and current of the bias
supply for the transistor. The biasing point for the LNA should have high gain, low
noise figure, linear, good input and output matching and stable at the lowest current
drain from the supply. In designing LNAs, stability of the circuit is important pa-
rameter. This stability characteristic means that the device does not oscillate over a
range of frequencies with any combination of source and load impedance. The next
figure describes the block diagram of LNA [121, 122] (Fig. 5.7).
BJT technology is selected to design the wideband low noise amplifier due to the
higher gain at low power consumption, with reasonable low noise figure.
The RC feedback is the one of the most popular techniques to be used in
amplifiers circuit for its wideband input match and good linearity. The schematic of
the LNA is shown in the below figure. The transistor is self biased with the biasing
resistor of R1 and R2 and designed to low power product application. (Fig. 5.8).
The LNA design has implemented RLC feedback (L1, R3, C1) in order to lower
the gain at the lower frequencies and hence improve the stability of the circuit. We
need to tune the RLC feedback banch in order to meet the design specifications.
The LNA design also employs output resistive loading in stabilizing the circuit. The
initial output resistor value R4 is set to less than 50 Ω because high output resistor
value may result in huge decrease of gain and P1 dB point. L2 and L3 in the circuit
acts as RF choke which separate RF and DC path in the circuit. The LNA is
matched using lumped element as it is simple and compact (L matching network).
Typical LC matching network include the use of capacitors and inductors in either
series or shunt configuration. Circuit stability analysis is done by considering BJT
Small Signal (SS) equivalent circuit model. We consider “AC ground“ in the
circuit. Since the voltage at this terminal is held constant at Vcc, there is no time
5.4 Wideband Low Noise Amplifier (LNA) with Negative … 575
Fig. 5.7 Block diagram of LNA
Fig. 5.8 Wideband Low Noise Amplifier (LNA) with input and output matching network
schematics
variation of the voltage. Consequently. We can set this terminal to be an “AC

ground” in the small signal circuit. For AC ground, we “kill” the DC sources at that
terminal (short circuit voltage sources and open circuit current sources). Input and
output matching networks are L-type. Under AC and Small Signal (SS) conditions,
the BJT can be replaced with linear Hybrid Pi model. Let’s verify that this circuit
incorporates all the necessary Small Signal (SS) characteristics of BJT: ib ¼
vbe =rp ; ic ¼ gm vbe ; ib þ ic ¼ ie [24–26] (Fig. 5.9, Table 5.3).
Fig. 5.9 Wideband Low Noise Amplifier (LNA) with SS hybrid PI model and matching networks
Table 5.3 Wideband Low Noise Amplifier (LNA) with SS hybrid PI model and matching
networks KCL@ node and expressions
KCL @ node Expression
A1 IRs ¼ ICin þ ILin
A2 ILin ¼ Irp þ IL1 þ IL2
A3 IL1 ¼ IR3
A4 IR3 ¼ IC1
A5 IC1 þ IL3 ¼ IR4 þ gm vBE
A6 IL2 ¼ IR2
A7 IR2 ¼ IR1 þ IL3
A8 IR4 ¼ ICout þ ILout
A9 ILout ¼ IRload
Vs ðtÞ VA1 dVA1 dIL VA

IRs ¼ ; ICin ¼ Cin ; VLin ¼ VA1 VA2 ¼ Lin in ; Irp ¼ 2
Rs dt dt rp
dIL1 VA3 VA4 d dVC1
VL1 ¼ VA2 VA3 ¼ L1 ; IR3 ¼ ; IC1 ¼ C1 ðVA4 VA5 Þ ¼ C1
dt R3 dt dt
dIL2 VA6 VA7 VA7 dIL
VL2 ¼ VA2 VA6 ¼ L2 ; IR2 ¼ ; IR1 ¼ ; VL3 ¼ VA7 VA5 ¼ L3 3
dt R2 R1 dt
VA5 VA8 dVA8 dIL
IL1 ¼ IR3 ¼ IC1 ; IL2 ¼ IR2 ; IR4 ¼ ; ICout ¼ Cout ; VLout ¼ VA8 VA9 ¼ Lout out
R4 dt dt
Vout VA9
IRload ¼ ¼ ; IL ¼ IRload
Rload Rload out
Vs ðtÞ VA1
IRs ¼ ) IRs Rs ¼ Vs ðtÞ VA1 ) VA1 ¼ Vs ðtÞ IRs Rs
Rs
d dVs ðtÞ dIRs dIL
ICin ¼ Cin ðVs ðtÞ IRs Rs Þ ¼ Cin ð Rs Þ; VA1 VA2 ¼ Lin in
dt dt dt dt
dILin dILin VA2
VA2 ¼ VA1 Lin ¼ Vs ðtÞ IRs Rs Lin ; I rp ¼ ) V A 2 ¼ I rp r p
dt dt rp
dIL dIL dIL
Irp rp ¼ Vs ðtÞ IRs Rs Lin in ; VA2 VA3 ¼ L1 1 ) VA3 ¼ VA2 L1 1
dt dt dt
dIL1 VA 3 VA 4
VA3 ¼ Irp rp L1 ; IR 3 ¼
dt R3
) IR3 R3 ¼ VA3 VA4 ) VA4 ¼ VA3 IR3 R3
dIL d
VA4 ¼ Irp rp L1 1 IR3 R3 ; IC1 ¼ C1 ðVA4 VA5 Þ
dt dt
d dIL
¼ C1 ðIrp rp L1 1 IR3 R3 VA5 Þ
dt dt
dIrp d 2 IL dIR dVA5 dIL

IC1 ¼ C1 ð r p L1 2 1 3 R 3 Þ; VA2 VA6 ¼ L2 2 ) VA6
dt dt dt dt dt
dIL
¼ VA2 L2 2
dt
dIL2 VA VA7
VA6 ¼ Irp rp L2 ; IR 2 ¼ 6
dt R2
dIL2
) IR2 R2 ¼ VA6 VA7 ¼ Irp rp L2 IR1 R1
dt
VA7
IR1 ¼ ) VA7 ¼ IR1 R1 ; IR2 R2
R1
dIL dIL
¼ Irp rp L2 2 IR1 R1 ; VA7 VA5 ¼ L3 3
dt dt
dIL3 dIr d 2 IL dIR dVA7 d 2 IL
VA 5 ¼ VA 7 L 3 ; IC1 ¼ C1 ð p rp L1 2 1 3 R3 þ L3 2 3 Þ
dt dt dt dt dt dt
dIrp d 2 IL1 dIR3 dIR1 d 2 IL3
VA 7 ¼ IR1 R1 ) IC1 ¼ C1 ð rp L1 2 R3 R1 þ L3 2 Þ
dt dt dt dt dt
VA5 VA8 dVA8
IR 4 ¼ ) IR4 R4 ¼ VA5 VA8 ) VA8 ¼ VA5 IR4 R4 ; ICout ¼ Cout
R4 dt
dVA5 dIR4
¼ Cout ð R4 Þ
dt dt
dIL3 dIR d 2 IL dIR

VA5 ¼ IR1 R1 L3 ) ICout ¼ Cout ð 1 R1 L3 2 3 4 R4 Þ
dt dt dt dt
dILout dILout VA 9
VA8 VA9 ¼ Lout ) VA8 ¼ VA9 þ Lout ; IRload ¼ ) VA9 ¼ IRload Rload
dt dt Rload
dIL dVA8 dIR d 2 ILout
VA8 ¼ IRload Rload þ Lout out ; ICout ¼ Cout ¼ Cout ð load Rload þ Lout Þ
dt dt dt dt2
dVA8 dIL d 2 ILout
IRload ¼ ILout ) ICout ¼ Cout ¼ Cout ð out Rload þ Lout Þ
dt dt dt2
Additional equations: IL1 ¼ IR3 ¼ IC1 ; IL2 ¼ IR2 :

We can summary our wideband LNA with negative feedback circuit differential
equation:
[I] ICin ¼ Cin ðdVdts ðtÞ dIdtRs Rs Þ, [II] Irp rp ¼ Vs ðtÞ IRs Rs Lin dtLin :
dI
dIR2
[III] IR2 R2 ¼ Irp rp L2 dt IR1 R1
dIrp d2 I dI dI d2 I
[IV] IC1 ¼ C1 ð dt rp L1 dtC2 1 dtC1 R3 dtR1 R1 þ L3 dt2L3 Þ
dI d2 I dI dI
[V] ICout ¼ Cout ð dtR1 R1 L3 dt2L3 dtR4 R4 Þ, [VI] ICout ¼ Cout ð dtLout Rload þ
d 2 ILout
Lout dt2 Þ
d 2 IL3 dIR1 dIR4
[IV] + [V] L3 dt2 ¼ dt R1 dt R4 ICout C1out
dIrp d 2 IC dIC1 dIR 1

I C1 ¼ C 1 ð r p L1 2 1 R3 4 R4 ICout Þ
dt dt dt dt Cout
IRs ¼ ICin þ ILin ) ICin ¼ IRs ILin ; IC1 þ IR2 IR1 ¼ IR4 þ gm Irp rp
IR1 ¼ IC1 þ IR2 gm Irp rp IR4 ; IR4 ¼ ICout þ ILout ) ICout ¼ IR4 ILout

[I] IRs ILin ¼ Cin ðdVdts ðtÞ dIdtRs Rs Þ, [II] Irp rp ¼ Vs ðtÞ IRs Rs Lin dtLin :
dI
dIR2
[III] IR2 R2 ¼ Irp rp L2 dt ½IC1 þ IR2 gm Irp rp IR4 R1
d 2 IC1 dIC1 dIR4
[IV] + [V] IC1 ¼ C1 ðdIdtrp rp L1 dt2 dt R3 dt R4 ½IR4 ILout C1out Þ
dI d2 I
[VI] IR4 ILout ¼ Cout ð dtLout
Rload þ Lout dtL2out Þ
dI d2 I dI dI
We define new variables: Y1 ¼ dIdtrp ; Y2 ¼ dtC1 ; dYdt2 ¼ dtC2 1 ; Y3 ¼ dtR4 ; Y4 ¼ dtLout
[I] IRs ILin ¼ Cin ðdVdts ðtÞ dIdtRs Rs Þ, [II] Irp rp ¼ Vs ðtÞ IRs Rs Lin dtLin :
dI
dI
[III] IR2 R2 ¼ Irp rp L2 dtR2 ½IC1 þ IR2 gm Irp rp IR4 R1
[IV] + [V] IC1 ¼ C1 ðY1 rp L1 dYdt2 Y2 R3 Y3 R4 IR4 C1out þ ILout C1out Þ
[VI] IR4 ILout ¼ Cout ðY4 Rload þ Lout dYdt4 Þ
dIrp dIC dIR dIL dIR

¼ Y1 ; 1 ¼ Y2 ; 4 ¼ Y3 ; out ¼ Y4 ; s
dt dt dt dt dt
1 dVs ðtÞ 1
¼ ðIRs ILin Þ
Rs dt Cin Rs
dILin 1 Rs rp dY4 1 Rload

¼ Vs ðtÞ IRs Ir p ; ¼ ðIR4 ILout Þ Y4
dt Lin Lin Lin dt Cout Lout Lout
dIR2 R2 rp R1
¼ IR2 þ Irp ½IC1 þ IR2 gm Irp rp IR4
dt L2 L2 L2
dY2 1 rp R3 R4 1 1
¼ IC1 þ Y1 Y2 Y3 IR4 þ ILout
dt C1 L1 L1 L1 L1 Cout L1 Cout L1
dIrp dIC1 dIR4 dILout

At fixed points (equilibrium points): dt ¼ 0; dt ¼ 0; dt ¼ 0; dt ¼ 0;
dIRs
dt ¼0
dILin dIR dY2

¼ 0; 2 ¼ 0; ¼ 0; Y1 ¼ 0; Y2 ¼ 0; Y3 ¼ 0; Y4 ¼ 0; IRs ILin
dt dt dt
dVs ðtÞ
¼ Cin
dt
IRs Rs þ Irp rp ¼ Vs ðtÞ; IR4 ILout

1 1 1
¼ Y4 Rload Cout ; IC 1 IR4 þ ILout ¼0
C1 Cout Cout
IR2 R2 þ Irp rp ð1 þ gm R1 Þ IC 1 R1 IR2 R1 þ IR4 R1 ¼ 0
If dVdts ðtÞ ! e then IRs ¼ ILin :

librium points of the wideband Low Noise Amplifier (LNA) with negative feedback
circuit equivalent circuit consists in adding to coordinate ½IRs ; ILin ; Irp ; IR2 ; IC1 ; IR4 ; ILout ;
Y1 ; Y2 ; Y3 ; Y4 arbitrarily small increments of exponentially form ½iRs ; iLin ; irp ; iR2 ; iC1 ;
iR4 ; iLout ; y1 ; y2 ; y3 ; y4 ekt and retaining the first order terms in IRs ; ILin ; Irp ; IR2 ; IC1 ; IR4 ;
ILout ; Y1 ; Y2 ; Y3 ; Y4 . The system of homogenous equations leads to a polynomial
characteristic equation in the eigenvalues. The polynomial characteristic equations
accept by set of the below circuit variables, circuit variables derivative and circuit
variables second order derivative with respect to time into wideband Low Noise
Amplifier (LNA) with negative feedback [2–4]. Our wideband Low Noise Amplifier
(LNA) with negative feedback equivalent circuit fixed values with arbitrarily small
increments of exponential form ½iL1 ; iRb ; iRload ; iCout ; iCin ; y1 ; y2 ; y3 ; y4 ekt are: j = 0
(first fixed point), j = 1 (second fixed point), j = 2 (third fixed point), etc.,
ðjÞ ðjÞ ðjÞ

Y1 ðtÞ ¼ Y1 þ y1 ekt ; Y2 ðtÞ ¼ Y2 þ y2 ekt ; Y3 ðtÞ ¼ Y3 þ y3 ekt ;
ðjÞ
Y4 ðtÞ ¼ Y4 þ y4 ekt
ðjÞ ðjÞ ðjÞ
IRs ðtÞ ¼ IRs þ iRs ekt ; ILin ðtÞ ¼ ILin þ iLin ekt ; Irp ðtÞ ¼ IrðjÞ
p
þ iRp ekt ; IR2 ðtÞ ¼ IR2 þ iR2 ekt
ðjÞ ðjÞ ðjÞ dIrp ðtÞ
IC1 ðtÞ ¼ IC1 þ iC1 ekt ; IR4 ðtÞ ¼ IR4 þ iR4 ekt ; ILout ðtÞ ¼ ILout þ iLout ekt ; ¼ iRp k ekt
dt
dILin ðtÞ dIR ðtÞ dY2 ðtÞ dY4 ðtÞ
¼ iLin k ekt ; 2 ¼ iR2 k ekt ; ¼ y2 k ekt ; ¼ y4 k ekt
dt dt dt dt
dIC1 ðtÞ dIR ðtÞ dIL ðtÞ dIr ðtÞ
¼ iC1 k ekt ; 4 ¼ iR4 k ekt ; out ¼ iLout k ekt ; s ¼ iRs k ekt
dt dt dt dt
ðjÞ ðjÞ ðjÞ

iRp k ekt ¼ Y1 þ y1 ekt ; Y1 ¼ 0 ) iRp k þ y1 ¼ 0; iC1 k ekt ¼ Y2 þ y2 ekt ;
ðjÞ
Y2 ¼ 0 ) iC1 k þ y2 ¼ 0
ðjÞ ðjÞ ðjÞ

iR4 k ekt ¼ Y3 þ y3 ekt ; Y3 ¼ 0 ) iR4 k þ y3 ¼ 0; iLout k ekt ¼ Y4 þ y4 ekt ;
ðjÞ
Y4 ¼ 0 ) iLout k þ y4 ¼ 0
1 dVs ðtÞ ðjÞ 1 ðjÞ 1

iRs k ekt ¼ ½IRs þ iRs ekt þ ½ILin þ iLin ekt
Rs dt Cin Rs Cin Rs
1 dVs ðtÞ ðjÞ 1 ðjÞ 1 1

iRs k ekt ¼ IR s þ ILin iRs ekt þ iLin
Rs dt Cin Rs Cin Rs Cin Rs
1
ekt
Cin Rs
ðjÞ ðjÞ
At fixed point: 1
Rs dVdts ðtÞ IRs Cin1Rs þ ILin Cin1Rs ¼ 0
1 1
iRs k iRs þ iLin ¼0
Cin Rs Cin Rs
1 ðjÞ Rs rp
iLin k ekt ¼ Vs ðtÞ ½IRs þ iRs ekt ½IrðjÞ
p
þ iRp ekt
Lin Lin Lin
1 ðjÞ Rs rp Rs kt rp kt

iLin k ekt ¼ Vs ðtÞ IRs IrðjÞ
p
iRs e iRp e
Lin Lin Lin Lin Lin
ðjÞ ðjÞ
At fixed point: L1in Vs ðtÞ IRs LRins Irp Lrpin ¼ 0; iLin k iRs LRins iRp Lrpin ¼ 0
ðjÞ ðjÞ 1 ðjÞ Rload

y4 k ekt ¼ ðIR4 þ iR4 ekt ILout iLout ekt Þ ½Y4 þ y4 ekt
Cout Lout Lout
ðjÞ ðjÞ 1 ðjÞ Rload

y4 k ekt ¼ ðIR4 ILout Þ Y4
Cout Lout Lout
1 Rload kt
þ ðiR4 ekt iLout ekt Þ y4 e
Cout Lout Lout
ðjÞ ðjÞ ðjÞ
At fixed point: ðIR4 ILout Þ Cout1Lout Y4 RLload
out
¼0
1 1 Rload
y4 k þ iR4 iLout y4 ¼0
Cout Lout Cout Lout Lout
ðjÞ R2 rp ðjÞ
iR2 k ekt ¼ ½IR2 þ iR2 ekt þ ½IrðjÞ
p
þ iRp ekt ½IC1 þ iC1 ekt
L2 L2
ðjÞ ðjÞ R1
þ IR2 þ iR2 ekt gm ðIrðjÞ
p
þ iRp ekt Þ rp IR4 iR4 ekt
L2
ðjÞ R2 rp ðjÞ ðjÞ ðjÞ R1 R2

iR2 k ekt ¼ IR2 þ IrðjÞ
p
½IC1 þ IR2 gm IrðjÞ
p
rp IR4 iR2 ekt
L2 L2 L2 L2
rp kt R1
þ iRp e ½iC1 ekt þ iR2 ekt gm iRp ekt rp iR4 ekt
L2 L2
ðjÞ ðjÞ ðjÞ ðjÞ ðjÞ ðjÞ

At fixed point: IR2 RL22 þ Irp Lrp2 ½IC1 þ IR2 gm Irp rp IR4 RL21 ¼ 0
R2 rp R1
iR2 k iR2 þ iRp ½iC1 þ iR2 gm iRp rp iR4 ¼0
L2 L2 L2
R2 R1 rp R1 R1 R1
iR2 k iR2 iR2 þ iRp þ iRp gm rp iC1 þ iR ¼ 0
L2 L2 L2 L2 L2 L2 4
1 X 2
1 R1 R1
iR2 k iR2 Rk þ iRp rp ð1 þ gm R1 Þ iC1 þ iR ¼ 0
L2 k¼1 L2 L2 L2 4
ðjÞ 1 ðjÞ rp ðjÞ R3

y2 k ekt ¼ ½IC1 þ iC1 ekt þ ½Y1 þ y1 ekt ½Y2 þ y2 ekt
C1 L1 L1 L1
ðjÞ kt R4 ðjÞ kt 1 ðjÞ kt 1
½Y3 þ y3 e ½IR4 þ iR4 e þ ½ILout þ iLout e
L1 Cout L1 Cout L1
ðjÞ 1 ðjÞ rp ðjÞ R3 ðjÞ R4

y2 k ekt ¼ IC1 þ Y1 Y2 Y3
C1 L1 L1 L1 L1
ðjÞ 1 ðjÞ 1 1
IR4 þ ILout i C1 ekt
Cout L1 Cout L1 C1 L1
rp R3 R4
þ y1 ekt y2 ekt y3 ekt
L1 L1 L1
1 1
iR4 ekt þ iLout ekt
Cout L1 Cout L1
At fixed point:
ðjÞ 1 ðjÞ rp ðjÞ R3 ðjÞ R4 ðjÞ 1 ðjÞ 1

IC1 þ Y1 Y2 Y3 IR4 þ ILout
C1 L1 L1 L1 L1 Cout L1 Cout L1
¼0
1 rp R3 R4
y2 k ekt ¼ iC1 ekt þ y1 ekt y2 ekt y3 ekt iR4
C1 L1 L1 L1 L1
1 kt 1 kt
e þ iLout e
Cout L1 Cout L1
1 rp R3 R4 1 1
y2 k iC1 þ y1 y2 y3 iR4 þ iLout
C1 L1 L1 L1 L1 Cout L1 Cout L1
¼0
We build our system matrices ðiRp ; iC1 ; iR4 ; iLout ; iRs ; iLin ; y4 ; iR2 ; y2 Þ and ðy1 ; y2 ; y3 Þ
iRp k þ y1 ¼ 0; iC1 k þ y2 ¼ 0; iR4 k þ y3 ¼ 0; iLout k þ y4 ¼ 0
1 1 Rs rp
iRs k iRs þ iLin ¼ 0; iLin k iRs iRp ¼0
Cin Rs Cin Rs Lin Lin
Rload 1 1
y4 k y4 þ iR4 iLout ¼0
Lout Cout Lout Cout Lout
1 X 2
1 R1 R1
iR2 k iR2 Rk þ iRp rp ð1 þ gm R1 Þ iC1 þ iR ¼ 0
L2 k¼1 L2 L2 L2 4
R3 1 rp R4 1 1
y2 k y2 iC1 þ y1 y3 iR4 þ iLout
L1 C1 L1 L1 L1 Cout L1 Cout L1
¼0
0 1 0 1
iRp m11 m12
B i C B m21 m22 C
B C1 C B C
B C
B iR4 C B m31 m32 C
B C B C
0 1 B C B m41 m42
C
i11 . . . i19 B Lout C B
i C !
B .. .. C B C B C
.. B iRs C þ B C
m51 m52 y1
@ . . . A B C B C ¼ 0; i11 ¼ k; i12 ¼
B C B C y3
i91 . . . i99 B iLin C B m61 m62 C
B C B C
B y4 C B m71 m72 C
B C B C
B C B C
@ iR A @
2
m81 m82 A
y2 m91 m92
¼ i19 ¼ 0; i21 ¼ 0; i22 ¼ k
i23 ¼ ¼ i28 ¼ 0; i29 ¼ 1; i31 ¼ i32 ¼ 0; i33 ¼ k; i34 ¼ ¼ i39 ¼ 0; i41 ¼ i42
¼ i43 ¼ 0; i44 ¼ k
1
i45 ¼ i46 ¼ 0; i47 ¼ 1; i48 ¼ i49 ¼ 0; i51 ¼ i54 ¼ 0; i55 ¼ k ; i56
Cin Rs
1
¼ ; i57 ¼ i58 ¼ i59 ¼ 0
Cin Rs
rp Rs
i61 ¼ ; i62 ¼ ¼ i64 ¼ 0; i65 ¼ ; i66 ¼ k; i67 ¼ ¼ i69 ¼ 0; i71
Lin Lin
¼ i72 ¼ 0
1 1 Rload
i73 ¼ ; i74 ¼ ; i75 ¼ i76 ¼ 0; i77 ¼ k ; i78 ¼ i79 ¼ 0
Cout Lout Cout Lout Lout
1 R1 R1
i81 ¼ rp ð1 þ gm R1 Þ; i82 ¼ ; i83 ¼ ; i84 ¼ i85 ¼ i86 ¼ i87 ¼ 0; i88
L2 L2 L2
1 X 2
¼ k Rk
L2 k¼1
1 1 1
i89 ¼ 0; i91 ¼ 0; i92 ¼ ; i93 ¼ ; i94 ¼ ; i95 ¼ i96 ¼ i97
C1 L1 Cout L1 Cout L1
¼0
R3
i98 ¼ 0; i99 ¼ k ; m11 ¼ 1; m12 ¼ 0; m21 ¼ 0; m22 ¼ 0; m31 ¼ 0; m32 ¼ 1
L1
m41 ¼ 0; m42 ¼ 0; m51 ¼ 0; m52 ¼ 0; m61 ¼ 0; m62 ¼ 0; m71 ¼ 0; m72 ¼ 0

rp R4
m81 ¼ 0; m82 ¼ 0; m91 ¼ ; m92 ¼
L1 L1
Assumption mlk ; l ¼ 1; . . .; 9; k ¼ 1; 2 elements are neglected compar to ilk ele-

ments l ¼ 1; . . .; 9; k ¼ 1; . . .; 9.
0 1 0 1
i11 ... i19 i11 . . . i19
B .. .. .. C B .. .. . C
ðA k IÞ ¼ @ . . . A; detðA k IÞ ¼ 0 ) det@ . . .. A ¼ 0
i91 ... i99 i91 . . . i99

criterion for when the equation det ðA k IÞ ¼ 0 has a zero in the left half plane,
without calculating the eigenvalues explicit. We use criterion of Routh-Hurwitz
[2–4].
BJT Small Signal (SS) equivalent circuit models:
In order to develop these BJT small signal models, there are two small signal
resistances that determine. These are rp the small signal, active mode input resis-
tance between the base and emitter, as “seen looking into the base” and re the small
signal, active mode output resistance between the base and emitter, “as looking into
the emitter”. These resistances are not the same because the transistor is not a
reciprocal device. The behavior of the BJT in the circuit changes if we interchange
the terminals. Determine rp : assuming the transistor in this circuit is operating in the
active mode, then we get BJT base current iB ¼ biC ¼ b1 ðIC þ VICT vbe Þ; ib ¼
f f
IC
bf VT vbe ¼ gbm vbe . The iC DC section is IC and the AC section is IC
VT vbe [33, 34]
f
(Fig. 5.10).
Fig. 5.10 BJT transistor

circuit with biasing voltages
Fig. 5.11 BJT transistor AC

small signal equivalent circuit
(version 1)
The AC small signal equivalent circuit is as follow: (Fig. 5.11).

Since the voltage at circuit terminal is held constant at Vcc, there is no time
variation of the voltage. Consequently, this terminal ca be an “AC ground” in the
small signal circuit. For AC grounds, we “kill” the DC sources at that terminal:
short circuit voltage sources an open current sources.
b
We get rp ¼ vibeb ¼ gmf ½X, this rp is the BJT active mode small signal input
resistance of the BJT between the base and the emitter as seen looking into the base
terminals. Determine re : We determine re following a similar procedure as for rp
but start with iE ¼ aiCf ¼ IaCf þ aicf ; ie ¼ aicf ¼ afIV
C
T
vbe ; IE ¼ IaCf ; ie ¼ VIET vbe ; iE ¼
IE þ ie . re is the BJT small signal resistance between the emitter and base seen
looking into the emitter. Mathematically, this is stated as re i ve
e
) ve ¼
a I a a
vbe ; re vibee ; re ¼ VIET ; gm ¼ VICT ¼ Vf T E ) VIET ¼ gmf ; re ¼ gmf g1m ½X.
It can be shown that rp ¼ ðbf þ 1Þ re ½X; rp 6¼ re . The active mode BJT is a non
reciprocal device (Fig. 5.12).
Fig. 5.12 BJT transistor AC

small signal (version 2)
There are two families of equivalent active mode BJT small signal circuit:
Hybrid Pi model and T model. Both are equally valid models but choosing one over
the other sometimes leads to simpler analysis of certain circuits.
Hybrid Pi model Version A: The circuit incorporates all of the necessary small
signal characteristics of the BJT ib ¼ vrbep ; ic ¼ gm vbe ; ib þ ic ¼ ie ; ie ¼ vrbee
(Fig. 5.13).
Hybrid Pi model Version B: The second equivalent circuit is constructed by
using the following notation: gm vbe ¼ gm ðib rp Þ ¼ gm rp ib ¼ bf ib
(Fig. 5.14).
T model: The hybrid Pi model is the most popular small signal model for the
BJT. The alternative is the T model, which is useful in certain situations.
T model Version A: (Fig. 5.15).
T model Version B: (Fig. 5.16).
The small signal model for PNP BJTs are identically the same as for NPN
transistors. There is no change in any polarities (voltage or current) for the PNP
models relative to the NPN models. These small signal models are identically the
same.
Fig. 5.13 Equivalent active

mode BJT small signal circuit
Hybrid Pi model (version A)

mode BJT small signal circuit
Hybrid Pi model (version B)

model BJT small signal
circuit, T model (version A)

model BJT small signal
circuit, T model (version B)
Exercises
1. We have class AB push-pull variation amplifier circuit. The active elements

used in this circuit (Q1, D3 (Opto-coupler), and Q2) are almost complementary
symmetric transistors. Q1 is NPN phototransistor and Q2 is PNP transistor.
Pre-biasing is given to the transistor Q2 and LED D3 using the network com-
prising of resistors R1, R2 and the biasing diodes D1 and D2. Resistor rp is a
parasitic resistance between diodes D1 and D2. We use in our circuit analysis
BJT transistor Ebers-Moll model. The coupling coefficient between LED D3 and
photo transistor Q1 is “k” ðIBQ1 ¼ k ID3 Þ. C1 and C2 are input DC decoupling
capacitors and input RF source Vs(t) with series resistance Rs. Hint: we use
analog optocoupler in our circuit.
1:1 Draw circuit class AB push-pull variation amplifier equivalent circuit and
write the related circuit differential equations (Use transistor Q1 and Q2
Ebers-Moll model).
1:2 Find circuit fixed points and discuss stability by parameters variation.
1:3 We short diode D1, How it influences circuit behavior and stability?
1:4 We define coupling function between LED D3 current
and photo transistor
PM k
Q1 base current IBQ1 ¼ k¼1 ID3 ak ; M 2 N . How parameters a1 ; a2 ;
. . .; aM influence on circuit behavior stability?
1:5 Diode D3 is disconnected, Find circuit differential equations and fixed
points. Discuss stability.
Exercises 589
2. We have class AB push-pull variation amplifier circuit. The active elements

used in this circuit are Q1, Q2, and D3, Q3 (Opto coupler). Q1 is NPN transistor
and Q2 is NPN photo transistor, Q3 is PNP transistor. Pre-biasing is given to the
transistor Q1, LED D3, and transistor Q3 using the network comprising of
resistors R1, R2, R3 and the biasing diodes D1 and D2. Resistor rp is the parasitic
resistance between diodes D1 and D2. We use in our analysis BJT transistor
Ebers-Moll model. The coupling coefficient between LED D3 and photo tran-
sistor Q2 is “k” parameter ðIBQ2 ¼ k ID3 Þ.C1 and C2 are input DC decoupling
capacitors and input RF source Vs(t) with series resistance Rs. Hint: we use
analog optocoupler in our circuit.
2:1 Draw circuit class AB push-pull variation amplifier equivalent circuit and
write the related circuit differential equations. (Use transistors Q1, Q2, and
Q3 Ebers-Moll model.)
2:2 Find circuit fixed points and discuss stability by circuit parameters variation.
2:3 Resistor rp → ε, How it influences on circuit behavior and stability?
2:4 We define
coupling function between diode D3 and
photo transistor Q2 base
PP Ql
current IBQ2 ¼ l¼1 ½bl k¼1 ID3 ak ; P 2 N . How parameters ak ; bl
k
influence on circuit behavior and stability?

2:5 We short diode D1, How it influence circuit behavior and stability? Find
circuit differential equations and fixed points.
3. We have power amplifier with two internal resonant circuits. The first resonant
circuit is connected to transistor Q1’s emitter (C1 and L1) and the second res-
onant circuit is connected to transistor Q1’s collector (C2 and L2). Resistor Rb is
transistor Q1’s base resistor and Cin, Cout are input and output blocking
capacitors. Input RF source is Vs(t) with series resistor Rs. Since the input RF
signal is a large signal, we use in our analysis the Ebers-Moll BJT model for
transistor Q1. The values of L1 and C1 or L2 and C2 are so selected that the
resonant circuit oscillates in the frequency of the input signal. Capacitors Cin,
Cout are input and output blocking capacitors. Assumption: dVdts ðtÞ ! e.
3:1 Write circuit equations and differential equations. Find fixed points and
discuss stability.
3:2 Inductor L2 is disconnected. How it influences our circuit behavior?
Find fixed points and circuit differential equations. Discuss stability.
3:3 Resistor Rb is changed by Γ1 multiplication factor ðRb ! Rb C1 ; C1 2 RÞ.
We have two cases: (a) 0\C1 \1(b) C1 [ 1. How the dynamical behavior
Exercises 591
of the circuit is changed for cases (a) and (b)? Discuss stability for both
cases.
3:4 Capacitor C1 is disconnected. How it influences our circuit behavior? Find
fixed points and circuit differential equations. Discuss stability.
3:5 Capacitor C1 and inductor L1 are changed according the following trans-
formation: C1 ! C1 C2 ; L1 ! L1 C22 ; C2 [ 1; C2 [ 0; C2 2 R. How the
dynamical behavior of the circuit is changed for different values of
parameter Γ2? Discuss stability.
4. We have power amplifier with two internal resonant circuits. The first resonant
circuit is connected between the Vcc and Q1’s emitter (C1 and L1) and the second
resonant circuit is connected between Q1’s emitter and ground (C2 and L2). Rb is
transistor Q1’s base resistor and Cin, Cout are input and output blocking
capacitors. Input RF source is Vs(t) with series resistance Rs. Since the input RF
signal is a large signal, we use in our analysis the Ebers-Moll BJT model of Q1
transistor. Assumption: dVdts ðtÞ ! e.
discuss stability.
4:2 Capacitor C2 is disconnected. Find fixed points and circuit differential
equations. Discuss stability.
4:3 Transistor Q1’s af ; ar parameters are changed by multiplication factors Γ1
and Γ2 respectively af ! af C1 ; ar ! ar C2 ; C1 ; C2 2 R. af : 0:95 !
0:99; 0:95
af C1
0:99; ar : 0 ! 0:95; 0\ar C2 \0:95; af [ ar .
How the dynamical behavior of the circuit is changed for different values of
Γ1 and Γ2 factors? Discuss stability for different values of Γ1 and Γ2 factors

C1 ; C2 [ 0.
4:4 Inductor L1 is disconnected. How it influences our circuit behavior? Find
fixed points and circuit differential equations. Discuss stability.
4:5 Capacitors C2 and inductor L2 are changed according to the following
pffiffiffiffiffiffi
transformations: C2 ! C2 C3 ; L1 ! L1 ð C3 þ C33 Þ; C3 [ 0; C3 2 R.
5. We have power amplifier with resonant circuit (C1 and L1) between transistor
Q1 emitter-collector, Q1’s emitter branch can be connected to resistor Re1 or
resistor Re2 and inductor Le (you choose it by S1 switch). Rb is the transistor Q1
base resistor and Cin, Cout are input and output blocking capacitors. Input RF
source is Vs(t) with series resistance Rs. Since the input RF signal is a large
signal we use in our analysis the Ebers-Moll BJT model for Q1. The values of
capacitors C1 and inductor L1 are selected that the resonant circuit is oscillated
in the frequency of the input signal. We have two cases in our circuit, case (a):
switch S1 at position (1), case (b): switch S2 at position (2). Assumption:
dVs ðtÞ
dt ! e
discuss stability [two cases, S1 in position (1) and (2)].
5:2 Resistor Rc is changed by factor Ω1 ðRc ! Rc X1 ; X1 [ 0; X1 2 RÞ. How
the dynamical behavior of the circuit is changed for different values of Ω1
parameter? [Two cases: S1 in position (1) and (2)].
Exercises 593
5:3 Transistor Q1’s af ; ar parameters are changed by multiplication Γ1 and Γ2

pffiffiffiffiffiffi
factors respectively af ! af C1 ; ar ! ar C2 ; C1 ; C2 2 R. af : 0:95 !
0:99; 0:95
af C1
0:99; ar : 0 ! 0:95; 0\ar C2 \0:95; af [ ar . How
the dynamical behavior of the circuit is changed for different values of Γ1
and Γ2 factors? Discuss stability for different values of Γ1 and Γ2 factors
C1 ; C2 [ 0:
5:4 Capacitors C1 and inductor L1 are changed according to the following
pffiffiffiffiffiffi pffiffiffiffiffiffi
transformation: C1 ! C1 C3 ; L1 ! L1 ð 3 C3 þ C23 Þ C3 [ 0; C3 2 R.
5:5 Inductor L1 is disconnected. How the dynamical behavior of the circuit is
changed? [Cases: (1) and (2)]. Discuss stability.
6. We have power amplifier system which constructed from two BJT NPN tran-
sistors Q1 and Q2, peripheral components. Capacitors Cin and Cout are input and
output blocking capacitors. Transistor Q2 is RF transistor, L2 is radio frequency
choke which isolated the RF form the DC source Vcc. DC voltage source Vcc is
between 2v to 48v. The bias input at junction A to transistor Q2 having the
desired DC and RF impedance characteristics which will allow linear amplifi-
cation of RF frequencies over a wide power range. The biasing circuit has two
portions: one functioning essentially only at low power level and another por-
tion with the one portion functioning at increasing to high power level with a
smooth and continuous transition between them. Diode D2 is a constant current
diode, resistor R3 (low power level portion). Additionally resistor R1, inductor
L1, and Ferrite bead F1. When we increase to high power level portion of the
biasing circuit, its done additionally to low power level portion. Q1 biasing
series circuit: resistor R2, diode D1, and resistor Rb1. The series circuit provides
bias to the Q1’s base. The emitter-collector of transistor Q1 is a bypass around
resistor R1 and at these two components provides variable impedance for giving
the varying bias needed, as between low power inputs and high power inputs.
The purpose of diode D2 is to provide a high impedance constant current
through resistor R1. The circuit of constant current is through diode D2 con-
nected to Vcc through resistor R3, ferrite bead F1, RF inductor L1, and resistor
R1. Diode D2 provides a voltage of about 0.5v at point A under DC static
condition. The voltage at point A is the base voltage Vb of the transistor Q1. The
transistor Q1 having turn ON base voltage of about 0.6v. The DC bias voltage
value 0.5v initially maintains the transistor Q1 in a non-conducting state and the
amplifier is, in effect, operating in a class B mode. Resistor R3 is a power
dissipation limiting resistor and has a typical value of 1.8 kΩ in order to provide
large impedance.
Biasing circuit of Q2 at high power level: Diode D1, resistor R2, and resistor Rb1
rectify the RF supplied from the input and develops DC bias from transistor Q1
which is proportional to the RF input drive level. At low RF input level, Q1 is
biased OFF and allowing resistor R1 to develop the bias for RF amplifier Q2 which
operating in a class B. When RF input is increased sufficiently, transistor Q1 begin
to saturate, shunting resistor R1 and changing the operating mode of amplifier Q2
from class B to class C. At high drive level conditions, the DC dynamic impedance
between the base-emitter of Q2 is very low and promoting maximum transistor gain
and efficiency. As the drive level is increased or decreased from an intermediate
level (V ± ΔV), the dynamic impedance changes in a nonlinear manner. At low
power input RF transistor Q2 operates at class B and at higher power levels tran-
sistor Q2 operates at class C.
Constant current diode: Constant current diode is an electric device that limits
current to a maximum specified value for device (CLD—Current Limiting Diode,
CRD—Current Regulating Diode). This diode consist of a n-channel JFET tran-
sistor with the gate shorted to the source, which functions like a two terminal
current limiter or current source (analog to voltage limiting Zener diode). It allows a
current through to rise to a certain value, and then level OFF at specific value. This
Exercises 595
diode keeps the current constant. This device keeps the current flowing through it
unchanged when the voltage changes.
VL—Limiting voltage: measured at IL, VL together with knee AC impedance

(Zk), indicates the knee characteristics of the device.
POV = Peak Operating Voltage: maximum voltage to be applied to device.
In operation the CLD regulates the amount of current that can flow over a
voltage range of about 1–100 V. The equivalent circuit of the CLD is a current
generator in series with a parallel combination of the dynamic impedance and the
junction capacitance. The shunt capacitance of Central’s CLD is about 4–10 pF
over the useful operating voltage range.
Ferrite bead circuit model: A ferrite bead is a passive device that suppresses
high frequency noise in electric circuit. It is a specific type of electronic choke.
Ferrite beads prevent interference in two directions, from a device or to a device.
The equivalent ferrite bead is as follow:
In circuit analysis we need to use for low power level input RF signal—BJT
transistor Small Signal (SS) model (Hybrid Pi) and for high power level input RF
signal—BJT transistor Gummel-Poon model.
6:1 Write circuit equations and differential equations for low power level and
high power level input RF signals. Find fixed points and discuss stability.
6:2 Resistor R1 is disconnected. How the dynamical behavior of the circuit is
changed? Find circuit differential equations and fixed points. Discuss
stability.
6:3 We replace constant current diode (D2) by regular diode. How the
dynamical behavior of the circuit is changed? Find circuit differential
equations and fixed points. Discuss stability.
6:4 We replace Ferrite bead F1 by regular RF choke (LF). How the dynamical
behavior of the circuit is changed? Find circuit differential equations and
fixed points. Discuss stability.
6:5 We replace diode D1 by D2 and diode D2 by D1. How the dynamical
behavior of the circuit is changed? Find circuit differential equations and
fixed points. Discuss stability.
7. We have wideband Low Noise Amplifier (LNA) with negative feedback net-
work. The feedback network consists of capacitors, resistors, and inductors as
describe in the below circuit. Input and output matching networks are Pi-type.
Input RF signal is Vs(t), Vs ðtÞ ¼ A0 þ nðtÞ; A0 nðtÞ.
Inductors L2, L3 in the circuit act as the RF choke which blocks the DC current
from entering the RF path. The LNA design also employs output resistive loading
in stabilizing the circuit. The initial output resistor value, R4 is set to less than
100 Ω (R4 < 100 Ω) because high output resistor value may result in huge decrease
of gain and P1 dB point. The BJT transistor (Q1) is self-biased with the biasing
resistors of R2 and R3.
Exercises 597
discuss stability.
7:2 Capacitor C2 is disconnected. How the dynamical behavior of the circuit is
changed? Discuss stability and find circuit differential equations and fixed
points.
7:3 We change the input and output matching networks from Pi-type to T-type.
How the dynamical behavior of the circuit is changed? Discuss stability.
7:4 Resistor R4 is changed by factor Γ ðR4 ! R4 C; C [ 0; C 2 R þ Þ. How
the dynamical behavior of the circuit is changed for different values of Γ
parameter? Discuss stability for the cases: (a) 0\C
1, (b) C [ 0.
qffiffiffiffiffiffi
7:5 Transistor Q1’s αf and αr parameters are changed by factor Ω1 and X21
3
qffiffiffiffiffiffi
respectively ðaf ! af X1 ; ar ! ar X21 ; X1 [ 0; X1 2 R þ Þ. How the
3
dynamical behavior of the circuit is changed for different values of Ω1

parameter? Discuss stability for different value of Ω1 parameter.
af : 0:95 ! 0:99; 0:95

af X1
0:99; ar : 0
qffiffiffiffiffiffi
! 0:95; 0\ar X21 \0:95; af [ ar
3
8. We have Gummel-Poon model equivalent circuit for BJT. Model’s diode D1 is

disconnected, how it influences our circuit dynamic? Consider single ended
class B amplifier. Write circuit differential equations and find fixed points,
discuss stability. Model’s capacitors Cjx and Ccs are disconnected. Find circuit
differential equations, fixed points and discuss stability. Consider single ended
class B amplifier.
9. We have two wideband LNAs with negative feedback networks which are
connected in cascade. The first wideband LNA circuit feeds the second wide-
band LNA circuit. The feedback networks include mixture of capacitors,
inductors, and resistors. There are three circuit’s matching networks, input
matching network Pi-type, cascade matching network T-type, and output
matching network Pi-type. Input RF signal is Vs(t), Vs ðtÞ ¼ A0 þ
nðtÞ; A0 nðtÞ. Inductors L3, L4 in the first LNA circuit and inductors L7, L8
in the second LNA circuit act as the RF choke which blocks the DC current
from entering the RF part in each LNA circuit. The LNAs design also employs
output resistive loadings (R4.R5 and R9) in stabilizing the circuit. The initial
output resistors values, R4.R5 and R9 are set to less than 100 Ω
ðR4 ; R5 ; R9 \100 XÞ because high output resistor values may result in huge
decrease of gain and P1dB point. The BJT transistors (Q1 and Q2) are
self-biased with the biasing resistors R1, R2 and R8, R7 respectively. Switch S1
connects and disconnects the second LNA negative feedback network.
Exercises 599
discuss stability (switch S1 is close).
9:2 How the circuit dynamic is changed if S1 switch is open? Find circuit
9:3 We change the input and output matching networks to T-type. The cascade
matching network is L-type. Switch S1 is in close state. Find circuit dif-
ferential equations and fixed points. Discuss stability.
9:4 Transistor Q2’s αf and αr parameters are changed by factor Ω1 and
qffiffiffiffiffiffiffiffiffiffiffiffiffiffi qffiffiffiffiffiffiffiffiffiffiffiffiffiffi
X31 þ 1 (af ! af X1 ; ar ! ar X31 þ 1; X1 [ 0;
5 5
respectively
X1 2 R þ ).
Switch S1 is open. How the dynamical behavior of the circuit is changed for
different values of Ω1 parameter. Discuss stability for different values of Ω1
parameter.
af : 0:95 ! 0:99; 0:95

af X1
0:99; ar : 0
qffiffiffiffiffiffiffiffiffiffiffiffiffiffi
! 0:95; 0\ar X31 þ 1\0:95; af [ ar
5
9:5 We disconnect capacitor C1. How the dynamical behavior of the circuit is
changed? Discuss stability and find circuit differential equations, fixed
points (switch S1 is close).
10. We have wideband Low Noise Amplifier (LNA) with possible four options of
negative feedback networks. The feedback network consists of capacitors,
resistors, inductors, and two SPD2 switches. Input and output matching net-
works are T-type. Input RF signal is Vs(t), Vs ðtÞ ¼ A0 þ nðtÞ; A0 nðtÞ.
Inductors L1, L2 in the circuit act as the RF choke which blocks the DC current
from entering the RF path. The LNA design also employs output resistive
loading in stabilizing the circuit. The initial output resistor value, R3 is set to
less than 100 Ω (R3 < 100 Ω) because high output resistor value may result in
huge decrease of gain and P1 dB point. The BJT transistor (Q1) is self-biased
with the biasing resistors R1 and R2. The possible negative feedback networks
options are as follow: (1) S1(a1) & S2(a2), (2) S1(a1) & S2(b2), (3) S1(b1) &
S2(a2), (4) S1(b1) & S2(b2).
10:1 Write circuit equations and circuit differential equations for each of the
negative feedback network options. Find fixed points and discuss stability
for each network option.
10:2 We short inductor L1, How it influences the circuit dynamics? Find circuit
differential equations and fixed points. Discuss stability (all network
options cases).
10:3 We short resistor R3, How it influences circuit dynamics? Find circuit
differential equations, fixed points and discuss stability (all network
options cases).
10:4 Return (10.1), (10.2), and (10.3) when the input and output matching
networks are L-type.
10:5 Return (10.1), (10.2), and (10.3) when we replace Q1 by n-Darlington
transistor (two BJT transistors (Qa, Qb), two resistors Ra, Rb and diode
Da). Consider that Qa and Qb transistors internal parameters are the same
afa ¼ afb ; ara ¼ arb ; etc. . . .
Chapter 6
Microwave/RF Oscillator Systems
Stability Analysis
An electronic oscillator is an electronic circuit that produces a periodic, oscillating

electronic signal, often a sine wave or a square wave. Oscillators are class of circuits
with one terminal or port, which produce a periodic electrical output upon power
up. Oscillators can be classified into two types: (a) Relaxation and (b) Harmonic
oscillators. Relaxation oscillators (also called unstable multi-vibrator) is a class of
circuits with two unstable states. The circuit switches back and forth between these
states. The output is generally square waves. Harmonic oscillators are capable of
producing near sinusoidal output, and are based on positive feedback approach. In
microwave the purpose of a microwave oscillator is to generate a microwave signal.
An oscillator consists of two parts: An active device to generate the power and a
resonator to control the frequency of the microwave signal. The oscillations are
made using feedback or negative resistance. Important issues in oscillators are
frequency stability, frequency tuning, and phase noise. Oscillator is a non-linear
circuit, initially upon power up the condition of oscillation to start up will prevail.
As the magnitudes of voltages and currents in the circuit increase, the amplifier in
the oscillator begins to saturate, reducing the gain, until the loop gain becomes one.
A steady state condition is reached when loop gain is equal to one. We can view an
oscillator as an amplifier that produces an output when there is no input. Thus is an
unstable amplifier that becomes an oscillator. An amplifier can be made unstable by
providing some kind of local positive feedback. Two favorite transistor amplifier
configurations used for oscillator design are the Common-Base (CB) configuration
with base feedback and Common-Emitter (CE) configuration with emitter degen-
eration. Oscillator performance requirements are frequency, frequency stability,
tuning, phase noise, power and efficiency.

DOI 10.1007/978-3-319-45427-6_6
602 6 Microwave/RF Oscillator Systems Stability Analysis
6.1 A Resonator Circuit 180° Phase Shift from Its Input

to Output Stability Analysis Under Delayed Variables
in Time
A phase-shift oscillator is a linear electronic oscillator circuit that produces a sine

wave output. It consists of an inverting amplifier element such as a transistor or op
amp with its output fed back to its input through a phase-shift network consisting of
resistors and capacitors in a ladder network. The feedback network ‘shifts’ the
phase of the amplifier output by 180° at the oscillation frequency to give positive
feedback. Phase-shift oscillators are often used at audio frequency as audio oscil-
lators. A phase shift oscillator consists of a single stage of amplifier that amplifies
the input signal and produces a phase shift of 180° the input and its output signal. If
a part of this output is taken and feedback to input, it results in negative feedback
causing the output voltage to decrease. We require positive feedback which means
that the voltage signal feedback should be in phase with the input signal. The output
of the amplifier should take through a phase shift network to provide it an additional
phase shift of 180°. Amplifier provides a phase shift of 180° and the phase shift
network also gives a 180° and therefore, a total phase-shift of 360° (which is
equivalent to 0°) results [25, 26] (Fig. 6.1).
The RC network provides the required phase shift by using three RC. Each having
some value of R and C. These values are selected so as to produce 60° phase shift per
section, resulting in total of 180° phase shift as desired. Practically each RC section does
not provide the same phase shift because each section leads the previous one but the
overall phase shift is 180° which is the requirement. The frequency at which phase shift
Fig. 6.1 Resonator circuit 180° (degree) phase shift

6.1 A Resonator Circuit 180° Phase Shift … 603
1 pffiffi
is 180° is f ¼ 2pRC 6
. The circuit generates a sinusoidal wave at its output. The RC
phase shift network is used to form a positive feedback loop. Rb and Rc resistors
polarize the transistor Q1 for DC current, Rs is a separating resistor. R1 ¼ R2 ¼ R;
C1 ¼ C2 ¼ C3 ¼ C. Phase shift resonant circuit stability analysis is done by con-
sidering BJT Small Signal (SS) equivalent circuit model. We consider “AC ground” in
the circuit. Since the voltage at this terminal is held constant at Vcc, there is no time
variation of the voltage consequently. We can set this terminal to be an “AC ground”
in the small signal circuit. For AC ground, we “kill” the DC sources at that terminal
(short circuit voltage sources and open circuit current sources) [33, 34] (Fig. 6.2).
Vcc VA4 Vcc VA1 VA VA5 VA5

IRc ¼ ; IR b ¼ ; IR s ¼ 4 ; IRload ¼ ; Ir pi
Rc Rb Rs Rload
VA d
¼ 1 ; IC3 ¼ C3 ðVA4 VA3 Þ
r pi dt
VA3 d VA
IR 2 ¼ ; IC3 ¼ IC2 þ IR2 ; IC2 ¼ C2 ðVA3 VA2 Þ; IR1 ¼ 2 ; IC2 ¼ IC1 þ IR1
R3 dt R1
d
IC1 ¼ C1 ðVA2 VA1 Þ; IRb þ IC1 ¼ Ir pi ; IRc ¼ IC3 þ IRs þ gm vBE ; vBE ¼ VA1
dt
IRc ¼ IC3 þ IRs þ gm VA1 ¼ IC3 þ IRs þ gm ½Vcc IRb Rb
Fig. 6.2 Phase shift resonator equivalent circuit (BJT small signal)
Vcc VA4
IRc ¼ ) IRc Rc ¼ Vcc VA4 ) VA4 ¼ Vcc IRc Rc
Rc
Vcc VA1
IRb ¼ ) IRb Rb ¼ Vcc VA1 ) VA1 ¼ Vcc IRb Rb
Rb
VA5
IRload ¼ ; IRload ¼ IRs ; VA5 ¼ IRload Rload ¼ IRs Rload
Rload
VA VA5
IR s ¼ 4 ) IRs Rs ¼ VA4 VA5 ¼ Vcc IRc Rc IRs Rload
Rs
V A1
IRs ðRs þ Rload Þ ¼ Vcc IRc Rc ; Ir pi ¼ ) V A1 ¼ I r pi r pi; Ir pi r pi ¼ Vcc IRb Rb
r pi
VA3 VA d
I R2 ¼ ) VA3 ¼ IR2 R3 ; IR1 ¼ 2 ) VA2 ¼ IR1 R1 ; IC3 ¼ C3 ðVcc IRc Rc IR2 R3 Þ
R3 R1 dt
dVcc dIRc dIR dVcc

IC3 ¼ C3 ð Rc 2 R3 Þ; ! e;
dt dt dt dt
dIR dIR2
IC3 ¼ C3 ð c Rc þ R3 Þ
dt dt
d d
IC2 ¼ C2 ðVA3 VA2 Þ ¼ C2 ðIR2 R3 IR1 R1 Þ;
dt dt
dIR2 dIR1
I C2 ¼ C2 ð R3 R1 Þ
dt dt
d d
I C1 ¼ C 1 ðVA2 VA1 Þ ¼ C1 ðIR1 R1 Vcc þ IRb Rb Þ;
dt dt
dIR1 dVcc dIRb
I C1 ¼ C 1 ð R1 þ Rb Þ
dt dt dt
dVcc dIR dIRb
! e; IC1 ¼ C1 ð 1 R1 þ Rb Þ
dt dt dt
We can write KCL for the above circuit (Table 6.1):
Table 6.1 Phase shift KCL @ node Expression

resonator equivalent circuit
(BJT small signal) KCL@ A1 IRb þ IC1 ¼ Ir pi
nodes and expressions A2 IC2 ¼ IC1 þ IR1
A3 IC3 ¼ IC2 þ IR2
A4 IRc ¼ IC3 þ IRs þ gm vBE
A5 IRload ¼ IRs
IR b þ IC 1 ¼ Ir pi ) I C1 ¼ I r pi IRb ; IC2 ¼ IC1 þ IR1 ) IC2 ¼ Ir pi IRb þ IR1
X
2
IC3 ¼ IC2 þ IR2 ¼ Ir pi IRb þ IR1 þ IR2 ¼ Ir pi IRb þ IRk ;
k¼1
X
2
I C3 ¼ I r pi IRb þ IR k
k¼1
We can summery our system three differential equations:
dIR1 dIRb
Ir pi IRb ¼ C1 ð R1 þ Rb Þ; Ir pi IR b þ IR 1
dt dt
dIR2 dIR1
¼ C2 ð R3 R1 Þ
dt dt
X
2
dIRc dIR2
Ir pi IRb þ IRk ¼ C3 ð Rc þ R3 Þ
k¼1
dt dt
Vcc Rc
IRs ðRs þ Rload Þ ¼ Vcc IRc Rc ) IRs ¼ IR c
ðRs þ Rload Þ ðRs þ Rload Þ
X
2
Vcc Rc
IRc ¼ Ir pi IR b þ IRk þ IR c þ gm ½Vcc IRb Rb
k¼1
ðRs þ Rload Þ ðRs þ Rload Þ
Rc X
2
Vcc
IRc ð1 þ Þ ¼ Ir pi IR b þ IRk þ þ gm
ðRs þ Rload Þ k¼1
ðRs þ Rload Þ
½Vcc IRb Rb
Vcc Rb
Ir pi r pi ¼ Vcc IRb Rb ) Ir pi ¼ IRb
r pi r pi
Rc Vcc Rb X2
Vcc
IRc ð1 þ Þ¼ IRb IR b þ IRk þ
ðRs þ Rload Þ r pi r pi k¼1
ðRs þ Rload Þ
þ gm ½Vcc IRb Rb
Rc 1 1 Rb
IRc ð1 þ Þ ¼ Vcc ½ þ þ gm IRb ½ þ1
ðRs þ Rload Þ r pi ðRs þ Rload Þ r pi
X
2
þ gm Rb þ IRk
k¼1
½r 1pi þ ðRs þ Rload Þ þ gm

1
½r Rbpi þ 1 þ gm Rb 1
IRc ¼ Vcc IR b þ
ð1 þ ðRs þRRc load ÞÞ ð1 þ ðRs þ Rload ÞÞ
Rc
ð1 þ ðRs þ Rload ÞÞ
Rc
X
2
IRk
k¼1
½r 1pi þ ðRs þ Rload Þ þ gm

1
C1 ðr pi; Rs ; Rload ; . . .Þ ¼ ; C1 ¼ C1 ðr pi; Rs ; Rload ; . . .Þ
ð1 þ ðRs þRRc load ÞÞ
½r Rbpi þ 1 þ gm Rb
C2 ðRb ; r pi; Rb ; . . .Þ ¼ ; C2 ¼ C2 ðRb ; r pi; Rb ; . . .Þ
Rc
1
C3 ðRc ; Rs ; . . .Þ ¼ ; C3 ¼ C3 ðRc ; Rs ; . . .Þ
Rc
IRc ¼ Vcc C1 ðr pi; Rs ; Rload ; . . .Þ IRb C2 ðRb ; r pi; Rb ; . . .Þ þ C3 ðRc ; Rs ; . . .Þ

X
2
IR k
k¼1
dIRc dVcc dIR

¼ C1 ðr pi; Rs ; Rload ; . . .Þ b C2 ðRb ; r pi; Rb ; . . .Þ þ C3 ðRc ; Rs ; . . .Þ
dt dt dt
X2
dIRk

k¼1
dt
dVcc dVcc dIRc dIR X2

dIRk
!e) C1 ðr pi; Rs ; Rload ; . . .Þ ! e; ¼ b C2 þ C3
dt dt dt dt k¼1
dt
We can summery our system three differential equations with variables:

R1 ; R2 ; Rb .
Vcc Rb dIR dIRb

ðÞ IRb IRb ¼ C1 ð 1 R1 þ Rb Þ
r pi r pi dt dt
Vcc Rb dIR dIR

ðÞ IR b ½ þ 1 þ IR1 ¼ C2 ð 2 R3 1 R1 Þ
r pi r pi dt dt
Vcc Rb X2
dIR
ð Þ IRb IRb þ IRk ¼ C3 ð b C2 Rc þ C3 Rc
r pi r pi k¼1
dt
X dIR
2
dIR2
k
þ R3 Þ
k¼1
dt dt
Another representation of system three differential equations:
Vcc Rb dIR dIRb

ðÞ IRb IRb ¼ C1 ð 1 R1 þ Rb Þ
r pi r pi dt dt
Vcc Rb dIR dIR1

IR b ½ þ 1 b C1 Rb ¼ C1 R1
r pi r pi dt dt
Vcc Rb dIR dIR

ðÞ IR b ½ þ 1 þ IR1 ¼ C2 ð 2 R3 1 R1 Þ
r pi r pi dt dt
Vcc Rb dIR dIR

IRb ½ þ 1 þ IR1 ¼ C2 ð 2 R3 1 R1 Þ
r pi r pi dt dt
Vcc Rb X2
dIR
ð Þ IRb IRb þ IRk ¼ C3 ð b C2 Rc þ C3 Rc
r pi r pi k¼1
dt
X2
dIRk dIR2
þ R3 Þ
k¼1
dt dt
Vcc Rb X2
dIRb dIR dIR
IR b ½ þ 1 þ IRk ¼ C2 C3 Rc C3 Rc C3 1 2
r pi r pi k¼1
dt dt dt
C3 ½C3 Rc þ R3
And we get the following differential equations:
dIR1 Vcc ½r Rbpi þ 1 dIRb Rb

ðÞ ¼ IRb
dt r pi C1 R1 C1 R1 dt R1
ðÞ ! ðÞ
Vcc Rb
IRb ½ þ 1 þ IR1
r pi r pi
¼ C2 ð R3 ½ IR b R1 Þ
Vcc Rb dIR2 Vcc C2

IRb ½ þ 1 þ IR1 ¼ C2 R3 þ IRb
r pi r pi dt r pi C1
½r pi þ 1 C2 dIRb
Rb
þ Rb C2
C1 dt
Vcc dIR2 Vcc C2 Rb C2 dIRb

þ IR1 ¼ C2 R3 þ IRb ½ þ 1 ½ þ 1 þ Rb C2
r pi dt r pi C1 r pi C1 dt
ðÞ ! ð Þ
Vcc Rb X2
dIRb
IR b ½ þ 1 þ IR k ¼ C2 C3 Rc
r pi r pi k¼1
dt
Vcc ½r Rbpi þ 1 dIRb Rb dIR
C3 Rc C3 ½ IR b 2 C3 ½C3 Rc þ R3
r pi C1 R1 C 1 R1 dt R1 dt
Vcc Rb X2
dIRb C3 Rb
IRb ½ þ 1 þ IRk ¼ Rc C3 ½C2 þ
r pi r pi k¼1
dt R1
C3 Rc C3 Vcc C3 Rc C3 ½r Rbpi þ 1 dIR2
þ IRb C3 ½C3 Rc þ R3
r pi C1 R1 C1 R1 dt
We define new global parameters for simplicity:
C2 Rb C2 Rb
X1 ¼ ; X2 ¼ ½ þ 1 ½ þ 1; X3 ¼ ½ þ 1;
r pi C1 r pi C1 r pi
C3 Rb
X4 ¼ Rc C3 ½C2 þ
R1
C3 Rc C3 C3 Rc C3 ½r Rb
pi þ 1
X5 ¼ ; X6 ¼ ; X7 ¼ C3 ½C3 Rc þ R3
r pi C1 R1 C1 R1
Vcc dIR2 dIRb

þ IR1 ¼ C2 R3 Vcc X1 þ IRb X2 þ Rb C2
r pi dt dt
Vcc X2
dIRb dIR
IR b X 3 þ IR k ¼ X4 Vcc X5 þ IRb X6 2 X7
r pi k¼1
dt dt
Vcc dIR2 dIRb

þ IR1 ¼ C2 R3 Vcc X1 þ IRb X2 þ Rb C2
r pi dt dt
dIRb ½r 1pi þ X1 1 R3 dIR X2

¼ Vcc þ IR1 2 IRb
dt Rb C2 Rb C2 Rb dt Rb C2
Vcc X2
dIRb dIR
IR b X 3 þ IR k ¼ X4 Vcc X5 þ IRb X6 2 X7
r pi k¼1
dt dt
Vcc X2 ½r 1pi þ X1 1 R3 dIR X2

IR b X 3 þ IRk ¼ ½Vcc þ IR 1 2 IR b X4
r pi k¼1
R b C 2 R b C 2 Rb dt R b C2
dIR2
Vcc X5 þ IRb X6 X7
dt
R3 X4 dIR 1 ½r 1
pi þ X1 X4 X2 X4
½ þ X7 2 ¼ Vcc ð þ X 5 Þ þ IR b ½ X6 X3
Rb dt r pi Rb C 2 Rb C 2
X4
þ ½1 I R 1 þ IR 2
Rb C 2
½r 1
pi þ X1 X4
dIR2 ð Rb C2 r 1pi X5 Þ ½X6 þ X3 XRb2C
X4

¼ Vcc þ IRb 2
dt ½R3RXb 4 þ X7 ½R3RXb 4 þ X7
½RbXC4 2 1 1
þ IR1 IR2
½R3RXb 4 þ X7 ½R3RXb 4 þ X7
½ 1
pi þ X1 X4
ðr Rb C2 r 1pi X5 Þ
N1 ¼ N1 ðr pi; R3 ; Rb ; X1 ; X4 ; . . .Þ ¼
½R3RXb 4 þ X7
½X6 þ X3 XRb2C
X4

N2 ¼ N2 ðRb ; R3 ; X2 ; X3 ; X4 ; . . .Þ ¼ 2
; N3 ¼ N3 ðRb ; R3 ; X4 ; . . .Þ
½R3RXb 4 þ X7
½RbXC4 2 1
¼
½R3RXb 4 þ X7
1
N4 ¼ N4 ðR3 ; Rb ; X4 ; . . .Þ ¼
½R3RXb 4 þ X7
dIR2
¼ Vcc N1 þ IRb N2 þ N3 IR1 þ N4 IR2
dt
dIRb ½r pi þ X1
1
1 R3
¼ Vcc þ IR1 ½Vcc N1 þ IRb N2 þ N3 IR1
dt Rb C2 Rb C2 Rb
X2
þ N4 IR2 IRb
Rb C2
1 ½r pi þ X1
1
dIRb 1 1 R3
¼ Vcc ½ R3 N1 þ ½ N3 R3 IR1 N4 IR2
dt Rb C2 C2 Rb Rb
1 X2
IRb ½R3 N2 þ
Rb C2
1 ½r pi þ X1
1
N5 ¼ N5 ðr pi; Rb ; R3 ; . . .Þ ¼ ½ R3 N1
Rb C2
1 1 R3
N6 ¼ N6 ðN3 ; R3 ; Rb ; . . .Þ ¼ ½ N3 R3 ; N7 ¼ N7 ðN4 ; R3 ; . . .Þ ¼ N4
C2 Rb Rb
1 X2
N8 ¼ N8 ðR3 ; N2 ; . . .Þ ¼ ½R3 N2 þ ;
Rb C2
dIRb
¼ Vcc N5 þ N6 IR1 þ N7 IR2 þ IRb N8
dt

¼ IRb
dIR1 Vcc ½r Rbpi þ 1 Rb

¼ IR b ½Vcc N5 þ N6 IR1 þ N7 IR2 þ IRb N8
dt r pi C1 R1 C 1 R1 R1
dIR1 1 Rb Rb Rb 1
¼ Vcc ½ N5 N6 IR1 N7 IR2 IRb
dt r pi C1 R1 R1 R1 R1 R1
½r Rbpi þ 1
½N8 Rb þ
C1
1 Rb
N9 ¼ N9 ðR1 ; Rb ; r pi; . . .Þ ¼ N5 ; N10 ¼ N10 ðRb ; R1 ; . . .Þ
r pi C1 R1 R1
Rb
¼ N6
R1
Rb
N11 ¼ N11 ðN7 ; Rb ; . . .Þ ¼ N7 ; N12 ¼ N12 ðN8 ; R1 ; Rb ; . . .Þ
R1
1 ½r Rbpi þ 1
¼ ½N8 Rb þ
R1 C1
dIR1
dt
We can summery our system set of differential equations:
dIR1 dIR
¼ Vcc N9 þ N10 IR1 þ N11 IR2 þ IRb N12 ; 2
dt dt
¼ Vcc N1 þ IRb N2 þ N3 IR1 þ N4 IR2
dIRb
dt
Phase shift resonator circuit stability analysis is done by considering BJT Small
Signal (SS) equivalent circuit model. We consider “AC ground” in the circuit. Since
the voltage at this terminal is held constant at Vcc, there is no time variation of the
voltage. Consequently, we can set this terminal to be an “AC ground” in the small
signal circuit (Vcc = 0). Under AC and Small Signal (SS) conditions, the BJT is
replaced with linear Hybrid Pi model [26, 27].
dIR1 dIR
¼ N10 IR1 þ N11 IR2 þ IRb N12 ; 2 ¼ IRb N2 þ N3 IR1 þ N4 IR2
dt dt
dIRb
¼ N6 IR1 þ N7 IR2 þ IRb N8
dt
dIR1 dIR2 dIRb
At fixed points: dt ¼ 0; dt ¼ 0; dt ¼0
N10 IR1 þ N11 IR2 þ IRb N12 ¼ 0; N3 IR1 þ N4 IR2 þ IRb N2 ¼ 0;

N6 IR1 þ N7 IR2 þ IRb N8 ¼ 0
EðIR1 ; IR2 ; IRb Þ ¼ ð0; 0; 0Þ. If Vcc 6¼ 0 then we get the following three equations
for system fixed points (equilibrium points):
N10 IR1 þ N11 IR2 þ IRb N12 ¼ Vcc N9 ; N3 IR1 þ N4 IR2 þ IRb N2 ¼ Vcc N1
N6 IR1 þ N7 IR2 þ IRb N8 ¼ Vcc N5
To use determinants to solve our system fixed points equations ðVcc 6¼ 0Þ with
three fixed points (Cramer’s Rule), IR1 ; IR2 ; IRb , four determinants must be formed
following the procedure:
(I) Write all fixed points equations in standard form.
(II) Create the denominator determinant, D, by using the coefficients of
IR1 ; IR2 ; IRb from the equations and evaluate it.
(III) Create the IR1 —numerator determinant, DIR , the IR2 —numerator determinant,
1
DIR , and the IRb —numerator determinant, DIR , by replacing the respective
2 b
IR1 ; IR2 , and IRb coefficients with the constants from the equations in standard
form and evaluate each determinant.
DI DI DI
The answers for IR1 ; IR2 , and IRb are as follow: IR1 ¼ DR1 ; IR2 ¼ DR2 ; IRb ¼ D
Rb
.
We solve this system of fixed point’s equations, using Cramer’s Rule.
0 1
N10 N11 N12
B C N4 N2 N3 N2 N3 N4
D ¼ det@ N3 N4 N2 A ¼ N10 det N11 det þ N12 det
N7 N8 N6 N8 N6 N7
N6 N7 N8
D ¼ N10 ðN4 N8 N7 N2 Þ N11 ðN3 N8 N6 N2 Þ þ N12 ðN3 N7 N6 N4 Þ
D ¼ N10 N4 N8 N10 N7 N2 N11 N3 N8 þ N11 N6 N2 þ N12 N3 N7 N12 N6 N4
We use the constants to replace the IR1 —coefficients.

0 1
Vcc N9 N11 N12
B C N4 N2
D IR ¼ det@ Vcc N1 N4 N2 A ¼ Vcc N9 det
1 N7 N8
Vcc N5 N7 N8

N11 N12 N11 N12
þ Vcc N1 det Vcc N5 det
N7 N8 N4 N2
DIR ¼ Vcc N9 ðN4 N8 N7 N2 Þ þ Vcc N1 ðN11 N8 N7 N12 Þ Vcc N5 ðN11 N2 N4 N12 Þ

1
DIR ¼ Vcc N9 N4 N8 þ Vcc N9 N7 N2 þ Vcc N1 N11 N8 Vcc N1 N7 N12

1
Vcc N5 N11 N2 þ Vcc N5 N4 N12
We use the constants to replace the IR2 —coefficients.

0 1
N10 Vcc N9 N12
B C Vcc N1 N2
DIR ¼ det@ N3 Vcc N1 N2 A ¼ N10 det
2 Vcc N5 N8
N6 Vcc N5 N8

N3 N2 N3 Vcc N1
þ Vcc N9 N12
N6 N8 N6 Vcc N5
DIR ¼ N10 ðVcc N1 N8 þ Vcc N5 N2 Þ þ Vcc N9 ðN3 N8 N6 N2 Þ

2
N12 ðVcc N5 N3 þ Vcc N1 N6 Þ

2
þ Vcc N5 N12 N3 Vcc N1 N12 N6
We use the constants to replace the IRb —coefficients.

0 1
N10 N11 Vcc N9
B C N4 Vcc N1
DIR ¼ det@ N3 N4 Vcc N1 A ¼ N10 det
b N7 Vcc N5
N6 N7 Vcc N5

N3 Vcc N1 N3 N4
N11 Vcc N9
N6 Vcc N5 N6 N7
DIR ¼ N10 ðVcc N5 N4 þ Vcc N1 N7 Þ N11 ðVcc N5 N3 þ Vcc N1 N6 Þ

b
Vcc N9 ðN3 N7 N6 N4 Þ

b
Vcc N9 N3 N7 þ Vcc N9 N6 N4
DI DI DI
Therefore, IR1 ¼ DR1 ; IR2 ¼ DR2 ; IRb ¼ DRb . If the denominator determinant, D,
has a value of zero, then system is either inconsistent or dependent. The system is
dependent if all the determinants have a value of zero. The system is inconsistent if
at least one of the determinants, DIR ; DIR or DIR has a value not equal to zero and
1 2 b
the denominator determinant has a value of zero.
Stability discussion:
0 dIR 1 0 1 0 1 0 1 0 1
N10 N11 N12 b1 ðIR1 ; tÞ IR1 ðt ¼ 0Þ
1
dt IR 1
B dIR2 C @
@ dt A ¼ N3 N4 N2 A @ IR2 A þ @ b2 ðIR2 ; tÞ A; @ IR2 ðt ¼ 0Þ A ¼ a0
dIRb N6 N7 N8 IR b b3 ðIRb ; tÞ IRb ðt ¼ 0Þ
dt
1 0
IR 1
We define the following notation: X ¼ @ IR2 A; X ¼ ðIR1 ; IR2 ; IRb Þ?
IR b
0 1
N10 N11 N12
A ¼ @ N3 N4 N2 A; b ¼ ðarray20cb1 b2 b3 Þ; b ¼ ðb1 ; b2 ; b3 Þ?
N6 N7 N8
A is a real, time independent (3 × 3)-matrix and the components of

b ¼ ðb1 ; b2 ; b3 Þ? 0 1
IR1
Are real C1—functions for all X ¼ @ IR2 A in a neighborhood of the origin for
IRb
0 1 0 1
b1 ðIR1 ¼ 0; tÞ IR1
all times t ≥ 0. Moreover, let @ b2 ðIR2 ¼ 0; tÞ A ¼ 0 . Hence X ¼ @ IR2 A 0 is a
b3 ðIRb ¼ 0; tÞ IRb
solution of our system linear differential equation which corresponds to a point of

equilibrium of the system. The equilibrium point is at X = 0.
Stability: The equilibrium point EðIR1 ; IR2 ; IRb Þ ¼ ð0; 0; 0Þ is stable if and only if
for each ε > 0 there is a number δ > 0 such that from jtj\d, the existence of a
unique solution X ¼ XðtÞ ) ðIR1 IR2 IRb Þ ¼ ðIR1 ðtÞIR2 ðtÞIRb ðtÞÞ of our system linear
differential equation follows, with jXðtÞj\e for all times t ≥ 0. This means that
sufficiently small perturbations of the equilibrium configuration at X ¼
ðIR1 IR2 IRb Þ ¼ 0 remain small for all t ≥ 0.
Asymptotic stability: The equilibrium point EðIR1 ; IR2 ; IRb Þ ¼ ð0; 0; 0Þ is
asymptotically stable if and only if it is stable and in addition there is a number
d [ 0 such that, for each solution with jXð0Þj\d , we have the limit rela-
tion lim XðtÞ ¼ 0. Sufficient small perturbations of the equilibrium configuration at
t!1
time t = 0 return to their starting configuration after a sufficiently long term.
Instability: The equilibrium point EðIR1 ; IR2 ; IRb Þ ¼ ð0; 0; 0Þ is instable if and
only if it is not stable.
We can implement in our system the theorem on stability by Liapunov:
dIR1 dIR2 dIRb
Suppose a perturbation b of the linear system ð dt dt dt Þ ¼
0 1
N10 N11 N12
@ N3 N4 N2 A ðIR1 IR2 IRb Þ
N6 N7 N8
With constant coefficients is sufficiently small, i.e., one has
lim ðsup jbðX;tÞj
jXj Þ ¼ 0
jXj!1 t 0
The equilibrium point EðIR1 ; IR2 ; IRb Þ ¼ ð0; 0; 0Þ is asymptotically stable, if all
eigenvalues λ1, λ2, … of the matrix A are in the left half plane, have negative real
part for all j. The equilibrium point EðIR1 ; IR2 ; IRb Þ ¼ ð0; 0; 0Þ is unstable, if an
eigenvalue of A is in the right half plane, i.e., one has Re kj [ 0 for some j. If an
eigenvalue of A is on the imaginary axis, then the method of the center manifold
must be applied. To apply the stability criterion of Liapunov to our system, the
equation detðA k IÞ ¼ 0 has a zero in the left half plane, without calculating the
eigenvalues explicitly.
Eigenvalues stability discussion: Our phase shift oscillation system involving N
variables (N > 2, N = 3), the characteristic equation is of degree N = 3 and must
often be solved numerically. Expect in some particular cases, such an equation has
(N = 3) distinct roots that can be real or complex. These values are the eigenvalues
of the 3 × 3 Jacobian matrix (A). The general rule is that the Steady State (SS) is
stable if there is no eigenvalue with positive real part. It is sufficient that one
eigenvalue is positive for the steady state to be unstable. Our 3-variables
ðIR1 ; IR2 ; IRb Þ system has three eigenvalues. The type of behavior can be
characterized as a function of the position of these eigenvalues in the Re/Im plane.

Five non-degenerated cases can be distinguished: (1) the three eigenvalues are real
and negative (stable steady state), (2) the three eigenvalues are real, two of them are
negative (unstable steady state), (3) and (4) two eigenvalues are complex conjugates
with a negative real part and the third one real is negative (stable steady state), two
cases can be distinguished depending on the relative value of the real part of the
complex eigenvalues and of the real one, (5) two eigenvalues are complex conju-
gates with a negative real part and the third one real is positive (unstable steady
state) [2, 3, 4].
0 1
N10 N11 N12
A ¼ @ N3 N4 N2 A; detðA k IÞ ¼ 0
N0
6 N7 N8 1
N10 k N11 N12
) det@ N3 N4 k N2 A ¼ 0
N6 N7 N8 k
0 1
N10 k N11 N12
B C N4 k N2
det@ N3 N4 k N2 A ¼ ðN10 kÞ det
N7 N8 k
N6 N7 N8 k

N3 N2
N11 det
N6 N8 k

N3 N4 k
þ N12 det
N6 N7
detðA k IÞ ¼ ðN10 kÞ ½ðN4 kÞ ðN8 kÞ N7 N2 N11 ½N3 ðN8 kÞ N6 N2

þ N12 ½N3 N7 N6 ðN4 kÞ
detðA k IÞ ¼ ðN10 kÞ ½ðN4 N8 N7 N2 Þ ðN4 þ N8 Þ k þ k2

N11 ½ðN3 N8 N6 N2 Þ N3 k þ N12 ½ðN3 N7 N6 N4 Þ þ N6 k
detðA k IÞ ¼ N10 ðN4 N8 N7 N2 Þ N10 ðN4 þ N8 Þ k þ N10 k2

k ðN4 N8 N7 N2 Þ þ ðN4 þ N8 Þ k2 k3 N11 ðN3 N8 N6 N2 Þ
þ N11 N3 k þ N12 ðN3 N7 N6 N4 Þ þ N6 N12 k
detðA k IÞ ¼ k3 þ ½N10 þ N4 þ N8 k2

þ ½N11 N3 þ N6 N12 ðN4 N8 N7 N2 Þ N10 ðN4 þ N8 Þ k
þ N10 ðN4 N8 N7 N2 Þ N11 ðN3 N8 N6 N2 Þ þ N12 ðN3 N7 N6 N4 Þ
We define new parameters: !3 ¼ 1; !2 ¼ N10 þ N4 þ N8
!1 ¼ N11 N3 þ N6 N12 ðN4 N8 N7 N2 Þ N10 ðN4 þ N8 Þ

!0 ¼ N10 ðN4 N8 N7 N2 Þ N11 ðN3 N8 N6 N2 Þ þ N12 ðN3 N7 N6 N4 Þ
X
3 X
3
detðA k IÞ ¼ !k kk ; detðA k IÞ ¼ 0 ) ! k kk ¼ 0
k¼0 k¼0
P
We get zeros of the polynomial 3k¼0 !k kk ¼ 0 with real coefficients !k .
P
The characteristic equation is a polynomial of degree N = 3: 3k¼0 !k kk ¼ 0:
We suppose k1 ; k2 ; k3 all (known) eigenvalues of the linearized system:
dXðtÞ
dt ¼ A XðtÞ; XðtÞ ¼ ðIR1 ðtÞIR2 ðtÞIRb ðtÞÞ. If one eigenvalue kk ; k ¼ 1; 2; 3 has
a real part greater than zero kk [ 0; k ¼ 1; 2; 3; the perturbation will ultimately
increase and the steady state is thus unstable. To determine if all the eigenvalue
have a negative real part can be done by checking some conditions, known as the
Routh-Hurwitz criteria. In our case we define N = 3 matrices as follow:
Coefficients !k are real and fulfilling !k¼3 lie in the left half plane, if and only if,
all determinants: H1, H2, and H3 (with !m ¼ 0 for m [ 3; then !4 ¼ 0; !5 ¼ 0Þ are
positive. It is the condition for stability.
0 1
!1 !0 0
!1 !0
H1 ¼ ð!1 Þ; H2 ¼ ; H3 ¼ @ !3 !2 !1 A
!3 !2
!5 ¼ 0 !4 ¼ 0 !3
Stability criteria:

!1 !0
detðH1 Þ ¼ detð!1 Þ [ 0 ) !1 [ 0 ; det(H2 Þ ¼ det [0
!3 !2
) !1 !2 !3 !0 [ 0 ) !1 !2 [ !3 !0
0 1
!1 !0 0
B C
detðH3 Þ ¼ det@ !3 !2 !1 A [ 0 ) !1 !2 !3 !0 !23
!5 ¼ 0 !4 ¼ 0 !3
Y
3
¼ !k !0 !23 [ 0
k¼1
Y
3
!k !0 !23 ¼ !3 ð!1 !2 !0 !3 Þ [ 0
k¼1
We have two cases for detðH3 Þ [ 0: Case (I) !3 [ 0 & !1 !2 [ !0 !3

Case (II) !3 \0 and !1 !2 \!0 !3 .
We are interesting to get oscillations from the phase shift resonator circuit. To get
oscillations, we need eigenvalue of A is on imaginary axis, and then the method of
the center manifold is applied. The center manifold of an equilibrium point (fixed
point) of our dynamical system consists of orbits whose behavior around the
equilibrium point is not controlled by either the attraction of the stable manifold or
the repulsion of the unstable manifold. Phase shift resonator eigenvalues corre-
sponding to eigenvalues with negative real part form the stable Eigen space, which
gives rise to the stable manifold. Eigenvalue with positive real part yield the
unstable manifold. If there are eigenvalues whose real part is zero, then these give
rise to the center manifold. If the eigenvalues are precisely zero, then these more
specifically give rise to a slow manifold [3, 4].
The system matrix A defines three main subspaces:
(I) Stable subspace, which is spanned by our system generalized eigenvectors
corresponding to the eigenvalues λ with Re λ < 0.
(II) Unstable subspace, which is spanned by our system generalized eigenvectors
corresponding to the eigenvalues λ with Re λ > 0.
(III) Center subspace, which is spanned by our generalized eigenvectors corre-
sponding to the eigenvalues λ with Re λ = 0.
The options for sub spaces of interest include center-stable, center-unstable,
sub-center, slow, and fast subspaces. These subspaces are all invariant subspaces of
the system’s linear equations.
Center manifold theorem: The neighborhood may be chosen so that all solutions
of the system staying in the neighborhood tend exponentially to some solution on
the center manifold. A wide variety of initial conditions yields to solutions of the
full system which decay exponentially quickly to a solution on the relatively low
dimensional center manifold.
6.2 Closed Loop Functioning Oscillator Stability Analysis

Under Parameters Variations
A good oscillator is stable in that its frequency and amplitude of oscillation do not
vary appreciably with temperature, process, power supply and external distur-
bances. The amplitude of oscillation is particularly stable, always returning to the
same value. LC tank oscillator is not a good oscillator. Due to loss, no matter how
small, the amplitude of the oscillator decays. Many oscillators can be viewed as
feedback systems. The oscillation is sustained by feeding back a fraction of the
output signal, using an amplifier to gain the signal, and then injecting the energy
back into the tank. The transistor “pushes” the LC tank with just about enough
energy to compensate for the loss. Typical oscillator feedback system is describes in
the below figure (Fig. 6.3).
Fig. 6.3 Typical closed loop oscillator feedback system
Fig. 6.4 LC tank circuit and

active device as a negative
resistance generator
Another option is to view the active device as a negative resistance generator. In

steady state, the losses in the tank due to conductance G are balanced by the power
drawn from the active device through the negative conductance—G (Fig. 6.4).
In a real oscillator, the amplitude of oscillation initially grows exponentially as
linear system theory predicts. The oscillator amplitude is initially very small. But as
the oscillations become more vigorous, the non-linearity of the system comes into
play. We will analyze by using nonlinear dynamic the steady-state behavior, where
the system is non-linear but periodically time-varying. Typical circuit is BJT NPN
transistor LC oscillator. The base of the NPN transistor (Q1) is conveniently biased
through the transformer windings. The transistor Q1’s emitter resistor is bypassed
by a large capacitor at AC frequencies. The LC oscillator uses a transformer for
feedback. Since the amplifier has a phase shift of 180°, the feedback transformer
needs to provide an additional phase shift of 180° to provide positive feedback
[33, 34]. (Fig. 6.5).
6.2 Closed Loop Functioning Oscillator Stability Analysis … 619
Fig. 6.5 Closed loop functioning oscillator with transformer T1
We use for our stability analysis the BJT NPN transistor Small Signal
(SS) model equivalent circuit (Hybrid Pi model Version A). We consider “AC
ground” in the circuit. Since the voltage at this terminal is held constant at Vcc, there
is no time variation of the voltage. Consequently, we can set this terminal to be an
“AC ground” in the small signal circuit. For AC ground, we “kill” the DC sources
at that terminal (short circuit voltage sources and open circuit current sources).
Transformer T1: vp—transformer primary voltage, vs—transformer secondary
voltage. ϕ is the flux through a one turn coil located anywhere on the transformer
core. np and ns are the number of turns of the coil 1 (primary) and 2 (secondary),
respectively, then the total flux ϕp and ϕs through coils 1 and 2 respectively
d/p d/s
are /p ¼ np /; /s ¼ ns /; vp ¼ dt ; vs ¼ dt ; vp ¼ np dt ; vs ¼ ns dt ; vp ðtÞ ¼
d/ d/
v ðtÞ n n v ðtÞ n
vp ; vs ðtÞ ¼ vs ; vps ðtÞ ¼ nps ; a ¼ nps . vps ðtÞ ¼ nps for all times t and for all voltages vp and vs
n
. We define the ratio between np and ns as “a” ða ¼ nps Þ. Step down transformer
a > 1 and step up transformer a < 1. By law of conservation of energy, apparent
real and reactive powers are each conserved in the input and output
qffiffiffiffi
v n L
ðS ¼ Ip vp ¼ Is vs Þ. Ideal transformer identity: vps ¼ IIps ¼ nps ¼ Lps ¼ a. np ¼ n;
vA2
¼ vn 0 ¼
v vp
ns ¼ 1; vps ¼ n; n [ 1 ) vs ¼ n n .
Fig. 6.6 Closed loop functioning oscillator with transformer T1 and Hybrid Pi model (version A)
v
vp ¼ v0 ¼ vA2 ; vs ¼ vi vA3 ; vi vA3 ¼ vn 0 ¼ nA2 (Reference directions are
in the top of each transformer’s coil). np ip þ ns is ¼ 0; ip ¼ ip ðtÞ; is ¼ is ðtÞ;
ip ðtÞ
is ðtÞ¼ nnps .
For all t and all currents ip and is . The voltage vp across coil 1 does not depend
on ip or on is ; it depends only on vs . Similarly the current ip depends only on is and
is independent of vp ; vs (Fig. 6.6).
Remark: When we have two coils of wire in close physical proximity to one
another, no importance whether or not the coils are wrapped around a common core
of magnetic material. We assume that the coils do not move with respect to one
another or with respect to a core they might be wrapped around. If we have some
ferromagnetic material in the magnetic circuit of the two coils, then when the
current sufficiently large, the relation between the fluxes /p ; /s and the currents
ip ; is are no longer linear. In this case the equations have the following form:
/p ¼ fp ðip ; is Þ; /s ¼ fs ðip ; is Þ, where fp ðip ; is Þ; fs ðip ; is Þ are nonlinear functions of the
currents ip ; is , By Faraday’s law we get the following:
d/p @fp dip @fp dis

vp ¼ ¼ þ ; fp ¼ fp ðip ; is Þ; fs ¼ fs ðip ; is Þ
dt @ip dt @is dt
d/s @fs dip @fs dis

vs ¼ ¼ þ ; fp ¼ fp ðip ; is Þ; fs ¼ fs ðip ; is Þ
dt @ip dt @is dt
@f @f @fs @fs
The four partial derivatives ð@ipp ; @ips ; @i ; Þ are function of ip ; is (Fig. 6.7).
p @is
Fig. 6.7 Closed loop

functioning oscillator’s T1
transformer circuit
dðVA1 VA4 Þ VA1 VA4 VA dVA2

ICpi ¼ Cpi ; Ir pi ¼ ; IRL ¼ 2 ; ICL ¼ CL ;
dt r pi RL dt
vbe ¼ VA1 VA4
VA4 dVA4 VA VA dVA3

IRe ¼ ; ICe ¼ Ce ; IR1 ¼ 3 ; IR2 ¼ 3 ; IC1 ¼ C1
Re dt R1 R2 dt
KCL at nodes A1, A2, A3, and A4 gives the below results (Table 6.2):
np ip þ ns is ¼ 0; np ¼ n; ns ¼ 1; n ip þ is ¼ 0 ) is ¼ n ip ; n [ 1
dðVA1 VA4 Þ VA1 VA4

ICpi ¼ Cpi ; Ir pi ¼ ) VA1 VA4 ¼ Ir pi r pi;
dt r pi
dðIr pi r piÞ
ICpi ¼ Cpi
dt
dIr pi VA dðIRL RL Þ
ICpi ¼ Cpi r pi ; IRL ¼ 2 ) VA2 ¼ IRL RL ; ICL ¼ CL
dt RL dt
Table 6.2 Closed loop KCL @ node Expression

Hybrid Pi model microwave
RF oscillator KCL@ nodes A1 is þ ICpi þ Ir pi ¼0
and expressions A2 gm vbe þ IRL þ ICL ip ¼ 0
A3 IR2 þ IC1 þ IR1 is ¼ 0
A4 gm vbe þ Ir pi þ ICpi ¼ IRe þ ICe
dIRL VA dVA4 dðIRe Re Þ

I CL ¼ C L R L ; IRe ¼ 4 ) VA4 ¼ IRe Re ; ICe ¼ Ce ¼
dt Re dt dt
dIRe VA VA
ICe ¼ Ce Re ; IR1 ¼ 3 ) VA3 ¼ IR1 R1 ; IR2 ¼ 3 ) VA3 ¼ IR2 R2
dt R1 R2
dVA3 dðIR1 R1 Þ dðIR2 R2 Þ dIR

IC1 ¼ C1 ¼ C1 ¼ C1 ; IC1 ¼ C1 R1 1 ;
dt dt dt dt
dIR2
IC1 ¼ C1 R2
dt
v p np
¼ ¼ n ) vp ¼ vs n; vs ¼ vi VA3 ¼ VA1 VA3 ; vp ¼ VA2 ; VA2
v s ns
¼ ðVA1 VA3 Þ n
IRL RL ¼ ðVA1 IR1 R1 Þ n; IRL RL ¼ ðVA1 IR2 R2 Þ n; Ir pi

VA IRe Re
¼ 1
r pi
VA1 IRe Re
Ir pi ¼ ) VA1 ¼ Ir pi r pi þ IRe Re ; IRL RL
r pi
¼ ðIr pi r pi þ IRe Re IR2 R2 Þ n
r pi Re RL
IR2 ¼ Ir pi þ IRe þ IR L ; is ¼ n ip ; n [ 1
R2 R2 n R2
KCL @ node 1: is þ ICpi þ Ir pi ¼0
dIr pi dIr pi n 1
n ip þ Cpi r pi þ Ir pi ¼0) ¼ ip Ir pi
dt dt Cpi r pi Cpi r pi
KCL @ node 2: gm vbe þ IRL þ ICL ip ¼ 0; vbe ¼ VA1 VA4 ; vbe ¼ Ir pi r pi
dIRL dIRL
gm I r pi r pi þ IRL þ CL RL ip ¼ 0 )
dt dt
gm r pi 1
¼ Ir pi
CL RL CL RL
1
IRL þ ip
CL RL
KCL @ node 3: IR2 þ IC1 þ IR1 is ¼ 0
r pi Re RL dIR
IR 1 þ Ir pi þ IRe þ IR L þ C1 R1 1 þ n ip ¼ 0
R2 R2 n R2 dt
dIR1 r pi Re RL
C1 R1 ¼ IR1 Ir pi IRe IRL n ip
dt R2 R2 n R2
dIR1 1 r pi Re RL
¼ IR1 Ir pi IRe IRL
dt C1 R1 Q
2 Q
2 Q
2
C1 Rk C1 Rk n C1 Rk
k¼1 k¼1 k¼1
n
ip
C1 R1
KCL @ node 4: gm vbe þ Ir pi þ ICpi ¼ IRe þ ICe
dIr pi dIR
gm I r pi r pi þ Ir pi þ Cpi r pi ¼ IRe þ Ce Re e
dt dt
n 1
gm Ir pi r pi þ Ir pi þ Cpi r pi ½ ip Ir pi
Cpi r pi Cpi r pi
dIRe
¼ IRe þ Ce Re
dt
dIRe gm r pi n 1
¼ Ir pi þ ip IRe
dt Ce Re Ce Re Ce Re
dIr pi n 1
¼ ip Ir pi
dt Cpi r pi Cpi r pi
dIRL gm r pi 1 1
¼ Ir pi IRL þ ip
dt CL RL CL RL CL RL
dIR1 1 r pi Re RL
dt C1 R1 Q
2 Q
2 Q
2
C1 Rk C1 Rk n C1 Rk
k¼1 k¼1 k¼1
n
ip
C1 R1
dIRe gm r pi n 1
¼ Ir pi þ ip IRe
dIr dIRL dIR1
At fixed points (equilibrium points): dt
pi
¼ 0; dt ¼ 0; dt ¼ 0; dIdtRe ¼ 0
gm r pi 1 1
Ir ¼ n ip ; I IRL þ i ¼ 0
pi
CL RL r pi
CL RL CL RL p
r pi Re RL
IRL ¼ ip ð1 gm r pi nÞ; IR1 Ir IRe IRL n ip ¼ 0
pi
R2 R2 n R2
Re r pi RL
IR1 þ IRe ¼ ip ½n þ ð1 gm r pi nÞ þ n
R2 R2 n R2
gm r pi n 1
I þ i I ¼ 0
Ce Re r pi
Ce Re p Ce Re Re
gm r pi n 1
n ip þ ip I ¼ 0 ) IRe ¼ ip n ½gm r pi þ 1
Ce Re Ce Re Ce Re Re
Re r pi RL
IR1 þ IRe ¼ ip ½n þ ð1 gm r pi nÞ þ n
R2 R2 n R2
r pi RL Re
IR1 ¼ ip fn þ ð1 gm r pi nÞ þ n þ n ½gm r pi þ 1 g
R2 n R2 R2
We can summery our system fixed points: Eðip ; Ir

pi ; IRL ; IR1 ; IRe Þ
Eðip ; Ir
¼ fip ; n ip ; ip ð1 gm r pi nÞ;
r pi RL Re
ip fn þ ð1 gm r pi nÞ þ n þ n ½gm r pi þ 1 g;
R2 n R2 R2
ip n ½gm r pi þ 1g
Eðip ; Ir
¼ ip f1; n; ð1 gm r pi nÞ;
r pi RL Re
fn þ ð1 gm r pi nÞ þ n þ n ½gm r pi þ 1 g;
R2 n R2 R2
n ½gm r pi þ 1g
equilibrium points of the closed loop functioning oscillator circuit (BJT transistor
small signal model) consists in adding to coordinate ½ip ; Ir pi ; IRL ; IR1 ; IRe arbitrarily
small increments of exponentially form ½i0p ; ir pi ; iRL ; iR1 ; iRe ekt and retaining the
first order terms in ip ; Ir pi ; IRL ; IR1 ; IRe . The system of homogenous equations leads
to a polynomial characteristic equation in the eigenvalues [4]. The polynomial
characteristic equations accept by set of the below circuit variables, circuit variables
closed loop functioning oscillator circuit [2, 3, 4]. Our closed loop functioning
oscillator circuit fixed values with arbitrarily small increments of exponential form
½i0p ; ir pi ; iRL ; iR1 ; iRe ekt are: j = 0 (first fixed point), j = 1 (second fixed point),
j = 2(third fixed point), etc.,
ðjÞ ðjÞ
ip ðtÞ ¼ ipðjÞ þ i0p ekt ; Ir pi ðtÞ ¼ Ir pi þ ir pi ekt ; IRL ðtÞ ¼ IRL þ iRL ekt ; IR1 ðtÞ
ðjÞ
¼ IR1 þ iR1 ekt
ðjÞ dIr pi ðtÞ dIRL ðtÞ

IRe ðtÞ ¼ IRe þ iRe ekt ; ¼ ir pi k ekt ; ¼ iRL k ekt ;
dt dt
dIR1 ðtÞ
¼ iR1 k ekt
dt
dIRe ðtÞ
¼ iRe k ekt
dt
dIr pi n 1
¼ ip Ir pi
dt Cpi r pi Cpi r pi
n 1 ðjÞ
ir pi k ekt ¼ ½iðjÞ þ i0p ekt ½I þ ir pi ekt
Cpi r pi p Cpi r pi r pi
n 1 ðjÞ n
ir pi k ekt ¼ iðjÞ I þ i0 ekt
Cpi r pi p Cpi r pi r pi Cpi r pi p
1
ir pi ekt
Cpi r pi
ðjÞ ðjÞ
At fixed point n
Cpi r pi ip Cpi r1 pi Ir pi ¼0
n 1
i0 ir pi ir pi k¼0
Cpi r pi p Cpi r pi
dIRL gm r pi 1 1
¼ Ir pi IRL þ ip
dt CL RL CL RL CL RL
gm r pi ðjÞ 1 ðjÞ 1
iRL k ekt ¼ ½Ir pi þ ir pi ekt ½I þ iRL ekt þ
CL RL CL RL RL CL RL
½ip þ ip ekt
ðjÞ 0
gm r pi ðjÞ 1 ðjÞ 1
iRL k ekt ¼ Ir pi IR L þ iðjÞ
CL RL CL RL CL RL p
gm r pi 1 1
ir pi ekt iR ekt þ i0 ekt
CL RL CL RL L CL RL p
ðjÞ ðjÞ ðjÞ
At fixed point gCm rL RLpi Ir pi CL1RL IRL þ 1
CL RL ip ¼ 0
gm r pi 1 1
ir pi þ i0 iR iRL k ¼ 0
CL RL CL RL p CL RL L
dIR1 1 r pi Re RL
dt C1 R1 Q
2 Q
2 Q
2
C1 Rk C1 Rk n C1 Rk
k¼1 k¼1 k¼1
n
ip
C1 R1
1 ðjÞ ðjÞ r pi
iR1 k ekt ¼ ½I þ iR1 ekt ½Ir pi þ ir pi ekt
C1 R1 R1 Q
2
C1 Rk
k¼1
ðjÞ Re ðjÞ RL
½IRe þ iRe ekt ½IRL þ iRL ekt
Q
2 Q
2
C1 Rk n C1 Rk
k¼1 k¼1
n
½iðjÞ þ i0p ekt
C1 R1 p
1 ðjÞ 1 ðjÞ r pi
iR1 k ekt ¼ IR1 iR ekt Ir pi
C1 R1 C1 R1 1 Q
2
C1 Rk
k¼1
r pi ðjÞ Re Re
ir pi ekt IRe iRe ekt
Q
2 Q
2 Q
2
C1 Rk C1 Rk C1 Rk
k¼1 k¼1 k¼1
ðjÞ RL RL
IRL iRL ekt
Q
2 Q
2
n C1 Rk n C1 Rk
k¼1 k¼1
n n
iðjÞ i0 ekt
C1 R1 p C1 R1 p
1 ðjÞ ðjÞ r pi ðjÞ Re ðjÞ RL
iR1 k ekt ¼ I Ir pi IR e IR L
C1 R1 R1 Q
2 Q
2 Q
2
C1 Rk C1 Rk n C1 Rk
k¼1 k¼1 k¼1
n 1 r pi
iðjÞ iR ekt ir pi ekt
C 1 R1 p C 1 R1 1 Q
2
C1 Rk
k¼1
Re kt RL n
iRe e iRL ekt i0 ekt
Q
2 Q
2 C 1 R1 p
C1 Rk n C1 Rk
k¼1 k¼1
At fixed point
1 ðjÞ ðjÞ r pi ðjÞ Re

I Ir pi IRe
C1 R1 R1 Q
2 Q
2
C1 Rk C1 Rk
k¼1 k¼1
ðjÞ RL n
IRL iðjÞ ¼ 0
Q
2 C1 R1 p
n C1 Rk
k¼1
1 r pi Re
iR iR1 k ir pi iRe
C1 R1 1 Q
2 Q
2
C1 Rk C1 Rk
k¼1 k¼1
RL n
iRL i0 ¼ 0
Q
2 C1 R1 p
n C1 Rk
k¼1
dIRe gm r pi n 1
¼ Ir pi þ ip IRe
gm r pi ðjÞ n 1
iRe k ekt ¼ ½Ir pi þ ir pi ekt þ ½ipðjÞ þ i0p ekt
Ce Re Ce Re Ce Re
ðjÞ
½IRe þ iRe ekt
gm r pi ðjÞ gm r pi n n
iRe k ekt ¼ Ir pi þ ir pi ekt þ ipðjÞ þ i0 ekt
C e Re Ce Re Ce Re Ce Re p
1 ðjÞ 1
IRe iR ekt
C e Re C e Re e
gm r pi ðjÞ n 1 ðjÞ
iRe k ekt ¼ I þ iðjÞ I
Ce Re r pi Ce Re p Ce Re Re
gm r pi n 1
þ ir pi ekt þ i0p ekt iR ekt
Ce Re Ce Re Ce Re e
gm r pi ðjÞ ðjÞ ðjÞ

At fixed point Ce Re Ir pi þ n
Ce Re ip Ce1Re IRe ¼ 0
gm r pi n 1
ir pi þ i0 iR iRe k ¼ 0
Ce Re Ce Re p Ce Re e
We can summery our arbitrarily small increments equations:
n 1
i0p ir pi ir pi k¼0
Cpi r pi Cpi r pi
gm r pi 1 1
ir pi þ i0 iR iRL k ¼ 0
CL RL CL RL p CL RL L
1 r pi Re
iR iR1 k ir pi iRe
C1 R1 1 Q
2 Q
2
C1 Rk C1 Rk
k¼1 k¼1
RL n
iRL i0 ¼ 0
Q
2 C1 R1 p
n C1 Rk
k¼1
gm r pi n 1
ir pi þ i0 iR iRe k ¼ 0
Ce Re Ce Re p Ce Re e
0 n 1
0 1 Cpi r pi
N11 ... N14 B 1 C
B . .. .. C B C R C 0
B . C B L L C i ¼ 0;
@ . . . A ir pi iRL iR1 iRe þ B n C p
@ C1 R1 A
N41 N44 n
Ce Re
0 1 0 ir pi 1
N11 ... N14
B . B i C
B . .. .. C
CB RL C
@ . . . A B C0
@ iR1 A
N41 N44 i Re
1 gm r pi
N11 ¼ k; N12 ¼ N13 ¼ N14 ¼ 0; N21 ¼ ;
Cpi r pi CL RL
1
N22 ¼ k
CL RL
r pi RL 1
N23 ¼ N24 ¼ 0; N31 ¼ ; N32 ¼ ; N33 ¼ k
Q
2 Q
2 C1 R1
C1 Rk n C1 Rk
k¼1 k¼1
Re gm r pi 1
N34 ¼ ; N41 ¼ ; N42 ¼ N43 ¼ 0; N44 ¼ k
Q
2 Ce Re Ce Re
C1 Rk
k¼1
Assumption: Arbitrarily small increment i0p is very small compare to other

system arbitrarily small increments ðir pi ; iRL ; iR1 ; iRe Þ.
0 1
n
Cpi r pi
0 1
B 1 C N11 . . . N14
B C R C 0 B . .. .. C
B L L C i ! e; ðA k IÞ ¼ B . C
B n C p @ . . . A;
@ C1 R1 A
n N41 N44
Ce Re
0 1
N11 . . . N14
B . .. .. C
detðA k IÞ ¼ detB
@ .. .
C
. A
N41 N44
1 1 1 1
detðA k IÞ ¼ k4 þ k3 ½ þ þ þ
C1 R1 Ce Re Cpi r pi CL RL
1 1 1 1 1
þ k2 ½ þð þ Þð þ Þ
C1 R1 Ce Re Cpi r pi CL RL C 1 R1 C e Re
1 1 1 1
þ þ k f½ þ
Cpi r pi CL RL Cpi r pi CL RL C1 R1 Ce Re
1 1 1
þ½ þ g
1
þ
Cpi r pi CL RL C1 R1 Ce Re
1 1 1 1
!4 ¼ 1; !3 ¼ þ þ þ
1 1 1
!2 ¼ þð þ Þ
1 1 1
ð þ Þþ
1 1 1 1 1
!1 ¼ ½ þ þ½ þ
Cpi r pi CL RL C1 R1 Ce Re C1 R1 Ce Re
1

Cpi r pi CL RL
1 X
4
!0 ¼ ; detðA k IÞ ¼ !k kk
Cpi r pi CL RL C1 R1 Ce Re k¼0
Eigenvalues stability discussion: Our closed loop functioning oscillator circuit

(BJT transistor small signal model) involving N variables (N > 2, N = 5, arbitrarily
small increments), the characteristic equation is of degree N = 4 (we exclude small
increment i0p ) and must often be solved numerically. Expect in some particular
cases, such an equation has (N = 4) distinct roots that can be real or complex. These
values are the eigenvalues of the 4 × 4 Jacobian matrix (A).
The general rule is that the closed loop functioning oscillator circuit is stable if
there is no eigenvalue with positive real part. It is sufficient that one eigenvalue is
positive for the steady state to be unstable. Our 5-variables ðip ; Ir pi ; IRL ; IR1 ; IRe Þ
system has four eigenvalues (reduce to four system arbitrarily small increments).
The type of behavior can be characterized as a function of the position of these
eigenvalues in the Re/Im plane. Five non-degenerated cases can be distinguished:
(1) the four eigenvalues are real and negative (stable steady state), (2) the four
eigenvalues are real, two of them are negative (unstable steady state), (3) and
(4) two eigenvalues are complex conjugates with a negative real part and other two
real are negative (stable steady state), two cases can be distinguished depending on
the relative value of the real part of the complex eigenvalues and of the real one,
(5) two eigenvalues are complex conjugates with a negative real part and the other
two real are positive (unstable steady state) [2, 3].
X
4 X
4
k¼0 k¼0
We suppose k1 ; k2 ; k3 ; k4 all (known) eigenvalues of the linearized system.

We are interesting to get oscillations from the closed loop functioning oscillator
circuit. To get oscillations, we need eigenvalue of A is on imaginary axis, and then
the method of the center manifold is applied. The center manifold of an equilibrium
point (fixed point) of our dynamical system consists of orbits whose behavior
around the equilibrium point is not controlled by either the attraction of the stable
manifold or the repulsion of the unstable manifold. Closed loop functioning
oscillator circuit eigenvalues which corresponding to eigenvalues with negative real
part form the stable Eigen space, which gives rise to the stable manifold.
Eigenvalue with positive real part yield the unstable manifold. If there are eigen-
values whose real part is zero, then these give rise to the center manifold. If the
eigenvalues are precisely zero, then these more specifically give rise to a slow
manifold.
6.3 Hartley Oscillator Stability Analysis
There are types of transistor oscillators which use feedback and lumped inductance
and capacitance resonators. These oscillators are like their low frequency coun-
terparts and very small values of inductance and capacitance must be used to make
them resonate at microwave frequencies. A common base configuration is used for
oscillators, although common emitter configurations can be used. The transistor
input and output are matched and feedback is supplied from the output to the input.
The feedback circuit contains the resonator, which controls the oscillation fre-
quency. There are three types of transistor LC oscillators, Colpitts, Hartley and
Clapp. These oscillators differ only in the way that the feedback is applied. In the
6.3 Hartley Oscillator Stability Analysis 631
Hartley oscillator, the feedback is supplied by the inductive divider formed by two
inductors. The advantages of the Hartley oscillator are oscillator’s frequency may
be adjusted using a single variable capacitor (one side of which can be earthed),
oscillator’s output amplitude remains constant over the frequency range, either a
tapped coil or two fixed inductors are needed, and easy to create an accurate fixed
frequency crystal oscillator variation (it is done by replacing the capacitor with a
quartz crystal). The disadvantage of Hartley oscillator is the fact that harmonic rich
output if taken from the amplifier and not directly from the LC circuit. Basic LC
oscillator circuits have no means of controlling the amplitude of the oscillations and
it is difficult to tune the oscillator to the required frequency. Hartley Oscillator
configuration has a tuned tank circuit with its resonant coil tapped to feed a fraction
of the output signal back to the emitter of the transistor. The output of the transistors
emitter is always “in-phase” with the output at the collector, this feedback signal is
positive. The oscillating frequency which is a sine-wave voltage is determined by
the resonance frequency of the tank circuit [9, 10] (Fig. 6.8).
Resistor R1 provide the usual stabilizing DC bias for the transistor. L1 is the Radio
Frequency Coil (RFC). It is an RF Choke which has a high reactance at the fre-
quency of oscillations so most of the RF current is applied to the LC tanking circuit
Fig. 6.8 Hartley oscillator circuit

via capacitor C2 and the DC component passes through L1 to the power supply. We
use for our stability analysis the BJT NPN transistor Small Signal (SS) model
equivalent circuit (Hybrid Pi model Version A). We consider “AC ground” in the
circuit. Since the voltage at this terminal is held constant at Vcc, there is no time
variation of the voltage. Consequently, we can set this terminal to be an “AC
ground” in the small signal circuit. For AC ground, we “kill” the DC sources at that
terminal (short circuit voltage sources and open circuit current sources). (Fig. 6.9).
VL1 ¼ VA1 ; VL3 ¼ VA3 ; VL2 ¼ VA4 ; VC3 ¼ VA4 VA3 ; vbe ¼ VA2 ;
dIL
VA1 ¼ VL1 ¼ L1 1 ; VC2 ¼ VA1 VA4
dt
dðVA1 VA4 Þ dIL
I C2 ¼ C 2 ; VL2 ¼ VA4 ¼ L2 2 ;
dt dt
dIL3
V L3 ¼ VA3 ¼ L3 ; VC1 ¼ VA2 VA3
dt
Fig. 6.9 Hartley oscillator Hybrid Pi model (Version A) schematic

dðVA2 VA3 Þ dVA2

IC1 ¼ C1 ; VCpi ¼ VA2 ; ICpi ¼ Cpi ;
dt dt
VA VA VA2
Ir pi ¼ 2 ; VR1 ¼ VA1 VA2 ; IR1 ¼ 1
r pi R1
KCL @ node A1: IR1 þ gm vbe þ IC2 þ IL1 ¼ 0, KCL @ node A2:
IR1 ¼ IC1 þ ICpi þ Ir pi
KCL @ node A3: IC1 þ IC3 ¼ IL3 , KCL @ node A4: IC2 ¼ IL2 þ IC3
Z
dIL1 dðVA1 VA4 Þ 1
VA1 ¼ VL1 ¼ L1 ; IC2 ¼ C2 ) VA1 VA4 ¼ IC2 dt;
dt dt C2
Z
1
VA4 ¼ VA1 IC2 dt
C2
Z
d 1 dVA4 dVA1 1
fVA4 ¼ VA1 IC2 dtg ) ¼ I C2 ;
dt C2 dt dt C2
Z
dIL 1
VA4 ¼ L1 1 IC2 dt
dt C2
Z
dIL2 dIL dIL 1 d 2 I L2
VA4 ¼ L2 ) L2 2 ¼ L1 1 IC2 dt; L2
dt dt dt C2 dt2
2
d IL 1
¼ L1 2 1 I C2
dt C2
Z
dIL3 dðVA2 VA3 Þ 1
VA3 ¼ L3 ; IC1 ¼ C1 ) VA2 VA3 ¼ IC1 dt;
dt dt C1
Z
1
VA2 ¼ VA3 þ IC1 dt
C1
Z Z
dIL3 1 dVA2 1
VA2 ¼ L3 þ IC1 dt; ICpi ¼ Cpi ) VA2 ¼ ICpi dt
dt C1 dt Cpi
Z Z Z
1 dIL3 1 d 1
ICpi dt ¼ L3 þ IC dt; f ICpi dt
Cpi dt C1 Z 1 dt Cpi
dIL3 1
¼ L3 þ IC1 dtg
dt C1
1 d 2 IL 1 VA2
ICpi ¼ L3 2 3 þ I C1 ; I r pi ¼ ) VA2 ¼ Ir pi r pi;
Cpi dt C1 r pi
VA VA2
IR1 ¼ 1
R1
VA1 VA2 ¼ IR1 R1 ) VA1 ¼ VA2 þ IR1 R1 ; VA1 ¼ Ir pi r pi þ IR1 R1 ;

VA1 ¼ Ir pi r pi þ IR1 R1
dIL1 dIL dIL1 r pi R1

VA1 ¼ L1 ; L1 1 ¼ Ir pi r pi þ IR1 R1 ; ¼ Ir pi þ IR1
dt dt dt L1 L1
d 2 IL1 dIr pi r pi dIR1 R1 d 2 IL2

¼ þ ; L2
dt2 dt L1 dt L1 dt2
dIr pi r pi dIR1 R1 1
¼ L1 ½ þ IC2
dt L1 dt L1 C2
d 2 IL2 dIr pi dIR1 1

L2 ¼ r pi þ R1 IC2 ;
dt2 dt dt C2
d 2 IL2 dIr pi r pi dIR1 R1 1
¼ þ I C2
dt2 dt L2 dt L2 C2 L2
We can summery our Hartley oscillator differential equation:
d 2 IL2 dIr pi dIR1 1

L2 2
¼ r pi þ R1 I C2
dt dt dt C2
dIL1 r pi R1 1 d 2 IL 1
¼ Ir pi þ IR1 ; ICpi ¼ L3 2 3 þ I C1
dt L1 L1 Cpi dt C1
dIL2 dIr
We define for simplicity the following new variables: dt ¼ Y1 ; dt
pi
¼ Y2
dIR1 dIL d 2 IL dY1 d 2 IL3 dY4

¼ Y3 ; 3 ¼ Y4 ; 2 2 ¼ ; ¼
dt dt dt dt dt2 dt
dY1 r pi R1 1 dY4 1 1
¼ Y2 þ Y3 I C2 ; ¼ ICpi I C1
dt L2 L2 C2 L2 dt Cpi L3 C1 L3
Our circuit KCLs @ nodes A1 – A4: IR1 þ gm Ir pi r pi þ IC2 þ IL1 ¼ 0
IR1 ¼ IC1 þ ICpi þ Ir pi ; IC1 þ IC3 ¼ IL3 ; IC2 ¼ IL2 þ IC3
IC2 ¼ IL1 gm r pi Ir pi IR1 ; ICpi ¼ IR1 IC1 Ir pi ; IC1 ¼ IL3 IC3 ; IC3 ¼ IC2 IL2
IC1 ¼ IL3 ½IC2 IL2 ¼ IL3 IC2 þ IL2 ; IC1 ¼ IL3 þ IL2 ½IL1 gm r pi Ir pi IR1
X
3
IC1 ¼ IL3 þ IL2 þ IL1 þ gm r pi Ir pi þ IR1 ; IC1 ¼ ILk þ gm r pi Ir pi þ IR1
k¼1
X3
ICpi ¼ IR1 Ir pi ½ ILk þ gm r pi Ir pi þ IR1 ;
k¼1
X
3
ICpi ¼ Ir pi ½1 þ gm r pi I Lk
k¼1
We can summery our intermediate results:

X
3 X
3
ICpi ¼ Ir pi ½1 þ gm r pi I Lk ; I C1 ¼ ILk þ gm r pi Ir pi þ IR 1
k¼1 k¼1
IC2 ¼ IL1 gm r pi Ir pi IR1
&&&
dY1 r pi R1 1
¼ Y2 þ Y3 I C2
dt L2 L2 C2 L2
r pi R1 1
¼ Y2 þ Y3 ½IL1 gm r pi Ir pi IR1
L2 L2 C2 L2
dY1 r pi R1 1 1 1
¼ Y2 þ Y3 þ I L1 þ gm r pi Ir pi þ IR1
dt L2 L2 C2 L2 C2 L2 C2 L2
dY4 1 X
3
¼ ½Ir pi ½1 þ gm r pi I Lk
dt Cpi L3 k¼1
1 X
3
½ IL þ gm r pi Ir pi þ IR1
C1 L3 k¼1 k
dY4 Ir pi ½1 þ gm r pi gm r pi 1 1 1 X 3
¼ ð þ Þ½ þ IL
dt L3 Cpi C1 Cpi C1 L3 k¼1 k
1
IR1
C1 L3

dY1 r pi R1 1 1 1
¼ Y2 þ Y3 þ I L1 þ gm r pi Ir pi þ IR1
dt L2 L2 C2 L2 C2 L2 C2 L2
¼ ð þ Þ½ þ IL
1
IR1
C1 L3
dIL2 dIr pi dIR dIL dIL r pi R1

¼ Y1 ; ¼ Y2 ; 1 ¼ Y3 ; 3 ¼ Y4 ; 1 ¼ Ir pi þ IR1
dt dt dt dt dt L1 L1
At fixed points:
dY1 dY4 dIL dIr pi dIR dIL dIL

¼ 0; ¼ 0; 2 ¼ 0; ¼ 0; 1 ¼ 0; 3 ¼ 0; 1 ¼ 0
R1
Y1 ¼ 0; Y2 ¼ 0; Y3 ¼ 0; Y4 ¼ 0; r pi Ir pi þ IR1 R1 ¼ 0 ) Ir pi ¼ IR1
r pi
1 1 1
I þ gm r pi Ir þ I ¼ 0
C2 L2 L1 C2 L2 pi
C2 L2 R1
Ir pi ½1 þ gm r pi gm r pi 1 1 1 X 3
1
ð þ Þ½ þ I I ¼ 0
L3 Cpi C1 Cpi C1 L3 k¼1 Lk C1 L3 R1
We get two fixed points equations:
IL1 þ IR1 ð1 gm R1 Þ ¼ 0
R1 ½1 þ gm r pi gm r pi 1 1 1 1 X 3
IR1 f ð þ Þ g½ þ I
r pi L3 Cpi C1 C1 L3 Cpi C1 L3 k¼1 Lk
¼0
R1 ½1 þ gm r pi gm r pi 1 1 1 1
IR1 f ð þ Þ g½ þ
r pi L3 Cpi C1 C1 L 3 Cpi C1 L3
X3
½IL1 þ ILk ¼ 0
k¼2
X
3 X
3
ILk ¼ IL1 þ ILk ; IL1 þ IR1 ð1 gm R1 Þ ¼ 0 ) IL1 ¼ IR1 ð1 gm R1 Þ
k¼1 k¼2
R1 ½1 þ gm r pi gm r pi 1 1 1 1
IR1 f ð þ Þ g½ þ IL1
r pi L3 Cpi C1 C1 L3 Cpi C1 L3
1 1 1 X 3
½ þ I ¼0
Cpi C1 L3 k¼2 Lk
R1 ½1 þ gm r pi gm r pi 1 1 1 ð1 gm R1 Þ
IR1 f ð þ Þ gþ½ þ IR1
r pi L3 Cpi C1 C1 L 3 Cpi C1 L3
1 1 1 X 3
½ þ I ¼ 0
Cpi C1 L3 k¼2 Lk
R1 ½1 þ gm r pi gm r pi 1 1 1 ð1 gm R1 Þ
IR1 f ð þ Þ þ½ þ g
r pi L3 Cpi C1 C1 L3 Cpi C1 L3
1 1 1 X 3
½ þ I ¼ 0
Cpi C1 L3 k¼2 Lk
P
3
½C1pi þ C1
1
L13 ILk
IR1
k¼2
¼ R1
r piL3 ð½1 þ gCmpir pi
þ gm r pi
C1 Þ 1
C1 L3 þ ½C1pi þ C1
1
ð1gLm3 R1 Þ
P
3
ð1 gm R1 Þ ½C1pi þ C1
1
L13 ILk
IL1 ¼
k¼2
R1
þ gm r pi
C1 Þ C11L3 þ ½C1pi þ C1
1
ð1gLm3 R1 Þ
P
3
r R1pi ½C1pi þ C1
1
L13 ILk
Ir
k¼2
¼
ð½1 þ gCmpir gm r pi
ð1gLm3 R1 Þ
pi pi
R1
r piL3 þ C1 Þ C11L3 þ ½C1pi þ C1
1
We can define our system fixed points as E ðY1 ; Y2 ; Y3 ; Y4 ; Ir
pi ; IR1 ; IL1 ; IL2 ; IL3 Þ
We define three global parameters in our system: C1 ; C2 ; C3
C1 ¼ C1 ðR1 ; r pi; C1 ; Cpi ; . . .Þ; C2 ¼ C1 ðR1 ; r pi; C1 ; Cpi ; . . .Þ;

C3 ¼ C3 ðR1 ; r pi; C1 ; Cpi ; . . .Þ
½C1pi þ C1
1
L13
C1 ¼ R1
þ gm r pi
1
ð1gLm3 R1 Þ
ð1 gm R1 Þ ½C1pi þ C1
1
L13
C2 ¼ R1
þ gm r pi
1
ð1gLm3 R1 Þ
r R1pi ½C1pi þ C1
1
L13
C3 ¼ R1
þ gm r pi
1
ð1gLm3 R1 Þ
X
3 X
3
IR1 ¼ C1 ðR1 ; r pi; C1 ; Cpi ; . . .Þ ILk ; IL1 ¼ C2 ðR1 ; r pi; C1 ; Cpi ; . . .Þ ILk
k¼2 k¼2
X
3
Ir pi ¼ C3 ðR1 ; r pi; C1 ; Cpi ; . . .Þ ILk
k¼2
X
3
E ðY1 ; Y2 ; Y3 ; Y4 ; Ir
pi ; IR1 ; IL1 ; IL2 ; IL3 Þ ¼ ð0; 0; 0; 0; C3 ðR1 ; r pi; C1 ; Cpi ; . . .Þ ILk ;
k¼2
X
3 X
3
C1 ðR1 ; r pi; C1 ; Cpi ; . . .Þ ILk ; C2 ðR1 ; r pi; C1 ; Cpi ; . . .Þ ILk ; IL2 ; IL3 Þ
k¼2 k¼2
equilibrium points of the Hartley oscillator circuit (BJT transistor small signal
model) consists in adding to coordinate [Y1 ; Y2 ; Y3 ; Y4 ; Ir pi ; IR1 ; IL1 ; IL2 ; IL3 ] arbi-
trarily small increments of exponentially form ½y1 ; y2 ; y3 ; y4 ; ir pi ; iR1 ; iL1 ; iL2 ; iL3
ekt and retaining the first order terms in Y1 ; Y2 ; Y3 ; Y4 ; Ir pi ; IR1 ; IL1 ; IL2 ; IL3 . The
system of homogenous equations leads to a polynomial characteristic equation in
the eigenvalues [4]. The polynomial characteristic equations accept by set of the
below circuit variables, circuit variables derivative and circuit variables second
order derivative with respect to time into Hartley oscillator circuit [2, 3, 4]. Our
Hartley oscillator fixed values with arbitrarily small increments of exponential form
½y1 ; y2 ; y3 ; y4 ; ir pi ; iR1 ; iL1 ; iL2 ; iL3 ekt are: j = 0(first fixed point), j = 1(second
fixed point), j = 2(third fixed point), etc.,
ðjÞ ðjÞ
Ir pi ðtÞ ¼ Ir pi þ ir pi ekt ; IR1 ðtÞ ¼ IR1 þ iR1 ekt ;
ðjÞ ðjÞ
IL1 ðtÞ ¼ IL1 þ iL1 ekt ; IL2 ðtÞ ¼ IL2 þ iL2 ekt
ðjÞ ðjÞ ðjÞ

IL3 ðtÞ ¼ IL3 þ iL3 ekt ; Y1 ðtÞ ¼ Y1 þ y1 ekt ; Y2 ðtÞ ¼ Y2 þ y2 ekt ;
ðjÞ
Y3 ðtÞ ¼ Y3 þ y3 ekt
ðjÞ dY1 ðtÞ

Y4 ðtÞ ¼ Y4 þ y4 ekt ; ¼ y1 k ekt ;
dt
dY4 ðtÞ dIr pi ðtÞ
¼ y4 k ekt ; ¼ ir pi k ekt
dt dt
dIR1 ðtÞ dIL ðtÞ
¼ iR1 k ekt ; 1 ¼ iL1 k ekt ;
dt dt
dIL2 ðtÞ kt dIL3 ðtÞ
¼ i L2 k e ; ¼ iL3 k ekt
dt dt
&&&
dY1 r pi R1 1 1 1
¼ Y2 þ Y3 þ IL þ gm r pi Ir pi þ IR1
dt L2 L2 C2 L2 1 C2 L2 C2 L2
ðjÞ r pi ðjÞ R1 1 ðjÞ

y1 k ekt ¼ ½Y2 þ y2 ekt þ ½Y3 þ y3 ekt þ ½I þ iL1 ekt
L2 L2 C2 L2 L1
1 ðjÞ 1 ðjÞ
þ gm r pi ½Ir pi þ ir pi ekt þ ½I þ iR1 ekt
C2 L2 C2 L2 R1
r pi
ðjÞ ðjÞ R1 1 ðjÞ 1 ðjÞ
y1 k ekt ¼ Y2 þ Y3 þ I L1 þ gm r pi Ir pi
L2 L2 C2 L2 C2 L2
1 ðjÞ r pi kt R1 1
þ I þ y2 e þ y3 ekt þ iL ekt
C2 L2 R1 L2 L2 C2 L2 1
1 1
þ gm r pi ir pi ekt þ iR ekt
C2 L2 C2 L2 1
At fixed point:
ðjÞ r pi ðjÞ R1 1 ðjÞ 1 ðjÞ 1 ðjÞ

Y2 þ Y3 þ I L1 þ gm r pi Ir pi þ I ¼0
L2 L2 C2 L2 C2 L2 C2 L2 R1
r pi kt R1 1
y1 k ekt þ y2 e þ y3 ekt þ iL ekt
L2 L2 C2 L2 1
1 1
þ gm r pi ir pi ekt þ iR ekt ¼ 0
C2 L2 C2 L2 1
Dividing the two sides of the above equation by ekt term:
r pi R1 1 1
y1 k þ y2 þ y3 þ iL þ gm r pi ir pi
L2 L2 C2 L2 1 C2 L2
1
þ iR ¼ 0
C2 L2 1
¼ ð þ Þ½ þ IL
1
IR1
C1 L3
1 ðjÞ ½1 þ gm r pi gm r pi
y4 k ekt ¼ ½I ekt ð
þ ir pi þ Þ
L3 r pi Cpi C1
1 1 1 X 3
ðjÞ 1 ðjÞ
½ þ ðI þ iLk ekt Þ ½I þ iR1 ekt
Cpi C1 L3 k¼1 Lk C1 L3 R1
y4 k ekt ¼ Ir pi ð þ Þ
L3 Cpi C1
1 ½1 þ gm r pi gm r pi
ir pi ð þ Þ ekt
L3 Cpi C1
1 1 1 X 3
ðjÞ 1 1 1 X 3
½ þ I Lk ½ þ iL ekt
Cpi C1 L3 k¼1 Cpi C1 L3 k¼1 k
1 ðjÞ 1
IR1 iR ekt
C1 L3 C1 L3 1
y4 k ekt ¼ I ð þ Þ
L3 r pi Cpi C1
1 1 1 X 3
ðjÞ 1 ðjÞ
½ þ I I
Cpi C1 L3 k¼1 Lk C1 L3 R1
ir pi ð þ Þ ekt
L3 Cpi C1
1 1 1 X 3
1
½ þ iL ekt iR ekt
Cpi C1 L3 k¼1 k C1 L3 1
At fixed point:
1 ðjÞ ½1 þ gm r pi gm r pi 1 1 1 X 3
ðjÞ 1 ðjÞ
Ir pi ð þ Þ½ þ I I
L3 Cpi C1 Cpi C1 L3 k¼1 Lk C1 L3 R1
¼0
1 ½1 þ gm r pi gm r pi 1 1 1
y4 k ekt ¼ ir pi ð þ Þ ekt ½ þ
L3 Cpi C1 Cpi C1 L3
X3
1
iLk ekt iR1 ekt
k¼1
C 1 L 3
Dividing the two sides of the above equation by ekt term:
1 ½1 þ gm r pi gm r pi 1 1 1 X 3
y4 k ir pi ð þ Þ½ þ iL
L3 Cpi C1 Cpi C1 L3 k¼1 k
1
iR ¼ 0
C1 L3 1
dIL2 dIr pi dIR dIL

¼ Y1 ; ¼ Y2 ; 1 ¼ Y3 ; 3 ¼ Y4
dt dt dt dt
ðjÞ ðjÞ ðjÞ
iL2 k ekt ¼ Y1 þ y1 ekt ; ir pi k ekt ¼ Y2 þ y2 ekt ; iR1 k ekt ¼ Y3 þ y3 ekt
ðjÞ
iL3 k ekt ¼ Y4 þ y4 ekt
At fixed point:
Y1 ¼ 0; Y2 ¼ 0; Y3 ¼ 0; Y4 ¼ 0
iL2 k þ y1 ¼ 0; ir pi k þ y2 ¼ 0; iR1 k þ y3 ¼ 0; iL3 k þ y4 ¼ 0
dIL1 r pi R1
¼ Ir pi þ IR1 ; iL1 k ekt

dt L1 L1
ðjÞ kt r pi ðjÞ R1
¼ ½Ir pi þ ir pi e þ ½IR1 þ iR1 ekt
L1 L1
ðjÞ r pi ðjÞ R1 r pi kt R1

iL1 k ekt ¼ Ir pi þ IR1 þ ir pi e þ iR1 ekt
L1 L1 L1 L1
At fixed point :
ðjÞ r pi ðjÞ R1
Ir pi þ IR 1 ¼ 0:
L1 L1
We get iL1 k þ ir pi r L1pi þ iR1 RL11 ¼ 0.

We can summery our Hartley oscillator circuit (BJT transistor small signal
model) arbitrarily small increments equations:
r pi R1 1 1
y1 k þ y2 þ y3 þ iL þ gm r pi ir pi
L2 L2 C2 L2 1 C2 L2
1
þ iR ¼ 0
C2 L2 1
1 ½1 þ gm r pi gm r pi 1 1 1 X 3
y4 k ir pi ð þ Þ½ þ iL
L3 Cpi C1 Cpi C1 L3 k¼1 k
1
iR ¼ 0
C1 L3 1
iL2 k þ y1 ¼ 0; ir pi k þ y2 ¼ 0; iR1 k þ y3 ¼ 0; iL3 k þ y4 ¼ 0
r pi R1
iL1 k þ ir pi þ iR1 ¼0
L1 L1
0 1 0 r pi R1 1
y1 L2 L2
B y C B C
0 1 B
B
4 C
C B
B 0 0C
C
N11 ... N17 B ir pi C B 0C
B . B C B 1 C y
B . .. .. CCB C B C 2
@ . . . A B R1 C
B i þB 0 1 C ¼ 0;
C B C y3
B iL C B C
N71 N77 B 1 C B 0 0C
B C B C
@ i L2 A @ 0 0A
i L3 0 0
0 1
y1
B y C
0 1 B
B
4 C
C
N11 ... N17 B ir pi C
B . C B C
B . .. .. C B C
@ . . . AB
B
iR1 C 0
C
B iL C
N71 N77 B 1 C
B C
@ i L2 A
i L3
Assumption: arbitrarily small increments: y2 ! e; y3 ! e

0r pi R1
1
L2 L2
B C
B 0 0 C
B C
B C
B 1 0 C
B C y2
B 0 1 C ! e; N11 ¼ k; N12 ¼ 0;
B C y
B C 3
B 0 0C
B C
@ 0 0A
0 0
1 1
N13 ¼ gm r pi; N14 ¼
C2 L2 C2 L2
1
N15 ¼ ; N16 ¼ N17 ¼ 0; N21 ¼ 0; N22 ¼ k;
C2 L2
N23 ¼ ð þ Þ
L3 Cpi C1
1 1 1 1
N24 ¼ ; N25 ¼ N26 ¼ N27 ¼ ½ þ ; N31 ¼ N32 ¼ 0; N33 ¼ k
C1 L3 Cpi C1 L3
N34 ¼ N35 ¼ N36 ¼ N37 ¼ 0; N41 ¼ N42 ¼ N43 ¼ 0; N44 ¼ k; N45 ¼ N46 ¼ N47
¼0
r pi R1
N51 ¼ N52 ¼ 0; N53 ¼ ; N54 ¼ ; N55 ¼ k; N56 ¼ N57 ¼ 0; N61 ¼ 1; N62
L1 L1
¼0
N63 ¼ N64 ¼ N65 ¼ 0; N66 ¼ k; N67 ¼ 0; N71 ¼ 0; N72 ¼ 1; N73 ¼ N74 ¼ N75
¼ N76 ¼ 0
0 1
N11 . . . N17
B .. .. .. C
N77 ¼ k; ðA k IÞ ¼ @ . . . A ; detðA k IÞ
0 N
171 N77
N11 . . . N17
B .. .. .. C
¼ det@ . . . A
N71 N77
1 1 1
detðA k IÞ ¼ k7 k5 ½ þ ; detðA k IÞ
Cpi C1 L3
1 1 1
¼ k5 ðk2 þ ½ þ Þ
Cpi C1 L3
If an eigenvalue of A is on the imaginary axis, then the method of the center

manifold must be applied. To apply the stability criterion of Liapunov to our
system, the equation detðA k IÞ ¼ 0 has a zero in the left half plane, without
calculating the eigenvalues explicitly [2, 3].
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
1 1 1 1 1 1
k1 ¼ k2 ¼ . . . ¼ k5 ¼ 0; k þ ½ þ ¼ 0 ) k6;7 ¼ j
2
½ þ
Cpi C1 L3 Cpi C1 L3
6.4 Colpitts Oscillator Stability Analysis
The Colpitt’s oscillator is designed for generation of high frequency sinusoidal

oscillations (radio frequencies ranging from 10 kHz to 100 MHz). They are widely
used in commercial signal generators up to 100 MHz. Colpitt’s oscillator is same as
Hartley oscillator except for one difference. Instead of using a tapped inductance,
Colpitt’s oscillator uses a tapped capacitance. The circuit diagram of Colpitt’s
oscillator using BJT. It consists of an R-C coupled amplifier using an NPN tran-
sistor in CE configuration. R1 is resistor which forms a voltage bias to the transistor.
We can connect resistor RE which stabilizes the circuit against temperature varia-
tion (not in our circuit). If we connect RE resistor then a capacitor CE is connected
parallel with RE, acts as a bypass capacitor provides a low reactive path to the
amplified AC signal. The coupling capacitor C2 blocks DC and provides an AC
path from collector to the tank circuit [25, 26] (Fig. 6.10).
The feedback network (tank circuit) consists of two capacitors C3, C4 (in series)
which placed across a common inductor L2. The Centre of the two capacitors is
tapped (grounded). The feedback network (C3, C4, and L2) determines the frequency
of oscillation of the oscillator. The two capacitors C3, C4 form the potential divider
led for providing the feedback voltage. The voltage developed across the capacitor
Fig. 6.10 Colpitts oscillator circuit

6.4 Colpitts Oscillator Stability Analysis 645
Fig. 6.11 Colpitts oscillator Hybrid Pi model (Version A) circuit
C4 provides regenerative feedback which is essential for sustained oscillators. We use

for our stability analysis the BJT NPN transistor Small Signal (SS) model equivalent
circuit (Hybrid Pi model Version A). We consider “AC ground” in the circuit. Since
the voltage at this terminal is held constant at Vcc, there is no time variation of the
voltage. Consequently, we can set this terminal to be an “AC ground” in the small
signal circuit. For AC ground, we “kill” the DC sources at that terminal (short circuit
voltage sources and open circuit current sources) (Fig. 6.11)..
VA1 VA2 dIL VA2 dVCpi

IR1 ¼ ; VL1 ¼ VA1 ¼ L1 1 ; vbe ¼ VA2 ; Ir pi ¼ ; IC ¼ Cpi
R1 dt r pi pi dt
dVA2
¼ Cpi
dt

IC2 ¼ C2 ; IC4 ¼ C4 ; IC3 ¼ C3 ; VC3 ¼ VA4 ; VC4
dt dt dt
dIL2
¼ VA3 ; VL2 ¼ L2
dt
VL2 ¼ VA4 VA3 ; VC1 ¼ VA2 VA3 ; VA2 ¼ VCpi ; VR1 ¼ VA1 VA2 ; VL1 ¼ VA1 ; IC1
dðVA2 VA3 Þ
¼ C1
dt
KCL @ node A1: IR1 þ gm vbe þ IL1 þ IC2 ¼ 0; KCL @ node A2:
IR1 ¼ IC1 þ ICpi þ Ir pi
KCL @ node A3: IC1 þ IL2 ¼ IC4 ; KCL @ node A4: IC2 ¼ IC3 þ IL2
VA2 VA1 VA2

Ir pi ¼ ) VA2 ¼ Ir pi r pi; IR1 ¼ ) VA1 VA2 ¼ IR1 R1 ; VA1
r pi R1
dIL
¼ L1 1
dt
Z
dIL1 dVA2 1
L1 Ir pi r pi ¼ IR1 R1 ; ICpi ¼ Cpi ) VA2 ¼ ICpi dt
dt dt Cpi
Z
dðVA1 VA4 Þ 1 dVA3
I C2 ¼ C 2 ) VA1 VA4 ¼ IC2 dt; IC4 ¼ C4 ) VA3
Z dt C2 dt
1
¼ IC4 dt
C4
Z
dVA4 1 dIL
I C3 ¼ C3 ) VA4 ¼ IC3 dt; VA4 VA3 ¼ L2 2
Zdt C 3Z dt
1 1 dIL2
) IC3 dt IC4 dt ¼ L2
C3 C4 dt
Z Z
d 1 1 dIL2 1 1 d 2 IL
f IC3 dt IC4 dt ¼ L2 g) IC 3 I C4 ¼ L 2 2 2
dt C3 C4 dt C3 C4 dt
Z
dðVA1 VA4 Þ 1
IC2 ¼ C2 ) VA1 VA4 ¼ IC2 dt; VA1 VA4
Z dt C2
1
¼ IC2 dt
C2
Z Z Z
dIL 1 1 d dIL1 1
L1 1 I C3 dt ¼ IC dt; fL1 IC3 dt
dt C3 C2 Z 2 dt dt C3
1
¼ IC2 dtg
C2
d 2 I L1 1 1 dðVA2 VA3 Þ
L1 IC3 ¼ IC ; IC ¼ C1 ) VA2 VA3
dt2 C3 C2 Z 2 1 dt
1
¼ IC1 dt
C1
Z Z Z
1 1 d 1
Ir pi r pi IC4 dt ¼ IC dt; fI r pi r pi IC4 dt
C4 C1 Z 1 dt C4
1
¼ IC1 dtg
C1
dIr pi 1 1
r pi IC4 ¼ I C1
dt C4 C1
We can summery our Colpitt’s oscillator differential equations:
dIL1 1 1 d 2 IL
L1 Ir pi r pi ¼ IR1 R1 ; I C3 IC4 ¼ L2 2 2
dt C3 C4 dt
d 2 I L1 1 1 dIr pi 1 1
L1 2
I C3 ¼ I C2 ; r pi IC4 ¼ I C1
dt C3 C2 dt C4 C1
KCL @ nodes A1 – A4: IR1 þ gm Ir pi r pi þ IL1 þ IC2 ¼ 0; IR1 ¼ IC1 þ

ICpi þ Ir pi
IC1 þ IL2 ¼ IC4 ; IC2 ¼ IC3 þ IL2 :
dIL1 R1 r pi d 2 IL dIR1 R1 dIr pi r pi

¼ IR1 þ Ir pi ) 21 ¼ þ
dt L1 L1 dt dt L1 dt L1
d 2 I L1 1 1 dIR R1 dIr pi r pi
¼ I C2 þ I C3 ; 1 þ
dt2 C2 L1 C3 L1 dt L1 dt L1
1 1
¼ I C2 þ I C3
C2 L1 C3 L1
dIR1 dIr pi 1 1 dIr pi

R1 þ r pi ¼ I C2 þ IC3 ;
dt dt C2 C3 dt
1 1
¼ I C1 þ I C4
C1 r pi C4 r pi
dIR1 1 1 1 1
R1 þ ½ I C1 þ IC4 r pi ¼ I C2 þ I C3
dt C1 r pi C4 r pi C2 C3
dIR1 1 1 1 1
¼ I C2 þ IC3 I C1 IC4
dt C2 R1 C3 R1 C1 R1 C4 R1
d 2 IL2 1 1 dIL2 dY1 1 1

L2 ¼ I C3 I C4 ; ¼ Y1 ; ¼ I C3 I C4
dt2 C3 C4 dt dt C3 L2 C4 L2
We can summery our system differential equations (version 1):
dIR1 1 1 1 1 dIL
¼ I C2 þ I C3 I C1 I C4 ; 1
dt C2 R1 C3 R1 C1 R1 C4 R1 dt
R1 r pi
¼ IR 1 þ Ir pi
L1 L1
dIr pi 1 1 dIL2 dY1

¼ I C1 þ I C4 ; ¼ Y1 ;
dt C1 r pi C4 r pi dt dt
1 1
¼ I C3 I C4
C3 L2 C4 L2
KCL @ nodes A1 – A 4: IC2 ¼ IR1 gm Ir pi r pi IL1 ; IC1 ¼

IR1 ICpi Ir pi
X
2
IC4 ¼ IR1 ICpi Ir pi þ IL2 ; IC3 ¼ IR1 gm Ir pi r pi I Lk :
k¼1
dIR1 1
¼ ½IR1 gm Ir pi r pi IL1
dt C 2 R1
1 X
2
þ ½IR1 gm Ir pi r pi ILk
C3 R1 k¼1
1 1
½IR1 ICpi Ir pi ½IR1 ICpi Ir pi þ IL 2
C1 R1 C 4 R1
dIR1 1 1 1
¼ IR1 gm Ir pi r pi IL
dt C2 R1 C2 R1 C2 R1 1
1 1
IR1 gm Ir pi r pi
C3 R1 C3 R1
1 X 2
1 1 1
ILk IR1 þ ICpi þ Ir pi
C3 R1 k¼1 C1 R1 C1 R1 C1 R1
1 1 1 1
IR1 þ ICpi þ Ir pi IL
C4 R1 C4 R1 C4 R1 C4 R1 2
dIR1 1 1 1 1
¼ IR1 IR1 IR1 IR1
dt C2 R1 C1 R1 C3 R1 C4 R1
1 1 1 1
gm Ir pi r pi gm Ir pi r pi þ Ir pi þ Ir pi
C 2 R1 C3 R 1 C1 R 1 C4 R 1
1 1 1 1 1 X
2
þ ICpi þ ICpi IL IL ILk
C1 R1 C4 R1 C2 R1 1 C4 R1 2 C3 R1 k¼1
dIR1 X 4
1 1 X4
1 X3
1 1 X4
1 1
¼ ½ IR 1 þ ½ ð Þ gm r pi Ir pi þð Þ ICpi
dt C R1
k¼1 k
C
k¼1 k
C
k¼2 k
R1 C R1
k¼1 k
k6¼2 k6¼2
k6¼3 k6¼3
X
3
1 1 X
4
1 1 dIL R1 r pi
½ IL ½ IL ; 1 ¼ IR 1 þ Ir pi
k¼2
Ck R1 1 C R1 2 dt
k¼3 k
L1 L1
dIr pi 1 1
¼ ½IR1 ICpi Ir pi þ ½IR1 ICpi Ir pi þ I L2
dt C1 r pi C4 r pi
dIr pi X4
1 1 X4
1 1 X4
1 1
¼½ IR1 ½ ICpi ½
dt C r pi
k¼1 k
C r pi
k¼1 k
C r pi
k¼1 k
k6¼2 k6¼2 k6¼2
k6¼3 k6¼3 k6¼3
1
Ir pi þ IL
C4 r pi 2
dIL2 dY1 1 X
2
1
¼ Y1 ; ¼ ½IR1 gm Ir pi r pi ILk ½IR1 ICpi Ir pi þ IL2
dt dt C 3 L2 k¼1
C 4 L2
dY1 1 1 1 1
¼ IR1 IR1 gm Ir pi r pi þ Ir pi
dt C3 L2 C4 L2 C3 L2 C4 L2
1 1 1 1
I L1 IL2 I L2 þ ICpi
C3 L2 C3 L2 C4 L2 C4 L2
dY1 X4
1 1 1 1 1 1
¼ ½ IR1 þ ½ gm r pi Ir pi IL
dt C
k¼3 k
L 2 C4 C 3 L2 C3 L2 1
X4
1 1 1
½ I L2 þ ICpi
C
k¼3 k
L2 C 4 L2
dIR1 X4
1 1 X4
1 X3
1 1 X4
1 1
¼ ½ IR 1 þ ½ ð Þ gm r pi Ir pi þð Þ ICpi
dt k¼1
C k R1 k¼1
C k k¼2
C k R 1 k¼1
C k R 1
k6¼2 k6¼2
k6¼3 k6¼3
X3
1 1 X4
1 1
½ IL1 ½ IL
C R1
k¼2 k
C R1 2
k¼3 k
dIr pi X4
1 1 X4
1 1 X4
1 1
¼½ IR1 ½ ICpi ½
dt C r pi
k¼1 k
C r pi
k¼1 k
C r pi
k¼1 k
k6¼2 k6¼2 k6¼2
k6¼3 k6¼3 k6¼3
1
Ir pi þ IL
C4 r pi 2
dY1 X4
1 1 1 1 1 1
dt C L2
k¼3 k
C4 C3 L2 C3 L2 1
X4
1 1 1
½ I L2 þ ICpi
C L2
k¼3 k
C4 L2
dIL1 R1 r pi dIL2
¼ IR1 þ Ir pi ; ¼ Y1
dt L1 L1 dt
At fixed points (equilibrium points):
dIR1 dIr pi dY1 dIL dIL

¼ 0; ¼ 0; ¼ 0; 1 ¼ 0; 2 ¼ 0
dt dt dt dt dt
dIL2 dIL R1 r pi
¼ 0 ) Y1 ¼ 0; 1 ¼ 0 ) IR1 þ Ir pi ¼0
dt dt L1 L1
dY1 X4
1 1 1 1 1 1
¼ 0 ) ½ IR1 þ ½ gm r pi Ir I
dt k¼3
Ck L2 C 4 C3 L 2
pi
C3 L2 L1
X4
1 1 1
½ I L2 þ IC pi ¼ 0
C
k¼3 k
L2 C 4 L2
dIr pi
¼0
dt
X4
1 1 X4
1 1 X4
1 1
)½ IR1 ½ IC pi ½
C r pi
k¼1 k
C r pi
k¼1 k
C r pi
k¼1 k
k6¼2 k6¼2 k6¼2
k6¼3 k6¼3 k6¼3
1
Ir þ I
pi
C4 r pi L2
¼0
dIR1 X4
1 1 X4
1 X 3
1 1
¼ 0 ) ½ IR1 þ ½ ð Þ gm r pi Ir pi
dt k¼1
Ck R 1 k¼1
Ck k¼2
C k R 1
k6¼2
k6¼3
X4
1 1 X3
1 1 X 4
1 1
þð Þ IC pi ½ IL1 ½ IL2 ¼ 0
C
k¼1 k
R 1 C
k¼2 k
R1 C
k¼3 k
R1
k6¼2
k6¼3
We can summery our system fixed points equations (version 1):
r pi
Y1 ¼ 0; IR1 R1 þ Ir pi r pi ¼ 0 ) IR1 ¼ Ir pi
R1
X 4
1 1 1 1 1 1 X 4
1 1
½ IR 1 þ ½ gm r pi Ir pi IL1 ½ IL2
k¼3
Ck L 2 C 4 C3 L2 C3 L2 C
k¼3 k
L2
1
þ I ¼ 0
C4 L2 Cpi
X4
1 1 X4
1 1 X4
1 1 1
½ IR1 ½ IC pi ½ Ir þ I ¼ 0
k¼1
C k r pi k¼1
Ck r pi k¼1
C k r pi pi
C4 r pi L2
k6¼2 k6¼2 k6¼2
k6¼3 k6¼3 k6¼3
X4
1 1 X4
1 X3
1 1 X4
1 1
½ IR1 þ ½ ð Þ gm r pi Ir pi þð Þ IC pi
k¼1
C k R 1 k¼1
C k k¼2
C k R 1 k¼1
Ck R 1
k6¼2 k6¼2
k6¼3 k6¼3
X3
1 1 X4
1 1
½ IL1 ½ IL2 ¼ 0
C
k¼2 k
R 1 C
k¼3 k
R 1
X4
1 r pi 1 1 1 1
ðÞf½ þ½ gm r pi g Ir I
C R1 L2
k¼3 k
C4 C3 L2 pi
C3 L2 L1
X4
1 1 1
½ I L2 þ IC pi ¼ 0
k¼3
Ck L2 C4 L2
X4
1 1 1 1 X4
1 1
ðÞ ½ ð þ Þ Ir þ IL2 ½ IC pi ¼ 0
C
k¼1 k
R1 r pi pi
C4 r pi C
k¼1 k
r pi
k6¼2 k6¼2
k6¼3 k6¼3
X4
1 r pi X4
1 X 3
1 1 X4
1 1
ð Þf½ þ½ ð Þ gm r pig Ir pi þð Þ IC pi
k¼1
C k R1 k¼1
C k k¼2
C k R 1 k¼1
C k R 1
k6¼2 k6¼2
k6¼3 k6¼3
X3
1 1 X4
1 1
½ IL1 ½ IL2 ¼ 0
k¼2
C k R 1 k¼3
C k R 1
X4
1 X4
1 C4 r pi
ðÞIL2 ¼ ½ C4 IC pi þ ½ ð þ C4 Þ Ir pi
C
k¼1 k
C
k¼1 k
R1
k6¼2 k6¼2
k6¼3 k6¼3
ðÞ ! ðÞ
X 4
1 r pi 1 1 1 1
f½ þ½ gm r pi g Ir I
k¼3
C k R 1 L2 C4 C 3 L2
pi
C3 L2 L1
X4
1 1 X4
1 X4
1 C4 r pi
½ f½ C4 IC pi þ ½ ð þ C4 Þ Ir pi g
C
k¼3 k
L2 C
k¼1 k
C
k¼1 k
R1
k6¼2 k6¼2
k6¼3 k6¼3
1
þ I ¼ 0
C4 L2 Cpi
X4
1 r pi 1 1 1 X 4
1 1 X 4
1 C4 r pi
f½ þ½ gm r pi ½ ½ ð þ C4 Þ Ir pi g
k¼3
C k R 1 L 2 C 4 C3 L 2 k¼3
Ck L 2 k¼1
C k R1
k6¼2
k6¼3
1 1 X4
1 X4
1 1
I þ f ½ ½ C4 g IC pi ¼ 0
C3 L2 L1 C4 C
k¼3 k
C
k¼1 k
L2
k6¼2
k6¼3
ðÞ ! ð Þ
X4
1 r pi X4
1 X3
1 1 X4
1 1
f½ þ½ ð Þ gm r pig Ir pi þð Þ I
C
k¼1 k
R1 C
k¼1 k
C
k¼2 k
R1 C R1 Cpi
k¼1 k
k6¼2 k6¼2
k6¼3 k6¼3
X 3
1 1 X4
1 1 X4
1 X4
1 C4 r pi
½ IL1 ½ f½ C4 IC pi þ ½ ð þ C4 Þ Ir pi g ¼0
k¼2
C k R1 k¼3
Ck R1 k¼1
Ck k¼1
C k R 1
k6¼2 k6¼2
k6¼3 k6¼3
X4
1 r pi X4
1 X3
1 1
ff½ þ½ ð Þ gm r pig
C
k¼1 k
R1 C
k¼1 k
C
k¼2 k
R 1
k6¼2
k6¼3
X4
1 1 X 4
1 C4 r pi
½ ½ ð þ C4 Þg Ir pi
k¼3
Ck R1 k¼1
C k R 1
k6¼2
k6¼3
X4
1 1 X4
1 1 X 4
1 X 3
1 1
þ fð Þ ½ ½ C4 g IC pi ½ IL1 ¼ 0
C R1
k¼1 k
C R1 k¼1 Ck
k¼3 k
C R1
k¼2 k
k6¼2 k6¼2
k6¼3 k6¼3
Remark: it is reader exercise to get the exact fixed points options in our case.
equilibrium points of the Colpitt’s oscillator circuit (BJT transistor small signal
model) consists in adding to coordinate ½Y1 ; Ir pi ; IR1 ; IL1 ; IL2 arbitrarily small
increments of exponentially form ½y1 ; ir pi ; iR1 ; iL1 ; iL2 ekt and retaining the first
order terms in Y1 ; Ir pi ; IR1 ; IL1 ; IL2 . The system of homogenous equations leads to a
polynomial characteristic equation in the eigenvalues [4]. The polynomial charac-
teristic equations accept by set of the below circuit variables, circuit variables
Colpitt’s oscillator circuit [2, 3, 4]. Our Colpitt’s oscillator fixed values with
arbitrarily small increments of exponential form ½y1 ; ir pi ; iR1 ; iL1 ; iL2 ekt are: j = 0
(first fixed point), j = 1(second fixed point), j = 2(third fixed point), etc.,
ðjÞ ðjÞ ðjÞ

Ir pi ðtÞ ¼ Ir pi þ ir pi ekt ; IR1 ðtÞ ¼ IR1 þ iR1 ekt ; IL1 ðtÞ ¼ IL1 þ iL1 ekt ; IL2 ðtÞ
ðjÞ kt
¼ I L2 þ i L2 e
ðjÞ dY1 ðtÞ dIr pi ðtÞ

Y1 ðtÞ ¼ Y1 þ y1 ekt ; ¼ y1 k ekt ; ¼ ir pi k ekt ; ICpi ðtÞ
dt dt
ðjÞ
¼ ICpi þ iCpi ekt
dIR1 ðtÞ dIL ðtÞ dIL ðtÞ dIC ðtÞ

¼ iR1 k ekt ; 1 ¼ iL1 k ekt ; 2 ¼ iL2 k ekt ; pi
dt dt dt dt
¼ iCpi k ekt
&&&
dIR1 X4
1 1 X4
1 X3
1 1 X4
1 1
¼½ IR 1 þ ½ ð Þ gm r pi Ir pi þð Þ ICpi
dt k¼1
C k R 1 k¼1
C k k¼2
C k R 1 k¼1
Ck R 1
k6¼2 k6¼2
k6¼3 k6¼3
X3
1 1 X4
1 1
½ IL 1 ½ IL2
k¼2
C k R 1 k¼3
C k R 1
X4
1 1 ðjÞ
iR1 k ekt ¼ ½ ½IR1 þ iR1 ekt
C R1
k¼1 k
X4
1 X3
1 1 ðjÞ
þ½ ð Þ gm r pi ½Ir pi þ ir pi ekt
k¼1
Ck k¼2
Ck R 1
k6¼2
k6¼3
X4
1 1 ðjÞ
X3
1 1 ðjÞ
þð Þ ½ICpi þ iCpi ekt ½ ½IL1 þ iL1 ekt
k¼1
Ck R 1 k¼2
Ck R 1
k6¼2
k6¼3
X4
1 1 ðjÞ
½ ½IL2 þ iL2 ekt
k¼3
Ck R1
X4
1 1 ðjÞ X 4
1 X3
1 1 ðjÞ
iR1 k ekt ¼ f½ IR 1 þ ½ ð Þ gm r pi Ir pi
k¼1
C k R 1 k¼1
Ck k¼2
Ck R 1
k6¼2
k6¼3
X4
1 1 ðjÞ X 3
1 1 ðjÞ X 4
1 1 ðjÞ
þð Þ ICpi ½ IL 1 ½ IL 2 g
C
k¼1 k
R 1 C
k¼2 k
R 1 C
k¼3 k
R 1
k6¼2
k6¼3
X4
1 1 X4
1 X 3
1 1
½ iR1 ekt þ ½ ð Þ gm r pi ir pi ekt
k¼1
C k R 1 k¼1
Ck k¼2
Ck R 1
k6¼2
k6¼3
X4
1 1 X 3
1 1 X 4
1 1
þð Þ iCpi ekt ½ iL1 ekt ½ iL2 ekt
C R1
k¼1 k
C R1
k¼2 k
C R1
k¼3 k
k6¼2
k6¼3
At fixed points:
X 4
1 1 ðjÞ X 4
1 X 3
1 1 ðjÞ X4
1 1 ðjÞ
½ IR1 þ ½ ð Þ gm r pi Ir pi þ ð Þ ICpi
C
k¼1 k
R 1 C
k¼1 k
C
k¼2 k
R 1 C
k¼1 k
R 1
k6¼2 k6¼2
k6¼3 k6¼3
X 3
1 1 ðjÞ X 4
1 1 ðjÞ
½ IL 1 ½ IL 2 ¼ 0
k¼2
Ck R 1 k¼3
C k R 1
X4
1 1 X4
1 X3
1 1 X4
1 1
i R1 k ½ i R1 þ ½ ð Þ gm r pi ir pi þð Þ iCpi
k¼1
Ck R 1 k¼1
C k k¼2
C k R 1 k¼1
C k R 1
k6¼2 k6¼2
k6¼3 k6¼3
X3
1 1 X4
1 1
½ iL 1 ½ iL2 ¼ 0
k¼2
Ck R 1 k¼3
C k R 1
dIr pi X4
1 1 X4
1 1 X4
1 1
¼½ IR1 ½ ICpi ½
dt k¼1
C k r pi k¼1
Ck r pi k¼1
C k r pi
k6¼2 k6¼2 k6¼2
k6¼3 k6¼3 k6¼3
1
Ir pi þ IL
C4 r pi 2
X4
1 1 ðjÞ
X4
1 1 ðjÞ
ir pi k ekt ¼ ½ ½IR1 þ iR1 ekt ½ ½ICpi þ iCpi ekt
k¼1
Ck r pi k¼1
C k r pi
k6¼2 k6¼2
k6¼3 k6¼3
X4
1 1 ðjÞ 1 ðjÞ
½ ½Ir pi þ ir pi ekt þ ½I þ iL2 ekt
C
k¼1 k
r pi C4 r pi L2
k6¼2
k6¼3
X4
1 1 ðjÞ
X4
1 1 ðjÞ
X4
1 1 ðjÞ
ir pi k ekt ¼ ½ IR 1 ½ ICpi ½ Ir pi
k¼1
C k r pi k¼1
C k r pi k¼1
C k r pi
k6¼2 k6¼2 k6¼2
k6¼3 k6¼3 k6¼3
1 ðjÞ
X4
1 1
þ IL2 þ ½ iR ekt
C4 r pi k¼1
C k r pi 1
k6¼2
k6¼3
X4
1 1 X4
1 1 1
½ iCpi ekt ½ ir pi ekt þ iL ekt
k¼1
C k r pi k¼1
Ck r pi C4 r pi 2
k6¼2 k6¼2
k6¼3 k6¼3
At fixed points:
X4
1 1 ðjÞ
X4
1 1 ðjÞ
X4
1 1 ðjÞ 1 ðjÞ
½ IR1 ½ ICpi ½ Ir pi þ I ¼0
C r pi
k¼1 k
C r pi
k¼1 k
C r pi
k¼1 k
C4 r pi L2
k6¼2 k6¼2 k6¼2
k6¼3 k6¼3 k6¼3
X4
1 1 X4
1 1 X4
1 1
ir pi kþ½ iR1 ½ iCpi ½ ir pi
C r pi
k¼1 k
C r pi
k¼1 k
C r pi
k¼1 k
k6¼2 k6¼2 k6¼2
k6¼3 k6¼3 k6¼3
1
þ iL ¼ 0
C4 r pi 2
dY1 X4
1 1 1 1 1 1
dt C
k¼3 k
L 2 C4 C 3 L2 C3 L2 1
X4
1 1 1
½ I L2 þ ICpi
C L2
k¼3 k
C4 L2
X4
1 1 ðjÞ 1 1 1 ðjÞ
y1 k ekt ¼ ½ ½IR1 þ iR1 ekt þ ½ gm r pi ½Ir pi þ ir pi ekt
C L2
k¼3 k
C4 C3 L2
1 ðjÞ
X4
1 1 ðjÞ
½IL1 þ iL1 ekt ½ ½IL2 þ iL2 ekt
C3 L2 k¼3
C k L 2
1 ðjÞ
þ ½I þ iCpi ekt
C4 L2 Cpi
X4
1 1 ðjÞ X4
1 1 1 1 1 ðjÞ
y1 k ekt ¼ ½ IR1 ½ iR1 ekt þ ½ gm r pi Ir pi
k¼3
C k L 2 k¼3
C k L 2 C 4 C 3 L 2
1 1 1 1 ðjÞ 1 X4
1 1 ðjÞ
þ½ gm r pi ir pi ekt I iL ekt ½ I
C4 C3 L2 C3 L2 L1 C3 L2 1 C L2 L2
k¼3 k
X4
1 1 1 ðjÞ 1
½ iL2 ekt þ ICpi þ iCpi ekt
k¼3
C k L 2 C 4 L2 C 4 L2
X4
1 1 ðjÞ 1 1 1 ðjÞ 1 ðjÞ
y1 k ekt ¼ f½ IR 1 þ ½ gm r pi Ir pi IL 1
k¼3
Ck L2 C 4 C3 L2 C3 L2
X4
1 1 ðjÞ 1 ðjÞ
X4
1 1
½ IL2 þ ICpi g ½ iR1 ekt
k¼3
Ck L2 C4 L2 k¼3
C k L2
1 1 1
þ½ gm r pi ir pi ekt
C4 C3 L2
1 X 4
1 1 1
iL1 ekt ½ iL2 ekt þ iCpi ekt
C3 L2 k¼3
C k L 2 C 4 L 2
At fixed points:
X4
1 1 ðjÞ 1 1 1 ðjÞ 1 ðjÞ
½ IR1 þ ½ gm r pi Ir pi I L1
k¼3
C k L2 C4 C3 L2 C3 L2
X 4
1 1 ðjÞ 1 ðjÞ
½ I L2 þ ICpi ¼ 0
k¼3
C k L 2 C 4 L2
X4
1 1 1 1 1 1
y1 k ½ iR1 þ ½ gm r pi ir pi iL
C
k¼3 k
L2 C 4 C3 L2 C3 L2 1
X4
1 1 1
½ i L2 þ iCpi ¼ 0
C L2
k¼3 k
C4 L2
dIL1 R1 r pi ðjÞ R1
¼ IR1 þ Ir pi ; iL1 k ekt ¼ ½IR1 þ iR1 ekt

dt L1 L1 L1
ðjÞ r pi ðjÞ R ðjÞ r pi R1
þ ½Ir pi þ ir pi ekt iL1 k ekt ¼ IR1 þ iR1 ekt
1
þ Ir pi
L1 L1 L1 L1
r pi kt
þ ir pi e
L1
At fixed points:
ðjÞ R1 ðjÞ r pi R1 r pi
IR1 þ Ir pi ¼ 0; iL1 k þ iR1 þ ir pi ¼0
L1 L1 L1 L1
dIL2 ðjÞ ðjÞ
dt ¼ Y1 ; iL2 k ekt ¼ Y1 þ y1 ekt . At fixed points: Y1 ¼ 0; iL2 k þ y1 ¼ 0
We can summery our Colpitt’s oscillator circuit arbitrarily small increments
equations:
X4
1 1 X4
1 X3
1 1 X4
1 1
i R1 k ½ i R1 þ ½ ð Þ gm r pi ir pi þð Þ iCpi
k¼1
Ck R 1 k¼1
C k k¼2
C k R 1 k¼1
C k R 1
k6¼2 k6¼2
k6¼3 k6¼3
X3
1 1 X4
1 1
½ iL 1 ½ iL2 ¼ 0
k¼2
Ck R 1 k¼3
C k R 1
X4
1 1 X4
1 1 X4
1 1
ir pi kþ½ iR1 ½ iCpi ½ ir pi
C
k¼1 k
r pi C
k¼1 k
r pi C
k¼1 k
r pi
k6¼2 k6¼2 k6¼2
k6¼3 k6¼3 k6¼3
1
þ iL ¼ 0
C4 r pi 2
X4
1 1 1 1 1 1
y1 k ½ iR1 þ ½ gm r pi ir pi iL
k¼3
Ck L2 C 4 C3 L2 C3 L2 1
X4
1 1 1
½ iL2 þ iCpi ¼ 0
C
k¼3 k
L2 C 4 L2
R1 r pi
iL1 k þ iR1 þ ir pi ¼ 0; iL2 k þ y1 ¼ 0
L1 L1
0 1
P
4
½ð 1
B k¼1 Ck R1 C Þ 1

B k6¼2 C
0 1 B
B k6¼3
C
C
0 1 iR B C
B
1
B
C B f½ P 4 C
N11 ... N15 B r pi C B
i 1

Ck C
C
B . .. .. C B C B k¼1 C
B . C B C
. A B y1 C þ B C iCpi ¼ 0;
¼
@ . .
k6 2
C B C
k6¼3
B B C
N51 N55 @ i L1 A B r 1pig C
B C
i L2 B 1 C
B C4 L2 C
B C
@ 0 A
0
0 1
0 1 iR1
N11 ... N15 B C
B ir pi C
B . .. C
.. C B B C
B . C
@ . . . A B y1 C 0
B C
N51 N55 @ i L1 A
i L2
0 1
P
4
B ½ð Ck Þ
R1 C
1 1
B k¼1 C
B k6k6¼¼32 C
B C
B P C
B 4
C
B f½ Ck
1
C P
B C 4
Assumption: B k¼1
C iCpi ! e; N11 ¼ k ½ C1k R11
B k6¼2
k6¼3 C
B C k¼1
B r 1pig C
B C
B 1 C
B C4 L2 C
@ 0 A
0
X4
1 X3
1 1
N12 ¼ ½ ð Þ gm r pi ; N13 ¼ 0;
C
k¼1 k
C
k¼2 k
R1
k6¼2
k6¼3
X3
1 1 X4
1 1
N14 ¼ ½ ; N15 ¼ ½
k¼2
Ck R1 k¼3
Ck R 1
X4
1 1 X4
1 1
N21 ¼ ½ ; N22 ¼ k ½ ;
C
k¼1 k
r pi C
k¼1 k
r pi
k6¼2 k6¼2
k6¼3 k6¼3
1
N23 ¼ 0; N24 ¼ 0; N25 ¼
C4 r pi
X4
1 1 1 1 1 1
N31 ¼ ½ ; N32 ¼ ½ gm r pi ; N33 ¼ k; N34 ¼
k¼3
Ck L2 C 4 C3 L2 C3 L2
X4
1 1 R1 r pi
N35 ¼ ½ ; N41 ¼ ; N42 ¼ ; N43 ¼ 0; N44 ¼ k; N45 ¼ 0
k¼3
Ck L2 L 1 L1
N51 ¼ N52 ¼ 0; N53 ¼ 1; N54 ¼ 0; N55 ¼ k

0 1 0 1
N11 . . . N15 N11 ... N15
B . .. .. C B .. .. .. C
ðA k IÞ ¼ @ .. . . A ; detðA k IÞ ¼ det @ . . . A
N51 N55 N51 N55
If an eigenvalue of A is on the imaginary axis, then the method of the center

manifold must be applied. To apply the stability criterion of Liapunov to our
system, the equation detðA k IÞ ¼ 0 has a zero in the left half plane, without
calculating the eigenvalues explicitly. We define new system parameters functions:
X4
1 1
n1 ¼ n1 ðC k
; R1 Þ ¼ ½ ; n1 ¼ n2 ðC k
;
C
k¼1 k
R 1
½k ¼ 1; 2; 3; 4 ½k ¼ 1; 2; 3; 4
X4
1 X 3
1 1
R1 ; r pi; gm Þ ¼ ½ ð Þ gm r pi
C
k¼1 k
C
k¼2 k
R1
k6¼2
k6¼3
X3
1 1 X4
1 1
n3 ¼ n3 ðC2 ; C3 ; R1 Þ ¼ ½ ; n4 ¼ n4 ðC3 ; C4 ; R1 Þ ¼ ½
k¼2
C k R 1 k¼3
Ck R 1
X4
1 1 X4
1 1
n5 ¼ n5 ðC1 ; C4 ; r piÞ ¼ ½ ; n6 ¼ n6 ðC1 ; C4 ; r piÞ ¼ ½
C
k¼1 k
r pi C
k¼1 k
r pi
k6¼2 k6¼2
k6¼3 k6¼3
1 X4
1 1
n7 ¼ n7 ðC4 ; r piÞ ¼ ; n8 ¼ n8 ðC3 ; C4 ; L2 Þ ¼ ½
C4 r pi C L2
k¼3 k
1 1 1 1
n9 ¼ n9 ðC3 ; C4 ; r pi; L2 Þ ¼ ½ gm r pi ; n10 ¼ n10 ðC3 ; L2 Þ ¼
C4 C3 L2 C3 L2
X4
1 1 R1
n11 ¼ n11 ðC3 ; C4 ; L2 Þ ¼ ½ ; n12 ¼ n12 ðR1 ; L1 Þ ¼ ;
k¼3
Ck L 2 L1
r pi
n13 ¼ n13 ðr pi; L1 Þ ¼
L1
detðA k IÞ ¼ k n1 fk4 k3 n6 k2 n11 þ k ½n11 n6 n7 n9 n7 n10 n13 g

n2 fk3 n5 k ½n11 n5 þ n7 n8 n7 n10 n12 g n3
fk3 n12 þ k2 ½n6 n12 n5 n13 þ k n11 n12 þ ½n11 n13 n5 n11
n6 n12 n7 n8 n13 þ n7 n9 n12 g þ n4 fk2 n8 þ k ½n5 n9
þ n10 n12 n6 n8 þ ½n10 n13 n5 n6 n10 n12 g
detðA k IÞ ¼ n1 k5 þ k4 n6 n1 þ k3 n1 n11 k2 n1 ½n11 n6 n7 n9

þ k n1 n7 n10 n13 þ k3 n2 n5 þ k n2 ½n11 n5 þ n7 n8
þ n2 n7 n10 n12 þ k3 n3 n12 k2 n3 ½n6 n12 n5 n13
k n11 n3 n12 n3 ½n11 n13 n5 n11 n6 n12 n7 n8 n13 þ n7 n9 n12
þ k2 n4 n8 þ k n4 ½n5 n9 þ n10 n12 n6 n8
þ n4 ½n10 n13 n5 n6 n10 n12
detðA k IÞ ¼ n1 k5 þ k4 n6 n1 þ k3 ½n1 n11 þ n2 n5 þ n3 n12

þ k2 fn4 n8 n1 ½n11 n6 n7 n9 n3 ½n6 n12 n5 n13 g
þ k fn1 n7 n10 n13 þ k n2 ½n11 n5 þ n7 n8
n11 n3 n12 þ n4 ½n5 n9 þ n10 n12 n6 n8 g
þ n2 n7 n10 n12 n3 ½n11 n13 n5 n11 n6 n12
n7 n8 n13 þ n7 n9 n12
þ n4 ½n10 n13 n5 n6 n10 n12
We define new global parameters: !5 ¼ n1 ; !4 ¼ n6 n1 ; !3 ¼ n1 n11 þ n2

n5 þ n3 n12
!2 ¼ n4 n8 n1 ½n11 n6 n7 n9 n3 ½n6 n12 n5 n13
!1 ¼ n1 n7 n10 n13 þ k n2 ½n11 n5 þ n7 n8 n11 n3 n12 þ n4

½n5 n9 þ n10 n12 n6 n8
!0 ¼ n2 n7 n10 n12 n3 ½n11 n13 n5 n11 n6 n12 n7 n8 n13 þ n7 n9 n12

þ n4 ½n10 n13 n5 n6 n10 n12
P
5 P
5
detðA k IÞ ¼ !k kk ¼ 0. Next is to find zeros of !k kk ¼ 0 and
k¼0 k¼0
establish stability according to criterion of Liapunov [2, 3, 4].
Exercises 661
Exercises
1. We have resonator circuit hð0\h\1Þ degree phase shift. The output of the
amplifier is taken through a phase shift network to provide it an additional
phase shift of hð0\h\1Þ degree. Amplifier provides a phase shift of 180°
and the phase shift network also gives a θ°. Therefore total phases shift of
(180 + θ) degree (it is a 60° phase shift per RC section). The phase shift
network is constructed from n RC sections (n [ 3; n 2 N).
Phase shift resonant circuit stability analysis is done by considering BJT Small
Signal (SS) equivalent circuit model. Consider in your analysis “AC ground” in
the Small Signal (SS) circuit.
1:1 Find resonator circuit differential equations.
1:2 Find resonator circuit fixed points.
1:3 Discuss stability of resonator circuit hð0\h\1Þ degree phase shift by
using Liapunov theory. How the stability is dependent on the number
(n) of RC sections?
1:4 Discuss the circuit behavior which characterized as a function of the
position of eigenvalues in the Re/Im plane. How the position of eigen-
values in the Re/Im plane is changed for different numbers (n) of RC
sections?
1:5 Discuss center manifold theorem in our circuit for different number (n) of
RC sections.
2. We have resonator circuit uð0\u\1Þ degree phase shift. The output of the
amplifier is taken through two phase shift networks in cascade (series). There is
intermediate RF choke (L) between the first and the second RC networks. The
phase shift of the first network is u1 ð0\u1 \1Þ degree and the phase shift of
P2
the second network is u2 ð0\u2 \1Þ degree u¼ uk ; 0\u\1. In the first
k¼1
RC network the number of RC sections is n. In the second RC network the
number of RC sections is m (n 6¼ m; n; m [ 3; n; m 2 N).
Phase shift resonator circuit stability analysis is done by considering BJT Small
Signal (SS) equivalent circuit model. Consider in your analysis “AC signal” in
the small signal circuit model.
2:1 Find resonator circuit differential equations.
2:2 Find resonator circuit fixed points.
2:3 Discuss stability of resonator circuit uð0\u\1Þ phase shift by using
Liapunov theory. How the stability is dependent on the number (n) of RC
Exercises 663
sections in the first network? and the number (m) of RC sections in the
second network? How the stability of resonator circuit is dependent on the
value of intermediate RF choke (L)?
position of eigenvalues in the Re/Im plane. How the position of eigen-
values in the Re/Im plane is changed for different number of RC sections
in the first network (n) and second network (m)?
2:5 Discuss center manifold theorem in our circuit for different number of RC
sections in the first network (n) and second network (m).
3. We have parallel resonator crystal oscillator circuit (BJT crystal oscillator). The
BJT’s collector is connected to RFC (Radio Frequency Coil) L1. Additionally
we have XTAL which is connected to circuit’s output. The circuit contains
biasing resistors R1 and R2 and additional elements (CB, RE, C1, C2). Our
circuit stability analysis is done by considering BJT Small Signal
(SS) equivalent model. Consider in your analysis “AC ground” in the Small
Signal (SS) circuit.
Hint: Replace the circuit’s crystal by equivalent circuit (resonator circuit).

3:1 Find parallel resonator crystal oscillator circuit differential equations.
3:2 Find parallel resonator crystal oscillator circuit fixed points.
3:3 Discuss stability of parallel resonator crystal oscillator circuit by using
Liapunov theorem. How the stability is changed if capacitor Cb is dis-
connected? How the stability is changed if we disconnect resistor Re?
3:4 Discuss circuit behavior which characterized as a function of the position

of eigenvalues in the Re/Im plane. How the position of eigenvalues in the
Re/Im plane changed if we connect two parallel crystals (XTAL1 and
XTAL2) to our circuit’s output? Consider that two parallel crystals
(XTAL1 and XTAL2) are not same.
3:5 Discuss center manifold theorem in our circuit for two cases. First case:
resistor R2 is disconnected and Second case: capacitor C2 is disconnected.
4. We have BJT based Hartley oscillator circuit. The circuit is constructed from
sub Tank circuit and peripheral components. The sub tank circuit is constructed
from capacitor C and primary transformer (T1) inductances (L1 and L2), total
P
Lp ¼ 2i¼1 Li . The secondary transformer (T1) inductance is Ls (Lp ≠ Ls).
Additionally there are L3 (RFC = Radio Frequency Coil), BJT transistor bias
resistors R1 and R2, Q1 transistor emitter resistor (Re) and capacitor (Ce),
feedback capacitor (Cc).
Transformer T1: vp—transformer primary voltage, vs—transformer secondary

voltage. ϕ is the flux through a one turn coil located anywhere on the transformer
core. np and ns are the number of turns of the coil 1 (primary) and 2 (secondary),
respectively, then the total flux ϕp and ϕs through coils 1 and 2 respectively are
d/ P2
/p ¼ np /; /s ¼ ns /; vp ¼ dtp ; vs ¼ d/dt ; vp ¼ np dt ; Lp ¼
s d/
i¼1 Li
Exercises 665
v ðtÞ np np v ðtÞ np
vs ¼ ns d/
dt ; vp ðtÞ ¼ vp ; vs ðtÞ ¼ vs ; vs ðtÞ ¼
p
ns ;a ¼ ns . vps ðtÞ ¼ ns for all times t and
n
for all voltages vp and vs . We define the ratio between np and ns as “a” ða ¼ nps Þ.
Step down transformer a > 1 and step up transformer a < 1. By law of con-
servation of energy, apparent real and reactive powers are each conserved in the
input and output ðS ¼ Ip vp ¼ Is vs ; vp ¼ VL1 þ VL2 Þ.
Remark: Coil 1 (primary) is constructed from two coils (L1 and L2 in series).
Hartley oscillator circuit stability analysis is done by considering BJT Small
Signal (SS) equivalent circuit model. Consider in your analysis “AC signal” in
the small signal circuit model.
4:1 Find Hartley oscillator circuit differential equations.
4:2 Find Hartley oscillator circuit fixed points.
4:3 Discuss stability of Hartley oscillator circuit by using Liapunov theorem.
How the stability is dependent on the number of turns of the coil 1
(primary) and 2 (secondary) of T1 transformer?
position of eigenvalues in the Re/Im plane. How the position of the
eigenvalues in the Re/Im plane is changed if we short resistor Re? And if
we disconnect capacitor Ce?
4:5 Discuss center manifold theorem in our circuit if we multiple the capac-
itance C by factor Γ (C ! C C; C 2 R þ ). How the circuit stability is
dependent on Γ parameter?
5. We have transistor Colpitts oscillator circuit. The circuit is constructed from
sub tank circuit and peripheral components. The sub tank circuit is constructed
from two capacitors C1 and C2, primary transformer (T1)—inductance L
(Lp = L). The secondary transformer (T1) inductance is Ls (Lp ≠ Ls).
Additionally there are L1 (RFC = Radio Frequency Coil), BJT transistor bias
resistors R1 and R2, Q1 transistor emitter resistor (Re) and capacitor (Ce),
feedback capacitor (Cc). Colpitts oscillator circuit stability analysis is done by
considering BJT Small Signal (SS) equivalent circuit model. Consider in your
analysis “AC signal” in the small signal circuit model. Transformer T1: vp—
transformer primary voltage, vs—transformer secondary voltage. ϕ is the flux
through a one turn coil located anywhere on the transformer core. np and ns are
the number of turns of the coil 1 (primary) and 2 (secondary), respectively, then
the total flux ϕp and ϕs through coils 1 and 2 respectively are /p ¼ np /; /s ¼
d/p d/s vp ðtÞ
ns /; vp ¼ dt ; vs ¼ dt ; vp ¼ np dt vs ¼ ns dt ; vp ðtÞ ¼ vp ; vs ðtÞ ¼ vs ; vs ðtÞ ¼
d/ d/
np n v ðtÞ n
ns ;a ¼ nps . vps ðtÞ ¼ nps for all times t and for all voltages vp and vs . We define the
n
ratio between np and ns as “a” (a ¼ nps ). Step down transformer a > 1 and step
up transformer a < 1. By law of conservation of energy, apparent real and
reactive powers are each conserved in the input and output
(S ¼ Ip vp ¼ Is vs ).
5:1 Find Colpitts oscillator circuit differential equations.

5:2 Find Colpitts oscillator circuit fixed points.
5:3 Discuss stability of Colpitts oscillator circuit by using Liapunov theorem.
How the stability is dependent on the number of turns of the coil 1
(primary) and 2 (secondary) of T1 transformer?
eigenvalues in the Re/Im plane is changed if we disconnect resistor Re?
And if we disconnect capacitor Ce?
itance Cc by factor Γ (Cc ! C Cc ; C 2 R þ ). How the circuit stability is
6. We have Colpitts crystal oscillator circuit. The crystal oscillator (XTAL) is
designed around a CC (Common Collector), emitter-follower amplifier. The R1
and R2 resistor network sets the DC bias level on the base while emitter resistor
Re sets the output voltage level. Resistor R2 is set as large as possible to prevent
loading to the parallel connected crystal. The type of transistor is NPN con-
nected in a common collector configuration and is capable of operating at high
switching speeds. Capacitors C1 and C2 shunt the output of the transistor which
Exercises 667
reduces the feedback signal. Therefore, the gain of the transistor limits the
maximum value of C1 and C2. The output amplitude should be kept low in
order to avoid excessive power dissipation in the crystal.
Our circuit stability analysis is done by considering BJT Small Signal

Signal (SS) circuit. Hint: Replace the circuit’s crystal by equivalent circuit
(resonator circuit).
6:1 Find Colpitts crystal oscillator circuit differential equations.
6:2 Find Colpitts crystal oscillator circuit fixed points.
6:3 Discuss stability of Colpitts crystal oscillator circuit by using Liapunov
theorem. How the stability is dependent on values of capacitors C1, C2,
and C3?
eigenvalues in the Re/Im plane is changed if we short resistor Re? And if
we disconnect capacitor C3?
itance C1 by factor Γ (C1 ! C C1 ; C 2 R þ ). How the circuit stability is
7. The Colpitts oscillator, like Hartley is capable of giving an excellent sine wave
shape, and also has the advantage of better stability at very high frequencies. It
can be recognized by always having a “tapped capacitor”. The circuit is a
buffered Colpitts oscillator which is a common solution to feed the oscillator
output into an emitter follower buffer amplifier. The oscillator section of this
circuit is a slightly different version of typical Colpitts oscillator. The RF choke
(L1) is the load impedance for Q1 and the tank circuit is isolated from Q1 by two
DC blocking capacitors, C1 and C4. The Colpitts oscillator circuit uses a tuned
feedback path rather than a tuned amplifier. The emitter follower stage (R4, Q2
and R5) has very high input impedance, thus having little loading effect on the
oscillator, and very low output impedance allowing it to drive loads of only a
few tens of ohms impedance. The frequency stability of oscillators can be
affected by variations in supply voltage.
(SS) equivalent model for Q1 and Q2. Consider in your analysis “AC ground”
in the Small Signal (SS) circuit.
7:1 Find buffered Colpitts oscillator circuit differential equations

7:2 Find buffered Colpitts oscillator circuit fixed points.
7:3 Discuss stability of buffered Colpitts oscillator circuit by using Liapunov
theorem. How the stability is dependent on values of capacitors C6, C2,
and C3?
eigenvalues in the Re/Im plane is changed if we short resistor R3? And if
we disconnect capacitor C5?
Exercises 669
itance C2 by factor Γ (C2 ! C C2 ; C 2 R þ ). How the circuit stability is
8. We have Colpitts double crystals oscillator circuit. The crystals oscillator
(XTAL1 and XTAL2) is designed around a CC (Common Collector),
emitter-follower amplifier. The R1 resistor sets the DC bias level on the base
while emitter resistor Re sets the output voltage level. The type of transistor is
NPN connected in a common collector configuration and is capable of oper-
ating at high switching speeds. Capacitor C1 can be tune in his value.

Signal (SS) circuit. Hint: Replace the circuit’s crystal by equivalent circuit
(resonator circuit).
8:1 Find Colpitts double crystals oscillator circuit differential equations.
8:2 Find Colpitts double crystals circuit fixed points.
8:3 Discuss stability of Colpitts double crystals circuit by using Liapunov
theorem. How the stability is dependent on values of capacitors C1, Re?
eigenvalues in the Re/Im plane is changed if we disconnect resistor Re?
And if we disconnect capacitor C3?
8:5 Discuss center manifold theorem in our circuit if capacitance C1 transform
according to the function C1 ! C1 ¼ nðC1 ; CÞ
pffiffiffiffiffiffiffiffiffiffiffiffiffiffi
(nðC1 ; CÞ ¼ C1 C þ C1 C3 ; C 2 R þ ). How the circuit stability is
9. We have voltage controlled RF oscillator (VCO) circuit. The VCO circuit may
be considered as an amplifier and a feedback loop. For the circuit to oscillate
the total phase shift around the loop must be 360° and the gain must be unity.
VCO circuit uses a common emitter circuit. This is itself produces a phase shift
of 180°, leaving the feedback network to provide a further 180°. VCO circuit
which uses a common base circuit where there is no phase shift between the
emitter and collector signals (using bipolar transistor) and the phase shift net-
work must provide either 0° or 360°. The system includes a resonator circuit to
ensure that the oscillation occurs on a given frequency. The resonator circuit
can be an LC resonator circuit in either series or parallel resonance depending
upon the circuit, or a quartz crystal. Resistors Rb1 and Rb2 are Q1’s transistor
biasing elements.
One of the options for VCO active device is bipolar transistor. The bipolar transistor has a
low input impedance and is current driven. To make a VCO, the oscillator needs to be tuned
by a voltage. This is done by variable capacitor from varactor diodes. The tune voltage for
the VCO can then be applied to the varactors. The control line from the phase detector is
isolated from the varactor diodes using a resistor R1. Inductor L1 is the RF choke (RFC).
The series capacitor C3 is used to block the DC from the inductor otherwise it would
provide a direct short to ground and upset the bias arrangements of the circuit. The value of
C3 is normally large in comparison with C1 and C2 and can be ignored from the resonance
perspective. If we use a varactor diodes within a voltage controlled oscillator, care must be
taken in the design of the circuit to ensure that the drive level in the tuned circuit is not too
high. In that case the varactor diodes may be driven into forward conduction, reducing the
Q and increasing the level of spurious signals.
Varactor diodes: In your stability analysis use the simple model of a packaged
varactor diode. A varactor diode is a P-N diode that changes its capacitance and
the series resistance as the bias applied to the diode is varied. The property of
Exercises 671
capacitance change is utilized to achieve a change in the frequency and/or the

phase of an electrical circuit. In the simple model of a packaged varactor diode
Cj(V) is the variable junction capacitance of the diode die, Rs(V) is the variable
series resistance of the diode die, and Cp is the fixed parasitic capacitance
arising from the installation of the die in a package. Package parasitic induc-
tance Lp. The contribution to the series resistance from the packaging is very
small and may be ignored. Similarly, the inductance associated with the die
itself is very small and may be ignored. Variation of the junction capacitance
and the junction series resistance as a function of applied reverse voltage is
reported in the individual varactor data sheets. D1 and D2 are varactor diodes,
consider that they identical for circuit analysis.
9:1 Find voltage controlled RF oscillator (VCO) circuit differential equations.
9:2 Find voltage controlled RF oscillator (VCO) circuit fixed points.
9:3 Discuss stability of voltage controlled RF oscillator (VCO) circuit by
using Liapunov theorem. How the stability is dependent on values of
capacitors Cc, Re?
9:4 We disconnect resistor Rb2, How the stability and dynamics of our circuit
is changed?
9:5 Discuss center manifold theorem in our circuit if capacitance C1 and C2
transform according to the functions C1 ! C1 ¼ n1 ðC1 ; CÞ
pffiffiffiffi qffiffiffiffiffiffiffiffiffiffiffiffiffiffi3
C2 ! C2 ¼ n2 ðC2 ; XÞ; ðn1 ðC1 ; CÞ ¼ C1 C þ C1 C ; C 2 R þ Þ:
pffiffiffiffiffiffi pffiffiffiffiffiffiffiffiffiffiffiffi
ðn2 ðC2 ; XÞ ¼ C2 X3 þ C2 X ; X 2 R þ Þ. How the circuit stability is
dependent on Γ and Ω parameters?
10. We have Common Base (CB) bipolar colpitts oscillator with varactor diode.
Common Base (CB) amplifier, input signal is inserted at emitter (E) and output
signal taken from the collector (C). The CB amplifier can operate as a voltage
amplifier for low input impedance circuits. The transistor is matched to the
load. One capacitor of the divider circuit is the varactor, whose voltage can be
changed by applying a tuning voltage through RF choke (L3) and a bypass
conductor. Note that the capacitance variation with voltage is nonlinear, with
the capacitance changing more at low values of bias voltage than at high values.
The typical capacitance variation of a varactor diode (D1) is as a function of the
reverse bias voltage. This leads to a nonlinear frequency versus voltage tuning
curve for the oscillator. If linearity is required, a compensation network must be
used to modify the tuning voltage before it is applied to varactor. Resistors
R1… R4 are circuit biasing elements. Output matching network can be Pi or T
type. Varactor diodes: In your stability analysis use the simple model of a
packaged varactor diode. A varactor diode is a P-N diode that changes its
capacitance and the series resistance as the bias applied to the diode is varied.

10:1 Find voltage Common Base (CB) bipolar colpitts oscillator circuit dif-
ferential equations for Pi and T matching networks.
10:2 Find Common Base (CB) bipolar colpitts oscillator circuit fixed points
for Pi and T matching networks.
10:3 Discuss stability of Common Base (CB) bipolar colpitts circuit by using
Liapunov theorem. How the stability is dependent on values of capaci-
tors C2, R4?
10:4 We disconnect capacitor C1, How the stability and dynamics of our
circuit is changed?
10:5 Discuss center manifold theorem in our circuit if capacitance C1 and C2
transform according to the functions C1 ! C1 ¼ n1 ðC1 ; CÞ C2 ! C2 ¼
pffiffiffiffi qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
pffiffiffiffiffiffi
n2 ðC2 ; XÞ; (n1 ðC1 ; CÞ ¼ C1 C þ C1 C3 ; C 2 R þ ). (n2 ðC2 ; XÞ ¼
pffiffiffiffiffiffi pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
pffiffiffiffi
C2 X3 þ C2 X; X 2 R þ ). How the circuit stability Is dependent
on Γ and Ω parameters?
Chapter 7
Filters Systems Stability Analysis
The target of analog and RF filtering is to modify the magnitude and phase of signal
frequency components. Many analog or radio frequency (RF) circuits perform fil-
tering on the signals passing through them. The analog and RF filters types are
defined on the criteria how they modify the magnitude and/or phase of sinusoidal
frequency components. The primary issue is magnitude response. In other cases
filters concerned with phase modifications. Filters are typically classified based on
how they modify the frequency spectrum. The four basic types of filters are; the low
pass filter, High pass filter, bandpass filter and band stop filter. Microwave and RF
filters pass a range of frequencies and reject other frequencies. Filters are widely
used at the input of a microwave receiver. The RF and microwave system’s fre-
quencies are picked up by the receiving antenna system and enter the radio receiver.
The first target of the filter is to pass only those frequencies in the assigned oper-
ation range. The second target of the filter is to reject all other frequencies.
Important graphs when analyzing the functionality of filter is the attenuation of a
microwave signal passing through the filter as a function of frequency. Good filters
have a large out of band attenuation and a low in band insertion loss. The selectivity
feature defines the frequency range over which the filter characteristics change from
passing the signal to blocking it. A Diplexer is a three port network that splits the
incoming signal on one end and directs it through two outputs to different lines,
dependent on frequency. A diplexer is the simplest form of a multiplexer, which can
split the signals from one common port into many different paths. Quartz crystals
are widely used In many filters applications. Quartz crystals have piezo electric
properties; they develop an electric potential upon the application of mechanical
stress. One of the most common piezoelectric uses of quartz is as a crystal oscil-
lator. The resonant frequency of a quartz crystal oscillator is changed by
mechanically loading it. A quartz crystal has two resonant frequencies, a series
resonant frequency and a parallel resonant frequency. A quartz crystal operating on
its parallel resonant frequency behaves like a parallel LC circuit. It has high
impedance at one frequency and other frequency only. The practical advantage of
quartz crystal is that it has an extremely high Q and result of an excellent high
DOI 10.1007/978-3-319-45427-6_7
674 7 Filters Systems Stability Analysis
selectivity filter. The crystal lattice filter can be a band pass filter, suitable for use as
a sideband filter in an SSB transmitter receiver or transmitter. The two resonant
frequencies of a quartz crystal are important when designing many RF systems.
A tunable third order bandpass filter using varactors is commonly used in many RF
applications. A tunable filter has greater functionality, better channel selectivity,
reduced size, and lower weight since the same hardware can be employed at
multiple bands. Practically tunable frequency filters are used as tracking filters for
multi band telecommunication systems, wideband radar systems and radiometers.
Tracking filters are mechanically tuned by adjusting the cavity dimensions of the
resonators or magnetically altering the resonant frequency. The tuning element is a
reverse-biased varactor diode [26, 33, 34, 42].
7.1 BPF Diplexer Without a Series Input Stability

Analysis
The terminology duplexers and diplexers are very important for wireless commu-
nication. Duplexer is when two band pass filters are duplexed, meaning one
common input, and two outputs (reverse is also correct, two inputs and one output).
Duplexer will duplex a receiving and transmitting signal using two band pass filter,
one common input (or output), and two outputs (or inputs), and is a three port
device. A diplexer will refer to a duplexed high pass and low pass where broad
bands transmit and receive is necessary, it is also a three port device with a common
input and two outputs. A diplexer is a passive device that implements frequency
domain multiplexing. Two ports (e.g., L and H) are multiplexed onto a third port
(e.g., S). The signals on ports L and H occupy disjoint frequency bands.
Consequently, the signals on L and H can coexist on port S without interfering with
each other. The signal on port L will occupy a single low frequency band and the
signal on port H will occupy a higher frequency band. In that situation, the diplexer
consists of a low pass filter connecting ports L and S and high pass filter connecting
ports H and S. Ideally, all the signal power on port L is transferred to the S port and
vice versa. None of the low band signal is transferred from S port to the H port.
Some power will be lost, and some signal power will leak to the wrong port. The
diplexer, being a passive device, is reciprocal; the device itself doesn’t have a
notion of input or output. The diplexer is a different device than a passive combiner
or splitter. The ports of a diplexer are frequency selective; the ports of a combiner
are not. This is also a power “loss” difference—a combiner takes all the power
delivered to the S port and divides it between the A and B ports. A diplexer
multiplexes two ports onto one port, but more than two ports may be multiplexed: a
three port to one port multiplexer is known as a triplexer. Our BPF diplexer without
a series input circuits splits transmit from receive frequency in an FDD (Frequency
Division Duplex) transceiver. A diplexer can be placed at the output of a frequency
RF source, where it functions as an absorptive filter [91, 92].
7.1 BPF Diplexer Without a Series Input Stability Analysis 675
Fig. 7.1 BPF diplexer without a series input block diagram with RF source, output matching
network and Rload
Filter 1 of the duplexer has a BPF that corresponds with the undesired frequency
band, which can pass right through with little attenuation, and is terminated within
the 50 X load. These undesired frequencies are blocked from entering filter 2 by
that filter’s stop band. Filter 2 BPF passes all the desired signals onto the load
through matching network with little attenuation. The undesired signals through the
diplexer are absorbed instead of being reflected as they would be in a typical filter.
This absorption will prevent any undesired frequency products that were created by
the RF source nonlinearity, from being bounced off of a reflective filters stopbands,
which would return to the RF source and cause increased Inter Modulation
Distortion (IMD) levels. The design of a diplexer is as two different frequency
filters with non-overlapping band pass. The full BPF diplexer without a series input
circuit is presented in the next figure. The matching network between diplexer unit
and load can be L-type, Pi-type, and T-type.
Fig. 7.2 BPF diplexer without a series input circuit

In actual microwave and RF system the diplexer unit is connected through micro
strips lines to RF source, 50 X resistor, and load through matching network. We
represent these micro strips parasitic effects as a delay lines in time. The delays are
related to the current which flows through micro strips and are sin, sout1, and sout2
respectively. We choose T-type matching network. Maximum power transfer is
achieved by using T matching network (passive) connected between diplexer unit
and load Rload. The circuit matching network no only designed to meet the
requirement of minimum power loss but to additional targets. The additional targets
of matching network are minimizing noise influence, maximizing power handling
capabilities, and linearizing the frequency response [107, 108].
Fig. 7.3 BPF diplexer without a series input full schematic with delay line elements
Vsin ¼ VA1 VA2 ; Vsout1 ¼ VA4 VA5 ; Vsout2 ¼ VA6 VA7 ; Vsin ! e; Vsout1 ! e; Vsout2
!e
IR1 ðtÞ ¼ IL1 ðt sout2 Þ þ IC1 ðt sout2 Þ; ILn1 ðtÞ ¼ IL2 ðt sout1 Þ; IRs ðt sin Þ
¼ IC2 ðtÞ þ IL1 ðtÞ þ IC1 ðtÞ
Vs ðtÞ VA1 dðVA2 VA3 Þ dIL

IR s ¼ ; IC2 ¼ C2 ; VL2 ¼ VA3 VA4 ¼ L2 2
Rs dt dt
dIL1 dðVA2 VA6 Þ VA

VL1 ¼ VA2 VA6 ¼ L1 ; IC1 ¼ C1 ; IR1 ¼ 7 ; VLn1
dt dt R1
dILn1
¼ VA5 VA8 ¼ Ln1
dt
dILn2 dVA8 VA9

VLn2 ¼ VA8 VA9 ¼ Ln2 ; ICn ¼ Cn ; IRload ¼ ; ILn2 ¼ IRload ; IC2
dt dt Rload
¼ I L2
Vs ðtÞ VA1
Ln1 ¼ ILn2 þ ICn ; IRs ¼ ) VA1 ¼ Vs ðtÞ IRs Rs ;
Rs
Z
dðVA2 VA3 Þ 1
I C2 ¼ C 2 ) VA2 VA3 ¼ IC2 dt
dt C2
dIL2 dIL1
VA3 VA4 ¼ L2 ; VA2 VA6 ¼ L1 ;
dt dt Z
dðVA2 VA6 Þ 1
I C1 ¼ C 1 ) VA2 VA6 ¼ IC1 dt
dt C1
VA dIL dIL
IR1 ¼ 7 ) VA7 ¼ IR1 R1 ; VA5 VA8 ¼ Ln1 n1 ; VA8 VA9 ¼ Ln2 n2
R1 dt dt
Z
dVA8 1 VA9
ICn ¼ Cn ) VA8 ¼ ICn dt; IRload ¼ ) VA9 ¼ IRload Rload
dt Cn Rload
dIL2
VA1 VA2 ; VA4 VA5 ; VA6 VA7 ; VA2 ¼ Vs ðtÞ IRs Rs ; VA3 VA5 ¼ L2
dt
Z Z
dIL1 1 1
VA2 VA7 ¼ L1 ; VA2 VA7 ¼ IC1 dt; VA3 ¼ VA2 IC2 dt
dt C1 C2
Z
1 dIL
VA3 ¼ Vs ðtÞ IRs Rs IC2 dt; VA3 VA5 ¼ L2 2 ) VA5
C2 dt
dIL2
¼ VA3 L2
dt
Z
1 dIL dIL
VA5 ¼ Vs ðtÞ IRs Rs IC2 dt L2 2 ; VA7 ¼ VA2 L1 1
C2 dt dt
Z
dIL 1
VA7 ¼ Vs ðtÞ IRs Rs L1 1 ; VA7 ¼ VA2 IC1 dt; VA7
Zdt C1
1
¼ Vs ðtÞ IRs Rs IC1 dt
C1
&&&
Z
1
IR1 R1 ¼ Vs ðtÞ IRs Rs IC1 dt; IR1 R1
C1 Z
dIL 1 dIL
¼ Vs ðtÞ IRs Rs L1 1 ; IC1 dt ¼ L1 1
dt C1 dt
dILn1 dIL
VA5 VA8 ¼ Ln1 ) VA8 ¼ VA5 Ln1 n1 ; VA8
dt Z dt
1 dIL dIL
¼ Vs ðtÞ IRs Rs IC2 dt L2 2 Ln1 n1
C2 dt dt
Z Z
1 1 dIL dIL 1
I Cn dt ¼ Vs ðtÞ IRs Rs IC2 dt L2 2 Ln1 n1 ;
Cn Z C2 dt dt Cn
ICn dt IRload Rload

dILn2
¼ Ln2
dt
We can summery our system differential equations (Version 1):

Z
1
IR1 R1 ¼ Vs ðtÞ IRs Rs IC1 dt; IR1 R1
C1 Z
dIL1 1 dIL
¼ Vs ðtÞ IRs Rs L1 ; IC1 dt ¼ L1 1
dt C1 dt
Z Z
1 1 dIL dIL
I Cn dt ¼ Vs ðtÞ IRs Rs IC2 dt L2 2 Ln1 n1
Cn C2 dt dt
Z
1 dILn2
ICn dt IRload Rload ¼ Ln2 ; ILn2 ¼ IRload ; IC2 ¼ IL2
Cn dt
dIR1 dVs ðtÞ dIRs 1 dIR1

R1 ¼ Rs IC1 ; R1
dt dt dt C1 dt
dVs ðtÞ dIRs d 2 IL 1 d 2 IL
¼ Rs L1 2 1 ; I C 1 ¼ L1 2 1
dt dt dt C1 dt
1 dVs ðtÞ dIRs 1 d 2 IL d 2 ILn1 1 dIL

ICn ¼ Rs IL2 L2 2 2 Ln1 2
; ICn n2
Cn dt dt C2 dt dt Cn dt
Rload
d 2 ILn2
¼ Ln2
dt2
IR1 ðtÞ ¼ IL1 ðt sout2 Þ þ IC1 ðt sout2 Þ; ILn1 ðtÞ ¼ IL2 ðt sout1 Þ; IRs ðt sin Þ
¼ IC2 ðtÞ þ IL1 ðtÞ þ IC1 ðtÞ
dIL1
IR1 R1 ¼ Vs ðtÞ IRs Rs L1 ; ½IL1 ðt sout2 Þ þ IC1 ðt sout2 Þ R1
dt
dIL1
¼ Vs ðtÞ IRs Rs L1
dt
d dVs ðtÞ dIRs 1 1
½IL1 ðt sout2 Þ þ IC1 ðt sout2 Þ R1 ¼ Rs I C1 ; I C1
dt dt dt C1 C1
d 2 IL
¼ L1 2 1
dt
1 dVs ðtÞ dIRs 1 d 2 IL d 2 IL2 ðt sout1 Þ

I Cn ¼ Rs IL2 L2 2 2 Ln1
Cn dt dt C2 dt dt2
1 dIL d 2 ILn2
ICn n2 Rload ¼ Ln2 ; IRs ðt sin Þ ¼ IL2 ðtÞ þ IL1 ðtÞ þ IC1 ðtÞ
Cn dt dt2
dI dI dI
We define new system variables: dtL1 ¼ X1 ; dtL2 ¼ X2 ; dtLn2 ¼ X3
½IL1 ðt sout2 Þ þ IC1 ðt sout2 Þ R1 ¼ Vs ðtÞ IRs Rs L1 X1
dIL1 ðt sout2 Þ dIC1 ðt sout2 Þ dVs ðtÞ dIRs 1 1

R1 þ R1 ¼ Rs I C1 ; I C1
dt dt dt dt C1 C1
dX1
¼ L1
dt
1 dVs ðtÞ dIRs 1 dX2 d 2 IL2 ðt sout1 Þ

I Cn ¼ Rs I L 2 L2 Ln1
Cn dt dt C2 dt dt2
1 dX3 dX3 1 Rload

ICn X3 Rload ¼ Ln2 ) ¼ ICn X3
Cn dt dt Cn Ln2 Ln2
IRs ðt sin Þ ¼ IC2 ðtÞ þ IL1 ðtÞ þ IC1 ðtÞ; ILn1 ¼ ILn2 þ ICn
Remark 1.0 IL1 ðt sout2 Þ ¼ IL1 þ iL1 ekt eksout2 ; IL1 ðtÞ ¼ IL1 þ iL1 ekt
dIL1 ðtÞ dIL1 ðt sout2 Þ dIL1 ðt sout2 Þ

¼ iL1 k ekt ; ¼ iL1 k ekt eksout2 ;
dt dt dt
dIL1 ðtÞ ksout2
¼ e
dt
dIL2 ðtÞ dIL2 ðt sout1 Þ dIL2 ðt sout1 Þ

¼ iL2 k ekt ; ¼ iL2 k ekt eksout1 ;
dt dt dt
dIL2 ðtÞ ksout1
¼ e
dt
d 2 IL2 ðtÞ d 2 IL2 ðt sout1 Þ d 2 IL2 ðt sout1 Þ

2
¼ iL2 k2 ekt ; 2
¼ iL2 k2 ekt eksout1 ;
dt dt dt2
d IL2 ðtÞ ksout1
2
¼ e
dt2
Remark 1.3 IC1 ðt sout2 Þ ¼ IC 1 þ iC1 ekt eksout2 ; IC1 ðtÞ ¼ IC 1 þ iC1 ekt
dIC1 ðtÞ dIC1 ðt sout2 Þ dIC1 ðt sout2 Þ

¼ iC1 k ekt ; ¼ iC1 k ekt eksout2 ;
dt dt dt
dIC1 ðtÞ ksout2
¼ e
dt
Remark 1.4 IRs ðt sin Þ ¼ IRs þ iRs ekt eksin ; IRs ðtÞ ¼ IRs þ iRs ekt
dIRs ðtÞ dIRs ðt sin Þ dIRs ðt sin Þ

¼ iRs k ekt ; ¼ iRs k ekt eksin ;
dt dt dt
dIRs ðtÞ ksin
¼ e
dt
dIL1 dIL1 ðtÞ dIL2 dIL2 ðtÞ dX2 d 2 IL2 d 2 IL2 ðtÞ dIC1 dIC1 ðtÞ
¼ ; ¼ ; ¼ 2 ¼ ; ¼
dt dt dt dt dt dt dt2 dt dt
&&& dX3
dt ¼ Cn L
1
n2
ICn X3 RLload
n2
dIRs ðt sin Þ dIL2 ðtÞ dIL1 ðtÞ dIC1 ðtÞ dIRs ðtÞ X2
dIC1 ksin
¼ þ þ ; ¼½ Xk eksin þ e
dt dt dt dt dt k¼1
dt
dIL1 ðt sout2 Þ dIC1 ðt sout2 Þ dVs ðtÞ dIRs 1 1

R1 þ R1 ¼ Rs I C1 ; I C1
dt dt dt dt C1 C1
dX1
¼ L1
dt
dIL1 ðtÞ ksout2 dIC1 ðtÞ ksout2 dVs ðtÞ dIRs 1 dX1
e R1 þ e R1 ¼ Rs I C1 ;
dt dt dt dt C1 dt
1
¼ I C1
L1 C1
dIC1 ksout2 dVs ðtÞ X2

dIC1 ksin
X1 eksout2 R1 þ e R1 ¼ f½ Xk eksin þ e g Rs
dt dt k¼1
dt
1
I C1
C1
dIC1 dVs ðtÞ X 2

1
½eksout2 R1 þ Rs eksin ¼ ½ Xk Rs eksin IC1 X1
dt dt k¼1
C1
eksout2 R1
dIC1 1 dVs ðtÞ X2a Rs

¼ ks X1 ks eksin
dt ðe out2 R1 þ Rs eksin Þ dt ðe out2 R1 þ Rs eksin Þ
1
ks
IC1 ; n1 ðR1 ; Rs ; sout2 ; sin ; kÞ
C1 ðe out2 R1 þ Rs eksin Þ
¼ eksout2 R1 þ Rs eksin
dIC1 1 dVs ðtÞ X2 Rs ksin 1

n1 ¼ n1 ðR1 ; Rs ; sout2 ; sin ; kÞ; ¼ X1 e I C1
dt n1 dt n1 C 1 n1
1 dVs ðtÞ dIRs 1 dX2 d 2 IL2 ðt sout1 Þ

I Cn ¼ Rs I L 2 L2 Ln1
Cn dt dt C2 dt dt2
1 dVs ðtÞ dIRs 1 dX2 d 2 IL2 ðtÞ ksout1

ICn ¼ Rs I L 2 L2 Ln1 e
Cn dt dt C2 dt dt2
1 dVs ðtÞ dIRs 1 dX2 dX2 ksout1

I Cn ¼ Rs I L 2 L2 Ln1 e
Cn dt dt C2 dt dt
dX2 dVs ðtÞ dIRs 1 1

½L2 þ Ln1 eksout1 ¼ Rs IL I Cn
dt dt dt C2 2 Cn
dX2 dVs ðtÞ X 2

dIC1 ksin 1
½L2 þ Ln1 eksout1 ¼ ½ Xk eksin Rs e Rs I L2
dt dt k¼1
dt C2
1
I Cn
Cn
dIL1 dIL2 dILn2 dX1 1

¼ X1 ; ¼ X2 ; ¼ X3 ; ¼ IC 1
dt dt dt dt L1 C1
dIC1 1 dVs ðtÞ X2 Rs ksin 1 dX3 1 Rload
¼ X1 e IC 1 ; ¼ ICn X3
dt n1 dt n1 C 1 n1 dt Cn Ln2 Ln2
P
dX2 1 dVs ðtÞ ½ 2k¼1 Xk eksin dIC1 Rs eksin
¼ R s
dt ½L2 þ Ln1 eksout1 dt ½L2 þ Ln1 eksout1 dt ½L2 þ Ln1 eksout1
1 1
IL ICn
C2 ½L2 þ Ln1 eksout1 2 Cn ½L2 þ Ln1 eksout1
dIRs X 2
1 dVs ðtÞ X2 Rs ksin 1
¼½ Xk eksin þ ½ X1 e IC1 eksin
dt k¼1
n 1 dt n 1 C 1 n 1
dIL1 dIL2 dILn2 dIRs

At fixed points: dt ¼ 0; dt ¼ 0; dt ¼ 0; dX1
dt ¼ 0; dX3
dt ¼ 0; dX2
dt ¼ 0; dt ¼0
dVs ðtÞ
Assumption dt !e
dVs ðtÞ
X1 ¼ 0; X2 ¼ 0; X3 ¼ 0; IC 1 ¼ 0; IC 1 ¼ C1 ! e; IC n ¼ 0; IL2
dt
dVs ðtÞ
¼ C2 !e
dt
ILn1 ¼ ILn2 þ ICn ) ICn ¼ ILn1 ILn2 ; ICn ¼ IL2 ðt sout1 Þ ILn2
dIL1 dIL2 dILn2 dX1 1

¼ X1 ; ¼ X2 ; ¼ X3 ; ¼ I C1
dt dt dt dt L1 C1
¼ X1 e IC1 ;
dt n1 dt n1 C1 n1
dX3 1 Rload
¼ ½IL2 ðt sout1 Þ ILn2 X3
dt Cn Ln2 Ln2
P
dX2 1 dVs ðtÞ ½ 2k¼1 Xk eksin
¼ Rs
dt ½L2 þ Ln1 eksout1 dt ½L2 þ Ln1 eksout1
1 dVs ðtÞ X2 Rs ksin 1 Rs eksin
½ X1 e IC1
n1 dt n1 C1 n1 ½L2 þ Ln1 eksout1
1 1
IL ½IL2 ðt sout1 Þ ILn2
dIRs X 2
1 dVs ðtÞ X2 Rs ksin 1
¼½ Xk eksin þ ½ X1 e IC1 eksin
dt k¼1
n 1 dt n 1 C 1 n1
equilibrium points of BPF diplexer circuit consists in adding to its coordinated
[X1 X2 X3 IL1 IL2 ILn2 IC1 IRS ] arbitrarily small increments of exponential terms
½x1 x2 x3 iL1 iL2 iLn2 iC1 iRS ekt , and retaining the first order terms in
x1 x2 x3 iL1 iL2 iLn2 iC1 iRS .
The system of eight homogeneous equations leads to a polynomial characteristic

equation in the eigenvalue k. The polynomial characteristic equation accepts by set
the BPF diplexer circuit equations. The BPF diplexer circuit fixed values with
arbitrarily small increments of exponential form ½x1 x2 x3 iL1 iL2 iLn2 iC1 iRS ekt are;
i = 0 (first fixed point), i = 1 (second fixed point), i = 2 (third fixed point), etc.,
[2–4].
ðiÞ ðiÞ
X1 ðtÞ ¼ X1 þ x1 ekt ; X2 ðtÞ ¼ X2 þ x2 ekt ;
ðiÞ ðiÞ
X3 ðtÞ ¼ X3 þ x3 ekt ; IL1 ðtÞ ¼ IL1 þ iL1 ekt
ðiÞ ðiÞ
IL2 ðtÞ ¼ IL2 þ iL2 ekt ; ILn2 ðtÞ ¼ ILn2 þ iLn2 ekt ;
ðiÞ ðiÞ
IC1 ðtÞ ¼ IC1 þ iC1 ekt ; IRs ¼ IRs þ iRs ekt
ðiÞ dIL1
IL2 ðt sout1 Þ ¼ IL2 þ iL2 ekðtsout1 Þ ; ¼ iL1 k ekt ;
dt
dIL2 dILn2
¼ iL2 k ekt ; ¼ iLn2 k ekt
dt dt
dX1 dX2 dX3 dIC1
¼ x1 k ekt ; ¼ x2 k ekt ; ¼ x3 k ekt ; ¼ iC1 k ekt ; 8 i
dt dt dt dt
¼ 0; 1; 2; . . .
For k < 0, t > 0 the selected fixed point is stable otherwise k > 0, t > 0 unstable.
Our BPF diplexer circuit tends to the selected fixed point exponentially for k < 0,
t > 0 otherwise go away from the selected fixed point exponentially. k is the
eigenvalue parameter which establish if the fixed point is stable or unstable,
additionally his absolute value |k| establish the speed of flow toward or away from
the selected fixed point [4].
dIL1 ðiÞ ðiÞ

¼ X1 ) iL1 k ekt ¼ X1 þ x1 ekt ; X1 ¼ X1 ¼ 0; iL1 k þ x1 ¼ 0
dt
dIL2 ðiÞ ðiÞ
¼ X2 ) iL2 k ekt ¼ X2 þ x2 ekt ; X2 ¼ X2 ¼ 0; iL2 k þ x2 ¼ 0
dt
dILn2 ðiÞ ðiÞ
¼ X3 ) iLn2 k ekt ¼ X3 þ x3 ekt ; X3 ¼ X3 ¼ 0; iLn2 k þ x3 ¼ 0
dt
dX1 1 1 ðiÞ ðiÞ
¼ IC1 ) x1 k ekt ¼ ½I þ iC1 ekt ; IC1 ¼ IC 1
dt L1 C1 L1 C1 C1
1
¼ 0; x1 k þ i C1 ¼ 0
L1 C1

¼ X1 e I C1
dt n1 dt n1 C1 n1
ðiÞ
1 dVs ðtÞ ðiÞ ½X þ x2 ekt Rs ksin 1
iC1 k ekt ¼ ½X1 þ x1 ekt 2 e
n1 dt n1 C1 n1
ðiÞ
½IC1 þ iC1 ekt
ðiÞ
1 dVs ðtÞ ðiÞ X Rs ksin 1 ðiÞ
iC1 k ekt ¼ X1 2 e I
n1 dt n1 C1 n1 C1
x2 ekt Rs ksin 1
x1 ekt e iC1 ekt
n1 C 1 n1
ðiÞ
ðiÞ X2 Rs ðiÞ
At fixed point: 1
n1 dVdts ðtÞ X1 n1 eksin C11n1 IC1 ¼ 0
Rs ksin 1
iC1 k x1 x2 e i C1 ¼ 0
n1 C 1 n1
dX3 1 Rload
¼ ½IL2 ðt sout1 Þ ILn2 ðtÞ X3 ; lim IL ðt sout1 Þ ¼ ILn2 ðtÞ; t
dt Cn Ln2 Ln2 t!1 2
sout1
1 ðiÞ ðiÞ ðiÞ

x3 k ekt ¼ ½I þ iL2 ekðtsout1 Þ ILn2 iLn2 ekt ½X3 þ x3 ekt
Cn Ln2 L2
Rload

Ln2
1 ðiÞ ðiÞ ðiÞ Rload 1

x3 k ekt ¼ ½I ILn2 X3 þ
Cn Ln2 L2 Ln2 Cn Ln2
Rload kt
½iL2 ekðtsout1 Þ iLn2 ekt x3 e
Ln2
ðiÞ ðiÞ ðiÞ
At fixed point: 1
Cn Ln2 ½IL2 ILn2 X3 RLload
n2
¼0
Rload 1 1
x3 k x3 þ iL eksout1 iL ¼ 0
Ln2 Cn Ln2 2 Cn Ln2 n2
P
dX2 1 dVs ðtÞ ½ 2k¼1 Xk eksin
¼ Rs
dt ½L2 þ Ln1 eksout1 dt ½L2 þ Ln1 eksout1
1 dVs ðtÞ X2 Rs ksin 1 Rs eksin
½ X1 e IC1
1 1
IL ½IL2 ðt sout1 Þ ILn2
P2 ðiÞ
1 dVs ðtÞ ½ ðXk þ xk ekt Þ eksin
x2 k ekt ¼ k¼1
Rs
½L2 þ Ln1 eksout1 dt ½L2 þ Ln1 eksout1
ðiÞ
1 dVs ðtÞ ðiÞ ½X þ x2 ekt Rs ksin
f ½X1 þ x1 ekt 2 e
n1 dt n1
1 ðiÞ Rs eksin
½IC1 þ iC1 ekt g
C 1 n1 ½L2 þ Ln1 eksout1
1 ðiÞ
½I þ iL2 ekt
C2 ½L2 þ Ln1 eksout1 L2
1 ðiÞ ðiÞ
½I þ iL2 ekðtsout1 Þ ILn2 iLn2 ekt
Cn ½L2 þ Ln1 eksout1 L2
P ðiÞ
kt 1 dVs ðtÞ ½ 2k¼1 Xk eksin
x2 k e ¼ Rs
ðiÞ
1 dVs ðtÞ ðiÞ X Rs ksin 1 ðiÞ Rs eksin
f X1 2 e IC1 g
n1 dt n1 C1 n 1 ½L2 þ Ln1 eksout1
1 ðiÞ 1 ðiÞ ðiÞ
I ½I ILn2
C2 ½L2 þ Ln1 eksout1 L2 Cn ½L2 þ Ln1 eksout1 L2
P
½ 2k¼1 xk eksin ekt x2 ekt Rs ksin
ks
Rs fx1 ekt e
½L2 þ Ln1 e out1 n1
1 Rs eksin
iC1 ekt g
C 1 n1 ½L2 þ Ln1 eksout1
1
iL ekt
C2 ½L2 þ Ln1 eksout1 2
1
½iL2 eksout1 iLn2 ekt
Cn ½L2 þ Ln1 eksout1
At fixed point:
P ðiÞ
1 dVs ðtÞ ½ 2k¼1 Xk eksin
Rs
ðiÞ
1 dVs ðtÞ ðiÞ X Rs ksin 1 ðiÞ Rs eksin
f X1 2 e I C1 g
ks
I L2 ½I ILn2 ¼ 0
C2 ½L2 þ Ln1 e out1 Cn ½L2 þ Ln1 eksout1 L2
P
½ 2k¼1 xk eksin x2 Rs ksin
x2 k Rs fx1 e
½L2 þ Ln1 eksout1 n1
1 Rs eksin
iC1 g
C1 n1 ½L2 þ Ln1 eksout1
1
iL
C2 ½L2 þ Ln1 eksout1 2
1
½iL2 eksout1 iLn2 ¼ 0
ðRn1s eksin 1Þ eksin

1 Rs eksin
x2 k þ x2 Rs þ
iC
½L2 þ Ln1 eksout1
C1 n1 ½L2 þ Ln1 eksout1 1
1 1 ksout1 1 1
ð þ e Þ iL þ iL ¼ 0
C2 Cn ½L2 þ Ln1 eksout1 2 Cn ½L2 þ Ln1 eksout1 n2
Remark: The last differential equation dIdtRs ¼ . . . is not essential for our circuit
stability analysis.
We can summery our BPF diplexer arbitrarily small increments equations.
iL1 k þ x1 ¼ 0; iL2 k þ x2 ¼ 0; iLn2 k þ x3 ¼ 0;

1 Rs
iC1 k iC1 x1 x2 eksin ¼ 0
C1 n1 n1
1 Rload
x1 k þ iC1 ¼ 0; x3 k x3
L1 C1 Ln2
1 1
þ iL eksout1 iL ¼ 0
Cn Ln2 2 Cn Ln2 n2

1 Rs eksin
x2 k þ x2 Rs þ
iC
½L2 þ Ln1 eksout1
C1 n1 ½L2 þ Ln1 eksout1 1
1 1 ksout1 1 1
ð þ e Þ iL þ iL ¼ 0
C2 Cn ½L2 þ Ln1 eksout1 2 Cn ½L2 þ Ln1 eksout1 n2
0 1
iL1
B i C
B L C
0 1 B 2C
N11 ... N17 B iLn2 C
B C
B .. .. .. C B C
@ . . . A B i C C ¼ 0 ; N11 ¼ k ; N12 ¼ N13 ¼ N14 ¼ 0 ; N15
B 1C
N71 N77 B x1 C
B C
B C
@ x2 A
x3
¼ 1 ; N16 ¼ N17 ¼ 0
N21 ¼ 0; N22 ¼ k; N23 ¼ N24 ¼ N25 ¼ 0; N26 ¼ 1; N27 ¼ 0; N31 ¼ N32
¼ 0; N33 ¼ k
N34 ¼ N35 ¼ N36 ¼ 0; N37 ¼ 1; N41 ¼ N42 ¼ N43 ¼ 0;

1
N44 ¼ k ; N45 ¼ 1
C 1 n1
Rs 1
N46 ¼ eksin ; N47 ¼ 0; N51 ¼ N52 ¼ N53 ¼ 0; N54 ¼ ;
n1 L1 C1
N55 ¼ k; N56 ¼ N57 ¼ 0
1 1 ksout1 1
N61 ¼ 0; N62 ¼ ð þ e Þ ; N63
C2 Cn ½L2 þ Ln1 eksout1
1
¼
1 Rs eksin
N64 ¼ ; N65 ¼ 0; N66
ksin ksin
ðn e
Rs
1Þ e
¼ k þ 1 Rs ; N67 ¼ 0
½L2 þ Ln1 eksout1
1 1
N71 ¼ 0; N72 ¼ eksout1 ; N73 ¼ ; N74 ¼ N75 ¼ N76 ¼ 0; N77
Cn Ln2 Cn Ln2
Rload
¼ k
Ln2
We define for simplicity global parameters in our system:
n1 ¼ n1 ðR1 ; Rs ; sout2 ; sin ; kÞ¼eksout2 R1 þ Rs eksin
ðC12 þ 1
Cn eksout1 Þ
w1 ¼ w1 ðC2 ; Cn ; L2 ; Ln1 ; sout1 ; kÞ ¼
½L2 þ Ln1 eksout1
1
w2 ¼ w2 ðCn ; L2 ; Ln1 ; sout1 ; kÞ ¼
1 Rs eksin
w3 ¼ w3 ðC1 ; R1 ; Rs ; L2 ; Ln1 ; sin ; sout1 ; sout2 ; kÞ ¼
1 Rs eksin
¼
C1 ½eksout2 R1 þ Rs eksin ½L2 þ Ln1 eksout1

w4 ¼ w4 ðRs ; R1 ; L2 ; Ln1 ; sout1 ; sout2 ; kÞ ¼ Rs
½L2 þ Ln1 eksout1
ð½eksout2 RRsþ R eksin eksin 1Þ eksin
¼ 1 s
Rs
½L2 þ Ln1 eksout1
N62 ¼ w1 ¼ w1 ðC2 ; Cn ; L2 ; Ln1 ; sout1 ; kÞ; N63 ¼ w2 ¼ w2 ðCn ; L2 ; Ln1 ; sout1 ; kÞ
N64 ¼ w3 ¼ w3 ðC1 ; R1 ; Rs ; L2 ; Ln1 ; sin ; sout1 ; sout2 ; kÞ; N66 ¼ w4

¼ w4 ðRs ; R1 ; L2 ; Ln1 ; sout1 ; sout2 ; kÞ
0 1 0 1
N11 . . . N17 N11 . . . N17
B . .. .. C B .. .. .. C
ðA k IÞ ¼ @ .. . . A ; detðA k IÞ ¼ det@ . . . A¼0
N71 N77 N71 N77
0 1
N11 . . . N17
B . .. .. C
B
detðA k IÞ ¼ det@ .. C
. . A ¼ ðkÞ fðkÞ ½ðkÞ P1 ðkÞ þ P2 ðkÞ
N71 N77
½k P3 ðkÞ þ P4 ðkÞg
Rload 1
P1 ðkÞ ¼ k4 þ k3 ½ w4 þ þ k2
Ln2 C1 n1
1 Rload Rload 1 Rs
½ ð w4 Þ w4 þ þ esin k w3
C1 n1 Ln2 Ln2 L1 C1 n1
Rload 1 w4 Rload Rload Rs ksin
þ k ½ð w4 Þ þ e w3
Ln2 L1 C1 C1 n1 Ln2 n1 Ln2
Rload
w4
Ln2 L1 C1
1 1 w 1 Rs
P2 ðkÞ ¼ ½k3 þ k2 ðw4 Þþk ð 4 eksin w3 Þ
Cn Ln2 C1 n1 C1 n1 L1 C1 n1
1
þ w4
L1 C1
1 1 1
P2 ðkÞ ¼ k3 þ k2 ðw4 Þ
Cn Ln2 Cn Ln2 C1 n1
1 w 1 Rs
þk ð 4 eksin w3 Þ
Cn Ln2 C1 n1 L1 C1 n1
1 1
þ w4
Cn Ln2 L1 C1
Table 7.1 BPF diplexer circuit cases for sin, sout1, sout2
Case no. sin sout1 sout2
1 0 0 0
2 s 0 0
3 0 s 0
4 0 0 s
5 s s 0
6 0 s s
7 s 0 s
8 s s s
Rload 1
P3 ðkÞ ¼ w1 k3 þ k2 w1 ð þ Þ þ k w1
Ln2 C1 n1
Rload 1 w1 Rload
ð þ Þþ
C1 n1 Ln2 L1 C1 L1 C1 Ln2
1 1
P4 ðkÞ ¼ k2 ½ w þ w2 eksout1
Cn Ln2 1 Cn Ln2
1 1
þk ½w1 þ w2 eksout1
C1 n1 Cn Ln2
1 1
þ ½w1 þ w2 eksout1
L1 C1 Cn Ln2
detðA k IÞ ¼ k3 P1 ðk; sin ; sout1 ; sout2 Þ þ k2 P2 ðk; sin ; sout1 ; sout2 Þ
þ k2 P3 ðk; sin ; sout1 ; sout2 Þ þ k P4 ðk; sin ; sout1 ; sout2 Þ
We analyze our BPF diplexer circuit for the following cases:

Case No. 1: sin ¼ 0; sout1 ¼ 0; sout2 ¼ 0:
n1 ðR1 ; Rs ; sout2 ¼ 0; sin ¼ 0; kÞ¼R1 þ Rs ; w1 ¼ w1 ðC2 ; Cn ; L2 ; Ln1 ; sout1 ¼ 0; kÞ

ð1 þ 1Þ
¼ C2 Cn
½L2 þ Ln1
1
w2 ¼ w2 ðCn ; L2 ; Ln1 ; sout1 ¼ 0; kÞ ¼
Cn ½L2 þ Ln1
w3 ¼ w3 ðC1 ; R1 ; Rs ; L2 ; Ln1 ; sin ¼ 0; sout1 ¼ 0; sout2 ¼ 0; kÞ

1 Rs 1 Rs
¼ ¼
C1 n1 ðsout2 ¼ 0; sin ¼ 0Þ ½L2 þ Ln1 C1 ½R1 þ Rs ½L2 þ Ln1
ðRn1s 1Þ
w4 ¼ w4 ðRs ; R1 ; L2 ; Ln1 ; sout1 ¼ 0; sout2 ¼ 0; kÞ ¼ Rs
½L2 þ Ln1
ð½R1 Rþs Rs 1Þ
¼ Rs
½L2 þ Ln1
1 1
P2 ðk; sin ¼ 0; sout1 ¼ 0; sout2 ¼ 0Þ ¼ k3 þ k2
Cn Ln2 Cn Ln2
1
ðw4 ð; sin ¼ 0; sout1 ¼ 0; sout2 ¼ 0Þ Þ
C1 n1 ð; sin ¼ 0; sout1 ¼ 0; sout2 ¼ 0Þ
1 w ð; sin ¼ 0; sout1 ¼ 0; sout2 ¼ 0Þ 1
þk ð 4
Cn Ln2 C1 n1 ð; sin ¼ 0; sout1 ¼ 0; sout2 ¼ 0Þ L1 C1
Rs
w ð; sin ¼ 0; sout1 ¼ 0; sout2 ¼ 0ÞÞ
n1 ð; sin ¼ 0; sout1 ¼ 0; sout2 ¼ 0Þ 3
1 1
þ w ð; sin ¼ 0; sout1 ¼ 0; sout2 ¼ 0Þ
Cn Ln2 4 L1 C1
Rload
P1 ðk; sin ¼ 0; sout1 ¼ 0; sout2 ¼ 0Þ ¼ k4 þ k3 ½ w4 ðsin ¼ 0; sout1 ¼ 0; sout2 ¼ 0Þ
Ln2
1 1
þ þ k2 ½
C1 n1 ðsin ¼ 0; sout1 ¼ 0; sout2 ¼ 0Þ C1 n1 ðsin ¼ 0; sout1 ¼ 0; sout2 ¼ 0Þ
Rload
ð w4 ðsin ¼ 0; sout1 ¼ 0; sout2 ¼ 0ÞÞ w4 ðsin ¼ 0; sout1 ¼ 0; sout2 ¼ 0Þ
Ln2
Rload 1 Rs
þ þ
Ln2 L1 C1 n1 ðsin ¼ 0; sout1 ¼ 0; sout2 ¼ 0Þ
esin k w3 ðsin ¼ 0; sout1 ¼ 0; sout2 ¼ 0Þ þ k
Rload
½ð w4 ðsin ¼ 0; sout1 ¼ 0; sout2 ¼ 0ÞÞ
Ln2
1 w ðsin ¼ 0; sout1 ¼ 0; sout2 ¼ 0Þ Rload
4
L1 C1 C1 n1 ðsin ¼ 0; sout1 ¼ 0; sout2 ¼ 0Þ Ln2
Rload Rs
þ
n1 ðsin ¼ 0; sout1 ¼ 0; sout2 ¼ 0Þ Ln2
eksin w3 ðsin ¼ 0; sout1 ¼ 0; sout2 ¼ 0Þ w4 ðsin ¼ 0; sout1 ¼ 0; sout2 ¼ 0Þ
Rload

Ln2 L1 C1
P3 ðk; sin ¼ 0; sout1 ¼ 0; sout2 ¼ 0Þ ¼ w1 ðsin ¼ 0; sout1 ¼ 0; sout2 ¼ 0Þ k3

þ k2 w1 ðsin ¼ 0; sout1 ¼ 0; sout2 ¼ 0Þ
Rload 1
ð þ Þ
Ln2 C1 n1 ðsin ¼ 0; sout1 ¼ 0; sout2 ¼ 0Þ
þ k w1 ðsin ¼ 0; sout1 ¼ 0; sout2 ¼ 0Þ
Rload 1
ð þ Þ
C1 n1 ðsin ¼ 0; sout1 ¼ 0; sout2 ¼ 0Þ Ln2 L1 C1
w ðsin ¼ 0; sout1 ¼ 0; sout2 ¼ 0Þ Rload
þ 1
L1 C1 Ln2
1
P4 ðk; sin ¼ 0; sout1 ¼ 0; sout2 ¼ 0Þ ¼ k2 ½ w ðsin ¼ 0; sout1 ¼ 0; sout2 ¼ 0Þ
Cn Ln2 1
1
þ w2 ðsin ¼ 0; sout1 ¼ 0; sout2 ¼ 0Þ
Cn Ln2
1 1
þk
C1 n1 ðsin ¼ 0; sout1 ¼ 0; sout2 ¼ 0Þ Cn Ln2
½w1 ðsin ¼ 0; sout1 ¼ 0; sout2 ¼ 0Þ
1 1
þ w2 ðsin ¼ 0; sout1 ¼ 0; sout2 ¼ 0Þ þ
L1 C1 Cn Ln2
½w1 ðsin ¼ 0; sout1 ¼ 0; sout2 ¼ 0Þ
þ w2 ðsin ¼ 0; sout1 ¼ 0; sout2 ¼ 0Þ
We can summery our w1 ; w2 ; w3 ; w4 expression for sin ¼ 0; sout1 ¼ 0; sout2 ¼ 0:
ðC12 þ Cn Þ
1
1
w1 ðsout1 ¼ 0Þ ¼ ; w2 ðsout1 ¼ 0Þ ¼ ; n ðsout2 ¼ 0; sin
½L2 þ Ln1 Cn ½L2 þ Ln1 1
¼ 0Þ ¼ R1 þ Rs
1 Rs
w3 ðsin ¼ 0; sout1 ¼ 0; sout2 ¼ 0Þ ¼ ; w ðsout1 ¼ 0; sout2
C1 ½R1 þ Rs ½L2 þ Ln1 4
ð½R1 Rþs Rs 1Þ
¼ 0Þ ¼ Rs
½L2 þ Ln1
&&&
Rload ð½R1 Rþs Rs 1Þ

P1 ðk; sin ¼ 0; sout1 ¼ 0; sout2 ¼ 0Þ ¼ k þ k ½
4 3
ð Rs Þ
Ln2 ½L2 þ Ln1
1 1 Rload ð½R Rþs R 1Þ
þ þ k2 ½ ð ð 1 s Rs ÞÞ
C1 ðR1 þ Rs Þ C1 ðR1 þ Rs Þ Ln2 ½L2 þ Ln1
ð½R1 Rþs Rs 1Þ Rload 1 Rs 1 Rs
ð Rs Þ þ þ ð Þ
½L2 þ Ln1 Ln2 L1 C1 ðR1 þ Rs Þ C1 ½R1 þ Rs ½L2 þ Ln1
ð Rs
1Þ
Rload ð½R Rþs R 1Þ 1 ð ½R½L12þþRsL n1 Rs Þ Rload
þ k ½ð ð 1 s Rs ÞÞ
Ln2 ½L2 þ Ln1 L1 C1 C1 ðR1 þ Rs Þ Ln2
Rload Rs 1 Rs
þ ð Þ
ðR1 þ Rs Þ Ln2 C1 ½R1 þ Rs ½L2 þ Ln1
ð½R1 Rþs Rs 1Þ Rload
ð Rs Þ
½L2 þ Ln1 Ln2 L1 C1
We define for simplicity new global parameters:
Rload ð½R Rþs R 1Þ 1

A4 ¼ 1; A3 ¼ ð 1 s Rs Þ þ ; A0
Ln2 ½L2 þ Ln1 C1 ðR1 þ Rs Þ
ð½R Rþs R 1Þ Rload
¼ ð 1 s Rs Þ
½L2 þ Ln1 Ln2 L1 C1
1 Rload ð½R Rþs R 1Þ ð½R Rþs R 1Þ Rload

A2 ¼ ð ð 1 s Rs ÞÞ ð 1 s Rs Þ
C1 ðR1 þ Rs Þ Ln2 ½L2 þ Ln1 ½L2 þ Ln1 Ln2
1 Rs 1 Rs
þ þ ð Þ
L1 C1 ðR1 þ Rs Þ C1 ½R1 þ Rs ½L2 þ Ln1
ð Rs
1Þ
Rload ð½R1 Rþs Rs 1Þ 1 ð ½R½L12þþRsL n1 Rs Þ Rload
A1 ¼ ð ð Rs ÞÞ
Ln2 ½L2 þ Ln1 L1 C1 C1 ðR1 þ Rs Þ Ln2
Rload Rs 1 Rs
þ ð Þ
ðR1 þ Rs Þ Ln2 C1 ½R1 þ Rs ½L2 þ Ln1
X
4
P1 ðk; sin ¼ 0; sout1 ¼ 0; sout2 ¼ 0Þ ¼ kk Ak
k¼0
¼ k4 jA4 ¼1 þ k3 A3 þ k2 A2 þ k A1 þ A0
1 1
P2 ðk; sin ¼ 0; sout1 ¼ 0; sout2 ¼ 0Þ ¼ k3 þ k 2
Cn Ln2 Cn Ln2
ð½R1 Rþs Rs 1Þ 1 1
ð Rs Þþk
½L2 þ Ln1 C1 ðR1 þ Rs Þ Cn Ln2
ð½RRs
1Þ
1 þ Rs
½L2 þ Ln1 Rs 1 Rs 1 Rs
ð Þ
C1 ðR1 þ Rs Þ L1 C1 ðR1 þ Rs Þ C1 ½R1 þ Rs ½L2 þ Ln1
1 ð½R1 Rþs Rs 1Þ 1
þ Rs
Cn Ln2 ½L2 þ Ln1 L1 C1
1 1 ð½R Rþs R 1Þ 1
B3 ¼ ; B2 ¼ ð 1 s Rs Þ
Cn Ln2 Cn Ln2 ½L2 þ Ln1 C1 ðR1 þ Rs Þ
ð½RRs
1Þ
½L þ L Rs
1 þ Rs
1 1 Rs 1 Rs
B1 ¼ ð 2 n1 Þ
Cn Ln2 C1 ðR1 þ Rs Þ L1 C1 ðR1 þ Rs Þ C1 ½R1 þ Rs ½L2 þ Ln1
1 ð½R Rþs R 1Þ 1
B0 ¼ 1 s Rs
Cn Ln2 ½L2 þ Ln1 L1 C1
X
3
P2 ðk; sin ¼ 0; sout1 ¼ 0; sout2 ¼ 0Þ ¼ kk Bk ¼ k3 B3 þ k2 B2 þ k B1 þ B0
k¼0
ðC12 þ Cn Þ
1
ðC12 þ Cn Þ
1
P3 ðk; sin ¼ 0; sout1 ¼ 0; sout2 ¼ 0Þ ¼ k 3 k2
½L2 þ Ln1 ½L2 þ Ln1
Rload 1 ð 1
þ 1
Þ Rload 1
ð þ Þ k C 2 Cn ð þ Þ
Ln2 C1 ðR1 þ Rs Þ ½L2 þ Ln1 C1 ðR1 þ Rs Þ Ln2 L1 C1
ðC1 þ C1n Þ
2
½L2 þ Ln1 Rload

L1 C1 Ln2
ðC12 þ Cn Þ
1
ðC12 þ Cn Þ
1
Rload 1
C3 ¼ ; C2 ¼ ð þ Þ
½L2 þ Ln1 ½L2 þ Ln1 Ln2 C1 ðR1 þ Rs Þ
ð1 þ 1
Þ
ðC12 þ Cn Þ ½L þ L Rload
1 C2 Cn
Rload 1
C1 ¼ ð þ Þ; C0 ¼ 2 n1
½L2 þ Ln1 C1 ðR1 þ Rs Þ Ln2 L1 C1 L1 C1 Ln2
X
3
P3 ðk; sin ¼ 0; sout1 ¼ 0; sout2 ¼ 0Þ ¼ kk Ck ¼ k3 C3 þ k2 C2 þ k C1 þ C0
k¼0
1 ð1 þ 1Þ
P4 ðk; sin ¼ 0; sout1 ¼ 0; sout2 ¼ 0Þ ¼ k2 ½ ð C2 Cn Þ
Cn Ln2 ½L2 þ Ln1
1 1 1 1 ð1 þ 1Þ
þ þk ½ C2 Cn
Cn ½L2 þ Ln1 Cn Ln2 C1 ðR1 þ Rs Þ Cn Ln2 ½L2 þ Ln1
1 1 1 ð þ Þ
1 1
1
þ þ ½ C2 Cn þ
Cn ½L2 þ Ln1 L1 C1 Cn Ln2 ½L2 þ Ln1 Cn ½L2 þ Ln1
1 ð1 þ 1Þ 1 1
D2 ¼ ð C 2 Cn Þ þ
Cn Ln2 ½L2 þ Ln1 Cn ½L2 þ Ln1 Cn Ln2
1 1 ð1 þ 1Þ 1
D1 ¼ ½ C2 Cn þ
C1 ðR1 þ Rs Þ Cn Ln2 ½L2 þ Ln1 Cn ½L2 þ Ln1
1 1 ð1 þ 1Þ 1
D0 ¼ ½ C2 Cn þ
L1 C1 Cn Ln2 ½L2 þ Ln1 Cn ½L2 þ Ln1
X
2
P4 ðk; sin ¼ 0; sout1 ¼ 0; sout2 ¼ 0Þ ¼ kk Dk ¼ k2 D2 þ k D1 þ D0
k¼0
X
4 X
3 X
3
detðA k IÞjsin ¼0;sout1 ¼0;sout2 ¼0 ¼ k3 kk Ak þ k2 kk B k þ k2 kk
k¼0 k¼0 k¼0
X
2
Ck þ k k Dk
k
k¼0
X
4 X
3 X
3
detðA k IÞjsin ¼0;sout1 ¼0;sout2 ¼0 ¼ kk þ 3 Ak þ kk þ 2 Bk þ kk þ 2
k¼0 k¼0 k¼0
X
2
Ck þ kk þ 1 Dk
k¼0
X
3 X
2 X
3 X
2
kk þ 2 Ck ¼ k5 C3 þ kk þ 2 C k ; kk þ 2 Bk ¼ k5 B3 þ kk þ 2 Bk
k¼0 k¼0 k¼0 k¼0
X
4 X
2
kk þ 3 Ak ¼ k7 A4 þ k6 A3 þ kk þ 3 A k
k¼0 k¼0
X
2
detðA k IÞjsin ¼0;sout1 ¼0;sout2 ¼0 ¼ k7 A4 k6 A3 kk þ 3 Ak þ k5 B3
k¼0
X
2 X
2 X
2
þ kk þ 2 Bk þ k5 C3 þ kk þ 2 Ck þ kk þ 1 D k
k¼0 k¼0 k¼0
detðA k IÞjsin ¼0;sout1 ¼0;sout2 ¼0 ¼ k7 A4 k6 A3 þ k5 ðB3 þ C3 Þ

X
2
½kk þ 2 Bk þ kk þ 2 Ck þ kk þ 1 Dk kk þ 3 Ak
k¼0
X
2
detðA k IÞjsin ¼0;sout1 ¼0;sout2 ¼0 ¼ k7 A4 k6 A3 kk þ 3 Ak þ k5 B3
k¼0
X
2 X
2 X
2
þ kk þ 2 Bk þ k5 C3 þ kk þ 2 Ck þ kk þ 1 D k
k¼0 k¼0 k¼0
X
2
kk þ 3 Ak ¼ k5 A2 þ k4 A1 þ k3 A0 ;
k¼0
X
2
kk þ 2 Bk ¼ k4 B2 þ k3 B1 þ k2 B0
k¼0
X
2
kk þ 2 Ck ¼ k4 C2 þ k3 C1 þ k2 C0 ;
k¼0
X
2
kk þ 1 D k ¼ k3 D 2 þ k2 D 1 þ k D 0
k¼0
detðA k IÞjsin ¼0;sout1 ¼0;sout2 ¼0 ¼ k7 A4 k6 A3 ½k5 A2 þ k4 A1 þ k3 A0

þ k5 B3 þ k4 B2 þ k3 B1 þ k2 B0 þ k5 C3 þ k4 C2
þ k3 C1 þ k2 C0 þ k3 D2 þ k2 D1 þ k D0
detðA k IÞjsin ¼0;sout1 ¼0;sout2 ¼0 ¼ k7 A4 k6 A3 þ k5 ½B3 A2 þ C3

þ k4 ½B2 A1 þ C2 þ k3 ½B1 A0 þ C1 þ D2 þ k2 ½B0 þ C0 þ D1 þ k D0
X
7 X
7
detðA k IÞ ¼ k kk ; detðA k IÞ ¼ 0 ) k kk ¼ 0
k¼0 k¼0
!7 ¼ 1; !6 ¼ A3 ; !5 ¼ B3 A2 þ C3 ; !4 ¼ B2 A1 þ C2 ; !3
¼ B1 A0 þ C1 þ D2
!2 ¼ B0 þ C0 þ D1 ; !1 ¼ D0 ; !0 ¼ 0
Eigenvalues stability discussion: Our BPF diplexer circuit involving N variables

(N > 2, N = 7), the characteristic equation is of degree N = 7 and must often be
solved numerically. Expect in some particular cases, such an equation has (N = 7)
distinct roots that can be real or complex. These values are the eigenvalues of the
7 7 Jacobian matrix (A). The general rule is that the Steady State (SS) is stable if
there is no eigenvalue with positive real part. It is sufficient that one eigenvalue is
positive for the steady state to be unstable. Our 7-variables
(X1 X2 X3 IL1 IL2 ILn2 IC1 ) BPF diplexer circuit has seven eigenvalues. The type of
behavior can be characterized as a function of the position of these eigenvalues in
the Re/Im plane. Five non-degenerated cases can be distinguished: (1) the seven
eigenvalues are real and negative (stable steady state), (2) the seven eigenvalues are
real, six of them are negative (unstable steady state), (3) and (4) two eigenvalues are
complex conjugates with a negative real part and the other eigenvalues are real and
negative (stable steady state), two cases can be distinguished depending on the
relative value of the real part of the complex eigenvalues and of the real one,
(5) two eigenvalues are complex conjugates with a negative real part and at least
one eigenvalue is positive (unstable steady state) [12, 13].
Remark It is reader exercise to analyze BPF diplexer circuit stability for cases 2–8
(at least one delay parameter sin ; sout1 ; sout2 is positive and real number). The sta-
bility analysis is done by using geometric stability switch criteria in delay differ-
ential systems (E. Beretta and Y. Kuang). It is a practical guideline that combines
graphical information with analytical work to effectively study the local stability of
BPF diplexer circuit model involving delay dependent parameters. The stability of
BPF diplexer circuit steady state is determined by the graphs of some functions of s
which can be expressed explicitly.
The general geometric criterion: The occurrence of any possible stability
switching resulting from the increase of value of the time delay s for our BPF
diplexer circuit characteristic equation.
Dðk; sin ; sout1 ; sout2 Þ ¼ detðA k IÞ; Dðk; sin ; sout1 ; sout2 Þ ¼ 0; Dðk; sÞ
¼ Pn ðk; sÞ þ Qm ðk; sÞ eks
X
n X
m
Pn ðk; sÞ ¼ pk ðsÞ kk ; Qm ðk; sÞ ¼ qk ðsÞ kk ; n; m 2 N0 ; n [ m
k¼0 k¼0
pk ð Þ; qk ð Þ : R þ 0 ! R are continuous and differentiable functions of s.
Pn ðk ¼ 0; sÞ þ Qm ðk ¼ 0; sÞ ¼ p0 ðsÞ þ q0 ðsÞ 6¼ 0 8 s 2 R þ 0
Pn ðk; sÞ; Qm ðk; sÞ are analytic functions in k and differentiable in s for which we
assume:
(I) If k ¼ i x; x 2 R then Pn ði x; sÞ þ Qm ði x; sÞ 6¼ 0; s 2 R.
(II) lim supfjQm ðk; sÞ=Pn ðk; sÞj : jkj ! 1; Rek
0g\1 for any s.
(III) Fðx; sÞ ¼ jPn ði x; sÞj2 jQm ði x; sÞj2 for each s has at most a finite
number of real zeros.
(IV) Each positive root xðsÞ of Fðx; sÞ ¼ 0 is continuous and differentiable in s
whenever it exists.
7.2 Dual Band Diplexer Filter Stability Analysis Under

A diplexer filters to pass two bands to separate ports, and its stability analysis under
parameters variation. In our RF and microwave system, two desired frequencies are
relative close together; design each separate filter to have its own band edge as far
as possible from the other filter. The desired frequency we want to pass is located at
the upper band edge of BPF1, while the desired frequency of BPF2 is at its lower
band edge.
The circuit of dual band diplexer filter contains capacitors and inductors, RF
source Vs(t) and series resistor Rs, two matching networks, and resistive loads
Rload1, Rload2. Three delay lines sin ; sout1 ; sout2 represent circuit micro strip lines
parasitic effects. In our stability analysis we consider fully match between Rload1
and BPF1 and between Rload2 and BPF2 (no matching networks) [25, 26].
Fig. 7.4 Dual band diplexer filter block diagram

Fig. 7.5 Dual band diplexer filter full schematic with delay lines
Vsin ! e; Vsout1 ! e; Vsout2 ! e; Vsin ¼ VA1 VA2 ; Vsout1 ¼ VA6 VA7 ; Vsout2
¼ VA11 VA12
Vs ðtÞ VA1 dIL dðVA3 VA4 Þ

IR s ¼ ; VL5 ¼ VA2 VA3 ¼ L5 5 ; IC1 ¼ C1 ; VA4
Rs dt dt
dIL
¼ VL2 ¼ L2 2
dt
dVA4 dIL dðVA5 VA6 Þ
IC2 ¼ C2 ; VL1 ¼ VA4 VA5 ¼ L1 1 ; IC3 ¼ C3 ; IRload1
dt dt dt
VA7
¼ ; VA1 VA2
Rload1
dIL dðVA8 VA9 Þ

VA6 VA7 ; VL6 ¼ VA2 VA8 ¼ L6 6 ; IC6 ¼ C6 ; VA9 ¼ VL4
dt dt
dIL4
¼ L4
dt
dVA9 dIL dðVA10 VA11 Þ
IC5 ¼ C5 ; VL3 ¼ VA9 VA10 ¼ L3 3 ; IC4 ¼ C4 ; IRload2
dt dt dt
VA12
¼ ; VA11 VA12
Rload2
X
6
IRs ðt sin Þ ¼ IL5 þ IL6 ¼ ILk ; IL5 ¼ IC1 ; IL6 ¼ IC6 ; IL1 ¼ IC3 ; IL3 ¼ IC4 ; IRload1 ¼ IC3 ðt sout1 Þ
k¼5
IRload1 ¼ IL1 ðt sout1 Þ; IRload2 ¼ IC4 ðt sout2 Þ; IRload2 ¼ IL3 ðt sout2 Þ
7.2 Dual Band Diplexer Filter Stability Analysis Under Parameters Variation 699
P
6
KCL @ node A2: IRs ðt sin Þ ¼ IL5 þ IL6 ¼ ILk , KCL @ node A4:
k¼5
P
2
IC1 ¼ IL2 þ IC2 þ IL1 ¼ IC2 þ I Lk
k¼1
P
4
KCL @ node A9: IC6 ¼ IL4 þ IC5 þ IL3 ¼ IC5 þ I Lk
k¼3
&&&
Vs ðtÞ VA1 dIL

IRs ¼ ) VA1 ¼ Vs ðtÞ IRs Rs ; VA2 VA3 ¼ L5 5
Rs dt
Z
dVA4 1
IC2 ¼ C2 ) VA4 ¼ IC2 dt
dt C2
Z
dðVA3 VA4 Þ 1 dIL
IC1 ¼ C1 ) VA3 VA4 ¼ IC1 dt; VA4 VA5 ¼ L1 1
dt C1 dt
dIL1
) VA5 ¼ VA4 L1
dt
Z
1 dIL dðVA5 VA6 Þ
VA5 ¼ IC2 dt L1 1 ; IC3 ¼ C3 ) VA5 VA6
C2 Z dt dt
1
¼ IC3 dt; VA7 ¼ IRload1 Rload1
C3
dIL dðVA8 VA9 Þ

VA2 VA8 ¼ L6 6 ; IC6 ¼ C6 ) VA8 VA9
Zdt dt
1 dIL
¼ IC6 dt; VA9 ¼ L4 4
C6 dt
Z
dVA9 1 dIL
IC5 ¼ C5 ) VA9 ¼ IC5 dt; VA9 VA10 ¼ L3 3
dt C5 dt
Z
dðVA10 VA11 Þ 1
IC4 ¼ C4 ) VA10 VA11 ¼ IC4 dt
dt C4
VA12 VA11
IRload2 ¼ ; VA11 VA12 ; IRload2 ¼ ) VA11 ¼ IRload2 Rload2
Rload2 Rload2
Vsin ! e; VA1 ! VA2 ; VA2 ¼ Vs ðtÞ IRs Rs ;

dIL dIL
VA3 ¼ VA2 L5 5 ¼ Vs ðtÞ IRs Rs L5 5
dt dt
dIL5 dIL5 dIL
VA2 VA3 ¼ L5 ) VA3 ¼ VA2 L5 ¼ Vs ðtÞ IRs Rs L5 5
dt dt dt
Z Z
1 dIL5 1
VA3 VA4 ¼ IC dt ) Vs ðtÞ IRs Rs L5 IC2 dt
C1 Z 1 dt C2
1
¼ IC1 dt
C1
Z Z
d dIL5 1 1
fVs ðtÞ IRs Rs L5 IC2 dt ¼ IC1 dtg
dt dt C2 C1
dVs ðtÞ dIRs d 2 IL 1 1

R s L5 2 5 I C2 ¼ I C1
dt dt dt C2 C1
Z Z
1 1
VA5 VA6 ¼ IC3 dt ) VA6 ¼ VA5 IC3 dt
C3 Z CZ3
1 dIL1 1
¼ IC2 dt L1 IC3 dt
C2 dt C3
VA6
Vsout1 ! e; VA6 ! VA7 ; IRload1 ¼ ) VA6 ¼ IRload1 Rload1
Rload1
Z Z
1 dIL 1
IC2 dt L1 1 IC3 dt ¼ IRload1 Rload1 ;
C2 dt C3
1 d 2 IL 1 dIRload1
I C 2 L1 2 1 I C3 ¼ Rload1
C2 dt C3 dt
dIL6 dIL dIL

VA2 VA8 ¼ L6 ) VA8 ¼ VA2 L6 6 ¼ Vs ðtÞ IRs Rs L6 6
dt dt dt
Z Z
1 1
C6 C6Z
dIL 1
¼ Vs ðtÞ IRs Rs L6 6 IC6 dt
dt C6
Z
dIL4 1 dIL
VA9 ¼ L4 ; VA9 ¼ IC5 dt; L4 4 ¼ Vs ðtÞ IRs Rs
dt C5 dt
Z
dIL6 1
L6 IC6 dt
dt C6
d 2 IL dVs ðtÞ dIRs d 2 IL 1
L4 2 4 ¼ Rs L6 2 6 I C6
dt dt dt dt C6
Z Z
1 dIL 1 1
IC5 dt ¼ Vs ðtÞ IRs Rs L6 6 IC6 dt; I C5
C5 dt C6 C5
dVs ðtÞ dIRs d 2 IL 1
¼ R s L6 2 6 I C6
dt dt dt C6
dIL3 dIL dIL dIL

VA9 VA10 ¼ L3 ) VA10 ¼ VA9 L3 3 ¼ L4 4 L3 3 ;
dt dt dt dt
1 d 2 I L4
I C5 ¼ L4 2
C5 dt
Z Z
1 1
C4 C4
Z
dIL4 dIL3 1
¼ L4 L3 IC4 dt
dt dt C4
Vsout2 ! e; VA11 ! VA12 ; VA11 ¼ IRload2Z Rload2 ; IRload2 Rload2

dIL4 dIL3 1
¼ L4 L3 IC4 dt
dt dt C4
Z
d dIL dIL 1 dIRload2
fIRload2 Rload2 ¼ L4 4 L3 3 IC4 dtg; Rload2
dt dt dt C4 dt
d 2 IL d 2 IL 1
¼ L 4 2 4 L3 2 3 IC4
dt dt C4
Summary of our circuit differential equations (Version 1):
dVs ðtÞ dIRs d 2 IL 1 1 1 d 2 IL 1

Rs L5 2 5 I C2 ¼ IC1 ; IC2 L1 2 1 I C3
dt dt dt C2 C1 C2 dt C3
dIRload1
¼ Rload1
dt
d 2 IL4 dVs ðtÞ dIRs d 2 I L6 1 1

L4 ¼ Rs L 6 I C6 ; I C5
dt2 dt dt dt2 C6 C5
dVs ðtÞ dIRs 2
d IL 1
¼ Rs L6 2 6 I C6
dt dt dt C6
dIRload2 d 2 IL d 2 IL 1 1 d 2 IL
Rload2 ¼ L4 2 4 L3 2 3 I C4 ; IC5 ¼ L4 2 4
dt dt dt C4 C5 dt
I C1 ! I L 5 ; I C6 ! I L 6 ; I L 1 ! I C3 ; I L 3 ! I C4
dVs ðtÞ dIRs d 2 IL 1 1 1 d 2 IC 1

R s L5 2 5 I C2 ¼ I L5 ; IC2 L1 2 3 I C3
dt dt dt C2 C1 C2 dt C3
dIRload1
¼ Rload1
dt
1 dVs ðtÞ dIRs d 2 IL 1 1 d 2 IL

I C5 ¼ Rs L6 2 6 I L6 ; IC5 ¼ L4 2 4
C5 dt dt dt C6 C5 dt
dIRload2 d 2 I L4 d 2 I C4 1
Rload2 ¼ L4 2 L3 2 I C4
dt dt dt C4
Remark 1.0 IRs ðt sin Þ ¼ IL5 þ IL6 ; IRload2 ¼ IC4 ðt sout2 Þ; IRload1 ¼ IC3 ðt sout1 Þ
dIRload2 dIC4 ðt sout2 Þ

¼ ; IC4 ðt sout2 Þ ¼ IC 4 þ iC4 ekt eksout2 ; IC4 ðtÞ
dt dt
¼ IC 4 þ iC4 ekt
dIC4 ðt sout2 Þ dIC4 ðtÞ dIC4 ðt sout2 Þ

¼ iC4 k ekt eksout2 ; ¼ iC4 k ekt ;
dt dt dt
dIC4 ðtÞ ksout2
¼ e
dt
dIRload1 dIC3 ðtsout1 Þ
Remark 1.1 dt ¼ dt ; IC3 ðt sout1 Þ ¼ IC 3 þ iC3 ekt eksout1 ; IC3 ðtÞ ¼
IC 3 þ iC3 ekt
dIC3 ðt sout1 Þ dIC3 ðtÞ dIC3 ðt sout1 Þ

¼ iC3 k ekt eksout1 ; ¼ iC3 k ekt ;
dt dt dt
dIC3 ðtÞ ksout1
¼ e
dt
dIRload1 dIC3 ðt sout1 Þ dIC3 ðtÞ ksout1
¼ ¼ e
dt dt dt
dIRs ðtsin Þ dIL5 dIL6
Remark 1.2 IRs ðt sin Þ ¼ IL5 þ IL6 ; dt ¼ dt þ dt
dIRs ðt sin Þ
IRs ðt sin Þ ¼ IRs þ iRs ekt eksin ; IRs ðtÞ ¼ IRs þ iRs ekt ;
dt
¼ iRs k ekt eksin
dIRs ðtÞ dIRs ðt sin Þ dIRs ðtÞ ksin dIRs ðtÞ ksin dIL5 dIL6
¼ iRs k ekt ; ¼ e ; e ¼ þ
dt dt dt dt dt dt
dIRs ðtÞ dIL5 ksin dIL6 ksin dIRload2 dIC4 ðt sout2 Þ dIC4 ðtÞ ksout2
¼ e þ e ; ¼ ¼ e
dt dt dt dt dt dt
dX2 d 2 IL5 dIL dX3 d 2 IC3 dIC3 dX4 d 2 IL6 dIL

¼ 2 ; X2 ¼ 5 ; ¼ 2
; X 3 ¼ ; ¼ 2 ; X4 ¼ 6
dt dt dt dt dt dt dt dt dt
dX5 d 2 IL4 dIL dX6 d 2 IC4 dIC4 dIRs ðtÞ

¼ 2 ; X5 ¼ 4 ; ¼ 2
; X6 ¼ ; ¼ X2 eksin þ X4 eksin
dVs ðtÞ dX2 1

½X2 eksin þ X4 eksin Rs L5 IC2
dt dt C2
1 1 dX3 1
¼ IL5 ; I C 2 L1 IC3 ¼ X3 eksout1 Rload1
C1 C2 dt C3
1 dVs ðtÞ dX4 1 1

I C5 ¼ ½X2 eksin þ X4 eksin Rs L6 I L6 ; I C5
C5 dt dt C6 C5
dX5 dX5 1
¼ L4 ) ¼ I C5
dt dt C5 L4
dX5 dX6 1
X6 eksout2 Rload2 ¼ L4 L3 I C4
dt dt C4
1 dX6 1
X6 eksout2 Rload2 ¼ L4 I C 5 L3 I C4
C5 L4 dt C4
dIL5 dIC dIL dIL dIC

¼ X2 ; 3 ¼ X3 ; 6 ¼ X4 ; 4 ¼ X5 ; 4 ¼ X6
dt dt dt dt dt
dX2 1 dVs ðtÞ Rs 1 1
¼ ½X2 eksin þ X4 eksin IC2 IL
dt L5 dt L5 C2 L5 C1 L5 5
dX3 1 1 Rload1 dX5 1
¼ I C2 IC3 X3 eksout1 ; ¼ IC5
dt C2 L1 C3 L1 L1 dt C5 L4

¼ ½X2 eksin þ X4 eksin I C2 IL
dt L5 dt L5 C2 L5 C1 L5 5
dX3 1 1 Rload1 dX5 1

¼ I C2 IC3 X3 eksout1 ; ¼ IC5
dt C2 L1 C3 L1 L1 dt C5 L4
dVs ðtÞ
At fixed points (equilibrium points): dt !e
dIL5 ðiÞ dIC ðiÞ dIL ðiÞ dIL ðiÞ

¼ 0 ) X2 ¼ 0; 3 ¼ 0 ) X3 ¼ 0; 6 ¼ 0 ) X4 ¼ 0; 4 ¼ 0 ) X5 ¼ 0
dt dt dt dt
dIC4 ðiÞ dX2 dVs ðtÞ 1 ðiÞ 1 ðiÞ

¼ 0 ) X6 ¼ 0; ¼0)f ! eg I I ¼0
dt dt dt C2 C2 C1 L5
dX3 1 ðiÞ 1 ðiÞ dX5 1 ðiÞ ðiÞ

¼0) I I ¼ 0; ¼0) I ¼ 0 ) I C5 ¼ 0
dt C2 C2 C3 C3 dt C5 L4 C5
dX4 dVs ðtÞ 1 ðiÞ 1 ðiÞ ðiÞ ðiÞ

¼0)f ! eg I L6 I ¼ 0; IC5 ¼ 0 ) IL6 ¼ 0
dt dt C6 C5 C5
dX6 1 ðiÞ 1 ðiÞ ðiÞ ðiÞ ðiÞ

¼0) I I ¼ 0; IC5 ¼ 0; IC4 ¼ 0; IL6 ¼ 0
dt C 5 C5 C 4 C4
We can summery our system fixed points:

Xk ¼ 0; IC5 ¼ 0; IC4 ¼ 0; IL6 ¼ 0
8k¼2;3;4;5;6
1 ðiÞ 1 ðiÞ 1 ðiÞ 1 ðiÞ

I þ I ¼ 0; I I ¼0
C 2 C2 C 1 L 5 C2 C2 C3 C3
equilibrium points of dual band diplexer filter circuit consists in adding to its
coordinated [X2 X3 X4 X5 X6 IL5 IC3 IL6 IL4 IC4 IC2 IC5 ] arbitrarily small increments
of exponential terms ½x2 x3 x4 x5 x6 iL5 iC3 iL6 iL4 iC4 iC2 iC5 ekt , and retaining the
first order terms in x2 x3 x4 x5 x6 iL5 iC3 iL6 iL4 iC4 iC2 iC5 . The system of ten ho-
valuek. The polynomial characteristic equation accepts by set the dual band
diplexer filter circuit equations. The dual band diplexer filter circuit fixed values
with arbitrarily small increments of exponential form ½x2 x3 x4 x5 x6 iL5 iC3 iL6
iL4 iC4 iC2 iC5 ekt are; i = 0 (first fixed point), i = 1 (second fixed point), i = 2
(third fixed point), etc., [2–4].
ðiÞ ðiÞ
X2 ðtÞ ¼ X2 þ x2 ekt ; X3 ðtÞ ¼ X3 þ x3 ekt ;
ðiÞ ðiÞ
X4 ðtÞ ¼ X4 þ x4 ekt ; X5 ðtÞ ¼ X5 þ x5 ekt
ðiÞ ðiÞ
X6 ðtÞ ¼ X6 þ x6 ekt ; IL5 ðtÞ ¼ IL5 þ iL5 ekt ;
ðiÞ ðiÞ
IC3 ðtÞ ¼ IC3 þ iC3 ekt ; IL6 ðtÞ ¼ IL6 þ iL6 ekt
ðiÞ ðiÞ ðiÞ

IL4 ðtÞ ¼ IL4 þ iL4 ekt ; IC4 ðtÞ ¼ IC4 þ iC4 ekt ; IC2 ðtÞ ¼ IC2 þ iC2 ekt ; IC5 ðtÞ
ðiÞ
¼ IC5 þ iC5 ekt
dX2 ðtÞ dX3 ðtÞ dX4 ðtÞ dX5 ðtÞ

¼ x2 k ekt ; ¼ x3 k ekt ; ¼ x4 k ekt ; ¼ x5 k ekt
dt dt dt dt
dX6 ðtÞ dIL ðtÞ dIC ðtÞ dIL ðtÞ

¼ x6 k ekt ; 5 ¼ iL5 k ekt ; 3 ¼ iC3 k ekt ; 6
dt dt dt dt
¼ iL6 k ekt
dIL4 ðtÞ dIC ðtÞ

¼ iL4 k ekt ; 4 ¼ iC4 k ekt
dt dt
Our dual band diplexer filter circuit tends to the selected fixed point exponentially
for k < 0, t > 0 otherwise go away from the selected fixed point exponentially. k is
the eigenvalue parameter which establish if the fixed point is stable or unstable,
the selected fixed point [4].
dIL5 ðiÞ ðiÞ
¼ X2 ) iL5 k ekt ¼ X2 þ x2 ekt ; X2 ¼ 0; iL5 k þ x2 ¼ 0
dt
dIC3 ðiÞ ðiÞ
¼ X3 ) iC3 k ekt ¼ X3 þ x3 ekt ; X3 ¼ 0; iC3 k þ x3 ¼ 0
dt
dIL6 ðiÞ ðiÞ
dt
dIL4 ðiÞ ðiÞ
dt
dIC4 ðiÞ ðiÞ
¼ X6 ) iC4 k ekt ¼ X6 þ x6 ekt ; X6 ¼ 0; iC4 k þ x6 ¼ 0
dt
dX2 Rs 1 1
¼ ½X2 þ X4 eksin I C2 IL
dt L5 C2 L5 C1 L5 5
ðiÞ ðiÞ Rs 1 ðiÞ

x2 k ekt ¼ ½X2 þ x2 ekt þ X4 þ x4 ekt eksin ½I þ iC2 ekt
L5 C2 L5 C2
1 ðiÞ
½I þ iL5 ekt
C1 L5 L5
ðiÞ ðiÞ Rs 1 ðiÞ 1 ðiÞ

x2 k ekt ¼ ½X2 þ X4 eksin I I ½x2 ekt þ x4 ekt
L5 C2 L5 C2 C1 L5 L5
Rs 1 1
eksin iC2 ekt iL ekt
L5 C2 L5 C1 L5 5

At fixed points: ½X2 þ X4 eksin LR5s C21L5 IC2 C11L5 IL5 ¼ 0
Rs 1 1
x2 k ekt ¼ ½x2 ekt þ x4 ekt eksin iC2 ekt iL ekt
L5 C2 L5 C1 L5 5
Rs Rs 1 1
x2 k x2 eksin x4 eksin i C2 iL ¼ 0
L5 L5 C2 L5 C1 L5 5
dX3 1 1 Rload1
¼ I C2 IC3 X3 eksout1
dt C2 L1 C3 L1 L1

x3 k ekt ¼ ½IC2 þ iC2 ekt ½I þ iC3 ekt ½X3 þ x3 ekt
C2 L1 C3 L1 C3
Rload1
eksout1
L1
1 ðiÞ 1 ðiÞ ðiÞ Rload1 1

x3 k ekt ¼ I C2 IC3 X3 eksout1 þ iC2 ekt
C2 L1 C3 L1 L1 C2 L1
1 Rload1
iC3 ekt x3 ekt eksout1
C3 L1 L1
ðiÞ ðiÞ ðiÞ

At fixed points: 1
C2 L1 IC2 C31L1 IC3 X3 eksout1 Rload1
L1 ¼ 0
1 1 Rload1
x3 k ekt ¼ iC2 ekt iC3 ekt x3 ekt eksout1
C2 L1 C3 L1 L1
1 1 Rload1
x3 k þ iC2 iC3 x3 eksout1 ¼0
C2 L1 C3 L1 L1
dX5 1 1 ðiÞ
¼ IC5 ) x5 k ekt ¼ ½I þ iC5 ekt ; x5 k ekt
dt C5 L4 C5 L4 C5
1 ðiÞ 1
¼ I C5 þ iC5 ekt
C5 L4 C5 L4
ðiÞ
At fixed points: 1
C5 L4 IC5 ¼ 0; x5 k þ 1
C5 L4 iC5 ¼ 0

¼ ½X2 eksin þ X4 eksin IL I C5
dt L6 dt L6 C6 L6 6 C5 L6
1 dVs ðtÞ ðiÞ ðiÞ Rs

x4 k ekt ¼ ½X2 þ x2 ekt þ X4 þ x4 ekt eksin
L6 dt L6
1 ðiÞ 1 ðiÞ
½I þ iL6 ekt ½I þ iC5 ekt
C6 L6 L6 C5 L6 C5
1 dVs ðtÞ ðiÞ ðiÞ Rs 1 ðiÞ 1 ðiÞ

x4 k ekt ¼ ½X2 þ X4 eksin I I
L6 dt L6 C6 L6 L6 C5 L6 C5
Rs 1 1
½x2 ekt þ x4 ekt eksin iL ekt iC5 ekt
L6 C6 L6 6 C5 L6
ðiÞ ðiÞ ðiÞ

At fixed points: 1
L6 fdVdts ðtÞ ! eg ½X2 þ X4 LR6s eksin C61L6 IL6 C51L6
ðiÞ
I C5 ¼ 0
Rs ksin 1 1
x4 k ekt ¼ ½x2 ekt þ x4 ekt e iL ekt iC5 ekt
L6 C6 L6 6 C5 L6
Rs ksin Rs 1 1
x4 k x2 e x4 eksin iL i C5 ¼ 0
L6 L6 C6 L6 6 C5 L6
dX6 1 1 Rload2
¼ I C5 IC4 X6 eksout2
dt C5 L3 C4 L3 L3

x6 k ekt ¼ ½IC5 þ iC5 ekt ½I þ iC4 ekt ½X6 þ x6 ekt
C5 L3 C4 L3 C4
Rload2
eksout2
L3
1 ðiÞ 1 ðiÞ ðiÞ Rload2 1

x6 k ekt ¼ I I X6 eksout2 þ iC5 ekt
C5 L3 C5 C4 L3 C4 L3 C5 L3
1 Rload2
iC4 ekt x6 ekt eksout2
C4 L3 L3
ðiÞ ðiÞ ðiÞ

At fixed points: 1
C5 L3 IC5 C41L3 IC4 X6 eksout2 Rload2
L3 ¼ 0
1 1 Rload2
x6 k ekt ¼ iC5 ekt iC4 ekt x6 ekt eksout2
C5 L3 C4 L3 L3
1 1 Rload2
C5 L3 C4 L3 L3
We can summery our dual band diplexer filter circuit arbitrarily small increments
equations:
iL5 k þ x2 ¼ 0; iC3 k þ x3 ¼ 0; iL6 k þ x4 ¼ 0; iL4 k þ x5

¼ 0; iC4 k þ x6 ¼ 0
Rs Rs 1 1
x2 k x2 eksin x4 eksin i C2 iL ¼ 0
L5 L5 C2 L5 C1 L5 5
1 1 Rload1
x3 k þ i C2 iC3 x3 eksout1
C2 L1 C3 L1 L1
1
¼ 0; x5 k þ iC5 ¼ 0
C5 L4
Rs ksin Rs 1 1
x4 k x2 e x4 eksin iL i C5 ¼ 0
L6 L6 C6 L6 6 C5 L6
1 1 Rload2
C5 L3 C4 L3 L3
0 1 0 1
iL5 !11 !12
B C B ! !22 C
B iC3 C B 21
C
B C B
B iL C B !31 !32 C C
B 6C B C
B C B ! ! C
0 B
1 B iL4 C BC 41 42 C !
N11 . . . N1 10 Bi C B B ! ! C
C
B .. .. .. C B C5 C þ B 51 52
C
i C2
@ . . . A B C
B x2 C B !61 !62 C
N10 1 N10 10 B C B C i C5
B x C B B ! ! C
C
B 3C B 71 72
C
B C B ! !82 C
B x4 C B 81 C
B C B C
B x C @ !91 ! A
@ 5A 92
x6 ! 10 1 ! 10 2
1
¼ 0; !kl ; !61 ¼ ; !62 ¼ 0
8 k¼1;2;3;4;5;l¼1;2 C2 L5
1 1
!71 ¼ ; !72 ¼ 0; !81 ¼ 0; !82 ¼ ; !91 ¼ 0; !92
C2 L1 C5 L6
1 1
¼ ; !10 1 ¼ 0; !10 2 ¼
C5 L4 C5 L3
Assumption
0 1 0 1
!11 12 i L5
B !21 22 C B C
B C B i C3 C
B ! B C
B 32 C
C B iL C
B
31
C B 6C
B !41 42 C B C
B C 0 1 B i L4 C
B ! N11 ... N1 B C
B 52 C
C i C2 B
10 B C
C B i C5 C
B
51
C ! e; @ .. .. ..
B !61 62 C . . . AB C0
B C i C5 B x2 C
B ! N10 N10 B C
B 71 72 C
C 1 10 B x C
B 3C
B C B C
B !81 82 C B x4 C
B C B C
B ! 92 C B x C
@ 91 A @ 5A
!10 1 10 2 x6
i C2
Arbitrarily small increments iC2 ; iC5 are very small and ð Þ ! e.
i C5
N11 ¼ k; N12 ¼ . . . ¼ N15 ¼ 0; N16 ¼ 1; N17 ¼ . . .N1 10 ¼ 0; N21 ¼ 0; N22

¼ k
N23 ¼ . . .N26 ¼ 0; N27 ¼ 1; N28 ¼ N29 ¼ N2 10 ¼ 0; N31 ¼ N32 ¼ 0; N33 ¼ k
N34 ¼ . . .N37 ¼ 0; N38 ¼ 1; N39 ¼ N3 10 ¼ 0; N41 ¼ N42 ¼ N43 ¼ 0; N44 ¼ k
N45 ¼ . . . ¼ N48 ¼ 0; N49 ¼ 1; N4 10 ¼ 0; N51 ¼ . . . ¼ N54 ¼ 0; N55 ¼ k; N56

¼ . . .N59 ¼ 0
1
N5 10 ¼ 1; N61 ¼ ; N62 ¼ 0; N63 ¼ N64 ¼ N65 ¼ 0; N66
C1 L5
Rs
¼ k eksin
L5
Rs 1
N67 ¼ 0; N68 ¼ eksin ; N69 ¼ 0; N6 10 ¼ 0; N71 ¼ 0; N72 ¼ ; N73
L5 C3 L1
¼0
Rload1
N74 ¼ N75 ¼ 0; N76 ¼ 0; N77 ¼ k eksout1 ; N78 ¼ N79 ¼ N7 10 ¼0
L1
1 Rs
N81 ¼ N82 ¼ 0; N83 ¼ ; N84 ¼ N85 ¼ 0; N86 ¼ eksin ; N87 ¼ 0
C6 L6 L6
Rs ksin
N88 ¼ k e ; N89 ¼ N8 10 ¼ 0; N91 ¼ . . . ¼ N98 ¼ 0; N99 ¼ k; N9 10
L6
¼0
1
N10 1 ¼ N10 2 ¼ N10 3 ¼ N10 4 ¼ 0; N10 5 ¼ ; N10 6 ¼ . . . ¼ N10 9
C4 L3
¼0
Rload2
N10 10 ¼ k eksout2
L3
We analyze dual band diplexer filter circuit for the following cases:
Table 7.2 Dual band diplexer filter circuit for cases of sin, sout1, sout2
Case no. sin sout1 sout2
1 0 0 0
2 s 0 0
3 0 s 0
4 0 0 s
5 s s 0
6 0 s s
7 s 0 s
8 s s s
Case No. 1: sin ¼ 0; sout1 ¼ 0; sout2 ¼ 0; N66 ¼ k LR5s ; N68 ¼ LR5s .
Rload1 Rs Rs Rload2
N77 ¼ k ; N86 ¼ ; N88 ¼ k ; N10 10 ¼ k
L1 L6 L6 L3
X
10 X
10
detðA k IÞj sin ¼0 ¼ !k kk ; detðA k IÞj sin ¼0 ¼ 0 ) !k kk j sin ¼0
sout1 ¼0 sout1 ¼0 sout1 ¼0
sout2 ¼0 k¼0 sout2 ¼0 k¼0 sout2 ¼0
¼0
Remark It is reader exercise to find above polynomial parameters !1 ; . . .; !10 .

Eigenvalues stability discussion: Our dual band diplexer filter circuit involving
N variables (N > 2, N = 10), the characteristic equation is of degree N = 10 and
must often be solved numerically. Expect in some particular cases, such an equation
has (N = 10) distinct roots that can be real or complex. These values are the
eigenvalues of the 10 10 Jacobian matrix (A). The general rule is that the Steady
State (SS) is stable if there is no eigenvalue with positive real part. It is sufficient
that one eigenvalue is positive for the steady state to be unstable. Our 10-variables
(X2 X3 X4 X5 X6 IL5 IC3 IL6 IL4 IC4 ) dual band diplexer filter circuit has ten eigen-
values. The type of behavior can be characterized as a function of the position of
these eigenvalues in the Re/Im plane. Five non-degenerated cases can be distin-
guished: (1) the ten eigenvalues are real and negative (stable steady state), (2) the
ten eigenvalues are real, nine of them are negative (unstable steady state), (3) and
eigenvalues are real and negative (stable steady state), two cases can be distin-
guished depending on the relative value of the real part of the complex eigenvalues
and of the real one, (5) two eigenvalues are complex conjugates with a negative real
part and at least one eigenvalue is positive (unstable steady state).
Remark It is reader exercise to analyze dual band diplexer filter circuit stability for
cases 2 to 8 (at least one delay parameter sin ; sout1 ; sout2 is positive and real number).
The stability analysis is done by using geometric stability switch criteria in delay
differential systems (E. Beretta and Y. Kuang). It is a practical guideline that

combines graphical information with analytical work to effectively study the local
stability of dual band diplexer filter circuit model involving delay dependent
parameters. The stability of dual band diplexer filter circuit steady state is deter-
mined by the graphs of some functions of s which can be expressed explicitly [5, 6].
switching resulting from the increase of value of the time delay s for our dual band
diplexer filter circuit characteristic equation.
Dðk; sin ; sout1 ; sout2 Þ ¼ detðA k IÞ; Dðk; sin ; sout1 ; sout2 Þ ¼ 0; Dðk; sÞ ¼ Pn ðk; sÞ þ Qm ðk; sÞ eks
Xn X
m
k¼0 k¼0
pk ð Þ; qk ð Þ : R þ 0 ! R are continuous and differentiable functions of s.
assume:
(II) lim supfjQm ðk; sÞ=Pn ðk; sÞj : jkj ! 1; Rek
0g\1 for any s.
whenever it exists.
7.3 A Crystal-Lattice BPF Circuit Stability Analysis
We use crystal in place of LC filter for low frequency applications. It is emphasis in

narrow bandwidth filtering. Crystals have a series and parallel resonant mode. Other
crystal mode is overtone or harmonic mode. There are lattice crystal filter, half
latticex and cascaded half lattice filters. The bandwidth of these filters is a function
of the frequency separation of the crystals. Another form of filter is the ladder
crystal filter. It has an asymmetrical response and it is called the “lower-sideband
ladder“ configuration. Ladder filters have some advantages: no need to pick crystals
for proper frequency separation, no need to match crystal pairs, simple filter
topology, simple construction methods, no adjustable components are required after
alignment is completed, benefits by the absence of coils, compact assembly, high
number of poles, and shape the filter response with great accuracy. The equivalent
circuit of a quartz crystal is describe in the below figure.
We need to measure the crystal parameters. The important parameters in our design
are: DF—frequency offset or deviation from the specific center frequency, r—series
resistance of the crystal, fL and fH—3 dB points required for the Q calculation, Lm—
motional inductance which is derived from the Q and r, Cp—parallel capacitance of the
crystal’s holder, Cm—motional capacitance. In our design the crystals can be matched
for Q, Lm and DF. Several factors influence the choice of bandwidth of a crystal filter:
the desired selectivity (narrow filters for contest work and wider filters for casual
rag-chewing), receiver sensitivity, dynamic range, and personal preference. The value
of the terminating resistance should be as low as possible to minimize the transfor-
mation ratio of the impedance matching transformers. The crystal can be forced to
resonate efficiency at odd harmonic intervals of its fundamental frequency. The
overtone modes force a crystal filter to have undesired reentrance modes (odd multiples
of the series resonant frequency). Crystals can be in larger combinations, within RF
filter packages. The crystal lattice filters contain several crystals within a single circuit
(adopted for use as a very sharp bandpass filter). The input and outputs employ RF
transformers (T1 and T2) with shunt capacitors (C1 and C2). Each set of crystals
XTAL1 plus XTAL2 and XTAL3 plus XTAL4 are cut to different frequencies. The
matched set of XTAL1 and XTAL2 having a lower resonant frequency than the other
Fig. 7.6 Equivalent circuit of a quartz crystal

7.3 A Crystal-Lattice BPF Circuit Stability Analysis 713
Fig. 7.7 Lattice crystal filter
matched set of XTAL3 and XTAL4. This structure circuit attains the desired band-
width and selectivity [107, 108].
Terminology: ip1 -input current to transformer T1’s primary coil, is1 -input current
to transformer T1’s secondary coil, ip2 -input current to transformer T2’s primary
coil, is2 -input current to transformer T2’s secondary coil
Transformer T1: vp1—transformer primary voltage, vs1—transformer secondary
voltage. /1 is the flux through a one turn coil located anywhere on the transformer
core. np1 and ns1 are the number of turns of the coil 1 (primary) and 2 (secondary),
respectively, then the total flux /p1 and /s1 through coils 1 and 2 respectively are
d/p1 d/s1
/p1 ¼ np1 /1 ; /s1 ¼ ns1 /; vp1 ¼ dt ; vs1 ¼ dt ; vp1 ¼ np1 dt
d/1
v ðtÞ n n v ðtÞ n p1
vs1 ¼ ns1 d/
dt ; vp1 ðtÞ ¼ vp1 ; vs1 ðtÞ ¼
1
vs1 ; vps1 ðtÞ ¼ nps1 ; a1 ¼ nps1 . vps1 ðtÞ ¼ n s1 for all
1 1 1 1
times t and for all voltages vp1 and vs1 . We define the ratio between np1 and ns1 as
n
“a1” (a1 ¼ nps1 ). Step down transformer a1 > 1 and step up transformer a1 < 1. By
1
law of conservation of energy, apparent real and reactive powers are each conserved
in the input and output (S1 ¼ Ip1 vp1 ¼ Is1 vs1 ). Ideal transformer identity:
qffiffiffiffiffi
v p1 I s1 np1 L p1
vs ¼ Ip ¼ ns ¼ Ls ¼ a 1 :
1 1 1 1
vp1 n1 n2 n2
np1 ¼ n1 ; ns1 ¼ n2 ; ¼ ; n1 ; n2 [ 1 ) vs1 ¼ vp1 ¼ VA1
vs1 n2 n1 n1
n2
¼ ½Vs ðtÞ Rs IRs
n1
vs1 ¼ VA1 nn21 ; vp1 ¼ Vs ðtÞ Rs IRs ; vs1 ¼ ½Vs ðtÞ Rs IRs nn21 (Reference direc-
tions are in the top of each transformer’s coil). np1 ip1 þ ns1 is1 ¼ 0; ip1 ¼ ip1 ðtÞ;
i ðtÞ n
is1 ¼ is1 ðtÞ; ips1 ðtÞ ¼ nps1
1 1
For all t and all currents ip1 and is1 . The voltage vp1 across coil 1 does not depend
on ip1 or on is1 ; it depends only on vs1 . Similarly the current ip1 depends only on is1
and is independent of vp1 ; vs1 . vs1 ¼ VA2 VA9
Transformer T2: vp2—transformer primary voltage, vs2—transformer secondary
voltage. /2 is the flux through a one turn coil located anywhere on the transformer
core. np2 and ns2 are the number of turns of the coil 1 (primary) and 2 (secondary),
respectively, then the total flux /p2 and /s2 through coils 1 and 2 respectively are
d/p2 d/ d/
/p2 ¼ np2 /2 ; /s2 ¼ ns2 /; vp2 ¼ ; vs2 ¼ s2 ; vp2 ¼ np2 2
dt dt dt
d/2 vp ðtÞ np np
vs2 ¼ ns2 ; vp2 ðtÞ ¼ vp2 ; vs2 ðtÞ ¼ vs2 ; 2 ¼ 2 ; a2 ¼ 2 :
dt vs2 ðtÞ ns2 ns 2
vp2 ðtÞ np2

vs2 ðtÞ ¼ ns2 for all times t and for all voltages vp2 and vs2 . We define the ratio
n p2
between np2 and ns2 as “a2” (a2 ¼ ns2 ). Step down transformer a2 > 1 and step up
transformer a2 < 1. By law of conservation of energy, apparent real and reactive
powers are each conserved in the input and output (S2 ¼ Ip2 vp2 ¼ Is2 vs2 ). Ideal
qffiffiffiffiffi
v I n L
transformer identity: vps2 ¼ Ips2 ¼ nps2 ¼ Lps2 ¼ a2 .
2 2 2 2
vs2 ¼ VA14 ¼ IRload Rload

v p 2 n3 n4 n4
np 2 ¼ n3 ; ns 2 ¼ n4 ; ¼ ; n3 ; n4 [ 1 ) vs2 ¼ vp2 ¼ ½VA6 VA3
v s 2 n4 n3 n3
vs2 ¼ ½VA6 VA3 nn43 ; vp2 ¼ VA6 VA3 ; vs2 ¼ vp2 nn43 (Reference directions are in
the top of each transformer’s coil). np2 ip2 þ ns2 is2 ¼ 0; ip2 ¼ ip2 ðtÞ; is2 ¼ is2 ðtÞ;
ip2 ðtÞ n
is2 ðtÞ ¼ nps2 .
2
For all t and all currents ip2 and is2 . The voltage vp2 across coil 1 does not depend
on ip2 or on is2 ; it depends only on vs2 . Similarly the current ip2 depends only on is2
and is independent of vp2 ; vs2 . We consider for two transformers a1 6¼ a2 [24].
Fig. 7.8 Lattice crystal filter equivalent circuit
Remark When we have two coils of wire in close physical proximity to one
another, no importance whether or not the coils are wrapped around a common core
of magnetic material. We assume that the coils do not move with respect to one
another or with respect to a core they might be wrapped around. If we have some
ferromagnetic material in the magnetic circuit of the two coils, then when the
current sufficiently large, the relation between the fluxes /p ; /s and the currents
ip ; is are no longer linear. In this case the equations have the following form:
/p ¼ fp ðip ; is Þ; /s ¼ fs ðip ; is Þ, where fp ðip ; is Þ; fs ðip ; is Þ are nonlinear functions of the
currents ip ; is , By Faraday’s law we get the following:
d/p @fp dip @fp dis

vp ¼ ¼ þ ; fp ¼ fp ðip ; is Þ; fs ¼ fs ðip ; is Þ
dt @ip dt @is dt
d/s @fs dip @fs dis

vs ¼ ¼ þ ; fp ¼ fp ðip ; is Þ; fs ¼ fs ðip ; is Þ
dt @ip dt @is dt
@f @f@fs @fs
The four partial derivatives (@ipp ; @ips ; @i ; ) are function of ip ; is . It is applicable
p @is
for the first transformer (T1) and the second transformer (T2) in our Lattice crystal
filter circuit.
Ir1 ¼ ICm1 ¼ ILm1 ; Ir2 ¼ ICm2 ¼ ILm2 ; Ir3 ¼ ICm3 ¼ ILm3 ;

Ir4 ¼ ICm4 ¼ ILm4 ; Irk ¼ ICmk ¼ ILmk 8 k ¼ 1; 2; 3; 4
dðVA9 VA2 Þ dðVA3 VA6 Þ
IRload ¼ is2 ; IRs ¼ ip1 ; IC1 ¼ C1 ; IC2 ¼ C2 ;
dt dt
dðVA3 VA2 Þ VA VA2 dðVA5 VA4 Þ
ICp3 ¼ Cp3 Ir3 ¼ 4 ; ICm3 ¼ Cm3 ;
dt r3 dt
dILm3 dðVA6 VA2 Þ
VA3 VA5 ¼ Lm3 ; ICp1 ¼ Cp1
dt dt
VA7 VA2 dðVA8 VA7 Þ dILm1

Ir1 ¼ ; ICm1 ¼ Cm1 ; VA6 VA8 ¼ Lm1 ; I Cp2
r1 dt dt
dðVA9 VA3 Þ
¼ Cp2
dt
VA9 VA10 dðVA10 VA11 Þ dILm2

Ir2 ¼ ; ICm2 ¼ Cm2 ; VA11 VA3 ¼ Lm2 ; I Cp4
r2 dt dt
dðVA9 VA6 Þ
¼ Cp4
dt
VA9 VA13 dðVA13 VA12 Þ dIL
Ir 4 ¼ ; ICm 4 ¼ Cm4 ; VA12 VA6 ¼ Lm4 m 4
r4 dt dt
vs 2 ¼ VA14 ¼ IRload Rload ; vs1 ¼ VA2 VA9 ; vp2 ¼ VA6 VA3 ; vp1 ¼ Vs ðtÞ IRs Rs
KCL @ A2: ICp3 þ Ir3 þ ICp1 þ Ir1 þ IC1 ¼ is1 ; KCL @ A9:
is1 ¼ IC1 þ ICp2 þ Ir2 þ ICp4 þ Ir4
KCL @ A 3: ICp3 þ ILm3 þ IC2 ¼ ip2 þ ICp2 þ ILm2 KCL @ A6:
ICp4 þ ILm4 þ IC2 ¼ ip2 þ ICp1 þ ILm1
Z
dðVA3 VA6 Þ 1 dðVA9 VA2 Þ
IC2 ¼ C2 ) VA3 VA6 ¼ IC2 dt; IC1 ¼ C1
dt Z C2 dt
1
) VA9 VA2 ¼ IC1 dt
C1
Z
dðVA3 VA2 Þ 1 VA4 VA2
ICp3 ¼ Cp3 ) VA3 VA2 ¼ ICp3 dt; Ir3 ¼
dt Cp3 r3
) VA4 VA2 ¼ Ir3 r3
Z
dðVA5 VA4 Þ 1
ICm3 ¼ Cm3 ) VA5 VA4 ¼ ICm3 dt; VA3 VA5
dt Cm3
dILm3
¼ Lm3
dt
Z
dðVA6 VA2 Þ 1 VA7 VA2
I Cp1 ¼ Cp1 ) VA6 VA2 ¼ ICp1 dt; Ir1 ¼
dt Cp1 r1
) VA7 VA2 ¼ Ir1 r1
Z
dðVA8 VA7 Þ 1
ICm1 ¼ Cm1 ) VA8 VA7 ¼ ICm1 dt; VA6 VA8
dt Cm1
dILm1
¼ Lm1
dt
Z
dðVA9 VA3 Þ 1
ICp2 ¼ Cp2 ) VA9 VA3 ¼ ICp2 dt;
dt Cp2
VA VA10
Ir2 ¼ 9 ) VA9 VA10 ¼ Ir2 r2
r2
Z
dðVA10 VA11 Þ 1
ICm2 ¼ Cm2 ) VA10 VA11 ¼ ICm2 dt;
dt Cm2
dILm2
VA11 VA3 ¼ Lm2
dt
Z
dðVA9 VA6 Þ 1 VA VA13
ICp4 ¼ Cp4 ) VA9 VA6 ¼ ICp4 dt; Ir4 ¼ 9 ) VA9 VA13 ¼ Ir4 r4
dt Cp4 r4
Z
dðVA13 VA12 Þ 1 dILm4
ICm4 ¼ Cm4 ) VA13 VA12 ¼ ICm4 dt; VA12 VA6 ¼ Lm4
dt Cm 4 dt
Lattice crystal filter differential equations group No. 1:

Z Z
1 1
VA9 VA2 ¼ IC1 dt; VA3 VA6 ¼ IC2 dt; VA3 VA2
C1 Z C2
1
¼ ICp 3 dt
Cp3
Z Z
1 1
VA5 VA4 ¼ ICm 3 dt; VA6 VA2 ¼ ICp 1 dt; VA8 VA7
Cm3 Z Cp1
1
¼ ICm 1 dt
Cm1
Z Z
1 1
VA9 VA3 ¼ IC dt; VA10 VA11 ¼ ICm 2 dt; VA9 VA6
Cp2 Z p 2 Cm2
1
¼ ICp 4 dt
Cp4
Z
1
VA13 VA12 ¼ ICm 4 dt; VA4 VA2 ¼ Ir3 r3 ; VA3 VA5
Cm4
dIL
¼ Lm3 m 3 ; VA7 VA2 ¼ Ir1 r1
dt
dILm 1 dIL
VA6 VA8 ¼ Lm1 ; VA9 VA10 ¼ Ir2 r2 ; VA11 VA3 ¼ Lm2 m 2 ; VA9 VA13
dt dt
¼ Ir 4 r4
dILm 4
VA12 VA6 ¼ Lm4
dt
&&&
Z Z
1 1
C1 C1
Z Z
1 1
VA3 VA2 ¼ ICp 3 dt ) VA3 ¼ VA2 þ ICp 3 dt
Cp3 Cp3
Z Z Z
1 1 1
VA3 ¼ VA9 IC1 dt þ ICp 3 dt; VA6 VA2 ¼ ICp 1 dt
C1 Cp 3 Cp1
Z Z Z
1 1 1
VA6 ¼ VA2 þ ICp 1 dt ¼ VA9 IC1 dt þ ICp 1 dt
Cp 1 C1 Cp 1
Z Z Z
1 1 1
VA3 VA6 ¼ IC2 dt ) ICp 3 dt ICp 1 dt
C2 Z Cp 3 Cp1
1
¼ IC2 dt
C2
Z Z Z
d 1 1 1 1 1
f ICp 3 dt ICp 1 dt ¼ IC2 dtg ) ICp 3 I Cp 1
dt Cp3 Cp 1 C2 Cp3 Cp 1
1
¼ I C2
C2
Z Z
1 1
VA9 VA3 ¼ ICp2 dt ) VA9 ¼ VA3 þ ICp2 dt
Cp2 Cp2
Z Z Z
1 1 1
VA9 ¼ VA9 IC1 dt þ ICp3 dt þ ICp2 dt
C1 Cp3 Cp2
Z Z Z
1 1 1
IC1 dt þ ICp3 dt þ ICp2 dt ¼ 0
C1 Cp3 Cp2
Z Z Z
d 1 1 1 1
f IC1 dt þ ICp3 dt þ ICp2 dt ¼ 0g; IC1
dt C1 Cp3 Cp2 C1
1 1
¼ I Cp3 þ I Cp2
Cp3 Cp2
Z Z Z
1 1 1
VA9 VA6 ¼ IC dt ) IC1 dt ICp1 dt
Cp4 Z p4 C1 Cp1
1
¼ ICp4 dt
Cp4
Z Z Z
d 1 1 1 1 1
f IC1 dt ICp1 dt ¼ ICp4 dtg ) IC1 I Cp1
dt C1 Cp1 Cp4 C1 Cp1
1
¼ I Cp4
Cp4
Z
1
VA4 VA2 ¼ Ir3 r3 ) VA4 ¼ VA2 þ Ir3 r3 ¼ VA9 IC1 dt þ Ir3 r3
C1
dILm3 dILm3
VA3 VA5 ¼ Lm3 ) VA5 ¼ VA3 Lm3
dt dt
Z Z
1 1 dILm3
¼ VA9 IC1 dt þ ICp3 dt Lm3
C1 Cp3 dt
Z
1
VA7 VA2 ¼ Ir1 r1 ) VA7 ¼ VA2 þ Ir1 r1 ¼ VA9 IC1 dt þ Ir1 r1
C1
dILm1 dILm1
VA6 VA8 ¼ Lm1 ) VA8 ¼ VA6 Lm1
dt dt
Z Z
1 1 dILm1
¼ VA9 IC1 dt þ ICp1 dt Lm1
C1 Cp1 dt
VA9 VA10 ¼ Ir2 r2 ) VA10 ¼ VA9 Ir2 r2 ;
dILm2 dILm2
VA11 VA3 ¼ Lm2 ) VA11 ¼ VA3 þ Lm2
dt dt
Z Z
1 1 dILm2
VA11 ¼ VA9 IC1 dt þ ICp3 dt þ Lm2 ;
C1 Cp3 dt
VA9 VA13 ¼ Ir4 r4 ) VA13 ¼ VA9 Ir4 r4
dILm4 dILm4
VA12 VA6 ¼ Lm4 ) VA12 ¼ VA6 þ Lm4
dt dt
Z Z
1 1 dILm4
¼ VA9 IC1 dt þ ICp1 dt þ Lm4
C1 Cp1 dt
Remark A
Z Z
1 1 dILm3
VA5 VA4 ¼ ICm3 dt; ICp3 dt Lm3 Ir3 r3
Cm3 Z Cp3 dt
1
¼ ICm3 dt
Cm3
Z Z
1 1 dILm3
VA9 IC1 dt þ ICp3 dt Lm3
C1 Z Cp3 dt
1
½VA9 IC1 dt þ Ir3 r3
Z C1
1
¼ ICm3 dt
Cm3
Z Z
d 1 dILm3 1
f ICp3 dt Lm3 Ir 3 r3 ¼ ICm3 dtg
dt Cp3 dt Cm3
1 d 2 ILm3 dIr3 1
) I C p 3 Lm 3 r3 ¼ ICm3
Cp3 dt2 dt Cm3
Remark B
Z Z
1 1 dILm1
VA8 VA7 ¼ ICm1 dt; ICp1 dt Lm1 Ir1 r1
Cm1 Z Cp1 dt
1
¼ ICm1 dt
Cm1
Z Z
1 1 dILm1
VA9 IC1 dt þ ICp1 dt Lm1
C1 Z Cp1 dt
1
½VA9 IC1 dt þ Ir1 r1
Z C1
1
¼ ICm1 dt
Cm1
Z Z
d 1 dILm1 1
f ICp1 dt Lm1 Ir 1 r1 ¼ ICm1 dtg
dt Cp1 dt Cm1
1 d 2 ILm1 dIr1 1
) I C p 1 Lm 1 r1 ¼ ICm1
Cp1 dt2 dt Cm1
Remark C
Z
1 dIr 1 1 d 2 ILm2
VA10 VA11 ¼ ICm2 dt; 2 r2 þ I C1 I C p 3 Lm 2
Cm2 dt C1 Cp3 dt2
1
¼ ICm2
Cm2
Z Z
1 1 dILm2
VA9 Ir2 r2 ½VA9 IC1 dt þ ICp3 dt þ Lm2
Z C1 Cp3 dt
1
¼ ICm2 dt
Cm2
Z Z Z
1 1 dILm2 1
Ir2 r2 þ I C1 dt ICp3 dt Lm2 ¼ ICm2 dt
C1 Cp3 dt Cm2
Z Z Z
d 1 1 dILm2 1
fIr2 r2 þ IC1 dt ICp3 dt Lm2 ¼ ICm2 dtg
dt C1 Cp3 dt Cm2
dIr2 1 1 d 2 ILm2 1
r2 þ I C1 ICp3 Lm2 ¼ ICm2
dt C1 Cp3 dt2 Cm2
Remark D
Z
1 dIr 1 1 d 2 I Lm 4
VA13 VA12 ¼ ICm4 dt; 4 r4 þ I C1 I C p 1 Lm 4
Cm4 dt C1 Cp1 dt2
1
¼ I Cm 4
Cm4
Z Z
1 1 dILm4
VA9 Ir4 r4 ½VA9 IC1 dt þ ICp1 dt þ Lm4
Z C1 Cp1 dt
1
¼ ICm4 dt
Cm4
Z Z Z
1 1 dILm4 1
Ir4 r4 þ I C1 dt ICp1 dt Lm4 ¼ ICm4 dt
C1 Cp1 dt Cm4
Z Z Z
d 1 1 dILm4 1
fIr4 r4 þ I C1 dt ICp1 dt Lm4 ¼ ICm4 dtg
dt C1 Cp1 dt Cm4
dIr4 1 1 d 2 ILm4 1
dt C1 Cp1 dt2 Cm4
1 1 1 1
ICp3 I Cp1 ¼ IC2 ; IC1
Cp3 Cp1 C2 C1
1 1 1 1 1
¼ I Cp3 þ ICp2 ; IC1 I Cp1 ¼ I Cp4
Cp3 Cp2 C1 Cp1 Cp4
1 d 2 ILm3 dIr3 1 1 d 2 ILm1 dIr1

ICp3 Lm3 r3 ¼ IC m
; IC p
L m 1
r1
Cp3 dt2 dt Cm3 3
Cp1 1
dt2 dt
1
¼ ICm1
Cm1
dIr2 1 1 d 2 ILm2 1
dt C1 Cp3 dt2 Cm2
dIr4 1 1 d 2 ILm4 1
dt C1 Cp1 dt2 Cm4
Lattice crystal filters Variables transformation:
Ir1 ¼ ICm1 ¼ ILm1 ; Ir2 ¼ ICm2 ¼ ILm2 ; Ir3 ¼ ICm3 ¼ ILm3 ;

Ir4 ¼ ICm4 ¼ ILm4 ; Irk ¼ ICmk ¼ ILmk 8 k ¼ 1; 2; 3; 4
Ir1 ! ILm1 ; ICm1 ! ILm1 ; Ir2 ! ILm2 ; ICm2 ! ILm2 ; Ir3 ! ILm3 ;
ICm3 ! ILm3 ; Ir4 ! ILm4 ; ICm4 ! ILm4
1 1 1 1
I Cp3 I Cp1 ¼ I C2 ; I C1
Cp3 Cp1 C2 C1
1 1 1 1 1
¼ I Cp3 þ I Cp2 ; I C1 I Cp1 ¼ I Cp4
Cp3 Cp2 C1 Cp1 Cp4
1 d 2 ILm3 dILm3 1 1 d 2 ILm1 dILm1

ICp3 Lm3 r 3 ¼ I Lm3 ; I Cp1 L m1
Cp3 dt2 dt Cm3 Cp1 dt2 dt
r1
1
¼ ILm1
Cm1
dILm2 1 1 d 2 ILm2 1
r2 þ I C1 I C p 3 Lm 2 2
¼ ILm2
dt C1 Cp3 dt Cm2
r4 þ I C1 I C p 1 Lm 4 ¼ ILm4
dt C1 Cp1 dt2 Cm4
KCL @ A2: ICp3 þ ILm3 þ ICp1 þ ILm1 þ IC1 ¼ is1 ; KCL @ A9:
is1 ¼ IC1 þ ICp2 þ ILm2 þ ICp4 þ ILm4
Lattice crystal filter circuit transformer T1’s secondary coil current differential
equation:
ip1 ðtÞ ns n2 n1
¼ 1 ¼ ) is1 ðtÞ ¼ ip1 ðtÞ ; is1 ¼ is1 ðtÞ; ip1 ¼ ip1 ðtÞ; is1
is1 ðtÞ np 1 n1 n2
n1
¼ ip1
n2
d
vs1 ¼ VC1 ¼ ðVA9 VA2 Þ ) VA9 VA2 ¼ vs1 ; IC1 ¼ C1 ðVA9 VA2 Þ; IC1
dt
dvs
¼ C1 1
dt
n2 n2 dip dis n2
ip1 ¼ IRs ; vs1 ¼ ½Vs ðtÞ Rs ip1 ; ip1 ¼ is1 ) 1 ¼ 1
n1 n1 dt dt n1
d½Vs ðtÞ Rs ip1 nn21

n2 dVs ðtÞ dip
IC1 ¼ C1 ¼ C1 ½ Rs 1 ;
dt n1 dt dt
n2 dVs ðtÞ dis1 n2
IC1 ¼ C1 ½ Rs ð Þ
n1 dt dt n1
n2 dVs ðtÞ n2 dis
IC1 ¼ C1 ½ þ Rs 1 ;
n1 dt n1 dt
n2 dis 1 n1 dVs ðtÞ
Rs 1 ¼ IC1
n1 dt C 1 n2 dt
dis1 1 n2 n1 1 dVs ðtÞ

¼ IC1 12
dt C1 Rs n2 n2 Rs dt
Lattice crystal filter circuit transformer T2’s primary coil current differential
equation:
ip2 ðtÞ ns n4 ip
IRload ¼ is2 ; vp2 ¼ VC2 ; ¼ 2 ¼ ; ip2 ¼ ip2 ðtÞ; is2 ¼ is2 ðtÞ; 2
is2 ðtÞ np 2 n3 is2
n4
¼
n3
n4 n4 n3 n3
vs2 ¼ vp2 ¼ ½VA6 VA3 ; VA6 VA3 ¼ vs2 ; VA3 VA6 ¼ vs2 ; vs2
n3 n3 n4 n4
¼ is2 Rload
d d n3
IC2 ¼ C2 ½VA VA6 ; vs2 ¼ VRload ; IC2 ¼ C2 ½vs2 ; IC2
dt 3 dt n4
n3 dvs2
¼ C2
n4 dt
ip2 n4 n3 dis dip n3 n3 dis

¼ ) is2 ¼ ip2 ; 2 ¼ 2 ; IC2 ¼ C2 Rload 2
is2 n3 n4 dt dt n4 n4 dt
dis2 n4 dip n3 n4 dip2

¼ I C2 ; 2 ¼ I C2 ;
dt n3 C2 Rload dt n4 n3 C2 Rload dt
n24 1
¼ 2 IC 2
n3 C2 Rload
&&&
1 1 1 1
I Cp3 I Cp1 ¼ I C2 ; I C1
Cp3 Cp1 C2 C1
1 1 1 1 1
¼ I Cp3 þ I Cp2 ; I C1 I Cp1 ¼ I Cp4
Cp3 Cp2 C1 Cp1 Cp4
KCL @ A2: ICp3 þ ILm3 þ ICp1 þ ILm1 þ IC1 ¼ is1 ; KCL @ A9:
is1 ¼ IC1 þ ICp2 þ ILm2 þ ICp4 þ ILm4
Find circuit variables ICp1 ; ICp2 ; ICp3 ; ICp4 as a function of circuit variables
ILm1 ; ILm2 ; ILm3 ; ILm4 and circuit parameters Version No. 1:
1 1 1 1
I Cp1 I Cp2 I Cp3 þ ICp4 ¼ 0;
Cp1 Cp2 Cp3 Cp4
ICp3 þ ICp1 ICp2 ICp4 ¼ ILm1 þ ILm2 ILm3 þ ILm4
C2 C2 C2
I C2 ¼ I Cp3 I Cp1 ; I Cp1
Cp3 Cp1 Cp1
C2
þ I Cp2 ½ þ 1 ICp3 ¼ ip2 ILm2 þ ILm3
Cp3
C2 C2 C2 C2
I C2 ¼ I Cp3 I Cp1 ; ½ þ 1 ICp1 ICp3 ICp4 ¼ ip2 ILm1 þ ILm4
Cp3 Cp1 Cp1 Cp3
X
4
ILmk ð1Þk ¼ ILm1 þ ILm2 ILm3 þ ILm4 ; ip2 ILm2 þ ILm3
k¼1
X
3
¼ ip2 þ ILmk ð1Þk þ 1
k¼2
X
4
ip2 ILm1 þ ILm4 ¼ ip2 þ ILmk ð1Þk
k¼1
k 6¼ 2
k 6¼ 3
1 1 1 1
I Cp1 I Cp2 I Cp3 þ ICp4 ¼ 0;
Cp1 Cp2 Cp3 Cp4
X4
I Cp3 þ I Cp1 I Cp2 I Cp4 ¼ ILmk ð1Þk
k¼1
C2 C2 X3
ICp1 þ ICp2 ½ þ 1 ICp3 ¼ ip2 þ ILmk ð1Þk þ 1
Cp1 Cp3 k¼2
C2 C2 X4
½ þ 1 ICp1 ICp3 ICp4 ¼ ip2 þ ILmk ð1Þk
Cp1 Cp3 k¼1
k¼
6 2
k¼6 3
0 1 0 1
1
C p1 ! Cp1 !
Cp1 3 1
B 1
C B 1 C C p4
@ A þ I Cp2 þ I Cp3 @ ½CCp23 þ 1 A þ ICp4
C2 2
I Cp1 C p1
1
1
1
0
C2 0 C 1
þ1 2
C p1 Cp3
0 0 1
P
4
B C
ILm ð1Þk
B C
k
k¼1
B P 3 C
B ip2 þ ILmk ð1Þk þ 1 C
¼B C
B k¼2
C
B ip þ P 4
k C
@ 2 ILmk ð1Þ
A
k¼1
k6¼2
k6¼3
1 1 1 1
I Cp1 ICp2 I Cp3 þ ICp4 ¼ 0 ) ICp1
Cp1 Cp2 Cp3 Cp4
Cp Cp Cp
¼ 1 ICp2 þ 1 ICp3 1 ICp4
Cp2 Cp3 Cp4
Cp1 Cp Cp X4
½ 1 ICp2 þ ½ 1 þ 1 ICp3 ½ 1 þ 1 ICp4 ¼ ILmk ð1Þk
Cp2 Cp3 Cp4 k¼1
C2 C2 X3
½ þ 1 ICp2 ICp3 ICp4 ¼ ip2 þ ILmk ð1Þk þ 1
Cp2 Cp4 k¼2
C2 Cp Cp C2 Cp
½ þ 1 1 ICp2 þ 1 ICp3 ð½ þ 1 1 þ 1Þ ICp4
Cp1 Cp2 Cp3 Cp1 Cp4
X 4
¼ ip2 þ ILmk ð1Þk
k¼1
k 6¼ 2
k 6¼ 3
We can summery our circuit variables ICp1 ; ICp2 ; ICp3 ; ICp4 as a function of circuit
variables ILm1 ; ILm2 ; ILm3 ; ILm4 and circuit parameters Version No. 4:
Cp1 Cp Cp
I Cp1 ¼ I Cp2 þ 1 I Cp3 1 I Cp4
Cp2 Cp3 Cp4
0 1 0 1
Cp
½Cp1 1 Cp ! Cp
½Cp1 þ 1
½Cp1 þ 1
B 2
C B 4
C
A I Cp2 þ I Cp3 þ @ A I Cp4
C 3 C
@ ½ 2 þ 1
Cp2 1
C p1
2
C p4
C Cp C Cp
½ 2 þ 1 1 C p3 ð½ 2 þ 1 1 þ 1Þ
0 1
Cp1 Cp2 Cp1 Cp4
P
4
ILmk ð1Þk
B k¼1 C
B C
B ip þ P3
C
B 2 ILm ð1Þ k þ 1
C
¼B k¼2
k
C
B P 4 C
B ip2 þ ILmk ð1Þk C
@ k¼1
A
k6¼2
k6¼3
We use Cramer’s rule for the solution of above linear equations. Each variable
(Cpk ; k ¼ 2; 3; 4) given by a quotient two determinants.
0 C C C 1
½Cpp1 1 ½Cpp1 þ 1 ½Cpp1 þ 1
B 2 3 4
C
D ¼ detB
@ ½CCp2 þ 1
2
1 CCp2
4
C
A
C Cp 1 C
½CCp2 þ 1 Cpp1 Cp 3 ð½CCp2 þ 1 Cpp1 þ 1Þ
1 2 1 4
0 1
Cp1 1 CCp2
D¼½ @
1 det Cp
4
A ½Cp1 þ 1
Cp1
Cp2 1
C p3 ð½Cp þ 1 Cp þ 1Þ
C 2 Cp3
0 C 1 4
1
½Cp2 þ 1 CCp2
@
det C
2 4
A ½Cp1 þ 1
Cp1 Cp1
½Cp þ 1 Cp ð½Cp þ 1 Cp þ 1Þ
2 C2 Cp4
0 1C 2
11 4
½Cp2 þ 1 1
@
det C
2
A; D ¼ DðCp1 ; . . .; Cp4 ; C2 Þ
C C
½Cp2 þ 1 Cpp1 Cpp1
1 2 3
0 1
Cp1
P
4 Cp ½ þ 1
B ILmk ð1Þk ½ 1 þ 1 Cp4 C
B Cp3 C
B k¼1 C
B C
B ip þ P IL ð1Þk þ 1
3
1 Cp2
C
C
DICp2 B
¼ detB 4 C
C
2 mk
k¼2
B C2 Cp C
B þ 1Þ C
B i þ P I ð1Þk þ 1 1
4 Cp1 ð½ C
@ p2 Lmk
Cp3
Cp1 Cp4 A
k¼1
k6¼2
k6¼3
!
X
4 C2 Cp
ð½ þ 1 1 þ 1Þ
C
k Cp2
DICp2 ¼ ð ILmk ð1Þ Þ det 1 C p1
4
Cp1 Cp4
k¼1 C p3
0 1
P3 C2
B i p2
þ ILmk ð1Þk þ 1 C
B k¼2 Cp4 C
Cp1 B C
½ þ 1 detB P
Cp1 C
Cp3 B 4 C2
ð½Cp þ 1 þ 1 ÞC
@ ip2 þ ILmk ð1Þk 1 Cp4 A
k¼1
k6¼2
k6¼3
0 1
P
3
kþ1
B ip2 þ ILmk ð1Þ 1 C
B k¼2 C
C B C
½Cpp1 þ 1 detB C p1 C
B ip þ P IL ð1Þk
4
4 C p3 C
@ 2 mk
k¼1
A
k6¼2
k6¼3
X
4
C2 Cp Cp C2
DICp2 ¼ ð ILmk ð1Þk Þ ½ð½ þ 1 1 þ 1Þ þ 1
k¼1
Cp1 Cp4 Cp3 Cp4
Cp C2 Cp C2 Cp
½ 1 þ 1 ½ip2 ð½ þ 1 1 þ 1Þ ð½ þ 1 1 þ 1Þ
Cp3 Cp1 Cp4 Cp1 Cp4
X3
C2 C2 X 4
Cp
ILmk ð1Þk þ 1 ip2 þ ILmk ð1Þk ½ 1 þ 1
k¼2
Cp4 Cp4 k¼1 Cp4
k6¼2
k6¼3
Cp1 Cp X 3 X4
½ip2 þ 1 ILmk ð1Þk þ 1 ip2 þ ILmk ð1Þk
Cp3 Cp3 k¼2 k¼1
k6¼2
k6¼3
X
4
C2 Cp Cp C2
DICp2 ¼ ð ILmk ð1Þk Þ ½ð½ þ 1 1 þ 1Þ þ 1
k¼1
Cp1 Cp4 Cp3 Cp4
Cp C2 Cp Cp C2 Cp
ip2 ½ 1 þ 1 ð½ þ 1 1 þ 1Þ þ ½ 1 þ 1 ð½ þ 1 1 þ 1Þ
Cp3 Cp1 Cp4 Cp3 Cp1 Cp4
X3
Cp C2 C2 Cp1 X4
ILmk ð1Þk þ 1 þ ip2 ½ 1 þ 1 ½ þ 1 ILmk ð1Þk
k¼2
C p3 Cp4 C p4 C p3 k¼1
k6¼2
k6¼3
Cp1 Cp Cp Cp X3
Cp
þ ip2 ½ þ 1 1 1 ½ 1 þ 1 ILmk ð1Þk þ 1 þ ½ 1 þ 1 ip2
Cp4 Cp3 Cp3 Cp4 k¼2
Cp4
C P
4
½Cpp1 þ 1 ILmk ð1Þk
4 k¼1
k6¼2
k6¼3
X
4
C2 Cp Cp C2 Cp
DICp2 ¼ ð ILmk ð1Þk Þ ½ð½ þ 1 1 þ 1Þ þ 1 þ ½ 1 þ 1
k¼1
Cp1 Cp4 Cp3 Cp4 Cp3
C2 Cp X3
C2 Cp1
ð½ þ 1 1 þ 1Þ ILmk ð1Þk þ 1 ½ þ 1
Cp1 Cp4 k¼2
C p4 Cp3
X
4
Cp1 Cp1 X3
Cp
ILmk ð1Þk ½ þ 1 ILmk ð1Þk þ 1 ½ 1 þ 1
k¼1
Cp3 Cp4 k¼2
Cp4
k6¼2
k6¼3
X
4
Cp C2 Cp C2 Cp
ILmk ð1Þk þ ip2 f½ 1 þ 1 ½ 1 þ 1 ð½ þ 1 1 þ 1Þ
k¼1
Cp3 Cp4 Cp3 Cp1 Cp4
k6¼2
k6¼3
C C C
þ ½Cpp1 þ 1 Cpp1 þ ½Cpp1 þ 1g
4 3 4
We can define DICp2 as DICp2 ¼ ip2 C1 þ w1 ðILmk ; k ¼ 1; 2; 3; 4; Cp1 ; Cp1 ; . . .Þ
Cp1 C2 Cp C2 Cp Cp
C1 ¼ ½ þ 1 ½ 1 þ 1 ðð þ 1Þ 1 þ 1Þ þ ½ 1 þ 1
Cp1 Cp1
þ½ þ 1
Cp3 Cp4
w1 ¼ w1 ðILmk ; k ¼ 1; 2; 3; 4; Cp1 ; Cp1 ; . . .Þ

X
4
C2 Cp Cp C2
w1 ðILmk ; k ¼ 1; 2; 3; 4; Cp1 ; Cp1 ; . . .Þ ¼ ð ILmk ð1Þk Þ ½ð½ þ 1 1 þ 1Þ þ 1
k¼1
Cp1 Cp4 Cp3 Cp4
Cp1 C2 Cp X
3
C2 Cp1 X4
þ½ þ 1 ð½ þ 1 1 þ 1Þ ILmk ð1Þk þ 1 ½ þ 1 ILmk ð1Þk
Cp3 Cp1 Cp4 k¼2
Cp4 Cp3 k¼1
k6¼2
k6¼3
Cp1 Cp1 X3
Cp X4
½ þ 1 ILmk ð1Þk þ 1 ½ 1 þ 1 ILmk ð1Þk
Cp3 Cp4 k¼2
C p4 k¼1
k6¼2
k6¼3
0 1
C P
4
k C
B 1 ½Cpp1 ILmk ð1Þ ½Cpp1 þ 1 C
B 2
k¼1 4 C
B P C
B C2 3 C
B ½ þ 1 ip2 þ ILmk ð1Þk þ 1 CCp2 C
DICp3 ¼ detB Cp2 C
B C
4
k¼2
B C P
4 C
B ½ 2 þ 1 Cp1 ip2 þ ILmk ð1Þk ð½CCp2 þ 1 Cpp1
C
þ 1Þ C
@ Cp 1 Cp2
k¼1 1 4
A
k6¼2
k6¼3
0 1
P3
B ip2 þ ILmk ð1Þk þ 1 CCp2 C
B k¼2 4 C
Cp B C
DICp3 ¼ ½ 1 1 detB C
B ip þ P IL ð1Þk
4
þ 1Þ C
Cp2 C
@ 2 mk
ð½CCp2 þ 1 Cpp1 A
k¼1 1 4
k6¼2
k6¼3
0 1
X
4 ½CCp2 þ 1 CCp2
ð ILmk ð1Þk Þ det@ A
2 4
C C
k¼1 ½CCp2 þ 1 Cpp1 ð½CCp2 þ 1 Cpp1 þ 1Þ
0 1 2 1 4
1
P3
kþ1
B ½CCp2
þ 1 ip2 þ ILmk ð1Þ C
B 2
k¼2 C
Cp1 B C
½ þ 1 detB P4 C
Cp4 B ½ C2 þ 1 Cp1 ip2 þ k C
ILmk ð1Þ A
@ Cp1 Cp2
k¼1
k6¼2
k6¼3
Cp1 C2 Cp C2 Cp X3
DICp3 ¼ ½ 1 fip2 ð½ þ 1 1 þ 1Þ ð½ þ 1 1 þ 1Þ ILmk ð1Þk þ 1
Cp2 Cp1 C p4 C p1 Cp4 k¼2
C2 C2 X 4 X4
C2 Cp C2 C2 C2 Cp
ip2 þ IL ð1Þk g ð ILmk ð1Þk Þ fð½ þ 1 1 þ 1Þ ½ þ 1 þ ½ þ 1 1 g
Cp4 Cp4 k¼1 mk k¼1
C p1 Cp4 Cp2 Cp4 Cp1 Cp2
k6¼2
k6¼3
Cp1 C2 C2 X4
C2 Cp C2 Cp X 3
½ þ 1 fip2 ½ þ 1 þ ½ þ 1 ILmk ð1Þk þ ½ þ 1 1 ip2 ½ þ 1 1 IL ð1Þk þ 1 g
Cp4 Cp2 Cp2 k¼1
Cp1 C p2 Cp1 Cp2 k¼2 mk
k6¼2
k6¼3
C p1 C2 Cp Cp C2 Cp
DICp3 ¼ ip2 ½ 1 ð½ þ 1 1 þ 1Þ ½ 1 1 ð½ þ 1 1 þ 1Þ
C p2 Cp1 Cp4 Cp2 Cp1 C p4
X3
Cp C2 Cp C2 X4
ILmk ð1Þk þ 1 ip2 ½ 1 1 þ ½ 1 1 ILmk ð1Þk
k¼2
Cp 2
Cp 4
Cp 2
Cp 4 k¼1
k6¼2
k6¼3
X
4
C2 Cp C2 C2 C2 Cp
ð ILmk ð1Þk Þ fð½ þ 1 1 þ 1Þ ½ þ 1 þ ½ þ 1 1 g
k¼1
Cp1 Cp4 Cp2 Cp4 Cp1 C p2
Cp1 C2 Cp C2 X4
þ i p2 ½ þ 1 ½ þ 1 ½ 1 þ 1 ½ þ 1 ILmk ð1Þk
Cp4 Cp2 C p4 C p2 k¼1
k6¼2
k6¼3
Cp1 C2 Cp Cp C2 Cp X 3
½ þ 1 ½ þ 1 1 ip2 þ ½ 1 þ 1 ½ þ 1 1 IL ð1Þk þ 1
Cp4 Cp1 C p2 Cp4 Cp1 Cp2 k¼2 mk
Cp1 C2 Cp Cp C2 Cp
DICp3 ¼ ip2 f½ 1 ð½ þ 1 1 þ 1Þ ½ 1 1 þ ½ 1 þ 1
C2 Cp1 C2 Cp1 Cp1 C2 Cp
½ þ 1½ þ 1 ½ þ 1 g½ 1 ð½ þ 1 1 þ 1Þ
Cp2 Cp4 Cp1 Cp2 Cp2 Cp1 Cp4
X3
Cp C2 X 4 X4
ILmk ð1Þk þ 1 þ ½ 1 1 ILmk ð1Þk ð ILmk ð1Þk Þ
k¼2
C p2 C p4 k¼1 k¼1
k6¼2
k6¼3
C2 Cp C2 C2 C2 Cp Cp C2
fð½ þ 1 1 þ 1Þ ½ þ 1 þ ½ þ 1 1 g ½ 1 þ 1 ½ þ 1
Cp1 Cp4 Cp2 Cp4 Cp1 Cp2 Cp4 Cp2
X4
Cp C2 Cp X 3
ILmk ð1Þk þ ½ 1 þ 1 ½ þ 1 1 ILmk ð1Þk þ 1
k¼1
C p4
C p1
C p2 k¼2
k6¼2
k6¼3
We can define DICp3 as DICp3 ¼ ip2 C2 þ w2 ðILmk ; k ¼ 1; 2; 3; 4; Cp1 ; Cp1 ; . . .Þ.
Cp1 C2 Cp Cp C2 Cp C2
C2 ¼ ½ 1 ð½ þ 1 1 þ 1Þ ½ 1 1 þ ½ 1 þ 1 ½ þ 1
Cp C2 Cp
½ 1 þ 1 ½ þ 1 1
Cp4 Cp1 Cp2
w2 ¼ w2 ðILmk ; k ¼ 1; 2; 3; 4; Cp1 ; Cp1 ; . . .Þ

Cp1 C2 Cp
w2 ðILm k ; k ¼ 1; 2; 3; 4; Cp1 ; Cp1 ; . . .Þ ¼ ½ 1 ð½ þ 1 1 þ 1Þ
Cp2 Cp1 Cp4
X
3
Cp C2 X 4 X 4
ILmk ð1Þk þ 1 þ ½ 1 1 ILmk ð1Þk ð ILmk ð1Þk Þ
k¼2
Cp2 Cp4 k¼1 k¼1
k6¼2
k6¼3
C2 Cp C2 C2 C2 Cp Cp
fð½ þ 1 1 þ 1Þ ½ þ 1 þ ½ þ 1 1 g ½ 1 þ 1
C2 X4
Cp C2 Cp X 3
½ þ 1 ILmk ð1Þk þ ½ 1 þ 1 ½ þ 1 1 IL ð1Þk þ 1
Cp2 k¼1
Cp4 C p1 Cp2 k¼2 mk
k6¼2
k6¼3
0 1
C C P
4
k
B 1
½Cpp1 ½Cpp1 þ 1 ILmk ð1Þ C
B 2 3
k¼1 C
B P C
B C2 3
kþ1 C
B ½ þ 1 1 ip2 þ ILmk ð1Þ C
DICp4 ¼ detB Cp2 C
B k¼2 C
B C P4 C
B ½ 2 þ 1 Cp1 C p1
ip2 þ ILmk ð1Þk C
@ Cp1 C p2 C p3
k¼1
A
k6¼2
k6¼3
0 1
P
3
kþ1
B 1 ip2 þ ILmk ð1Þ C
B k¼2 C
Cp1 B C
DICp4 ¼½ 1 detB Cp P4 C
Cp2 B 1 ip2 þ ILmk ð1Þk C
@ Cp3 k¼1
A
k6¼2
k6¼3
0 1
P
3
B ½CCp2 þ 1 ip2 þ ILmk ð1Þk þ 1 C
B 2
k¼2 C
Cp1 B C
½ þ 1 detB P
4 C
Cp3 B ½ C2 þ 1 Cp1 ip2 þ k C
ILmk ð1Þ A
@ Cp1 C p2
k¼1
k6¼2
k6¼3
! 0 1
X
4 ½CCp2 þ 1 1
þ ILmk ð1Þk det@ A
2
C Cp1
k¼1 ½CCp2 þ 1 Cpp1 Cp3
1 2
Cp1 X4
Cp Cp X 3
DICp4 ¼ ½ 1 fip2 ILmk ð1Þk þ ip2 1 1 IL ð1Þk þ 1 g
Cp2 k¼1
Cp3 Cp3 k¼2 mk
k6¼2
k6¼3
Cp1 C2 C2 X4
C2 Cp
½ þ 1 fip2 ½ þ 1 þ ½ þ 1 ILmk ð1Þk þ ip2 ½ þ 1 1
Cp3 Cp2 Cp2 k¼1
Cp1 Cp2
k6¼2
k6¼3
C2 Cp X 3 X 4
C2 Cp C2 Cp
½ þ 1 1 IL ð1Þk þ 1 g þ ð ILmk ð1Þk Þ f½ þ 1 1 þ ½ þ 1 1 g
Cp1 Cp2 k¼2 mk k¼1
C p2
C p 3
C p1
Cp 2
Cp Cp X4
Cp Cp Cp
DICp4 ¼ ½ 1 1 ip2 ½ 1 1 ILmk ð1Þk þ ip2 ½ 1 1 1 ½ 1 1
Cp2 Cp2 k¼1
C p2
C p3
Cp2
k6¼2
k6¼3
Cp1 X3
Cp C2 Cp C2
IL ð1Þk þ 1 þ ip2 ½ 1 þ 1 ½ þ 1 ½ 1 þ 1 ½ þ 1
Cp3 k¼2 mk Cp3 Cp2 Cp3 Cp2
X
4
Cp C2 Cp Cp C2 Cp
ILmk ð1Þk ip2 ½ 1 þ 1 ½ þ 1 1 þ ½ 1 þ 1 ½ þ 1 1
k¼1
k6¼2
k6¼3
X
3 X
4
C2 Cp C2 Cp
ILmk ð1Þk þ 1 þ ð ILmk ð1Þk Þ f½ þ 1 1 þ ½ þ 1 1 g
k¼2 k¼1
C p2 Cp3 C p1 C p2
Cp1 Cp Cp Cp C2 Cp
DICp4 ¼ ip2 f½ 1 þ ½ 1 1 1 þ ½ 1 þ 1 ½ þ 1 ½ 1 þ 1
C2 Cp Cp X4
Cp
½ þ 1 1 g ½ 1 1 ILmk ð1Þk ½ 1 1
Cp1 Cp2 Cp2 k¼1
Cp2
k6¼2
k6¼3
Cp1 X3
Cp C2 X4
ILmk ð1Þk þ 1 ½ 1 þ 1 ½ þ 1 ILmk ð1Þk
Cp3 k¼2 Cp3 Cp2 k¼1
k6¼2
k6¼3
Cp1 C2 Cp X 3 X4
þ½ þ 1 ½ þ 1 1 ILmk ð1Þk þ 1 þ ð ILmk ð1Þk Þ
Cp3 Cp1 Cp2 k¼2 k¼1
C2 Cp C2 Cp
f½ þ 1 1 þ ½ þ 1 1 g
Cp2 Cp3 Cp1 Cp2
We can define DICp4 as DICp4 ¼ ip2 C3 þ w3 ðILmk ; k ¼ 1; 2; 3; 4; Cp1 ; Cp1 ; . . .Þ.
Cp Cp Cp Cp C2 Cp C2
C3 ¼ ½ 1 1 þ ½ 1 1 1 þ ½ 1 þ 1 ½ þ 1 ½ 1 þ 1 ½ þ 1
Cp
1
Cp2
w3 ¼ w3 ðILmk ; k ¼ 1; 2; 3; 4; Cp1 ; Cp1 ; . . .Þ
C p1 X4
Cp
w3 ðILmk ; k ¼ 1; 2; 3; 4; Cp1 ; Cp1 ; . . .Þ ¼ ½ 1 ILmk ð1Þk ½ 1 1
C p2 k¼1
Cp2
k6¼2
k6¼3
Cp1 X3
Cp C2 X4
Cp C2
ILmk ð1Þk þ 1 ½ 1 þ 1 ½ þ 1 ILmk ð1Þk þ ½ 1 þ 1 ½ þ 1
Cp3 k¼2 Cp3 Cp2 k¼1
Cp 3 Cp1
k6¼2
k6¼3
Cp1 X3 X 4
C2 Cp C2 Cp
ILmk ð1Þk þ 1 þ ð ILmk ð1Þk Þ f½ þ 1 1 þ ½ þ 1 1 g
Cp2 k¼2 k¼1
Cp2 Cp3 Cp1 Cp2
We can summery our expressions for ILmk ; k ¼ 1; 2; 3; 4:
DICp2 C1 w1 DICp3 C2 w2 DICp4

I Cp2 ¼ ¼ ip2 þ ; I Cp3 ¼ ¼ ip2 þ ; I Cp4 ¼
D D D D D D D
C3 w3
¼ ip2 þ
D D
Cp1 C1 w1 Cp C2 w2 Cp C3 w3
I Cp1 ¼ ½ip2 þ þ 1 ½ip2 þ 1 ½ip2 þ
Cp2 D D Cp3 D D Cp4 D D
Cp1 C1 C2 C3 Cp w w w
ICp1 ¼ ip2 ½ þ þ 1 ½ 1 þ 2 3
D Cp2 Cp3 Cp4 D Cp2 Cp3 Cp4
We define for simplicity new parameter C4 and function w4 ðw1 ; w2 ; w3;...Þ
C1 C2 C3 w w w Cp Cp
C4 ¼ þ ; w ¼ 1 þ 2 3 ; ICp1 ¼ ip2 1 C4 þ 1 w4
Cp2 Cp3 Cp4 4 Cp2 Cp3 Cp4 D D
1 d 2 ILm3 dILm3 1 1 d 2 ILm1 dILm1

ICp3 Lm3 r 3 ¼ I L m3
; I C p1
L m 1

Cp3 dt2 dt Cm3 Cp1 dt2 dt
r1
1
¼ ILm1
Cm1
r2 þ I C1 I C p 3 Lm 2 2
¼ ILm2
dt C1 Cp3 dt Cm2
r4 þ I C1 I C p 1 Lm 4 ¼ ILm4
dt C1 Cp1 dt2 Cm4
dis1 1 n2 n1 1 dVs ðtÞ dip2 n2 1

¼ IC1 12 ; ¼ 42 I C2
dt C1 Rs n2 n2 Rs dt dt n3 C2 Rload
dILm1 dILm2 dILm3

We define for simplicity new variables Y1 ¼ dt ; Y2 ¼ dt ; Y3 ¼ dt ;
dILm4
Y4 ¼ dt
dY1 d 2 ILm1 dY2 d 2 ILm2 dY3 d 2 ILm3 dY4 d 2 ILm4

¼ ; ¼ ; ¼ ; ¼
dt dt2 dt dt2 dt dt2 dt dt2
1 dY3 1 1 dY1
ICp3 Lm3 Y3 r 3 ¼ ILm3 ; ICp1 Lm1 Y 1 r1
Cp3 dt Cm3 Cp1 dt
1
¼ ILm1
Cm1
1 1 dY2 1 dILm1 dILm2

Y2 r2 þ I C1 ICp3 Lm2 ¼ ILm2 ; ¼ Y1 ;
C1 Cp3 dt Cm2 dt dt
dILm3 dILm4
¼ Y2 ; ¼ Y3 ; ¼ Y4
dt dt
1 1 dY4 1
Y4 r4 þ I C1 I C p 1 Lm 4 ¼ ILm4
C1 Cp1 dt Cm4

¼ IC1 12 ; ¼ 42 I C2
dY3 1 r3 1 dY1
¼ I C p 3 Y3 ILm3 ;
dt Cp3 Lm3 Lm3 Cm3 Lm3 dt
1 r1 1
¼ I C p 1 Y1 ILm1
Cp1 Lm1 Lm1 Cm1 Lm1
dY2 r2 1 1 1
¼ Y2 þ I C1 I Cp3 ILm2
dt Lm2 C1 Lm2 Cp3 Lm2 Cm2 Lm2
dY4 r4 1 1 1
¼ Y4 þ I C1 I Cp1 ILm4
dt Lm4 C1 Lm4 Cp1 Lm4 Cm4 Lm4
dILm1 dILm2 dILm3 dILm4

¼ Y1 ; ¼ Y2 ; ¼ Y3 ; ¼ Y4
dt dt dt dt

¼ IC1 12 ; ¼ 42 I C2
dY3 1 C2 w2 r3 1
¼ ½ip2 þ Y3 ILm3
dt Cp3 Lm3 D D Lm3 Cm3 Lm3
dY1 1 Cp Cp r1 1
¼ ½ip2 1 C4 þ 1 w4 Y1 I Lm 1
dY2 r2 1 1 C2 w2 1
¼ Y2 þ I C1 ½ip2 þ ILm2
dt Lm2 C1 Lm2 Cp3 Lm2 D D Cm2 Lm2
dY4 r4 1 1 Cp Cp 1
¼ Y4 þ I C1 ½ip2 1 C4 þ 1 w4
dt Lm4 C1 Lm4 Cp1 Lm4 D D Cm4 Lm4
ILm4

¼ Y1 ; ¼ Y2 ; ¼ Y3 ; ¼ Y4
dt dt dt dt

¼ IC1 12 ; ¼ 42 I C2
dILm1
At fixed points (equilibrium points): dYk
dt ¼ 0 8 k = 1,2,3; 4; dt ¼0
dILm2 dILm3 dILm4 dis dip

¼ 0; ¼ 0; ¼ 0; 1 ¼ 0; 2 ¼ 0; Yk ¼ 0 8 k ¼ 1; 2; 3; 4
dt dt dt dt dt
dVs ðtÞ n2
Assumption dt ! e. IC1 C11Rs n12 ! e ) IC 1 ¼ 0; IC 2 ¼ 0:
2
dY1 1 Cp Cp 1
¼0) ½i 1 C4 þ 1 w4 ðILm ; k ¼ 1; 2; 3; 4; Cp1 ; Cp1 ; . . .Þ I ¼ 0
dt Cp1 Lm1 p2 D D k Cm1 Lm1 Lm1

dY2 1 C2 w2 ðILmk ; k ¼ 1; 2; 3; 4; Cp1 ; Cp1 ; . . .Þ 1
¼0) ½ip2 þ I ¼ 0
dt Cp3 Lm2 D D Cm2 Lm2 Lm2

dY3 1 C2 w2 ðILmk ; k ¼ 1; 2; 3; 4; Cp1 ; Cp1 ; . . .Þ 1
¼0) ½ip2 þ I ¼ 0
dt Cp3 Lm3 D D Cm3 Lm3 Lm3
dY4 1 Cp Cp 1
¼0) ½i 1 C4 þ 1 w4 ðILm ; k ¼ 1; 2; 3; 4; Cp1 ; Cp1 ; . . .Þ I ¼ 0
dt C p 1 L m4 p 2 D D k C m 4 L m 4 L m4
equilibrium points of Lattice crystal filter circuit consists in adding to its coordi-
nated [Y1 Y2 Y3 Y4 ILm1 ILm2 ILm3 ILm4 is1 ip2 ] arbitrarily small increments of expo-
nential terms ½y1 y2 y3 y4 iLm1 iLm2 iLm3 iLm4 i0s1 i0p2 ekt , and retaining the first order
terms in y1 y2 y3 y4 iLm1 iLm2 iLm3 iLm4 i0s1 i0p2 . The system of ten homogeneous equations
leads to a polynomial characteristic equation in the eigenvaluek. The polynomial
characteristic equation accepts by set the Lattice crystal filter circuit equations. The
Lattice crystal filter circuit fixed values with arbitrarily small increments of expo-
nential form ½y1 y2 y3 y4 iLm1 iLm2 iLm3 iLm4 i0s1 i0p2 ekt are; i = 0 (first fixed point),
i = 1 (second fixed point), i = 2 (third fixed point), etc.,
ðiÞ ðiÞ
Y1 ðtÞ ¼ Y1 þ y1 ekt ; Y2 ðtÞ ¼ Y2 þ y2 ekt ;
ðiÞ ðiÞ
Y3 ðtÞ ¼ Y3 þ y3 ekt ; Y4 ðtÞ ¼ Y4 þ y4 ekt
ðiÞ ðiÞ
ILm1 ðtÞ ¼ ILm þ iLm1 ekt ; ILm2 ðtÞ ¼ ILm þ iLm2 ekt ;
1 2
ðiÞ kt ðiÞ
ILm3 ðtÞ ¼ I Lm þ iLm3 e ; ILm4 ðtÞ ¼ I Lm þ iLm4 ekt
3 4
is1 ðtÞ ¼ isðiÞ1 þ i0s1 ekt ; ip2 ðtÞ ¼ ipðiÞ2 þ i0p2 ekt ;
ðiÞ ðiÞ
IC2 ðtÞ ¼ IC2 þ iC2 ekt ; IC1 ðtÞ ¼ IC1 þ iC1 ekt
dY1 ðtÞ dY2 ðtÞ dY3 ðtÞ dY4 ðtÞ

¼ y1 k ekt ; ¼ y2 k ekt ; ¼ y3 k ekt ; ¼ y4 k ekt
dt dt dt dt
dILm1 ðtÞ dILm2 ðtÞ dILm3 ðtÞ dILm4 ðtÞ

¼ iLm1 k ekt ; ¼ iLm2 k ekt ; ¼ iLm3 k ekt ;
dt dt dt dt
¼ iLm4 k ekt
dis1 ðtÞ dip ðtÞ

¼ i0s1 k ekt ; 2 ¼ i0p2 k ekt
dt dt
C2 C2 C2 C2
I C2 ¼ ICp3 ICp1 ¼ ip2 ½ C4
Cp3 Cp1 D Cp3
C2 1
þ ½ w ðIL ; k ¼ 1; 2; 3; 4; Cp1 ; Cp1 ; . . .Þ
D Cp3 2 mk
w4 ðILmk ; k ¼ 1; 2; 3; 4; Cp1 ; Cp1 ; . . .Þ
C1 C1 C1 C2 C1
I C1 ¼ ICp3 þ ICp2 ¼ ip2 ½ þ
Cp3 Cp2 D Cp3 Cp2
C1 1 1
þ ½ w ðIL ; k ¼ 1; 2; 3; 4; Cp1 ; Cp1 ; . . .Þ þ
D Cp3 2 mk Cp2
w1 ðILmk ; k ¼ 1; 2; 3; 4; Cp1 ; Cp1 ; . . .Þ
C1 C1 C1 C3
I C1 ¼ ICp4 þ ICp1 ¼ ip2 ½ þ C4
Cp4 Cp1 D Cp4
C1 1
þ ½ w ðIL ; k ¼ 1; 2; 3; 4; Cp1 ; Cp1 ; . . .Þ
D Cp4 3 mk
þ w4 ðILmk ; k ¼ 1; 2; 3; 4; Cp1 ; Cp1 ; . . .Þ
dVs ðtÞ
Lattice crystal filter differential equations group No. 8: dt !e
dY3 1 C2 w2 r3 1
¼ ½ip2 þ Y3 ILm3
dY1 1 Cp Cp r1 1
¼ ½ip2 1 C4 þ 1 w4 Y1 I Lm 1
dY2 r2 1 C1 C2 C1 C1 1 1
¼ Y2 þ ðip2 ½ þ þ ½ w þ w Þ
dt Lm2 C1 Lm2 D Cp3 Cp2 D Cp3 2 Cp2 1
1 C2 w2 1
½ip2 þ ILm2
Cp3 Lm2 D D Cm2 Lm2
dY4 r4 1 C1 C2 C1 C1 1 1
1 Cp Cp 1
½ip2 1 C4 þ 1 w4 ILm4
Cp1 Lm4 D D Cm4 Lm4

¼ Y1 ; ¼ Y2 ; ¼ Y3 ; ¼ Y4
dt dt dt dt
dis1 C1 C2 C1 C1 1 1 1
¼ fip2 ½ þ þ ½ w þ w g
dt D Cp3 Cp2 D Cp3 2 Cp2 1 C1 Rs
n21 n1 1 dVs ðtÞ

n22 n2 Rs dt
dip2 n2 1 C2 C2 C2 1
¼ 42 fip2 ½ C4 þ ½ w w4 g
dt n3 C2 Rload D Cp3 D Cp3 2
Remark
wk ðILmk ðtÞ ¼ ILm þ iLmk ekt ; k ¼ 1; 2; 3; 4; Cp1 ; Cp1 ; . . .Þ

k
¼ wk ðILm ; k ¼ 1; 2; 3; 4; Cp1 ; Cp1 ; . . .Þ

k
þ wk ðiLmk ekt ; k ¼ 1; 2; 3; 4; Cp1 ; Cp1 ; . . .Þ;

wk ðILmk Þ ¼ wk ðILm Þ þ wk ðiLmk ekt Þ; k ¼ 1; 2; 3; 4
k
wk ðiLmk ekt Þ ¼ ekt wk ðiLmk Þ 8 k ¼ 1; 2; 3; 4
Our Lattice crystal filter circuit tends to the selected fixed point exponentially for
k < 0, t > 0 otherwise go away from the selected fixed point exponentially. k is the
the selected fixed point [2–4].
dY1 1 Cp Cp r1 1
¼ ½ip2 1 C4 þ 1 w4 Y1 I Lm 1
1 Cp Cp
y1 k ekt ¼ ½ðipðiÞ2 þ i0p2 ekt Þ 1 C4 þ 1 ðw4 ðILm Þ þ w4 ðiLmk ekt ÞÞ
Cp1 Lm1 D D k
ðiÞ r1 1 ðiÞ
½Y1 þ y1 ekt ½I þ iLm1 ekt
Lm1 Cm1 Lm1 Lm1
1 Cp Cp Cp
y1 k ekt ¼ ½ipðiÞ2 1 C4 þ i0p2 1 C4 ekt þ 1 w4 ðILm Þ
Cp1 Lm1 D D D k
Cp1 ðiÞ r r 1
w4 ðiLmk ekt Þ Y1 ekt
1 1
þ y1
D Lm1 Lm1 Cm1 Lm1
ðiÞ 1
I Lm iL ekt
1 Cm1 Lm1 m1
1 Cp Cp ðiÞ r1 1 ðiÞ
y1 k ekt ¼ ½iðiÞ 1 C4 þ 1 w4 ðILm Þ Y1 I
Cp1 Lm1 p2 D D k Lm1 Cm1 Lm1 Lm1
1 Cp Cp
þ ½i0p2 1 C4 ekt þ 1 w4 ðiLmk ekt Þ
Cp1 Lm1 D D
r1 1
y1 ekt iL ekt
Lm1 Cm1 Lm1 m1
ðiÞ C p1 C p1 ðiÞ
At fixed points: 1
Cp1 Lm1 ½ip2 D C4 þ D w4 ðILm Þ Y1 Lrm1 Cm 1Lm
k 1 1 1
ðiÞ
I Lm ¼ 0
1
1 Cp Cp r1 1
y1 k þ ½i0 1 C4 þ 1 w4 ðiLmk Þ y1 iL
Cp1 Lm1 p2 D D Lm1 Cm1 Lm1 m1
¼0
dY2 r2 1 C1 C2 C1 C1 1 1
1 C2 w2 1
½ip2 þ I Lm 2
Cp3 Lm2 D D Cm2 Lm2
ðiÞ r2 1 C1 C2 C1
y2 k ekt ¼ ½Y2 þ y2 ekt þ ð½iðiÞ 0 kt
p2 þ ip2 e ½ þ
Lm2 C 1 L m2 D Cp3 Cp2
C1 1 1
þ ½ ½w2 ðILm Þ þ w2 ðiLmk ekt Þ þ fw1 ðILm Þ þ w1 ðiLmk ekt ÞgÞ
D Cp3 k Cp2 k
kt
1 C2 ½w2 ðILmk Þ þ w2 ðiLmk e Þ
½ðiðiÞ þ i 0
ekt
Þ þ
Cp3 Lm2 p2 p2
D D
1 ðiÞ
½I þ iLm2 ekt
Cm2 Lm2 Lm2
ðiÞr2 1 C1 C2 C1 C1 1
y2 k ekt ¼ Y2 þ ðiðiÞ ½ þ þ ½ w ðI Þ
Lm2 C1 Lm2 p2 D Cp3 Cp2 D Cp3 2 Lmk

1 1 C2 w2 ðILmk Þ 1 ðiÞ r2
þ w1 ðILm ÞÞ ½iðiÞ þ I y2 ekt
Cp2 k Cp3 Lm2 p 2
D D Cm2 Lm2 Lm2 Lm2
1 C1 C2 C1 C1 1 1
þ ði0 ekt ½ þ þ ½ w ðiL ekt Þ þ w ðiL ekt ÞÞ
C1 Lm2 p2 D Cp3 Cp2 D Cp3 2 mk Cp2 1 mk
1 C2 w2 ðiLmk ekt Þ 1
½i0p2 ekt þ iL ekt
Cp3 Lm2 D D C m 2 L m 2 m2
At fixed points:
r2
ðiÞ 1 C1 C2 C1 C1 1
Y2 þ ðipðiÞ2 ½ þ þ ½ w ðI Þ
Lm2 C1 Lm2 D Cp3 Cp2 D Cp3 2 Lmk
w2 ðILm Þ
1 1 ðiÞ C2 1 ðiÞ
þ w ðI ÞÞ ½i þ k
I ¼0
Cp2 1 Lmk Cp3 Lm2 p2 D D Cm2 Lm2 Lm2
r2 1 C1 C2 C1 C1
y2 k ekt ¼ y2 ekt þ ði0 ekt ½ þ þ
Lm2 C1 Lm2 p2 D Cp3 Cp2 D
1 1 1
½ w ðiL ekt Þ þ w ðiL ekt ÞÞ
Cp3 2 mk Cp2 1 mk Cp3 Lm2
kt
C2 w2 ðiLmk e Þ 1
½i0p2 ekt þ iL ekt
D D Cm2 Lm2 m2
kt kt
w1 ðiLmk e Þ ¼ e w1 ðiLmk Þ; w2 ðiLmk e Þ ¼ ekt w2 ðiLmk Þ
kt
r2 1 C1 1 1 1
y2 k y2 þ i0p2 þ ½ w2 ðiLmk Þ þ w ðiL Þ
Lm2 D Lm2 Cp2 Lm2 D Cp3 Cp2 1 mk
1 w2 ðiLmk Þ 1
iL ¼ 0
Cp3 Lm2 D Cm2 Lm2 m2
dY3 1 C2 w2 r3 1
¼ ½ip2 þ Y3 ILm3
kt
1 C2 ½w2 ðILmk Þ þ w2 ðiLmk e Þ
y3 k ekt ¼ ½ðipðiÞ2 þ i0p2 ekt Þ þ
Cp3 Lm3 D D
ðiÞ r3 1 ðiÞ
ðY3 þ y3 ekt Þ ðILm þ iLm3 ekt Þ
Lm3 Cm3 Lm3 3
w2 ðILm Þ
kt 1 ðiÞ C2 ðiÞ r3 1 ðiÞ
y3 k e ¼ ½i þ k
Y3 I
Cp3 Lm3 p2 D D Lm3 Cm3 Lm3 Lm3
kt
1 C2 w2 ðiLmk e Þ r3
þ ½i0p2 ekt þ y3 ekt
Cp3 Lm3 D D Lm3
1
iL ekt
Cm3 Lm3 m3
ðiÞ w2 ðILm Þ ðiÞ ðiÞ

At fixed points: 1
Cp3 Lm3 ½ip2 CD2 þ D
k
Y3 Lrm3 Cm 1Lm ILm ¼ 0
3 3 3 3
1 C2 w2 ðiLmk Þ r3 1
y3 k þ ½i0p2 þ y3 iL ¼ 0
Cp3 Lm3 D D Lm3 Cm3 Lm3 m3
dY4 r4 1 C1 C2 C1 C1 1 1
1 Cp Cp 1
½ip2 1 C4 þ 1 w4 ILm4
Cp1 Lm4 D D Cm4 Lm4
ðiÞ r4 1 C1 C2 C1
y4 k ekt ¼ ðY4 þ y4 ekt Þ þ ð½ipðiÞ2 þ i0p2 ekt ½ þ
Lm4 C1 Lm4 D C p3 Cp2
C1 1 1
þ ½ ½w2 ðILm Þ þ w2 ðiLmk ekt Þ þ fw1 ðILm Þ þ w1 ðiLmk ekt ÞgÞ
D Cp3 k Cp2 k
1 C C
½ðipðiÞ2 þ i0p2 ekt Þ 1 C4 þ 1 fw4 ðILm Þ þ w4 ðiLmk ekt Þg
p p

Cp1 Lm4 D D k
1 ðiÞ
ðILm þ iLm4 ekt Þ
Cm4 Lm4 4
ðiÞ r4 1 C1 C2 C1 C1 1 1
y4 k ekt ¼ Y4 þ ðiðiÞ ½ þ þ ½ w ðI Þ þ w ðI ÞÞ
Lm4 C1 Lm4 p2 D Cp3 Cp 2 D Cp 3 2 L m k Cp2 1 Lmk
1 Cp Cp 1 ðiÞ r4
½iðiÞ 1 C4 þ 1 w4 ðILm Þ I y4 ekt
Cp1 Lm4 p2 D D k Cm4 Lm4 Lm4 Lm4
1 C 1 C2 C1 C1 1 1
þ ði0 ekt ½ þ þ ½ w ðiL ekt Þ þ w ðiL ekt ÞÞ
C1 Lm4 p2 D Cp3 Cp 2 D Cp 3 2 m k Cp 2 1 m k
1 Cp Cp 1
½i0 ekt 1 C4 þ 1 w4 ðiLmk ekt Þ iL ekt
Cp1 Lm4 p2 D D Cm 4 L m 4 m 4
At fixed points:
ðiÞ r4 1 C1 C2 C1 C1 1 1
Y4 þ ðipðiÞ2 ½ þ þ ½ w2 ðILm Þ þ w ðI ÞÞ
Lm4 C1 Lm4 D Cp3 Cp2 D Cp3 k Cp2 1 Lmk
1 Cp Cp 1 ðiÞ
½iðiÞ 1 C4 þ 1 w4 ðILm Þ I ¼0
Cp1 Lm4 p2 D D k Cm4 Lm4 Lm4
r4 1 C1 C2 C1 C1
y4 k ekt ¼ y4 ekt þ ði0 ekt ½ þ þ
Lm4 C1 Lm4 p2 D Cp3 Cp2 D
1 1 1
½ w ðiL ekt Þ þ w ðiL ekt ÞÞ
Cp3 2 mk Cp2 1 mk Cp1 Lm4
C C 1
½i0p2 ekt 1 C4 þ 1 w4 ðiLmk ekt Þ iL ekt
p p
D D Cm4 Lm4 m4
r4 1 C1 C2 C1 C1 1
y4 k y4 þ ði0 ½ þ þ ½ w ðiL Þ
Lm4 C1 Lm4 p2 D Cp3 Cp2 D Cp3 2 mk
1 1 Cp Cp 1
þ w1 ðiLmk ÞÞ ½i0p2 1 C4 þ 1 w4 ðiLmk Þ iL ¼ 0
Cp2 Cp1 Lm4 D D Cm4 Lm4 m4
ðiÞ ðiÞ
iLm1 k ekt ¼ Y1 þ y1 ekt ; Y1 ¼ 0; iLm1 k þ y1 ¼ 0
ðiÞ ðiÞ
ðiÞ ðiÞ
ðiÞ ðiÞ
dis1 C1 C2 C1 C1 1 1 1 n2 n1 1
¼ ½ip2 ½ þ þ ½ w2 þ w1 12
dt D Cp3 Cp2 D Cp3 Cp2 C1 Rs n2 n2 Rs
dVs ðtÞ

dt
C1 C2 C1 C1 1
i0s1 k ekt ¼ fðipðiÞ2 þ i0p2 ekt Þ ½ þ þ ½ fw2 ðILm Þ
D Cp3 Cp2 D Cp3 k
1 1
þ w2 ðiLmk ekt Þg þ fw1 ðILm Þ þ w1 ðiLmk ekt Þgg
Cp2 k C1 Rs
n21 n1 1 dVs ðtÞ

n22 n2 Rs dt
C1 C2 C1 C1 1 1
i0s1 k ekt ¼ fðipðiÞ2 Þ ½ þ þ ½ w ðI Þ þ w ðI Þg
D Cp3 Cp2 D Cp3 2 Lmk Cp2 1 Lmk
1 n2 n1 1 dVs ðtÞ C1 C2 C1
12 f ! eg fi0p2 ekt ½ þ
C1 Rs n2 n2 Rs dt D Cp3 Cp2
C1 1 1 1 n2
þ ½ w2 ðiLmk ekt Þ þ w1 ðiLmk ekt Þg 12
D Cp3 Cp2 C1 Rs n2
ðiÞ
At fixed points: fðip2 Þ CD1 ½CCp2 þ C1
Cp2 þ D
C1
½C1p w2 ðILm Þ þ 1
Cp2 w1 ðILm Þg
3 3 k k
n21
1
C1 Rs n22
¼0
C1 C2 C1 C1 1 1 1 n2
i0s1 k ekt ¼ fi0p2 ekt ½ þ þ ½ w2 ðiLmk ekt Þ þ w1 ðiLmk ekt Þg 12
D C p3 C p2 D C p3 C p2 C1 Rs n2
C1 C 2 C1 C1 1 1 1 n2
i0s1 k fi0p2 ½ þ þ ½ w2 ðiLmk Þ þ w1 ðiLmk Þg 12 ¼ 0
D Cp3 Cp2 D Cp3 Cp2 C1 R s n 2
dip2 n2 1 C2 C2 C2 1
¼ 42 fip2 ½ C4 þ ½ w w4 g
dt n3 C2 Rload D Cp3 D Cp3 2
n24 1 C2 C2 C2 1
i0p2 k ekt ¼ fðipðiÞ2 þ i0p2 ekt Þ ½ C4 þ ½ fw2 ðILm Þ
n23 C2 Rload D Cp3 D Cp3 k
þ w2 ðiLmk ekt Þg fw4 ðILm Þ þ w4 ðiLmk ekt Þgg

k
n24 1 C2 C2 C2 1
i0p2 k ekt ¼ fipðiÞ2 ½ C4 þ ½ w ðI Þ w4 ðILm Þg
n23 C2 Rload D Cp3 D Cp3 2 Lmk k
n24 1 C2 C2 C2 1
fi0p2 ekt ½ C4 þ ½ w ðiL ekt Þ
n23 C2 Rload D Cp3 D Cp3 2 mk
w4 ðiLmk ekt Þg
n2 ðiÞ
At fixed points: n42 C2 R1 load fip2 CD2 ½CCp2 C4 þ C2
D ½C1p w2 ðILm Þ
3 3 3 k
w4 ðILm Þg ¼ 0
k
n24 1 C2 C2 C2 1
i0p2 k fi0p2 ½ C4 þ ½ w ðiL Þ w4 ðiLmk Þg
n3 C2 Rload
2 D Cp3 D Cp3 2 mk
¼0
We can summery our Lattice crystal filter arbitrarily small increments equations:
1 Cp Cp r1 1
y1 k þ ½i0 1 C4 þ 1 w4 ðiLmk Þ y1 iL ¼ 0
r2 1 C1 1 1 1
y2 k y2 þ i0p2 þ ½ w ðiL Þ þ w ðiL Þ
Lm2 D Lm2 Cp2 Lm2 D Cp3 2 mk Cp2 1 mk
1 w2 ðiLmk Þ 1
iL ¼ 0
Cp3 Lm2 D Cm2 Lm2 m2
1 C2 w2 ðiLmk Þ r3 1
y3 k þ ½i0 þ y3 iL ¼ 0
r4 1 C1 C2 C1 C1 1
y4 k y4 þ ði0 ½ þ þ ½ w ðiL Þ
Lm4 C1 Lm4 p2 D Cp3 Cp2 D Cp3 2 mk
1 1 Cp Cp 1
þ w ðiL ÞÞ ½i0 1 C4 þ 1 w4 ðiLmk Þ iL ¼ 0
Cp2 1 mk Cp1 Lm4 p2 D D Cm4 Lm4 m4
iLm1 k þ y1 ¼ 0; iLm2 k þ y2 ¼ 0; iLm3 k þ y3 ¼ 0; iLm4 k þ y4 ¼ 0
C1 C2 C1 C1 1 1 1 n2
i0s1 k fi0p2 ½ þ þ ½ w2 ðiLmk Þ þ w1 ðiLmk Þg 12 ¼ 0
D Cp3 Cp2 D Cp3 Cp2 C1 Rs n2
n24 1 C2 C2 C2 1
i0p2 k fi0p2 ½ C4 þ ½ w ðiL Þ w4 ðiLmk Þg ¼ 0
n23 C2 Rload D Cp3 D Cp3 2 mk
0 1
y1
B y C
B 2C
B C
B y3 C
B C
B C
0 1 BB y 4 C
N11 . . . N1 10 C
B iLm C
B .. .. B
.. C B C r1
C
1
@ . . . A B iLm C ¼ 0; N11 ¼ k L ; N12 ¼ N13 ¼ N14 ¼ 0
B 2C m1
N10 N10 10 B C
1
B iLm3 C
B C
Bi C
B L m4 C
B C
B i0 C
@ s1 A
i0p2
Remark Reader exercise to find N15 ; N16 ; N17 ; N18 . N19 ¼ 0; N1 10 ¼ Lm1 D C4
1
r2 1 C1
N21 ¼ 0; N22 ¼ k ; N23 ¼ N24 ¼ 0; N29 ¼ 0; N2 10 ¼
Lm2 D Lm2 Cp2
Remark Reader exercise to find N25 ; N26 ; N27 ; N28 .
r3 1 C2
N31 ¼ N32 ¼ 0; N33 ¼ k ; N34 ¼ 0; N39 ¼ 0; N3 10 ¼
Lm3 Cp3 Lm3 D

r4
N41 ¼ N42 ¼ N43 ¼ 0; N44 ¼ k ; N49 ¼ 0; N4 10
Lm4
1 C2 C1 1
¼ ½ þ C4
Lm4 D Cp3 Cp2 Lm4 D
N51 ¼ 1; N52 ¼ N53 ¼ N54 ¼ 0; N55 ¼ k; N56 ¼ . . . ¼ N5 10 ¼ 0; N61 ¼ 0; N62 ¼ 1
N63 ¼ N64 ¼ 0; N65 ¼ 0; N66 ¼ k; N67 ¼ . . . ¼ N6 10 ¼ 0; N71 ¼ N72 ¼ 0; N73 ¼ 1; N74 ¼ 0
N75 ¼ N76 ¼ 0; N77 ¼ k; N78 ¼ N79 ¼ N7 10 ¼ 0; N81 ¼ N82 ¼ N83 ¼ 0; N84 ¼ 1
N85 ¼ N86 ¼ N87 ¼ 0; N88 ¼ k; N89 ¼ N8 10 ¼ 0; N91 ¼ . . . ¼ N94 ¼ 0; N99 ¼ k
C1 C2 C1 1 n2
N9 10 ¼ ½ þ 12 :
D Cp3 Cp2 C1 Rs n2
Remark: Reader exercise to find N95 ; N96 ; N97 ; N98 .
n24 1 C2
N10 1 ¼ . . . ¼ N10 4 ¼ 0; N10 9 ¼ 0; N10 10 ¼ k ½ C4 :
n3 Rload D Cp3
2
Remark Reader exercise to find N10 5 ; N10 6 ; N10 7 ; N10 8 .

0 1
N11 ... N1 10
B C
ðA k IÞ ¼ @ ... ..
.
..
. A; detðA k IÞ
N010 1 N10 10 1
N11 . . . N1 10
B .. .. .. C ¼ 0
¼ det@ . . . A
N10 1 N10 10
X
10 X
10
detðA k IÞ ¼ !k kk ; !k kk ¼ 0:
k¼0 k¼0
Remark It is reader exercise to find the expressions for !k 8 k ¼ 0; 1; . . .; 10.

Eigenvalues stability discussion: Our Lattice crystal filter circuit involving N
variables (N > 2, N = 10, arbitrarily small increments), the characteristic equation
is of degree N = 10 and must often be solved numerically. Expect in some par-
ticular cases, such an equation has (N = 10) distinct roots that can be real or
complex. These values are the eigenvalues of the (10 10) Jacobian matrix (A).
The general rule is that the Lattice crystal filter circuit is stable if there is no
eigenvalue with positive real part. It is sufficient that one eigenvalue is positive for
the steady state to be unstable [3, 4]. Our 10-variables (y1 ; y2 ; y3 ; y4 ; iLm1 ; iLm2 ; iLm3 ;
iLm4 ; i0s1 ; i0p2 ) system has ten eigenvalues (ten system’s arbitrarily small increments).
The type of behavior can be characterized as a function of the position of these
eigenvalues in the Re/Im plane. Five non-degenerated cases can be distinguished:
(1) the ten eigenvalues are real and negative (stable steady state), (2) the ten
eigenvalues are real, at least one of them is positive (unstable steady state), (3) and
(4) two eigenvalues are complex conjugates with a negative real part and other
eigenvalues real are negative (stable steady state), two cases can be distinguished
depending on the relative value of the real part of the complex eigenvalues and of
the real one, (5) two eigenvalues are complex conjugates with a negative real part
and at least one of the other eigenvalues real is positive (unstable steady state).
X
10 X
10
detðA k IÞ ¼ !k kk ; detðA k IÞ ¼ 0 ) !k kk ¼ 0
k¼0 k¼0
7.4 A Tunable BPF Employing Varactor Diodes Stability

Analysis
We have circuit which represents a tunable BPF employing varactor diodes. It is

ideal for many diverse wireless applications. There are two types of tunable BPF
employing varactor diodes; top inductively coupled variable BPF and capacitive
coupled variable band pass filter. The best for wideband applications is the top
inductively coupled variable band pass filter. In that subchapter we discuss the
stability analysis of capacitive coupled variable BPF (employing varactor diodes).
The design of the basic top capacitive coupled BPF is based on selecting a center
frequency for the top capacitive coupled BPF at either the high and, low end, or
middle of the tunable range of the desired bandpass frequencies, depending on the
initial tuning voltages we supply to the tuning capacitors. The circuit includes
varactor diodes (Cv) and CT capacitors, bias resistors R1 and R2, R1 resistor isolates
the two varactors (Cv) from the effects of each other, and resistor R2 represents a
direct RF short to ground through Vtune. Capacitor CT blocks the DC inserted by

Vtune from being shorted by L1 or L2. Varactor diodes (Cv) supply the variable
tuning capacitance. Capacitor Cc couples the two tank circuits consisting of L1 and
CT/Cv, and L2 and CT/Cv. The capacitance of the series combination of CT and Cv
in series is CCTTþCCv v . The capacitor CT mainly functioning as a DC blocking capacitor,
while Cv, the varactor is supplying all of the tuning capacitance for the filter’s tanks.
By applying positive tuning voltage Vtune, we allow the varactor to either linearly
tune the filter to its maximum and minimum values or, by supplying Vtune with
discrete voltages it filter the incoming RF/Microwave signal in discrete steps. Due
to parasitic capacitances and inductances at these frequencies we need to optimize
our filter and investigate his stability. Input RF/Microwave signal is Vs(t) and his
series resistance Rs. Typical varactors are limited in the value of their maximum
capacitance, and in case we want to operate our tunable BPF at low frequencies
region, then we need to increase the capacitance of CT and Cv combination and it is
done by adding capacitor Cs in shunt with CT and Cv which will increase the
capacitance in each leg to Cs þ CCTTþCCv v . The circuit load resistance Rload is connected
in our tunable BPF circuit (parallel to inductor L2) [25, 26].
Varactor diode: Varactor diode is an electronic component whose reactance can

be varied, usually electronically. Varactor diode is a variable capacitance diode.
Other names of varactors are varicap, tuning diode, and voltage variable capacitor.
Varactors are integral part in many RF and microwave circuits. The capacitance of a
varactor can be controlled electronically and automatic circuit tuning becomes
practical. A varactor is a diode in which P and N regions are doped in such a
manner that the capacitance that normally forms near the PN junction can be
precisely controlled by a reverse bias voltage. There is an inverse relationship
between the capacitance of the varactor and the applied voltage; a small reverse bias
Fig. 7.9 Tunable BPF employing varactor diodes

7.4 A Tunable BPF Employing Varactor Diodes Stability Analysis 747
voltage gives a big capacitance, and a large reverse bias voltage gives a small
capacitance. The varactor capacitance is changed as a function of the bias voltage.
There are two circuits’ models of varactor diode, high frequency circuit model and
low frequency circuit model. High frequency circuit model: Ls is the lead induc-
tance and Cc is the package inductance. These components of the model only
become significant at very high frequencies. Cj is the junction capacitance which
varies with applied voltage according to equation Cj ðVÞ ¼ ð1 þC0V Þn . V0 is the junc-
V0
tion potential with no bias voltage applied and is usually in the range of 0.5–0.7 v.
It can be determined by measuring the voltage drop across the diode when it is
conducting under forward bias. C0 is the capacitance with zero bias voltage. The
exponent n is dependent on the doping profile: n = 1/3 for a graded junction,
n = 1/2 for an abrupt junction, and n = 1 to n = 2 for hyperabrupt junction. V is the
applied voltage and is positive for reverse bias and negative for forward bias.
The Cj ðVÞ ¼ . . . equation does apply for forward bias voltages but only up to
about V ¼ V20 . It is possible to derive an expression for C0 but in practice the value
of C0 usually has to be determined experimentally. The series resistance, Rs, is due
to the resistance of the semiconductor material of which the diode is made as well
as any lead and contact resistance. The part of the diode which is not part of the
depletion region contributes to this resistance. Rs is a function of the bias voltage.
As reverse bias is increased, the depletion region gets larger and Rs gets smaller and
vice versa. The parallel resistance Rp represents the reverse leakage current. It is in
general varying somewhat with the applied voltage and becoming smaller very
rapidly near the reverse breakdown voltage. The low frequency model for a varactor
is as follow:
Fig. 7.10 High frequency circuit model for a varactor
Fig. 7.11 Low frequency model for a varactor

Fig. 7.12 Tunable BPF employing varactor diodes equivalent circuit
The Tunable BPF (varactor diodes) equivalent circuit is present in the below
figure. We consider Vs(t) low frequency RF source. We use varactor low frequency
model in our analysis.
Vs ðtÞ VA1 dIL

IR s ¼ ; VA1 ¼ L1 1 ;
Rs dt
dVA1 dðVA1 VA2 Þ
ICs ðtank1Þ ¼ Cs ; ICT ðtank1Þ ¼ CT
dt dt
dðVA2 VA3 Þ VA2 VA3
ICj1 ¼ Cj1 ; IRp1 ¼ ;
dt Rp1
VA dðVA1 VA6 Þ
IRs1 ¼ 3 ; ICc ¼ Cc
R s1 dt
dðVA6 VA4 Þ dðVA4 VA5 Þ VA VA5

ICT ðtank2Þ ¼ CT ; ICj2 ¼ Cj2 ; IRp2 ¼ 4 ; IRs2
dt dt Rp2
VA5
¼
Rs2
dVA6 dIL VA6

ICs ðtank2Þ ¼ Cs ; VA6 ¼ L2 2 ; IRload ¼ ; VL1 ¼ VCs ðtank1Þ ¼ VA1
dt dt Rload
Vtune VA7 VA VA2

VCs ðtank2Þ ¼ VL2 ¼ VRload ¼ VA6 ; IR2 ¼ ; IR1 ðtank1Þ ¼ 7 ; IR1 ðtank2Þ
R2 R1
VA7 VA4
¼
R1
VA2 is the applied voltage to varactor diode D1. VA4 is the applied voltage to
varactor diode D2.
C0 C0
Cj1 ðVA2 Þ ¼ VA2 n ; Cj1 ðVA4 Þ ¼ VA4 n ; ICj 1 þ IRp 1 ¼ IRs 1 ; ICj 2 þ IRp 2 ¼ IRs 2
ð1 þ V0 Þ ð1 þ V0 Þ
X
2
IR2 ¼ IR1 ðtank1Þ þ IR1 ðtank2Þ ¼ IR1 ðtankiÞ
i¼1
KCL @ node A 1: IRs ¼ IL1 þ ICs ðtank1Þ þ ICT ðtank1Þ þ ICc .

KCL @ node A 6: ICc ¼ ICT ðtank2Þ þ ICs ðtank2Þ þ IL2 þ IRload
KCL @ node A 2: ICT ðtank1Þ þ IR1 ðtank1Þ ¼ ICj1 þ IRp1
KCL @ node A 4: ICT ðtank2Þ þ IR1 ðtank2Þ ¼ ICj2 þ IRp2
dVA1 d dIL d 2 IL
ICs ðtank1Þ ¼ Cs ¼ Cs ðL1 1 Þ ¼ Cs L1 2 1 ; VA1
Zdt dt dt dt
1
¼ ICs ðtank1Þ dt
Cs
Z
dðVA1 VA2 Þ 1
ICT ðtank1Þ ¼ CT ) VA1 VA2 ¼ ICT ðtank1Þ dt;
dt CT
Z
1
VA2 ¼ VA1 ICT ðtank1Þ dt
CT
Z
dIL 1 dðVA2 VA3 Þ
VA2 ¼ L1 1 ICT ðtank1Þ dt; ICj1 ¼ Cj1
dt CT dt
Z
1
) VA2 VA3 ¼ ICj1 dt
Cj1
Z Z Z
1 dIL1 1 1
VA3 ¼ VA2 ICj1 dt ¼ L1 ICT ðtank1Þ dt ICj1 dt
Cj1 dt CT Cj1
Z
VA VA3 1
IRp1 ¼ 2 ) VA2 VA3 ¼ IRp1 Rp1 ; IRp1 Rp1 ¼ ICj1 dt
Rp1 Cj1
Z
d 1 dIRp1 1 dIRp1 1
fIR Rp1 ¼ ICj1 dtg ) Rp1 ¼ ICj1 ) ¼ ICj1
dt p1 Cj1 dt Cj1 dt Cj1 Rp1
VA3
IRs1 ¼ ) VA3 ¼ IRs1 Rs1 ; IRs1 Rs1
Rs1 Z Z
dIL 1 1
¼ L1 1 ICT ðtank1Þ dt ICj1 dt
dt CT Cj1
dIRs1 d 2 IL 1 1 dIRs1 1 1
Rs1 ¼ L 1 2 1 ICT ðtank1Þ ICj1 ; ¼ ICs ðtank1Þ ICT ðtank1Þ
dt dt CT Cj1 dt C s Rs1 C T Rs1
1
ICj1
Cj1 Rs1
Z
dðVA1 VA6 Þ 1
ICc ¼ Cc ) VA 1 VA 6 ¼ ICc dt
dt Cc
Z Z Z
1 1 1
VA6 ¼ VA1 ICc dt ¼ ICs ðtank1Þ dt ICc dt
Cc Cs Cc
Z
dðVA6 VA4 Þ 1
ICT ðtank2Þ ¼ CT ) VA6 VA4 ¼ ICT ðtank2Þ dt;
dt CT
Z
1
VA4 ¼ VA6 ICT ðtank2Þ dt
C
Z T Z
1 1
VA4 ¼ ICs ðtank1Þ dt ICc dt
Cs Cc
Z
1
ICT ðtank2Þ dt
CT
Z
dðVA4 VA5 Þ 1
ICj2 ¼ Cj2 ) VA4 VA5 ¼ ICj2 dt ) VA5
dtZ Cj2
1
¼ VA4 ICj2 dt
Cj2
Z Z Z Z
1 1 1 1
VA5 ¼ ICs ðtank1Þ dt ICc dt ICT ðtank2Þ dt ICj2
Cs Cc CT Cj2
dt
VA4 VA5
IRp2 ¼ ) VA4 VA5 ¼ IRp2 Rp2 ) IRp2 Rp2
Rp2
Z
1 dIRp2 1
¼ ICj2 dt; ¼ ICj2
Cj2 dt Cj2 Rp2
Z
VA5 1
IRs2 ¼ ) VA5 ¼ IRs2 Rs2 ; IRs2 Rs2 ¼ ICs ðtank1Þ dt
Rs2 Cs
Z Z Z
1 1 1
ICc dt ICT ðtank2Þ dt ICj2 dt
Cc CT Cj2
Z Z
d 1 1
fIR Rs2 ¼ ICs ðtank1Þ dt ICc dt
dt s2 Cs Cc
Z Z
1 1
ICT ðtank2Þ dt ICj2 dtg
CT Cj2
dIRs2 1 1 1 1
Rs2 ¼ ICs ðtank1Þ ICc ICT ðtank2Þ ICj2
dt Cs Cc CT Cj2
Z Z
dVA6 1 1
ICs ðtank2Þ ¼ Cs ) VA6 ¼ ICs ðtank2Þ dt; ICs ðtank2Þ dt
dt Cs Cs
Z Z
1 1
¼ ICs ðtank1Þ dt ICc dt
C C
Z s Z c Z
d 1 1 1
f ICs ðtank2Þ dt ¼ ICs ðtank1Þ dt ICc dtg;
dt Cs Cs Cc
1 1 1
ICs ðtank2Þ ¼ ICs ðtank1Þ ICc
Cs Cs Cc
dVA6 d dIL d 2 IL
ICs ðtank2Þ ¼ Cs ¼ Cs ðL2 2 Þ ¼ Cs L2 2 2 ;
dt dt dt dt
VA6
IRload ¼ ) VA6 ¼ IRload Rload
Rload
dIL2 dIL Rload Vtune VA7

VA6 ¼ IRload Rload ¼ L2 ) 2 ¼ IRload ; IR 2 ¼ ) VA7
dt dt L2 R2
¼ Vtune IR2 R2
1 d 2 IL dIR d 2 IL dIR 1
ICs ðtank2Þ ¼ L2 2 2 ; load Rload ¼ L2 2 2 ; load ¼ ICs ðtank2Þ
Cs dt dt dt dt Cs Rload
VA7 VA2
IR1 ðtank1Þ ¼ ) IR1 ðtank1Þ R1 ¼ VA7 VA2 ;
R1
Z
dIL 1
IR1 ðtank1Þ R1 ¼ Vtune IR2 R2 L1 1 þ ICT ðtank1Þ dt
dt CT
Z
d dIL1 1
fIR ðtank1Þ R1 ¼ Vtune IR2 R2 L1 þ ICT ðtank1Þ dtg
dt 1 dt CT
dIR1 ðtank1Þ dVtune dIR2 d 2 IL 1

R1 ¼ R2 L1 2 1 þ ICT ðtank1Þ ; IR1 ðtank2Þ
dt dt dt dt CT
VA VA4
¼ 7
R1
VA7 VA4
IR1 ðtank2Þ ¼ ) VA7 VA4 ¼ IR1 ðtank2Þ R1
R1
dIR1 ðtank2Þ dVtune dIR2 1 1 1

R1 ¼ R2 ICs ðtank1Þ þ ICc þ ICT ðtank2Þ
dt dt dt Cs Cc CT
dIR1 ðtank1Þ dI
We consider that IR1 ðtank1Þ IR1 ðtank2Þ ; dt R1dtðtank2Þ . Vtune Is a DC voltage
dIR1 ðtank1Þ dIR2 d2 I
then dVtune
dt ¼ 0. dt R1 þ dt R2 ¼ L1 dt2L1 þ C1T ICT ðtank1Þ .
dIR1 ðtank2Þ dIR2 1 1 1

R1 þ R2 ¼ ICs ðtank1Þ þ I Cc þ ICT ðtank2Þ
dt dt Cs Cc CT
dIR1 ðtank1Þ dIR2 dIR1 ðtank2Þ dIR2

R1 þ R2 R1 þ R2
dt dt dt dt
d 2 IL 1 1 1 1
) L1 2 1 þ ICT ðtank1Þ ¼ ICs ðtank1Þ þ I Cc þ ICT ðtank2Þ
dt CT Cs Cc CT
d 2 I L1 1 1 1 1
L1 ICT ðtank1Þ ¼ ICs ðtank1Þ ICc ICT ðtank2Þ
dt2 CT Cs Cc CT
d 2 IL1 dIL1 d 2 IL1

We define new variables: ICs ðtank1Þ ¼ Cs L1 dt2 ; Y1 ¼ dt ; dY1
dt ¼ dt2
dY1 1
¼ ICs ðtank1Þ
dt Cs L1
Tunable BPF (varactor diodes) differential equations group No. 1:
dIL1 dY1 1 dIRp1 1

¼ Y1 ; ¼ ICs ðtank1Þ ; ¼ ICj1
dt dt Cs L1 dt Cj1 Rp1
dIRs1 1 1 1
¼ ICs ðtank1Þ ICT ðtank1Þ ICj1
dt Cs Rs1 CT Rs1 Cj1 Rs1
dIRp2 1 dIRs2 1
¼ ICj2 ; ¼ ICs ðtank1Þ
dt Cj2 Rp2 dt Cs Rs2
1 1 1
I Cc ICT ðtank2Þ ICj2
Cc Rs2 CT Rs2 Cj2 Rs2
dIL2 Rload dIRload 1
¼ IRload ; ¼ ICs ðtank2Þ
dt L2 dt Cs Rload
Tunable BPF (varactor diodes) KCL equations:
1 1 1
ICs ðtank2Þ ¼ ICs ðtank1Þ ICc ; IRs ¼ IL1 þ ICs ðtank1Þ þ ICT ðtank1Þ þ ICc
Cs Cs Cc
ICc ¼ ICT ðtank2Þ þ ICs ðtank2Þ þ IL2 þ IRload ; ICT ðtank1Þ þ IR1 ðtank1Þ ¼ ICj1 þ IRp1
ICT ðtank2Þ þ IR1 ðtank2Þ ¼ ICj2 þ IRp2 ; ICj1 þ IRp1 ¼ IRs1 ; ICj2 þ IRp2 ¼ IRs2
X
2
IR2 ¼ IR1 ðtank1Þ þ IR1 ðtank2Þ ¼ IR1 ðtankiÞ
i¼1
ICj1 þ IRp1 ¼ IRs1 ) ICj1 ¼ IRs1 IRp1 ; ICj2 þ IRp2 ¼ IRs2 ) ICj2 ¼ IRs2 IRp2
ICT ðtank1Þ þ IR1 ðtank1Þ ¼ ICj1 þ IRp1 ¼ IRs1 ; ICT ðtank2Þ þ IR1 ðtank2Þ ¼ ICj2 þ IRp2 ¼ IRs2
dIL1 dY1 1 dIRp1 1

¼ Y1 ; ¼ ICs ðtank1Þ ; ¼ ðIRs1 IRp1 Þ
dIRs1 1 1 1
¼ ICs ðtank1Þ ICT ðtank1Þ ðIRs1 IRp1 Þ;
dt Cs Rs1 CT Rs1 Cj1 Rs1
dIRp2 1
¼ ðIRs2 IRp2 Þ
dt Cj2 Rp2
dIRs2 1 1 1 1
¼ ICs ðtank1Þ I Cc ICT ðtank2Þ
dt Cs Rs2 Cc Rs2 CT Rs2 Cj2 Rs2
ðIRs2 IRp2 Þ

dt L2 dt Cs Rload
ICT ðtank1Þ þ IR1 ðtank1Þ ¼ IRs1 ) ICT ðtank1Þ ¼ IRs1 IR1 ðtank1Þ ; ICT ðtank2Þ þ IR1 ðtank2Þ ¼ IRs2
) ICT ðtank2Þ ¼ IRs2 IR1 ðtank2Þ
dIL1 dY1 1 dIRp1 1

¼ Y1 ; ¼ ICs ðtank1Þ ; ¼ ðIRs1 IRp1 Þ
dIRs1 1 1
¼ ICs ðtank1Þ ðIRs1 IR1 ðtank1Þ Þ
dt Cs Rs1 CT Rs1
1 dIRp2 1
ðIRs1 IRp1 Þ; ¼ ðIRs2 IRp2 Þ
Cj1 Rs1 dt Cj2 Rp2
dIRs2 1 1 1
¼ ICs ðtank1Þ I Cc ðIRs2 IR1 ðtank2Þ Þ
dt Cs Rs2 Cc Rs2 CT Rs2
1
ðIRs2 IRp2 Þ
Cj2 Rs2

dt L2 dt Cs Rload
&&&
IRs ¼ IL1 þ ICs ðtank1Þ þ ICT ðtank1Þ þ ICc ; ICc ¼ ICT ðtank2Þ þ ICs ðtank2Þ þ IL2 þ IRload ;
1 1 1
ICs ðtank2Þ ¼ ICs ðtank1Þ ICc
Cs Cs Cc
ICc ¼ IRs IL1 ICs ðtank1Þ ICT ðtank1Þ ¼ IRs IL1 ICs ðtank1Þ IRs1 þ IR1 ðtank1Þ
IRs IL1 ICs ðtank1Þ IRs1 þ IR1 ðtank1Þ ¼ IRs2 IR1 ðtank2Þ þ ICs ðtank2Þ þ IL2 þ IRload
IRs IL1 IL2 IRs1 IRs2 þ IR1 ðtank1Þ þ IR1 ðtank2Þ IRload ¼ ICs ðtank1Þ þ ICs ðtank2Þ
X
2 X
2 X
2
IR s I Lk IRsi þ IR1 ðtankiÞ IRload ¼ ICs ðtank1Þ þ ICs ðtank2Þ
k¼1 i¼1 i¼1
1 1 1
ICs ðtank2Þ ¼ ICs ðtank1Þ ðIRs IL1 ICs ðtank1Þ IRs1 þ IR1 ðtank1Þ Þ
Cs Cs Cc
1 1 1 1
IRs þ I L1 þ IRs1 IR1 ðtank1Þ
Cc Cc Cc Cc
1 1 1
¼ ICs ðtank2Þ ð þ Þ ICs ðtank1Þ
Cs Cs Cc
X
2 X
2 X
2
ICs ðtank1Þ ¼ IRs I Lk IRsi þ IR1 ðtankiÞ IRload ICs ðtank2Þ
k¼1 i¼1 i¼1
1 1 1 1 1
IRs þ IL þ IRs1 IR1 ðtank1Þ ¼ ICs ðtank2Þ
Cc Cc 1 Cc Cc Cs
1 1 X 2 X2 X2
ð þ Þ ½IRs I Lk IRsi þ IR1 ðtankiÞ IRload ICs ðtank2Þ
Cs Cc k¼1 i¼1 i¼1
1 1 1 1 1
IRs þ I L1 þ IRs1 IR1 ðtank1Þ ¼ ICs ðtank2Þ
Cc Cc Cc Cc Cs
1 1 1 1 X 2
1 1 X 2
1 1
ð þ Þ IRs þ ð þ Þ ILk þ ð þ Þ IR ð þ Þ
Cs Cc Cs Cc k¼1 Cs Cc i¼1 si Cs Cc
X
2
1 1 1 1
IR1 ðtankiÞ þ ð þ Þ IRload þ ð þ Þ ICs ðtank2Þ
i¼1
Cs Cc Cs Cc
2 1 1 1 1 1 1 1
ð þ Þ ICs ðtank2Þ ¼ IRs þ IL þ IRs1 IR1 ðtank1Þ þ ð þ Þ
Cs Cc Cc Cc 1 Cc Cc Cs Cc
X2
1 1 1 1 X 2
IR1 ðtankiÞ þ ð þ Þ IRs ð þ Þ IL
i¼1
Cs Cc Cs Cc k¼1 k
1 1 X 2
1 1
ð þ Þ IRsi ð þ Þ IRload
Cs Cc i¼1 Cs Cc
2 1 1 1 1 1
ð þ Þ ICs ðtank2Þ ¼ IRs þ IR1 ðtank1Þ þ IR1 ðtank2Þ ð þ Þ
Cs Cc Cs Cs Cs Cc
1 1 1 1 1 1
IL1 ð þ Þ IL2 IRs1 ð þ Þ
Cs Cs Cc Cs Cs Cc
1 1
IRs2 ð þ Þ IRload
Cs Cc
1 1 ðC1s þ Cc Þ
1
ICs ðtank2Þ ¼ IRs þ IR1 ðtank1Þ þ IR1 ðtank2Þ
ðC2s þ Cc Þ
1
Cs ðC2s þ Cc Þ
1
Cs ðC2s þ C1c Þ
1 ðC1s þ Cc Þ
1
1 ð1 þ Cc Þ
1
I L1 I L2 IRs1 C2s
Cs ðC2s þ Cc Þ
1
ðC2s þ Cc Þ
1
Cs ðC2s þ Cc Þ
1
ð Cs þ Cc Þ
1
ðC1s þ C1c Þ
IRs2 IRload
ðC2s þ C1c Þ
1 1 ðCc þ Cs Þ
ICs ðtank2Þ ¼ IRs þ IR1 ðtank1Þ þ IR1 ðtank2Þ
ð2 þ Cc Þ
Cs
ð2 þ Cc Þ
Cs ð2 Cc þ Cs Þ
1 ðCc þ Cs Þ 1
IL1 IL IR
ð2 þ CCcs Þ ð2 Cc þ Cs Þ 2 ð2 þ CCs Þ s1
c
ðCc þ Cs Þ ðCc þ Cs Þ
I R s2 IR
ð2 Cc þ Cs Þ ð2 Cc þ Cs Þ load
ðC1s þ Cc Þ
1
ðC1s þ Cc Þ
1
Cs Cc ðCc þ Cs Þ
C1 ¼ C1 ðCc ; Cs Þ ¼ ¼ ¼
ðC2s þ Cc Þ
1
ðC2s þ Cc Þ
1 Cs Cc ð2 Cc þ Cs Þ
1 1 1
ICs ðtank2Þ ¼ IRs þ IR1 ðtank1Þ þ IR1 ðtank2Þ C1 ðCc ; Cs Þ I L1
ð2 þ CCcs Þ ð2 þ CCcs Þ ð2 þ CCcs Þ
1
C1 ðCc ; Cs Þ IL2 IRs1 C1 ðCc ; Cs Þ IRs2 C1 ðCc ; Cs Þ IRload
ð2 þ CCcs Þ
1 1
ICs ðtank1Þ ¼ IRs þ ½ 1 IL1 þ ½C1 ðCc ; Cs Þ 1 IL2 þ ½ 1 IRs1
ð2 þ CCcs Þ ð2 þ CCcs Þ
1
þ ½C1 ðCc ; Cs Þ 1 IRs2 þ IR1 ðtank1Þ ½1 þ IR1 ðtank2Þ
ð2 þ CCcs Þ
1
½1 C1 ðCc ; Cs Þ þ ½C1 ðCc ; Cs Þ 1 IRload IRs
ð2 þ CCcs Þ
1 1
C2 ¼ C2 ðCs ; Cc Þ ¼ 1 Cs ; C2 ¼ C2 ðCs ; Cc Þ ¼ 1
ð2 þ Cc Þ ð2 þ CCcs Þ
C3 ¼ C3 ðCc ; Cs Þ ¼ 1 C1 ðCc ; Cs Þ; C3 ¼ C3 ðCc ; Cs Þ ¼ C1 ðCc ; Cs Þ 1
ICs ðtank1Þ ¼ IRs C2 IL1 C3 IL2 C2 IRs1 C3 IRs2 þ IR1 ðtank1Þ C2 þ IR1 ðtank2Þ
1
C3 C3 IRload IRs
ð2 þ CCcs Þ
1 1 1
ICs ðtank2Þ ¼ IRs þ IR1 ðtank1Þ þ IR1 ðtank2Þ C1 I L1
ð2 þ CCcs Þ ð2 þ CCcs Þ ð2 þ CCcs Þ
1
C1 IL2 IRs1 C1 IRs2 C1 IRload
ð2 þ CCcs Þ
dIL1 dY1 1
¼ Y1 ; ¼ fIRs C2 IL1 C3 IL2 C2 IRs1 C3 IRs2
dt dt Cs L1
1
þ IR1 ðtank1Þ C2 þ IR1 ðtank2Þ C3 C3 IRload IRs g;
ð2 þ CCcs Þ
dIRp1 1
¼ ðIRs1 IRp1 Þ
dt Cj1 Rp1
dIRs1 1
¼ ðIRs C2 IL1 C3 IL2 C2 IRs1 C3 IRs2 þ IR1 ðtank1Þ C2
dt Cs Rs1
1 1
þ IR1 ðtank2Þ C3 C3 IRload IR s Þ ðIRs1 IR1 ðtank1Þ Þ
ð2 þ CCcs Þ C T Rs1
1 dIRp2 1
Cj1 Rs1 dt Cj2 Rp2
dIRs2 1
¼ ðIRs C2 IL1 C3 IL2 C2 IRs1 C3 IRs2 þ IR1 ðtank1Þ C2 þ IR1 ðtank2Þ
dt Cs Rs2
1 1 1
C3 C3 IRload Cs IRs Þ C R ICc C R ðIRs2 IR1 ðtank2Þ Þ
ð2 þ Cc Þ c s2 T s2
1
ðIRs2 IRp2 Þ
Cj2 Rs2
dIL2 Rload dIRload 1 1 1

¼ IRload ; ¼ f IR þ IR1 ðtank1Þ
dt L2 dt Cs Rload ð2 þ CCs Þ s ð2 þ Cc Þ
Cs
c
1 1
þ IR1 ðtank2Þ C1 Cs IL1 C1 IL2
ð2 þ Cc Þ ð2 þ CCcs Þ
IRs1 C1 IRs2 C1 IRload g
ICc ¼ IRs IL1 ICs ðtank1Þ IRs1 þ IR1 ðtank1Þ

¼ IRs IL1 fIRs C2 IL1 C3 IL2 C2 IRs1
C3 IRs2 þ IR1 ðtank1Þ C2 þ IR1 ðtank2Þ C3 C3 IRload
1
IRs g IRs1 þ IR1 ðtank1Þ
ð2 þ CCcs Þ
ICc ¼ IRs IL1 IRs þ C2 IL1 þ C3 IL2 þ C2 IRs 1 þ C3 IRs 2 IR1 ðtank1Þ C2
1
IR1 ðtank2Þ C3 þ C3 IRload þ IRs IRs 1 þ IR1 ðtank1Þ
ð2 þ CCcs Þ
1
ICc ¼ ½1 þ IRs þ ½C2 1 IL1 IRs þ C3 IL2 þ ½C2 1 IRs 1 þ C3 IRs 2
ð2 þ CCcs Þ
IR1 ðtank2Þ C3 þ C3 IRload þ IR1 ðtank1Þ ½1 C2
dIL1 dY1 1
¼ Y1 ; ¼ fIRs C2 IL1 C3 IL2 C2 IRs1 C3 IRs2 þ IR1 ðtank1Þ C2
dt dt Cs L1
1 dIRp1 1
þ IR1 ðtank2Þ C3 C3 IRload IRs g; ¼ ðIRs1 IRp1 Þ
ð2 þ CCcs Þ dt Cj1 Rp1
dIRs1 1
¼ ðIRs C2 IL1 C3 IL2 C2 IRs1 C3 IRs2 þ IR1 ðtank1Þ
dt Cs Rs1
1 1 1
C2 þ IR1 ðtank2Þ C3 C3 IRload IRs Þ ðIRs1 IR1 ðtank1Þ Þ
ð2 þ CCcs Þ CT Rs1 Cj1 Rs1
dIRp2 1
dt Cj2 Rp2
dIRs2 1
¼ ðIRs C2 IL1 C3 IL2 C2 IRs1 C3 IRs2 þ IR1 ðtank1Þ
dt Cs Rs2
1 1 1
C2 þ IR1 ðtank2Þ C3 C3 IRload IRs Þ f½1 þ IRs
ð2 þ Cc Þ
Cs C c R s2 ð2 þ CCcs Þ
þ ½C2 1 IL1 IRs þ C3 IL2 þ ½C2 1 IRs1 þ C3 IRs2
1
IR1 ðtank2Þ C3 þ C3 IRload þ IR1 ðtank1Þ ½1 C2 g ðIRs2 IR1 ðtank2Þ Þ
CT Rs2
1
ðIRs2 IRp2 Þ
Cj2 Rs2
dIL2 Rload dIRload 1 1 1

¼ IRload ; ¼ f IR þ IR1 ðtank1Þ
dt L2 dt Cs Rload ð2 þ CCs Þ s ð2 þ CCs Þ
c c
1 1
þ IR1 ðtank2Þ C1 IL1 C1 IL2 IRs1 C1 IRs2 C1 IRload g
ð2 þ CCcs Þ ð2 þ CCcs Þ
dY1 1 1 1 1
¼ IR s C2 IL1 C 3 I L2 C2
dt Cs L1 Cs L1 Cs L1 Cs L1
1 1 1
IRs1 C3 IRs2 þ IR1 ðtank1Þ C2 þ IR1 ðtank2Þ
Cs L1 Cs L1 Cs L1
1 1 1
C3 C3 IRload IR
Cs L1 Cs L1 ð2 þ CCs Þ s
c
dIRp1 1 1 dIL
¼ IRs1 IRp1 ; 1 ¼ Y1
dt Cj1 Rp1 Cj1 Rp1 dt
dIRs1 1 1 1 1
¼ ½1 IRs C 2 I L1 C3 IL2
dt ð2 þ C Þ Cs Rs1
C s
c
Cs Rs1 Cs Rs1
1 1 1 1 1 1 1
½ C2 þ þ I R s1 C3 IRs2 þ ½ C2 þ
Cs CT Cj1 Rs1 Cs Rs1 Cs CT
1
IR1 ðtank1Þ þ IR1 ðtank2Þ
Rs1
1 1 1
C3 C3 IRload þ I R p1
Cs Rs1 Cs Rs1 Cj1 Rs1
dIRp2 1 1
¼ IRs2 IRp2
dt Cj2 Rp2 Cj2 Rp2
dIRs2 1 1 1 1 1 1 1
¼f ½1 þ Cs þ C g R IRs
dt Cs Cs ð2 þ CCs Þ Cc ð2 þ Cc Þ c s2
c
1 1 1 1 1 1
ð C2 þ ½C2 1Þ IL1 ð þ Þ C3 IL2
Cs Cc Rs2 Cs Cc Rs2
1 1 1 1 1 1 1
ð C2 þ ½C2 1Þ IRs1 ð C3 þ C3 þ þ Þ
Cs Cc Rs2 Cs Cc Cj2 CT
1 1 1 1 1 1 1
IRs2 þ ð C2 ½1 C2 Þ IR1 ðtank1Þ þ ð C3 þ þ C3 Þ
Rs2 Cs Cc Rs2 Cc CT Cs
1 1 1 1 1
IR1 ðtank2Þ ð þ Þ C3 IRload þ IRp2
Rs2 Cc Cs Rs2 Cj2 Rs2
dIL2 Rload dIRload 1 1 1 1

¼ IRload ; ¼ IRs þ IR1 ðtank1Þ
dt L2 dt Cs Rload ð2 þ C Þ
C s Cs Rload ð2 þ CCs Þ
c c
1 1 1 1
þ IR1 ðtank2Þ C1 IL C1 IL2
Cs Rload Cs Rload ð2 þ CCs Þ 1 Cs Rload
c
1 1 1 1
IRs1 C1 IRs2 C1 IRload
Cs Rload ð2 þ C Þ
C s Cs Rload Cs Rload
c
1 1 1 1 1
X1 ¼ ; X2 ¼ ; X3 ¼ ½1 ; X4 ¼ C2 ;
Cs L1 Cj1 Rp1 ð2 þ CCs Þ Cs Rs1 Cs Rs1
c
1 1 1 1 1 1 1 1
X5 ¼ C3 ; X6 ¼ ½ C2 þ þ ; X7 ¼ ½ C2 þ ;
Cs Rs1 Cs CT Cj1 Rs1 Cs CT Rs1
1
X8 ¼
Cj1 Rs1
1 1 1 1 1 1 1 1
X9 ¼ ; X10 ¼ f ½1 þ þ g
Cj2 Rp2 Cs Cs ð2 þ CCs Þ Cc ð2 þ Cs
Cc Þ C c R s2
c
1 1 1 1 1 1
X11 ¼ ð C2 þ ½C2 1Þ ; X12 ¼ ð þ Þ C3
Cs Cc Rs2 Cs Cc Rs2
1 1 1 1 1 1 1 1 1
X13 ¼ ð C3 þ C3 þ þ Þ ; X14 ¼ ð C3 þ þ C3 Þ
Cs Cc Cj2 CT Rs2 Cc CT Cs R s2
1 1 1 1
X15 ¼ ; X16 ¼ ; X17 ¼ C1
Cj2 Rs2 Cs Rload ð2 þ CCs Þ Cs Rload
c
dY1
¼ X1 IRs X1 C2 IL1 X1 C3 IL2 X1 C2 IRs1 X1 C3 IRs2
dt
1
þ X1 IR1 ðtank1Þ C2 þ X1 IR1 ðtank2Þ C3 X1 C3 IRload X1 IRs
ð2 þ CCcs Þ
dIRp1 dIL dIRp2 dIL

¼ X2 IRs1 X2 IRp1 ; 1 ¼ Y1 ; ¼ X9 IRs2 X9 IRp2 ; 2
dt dt dt dt
Rload
¼ IRload
L2
dIRs1
¼ X3 IRs X4 IL1 X5 IL2 X6 IRs1 X5 IRs2 þ X7
dt
IR1 ðtank1Þ þ IR1 ðtank2Þ X5 X5 IRload þ X8 IRp1
dIRs2
¼ X10 IRs X11 IL1 X12 IL2 X11 IRs1 X13 IRs2 þ X11 IR1 ðtank1Þ
dt
þ X14 IR1 ðtank2Þ X12 IRload þ X15 IRp2
dIRload
¼ X16 IRs þ IR1 ðtank1Þ X16 þ IR1 ðtank2Þ X17 X16 IL1 X17 IL2
dt
X16 IRs1 X17 IRs2 X17 IRload
dIRp1 dIL1 dIRp2 dIL2 dIRs1

At fixed points: dY1
dt ¼ 0; dt ¼ 0; dt ¼ 0; dt ¼ 0; dt ¼ 0; dt ¼0
dIRs2 dIR
¼ 0; load ¼ 0
dt dt
IRp ¼ IRs ; Y1 ¼ 0;IRp ¼ IRs ; IRload ¼ 0
1 1 2 2
1
½1 X1 IRs X1 C2 IL1 X1 C3 IL2 X1 C2 IRs X1 C3 IRs
ð2 þ CCcs Þ 1 2
þ X1 C2 IR1 ðtank1Þ þ X1 C3 IR1 ðtank2Þ ¼ 0
X3 IRs X4 IL1 X5 IL2 þ ½X8 X6 IRs X5 IRs þ X7 IR1 ðtank1Þ þ IR1 ðtank2Þ
1 2
X5 ¼ 0
X10 IRs X11 IL1 X12 IL2 X11 IRs þ ½X15 X13 IRs þ X11
1 2
IR1 ðtank1Þ þ X14 IR1 ðtank2Þ
¼0
X16 IRs þ IR1 ðtank1Þ X16 þ IR1 ðtank2Þ X17 X16 IL1 X17 IL2 X16 IRs X17
1
IRs
2
¼0
equilibrium points of Tunable BPF (varactor diodes) circuit consists in adding to its
coordinated [Y1 IRp1 IL1 IRs1 IRp2 IRs2 IL2 IRload IRs IR1 ðtank1Þ IR1 ðtank2Þ ] arbitrarily small
increments of exponential terms ½y1 iRp1 iL1 iRs1 iRp2 iRs2 iL2 iRload iRs iR1 ðtank1Þ
iR1 ðtank2Þ ekt , and retaining the first order terms in y1 iRp1 iL1 iRs1 iRp2 iRs2 iL2 iRload
iRs iR1 ðtank1Þ iR1 ðtank2Þ . The system of eight homogeneous equations leads to a
polynomial characteristic equation in the eigenvalue k. The polynomial charac-
teristic equation accepts by set the tunable BPF (varactor diodes) circuit equations.
The tunable BPF (varactor diodes) circuit fixed values with arbitrarily small incre-
ments of exponential form ½y1 iRp1 iL1 iRs1 iRp2 iRs2 iL2 iRload iRs iR1 ðtank1Þ iR1 ðtank2Þ ekt
are; i = 0 (first fixed point), i = 1 (second fixed point), i = 2 (third fixed point), etc.,
Y1 ðtÞ ¼ Y1 þ y1 ekt ; IRp1 ðtÞ ¼ IRp þ iRp1 ekt ; IL1 ðtÞ ¼ IL þ iL1 ekt ; IRs1 ðtÞ ¼ IRs þ iRs1 ekt
1 1 1
IRp2 ðtÞ ¼ IRp þ iRp2 ekt ; IRs2 ðtÞ ¼ IRs þ iRs2 ekt ; IL2 ðtÞ ¼ IL þ iL2 ekt ; IRload ðtÞ ¼ IRload þ iRload ekt
2 2 2
ðiÞ ðiÞ ðiÞ
IRs ðtÞ ¼ IRs þ iRs ekt ; IR1 ðtank1Þ ðtÞ ¼ IR1 ðtank1Þ þ iR1 ðtank1Þ ekt ; IR1 ðtank2Þ ðtÞ ¼ IR1 ðtank2Þ þ iR1 ðtank2Þ ekt
dY1 ðtÞ dIRp1 ðtÞ dIL ðtÞ dIRs1 ðtÞ

¼ y1 k ekt ; ¼ iRp1 k ekt ; 1 ¼ iL1 k ekt ;
dt dt dt dt
¼ iRs1 k ekt
dIRp2 ðtÞ dIRs2 ðtÞ dIL ðtÞ dIR ðtÞ

¼ iRp2 k ekt ; ¼ iRs2 k ekt ; 2 ¼ iL2 k ekt ; load
dt dt dt dt
¼ iRload k ekt
Our BPF (varactor diodes) circuit tends to the selected fixed point exponentially for
k < 0, t > 0 otherwise go away from the selected fixed point exponentially. k is the
the selected fixed point [2–4].
dY1
¼ X1 IRs X1 C2 IL1 X1 C3 IL2 X1 C2 IRs1 X1 C3 IRs2
dt
1
þ X1 IR1 ðtank1Þ C2 þ X1 IR1 ðtank2Þ C3 X1 C3 IRload X1 IRs
ð2 þ CCcs Þ
ðiÞ ðiÞ
y1 k ekt ¼ X1 ðIRs þ iRs ekt Þ X1 C2 ðIL þ iL1 ekt Þ X1 C3
1
ðiÞ ðiÞ ðiÞ
ðIL þ iL2 ekt Þ X1 C2 ðIRs þ iRs1 ekt Þ X1 C3 ðIRs þ iRs2 ekt Þ
2 1 2
ðiÞ ðiÞ
þ X1 C2 ðIR1 ðtank1Þ þ iR1 ðtank1Þ ekt Þ þ X1 C3 ðIR1 ðtank2Þ þ iR1 ðtank2Þ ekt Þ
ðiÞ 1 ðiÞ
X1 C3 ðIRload þ iRload ekt Þ X1 ðIRs þ iRs ekt Þ
ð2 þ CCcs Þ

y1 k ekt ¼ X1 IRs X1 C2 IL X1 C3 IL X1 C2 IRs X1 C3 IRs
1 2 1 2
ðiÞ ðiÞ ðiÞ 1

þ X1 C 2 IR1 ðtank1Þ þ X1 C 3 IR1 ðtank2Þ X1 C3 IRload X1
ð2 þ CCcs Þ
ðiÞ
IRs þ X1 iRs ekt X1 C2 iL1 ekt X1 C3 iL2 ekt X1 C2 iRs1
ekt X1 C3 iRs2 ekt þ X1 C2 iR1 ðtank1Þ ekt þ X1 C3 iR1 ðtank2Þ ekt
1
X1 C3 iRload ekt X1 iRs ekt
ð2 þ CCcs Þ
At fixed points:

X1 IRs X1 C2 IL X1 C3 IL X1 C2 IRs X1 C3 IRs
1 2 1 2
ðiÞ ðiÞ ðiÞ 1 ðiÞ

þ X1 C2 IR1 ðtank1Þ þ X1 C3 IR1 ðtank2Þ X1 C3 IRload X1 IR ¼ 0
ð2 þ CCcs Þ s
y1 k X1 C2 iL1 X1 C3 iL2 X1 C2 iRs1 X1 C3 iRs2

1
þ X1 C2 iR1 ðtank1Þ þ X1 C3 iR1 ðtank2Þ X1 C3 iRload þ X1 iRs ½1 ¼0
ð2 þ CCcs Þ
dIRp1
¼ X2 IRs1 X2 IRp1 ; iRp1 k ekt
dt
ðiÞ ðiÞ
¼ X2 ðIRs þ iRs1 ekt Þ X2 ðIRp þ iRp1 ekt Þ
1 1
ðiÞ ðiÞ
iRp1 k ekt ¼ X2 IRs X2 IRp þ X2 iRs1 ekt X2 iRp1 ekt
1 1
ðiÞ ðiÞ
At fixed points: X2 IRs X2 IRp ¼ 0 then iRp1 k þ X2 iRs1 X2 iRp1 ¼ 0
1 1
dIL1 ðiÞ ðiÞ

¼ Y1 ; iL1 k ekt ¼ Y1 þ y1 ekt ; Y1 ¼ 0 ) iL1 k þ y1 ¼ 0
dt
dIRp2
¼ X9 IRs2 X9 IRp2 ; iRp2 k ekt
dt
ðiÞ ðiÞ
¼ X9 ðIRs þ iRs2 ekt Þ X9 ðIRp þ iRp2 ekt Þ
2 2
ðiÞ ðiÞ
iRp2 k ekt ¼ X9 IRs X9 IRp þ X9 iRs2 ekt X9 iRp2 ekt
2 2
ðiÞ ðiÞ
At fixed points: X9 IRs X9 IRp ¼ 0 then iRp2 k þ X9 iRs2 X9 iRp2 ¼ 0
2 2
dIL2 Rload ðiÞ Rload

¼ IRload ; iL2 k ekt ¼ ðIRload þ iRload ekt Þ ; iL2 k ekt
dt L2 L2
ðiÞ Rload Rload kt
¼ IRload þ iRload e
L2 L2
ðiÞ
At fixed points: IRload RLload
2
¼ 0 then iL2 k þ iRload RLload
2
¼0
dIRs1
¼ X3 IRs X4 IL1 X5 IL2 X6 IRs1 X5 IRs2 þ X7
dt
IR1 ðtank1Þ þ IR1 ðtank2Þ X5 X5 IRload þ X8 IRp1
ðiÞ ðiÞ ðiÞ

iRs1 k ekt ¼ X3 ðIRs þ iRs ekt Þ X4 ðIL þ iL1 ekt Þ X5 ðIL þ iL2 ekt Þ
1 2
ðiÞ ðiÞ
X6 ðIRs þ iRs1 ekt Þ X5 ðIRs þ iRs2 ekt Þ þ X7
1 2
ðiÞ ðiÞ
ðIR1 ðtank1Þ þ iR1 ðtank1Þ ekt Þ þ ðIR1 ðtank2Þ þ iR1 ðtank2Þ ekt Þ X5
ðiÞ ðiÞ
X5 ðIRload þ iRload ekt Þ þ X8 ðIRp þ iRp1 ekt Þ
1
ðiÞ ðiÞ ðiÞ ðiÞ ðiÞ ðiÞ

iRs1 k ekt ¼ X3 IRs X4 IL X5 IL X6 IRs X5 IRs þ X7 IR1 ðtank1Þ
1 2 1 2
ðiÞ ðiÞ ðiÞ
þ IR1 ðtank2Þ X5 X5 IRload þ X8 IRp þ X3 iRs ekt X4 iL1 ekt
1
X5 iL2 ekt X6 iRs1 ekt X5 iRs2 ekt þ X7 iR1 ðtank1Þ

ekt þ iR1 ðtank2Þ ekt X5 X5 iRload ekt þ X8 iRp1 ekt
At fixed points:
ðiÞ ðiÞ ðiÞ ðiÞ ðiÞ ðiÞ ðiÞ

X3 IRs X4 IL X5 IL X6 IRs X5 IRs þ X7 IR1 ðtank1Þ þ IR1 ðtank2Þ
1 2 1 2
ðiÞ ðiÞ
X5 X5 IRload þ X8 IRp ¼ 0
1
iRs1 k þ X3 iRs X4 iL1 X5 iL2 X6 iRs1 X5 iRs2 þ X7

iR1 ðtank1Þ þ iR1 ðtank2Þ X5 X5 iRload þ X8 iRp1
¼0
dIRs2
¼ X10 IRs X11 IL1 X12 IL2 X11 IRs1 X13 IRs2 þ X11 IR1 ðtank1Þ
dt
þ X14 IR1 ðtank2Þ X12 IRload þ X15 IRp2
ðiÞ ðiÞ ðiÞ

iRs2 k ekt ¼ X10 ðIRs þ iRs ekt Þ X11 ðIL þ iL1 ekt Þ X12 ðIL þ iL2 ekt Þ
1 2
ðiÞ ðiÞ ðiÞ
X11 ðIRs þ iRs1 ekt Þ X13 ðIRs þ iRs2 ekt Þ þ X11 ðIR1 ðtank1Þ
1 2
ðiÞ
þ iR1 ðtank1Þ ekt Þ þ X14 ðIR1 ðtank2Þ þ iR1 ðtank2Þ ekt Þ
ðiÞ ðiÞ
X12 ðIRload þ iRload ekt Þ þ X15 ðIRp þ iRp2 ekt Þ
2

iRs2 k ekt ¼ X10 IRs X11 IL X12 IL X11 IRs X13 IRs þ X11
1 2 1 2
IR1 ðtank1Þ þ X14 IR1 ðtank2Þ X12 IRload þ X15 I Rp þ X10 iRs ekt
2
X11 iL1 ekt X12 iL2 ekt X11 iRs1 ekt X13 iRs2 ekt
þ X11 iR1 ðtank1Þ ekt þ X14 iR1 ðtank2Þ ekt X12 iRload ekt þ X15 iRp2 ekt
At fixed points:
ðiÞ ðiÞ ðiÞ ðiÞ ðiÞ ðiÞ ðiÞ
X10 IRs X11 IL X12 IL X11 IRs X13 IRs þ X11 IR1 ðtank1Þ þ X14 IR1 ðtank2Þ
1 2 1 2
ðiÞ ðiÞ
X12 IRload þ X15 IRp ¼ 0
2
iRs2 k þ X10 iRs X11 iL1 X12 iL2 X11 iRs1 X13 iRs2 þ X11 iR1 ðtank1Þ
þ X14 iR1 ðtank2Þ X12 iRload þ X15 iRp2 ¼ 0
dIRload
¼ X16 IRs þ IR1 ðtank1Þ X16 þ IR1 ðtank2Þ X17 X16 IL1 X17 IL2
dt
X16 IRs1 X17 IRs2 X17 IRload
ðiÞ ðiÞ
iRload k ekt ¼ X16 ðIRs þ iRs ekt Þ þ ðIR1 ðtank1Þ þ iR1 ðtank1Þ ekt Þ X16
ðiÞ ðiÞ
þ ðIR1 ðtank2Þ þ iR1 ðtank2Þ ekt Þ X17 X16 ðIL þ iL1 ekt Þ
1
ðiÞ ðiÞ ðiÞ
X17 ðIL þ iL2 ekt Þ X16 ðIRs þ iRs1 ekt Þ X17 ðIRs þ iRs2
2 1 2
ðiÞ
ekt Þ X17 ðIRload þ iRload ekt Þ

iRload k ekt ¼ X16 IRs þ IR1 ðtank1Þ X16 þ IR1 ðtank2Þ X17 X16 IL X17 IL
1 2
ðiÞ ðiÞ ðiÞ
X16 IRs X17 IR s X17 IRload þ X16 iRs e þ iR1 ðtank1Þ ekt
kt
1 2
X16 þ iR1 ðtank2Þ ekt X17 X16 iL1 ekt X17 iL2 ekt X16 iRs1
ekt X17 iRs2 ekt X17 iRload ekt
At fixed points:
ðiÞ ðiÞ ðiÞ ðiÞ ðiÞ ðiÞ

X16 IRs þ IR1 ðtank1Þ X16 þ IR1 ðtank2Þ X17 X16 IL X17 IL X16 IRs
1 2 1
ðiÞ ðiÞ
X17 IRs X17 IRload ¼ 0
2
iRload k þ X16 iRs þ iR1 ðtank1Þ X16 þ iR1 ðtank2Þ X17 X16 iL1 X17 iL2 X16
iRs1 X17 iRs2 X17 iRload
¼0
We can summery our BPF (varactor diodes) circuit arbitrarily small increments
equations:
y1 k X1 C2 iL1 X1 C3 iL2 X1 C2 iRs1 X1 C3 iRs2

þ X1 C2 iR1 ðtank1Þ þ X1 C3 iR1 ðtank2Þ X1 C3 iRload þ X1 iRs X1
1
iRs ¼ 0
ð2 þ CCcs Þ
iRp1 k X2 iRp1 þ X2 iRs1 ¼ 0; iL1 k þ y1 ¼ 0

iRs1 k X6 iRs1 þ X3 iRs X4 iL1 X5 iL2 X5 iRs2 þ X7

iR1 ðtank1Þ þ iR1 ðtank2Þ X5 X5 iRload þ X8 iRp1 ¼ 0
iRp2 k X9 iRp2 þ X9 iRs2 ¼ 0
iRs2 k X13 iRs2 þ X10 iRs X11 iL1 X12 iL2 X11 iRs1 þ X11 iR1 ðtank1Þ þ X14 iR1 ðtank2Þ
X12 iRload þ X15 iRp2 ¼ 0
Rload
iL2 k þ iRload ¼0
L2
iRload k X17 iRload þ X16 iRs þ iR1 ðtank1Þ X16 þ iR1 ðtank2Þ X17 X16 iL1 X17
iL2 X16 iRs1 X17 iRs2
¼0
0 1
y1
B C
B iRp1 C
B C
B iL1 C
0 1 B C 0 1 0 1
N11 ... N18 B C i11 ... i13 iRs
B iRs1 C
B .. .. .. C B C B .. .. .. C B i C
@ . . . A B Cþ@ . . . A @ R1 ðtank1Þ A ¼ 0; N11
B iRp2 C
N81 N88 B C i81 i83 iR1 ðtank2Þ
B C
B iRs2 C
B C
B iL2 C
@ A
iRload
¼ k; N12 ¼ 0; N13 ¼ X1 C2
N14 ¼ X1 C2 ; N15 ¼ 0; N16 ¼ X1 C3 ; N17 ¼ X1 C3 ; N18 ¼ X1 C3 ; N21 ¼ 0
N22 ¼ k X2 ; N23 ¼ 0; N24 ¼ X2 ; N25 ¼ N26 ¼ N27 ¼ N28 ¼ 0; N31 ¼ 1; N32 ¼ 0
N33 ¼ k; N34 ¼ . . .N38 ¼ 0; N41 ¼ 0; N42 ¼ X8 ; N43 ¼ X4 ; N44 ¼ k X6
N45 ¼ 0; N46 ¼ X5 ; N47 ¼ X5 ; N48 ¼ X5 ; N51 ¼ N52 ¼ N53 ¼ N54 ¼ 0; N55
¼ k X9
N56 ¼ X9 ; N57 ¼ N58 ¼ 0; N61 ¼ 0; N62 ¼ 0; N63 ¼ X11 ; N64 ¼ X11 ; N65 ¼ X15
N66 ¼ k X13 ; N67 ¼ X12 ; N68 ¼ X12 ; N71 ¼ . . . ¼ N76 ¼ 0; N77 ¼ k; N78
Rload
¼
L2
N81 ¼ N82 ¼ 0; N83 ¼ X16 ; N84 ¼ X16 ; N85 ¼ 0; N86 ¼ X17 ; N87 ¼ X17 ; N88
¼ k X17
1
i11 ¼ X1 ½1 ; i12 ¼ X1 C2 ; i13 ¼ X1 C3 ; i21 ¼ i22 ¼ i23 ¼ 0; i31
ð2 þ CCcs Þ
¼ i32 ¼ i33 ¼0
i41 ¼ X3 ; i42 ¼ X7 ; i43 ¼ X5 ; i51 ¼ i52 ¼ i53 ¼ 0; i61 ¼ X10 ; i62 ¼ X11 ; i63 ¼ X14
i71 ¼ i72 ¼ i73 ¼ 0; i81 ¼ X16 ; i82 ¼ X16 ; i83 ¼ X17
0 1 0 1
i11 ... i13 iRs
B .. .. .. C B i C
Assumption @ . . . A @ R1 ðtank1Þ A ! e
i81 i83 iR1 ðtank2Þ
0 1 0 1
N11 . . . N18 N11 . . . N18
B . .. .. C B .. .. .. C
ðA k IÞ ¼ @ .. . . A; detðA k IÞ ¼ det@ . . . A¼0
N81 N88 N81 N88
X
8 X
8
detðA k IÞ ¼ !k kk ; !k kk ¼ 0:
k¼0 k¼0
Remark It is reader exercise to find the expressions for !k 8 k ¼ 0; 1; . . .; 8:

Eigenvalues stability discussion: Our BPF (varactor diodes) circuit involving N
variables (N > 2, N = 11, arbitrarily small increments), the characteristic equation
is of degree N = 8 (reduced) and must often be solved numerically. Expect in some
particular cases, such an equation has (N = 8) distinct roots that can be real or
complex. These values are the eigenvalues of the (10 10) Jacobian matrix (A).
The general rule is that the BPF (varactor diodes) circuit is stable if there is no
eigenvalue with positive real part. It is sufficient that one eigenvalue is positive for
the steady state to be unstable. Our 8-variables (y1 iRp1 iL1 iRs1 iRp2 iRs2 iL2 iRload )
system has eight eigenvalues (eight system’s arbitrarily small increments). The type
of behavior can be characterized as a function of the position of these eigenvalues in
the Re/Im plane. Five non-degenerated cases can be distinguished: (1) the eight
eigenvalues are real and negative (stable steady state), (2) the eight eigenvalues are
real, at least one of them is positive (unstable steady state), (3) and (4) two
eigenvalues are complex conjugates with a negative real part and other eigenvalues
real are negative (stable steady state), two cases can be distinguished depending on
the relative value of the real part of the complex eigenvalues and of the real one,
(5) two eigenvalues are complex conjugates with a negative real part and at least
one of the other eigenvalues real is positive (unstable steady state) [12, 13].
X
8 X
8
k¼0 k¼0
Exercises
1. We have triplexer circuit. It is a four ports device with a common input and three
outputs. A triplexer is a passive device that implements frequency domain
multiplexing. Three ports (e.g., L, I, H) are multiplexed onto a fourth port (e.g.,
S). The signals on L, I, and H occupy disjoint frequency bands. Consequently,
the signals on L, I, and H can coexist on port S without interfering with each
other. The signal on port L (low band) will occupy a single low frequency band,
the signal on port I (intermediate band) will occupy a single intermediate fre-
quency band and the signal on port H will occupy a higher frequency band. In
that situation, the triplexer consists of a low pass filter connecting ports L and S,
and an intermediate pass filter connecting ports I and S, and a high pass filter
connecting ports H and S. The triplexer, being a passive device, is reciprocal;
the device itself doesn’t have a nothing of input or output. The ports of a
diplexer are frequency selective. A triplexer multiplexes three ports onto one
port. A triplexer can be placed at the output of a frequency RF source, where it
functions as an absorptive filter.
Exercises 769
Filter 2 (I) of the triplexer has a BPF that corresponds with the undesired
frequency band, which can pass right through with little attenuation, and is ter-
minated within to 50 X load. These undesired frequencies are blocked from
entering filter 2 (I) by that filter’s stop band. Filters 1(L) and 3(H) BPFs passes all
the desired signals onto load 1 or load 2 through matching networks (1 and 2) with
little attenuation. The undesired signals through the triplexer are absorbed instead of
being reflected as they would be in a typical filter. The design of a triplexer is as
three different frequency filters with non-overlapping bandpass. The matching
network between triplexer circuit unit and loads (Rload1 and Rload2) can be L-type,
Pi-type, and T-type.
Consider in our RF system which diplexer is an integrated unit, micro strip lines
between filter 1(L) and filter 3(H) to the matching networks. The parasitic effects of
micro strip line between filter 2(I) and 50 X resistor are neglected. We represent
these micro strip lines parasitic effects as a delay lines in time. The delays are
related to the current which flows through micro strips lines and are define as
sout1 ðTau-out1 Þ and sout2 ðTau-out2 Þ respectively. We neglect the triplexer input
micro strip line parasitic effects. s [ 0; s 2 R
1:1 Write circuit differential equations and find fixed points (matching network 1:
T type and matching network 2: L type). Assumption: dVdts ðtÞ ! e.
1:2 Discuss stability of our circuit for the following cases: (a) sout1 ¼ s; sout2 ¼ 0(b)
sout1 ¼ 0; sout2 ¼ s(c) sout1 ¼ C sout2 ; sout2 ¼ s; C 2 R þ How our circuit
behavior changes for different values of C and s Parameters?
1:3 How our circuit dynamic is changed if we disconnect inductor L2? Write
circuit differential equations and find fixed points. Discuss stability and sta-
bility switching.
1:4 We choose both circuits’ matching networks Pi type. Assumption: dVdts ðtÞ ! e .
Write circuit differential equations and find fixed points. Discuss stability and
stability switching.
1:5 We add additional filter to our triplexer (inductor L4 and capacitor C4 in series)
which terminated by 50 X resistor. Our circuit is Quad-plexer. Write circuit
differential equations and find fixed points (matching networks are L type).
2. We have a system of two diplexers in series. Diplexer (I): Two band pass filters
are duplexed, one common input, and two outputs. A diplexer (I) is referring to
a duplex high pass and low pass where broad bands transmit and receive is
necessary. It is a three ports device with common input and two outputs. It
implements frequency domain multiplexing two ports (e.g., LI and HI) are
multiplexed onto a third port (e.g., SI). The signal on port LI and HI occupy
disjoint frequency bands. Consequently, the signals on LI and HI can coexist on
port SI without interfering with each other. The signal on port LI will occupy a
single low frequency band and the signal on port HI will occupy a higher
frequency band. Diplexer (II): Two band pass filters are duplexed, one common
input, and two outputs. A diplexer (II) is referring to a duplex high pass and low
pass where broad bands transmit and receive is necessary. It is a three ports
device with common input and two outputs. It implements frequency domain
multiplexing two ports (e.g., LII and HII) are multiplexed onto a third port (e.g.,
SII). The signal on port LII and HII occupy disjoint frequency bands.
Consequently, the signals on LII and HII can coexist on port SII without inter-
fering with each other. The signal on port LII will occupy a single low frequency
band and the signal on port HII will occupy a higher frequency band. There is an
overlap between diplexer (I) port HI frequency band and diplexer (II) port HII
frequency band.
Diplexer (I)
LI HI
f(Hz)
Diplexer (II)
LII HII
f(Hz)
Exercises 771
Filter (II) of diplexer (I)—low band and filter (IV) of diplexer (II)—low band
have a BPFs that correspond with the undesired frequency bands, and are termi-
nated within the 50 X load. These undesired frequencies are blocked from entering
filters II and IV by that filter’s stop band. Diplexers (I) and (II) are connected in
series by micro strip lines and matching networks (I). The output load (Rload) is
connected to diplexer II’s output through matching network (II).
We represent these micro strips parasitic effects as a delay lines in time. The
delays are related to the current which flows through micro strips and are
sin ðTau-inÞ and sout ðTau-outÞ respectively. We choose T type matching network
(I) and L type matching network (II). We neglect in our analysis other circuit micro
strips parasitic effects and consider them as an ideal elements. s [ 0; s 2 R.
dVs ðtÞ
Assumption dt ! e.
2:2 Discuss stability of our circuit for the following cases: (a) sin ¼ s; sout ¼ 0,
pffiffiffi
(b) sin ¼ s; sout ¼ s2 , (c) sin ¼ s2 ; sout ¼ s s. How our circuit behavior
changes for different values of s parameter? Discuss stability switching for
different values of s parameter.
2:3 We disconnect diplexer (I)’s filter (II). How it influences our circuit behavior?
stability switching for the following cases: (a) sin ¼ s; sout ¼ 0,
(b) sin ¼ 0; sout ¼ s, (c) sin ¼ sout ¼ s.
2:4 We disconnect diplexer (II)’s filter (IV). How it influences our circuit
behavior? Write circuit differential equations and find fixed points. Discuss
pffiffiffi
stability and stability switching for the following cases: (a) sin ¼ s; sout ¼ 0,
p ffiffi
ffi
(b) sin ¼ 0; sout ¼ s, (c) sin ¼ sout ¼ s 3 s.
2:5 We disconnect inductor L3. How it influences our circuit dynamics? Write
circuit differential equations and find fixed points. Discuss stability and sta-
pffiffiffi
bility switching for the following cases: (a) sin ¼ s s; sout ¼ 0,
pffiffiffi p ffiffiffiffiffiffiffiffiffiffiffiffi
pffiffiffi
(b) sin ¼ 0; sout ¼ s, (c) sin ¼ sout ¼ s 3 s.
3. We have a system of two diplexers (I and II), two RF/Microwave sources, and
summation operator. The summation operator is implemented by using ideal op
amps (summing amplifier follows by inverting amplifier).
R
R1 ¼ R2 ¼ Rf1 ; Rfin2 ¼ 1; VC ¼ VA þ VB ; R2 Rs2 ; R2 þ Rs2 R2
Diplexer (I): Two band pass filters are duplexed, one common input, and two
outputs. A diplexer (I) is referring to a duplex high pass and low pass where broad
bands transmit and receive is necessary. It is a three ports device with common
input and two outputs. It implements frequency domain multiplexing two ports
(e.g., LI and HI) are multiplexed onto a third port (e.g., SI). The signal on port LI
and HI occupy disjoint frequency bands. Consequently, the signals on LI and HI
can coexist on port SI without interfering with each other. The signal on port LI will
occupy a single low frequency band and the signal on port HI will occupy a higher
frequency band. Diplexer (II): Two band pass filters are duplexed, one common
input, and two outputs. A diplexer (II) is referring to a duplex high pass and low
pass where broad bands transmit and receive is necessary. It is a three ports device
with common input and two outputs. It implements frequency domain multiplexing
two ports (e.g., LII and HII) are multiplexed onto a third port (e.g., SII). The signal
on port LII and HII occupy disjoint frequency bands. Consequently, the signals on
LII and HII can coexist on port SII without interfering with each other. The signal
on port LII will occupy a single low frequency band and the signal on port HII will
occupy a higher frequency band. There is an overlap between diplexer (I) port HI
frequency band and diplexer (II) port HII frequency band. The output of diplexer
Exercises 773
(I)’s filter (I) is connected to summation operator (port A). RF/Microwave source
Vs2(t) is connected through resistor Rs2 to summation operator (port B). Load
resistor Rload is connected to diplexer (II)’s filter (IV) through T type matching
network.
Filter (II) of diplexer (I)—low band and filter (III) of diplexer (II)—high band
have a BPFs that correspond with the undesired frequency bands, and are termi-
nated within the 50 X load. These undesired frequencies are blocked from entering
filters II and III by that filter’s stop band. Diplexers (I) and (II) are connected to
summation operator by micro strip lines (sin ðTau-inÞ; sout ðTau-outÞ) .Matching
networks (I) is connected to diplexer (II)’s filter (IV) by micro strip line
(sm ðTau-mÞ).
Assumption There is an overlaps between all circuit’s filters frequency bands, the
output signal from diplexer (II)’s filter (IV) is within specific frequency band and
target for our load resistance (Rload).
dV ðtÞ dV ðtÞ
Assumption dt s1
! e; dt s2
! e. All other micro strip lines in our system are
ideal with no parasitic effects in our circuit. s [ 0; s 2 R

3:2 Discuss stability of our circuit for the following cases: sm ¼ s
pffiffiffiffiffi pffiffiffi pffiffiffi
(a) sin ¼ s2 ; sout ¼ 0, (b) sin ¼ s; sout ¼ 3 s2 , (c) sin ¼ s2 ; sout ¼ 3 s s.
How our circuit behavior changes for different values of s parameter? Discuss
3:3 We disconnect diplexer (I)’s filter (II). How it influences our circuit behavior?
stability switching for the following cases: sm ¼ s (a) sin ¼ s; sout ¼ 0,
(b) sin ¼ 0; sout ¼ s, (c) sin ¼ sout ¼ s.
3:4 We disconnect diplexer (II)’s filter (III). How it influences our circuit
behavior? Write circuit differential equations and find fixed points. Discuss
pffiffiffi
stability and stability switching for the following cases: sm ¼ s
pffiffiffi pffiffiffi pffiffiffi
(a) sin ¼ s s; sout ¼ 0, (b) sin ¼ 0; sout ¼ 4s, (c) sin ¼ sout ¼ s 3 s.
3:5 We disconnect diplexer (I) from summation operator. How it influences our
circuit dynamics? Write circuit differential equations and find fixed points.
pffiffiffi
Discuss stability and stability switching for the following cases: sm ¼ s s
pffiffiffi pffiffiffi pffiffiffiffiffiffiffiffiffiffiffiffi
p ffiffiffi
(a) sin ¼ s s; sout ¼ 0, (b) sin ¼ 0; sout ¼ s, (c) sin ¼ sout ¼ s 3 s.
4. We have a system of triple band triplexer filter. The circuit of triple band
triplexer contains capacitors and inductors, RF source Vs(t) and series resistor
Rs, and resistive loads Rload1, Rload2 and Rload3. Three delays lines sout1 ; sout2 and
sout3 represent circuit micro strip lines parasitic effects. In our analysis we
consider full matching between Rload1 and BPF1, Rload2 and BPF2, Rload3 and
BPF3 (no matching networks). We consider all other micro strip lines in our
circuit are an ideal micro strips lines.
Load resistors Rload1 and Rload2 are connected to BPF1 and BPF2 through
inductors bridge (LA ; LB ; LC ; LD ; LA 6¼ LB 6¼ LC 6¼ LD ).
Exercises 775
4:1 Write circuit differential equations and find fixed points,
dVs ðtÞ
Assumption dt ! e.
4:2 Discuss stability of our circuit for the following cases:
pffiffiffi
(a) sout1 ¼ s; sout2 ¼ s; sout3 ¼ 0
pffiffiffi
(b) sout1 ¼ 0; sout2 ¼ s2 ; sout3 ¼ s s
p ffiffi
ffi
(c) sout1 ¼ s2 ; sout2 ¼ s; sout3 ¼ s:
4:3 We disconnect capacitor C2. How it influences our circuit dynamics? Write
circuit differential equations and find fixed points. Discuss stability of our
circuit for the following cases:
P
2
(a) sout1 ¼ s C1 ; sout2 ¼ s2out1 ; sout3 ¼ soutk
k¼1
pffiffiffi P2
(b) sout1 ¼ s2 C2 ; sout2 ¼ s; sout3 ¼ s2outk
k¼1
C1 ; C2 2 R þ ; C1 6¼ C2 ; s [ 0; s 2 R. Discuss stability switching for different

values of C1 ; C2 ; s parameters.
4:4 Return (4.3) if we disconnect inductors L2 and L4.
4:5 We disconnect Inductors Bridge’s LB, how it influences our circuit dynamics?
Write circuit differential equations and find fixed points. Discuss stability of
our circuit for the following cases:
pffiffiffi P
2
(a) sout1 ¼ s C1 ; sout2 ¼ s2out1 ; sout3 ¼ Ck soutk
k¼1
pffiffiffiffiffi pffiffiffi P
2
(b) sout1 ¼ 3 s2 C2 ; sout2 ¼ s s; sout3 ¼ Ck s2outk
k¼1

5. We have a system of triple band triplexer filter. The circuit of triple band
triplexer contains capacitors and inductors, RF source Vs(t) and series resistor
Rs, and resistive loads Rload1, Rload2. Three delays lines sout1 ; sout2 and sout3
represent circuit micro strip lines parasitic effects. In our analysis we consider
full matching between Rload1 and BPF1, Rload2 and BPF2 (no matching net-
works). We consider all other micro strip lines in our circuit are an ideal micro
strips lines. RF signals from BPF2 and BPF3 are added by using summation
operator and feed to capacitors and inductors bridge. Load resistors Rload1 and
Rload2 are connected to BPF1, BPF2, and BPF3 through inductors and
capacitors bridge (LA ; LB ; CA ; CB ; LA 6¼ LB ; CA 6¼ CB ). The summation oper-
ator is implemented by using ideal op amps (summing amplifier follows by
R
inverting amplifier). R1 ¼ R2 ¼ Rf1 ; Rfin2 ¼ 1; VC ¼ VA þ VB ; R2
Rs2 ; R2 þ Rs2 R2 (see question 3).
5:1 Write circuit differential equations and find fixed points, Assumption:
dVs ðtÞ
dt ! e.
5:2 Discuss stability of our circuit for the following cases:
pffiffiffi pffiffiffi
(a) sout1 ¼ 3 s; sout2 ¼ s; sout3 ¼ 0
p ffiffiffiffiffi pffiffiffi
(b) sout1 ¼ 0; sout2 ¼ s2 ; sout3 ¼ s2 s
3
p ffiffiffiffi
ffi pffiffiffi
(c) sout1 ¼ s2 ; sout2 ¼ s; sout3 ¼ s s:
3
5:3 We disconnect capacitor C5. How it influences our circuit dynamics? Write
circuit differential equations and find fixed points. Discuss stability of our
circuit for the following cases:
qffiffiffiffiffiffiffiffiffi P
2
(a) sout1 ¼ s C1 ; sout2 ¼ 3 s2out1 ; sout3 ¼ C21 soutk
k¼1
pffiffiffiffiffiffiffiffiffiffiffiffiffi pffiffiffi P
2
(b) sout1 ¼ 3 s2 C2 ; sout2 ¼ s s; sout3 ¼ s2outk
k¼1

5:4 Return (5.3) if we disconnect inductors L2 and L9.
Exercises 777
5:5 We disconnect Inductors Bridge’s CB, how it influences our circuit dynamics?
Write circuit differential equations and find fixed points. Discuss stability of
our circuit for the following cases:
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
pffiffiffi P
2
(a) sout1 ¼ s C1 ; sout2 ¼ s2out1 ; sout3 ¼ Ck soutk
k¼1
p ffiffiffiffiffi pffiffiffiffiffiffiffiffiffiffiffiffi
pffiffiffi P
2
(b) sout1 ¼ s2 C2 ; sout2 ¼ s s; sout3 ¼ Ck s2outk
3
k¼1

6. We have crystal-lattice BPF circuit with two output loads resistances Rload1 and
Rload2. Crystals have series and parallel resonant mode. The crystal lattice filter
contains several crystals within a single circuit. The input employs RF trans-
former (T1) with shunt capacitor (C1). The output employs five terminals
transformer (T2) with shunt capacitor (C2). Each set of crystals XTAL1 plus
XTAL2 and XTAL3 plus XTAL4 are cut to different frequencies. The matched
set of XTAL1 and XTAL2 having a lower resonant frequency than the other
matched set of XTAL3 and XTAL4. Transformer T1: Parameters np1 and ns1 are
the number of turns of coil 1 (primary) and 2 (secondary), respectively.
Transformer T2: Parameter np2 is the number of turns of coil 1 (primary).
Parameters ns2a and ns2b are the number of turns of coil 2 (secondary). In your
analysis use the equivalent circuit of a quartz crystal and use typical crystal
parameters.
6:1 Write circuit differential equations and find fixed points. Find arbitrarily small
increments equations.
6:2 Find circuit eigenvalues and discuss stability.
6:3 Capacitor C2 is disconnected. How it influences our circuit dynamics? Write

circuit differential equations and find fixed points. Discuss stability.
6:4 Load resistance Rload2 is disconnected. How it influences our circuit dynamic?
Write circuit differential equations and find fixed points. Discuss stability.
6:5 We multiple the value of capacitor C1 by C parameter (C1 ! C1 C),
C [ 0; C 2 R þ . How it influences our circuit dynamics. Discuss stability and
stability switching for different values of C parameter.
7. We have Half-lattice crystal filter circuit with two crystals and one load Rload.
Crystals have a series and parallel resonant mode. The Half lattice crystal filter
contains two crystals within a single circuit. The input employs RF transformer
(T1) with shunt capacitor (C1). The set of crystals XTAL1 and XTAL2 are cut to
different frequencies. Half-lattice crystal filter offers a flatter in band response.
The two crystals have different resonant frequencies. The response has a small
peak at either side of the center frequency and a small dip in the middle.
Transformer T1: np is the number of turns of coil 1 (primary). Nsa and nsb are
number of turns of coil 2 (secondary) and coil 3 (secondary). In our analysis, we
use the equivalent circuit of a quartz crystal and use typical crystal parameters.
increments equations.
7:3 Capacitor C1 is disconnected. How it influences our circuit dynamics? Write
7:4 We increase the number of coil 3 turns by C factor (nsb ! nsb C),
C [ 0; C 2 R þ . How it influences our circuit dynamics? Discuss stability
switching for different values of C parameter.
7:5 We short crystal XTAL2. How it influences circuit behavior? Write circuit
differential equation and find fixed points. Discuss stability.
Exercises 779
8. We have Half-lattice crystal filter circuit with two crystals and one load Rload.
Crystals have a series and parallel resonant mode. The Half lattice crystal filter
contains two crystals within a single circuit. The input employs RF trans-
former (T1) with shunt capacitor (C1). The RF transformer (T1)’s secondary is
connected to two capacitors (Ca and Cb). The set of crystals XTAL1 and
XTAL2 are cut to different frequencies. Half-lattice crystal filter offers a flatter
in band response. The two crystals have different resonant frequencies. The
response has a small peak at either side of the center frequency and a small dip
in the middle. Transformer T1: np is the number of turns of coil 1 (primary). ns
is number of turns of coil 2 (secondary). In our analysis, we use the equivalent
circuit of a quartz crystal and use typical crystal parameters.
increments Eqs.
8:3 Capacitor Cb is disconnected. How it influences our circuit dynamics? Write
8:4 We increase the number of coil 1 turns by C factor (np ! np C),
8:5 We short crystal XTAL1. How it influences circuit behavior? Write circuit
differential equation and find fixed points. Discuss stability.
9. We have circuit of a single crystal filter. It employs the very high Q of the
crystal. Its response is asymmetric and it is too narrow for most applications,
having a bandwidth of a hundred Hz or less. In the circuit there is a variable
capacitor (Cx) that is used to compensate for the parasitic capacitance in the
crystal. This capacitor was normally included as a front panel control. The input
employs RF transformer (T1) with shunt capacitor (C1). The RF transformer
(T1)’s secondary is connected to two capacitors (Ca and Cb). Transformer T1: np
is the number of turns of coil 1 (primary). ns is number of turns of coil 2
(secondary). In our analysis, we use the equivalent circuit of a quartz crystal and
use typical crystal parameters. The crystals XTAL1 is cut to different frequency.
increments equations
9:3 Capacitor Ca is disconnected. How it influences our circuit dynamics? Write
9:4 We increase the number of coil 1 turns by C factor (np ! np C),
9:5 We increase the Cx capacitor value by w parameter (Cx ! Cx w),
w [ 0; w 2 R þ . Write circuit differential equations and find fixed points.
Discuss stability and stability switching for different values of w parameter.
10. We have circuit of top inductively coupled variable BPF. For wideband
applications the top inductively coupled variable band pass filter is the best. The
circuit includes varactor diodes (Cv) and inductors L1, …, L5, bias resistors R1
and R2 isolated the two varactors (Cv) from the effects of each other, and
resistor R3 represents a direct RF short to ground through Vtune. Capacitors C2
and C3 block the DC inserted by Vtune from being shorted by L2 and L4.
Varactor diodes (Cv) supply the variable tuning capacitance. The capacitance of
the series combination of C2 and Cv in series is CC2 2þCCv v . The capacitance of the
series combination of C3 and Cv in series is CC3 3þCCv v . The capacitors C2 and C3
mainly functioning as a DC blocking capacitor, while Cv, the varactor is
Exercises 781
supplying all of the tuning capacitance for the filter tanks. By applying positive
Vtune we allow the varactor to either linearly tune the filter to its maximum and
minimum values or, by supplying Vtune with discrete voltages it filter the
incoming RF/Microwave signal in discrete steps. The circuit load resistance
Rload is connected in our voltage tunable inductor coupled bandpass filter. Input
RF/microwave signal voltage Vs(t) and his series resistance Rs.
Remark A varactor diode is a P-N junction diode that changes its capacitance and
the series resistance as the bias applied to the diode is varied. The property of
capacitance change is utilized to achieve a change in the frequency and/or phase of
our BPF. In your stability and circuit analysis use the simple model of a packaged
varactor diode. For normal operation, a varactor diode is always reverse biased.
Varactor diode is called voltage controlled capacitor.
10:2 Find arbitrarily small increments equations and discuss stability.
10:3 Resistor R1 is disconnected. How it influences our circuit dynamics? Write
10:4 We increase the value of L3 inductor by C parameter (L3 ! L3 C),
C [ 0; C 2 R þ . Write circuit differential equations and find fixed points.
Discuss stability and stability switching for different values of C parameter.
10:5 Inductor L4 is disconnected. How it influences our circuit dynamics? Find
circuit differential equations and fixed points. Discuss stability.
Chapter 8
Antennas System Stability Analysis
An antenna is a conductor or group of conductors used for radiating electromag-

netic energy into space or collecting electromagnetic energy from space. The radio
signal is generated in a transmitter and radiates through space to a receiver by
antenna. The transmitter signal energy is sent into space by a transmitting antenna
and the radio frequency energy is picked up from space by the receiving antenna.
As the electromagnetic field arrives at the receiving antenna, a voltage is induced
into the antenna and passed into the receiver. There are many types of antennas and
we discuss those antennas that operate at microwave frequencies. Microwave refer
to radio waves with wavelength ranging from as long as 1 m to as short as 1 mm
with frequencies between 300 MHz to 300 GHz. Microwave antenna is used for
radiating microwave signal into space and receiving microwave signal from space.
Microwave antenna is the transition region between free space and guiding struc-
ture. Antenna requirements include gain, receiving area, beam width, polarization,
bandwidth, and side lobes. The major function of the antenna used at the receiver
end of a RF/Microwave system is to collect as much of this transmitted power as
possible. It is important in many applications to make the receiving area or aperture
of the antenna as large as possible. Another antennas area is for RFID applications.
Complete RFID system includes RFID reader and transponder units. Electrical
current flowing through a conductor generates electromagnetic fields. We distin-
guish two fields regions related to RFID applications. The first region is far-field
region. In this region, the generated fields are radiated fields which energy prop-
agates through the space with an energy density proportional to the inverse of the
distance. The second region is the near-field region. In this region, radiated fields
are not prevalent. Instead, attenuating fields, in which the strength of the field
decreases with (1/r3) are dominant. Furthermore, the power in this region is reac-
tive. The boundaries between far-field and near-field regions depend on the rela-
tionship between the physical dimensions of the antenna and the wavelength of the

DOI 10.1007/978-3-319-45427-6_8
784 8 Antennas System Stability Analysis
propagating signal. In particular, the dimensions of the antenna should be com-

patible to the wavelength of the signal in order to achieve optimal performance. The
wavelength of signals operating in the LF (Low Frequency) region is 2.4 km, while
the wavelength for signals operating in the HF (High Frequency) region is 24 m.
Therefore, at these frequencies it would not be practical to build antennas with
dimensions similar to the wavelength of their signals. Any antennas that can be
built in a practical manner for RFID transponders operating in the LF or HF ranges
will be electrically small and therefore highly inefficient. RFID transponders
operating in the LF or HF frequencies cannot use dipole antennas because of the
mismatch in dimensions. The solution is to use a small loop antenna instead of a
dipole. A small loop antenna is a closed loop with a maximum dimension that is
less than about a tenth of the wavelength of the signal. The small loop antenna is the
dual equivalent of an ideal dipole and it is suitable for antennas incorporated in
transponders operating at LF or HF frequencies. A conductor of infinite length
carrying a magnitude of current of (I) amps. The magnetic field (B/) measured at a
distance of (r) meters from the conductor can be found using Ampere’s law as
l0 I Wb
B/ ¼ 2pr ð m2 Þ, where l0 is the permeability of the free space
l0 ¼ 4 p 107 m H . A conductor of infinite length is not realistic. Its practical
implementation is based on building a loop antenna by bending the original, finite
wire, which carries a current of (I) amps in a circle with a radius of (a) meters. In
practice, the wire is bent in such a way that produces a total of N turns as this allows
using a longer wire with a relatively small diameter. In this situation, the value of
the magnetic field in the (z) coordinate direction (Bz) for a point located at a
distance of (r) meters from the plane of the coil and located along the axis of the coil
can be found as Bz ¼ l02INa2 3=2 ðWb
2
2ða þ r Þ m2 Þ, where (a) is the radius of the loop in meters.

The other kind of N turns antenna is rectangular spiral antenna. Both antennas are
sensitive to their parameters variation and stability need to be investigated.
Additionally, micro strips lines in RFID system cause to system’s parasitic delays
and influence stability. Special antenna to many RF and microwave applications is
N-turn multilayer circular coil and there is an expression which define its induc-
tance as a function of overall parameters [7, 8]. The stability is inspected for
parameters variations and optimization under delayed electromagnetic interfer-
ences. Some antennas systems are straight thin film inductors antennas structure
(single turn square planar straight thin film inductors antenna system) and its sta-
bility is inspected for many RF applications. Helix (Helical) antenna is consisting of
a conducting wire wound in the form of a helix. Helical antennas are mounted over
a ground plane. The feed line is connected between the bottom of the helix and the
ground plane. Helical antennas can operate in one of two modes, normal mode or
axial mode. In each operation mode we can represent helical antenna as equivalent
circuit and inspect stability for parameters variations.
8.1 N-Turn Multilayer Circular Coil Antennas Transceiver System Stability … 785
8.1 N-Turn Multilayer Circular Coil Antennas

Transceiver System Stability Optimization Under
Delayed Electromagnetic Interferences
N-turn multilayer circular coil antennas can be integrated with RFID IC for com-
plete RFID tags. We investigate the system stability optimization under delayed
electromagnetic interference and parasitic effects. An N-turn multilayer circular coil
antenna is constructed from N-turn of circular coil with multilayer. Our system is
constructed from two antenna; each one, N-turn multilayer circular coil antenna.
Antennas are connected in series with micro strip line and to the RFID IC. An
N-turn multilayer circular coil antennas system is influenced by electromagnetic
interference, which effect their stability behavior. Additionally, micro strip line
which connected each antenna in series, has a parasitic effect, a delay in time Dl.
We inspect our system performances under electromagnetic interferences and micro
strip parasitic effects. Generally, N-turn multilayer circular coil antennas system is
good for many RF and microwave applications. The micro strip line feed technique
enhances the bandwidth of the simple micro strip antenna. Every N-turn multilayer
circular coil antenna has a parasitic DC resistance which needs to be calculated.
Index (i) indicates the first N-turn multilayer circular coil antenna (i = 1) or second
N-turn multilayer circular coil antenna (i = 2). We define RFID’s N-turn multilayer
coil antenna parameters, ai—Average radius of the coil in cm, Ni—number of turns,
bi—winding thickness in cm, Si—wire cross section area, mi—radius of the wire
and hi—winding height in cm. Integrating all those parameters gives the equations
for N-turn multilayer circular coil antenna inductance calculation [85].
0:31 ðai Ni Þ2
Lcalci ¼ ðlH):
6 ai þ 9 hi þ 10 bi
The N turn multilayer circular coil antenna length is calculated as follows: li is the
length of one turn li ¼ 2 p ai . lN is the length of N turn
lNi ¼ Ni li ¼ 2 p ai Ni . Assumption: ai bi ; ai þ bi ai (Fig. 8.1).
We consider system’s two N-turn multilayer circular coil antennas are not
identical (a1 6¼ a2 ; N1 6¼ N2 ; h1 6¼ h2 ; b1 6¼ b2 ; ai ; Ni ; hi ; bi 2 R þ ). The DC resis-
tance of the N-turn multilayer circular coil antenna:
l Ni 2 p ai Ni 2 ai Ni
RDCi ¼ ¼ ¼ :
r i Si r i Si ri m2i
lNi —total length of the wire, ri—conductivity of the wire (S/m), Si—wire cross
section area (p m2i ), mi—radius of the wire. Due to electromagnetic interference
there are differences in time delays with respect to the first (i = 1) and second (i = 2)
N-turn multilayer circular coil antenna voltages and voltages derivatives.
The delayed voltages are V1 ðt s1 Þ and V2 ðt s2 Þ respectively (s1 6¼ s2 ) and
delayed voltages derivatives are dV1 ðtD dt
1 Þ dV2 ðtD2 Þ
; dt respectively. Assumption:
Fig. 8.1 N-turn multilayer circular coil antennas transceiver system
D1 6¼ D2 ; s1 [ 0; s2 [ 0; D1 ; D2 0. Additionally, there is a delay in time for the

micro strip parasitic effects Dl . The stability of a given steady state is simply
determined by the graphs of some function of s1 ; s2 which can be expressed,
explicitly and thus can be easily depicted by MATLAB and other popular software.
We need only look at one such function and locate the zeros. At time delay
increases, stability changes from stable to unstable to stable. The analytical criteria
provided for the first and second order cases can be used to obtain some insightful
analytical statements and can be helpful for conducting simulations. N-turn mul-
tilayer circular coil antennas transceiver (RFID system) system can be represented
as two inductors (Lcalc1 and Lcalc2 ), parasitic resistances (RDC1 and RDC2 ) and
microstrip delay line. The N-turn multilayer circular coil antennas in series are
connected in parallel to RFID IC. The equivalent circuit of N-turn multilayer cir-
cular coil antennas transceiver (RFID system) is capacitor (C1) and resistor (R1) in
parallel with N-turn multilayer circular coil antennas in series. Element 2 Lm
represents the mutual inductance between Lcalc1 and Lcalc2 . Since two inductors
(Lcalc1 and Lcalc2 ) are in series and there is a mutual inductance between Lcalc1
and Lcalc2 , the total antenna inductance LT: LT ¼ Lcalc1 þ Lcalc2 þ 2 Lm and
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
Q2
Lm ¼ K i¼1 Lcalci , Lm is the mutual inductance between Lcalc1 and Lcalc2 , K
is the coupling coefficient of two inductors 0 K 1. Variable I(t) is the current
that flows through a N-turn multilayer circular coil antennas transceiver system for
Dl ! e. The V1 and V2 are the voltages on Lcalc1 and Lcalc2 respectively, Vm is
the voltage on N-turn multilayer circular coil antennas mutual inductance element.
We neglect the voltage on microstrip delay line Vl ! e. The delay which is related
to microstrip element is on current ILcalc2 ðtÞ ¼ ILcalc1 ðt Dl Þ and
IRDCi ¼ ILcalci ; i ¼ 1; 2. N-turn multilayer circular coil antennas RFID system
equivalent circuit is present in the next figure (Fig. 8.2).
Fig. 8.2 N-turn multilayer

circular coil antennas
equivalent circuit
ILcalc2 ðtÞ ¼ Iðt Dl Þ; IRDC2 ¼ Iðt Dl Þ; ILm ¼ Iðt Dl Þ

VAB VA VB dVAB dðVA VB Þ
IR 1 ¼ ¼ ; I C1 ¼ C 1 ¼ C1 ; IðtÞ þ IR1 þ IC1 ¼ 0
R1 R1 dt dt
VAB ¼ VR1 ¼ VC1 ; VAB ¼ VRDC1 þ VLcalc1 þ VRDC2 þ VLcalc2 þ VLm ;
V1 ðtÞ ¼ VLcalc1 ¼ VLcalc1 ðtÞ
Z
1
V2 ðtÞ ¼ VLcalc2 ¼ VLcalc2 ðtÞ; VAB ¼ VA VB ¼ R1 IR1 ; VAB ¼ IC1 dt;
C1
VAB dVAB
IðtÞ þ þ C1 ¼0
R1 dt
VAB ¼ VRDC1 þ VLcalc1 þ VRDC2 þ VLcalc2 þ Vm þ ðVl ! eÞ; Vm ¼ VLm ;
X
2 X
2
VAB ¼ VRDCi þ VLcalci þ Vm þ ðVl ! eÞ
i¼1 i¼1
2 a1 N1
VRDC1 ¼ IðtÞ RDC1 ¼ IðtÞ ;
r1 m21
2 a2 N2
VRDC2 ¼ Iðt Dl Þ RDC2 ¼ Iðt Dl Þ
r2 m22
dIðtÞ 0:31 ða1 N1 Þ2 dIðtÞ
VLcalc1 ¼ Lcalc1 ¼ ;
dt 6 a1 þ 9 h1 þ 10 b1 dt
dIðt Dl Þ 0:31 ða2 N2 Þ2 dIðt Dl Þ
VLcalc2 ¼ Lcalc2 ¼
dt 6 a2 þ 9 h2 þ 10 b2 dt
vffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
u 2
dIðt Dl Þ uY dIðt Dl Þ
Vm ¼ 2 Lm ¼ 2 K ðt Lcalci Þ ; Lcalc1 6¼ Lcalc2
dt i¼1
dt
VC1 dVC1
IðtÞ þ þ C1 ¼ 0;
R1 dt
u 2
X
2 X
2 uY dIðt Dl Þ
V C1 ¼ VRDCi þ VLcalci þ 2 K ðt Lcalci Þ þ ðVl ! eÞ
i¼1 i¼1 i¼1
dt
u 2
dVR1 dVC1 X 2
dVRDCi X 2
dVLcalci uY d 2 Iðt Dl Þ dVl
¼ ¼ þ þ 2 K ðt Lcalci Þ 2
þð ! eÞ
dt dt i¼1
dt i¼1
dt i¼1
dt dt
Z
dIðtÞ 1
VLcalc1 ¼ Lcalc1 ) IðtÞ ¼ VLcalc1 dt;
dt L
Z calc1
6 a1 þ 9 h1 þ 10 b1
IðtÞ ¼ ð Þ VLcalc1 dt
0:31 ða1 N1 Þ2
Z
dIðt Dl Þ 1
VLcalc2 ¼ Lcalc2 ) Iðt Dl Þ ¼ VLcalc2 dt;
dt Lcalc2
Z
6 a2 þ 9 h2 þ 10 b2
Iðt Dl Þ ¼ VLcalc2 dt
0:31 ða2 N2 Þ2
dIðtÞ VLcalc1 dIðt Dl Þ VLcalc2

¼ ; ¼ ;
dt Lcalc1 dt Lcalc2
dIðt Dl Þ dIðtÞ dIðt Dl Þ dIðtÞ
j j X; X ! e;
dt dt dt dt
VLcalc1 VL dIðt Dl Þ VL
calc2 ; Vm ¼ 2 Lm ¼ 2 Lm calc2 ;
Lcalc1 Lcalc2 dt Lcalc2
u 2
uY VL
Vm ¼ 2 K ðt Lcalci Þ calc2
i¼1
Lcalc2
rffiffiffiffiffiffiffiffiffiffiffiffiffi rffiffiffiffiffiffiffiffiffiffiffiffiffi
Lcalc1 dVm Lcalc1 dVLcalc2
Vm ¼ 2 K ð Þ VLcalc2 ; ¼2K ð Þ
Lcalc2 dt Lcalc2 dt
We get the following differential equation respect to VLcalc1 variable:

Z
1 VAB dVAB
VLcalc1 dt þ þ C1 ¼ 0;
Lcalc1 R1 dt
VRDC1 ¼ IðtÞ Rdc1 ; VRDC2 ¼ Iðt Dl Þ Rdc2
Z
1 1 X 2 X2
VLcalc1 dt þ ð VRDCi þ VLcalci þ Vm Þ
Lcalc1 R1 i¼1 i¼1
d X 2 X2
þ C1 ð VRDCi þ VLcalci þ Vm Þ ¼ 0
dt i¼1 i¼1
Z
1 1 X 2 X2
VLcalc1 dt þ ð VRDCi þ VLcalci þ Vm Þ
Lcalc1 R1 i¼1 i¼1
X
2
dVR X
2
dVL dVm
þ C1 ð DCi
þ calci
þ Þ¼0
i¼1
dt i¼1
dt dt
Z X2
1 1
VLcalc1 dt þ fIðtÞ Rdc1 þ Iðt Dl Þ Rdc2 þ VLcalci
Lcalc1 R1 i¼1
rffiffiffiffiffiffiffiffiffiffiffiffiffi
Lcalc1 dIðtÞ dIðt Dl Þ
þ2 K ð Þ VLcalc2 g þ C1 f Rdc1 þ Rdc2
Lcalc2 dt dt
X2
dVLcalci Lcalc1 dVLcalc2
þ þ 2 K ð Þ g¼0
i¼1
dt Lcalc2 dt
dIðtÞ VLcalc1 dIðt Dl Þ VLcalc2

¼ ; ¼
dt Lcalc1 dt Lcalc2
Z Z Z
1 1 1 1
VLcalc1 dt þ
f Rdc1 VLcalc1 dt þ Rdc2 VLcalc2 dt
Lcalc1 R1 Lcalc1 Lcalc2
X 2
Lcalc1 VLcalc1 VL
þ VLcalci þ 2 K ð Þ VLcalc2 g þ C1 f Rdc1 þ calc2 Rdc2
i¼1
L calc2 L calc1 Lcalc2
X 2
dVLcalci Lcalc1 dVLcalc2
þ þ2K ð Þ g¼0
i¼1
dt Lcalc2 dt
We derive in time the two sides of the above equations:
1 1 1 1
VLcalc1 þ f Rdc1 VLcalc1 þ Rdc2 VLcalc2
X 2
dVLcalci Lcalc1 dVLcalc2 dVL Rdc1
þ þ2 K ð Þ g þ C1 f calc1
i¼1
dt L calc2 dt dt Lcalc1
dVLcalc2 Rdc2 X2 2
d VLcalci Lcalc1 d VLcalc2 2
þ þ 2
þ2 K ð Þ g¼0
dt Lcalc2 i¼1
dt Lcalc2 dt2
VLcalc1 VL Lcalc2
’ calc2 ) VLcalc2 ¼ VLcalc1
Lcalc1 Lcalc2 Lcalc1
Then we get the following expression:
1 1 1 1 Lcalc2 X 2
dVLcalci
VLcalc1 þ f Rdc1 VLcalc1 þ Rdc2 VLcalc1 þ
Lcalc1 R1 Lcalc1 Lcalc2 Lcalc1 i¼1
dt
Lcalc1 dVLcalc1 Lcalc2 dVL Rdc1 dVLcalc1 Rdc2
þ2 K ð Þ g þ C1 f calc1 þ
Lcalc2 dt Lcalc1 dt Lcalc1 dt Lcalc1
r ffiffiffiffiffiffiffiffiffiffiffiffiffi
X 2 2
d VLcalci Lcalc1 Lcalc2 d VLcalc12
þ þ2 K ð Þ g¼0
i¼1
dt2 Lcalc2 Lcalc1 dt2
X
2
d 2 VL d 2 VLcalc1 d 2 VLcalc2 d 2 VLcalc1 Lcalc2
calci
¼ þ ¼ ð1 þ Þ;
i¼1
dt2 dt 2 dt 2 dt 2 Lcalc1
X
2
dVL dVLcalc1 dVLcalc2 dVLcalc1 Lcalc2
calci
¼ þ ¼ ð1 þ Þ
i¼1
dt dt dt dt Lcalc1
1 1 1 1
VLcalc1 þ f Rdc1 VLcalc1 þ Rdc2 VLcalc1
dVLcalc1 Lcalc2 Lcalc1 dVLcalc1 Lcalc2
þ ð1 þ Þþ2 K ð Þ g
dt Lcalc1 Lcalc2 dt Lcalc1
dVL Rdc1 dVLcalc1 Rdc2 d 2 VLcalc1 Lcalc2
þ C1 f calc1 þ þ ð1 þ Þ
dt Lcalc1 dt Lcalc1 dt2 Lcalc1
Lcalc1 Lcalc2 d 2 VLcalc1
þ2 K ð Þ g¼0
Lcalc2 Lcalc1 dt2
1 1 ðRdc1 þ Rdc2 Þ 1 Lcalc2 1 Lcalc2
½ þ VLcalc1 þ ½ ð1 þ Þþ2 K
Lcalc1 R1 Lcalc1 R1 Lcalc1 R1 Lcalc1
Rdc1 þ Rdc2 dVLcalc1 d 2 VLcalc1 Lcalc2 Lcalc2
þ C1 ð Þ þ C1 ½1 þ þ2K ¼0
Lcalc1 dt dt2 Lcalc1 Lcalc1
We define the following global parameters:
1 1 ðRdc1 þ Rdc2 Þ
g1 ¼ þ ;
Lcalc1 R1 Lcalc1
1 Lcalc2 1 Lcalc2 Rdc1 þ Rdc2
g2 ¼ ð1 þ Þþ2 K þ C1 ð Þ
R1 Lcalc1 R1 Lcalc1 Lcalc1
Lcalc2 Lcalc2
g3 ¼ C1 ½1 þ þ2 K ;
Lcalc1 Lcalc1
dVLcalc1 d 2 VLcalc1
g1 VLcalc1 þ g2 þ g3 ¼0
dt dt2
We get the following differential equation respect to VLcalc1 :
g1 VLcalc1 þ g2 þ g3 ¼ 0; g1 ¼ g1 ðRdc1 ; Rdc2 ; Lcalc1 ; R1 Þ
dt dt2
g2 ¼ g2 ðRdc1 ; Rdc2 ; Lcalc1 ; Lcalc2 ; R1 ; K; C1 Þ; g3 ¼ g3 ðLcalc1 ; Lcalc2 ; C1 ; KÞ
We define new variables: Y1 ¼ dt ; dYdt1 ¼ dt2
dY1 dY1 g g
g1 VLcalc1 þ g2 Y1 þ g3 ¼ 0; ¼ 1 VLcalc1 2 Y1 ;
dt dt g3 g3
g1 ¼ g1 ðRdc1 ; Rdc2 ; Lcalc1 ; R1 Þ
In the same manner we find VLcalc2 differential equation:
1 1 1 1
VLcalc1 þ
f Rdc1 VLcalc1 þ Rdc2 VLcalc2
X 2
dVLcalci Lcalc1 dVLcalc2 dVL Rdc1
þ þ2 K ð Þ g þ C1 f calc1
i¼1
dt L calc2 dt dt Lcalc1
dVLcalc2 Rdc2 X 2
d 2 VLcalci Lcalc1 d 2 VLcalc2
þ þ 2
þ2 K ð Þ g¼0
dt Lcalc2 i¼1
dt Lcalc2 dt2
VLcalc1 VL Lcalc1
’ calc2 ) VLcalc1 ¼ VLcalc2
Lcalc1 Lcalc2 Lcalc2
VLcalc2 1 Rdc1 Rdc2 X 2

dVLcalci
þ fVLcalc2 þ VLcalc2 þ
Lcalc2 R1 Lcalc2 Lcalc2 i¼1
dt
Lcalc1 dVLcalc2 dVL Rdc1 dVLcalc2 Rdc2
Lcalc2 dt dt Lcalc2 dt Lcalc2
X d 2 VL
2
Lcalc1 d 2 VLcalc2
þ calci
2
þ2 K ð Þ g¼0
i¼1
dt Lcalc2 dt2
X
2
d 2 VL d 2 VLcalc1 d 2 VLcalc2 d 2 VLcalc2 Lcalc1
calci
¼ þ ¼ ð þ 1Þ;
i¼1
dt2 dt 2 dt2 dt2 Lcalc2
X
2
dVL dVLcalc1 dVLcalc2 dVLcalc2 Lcalc1
calci
¼ þ ¼ ð þ 1Þ
i¼1
dt dt dt dt Lcalc2
VLcalc2 1 Rdc1 Rdc2 dVLcalc2 Lcalc1

þ fVLcalc2 þ VLcalc2 þ ð þ 1Þ
Lcalc2 R1 Lcalc2 Lcalc2 dt Lcalc2
Lcalc1 dVLcalc2 dVL Rdc1 dVLcalc2 Rdc2
Lcalc2 dt dt Lcalc2 dt Lcalc2
2
r ffiffiffiffiffiffiffiffiffiffiffiffiffi 2
d VLcalc2 Lcalc1 Lcalc1 d VLcalc2
þ ð þ 1Þ þ 2 K ð Þ g¼0
VLcalc2 1 Rdc1 1 Rdc2 dVLcalc2 1 Lcalc1

þ VLcalc2 þ VLcalc2 þ ð þ 1Þ
Lcalc2 R1 Lcalc2 R1 Lcalc2 dt R1 Lcalc2
1 Lcalc1 dVLcalc2 dVLcalc2 C1 Rdc1 dVLcalc2 C1 Rdc2
þ2 K ð Þ þ þ
R1 Lcalc2 dt dt Lcalc2 dt Lcalc2
2
r ffiffiffiffiffiffiffiffiffiffiffiffiffi 2
d VLcalc2 Lcalc1 Lcalc1 d VLcalc2
þ C1 ð þ 1Þ þ 2 K C1 ð Þ ¼0
1 1 Rdc1 þ Rdc2 1 Lcalc1 1 Lcalc1
½ þ ð Þ VLcalc2 þ ½ ð þ 1Þ þ 2 K
Lcalc2 R1 Lcalc2 R1 Lcalc2 R1 Lcalc2
Rdc1 þ Rdc2 dVLcalc2 d 2 VLcalc2 Lcalc1 Lcalc1
þ C1 ð Þ þ C1 ½ þ1þ2 K ¼0
Lcalc2 dt dt2 Lcalc2 Lcalc2
þ Rdc2
We define the following global parameters: n1 ¼ Lcalc2
1
þ 1
R1 ðRdc1
Lcalc2 Þ
1 Lcalc1 1 Lcalc1 Rdc1 þ Rdc2
n2 ¼ ð þ 1Þ þ 2 K þ C1 ð Þ;
R1 Lcalc2 R1 Lcalc2 Lcalc2
Lcalc1 Lcalc1
n3 ¼ C1 ½ þ1þ2 K
Lcalc2 Lcalc2
n1 ¼ n1 ðRdc1 ; Rdc2 ; Lcalc2 ; R1 Þ; n2 ¼ n2 ðLcalc1 ; Lcalc2 ; Rdc1 ; Rdc2 ; R1 ; K; C1 Þ;

n3 ¼ n3 ðLcalc1 ; Lcalc2 ; K; C1 Þ
n1 VLcalc2 þ n2 dt þ dt2 n3 ¼ 0. We define new variables:
Y2 ¼ dt ; dYdt2 ¼ dt2
n1 VLcalc2 þ n2 þ n3 ¼ 0;
dt dt2
dY2 dY2 n n
n1 VLcalc2 þ n2 Y2 þ n3 ¼ 0 ) ¼ 1 VLcalc2 2 Y2
dt dt n3 n3
Summary: We can get our N-turn multilayer circular coil antennas (RFID
system) system.
dVLcalc1 dY1 g g
¼ Y1 ; ¼ 1 VLcalc1 2 Y1 ;
dt dt g3 g3
dVLcalc2 dY2 n n
¼ Y2 ; ¼ 1 VLcalc2 2 Y2
dt dt n3 n3
0 1
dY1 0 1 0 Y 1
dt
B dVLcalc1 C C11 ... C14 1
B dt C B .. .. .. C B B VLcalc1 C
C g2 g1
B dY2 C ¼ @ . . . A @ Y2 A; C11 ¼ g ; C12 ¼ g ; C13
@ dt A 3 3
dVLcalc2 C41 C44 VLcalc2
dt
¼ 0; C14 ¼ 0
n2 n
C21 ¼ 1; C22 ¼ C23 ¼ C24 ¼ 0; C31 ¼ C32 ¼ 0; C33 ¼ ; C34 ¼ 1 ; C41 ¼ C42 ¼ 0
n3 n3
C43 ¼ 1; C44 ¼ 0
N-turn multilayer circular coil antennas (RFID system) system is composed from
two N-turn thin wire multilayer and circular. Units are all in centimeters (cm),
ai ; hi ; bi 8 i ¼ 1; 2; Ni 2 R þ . Inductors Lcalc1 ; Lcalc2 units are in lH. Due to
electromagnetic interferences there are different in time delays respect to first and
second N-turn thin wire multilayer and circular antennas voltages and voltages
derivatives. Additionally, there is a delay in time Dl for the current that
flows through inductor Lcalc2 and mutual inductance. The delay voltages are
V1 ðt s1 Þ; V2 ðt s2 Þ respectively (s1 6¼ s2 ) and delayed voltages derivatives are
dV1 ðts1 Þ dV2 ðts2 Þ
dt ; dt respectively (D1 6¼ D2 ; s1 0; s2 0; D1 ; D2 2 R þ ).
V1 ðtÞ ¼ VLcalc1 ¼ VLcalc1 ðtÞ; V2 ðtÞ ¼ VLcalc2 ¼ VLcalc2 ðtÞ;

VLcalc1 ðtÞ ! VLcalc1 ðt s1 Þ; VLcalc2 ðtÞ ! VLcalc2 ðt s2 Þ
dV dV
We consider no delay effect on dYdt1 ; Ldt ; dt ; Ldt
calc1 dY2 calc2
. The N-turn multilayer
circular coil antennas system differential equations under electromagnetic inter-
ferences (delay terms) influence only N-turn multilayer circular coil antennas sys-
tem first and second N-turn multilayer circular coil voltages VLcalc1 ðtÞ; VLcalc2 ðtÞ and
voltages derivatives Y1 ðtÞ and Y2 ðtÞ respect to time, there is no influence on
dY1 ðtÞ dVLcalc1 ðtÞ dY2 ðtÞ dVLcalc2 ðtÞ
dt ; dt ; dt ; dt [85].
0 1
dY1 0 1 0 Y ðt D Þ 1
dt
B dVLcalc1 C C11 . . . C14 1 1
B dt C B .. .. .. C B B VLcalc1 ðt s1 Þ C
C
B dY2 C ¼ @ . . . A @ Y2 ðt D2 Þ A
@ dt A
dVLcalc2 C41 C44 VLcalc2 ðt s2 Þ
dt
To find equilibrium points (fixed points) of N-turn multilayer circular coil

antennas RFID system is by
lim Y1 ðt D1 Þ ¼ Y1 ðtÞ; lim VLcalc1 ðt s1 Þ ¼ VLcalc1 ðtÞ

t!1 t!1
lim Y2 ðt D2 Þ ¼ Y2 ðtÞ; lim VLcalc2 ðt s2 Þ ¼ VLcalc2 ðtÞ:
t!1 t!1
dY1 dVLcalc1 dY2 dVLcalc2

¼ 0; ¼ 0; ¼ 0; ¼ 0; t s1 ; t s2 ; t D1 ; t D2
dt dt dt dt
ð0Þ ð0Þ
We get four equations and the only fixed point is Eð0Þ ðY1 ; VLcalc1 ;
ð0Þ ð0Þ
Y2 ; VLcalc2 Þ ¼ ð0; 0; 0; 0Þ.
1 Lcalc2
g3 6¼ 0 & g2 6¼ 0 ) ð1 þ Þ
R1 Lcalc1
1 Lcalc2 Rdc1 þ Rdc2
þ2 K þ C1 ð Þ 6¼ 0; C11 6¼ 0
R1 Lcalc1 Lcalc1
1 1 ðRdc1 þ Rdc2 Þ
g3 6¼ 0 & g1 6¼ 0 )C12 6¼ 0; þ 6¼ 0
Lcalc1 R1 Lcalc1
1 1 Rdc1 þ Rdc2
n3 6¼ 0; n1 6¼ 0 ) þ ð Þ 6¼ 0; C34 6¼ 0
Lcalc2 R1 Lcalc2
1 Lcalc1 1 Lcalc1
n3 6¼ 0; n2 6¼ 0 ) ð þ 1Þ þ 2 K
R1 Lcalc2 R1 Lcalc2
Rdc1 þ Rdc2
þ C1 ð Þ 6¼ 0; C33 6¼ 0
Lcalc2
equilibrium points (fixed points) of N-turn multilayer circular coil antennas RFID
system consists in adding to coordinates [Y1 VLcalc1 Y2 VLcalc2 ] arbitrarily small
increments of exponential form ½y1 vLcalc1 y2 vLcalc2 ekt , and retaining the first order
terms in Y1 VLcalc1 Y2 VLcalc2 . The system of four homogeneous equations leads to a
polynomial characteristics equation in the eigenvalues k. The polynomial charac-
teristics equations accept by set the below voltages and voltages derivative respect
to the time into N-turn multilayer circular coil antennas RFID system equations.
N-turn multilayer circular coil antennas RFID system fixed values with arbitrarily
small increments of exponential form ½Y1 VLcalc1 Y2 VLcalc2 ekt are: i = 1 (first fixed
point), i = 2 (second fixed point), i = 3 (third fixed point), etc.,
ðiÞ ðiÞ
Y1 ðtÞ ¼ Y1 þ y1 ekt ; VLcalc1 ðtÞ ¼ VLcalc1 þ vLcalc1 ekt ;
ðiÞ ðiÞ
Y2 ðtÞ ¼ Y2 þ y2 ekt ; VLcalc2 ðtÞ ¼ VLcalc2 þ vLcalc2 ekt
We choose the above expressions for our Y1 ðt),VLcalc1 ðt),Y2 ðt),VLcalc2 ðt) as small
displacement ½y1 vLcalc1 y2 vLcalc2 from the system fixed points at time t = 0.
ðiÞ ðiÞ
Y1 ðt ¼ 0Þ ¼ Y1 þ y1 ; VLcalc1 ðt ¼ 0Þ ¼ VLcalc1 þ vLcalc1 ;
ðiÞ ðiÞ
Y2 ðt ¼ 0Þ ¼ Y2 þ y2 ; VLcalc2 ðt ¼ 0Þ ¼ VLcalc2 þ vLcalc2
For t > 0, k < 0 the selected fixed point is stable otherwise t > 0, k > 0 is
unstable. Our system tends to the selected fixed point exponentially for t > 0, k < 0
otherwise go away from the selected fixed point exponentially. Eigenvalue k is the
parameter which establish if the fixed point is stable or unstable, additionally his
absolute value (|k|) establish the speed of flow toward or away from the selected
fixed point. Table 8.1 describes N-turn multilayer circular coil antennas RFID
system variables for different eigenvalue k and t values.
The speeds of flow toward or away from the selected fixed point for N-turn
multilayer circular coil antennas RFID system voltages and voltages derivatives
respect to time are as follow:
dY1 ðtÞ dVLcalc1 ðtÞ

¼ y1 k ekt ; ¼ vLcalc1 k ekt ;
dt dt
dY2 ðtÞ dVLcalc2 ðtÞ
¼ y2 k ekt ; ¼ vLcalc2 k ekt
dt dt
Table 8.1 N-turn multilayer circular coil antennas RFID system variables for different eigenvalue
k and time
k<0 k>0
t=0 ðiÞ ðiÞ
Y1 ðt ¼ 0Þ ¼ Y1 þ y1 Y1 ðt ¼ 0Þ ¼ Y1 þ y1
ðiÞ ðiÞ
VLcalc1 ðt ¼ 0Þ ¼ VLcalc1 þ vLcalc1 VLcalc1 ðt ¼ 0Þ ¼ VLcalc1 þ vLcalc1
ðiÞ ðiÞ
Y2 ðt ¼ 0Þ ¼ Y2 þ y2 Y2 ðt ¼ 0Þ ¼ Y2 þ y2
ðiÞ ðiÞ
ðiÞ ðiÞ
t>0 Y1 ðtÞ ¼ Y1 þ y1 ejkjt Y1 ðtÞ ¼ Y1 þ y1 ejkjt
ðiÞ ðiÞ
VLcalc1 ðtÞ ¼ VLcalc1 þ vLcalc1 ejkjt VLcalc1 ðtÞ ¼ VLcalc1 þ vLcalc1 ejkjt
ðiÞ ðiÞ
Y2 ðtÞ ¼ Y2 þ y2 ejkjt Y2 ðtÞ ¼ Y2 þ y2 ejkjt
ðiÞ ðiÞ
t!∞ Y1 ðt ! 1Þ ¼ Y1
ðiÞ
Y1 ðt ! 1Þ ’ y1 ejkjt
VLcalc1 ðt ! 1Þ ¼ VLcalc1
ðiÞ VLcalc1 ðt ! 1Þ ’ vLcalc1 ejkjt
Y2 ðt ! 1Þ ¼ Y2
ðiÞ Y2 ðt ! 1Þ ’ y2 ejkjt
ðiÞ VLcalc2 ðt ! 1Þ ’ vLcalc2 ejkjt
dVLcalc1 ðiÞ ðiÞ

¼ Y1 ) vLcalc1 k ekt ¼ Y1 þ y1 ekt ; Y1 ¼ 0; vLcalc1 k þ y1 ¼ 0
dt
dY1 g g
¼ 1 VLcalc1 2 Y1
dt g3 g3
g1 ðiÞ g ðiÞ
) y1 k ekt ¼ ðVLcalc1 þ vLcalc1 ekt Þ 2 ðY1 þ y1 ekt Þ
g3 g3
g1 ðiÞ g ðiÞ g g
y1 k ekt ¼ VLcalc1 2 Y1 1 vLcalc1 ekt 2 y1 ekt ;
g3 g3 g3 g3
g1 ðiÞ g2 ðiÞ
VLcalc1 Y1 ¼ 0
g3 g3
g1 g
y1 k vLcalc1 2 y1 ¼ 0;
g3 g3
¼ Y2 ) vLcalc2 k ekt ¼ Y2 þ y2 ekt ; Y2 ¼ 0
dt
dY2 n n
vLcalc2 k þ y2 ¼ 0; ¼ 1 VLcalc2 2 Y2
dt n3 n3
n ðiÞ n ðiÞ
) y2 k ekt ¼ ðVLcalc2 þ vLcalc2 ekt Þ 2 ðY2 þ y2 ekt Þ
1
n3 n3
n1 ðiÞ n ðiÞ n n
y2 k ekt ¼ V 2 Y 1 vLcalc2 ekt 2 y2 ekt ;
n3 Lcalc2 n3 2 n3 n3
n1 ðiÞ n2 ðiÞ
V Y2 ¼ 0
n3 Lcalc2 n3
n1 n
y2 k vLcalc2 2 y2 ¼ 0
n3 n3
We can summary our N-turn multilayer circular coil antennas RFID system
arbitrarily small increments equations:
g2 g
vLcalc1 k þ y1 ¼ 0; y1 k y1 1 vLcalc1 ¼ 0; vLcalc2 k þ y2 ¼ 0;
g3 g3
n2 n
y2 k y2 1 vLcalc2 ¼ 0
n3 n3
0 1 0 y1 1
!11 . . . !14
B . Bv C
B . .. .. C
C B Lcalc1 C g2 g1
@ . . . AB
@ y2 A
C ¼ 0; !11 ¼ k ; !12 ¼ ;
g3 g3
!41 !44 v Lcalc2
!13 ¼ 0; !14 ¼ 0; !21 ¼ 1
n2
!22 ¼ k; !23 ¼ 0; !24 ¼ 0; !31 ¼ 0; !32 ¼ 0; !33 ¼ k ;
n3
n1
!34 ¼ ; !41 ¼ 0
n3
!42 ¼ 0; !43 ¼ 1; !44 ¼ k
0 1
!11 . . . !14
B . .. .. C X4
AkI ¼B @ .
. . .
C; detðA k IÞ ¼ 0 )
A kk Pk ¼ 0;
k¼0
!41 !44
g n g n g n
P0 ¼ 1 1 ; P1 ¼ 2 1 þ 1 2
g3 n3 g3 n3 g3 n3
n g n g n g
P2 ¼ 1 þ 2 2 þ 1 ; P3 ¼ 2 þ 2 ; P4 ¼ 1
n3 g3 n3 g3 n3 g3
Eigenvalues stability discussion: Our N-turn multilayer circular coil antennas

RFID system involving Na variables (Na > 2, Na = 4), the characteristic equation is
P
of degree Na = 4 ( 4k¼0 kk Pk ¼ 0) and must often be solved numerically. Expect
in some particular cases, such an equation has (Na = 4) distinct roots that can be real
or complex. These values are the eigenvalues of the 4 4 Jacobian matrix (A). The
general rule is that the Steady State (SS) is stable if there is no eigenvalue with
positive real part. It is sufficient that one eigenvalue is positive for the steady state to
be unstable. Our 4-variables (Y1 VLcalc1 Y2 VLcalc2 ) N-turn multilayer circular coil
antennas RFID system has four eigenvalues. The type of behavior can be charac-
terized as a function of the position of these eigenvalues in the Re/Im plane. Five
non-degenerated cases can be distinguished: (1) the four eigenvalues are real and
negative (stable steady state), (2) the four eigenvalues are real, three of them are
negative (unstable steady state), (3) and (4) two eigenvalues are complex conjugates
with a negative real part and the other eigenvalues are real and negative (stable
steady state), two cases can be distinguished depending on the relative value of the
real part of the complex eigenvalues and of the real one, (5) two eigenvalues are
complex conjugates with a negative real part and at least one eigenvalue is positive
(unstable steady state) [2, 3].
We define the following N-turn multilayer circular coil antennas RFID system
variables delayed in time:
ðiÞ ðiÞ
Y1 ðt D1 Þ ¼ Y1 þ y1 ekðtD1 Þ ; VLcalc1 ðt s1 Þ ¼ VLcalc1 þ vLcalc1 ekðts1 Þ ;
ðiÞ ðiÞ
Y2 ðt D2 Þ ¼ Y2 þ y2 ekðtD2 Þ ; VLcalc2 ðt s2 Þ ¼ VLcalc2 þ vLcalc2 ekðts2 Þ
We get four delayed differential equations respect to coordinates

[Y1 VLcalc1 Y2 VLcalc2 ] arbitrarily small increments of exponential
½y1 vLcalc1 y2 vLcalc2 e .
kt
ðiÞ ðiÞ
vLcalc1 k ekt ¼ Y1 þ y1 ekðtD1 Þ ; Y1 ¼ 0; vLcalc1 k þ y1 ekD1 ¼ 0
g ðiÞ g ðiÞ
y1 k ekt ¼ 1 ðVLcalc1 þ vLcalc1 ekðts1 Þ Þ 2 ðY1 þ y1 ekðtD1 Þ Þ
g3 g3
g ðiÞ g ðiÞ g g
y1 k ekt ¼ 1 VLcalc1 2 Y1 1 vLcalc1 ekðts1 Þ 2 y1 ekðtD1 Þ
g3 g3 g3 g3
g1 ðiÞ g2 ðiÞ g1 ks1 g
VLcalc1 Y1 ¼ 0; y1 k vLcalc1 e 2 y1 ekD1 ¼ 0
g3 g3 g3 g3
¼ Y2 ; vLcalc2 k ekt ¼ Y2 þ y2 ekðtD2 Þ ; Y2 ¼ 0; vLcalc2 k þ y2 ekD2 ¼ 0
dt
n ðiÞ n ðiÞ n
y2 k ekt ¼ 1 ðVLcalc2 þ vLcalc2 ekðts2 Þ Þ 2 Y2 2 y2 ekðtD2 Þ
n3 n3 n3
n ðiÞ n ðiÞ n n
y2 k ekt ¼ VLcalc2 Y2 vLcalc2 ekðts2 Þ 2 y2 ekðtD2 Þ
1 2 1
n3 n3 n3 n3
n1 ðiÞ n2 ðiÞ n1 n
VLcalc2 Y2 ¼ 0; y2 k vLcalc2 eks2 2 y2 ekD2 ¼ 0
n3 n3 n3 n3
The small increments Jacobian of our N-turn multilayer circular coil antennas
RFID system is as follow:
g2 kD1 g
½k e y1 1 eks1 vLcalc1 ¼ 0; vLcalc1 k þ y1 ekD1 ¼ 0
g3 g3
n n
½k 2 ekD2 y2 1 eks2 vLcalc2 ¼ 0; vLcalc2 k þ y2 ekD2 ¼ 0
n3 n
0 3 1
0 1 y1
i11 . . . i14 B C
B . . . C B B vLcalc1 C
C g2
B . . . .. C C ¼ 0; i11 ¼ k ekD1 ;
@ . AB B C g3
@ y2 A
i41 i44
vLcalc2
g1 ks1
i12 ¼ e ; i13 ¼ i14 ¼ 0
g3
n2 kD2 n
i21 ¼ ekD1 ; i22 ¼ k; i23 ¼ i24 ¼ 0; i31 ¼ i32 ¼ 0; i33 ¼ k e ; i34 ¼ 1 eks2
n3 n3
i41 ¼ i42 ¼ 0; i43 ¼ ekD2 ; i44 ¼ k
0 1
i11 . . . i14
B . . C
AkI ¼B @ .. . . ... C
A; det jA k Ij ¼ 0; D ¼ Dðk; s1 ; s2 ; D1 ; D2 Þ
i41 i44
n n g g
C33 ¼ 2 ; C34 ¼ 1 ; C11 ¼ 2 ; C12 ¼ 1 ; i33 ¼ k þ C33 ekD2 ;
n3 n3 g3 g3
i34 ¼ C34 eks2
i11 ¼ k þ C11 ekD1 ; i12 ¼ C12 eks1
Dðk; s1 ; s2 ; D1 ; D2 Þ ¼ k4 k3 C33 ekD2 k2 C34 ekðD2 þ s2 Þ k3 C11 ekD1

P 2 P2
k Di kðs2 þ Di Þ
þ k2 C11 C33 e i¼1 þ k C11 C34 e i¼1 k2 C12 ekðs1 þ D1 Þ
P
2 P
2 P
2
kðs1 þ Di Þ kð si þ Di Þ
þ k C12 C33 e i¼1 þ C12 C34 e i¼1 i¼1
P
2 P
2 P
2
kð si þ Di Þ kðs2 þ Di Þ
Dðk; s1 ; s2 ; D1 ; D2 Þ ¼ k þ C12 C34 e
4 i¼1 i¼1 þ k ½C11 C34 e i¼1
P2 P
2
kðs1 þ Di Þ k Di
þ C12 C33 e i¼1 þ k2 ½C11 C33 e i¼1 C12 ekðs1 þ D1 Þ C34 ekðD2 þ s2 Þ
kD2 kD1
k ½C33 e
3
þ C11 e
We have three stability cases: (1) s1 ¼ s2 ¼ s & D1 ¼ 0; D2 ¼ 0 (2) s1 ¼ s2 ¼ 0

& D1 ¼ D; D2 ¼ D (3) s1 ¼ s2 ¼ sD & D1 ¼ sD ; D2 ¼ sD . Otherwise s1 6¼ s2 and
D1 6¼ D2 , they are positive parameters. There are other possible simple stability
cases: (4) s1 ¼ s; s2 ¼ 0; D1 ¼ D2 ¼ 0 (5) s1 ¼ 0; s2 ¼ s; D1 ¼ D2 ¼ 0 (6) s1 ¼ 0;
s2 ¼ 0; D1 ¼ D; D2 ¼ 0 (7) s1 ¼ 0; s2 ¼ 0; D1 ¼ 0; D2 ¼ D.
the value of time delays s; D; sD for the general characteristic equation
Dðk; s=D=sD Þ. If we choose s parameter then Dðk; sÞ ¼ Pn ðk; sÞ þ Qm ðk; sÞ eks .
X
n
Pn ðk; sÞ ¼ pk ðsÞ kk ¼ p0 ðsÞ þ p1 ðsÞ k þ p2 ðsÞ k2 þ . . .
k¼0
The expression for Qm ðk; sÞ is
X
m
Qm ðk; sÞ ¼ qk ðsÞ kk ¼ q0 ðsÞ þ q1 ðsÞ k þ q2 ðsÞ k2 þ . . .
k¼0
First we analyze the case when there is delay in first (i = 1) RFID’s N-turn
multilayer coil antenna voltage (s1 ¼ s) and there is no delay in second (i = 2)
RFID’s N-turn multilayer coil antenna voltage (s2 ¼ s). Additionally there is
no delay in RFID’s N-turn multilayer coil antenna voltages derivatives
(D1 ¼ 0; D2 ¼ 0).
The general characteristic equation Dðk; sÞ is as follow:
Dðk; s1 ¼ s; s2 ¼ 0; D1¼0 ; D2 ¼ 0Þ ¼ k4 þ C12 C34 eks þ k ½C11 C34

þ C12 C33 eks þ k2 ½C11 C33 C34 C12 eks k3 ½C33 þ C11
Dðk; s1 ¼ s; s2 ¼ 0; D1¼0 ; D2 ¼ 0Þ ¼ k4 k3 ½C33 þ C11 þ k2 ðC11 C33 C34 Þ

þ k C11 C34 þ ðk2 C12 þ k C12 C33 þ C12 C34 Þ eks
X
n¼4
Pn¼4 ðk; sÞ ¼ pk ðsÞ kk ¼ p0 ðsÞ þ p1 ðsÞ k þ p2 ðsÞ k2 þ p3 ðsÞ k3 þ p4 ðsÞ k4
k¼0
p0 ðsÞ ¼ 0; p1 ðsÞ ¼ C11 C34 ; p2 ðsÞ ¼ C11 C33 C34 ; p3 ðsÞ ¼ ½C33 þ C11 ; p4 ðsÞ ¼ 1
The expression for Qm ðk; sÞ:
X
m¼2
Qm¼2 ðk; sÞ ¼ qk ðsÞ kk ¼ q0 ðsÞ þ q1 ðsÞ k þ q2 ðsÞ k2
k¼0
q0 ðsÞ ¼ C12 C34 ; q1 ðsÞ ¼ C12 C33 ; q2 ðsÞ ¼ C12 . The homogeneous system
for Y1 VLcalc1 Y2 VLcalc2 leads to a characteristic equation for the eigenvalue k having
P P
the form PðkÞ þ QðkÞ eks ; PðkÞ ¼ 4j¼0 aj k j ; QðkÞ ¼ 2j¼0 cj k j and the
coefficients faj ðqi ; qk Þ; cj ðqi ; qk Þg 2 R depend on qi ; qk but not on s, qi ; qk are any
RFID’s N-turn multilayer coil antennas system’s parameters, other parameters keep
as a constant.
a0 ¼ 0; a1 ¼ C11 C34 ; a2 ¼ C11 C33 C34 ; a3 ¼ ðC33 þ C11 Þ; a4 ¼ 1;

c0 ¼ C12 C34 ; c1 ¼ C33 C12 ; c2 ¼ C12 :
Unless strictly necessary, the designation of the variation arguments ðqi ; qk Þ will
subsequently be omitted from P, Q, aj, and cj. The coefficients aj, and cj are
continuous, and differential functions of their arguments, and direct substitution
shows that a0 þ c0 6¼ 0 8 qi ; qk 2 R þ (C12 C34 6¼ 0). k = 0 is not a of
PðkÞ þ QðkÞ eks ¼ 0. Furthermore, PðkÞ; QðkÞ are analytic functions of k, for
present case [4–6]:
(b) jQðkÞ
PðkÞ j is bounded for jkj ! 1, Rek 0. No roots bifurcation from 1.
polynomial in x.
(d) Each positive root xðqi ; qk Þ of FðxÞ ¼ 0 is continuous and differentiable
We assume that Pn ðk; sÞ ¼ Pn ðkÞ and Qm ðk; sÞ ¼ Qm ðkÞ cannot have common
imaginary roots. That is for any real number x: Pn ðk ¼ i x; sÞ þ
Qm ðk ¼ i x; sÞ 6¼ 0.
Pn ðk ¼ i x; sÞ ¼ ðC11 C34 x þ ½C33 þ C11 x3 Þ i ðC11 C33 C34 Þ x2 þ x4

Qm ðk ¼ i x; sÞ ¼ C12 C34 þ C12 x2 þ C12 C33 i x
Pn ðk ¼ i x; sÞ þ Qm ðk ¼ i x; sÞ ¼ C12 C34 þ ðC12 C11 C33 þ C34 Þ x2 þ x4

þ ðC11 C34 þ C12 C33 Þ i x þ i ½C33 þ C11 x3 6¼ 0
jPn ðk ¼ i x; sÞj2 ¼ ðC11 C34 x þ ½C33 þ C11 x3 Þ2 þ ððC11 C33 C34 Þ x2 þ x4 Þ2

¼ C211 C234 x2 þ ½C33 þ C11 2 x6 þ 2 C11 C34 ½C33 þ C11 x4
þ ðC11 C33 C34 Þ2 x4 þ x8 2 ðC11 C33 C34 Þ x6
jPn ðk ¼ i x; sÞj2 ¼ jPði xÞj2 ¼ x8 þ f½C33 þ C11 2 2 ðC11 C33 C34 Þg x6

þ ð2 C11 C34 ½C33 þ C11 þ ðC11 C33 C34 Þ2 Þ x4 þ C211 C234 x2
jQm ðk ¼ i x; sÞj2 ¼ ðC12 C34 þ C12 x2 Þ2 þ C212 C233 x2 ¼ C212 C234 þ C212 x4
þ 2 C212 C34 x2 þ C212 C233 x2
jQm ðk ¼ i x; sÞj2 ¼ jQði xÞj2 ¼ C212 C234 þ ð2 C34 þ C233 Þ C212 x2 þ C212 x4
FðxÞ ¼ jPði xÞj2 jQði xÞj2 ¼ x8 þ f½C33 þ C11 2 2 ðC11 C33 C34 Þg x6
þ ð2 C11 C34 ½C33 þ C11 þ ðC11 C33 C34 Þ2 Þ x4 þ C211 C234 x2 C212 C234
ð2 C34 þ C233 Þ C212 x2 C212 x4
FðxÞ ¼ jPði xÞj2 jQði xÞj2 ¼ x8 þ f½C33 þ C11 2 2 ðC11 C33 C34 Þg x6
þ fð2 C11 C34 ½C33 þ C11 þ ðC11 C33 C34 Þ2 Þ C212 g x4
þ fC211 C234 ð2 C34 þ C233 Þ C212 g x2 C212 C234
We define the following parameters for simplicity: N0 ¼ C212 C234
N2 ¼ C211 C234 ð2 C34 þ C233 Þ C212 ; N4 ¼ 2 C11 C34 ½C33 þ C11

þ ðC11 C33 C34 Þ2 C212
N6 ¼ ½C33 þ C11 2 2 ðC11 C33 C34 Þ; N8 ¼ 1
FðxÞ ¼ jPði xÞj2 jQði xÞj2 ¼ N8 x8 þ N6 x6 þ N4 x4 þ N2 x2 þ N0

X4
¼ N2k x2k
k¼0
P
Hence FðxÞ ¼ 0 implies 4k¼0 N2k x2k ¼ 0 and its roots are given by solving
the above polynomial. Furthermore PR ði x; sÞ ¼ ðC11 C33 C34 Þ x2 þ x4
PI ði x; sÞ ¼ C11 C34 x þ ½C33 þ C11 x3 ; QR ði x; sÞ

¼ C12 C34 þ C12 x2 ; QI ði x; sÞ ¼ C12 C33 x
Hence
sin hðsÞ ¼
jQði x; sÞj2
And
cos hðsÞ ¼
jQði x; sÞj2
fðC11 C33 C34 Þ x2 þ x4 g C12 C33 x

þ fC11 C34 x þ ½C33 þ C11 x3 g fC12 C34 þ C12 x2 g
sin hðsÞ ¼
C212 C234 þ C212 x4 þ 2 C212 C34 x2 þ C212 C233 x2
fðC11 C33 C34 Þ x2 þ x4 g fC12 C34 þ C12 x2 g

þ fC11 C34 x þ ½C33 þ C11 x3 g C12 C33 x
cos hðsÞ ¼
C212 C234 þ C212 x4 þ 2 C212 C34 x2 þ C212 C233 x2
P
Which jointly with FðxÞ ¼ 0 ) 4k¼0 N2k x2k ¼ 0 that are continuous and
differentiable in s based on Lemma 1.1. Hence we use Theorem 1.2 and this prove
the Theorem 1.3.
The second case we analyze is when there is a delay in first (i = 1) and second
(i = 2) RFID’s N-turn multilayer coil antennas voltages (s1 ¼ s; s2 ¼ s) and no
delay in RFID’s N-turn multilayer coil antennas voltages derivatives D1 ¼ 0

D2 ¼ 0. The general characteristic equation Dðk; sÞ is as follow:
P
2 P
2 P
2
Dðk; s1 ; s2 ; D1 ; D2 Þ ¼ k þ C12 C34 e
4 i¼1 i¼1 þ k ½C11 C34 e i¼1
P2 P
2
kðs1 þ Di Þ k Di
þ C12 C33 e i¼1 þ k2 ½C11 C33 e i¼1 C12 ekðs1 þ D1 Þ
C34 ekðD2 þ s2 Þ k3 ½C33 ekD2 þ C11 ekD1
Dðk; s1 ¼ s; s2 ¼ s; D1 ¼ 0; D2 ¼ 0Þ ¼ k4 þ C12 C34 ek2s þ k ½C11 C34 eks

þ C12 C33 eks þ k2 ½C11 C33 C12 eks C34 eks k3 ½C33 þ C11
Dðk; s1 ¼ s; s2 ¼ s; D1 ¼ 0; D2 ¼ 0Þ ¼ k4 k3 ½C33 þ C11 þ k2 C11 C33

þ ðC12 C34 eks þ k ½C11 C34 þ C12 C33 k2 ½C12 þ C34 Þ eks
Under Taylor series approximation: eks 1 k s þ 12 k2 s2 . The Maclaurin

series is a Taylor series expansion of a expðk sÞ function about zero (0). We get
the following general characteristic equation D(k,s) under Taylor series approxi-
mation: eks 1 k s þ 12 k2 s2 .

1
þ ðC12 C34 ½1 k s þ k2 s2 þ k ½C11 C34 þ C12 C33
2
ks
k ½C12 þ C34 Þ e
2

1
þ fk2 ð C12 C34 s2 C12 C34 Þ þ k ðC11 C34 þ C12 C33 C12 C34 sÞ
2
þ C12 C34 g eks
X
n¼4
Pn ðk; sÞ ¼ k4 k3 ½C33 þ C11 þ k2 C11 C33 ; Pn¼4 ðk; sÞ ¼ pk ðsÞ kk
k¼0
X
n¼4
pk ðsÞ kk ¼ p0 ðsÞ þ p1 ðsÞ k þ p2 ðsÞ k2 þ p3 ðsÞ k3 þ p4 ðsÞ k4
k¼0
p0 ðsÞ ¼ 0; p1 ðsÞ ¼ 0; p2 ðsÞ ¼ C11 C33 ; p3 ðsÞ ¼ ½C33 þ C11 ; p4 ðsÞ ¼ 1
1
Qm¼2 ðk; sÞ ¼ k2 ð C12 C34 s2 C12 C34 Þ þ k ðC11 C34 þ C12 C33 C12
2
C34 sÞ þ C12 C34
X
k¼2
Qm¼2 ðk; sÞ ¼ qk ðsÞ kk ¼ q0 þ q1 k þ q2 k2 ; q0 ðsÞ ¼ C12 C34
k¼0
1
q2 ðsÞ ¼ C12 C34 s2 C12 C34 ; q1 ðsÞ ¼ C11 C34 þ C12 C33 C12 C34 s
2
The homogeneous system for Y1 VLcalc1 Y2 VLcalc2 leads to characteristic equation

P
for the eigenvalue k having the form Pðk; sÞ þ Qðk; sÞ eks , Pðk; sÞ ¼ 4j¼0 aj ðsÞ
P
k j QðkÞ ¼ 2j¼0 cj ðsÞ k j and the coefficients faj ðqi ; qk ; sÞ; cj ðqi ; qk ; sÞg 2 R
depend on qi ; qk and delay s, qi ; qk are any RFID’s N-turn multilayer coil antennas
system’s parameters, other parameters keep as a constant.
a0 ¼ 0; a1 ¼ 0; a2 ¼ C11 C33 ; a3 ¼ ½C33 þ C11 ; a4 ¼ 1; c0 ¼ C12 C34

1
c1 ¼ C11 C34 þ C12 C33 C12 C34 s; c2 ¼ C12 C34 s2 C34 C12
2
Unless strictly necessary, the designation of the variation arguments (qi ; qk ; s)

will subsequently be omitted from P, Q, aj, and cj. The coefficients aj, cj are
continuous and differentiable functions of their arguments, and direct substitution
shows that a0 þ c0 6¼ 0; C12 C34 6¼ 0 8 qi ; qk 2 R þ , i.e. k ¼ 0 is not a
Pðk; sÞ þ Qðk; sÞ eks . We assume that Pn ðk; sÞ and Qm ðk; sÞ can’t have common
imaginary roots. That is for any real number x: Pn ðk ¼ i x; sÞ þ
Qm ðk ¼ i x; sÞ 6¼ 0.
Pn ðk ¼ i x; sÞ ¼ x4 x2 C11 C33 þ i x3 ½C33 þ C11

1
Qm ðk ¼ i x; sÞ ¼ x2 ð C12 C34 s2 C12 C34 Þ þ C12 C34
2
þ i x ðC11 C34 þ C12 C33 C12 C34 sÞ
1
Pn ðk ¼ i x; sÞ þ Qm ðk ¼ i x; sÞ ¼ x4 x2 ½ C12 C34 s2 C12 C34 þ C11 C33
2
þ C12 C34 þ i fx3 ½C33 þ C11 þ x ðv11 C34 þ C12 C33 C12 C34 sÞg 6¼ 0
jPn ðk ¼ i x; sÞj2 ¼ jPði xÞj2 ¼ ðx4 x2 C11 C33 Þ2 þ x6 ½C33 þ C11 2

¼ x8 þ x6 ð½C33 þ C11 2 2 C11 C33 Þ þ x4 C211 C233
jPn ðk ¼ i x; sÞj2 ¼ jPði xÞj2 ¼ x8 þ x6 ð½C33 þ C11 2 2 C11 C33 Þ þ x4 C211 C233
1
jQm ðk ¼ i x; sÞj2 ¼ jQði xÞj2 ¼ ðx2 ð C12 C34 s2 C12 C34 Þ þ C12 C34 Þ2
2
þ x2 ðC11 C34 þ C12 C33 C12 C34 sÞ2
1
jQm ðk ¼ i x; sÞj2 ¼ jQði xÞj2 ¼ x4 ð C12 C34 s2 C12 C34 Þ2
2
þ x fðC11 C34 þ C12 C33 C12 C34 sÞ2
2
1
2 ð C12 C34 s2 C12 C34 Þ C12 C34 g þ C212 C234
2
We need to find the expression for FðxÞ ¼ jPði xÞj2 jQði xÞj2
FðxÞ ¼ jPði xÞj2 jQði xÞj2 ¼ x8 þ x6 ð½C33 þ C11 2 2 C11 C33 Þ

1
þ x4 fC211 C233 ð C12 C34 s2 C12 C34 Þ2 g
2
x2 fðC11 C34 þ C12 C33 C12 C34 sÞ2
1
2 ð C12 C34 s2 C12 C34 Þ C12 C34 g C212 C234
2
N0 ¼ C212 C234 ; N2 ¼ ðC11 C34 þ C12 C33 C12 C34 sÞ2

1
þ 2 ð C12 C34 s2 C12 C34 Þ C12 C34
2
1
N4 ¼ C211 C233 ð C12 C34 s2 C12 C34 Þ2 ;
2
N6 ¼ ½C33 þ C11 2 2 C11 C33 ; N8 ¼ 1
X
4
Fðx; sÞ ¼ jPði xÞj2 jQði xÞj2 ¼ N2k x2k
k¼0
¼ N0 þ N2 x2 þ N4 x4 þ N6 x6 þ N8 x8
P4
Hence Fðx; sÞ ¼ 0 implies k¼0 N2k x
2k
¼ 0 and its roots are given by
solving the above polynomial. Furthermore PR ði x; sÞ ¼ x4 x2
C11 C33 ; PI ði x; sÞ ¼ x ½C33 þ C11
3
1
QR ði x; sÞ ¼ x2 ð C12 C34 s2 C12 C34 Þ þ C12 C34 ;
2
QI ði x; sÞ ¼ x ðC11 C34 þ C12 C33 C12 C34 sÞ
Hence
sin hðsÞ ¼
jQði x; sÞj2
And
cos hðsÞ ¼
jQði x; sÞj2
ðx4 x2 C11 C33 Þ ½x ðC11 C34 þ C12 C33 C12 C34 sÞ

þ x3 ½C33 þ C11 ½x2 ð12 C12 C34 s2 C12 C34 Þ þ C12 C34
sin hðsÞ ¼
x4 ð12 C12 C34 s2 C12 C34 Þ2 þ x2 fðC11 C34 þ C12 C33 C12 C34 sÞ2
2 ð12 C12 C34 s2 C12 C34 Þ C12 C34 g þ C212 C234
ðx4 x2 C11 C33 Þ ½x2 ð12 C12 C34 s2 C12 C34 Þ þ C12 C34
þ x4 ½C33 þ C11 ðC11 C34 þ C12 C33 C12 C34 sÞ
cos hðsÞ ¼ 4 1
x ð2 C12 C34 s2 C12 C34 Þ2 þ x2 fðC11 C34 þ C12 C33 C12 C34 sÞ2
2 ð12 C12 C34 s2 C12 C34 Þ C12 C34 g þ C212 C234
Those are continuous and differentiable in s based on Lemma 1.1. Hence we use
Theorem 1.2 and this approve the Theorem 1.3.
The third case we analyze is when there is a delay in first (i = 1) and second (i = 2)
RFID’s N-turn multilayer coil antennas voltages and antennas voltages derivatives
(s1 ¼ s2 ¼ sD ; D1 ¼ D2 ¼ sD ). The general characteristic equation Dðk; sD Þ is as
follow:
P
2 P
2 P
2
Dðk; s1 ; s2 ; D1 ; D2 Þ ¼ k þ C12 C34 e
4 i¼1 i¼1 þ k ½C11 C34 e i¼1
P2 P
2
kðs1 þ Di Þ k Di
þ C12 C33 e i¼1 þ k2 ½C11 C33 e i¼1 C12 ekðs1 þ D1 Þ
C34 ekðD2 þ s2 Þ
k3 ½C33 ekD2 þ C11 ekD1
Dðk; s1 ¼ s2 ¼ sD ; D1 ¼ D2 ¼ sD Þ ¼ k4 þ C12 C34 ek4sD þ k ½C11 C34 ek3sD

þ C12 C33 ek3sD þ k2 ½C11 C33 ek2sD C12 ek2sD C34 ek2sD
k3 ½C33 eksD þ C11 eksD
Dðk; s1 ¼ s2 ¼ sD ; D1 ¼ D2 ¼ sD Þ
¼ k4 þ C12 C34 ek4sD þ k ½C11 C34 þ C12 C33 ek3sD
þ k2 ½C11 C33 C12 C34 ek2sD k3 ½C33 þ C11 eksD
Dðk; s1 ¼ s2 ¼ sD ; D1 ¼ D2 ¼ sD Þ
¼ k4 þ ðC12 C34 ek3sD þ k ½C11 C34 þ C12 C33 ek2sD
þ k2 ½C11 C33 C12 C34 eksD k3 ½C33 þ C11 Þ eksD
The Maclaurin series is a Taylor series expansion of ek3sD ; ek2sD ; eksD

functions about zero (0). We get the following general characteristic equation
Dðk; sD Þ under Taylor series approximation:
ek3sD ’ 1 k 3 sD ; ek2sD ’ 1 k 2 sD ;
eksD ’ 1 k sD :
Dðk; s1 ¼ s2 ¼ sD ; D1 ¼ D2 ¼ sD Þ ¼ k4 þ ðC12 C34 ð1 k 3 sD Þ

þ k ½C11 C34 þ C12 C33 ð1 k 2 sD Þ þ k2 ½C11 C33 C12 C34 ð1 k sD Þ
k3 ½C33 þ C11 Þ eksD
Dðk; s1 ¼ s2 ¼ sD ; D1 ¼ D2 ¼ sD Þ ¼ k4 þ ðC12 C34 þ k ðC11 C34 þ C12 C33

C12 C34 3 sD Þ þ k2 ðC11 C33 C12 C34 ½C11 C34 þ C12 C33 2 sD Þ
k3 f½C11 C33 C12 C34 sD þ C33 þ C11 gÞ eksD
Dðk; s1 ¼ s2 ¼ sD ; D1 ¼ D2 ¼ sD Þ ¼ k4 þ ðC12 C34 þ k ðC11 C34 þ C12 C33

C12 C34 3 sD Þ þ k2 ðC11 C33 C12 C34 ½C11 C34 þ C12 C33 2 sD Þ
þ k3 f½C11 C33 þ C12 þ C34 sD C33 C11 gÞ eksD
Dðk; sD Þ ¼ Pn ðk; sD Þ þ Qm ðk; sD Þ eksD ; n ¼ 4; m ¼ 3; n [ m
The expression for Pn ðk; sD Þ:
p0 ðsD Þ ¼ 0; p1 ðsD Þ ¼ 0; p2 ðsD Þ ¼ 0; p3 ðsD Þ ¼ 0; p4 ðsD Þ ¼ 1
X
n¼4
Pn¼4 ðk; sD Þ ¼ pk ðsD Þ kk ¼ p0 ðsD Þ þ p1 ðsD Þ k þ p2 ðsD Þ k2 þ p3 ðsD Þ k3 þ p4 ðsD Þ k4
k¼0
Pn ðk; sD Þ ¼ k4
The expression for Qm ðk; sD Þ:
q0 ðsD Þ ¼ C12 C34 ; q1 ðsD Þ ¼ C11 C34 þ C12 C33 C12 C34 3 sD
q2 ðsD Þ ¼ C11 C33 C12 C34 ½C11 C34 þ C12 C33 2 sD ;

q3 ðsD Þ ¼ ½C11 C33 þ C12 þ C34 sD C33 C11
X
m¼3
Qm¼3 ðk; sD Þ ¼ qk ðsD Þ kk ¼ q0 ðsD Þ þ q1 ðsD Þ k þ q2 ðsD Þ k2 þ q3 ðsD Þ k3
k¼0
Qm ðk; sD Þ ¼ C12 C34 þ k ðC11 C34 þ C12 C33 C12 C34 3 sD Þ

þ k3 f½C11 C33 þ C12 þ C34 sD C33 C11 g
The homogenous system for Y1 VLcalc1 Y2 VLcalc2 lead to a characteristic

equation for the eigenvalue k having the form Pðk; sD Þ þ Qðk; sD Þ eksD ¼ 0
P
4 P
3
Pðk; sD Þ ¼ aj ðsD Þ k j ; Qðk; sD Þ ¼ cj ðsD Þ k j and the coefficients faj ðqi ; qk ; sÞ;
j¼0 j¼0
cj ðqi ; qk ; sÞg 2 R depend on qi ; qk and delay sD , qi ; qk are any RFID’s N-turn
multilayer coil antennas system’s parameters, other parameters keep as a constant.
a0 ¼ 0; a1 ¼ 0; a2 ¼ 0; a3 ¼ 0; a4 ¼ 1; c0 ¼ C12 C34 ;
c1 ¼ C11 C34 þ C12 C33 C12 C34 3 sD ;
c2 ¼ C11 C33 C12 C34 ½C11 C34 þ C12 C33 2 sD ;
c3 ¼ ½C11 C33 þ C12 þ C34 sD C33 C11
Unless strictly necessary, the designation of the variation arguments (qi ; qk ) will
subsequently be omitted from P, Q, aj, and cj. The coefficients aj, cj are continuous
and differentiable functions of their arguments, and direct substitution shows that
a0 þ c0 6¼ 0 ) C12 C34 6¼ 0 8 qi ; qk 2 R þ , k = 0 is not a Pðk; sD Þ þ
Qðk; sD Þ eksD ¼ 0. We assume that Pn ðk; sD Þ; Qm ðk; sD Þ can’t have common
imaginary roots. That is for any real number x: Pn ðk ¼ i x; sD Þ;
Qm ðk ¼ i x; sD Þ 6¼ 0.
Pn ðk; sD Þ ¼ x4 ;Qm ðk; sD Þ ¼ C12 C34 x2 ðC11 C33 C12 C34

½C11 C34 þ C12 C33 2 sD Þ þ i fx ðC11 C34 þ C12 C33 C12 C34 3 sD Þ
x3 ð½C11 C33 þ C12 þ C34 sD C33 C11 Þg
Pn ðk; sD Þ þ Qm ðk; sD Þ ¼ C12 C34 x ðC11 C33 C12 C34
2
½C11 C34 þ C12 C33 2 sD Þ þ x4 þ i fx ðC11 C34 þ C12 C33 C12 C34 3 sD Þ
x3 ð½C11 C33 þ C12 þ C34 sD C33 C11 Þg 6¼ 0
jPði x; sD Þj2 ¼ x8
jQði x; sD Þj2 ¼ fC12 C34 x2 ðC11 C33 C12 C34
½C11 C34 þ C12 C33 2 sD Þg2 þ fx ðC11 C34 þ C12 C33
C12 C34 3 sD Þ x3 ð½C11 C33 þ C12 þ C34 sD C33 C11 Þg2
jQði x; sD Þj2 ¼ C212 C234 þ x2 fðC11 C34 þ C12 C33 C12 C34 3 sD Þ2
2 ðC11 C34 þ C12 C33 C12 C34 3 sD Þ ð½C11 C33 þ C12
þ C34 sD C33 C11 Þg þ x6 ð½C11 C33 þ C12 þ C34 sD C33 C11 Þ2
We need to find the expression: Fðx; sD Þ ¼ jPði xÞj2 jQði xÞj2
Fðx; sD Þ ¼ jPði xÞj2 jQði xÞj2 ¼ C212 C234 x2 fðC11 C34 þ C12 C33 C12 C34 3 sD Þ2
2 ðC11 C34 þ C12 C33 C12 C34 3 sD Þ ð½C11 C33 þ C12 þ C34 sD C33 C11 Þg
x6 ð½C11 C33 þ C12 þ C34 sD C33 C11 Þ2 þ x8
We define the following parameters for simplicity: N0 ¼ C212 C234
N2 ¼ fðC11 C34 þ C12 C33 C12 C34 3 sD Þ2 2 C12 C34 ðC11 C33 C12 C34
½C11 C34 þ C12 C33 2 sD Þg
N4 ¼ fðC11 C33 C12 C34 ½C11 C34 þ C12 C33 2 sD Þ2
N6 ¼ ð½C11 C33 þ C12 þ C34 sD C33 C11 Þ2 ; N8 ¼ 1
X
4
Fðx; sD Þ ¼ jPði xÞj2 jQði xÞj2 ¼ N2k x2k
k¼0
¼ N0 þ N2 x2 þ N4 x4 þ N6 x6 þ N8 x8
P
4
Hence Fðx; sD Þ ¼ 0 implies N2k x2k ¼ 0 and its roots are given by solving
k¼0
the above polynomial. Furthermore, PR ði x; sD Þ ¼ x4 ; PI ði x; sD Þ ¼ 0
QR ði x; sD Þ ¼ C12 C34 x2 ðC11 C33 C12 C34 ½C11 C34 þ C12 C33 2 sD Þ
QI ði x; sD Þ ¼ x ðC11 C34 þ C12 C33 C12 C34 3 sD Þ
x3 ð½C11 C33 þ C12 þ C34 sD C33 C11 Þ
Hence
PR ði x; sD Þ QI ði x; sD Þ þ PI ði x; sD Þ QR ði x; sD Þ
sin hðsD Þ ¼
jQði x; sD Þj2
And
PR ði x; sD Þ QR ði x; sD Þ þ PI ði x; sD Þ QI ði x; sD Þ
cos hðsD Þ ¼
jQði x; sD Þj2
x4 fx ðC11 C34 þ C12 C33 C12 C34 3 sD Þ

x3 ð½C11 C33 þ C12 þ C34 sD C33 C11 Þg
sin hðsD Þ ¼ 2
þ x6 ð½C11 C33 þ C12 þ C34 sD C33 C11 Þ2
x4 fC12 C34 x2 ðC11 C33 C12 C34 ½C11 C34 þ C12 C33 2 sD Þg
cos hðsD Þ ¼
þ x6 ð½C11 C33 þ C12 þ C34 sD C33 C11 Þ2
Above expressions are continuous and differentiable in sD based on Lemma 1.1.

Hence we use Theorem 1.2 and this prove Theorem 1.3.
definition k ! j; pk ðsD Þ ! aj ; qk ðsD Þ ! cj ; n ¼ 4; m ¼ 3; n [ m additionally Pn ðk; sD Þ !
P P
Pðk; sD Þ; Qm ðk; sD Þ ! Qðk; sD Þ then Pðk; sD Þ ¼ 4j¼0 aj k j ; Qðk; sD Þ ¼ 2j¼0 cj k j .
PðkÞ ¼ k4 ; Qðk; sD Þ ¼ C12 C34 þ k ðC11 C34 þ C12 C33 C12 C34 3 sD Þ
þ k3 f½C11 C33 þ C12 þ C34 sD C33 C11 g

tion of sD such that a0 þ c0 6¼ 0. In the following “_” denotes complex and con-
jugate. PðkÞ; Qðk; sD Þ are analytic functions in k and differentiable in sD . The
coefficients:
faj ðC1 ; R1 ; RFID’sN-turn multilayer coil antennas system0s parametersÞ &

cj ðC1 ; R1 ; sD ; RFID’sN-turn multilayer coil antennas system0s parameters) g 2 R
depend on RFID’s N-turn multilayer coil antennas system’s parameters C1, R1, sD
values and antennas parameters. Unless strictly necessary, the designation of the
variation arguments. Resistor R1, capacitor C1, sD and N-turn multilayer coil
antennas system’s parameters will subsequently be omitted from P, Q, aj, cj. The
direct substitution shows that a0 þ c0 6¼ 0; C12 C34 6¼ 0.
g1 n1
C12 C34 6¼
g3 n3
P
2 P
2
ð Rdci Þ Rdci
1
Lcalc1 þ 1
R1
Lcalc1
i¼1 1
þ 1
ð i¼1
Lcalc2 Þ
¼½ qffiffiffiffiffiffiffiffiffiffi ½ Lcalc2 R1
qffiffiffiffiffiffiffiffiffiffi 6¼ 0
C1 ½1 þ Lcalc2
Lcalc1 þ 2 K LLcalc2
calc1
C1 ½Lcalc1
Lcalc2 þ 1 þ 2 K Lcalc2
Lcalc1
8C1 ; R1 ; RFID’sN-turn multilayer coil antennas system’s parameters 2 R þ
i.e. k ¼ 0 is not a root of characteristic equation. Furthermore PðkÞ; Qðk; sD Þ are

analytic function of k for which the following requirements of the analysis (see
kuang, 1993, Sect. 3.4) can also be verified in the present case [2–4].
(a) If k ¼ i x, x 2 R then Pði xÞ þ Qði xÞ 6¼ 0,
i.e. P and Q have no common imaginary roots. This condition was verified
numerically in the entire (R1, C1, N-turn multilayer coil antennas system
parameters) domain of interest.
(b) jQ(kÞ=PðkÞj is bounded for jkj ! 1, Rek 0. No roots bifurcation from 1.
C12 C34 þ k ðC11 C34 þ C12 C33 C12 C34 3 sD Þ þ k2 ðC11 C33
Q(kÞ C12 C34 ½C11 C34 þ C12 C33 2 sD Þ þ k3 f½C11 C33 þ C12 þ C34 sD C33 C11 g
j j¼j j
PðkÞ k4
Fðx; sD Þ ¼ jPði xÞj2 jQði xÞj2

¼ C212 C234 x2 fðC11 C34 þ C12 C33 C12 C34 3 sD Þ2
2 ðC11 C34 þ C12 C33 C12 C34 3 sD Þ
ð½C11 C33 þ C12 þ C34 sD C33 C11 Þg
x6 ð½C11 C33 þ C12 þ C34 sD C33 C11 Þ2 þ x8

in x8 ).
(d) Each positive root x (R1, C1, sD, N-turn multilayer coil antennas system
parameters) of FðxÞ ¼ 0 is continuous and differentiable with respect to
R1 ; C1 ; sD ; N-turn multilayer coil antennas system parameters. This condition
can only be assessed numerically.
Pði xÞ ¼ Pði xÞ, and Qði xÞ ¼ Qði xÞ thus k ¼ i x, x [ 0 may be on
of characteristic equation situated on the imaginary axis of the complex k-plane,
where by increasing the parameters R1, C1, sD, and N-turn multilayer coil antennas
system parameters, Rek may, at the crossing, Change its sign from (−) to (+), i.e.
ð0Þ ð0Þ ð0Þ ð0Þ
from a stable focus Eð0Þ ðY1 ; VLcalc1 ; Y2 ; VLcalc2 Þ ¼ ð0; 0; 0; 0Þ to an unstable one,
or vice versa. This feature may be further assessed by examining the sign of the
partial derivatives with respect to C1 ; R1 ; sD and N-turn multilayer coil antennas
parameters. ^1 ðR1 Þ ¼ ð@@Re
R Þk¼ix ; C1 ; sD ; N-turn multilayer coil antennas system
k
1
parameters are constant.
@Rek
^1 ðLcalc1 Þ ¼ ð Þ ; C1 ; R1 ; sD ¼ const;
@Lcalc1 k¼ix
@Rek
^1 ðLcalc2 Þ ¼ ð Þ ; C1 ; R1 ; sD ¼ const
@Lcalc2 k¼ix
@Rek
^1 ðsD Þ ¼ ð Þ ; C1 ; R1 ; N-turn multilayer coil antennas system’s parameters
@sD k¼ix
¼ const where x 2 R þ :
inserting k ¼ i x Into RFID N-turn multilayer coil antennas system’s character-
istic equation, x must satisfy the following:
sin x sD ¼ gðxÞ ¼
jQði xÞj2
and
cos x sD ¼ hðxÞ ¼ :
jQði xÞj2
Where jQði xÞj2 6¼ 0 in view of requirement (a) above, ðg; hÞ 2 R.

Furthermore, it follows above equations sin x sD and cos x sD that, by squaring
and adding the sides, x must be a positive root of FðxÞ ¼ jPði xÞj2
jQði xÞj2 ¼ 0. Note FðxÞ is dependent of sD . Now it is important to notice that if
sD 62 I (assume that I
R þ 0 is the set where xðsD Þ is a positive root of FðxÞ and for
sD 62 I; xðsD Þ is not define. Then for all sD in I xðsD Þ is satisfies that Fðx; sD Þ ¼ 0).
Then there are no positive xðsD Þ solutions for Fðx; sD Þ ¼ 0, and we cannot have
stability switches. For any sD 2 I, where xðsD Þ is a positive solution of
Fðx; sD Þ ¼ 0, we can define the angle hðsD Þ 2 ½0; 2 p as the solution
of sin hðsD Þ ¼ PR ðixÞQI ðixÞ þ PI ðixÞQR ðixÞ
jQðixÞj2
; cos hðsD Þ ¼ PR ðixÞQR ðixÞ þ PI ðixÞQI ðixÞ
jQðixÞj2
and the relation between the argument hðsD Þ and xðsD Þ sD for sD 2 I must be
xðsD Þ sD ¼ hðsD Þ þ n 2 p 8 n 2 N0 . Hence we can define the maps sDn : I !
Þ þ n2p
R þ 0 given by sDn ðsD Þ ¼ hðsDxðs DÞ
; n 2 N0 ; sD 2 I. Let us introduce the functions
I ! R; Sn ðsÞ ¼ sD sDn ðsD Þ; sD 2 I; n 2 N0 that are continuous and differentiable
in s. In the following, the subscripts k; x; R1 ; C1 and N-turn multilayer coil
antennas system parameters (a1 6¼ a2 ; N1 6¼ N2 ; h1 6¼ h2 ; b1 6¼ b2 ; ai ; Ni ; hi ;
bi 2 R þ ) indicate the corresponding partial derivatives. Let us first concentrate on
^ðxÞ, remember in
kða1 6¼ a2 ; N1 6¼ N2 ; h1 6¼ h2 ; b1 6¼ b2 ; ai ; Ni ; hi ; bi 2 R þ Þ
xða1 6¼ a2 ; N1 6¼ N2 ; h1 6¼ h2 ; b1 6¼ b2 ; ai ; Ni ; hi ; bi 2 R þ Þ
and keeping all parameters except one (x) and sD . The derivation closely follows
that in reference [BK]. Differentiating N-turn multilayer coil antennas system
characteristic equation PðkÞ þ QðkÞ eksD ¼ 0 with respect to specific parameter
(x), and inverting the derivative, for convenience, one calculates:
Remark:
x ¼ ai ; Ni ; hi ; bi ; a1 6¼ a2 ; N1 6¼ N2 ; h1 6¼ h2 ; b1 6¼ b2 ; ai ; Ni ; hi ; bi 2 R þ
@k 1 Pk ðk; xÞ Qðk; xÞ þ Qk ðk; xÞ Pðk; xÞ sD Pðk; xÞ Qðk; xÞ

ð Þ ¼
Where Pk ¼ @P
@k ; . . . etc., Substituting k ¼ i x, and bearing Pði xÞ ¼ Pði xÞ;
Qði xÞ ¼ Qði xÞ then i Pk ði xÞ ¼ Px ði xÞ and i Qk ði xÞ ¼ Qx ði xÞ that
on the surface jPði xÞj2 ¼ jQði xÞj2 , one obtains
@k i Px ði x; xÞ Pði x; xÞ þ i Qk ði x; xÞ Qðk; xÞ sD jPði x; xÞj2

ð Þ1 jk¼ix ¼ ð Þ:

Q ¼ QR þ i QI

¼ P2R þ P2I :
When (x) can be any N-turn multilayer coil antennas system parameters R1, C1,
and time delay sD etc. Where for convenience, we have dropped the arguments
ði x; xÞ, and where Fx ¼ 2 ½ðPRx PR þ PIx PI Þ ðQRx QR þ QIx QI Þ
Fx ¼ 2 ½ðPRx PR þ PIx PI Þ ðQRx QR þ QIx QI Þ ; xx ¼ Fx =Fx .
We define U and V:

We choose our specific parameter as time delay x = sD.
PR ði x; sD Þ ¼ x4 ; PI ði x; sD Þ ¼ 0
QR ði x; sD Þ ¼ C12 C34 x2 ðC11 C33 C12 C34
½C11 C34 þ C12 C33 2 sD Þ
QI ði x; sD Þ ¼ x ðC11 C34 þ C12 C33 C12 C34 3 sD Þ
x3 ð½C11 C33 þ C12 þ C34 sD C33 C11 Þ
PRx ¼ 4 x3 ; PIx ¼ 0; PRsD ¼ 0; PIsD ¼ 0; QRsD ¼ x2 ½C11 C34 þ C12 C33 2
QIsD ¼ x C12 C34 3 x3 ½C11 C33 þ C12 þ C34 ; PRx PR ¼ 4 x7 ;

PIx PI ¼ 0; xsD ¼ FsD =Fx
QRx ¼ 2 x ðC11 C33 C12 C34 ½C11 C34 þ C12 C33 2 sD Þ
QIx ¼ ðC11 C34 þ C12 C33 C12 C34 3 sD Þ

3 x2 ð½C11 C33 þ C12 þ C34 sD C33 C11 Þ
QRx QR ¼ 2 x ðC11 C33 C12 C34 ½C11 C34 þ C12 C33 2 sD Þ
fC12 C34 x2 ðC11 C33 C12 C34 ½C11 C34 þ C12 C33 2 sD Þg
QIx QI ¼ fðC11 C34 þ C12 C33 C12 C34 3 sD Þ

3 x2 ð½C11 C33 þ C12 þ C34 sD
C33 C11 Þg fx ðC11 C34 þ C12 C33 C12 C34 3 sD Þ
x3 ð½C11 C33 þ C12 þ C34 sD C33 C11 Þg
FsD ¼ 2 ½ðPRsD PR þ PIsD PI Þ ðQRsD QR þ QIsD QI Þ ; PRsD ¼ 0; PIsD ¼ 0

FsD ¼ 2 fx2 ½C11 C34 þ C12 C33 2 ½C12 C34 x2 ðC11 C33 C12 C34
½C11 C34 þ C12 C33 2 sD Þ þ ðx C12 C34 3 x3 ½C11 C33 þ C12
þ C34 Þ ½x ðC11 C34 þ C12 C33 C12 C34 3 sD Þ
x3 ð½C11 C33 þ C12 þ C34 sD C33 C11 Þ g
PR PIx ¼ 0; PI PRx ¼ 0; QR QIx ¼ fC12 C34 x2 ðC11 C33 C12 C34

½C11 C34 þ C12 C33 2 sD Þg fðC11 C34 þ C12 C33 C12 C34 3 sD Þ
3 x2 ð½C11 C33 þ C12 þ C34 sD C33 C11 Þg
QI QRx ¼ 2 x fx ðC11 C34 þ C12 C33 C12 C34 3 sD Þ

x3 ð½C11 C33 þ C12 þ C34 sD C33 C11 Þg fðC11 C33 C12 C34
½C11 C34 þ C12 C33 2 sD Þg
V ¼ ðPR PIsD PI PRsD Þ ðQR QIsD QI QRsD Þ; PR PIsD ¼ 0; PI PRsD ¼ 0
QR QIsD ¼ fC12 C34 x2 ðC11 C33 C12 C34

½C11 C34 þ C12 C33 2 sD Þg
x fC12 C34 3 þ x2 ½C11 C33 þ C12 þ C34 g
QI QRsD ¼ fx ðC11 C34 þ C12 C33 C12 C34 3 sD Þ
x3 ð½C11 C33 þ C12 þ C34 sD
C33 C11 Þg fx2 ½C11 C34 þ C12 C33 2g;Fðx; sD Þ ¼ 0
Differentiating with respect to sD and we get
@x @x Fs
Fx þ FsD ¼ 0; sD 2 I ) ¼ D
@sD @sD Fx
1 @Rek 1 2 ½U þ sD jPj2 þ i Fx @x Fs
^ ðsD Þ ¼ ð Þk¼ix ; ^ ðsD Þ ¼ Ref g; ¼ xs D ¼ D
@sD FsD þ i 2 ½V þ x jPj2 @sD Fx
@Rek
signf^1 ðsD Þg ¼ signfð Þ g;
@sD k¼ix
@x
@x U @s þV
signf^1 ðsD Þg ¼ signfFx g signfsD þxþ D
g
@sD jPj2

in a N-turn multilayer coil antennas system, about the equilibrium point
E ð0Þ ðY1 ; VLcalc1 ; Y2 ; VLcalc2 Þ as a result of a variation of delay parameter sD. The
uated on the imaginary axis of the complex k-plane where by increasing the delay
parameter sD, Re k may at the crossing, change its sign from − to +, i.e. from a
assessed by examining the sign of the partial derivatives with respect to sD,
^1 ðsD Þ ¼ ð@@sRekÞ
D k¼ix , C1, R1 and N-turn multilayer coil antennas system
parameters are constant where x 2 R þ . We need to plot the stability switch dia-
gram based on different delay values of our N-turn multilayer coil antennas system.
Since it is a very complex function we recommend to solve it numerically rather
than analytic.
@Rek 2 ½U þ sD jPj2 þ i Fx
^1 ðsD Þ ¼ ð Þk¼ix ¼ Ref g
@sD FsD þ i 2 ½V þ x jPj2
@Rek 2 fFx ðV þ x P2 Þ FsD ðU þ sD P2 Þg
^1 ðsD Þ ¼ ð Þk¼ix ¼
@sD Fs2 þ 4 ðV þ x P2 Þ2
D
The stability switch occurs only on those delay values (sD ) which fit the equa-
tion: sD ¼ xh þþ ðsðsDDÞÞ and h þ ðsD Þ is the solution of sin hðsÞ ¼ . . .; cos hðsÞ ¼ . . . when
x ¼ x þ ðsD Þ if only x þ is feasible. Additionally when all N-turn multilayer coil
antennas system parameters are known and the stability switch due to various time
delay values sD is describe in the following expression:
signf^1 ðsD Þg ¼ signfFx ðxðsD Þ; sD Þg signfsD xsD ðxðsD ÞÞ þ xðsD Þ

UðxðsD ÞÞ xsD ðxðsD ÞÞ þ VðxðsD ÞÞ
þ g
jPðxðsD ÞÞj2
Remark: we know Fðx; sD Þ ¼ 0 implies it roots xi ðsD Þ and finding those delays
values sD which xi is feasible. There are sD values which xi are complex or
imaginary numbers, then unable to analyse stability.
Lemma 1.1 Assume that xðsÞ is a positive and real root of Fðx; sÞ ¼ 0 defined for
s 2 I, which is continuous and differentiable. Assume further that if k ¼ i x,
x 2 R, then Pn ði x; sÞ þ Qn ði x; sÞ 6¼ 0; s 2 R hold true. The functions
Sn ðsÞ; n 2 N0 , are continuous and differentiable on I.
s 2 I; I
R þ 0 , and at some s 2 I, Sn ðs Þ ¼ 0 for some n 2 N0 then a pair of simple
conjugate pure imaginary roots k þ ðs Þ ¼ i xðs Þ; k ðs Þ ¼ i xðs Þ of
dRek dSn ðsÞ

dðs Þ ¼ signf j
g ¼ signfFx ðxðs Þ; s Þg signf j g
ds k¼ixðs Þ ds s¼s
Theorem 1.3 The characteristic equation has a pair of simple and conjugate pure
imaginary roots k ¼ xðs Þ; xðs Þ real at s 2 I if Sn ðs Þ ¼ s sn ðs Þ ¼ 0 for
some n 2 N0 . If xðs Þ ¼ x þ ðs Þ, this pair of simple conjugate pure imaginary
roots crosses the imaginary axis from left to right if d þ ðs Þ [ 0 and crosses the
imaginary axis from right to left if d þ ðs Þ\0 where d þ ðs Þ ¼
signfdRe
ds jk¼ix þ ðs Þ g
k
dRek dSn ðsÞ

d þ ðs Þ ¼ signf jk¼ix þ ðs Þ g ¼ signf j g
ds ds s¼s
If xðs Þ ¼ x ðs Þ, this pair of simple conjugate pure imaginary roots cross the
imaginary axis from left to right, if d ðs Þ [ 0 and crosses the imaginary axis from
dSn ðsÞ
right to left. If d ðs Þ\0 where d ðs Þ ¼ signfdRe ds jk¼ix ðs Þ g ¼ signf ds js¼s g
k
If x ðs Þ ¼ x ðs Þ ¼ xðs Þ then Dðs Þ ¼ 0 and signfdRek j

þ ds k¼ixðs Þ g ¼ 0, the
same is true when S0n ðs Þ ¼ 0 the following result can be useful in identifying
values of s where stability switches happened.
Remark: Lemma 1.1 and Theorems 1.2, 1.3: In the first and second cases
we discuss delay parameter s and in the third case we discuss delay parameter sD
[18, 19].
8.2 Double Rectangular Spiral Coils Antennas System

Stability Optimization Under Delayed
Electromagnetic Interferences and Parasitic Effects
A double rectangular spiral antennas are constructed from two antennas, each
antenna is a rectangular spiral antenna. Antennas are connected in series with micro
strip line and to the RFID IC. A double rectangular spiral antennas system influence
by electromagnetic interferences which effect there stability behavior. Additionally,
micro strip line which connected antennas in the series has parasitic effect, delay in
time Dl . We inspect our system performances under electromagnetic interferences
and micro strip parasitic effects. Generally double spiral micro strip antenna over
rectangular patch improved the bandwidth of Mobile, WiMAX applications. Micro
strip line feed technique enhances the bandwidth of the simple micro strip antenna.
Every rectangular spiral antenna has a parasitic DC resistance which needs to be
calculated. The below figure describes the double rectangular spiral antenna system
[1, 7, 8] (Fig. 8.3).
Index (i) indicates first rectangular spiral antenna (i = 1) or second rectangular
spiral antenna (i = 2). We define RFID’s coil dimensional parameters. A0i, B0i—
Overal dimensions of the coil. Aavgi, Bavgi—Average dimensions of the coil. ti—
track thickness. wi—track width. gi—gaps between tracks. Nci—number of turns.
di—equivalent diameter of the track. Average coil area Aci = Aavgi ∙ Bavgi. Pi—
coil manufacturing technology parameter. Integrating all those parameters gives the
equations for rectangular spiral antenna inductance calculation:
Fig. 8.3 Double rectangular spiral coils antennas system

8.2 Double Rectangular Spiral Coils Antennas System Stability … 819
l0 X4
2 Aavgi Bavgi
Lcalci ¼ ð Xki X3 Þ NciPi ; X1i ¼ Aavgi lnð qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi Þ
p k¼1;k6¼3 di ðAavgi þ A2avgi þ B2avgi Þ
2 Aavgi Bavgi qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
X2i ¼ Bavgi lnð qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi Þ; X3i ¼ 2 ðAavgi þ Bavgi A2avgi þ B2avgi Þ
di ðBavgi þ A2avgi þ B2avgi Þ
ðAavgi þ Bavgi Þ 2 ðti þ wi Þ
X4i ¼ ; di ¼ ; Aavgi ¼ A0i Nci ðgi þ wi Þ;
4 p
Bavgi ¼ B0i Nci ðgi þ wi Þ
The rectangular spiral antenna length is calculated as follows: l0i is the length of
the first turn l0i ¼ 2 ðA0i þ B0i Þ ðwi þ gi Þ. lk is the length of turn k + 1. We
define the following:
NX
ci 1
lTi ¼ l0i þ fA0i ½1 þ ðk 1Þ 2 ðwi þ gi Þ þ B0i ½2 þ ðk 1Þ 2 ðwi þ gi Þ

k¼1
þ A0i ½2 þ ðk 1Þ 2 ðwi þ gi Þ þ B0i ½3 þ ðk 1Þ 2 ðwi þ gi Þg
NX
C 1 NX
C 1
LTi ¼ L0i þ 2 ðA0i þ B0i Þ ðNCi 1Þ 8 ðwi þ gi Þ k; k ¼ NCi 1

k¼1 k¼1
LTi ¼ 2 fðA0i þ B0i Þ ð1 þ NCi Þ ðwi þ gi Þ ½4 NCi 3 g
The DC resistance of rectangular spiral RFID antenna: RDCi ¼ rliTiSi ¼ ri pa

lTi
2.
i
lTi—total length of the wire. ri—conductivity of the wire (mX/m). Si—Cross
section area p a2i . ai—radius of the wire.
Remark: a2i ¼ w2i .
2 fðA0i þ B0i Þ ð1 þ NCi Þ ðwi þ gi Þ ½4 NCi 3 g

RDCi ¼
ri p w2i
Due to electromagnetic interferences there are different in time delays respect to

first and second rectangular spiral antennas voltages and voltages derivatives.
The delayed voltages are V1(t − s1) and V2(t − s2) respectively (s1 6¼ s2) and
delayed voltages derivatives are dV1(t − D1)/dt, dV2(t − D2)/dt respectively
ðD1 6¼ D2 ; s1 0Þ;ðs2 0; D1 ; D2 0Þ. Additionally, there is a delay in time for the
micro strip parasitic effects Dl . The double rectangular spiral antenna system
equivalent circuit can represent as delayed differential equations which, depending
on variable parameters and delays. Our double rectangular spiral antennas system
delay differential and delay different model can be analytically by using delay
differential equations in dynamically. The need of the incorporation of a time delay
is often of the existence of any stage structure. It is often difficult to analytically
study models with delay dependent parameters, even if only a single discrete delay
is present. There are practical guidelines that combine graphical information with
analytical work to effectively study the local stability of models involving delay
dependent parameters. The stability of a given steady state is simply determined by
the graphs of some function of s1, s2 which can be expressed, explicitly and thus
can be easily depicted by MATLAB and other popular software. We need only look
at one such function and locate the zeros. This function often has only two zeros,
providing thresholds for stability switches. As time delay increases, stability
changes from stable to unstable to stable. We emphasize the local stability aspects
of some models with delay dependent parameters, additionally there is a general
geometric criterion that, theoretically speaking, can be applied to models with many
delays, or even distributed delays. The simplest case of a first order characteristic
equation, providing more user friendly geometric and analytic criteria for stability
switches. The analytical criteria provided for the first and second order cases can be
used to obtain some insightful analytical statements and can be helpful for con-
ducting simulations. Double rectangular spiral antennas can be represented as a two
inductors in series (Lcalc-1 and Lcalc-2), parasitic resistances (RDC-1 and RDC-2) and
micro strip delay line. The rectangular spiral antennas in series are connected in
parallel to RFID TAG IC. The Equivalent Circuit of Passive RFID TAG with
double rectangular antennas is Capacitor (C1) and Resistor (R1) in parallel with
double rectangular antennas in the series [85] (Fig. 8.4).
Lcalc-1 and Lcalc-2 are mostly formed by traces on planar PCB. 2 ∙ Lm element
represents the mutual inductance between Lcalc-1 and Lcalc-2. Since two inductors
(Lcalc-1, Lcalc-2) are in series and there is a mutual inductance between Lcalc-1 and
Lcalc-2, the total antenna inductance LT: LT = Lcalc-1 + Lcalc-2 +2 ∙ Lm and
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
Lm ¼ K Lcalc1 Lcalc2 . Lm is the mutual inductance between Lcalc-1 and Lcalc-2.
K is the coupling coefficient of two inductors 0 K 1. I(t) is the current that flow
through a double rectangular spiral antenna. V1 and V1 are the voltages on Lcalc-1
and Lcalc-2 respectively. Vm is the voltage on double loop antenna mutual induc-
tance element. We neglect the voltage on microstrip delay line Vl ! e. ILcalc1 ¼
IðtÞ ILcalc2 ¼ Iðt Dl Þ; IRdc2 ¼ Iðt Dl Þ; ILm ¼ Iðt Dl Þ.
Fig. 8.4 Double rectangular

spiral coils antennas system
equivalent circuit
V1 ðtÞ ¼ VLcalc1 ¼ VLcalc1 ðtÞ; V2 ðtÞ ¼ VLcalc2 ¼ VLcalc2 ðtÞ

Z
1 dVAB
VAB ¼ VR1 ¼ VC1 ; VAB ¼ R1 IR1 ; VAB ¼ IC1 dt; IC1 ¼ C1 ; IðtÞ þ IC1 þ IR1 ¼ 0
C1 dt
VAB dVAB
IðtÞ þ þ C1 ¼ 0; VAB ¼ VLcalc1 þ VRdc1 þ VLcalc2 þ VRdc2 þ Vm þ ðVl ! eÞ
R1 dt
lT1 lT2
VRdc1 ¼ IðtÞ Rdc1 ¼ IðtÞ ; VRdc2 ¼ Iðt Dl Þ Rdc2 ¼ Iðt Dl Þ
r1 p a21 r2 p a22
dIðtÞ dIðt Dl Þ
VLcalc1 ¼ Lcalc1 ; VLcalc2 ¼ Lcalc2 ;
dt dt
dIðt Dl Þ
Vm ¼ 2 L m ; Lcalc1 6¼ Lcalc2
dt
dVC1 VC
C1 þ 1 þ IðtÞ ¼ 0; VC1 ¼ VLcalc1 þ VRdc1 þ VLcalc2 þ VRdc2 þ Vm
dt R1
dVR1 dVC1 dVLcalc1 dVRdc1 dVLcalc2 dVRdc2 dVm
¼ ¼ þ þ þ þ
dt dt Z dt dt dt Z dt dt
1 1
IðtÞ ¼ VLcalc1 dt; Iðt Dl Þ ¼ VLcalc2 dt
Lcalc1 Lcalc2
dIðtÞ VLcalc1 dIðt Dl Þ VLcalc2 dIðt Dl Þ dIðtÞ
¼ ; ¼ ;j jX
dt Lcalc1 dt Lcalc2 dt dt
dIðt Dl Þ dIðtÞ VL VL
X ! e; ) calc1 ¼ calc2 ;
dt dt Lcalc1 Lcalc2
dIðt Dl Þ VL
Vm ¼ 2 Lm ¼ 2 Lm calc2
dt Lcalc2
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi VLcalc2 Lcalc1
Vm ¼ 2 K f Lcalc1 Lcalc2 g ¼2K VLcalc2
Lcalc2 Lcalc2
dVm pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi dVLcalc2 =dt Lcalc1 dVLcalc2
¼ 2 K f Lcalc1 Lcalc2 g ¼2K
dt Lcalc2 Lcalc2 dt
We get the following differential equation respect to VLcalc1 variable:

Z
1 VAB dVAB
VLcalc1 dt þ þ C1 ¼0
Lcalc1 R1 dt
Z
1 1
VLcalc1 dt þ ½VLcalc1 þ VRdc1 þ VLcalc2 þ VRdc2 þ Vm
Lcalc1 R1
dVL dVRdc1 dVLcalc2 dVRdc2 dVm
þ C1 ½ calc1 þ þ þ þ ¼0
dt dt dt dt dt
VRdc1 ¼ IðtÞ Rdc1 ; VRdc2 ¼ Iðt Dl Þ Rdc2
Z
1 1
VLcalc1 dt þ
½VLcalc1 þ IðtÞ Rdc1 þ VLcalc2
Lcalc1 R1
dVL dIðtÞ
þ Iðt Dl Þ Rdc2 þ Vm þ C1 ½ calc1 þ Rdc1
dt
dt rffiffiffiffiffiffiffiffiffiffiffiffiffi
dVLcalc2 dIðt Dl Þ Lcalc1 dVLcalc2
þ þ Rdc2 þ 2 K ¼0
dt dt Lcalc2 dt
VLcalc1 dIðtÞ VLcalc2 dIðt Dl Þ

¼ ; ¼
Lcalc1 dt Lcalc2 dt
Z Z Z
1 1 Rdc1 Rdc2

VLcalc1 dt þ ½VLcalc1 þ VLcalc1 dt þ VLcalc2 þ VLcalc2 dt
Lcalc1 dVLcalc1 VLcalc1 dVLcalc2 VL
þ2 K VLcalc2 þ C1 ½ þ Rdc1 þ þ calc2 Rdc2
Lcalc2 dt Lcalc1 dt Lcalc2
Lcalc1 dVLcalc2
þ2 K ¼0
Lcalc2 dt
We derivative in time the two sides of the above equation.
VLcalc1 1 dVLcalc1 Rdc1 dVLcalc2 Rdc2

þ ½ þ VLcalc1 þ þ VLcalc2
Lcalc1 R1 dt Lcalc1 dt Lcalc2
Lcalc1 dVLcalc2 d 2 VLcalc1 Rdc1 dVLcalc1 d 2 VLcalc2
þ2 K þ C1 ½ þ þ
Lcalc2 dt dt2 Lcalc1 dt dt2
rffiffiffiffiffiffiffiffiffiffiffiffiffi 2
Rdc2 dVLcalc2 Lcalc1 d VLcalc2
þ þ2 K ¼0
Lcalc2 dt Lcalc2 dt2
Lcalc2
VLcalc2 ¼ VLcalc1
Lcalc1
VLcalc1 1 dVLcalc1 Rdc1 Lcalc2 dVLcalc1 Rdc2

þ þ VLcalc1 þ þ VLcalc1
Lcalc1 R1 dt R1 Lcalc1 R1 Lcalc1 dt R1 Lcalc1
Lcalc2 Lcalc1 dVLcalc1 d 2 VLcalc1 C1 Rdc1 dVLcalc1
þ2K þ C1 þ
R1 Lcalc1 Lcalc2 dt dt2 Lcalc1 dt
C1 Lcalc2 d 2 VLcalc1 C1 Lcalc2 Rdc2 dVLcalc1
þ þ
Lcalc1 dt2 L Lcalc2 dt
rffiffiffiffiffiffiffiffiffiffiffiffiffi calc1
C1 Lcalc2 Lcalc1 d 2 VLcalc1
þ2K ¼0
Lcalc1 Lcalc2 dt2
1 ðRdc1 þ Rdc2 Þ 1 Lcalc2 1 Lcalc2
½ þ VLcalc1 þ ½ þ þ2 K
Lcalc1 R1 Lcalc1 R1 R1 Lcalc1 R1 Lcalc1
C1 ðRdc1 þ Rdc2 Þ dVLcalc1 Lcalc2 Lcalc2 d 2 VLcalc1
þ þ C1 ½1 þ þ2K ¼0
Lcalc1 dt Lcalc1 Lcalc1 dt2
We get the following differential equation respect to VLcalc1 variable, g1 ; g2 ; g3

g1 VLcalc1 þ g2 þ g3 ¼0
dt dt2
1 ðRdc1 þ Rdc2 Þ
g1 ¼ þ ;
Lcalc1 R1 Lcalc1
Lcalc2 1 Lcalc2 C1 ðRdc1 þ Rdc2 Þ 1
g2 ¼ þ2 K þ þ
R1 Lcalc1 R1 Lcalc1 Lcalc1 R1
Lcalc2 Lcalc2
g3 ¼ C1 ½1 þ þ2 K g ;
Lcalc1 Lcalc1
g1 ¼ g1 ðR1 ; Lcalc1 Rdc1 ; Rdc2 Þ
g2 ¼ g2 ðLcalc1 ; Lcalc2 ; R1 ; C1 ; K; Rdc1 ; Rdc2 Þ;
g3 ¼ g3 ðLcalc1 ; Lcalc2 ; C1 ; KÞ
dVLcalc1 dX1 d 2 VLcalc1 dX1 g g dVLcalc1

X1 ¼ ; ¼ ; ¼ 1 VLcalc1 2 X1 ; ¼ X1
dt dt dt2 dt g3 g3 dt
In the same manner we find our VLcalc2 differential equation. We get the fol-
lowing differential equation respect to VLcalc1 variable, n1 ; n2 ; n3 are global
parameters.
VLcalc1 1 dVLcalc1 Rdc1 dVLcalc2 Rdc2

þ ½ þ VLcalc1 þ þ VLcalc2
Lcalc1 R1 dt Lcalc1 dt Lcalc2
Lcalc1 dVLcalc2 d 2 VLcalc1 Rdc1 dVLcalc1
þ2 K þ C1 ½ 2
þ
Lcalc2 dt dt Lcalc1 dt
d 2 VLcalc2 Rdc2 dVLcalc2 Lcalc1 d 2 VLcalc2
þ þ þ2 K ¼0
dt2 Lcalc2 dt Lcalc2 dt2
Lcalc1
VLcalc1 ¼ VLcalc2
Lcalc2
1 1 Lcalc1 dVLcalc2 Rdc1 dVLcalc2 Rdc2
VLcalc2 þ ½ þ VLcalc2 þ þ VLcalc2
Lcalc2 R1 Lcalc2 dt Lcalc2 dt Lcalc2
Lcalc1 dVLcalc2 Lcalc1 d 2 VLcalc2 Rdc1 dVLcalc2 d 2 VLcalc2
þ2 K þ C1 ½ 2
þ þ
Lcalc2 dt Lcalc2 dt Lcalc2 dt dt2
Rdc2 dVLcalc2 Lcalc1 d VLcalc2
þ þ2 K ¼0
1 Lcalc1 dVLcalc2 Rdc1 1 dVLcalc2 Rdc2

VLcalc2 þ þ VLcalc2 þ þ VLcalc2
Lcalc2 R1 Lcalc2 dt R1 Lcalc2 R1 dt R1 Lcalc2
2K Lcalc1 dVLcalc2 C1 Lcalc1 d 2 VLcalc2 C1 Rdc1 dVLcalc2 d 2 VLcalc2
þ þ þ þ C1
R1 Lcalc2 dt Lcalc2 dt2 Lcalc2 dt dt2
C1 Rdc2 dVLcalc2 Lcalc1 d VLcalc2
þ þ 2 K C1 ¼0
1 ðRdc1 þ Rdc2 Þ Lcalc1 1 2K Lcalc1
½ þ VLcalc2 þ ½ þ þ
Lcalc2 R1 Lcalc2 R1 Lcalc2 R1 R1 Lcalc2
C1 ðRdc1 þ Rdc2 Þ dVLcalc2 Lcalc1 Lcalc1 d 2 VLcalc2
þ þ C1 ½1 þ þ2 K ¼0
Lcalc2 dt Lcalc2 Lcalc2 dt2
1 ðRdc1 þ Rdc2 Þ Lcalc1 1 2K Lcalc1
n1 ¼ þ ; n2 ¼ þ þ
Lcalc2 R1 Lcalc2 R1 Lcalc2 R1 R1 Lcalc2
C1 ðRdc1 þ Rdc2 Þ
þ
Lcalc2
Lcalc1 Lcalc1
n3 ¼ C1 ½1 þ þ2 K ; n ¼ ðLcalc1 ; Lcalc2 ; R1 ; Rdc1 ; Rdc2 Þ
Lcalc2 Lcalc2 1
n2 ¼ ðLcalc1 ; Lcalc2 ; R1 ; Rdc1 ; Rdc2 ; C1 ; KÞ; n3 ¼ ðLcalc1 ; Lcalc2 ; C1 ; KÞ
We get the following differential equation respect to VLcalc2 variable, n1 ; n2 ; n3

dVLcalc2 d 2 VLcalc2 dVLcalc2 dX2 d 2 VLcalc2

n1 VLcalc2 þ n2 þ n3 ¼ 0; X2 ¼ ; ¼
dt dt2 dt dt dt2
dX2 n1 n2 dVLcalc2
¼ VLcalc2 X2 ; ¼ X2
dt n3 n3 dt
Summary: We get our RFID TAGs with double rectangular spiral antenna
system’s four differential equations.
dX1 g g dVLcalc1 dX2

¼ 1 VLcalc1 2 X1 ; ¼ X1 ;
dt g3 g3 dt dt
n1 n2 dVLcalc2
¼ VLcalc2 X2 ; ¼ X2
n3 n3 dt
0 1
dX1
dt 0 1 0 X1 1
B dVL C C11 ... C14
B calc1 C B BV C
B dt C B .. .. .. C
C B Lcalc1 C g2 g1 n2
B dX2 C¼@ . . . AB C; C11 ¼ ; C12 ¼ ; C33 ¼ ;
B C @ X2 A g3 g3 n3
@ dt A
dVLcalc2 C41 C44 VLcalc2
dt
n1
C3 4 ¼ ; C21 ¼ C43 ¼ 1
n3
C13 ¼ C14 ¼ C22 ¼ C23 ¼ C24 ¼ C31 ¼ C32 ¼ C41 ¼ C42 ¼ C44 ¼ 0
The RFID TAGs with double rectangular spiral antenna system’s first and
second rectangular spiral antenna are composed of a thin wire or a thin plate
element. Units are all in cm, and a1, a2 are radiuses of the first and second wires in
cm. There inductances can be calculated by the following formulas:
l0 X
4
l X
4
Lcalc1 ¼ ð Xk1 X3 Þ Nc1
P1
; Lcalc2 ¼ 0 ð Xk2 X3 Þ Nc2
P2
p k¼1;k6¼3 p k¼1;k6¼3

first and second rectangular spiral antennas voltages and voltages derivatives. The
delayed voltages are V1(t − s1) and V2(t − s2) respectively (s1 6¼ s2) and delayed
voltages derivatives are dV1(t − D1)/dt, dV2(t − D2)/dt respectively.
ðD1 6¼ D2 ; s1 0; s2 0; D1 ; D2 0Þ; V1 ðtÞ ¼ VLcalc1 ¼ VLcalc1 ðtÞ;

V2 ðtÞ ¼ VLcalc2 ¼ VLcalc2 ðtÞ
VLcalc1 ðtÞ ! VLcalc1 ðt s1 Þ; VLcalc2 ðtÞ ! VLcalc2 ðt s2 Þ;
X1 ðtÞ ! X1 ðt D1 Þ; X2 ðtÞ ! X2 ðt D2 Þ
dV dV
We consider no delay effect on dX dt ; ; dt ; Ldt
Lcalc1 dX2
1
dt
calc2
. The RFID TAGs with
double rectangular spiral antenna system differential equations under electromag-
netic interferences (delays terms) influence only RFID first and second rectangular
spiral antenna voltages VLcalc1 ðtÞ; VLcalc2 ðtÞ and voltages derivatives X1 ðtÞ and X2 ðtÞ
dX1 ðtÞ dVLcalc1 ðtÞ dX2 ðtÞ dVLcalc2 ðtÞ
respect to time, there is no influence on dt ; dt ; dt ; dt .
0 1
dX1 0 1 0 X ðt D Þ 1
dt
B dVLcalc1 C C11 . . . C14 1 1
B dt C B .. .. .. C B B VLcalc1 ðt s1 Þ C
C
B dX2 C¼@ . . . A @ X2 ðt D2 Þ A
@ dt A
dVLcalc2 C41 C44 VLcalc2 ðt s2 Þ
dt
To find equilibrium points (fixed points) of the RFID TAGs with double rect-
angular spiral antenna system is by
lim VLcalc1 ðt s1 Þ ¼ VLcalc1 ðtÞ; lim VLcalc2 ðt s2 Þ ¼ VLcalc2 ðtÞ

t!1 t!1
lim X1 ðt D1 Þ ¼ lim X1 ðtÞ; lim X2 ðt D2 Þ ¼ lim X2 ðtÞ
t!1 t!1 t!1 t!1
dX1 ðtÞ dVLcalc1 ðtÞ dX2 ðtÞ dVLcalc2 ðtÞ
¼ 0; ¼ 0; ¼ 0; ¼ 0; t s1 ; t s2 ; t D1 ; t D2
dt dt dt dt

We get four equations and the only fixed point is Eð0Þ ðX1 ; VLcalc1 ; X2 ; VLcalc2 Þ
E ð0Þ ðX1 ; VLcalc1 ; X2 ; VLcalc2 Þ ¼ ð0; 0; 0; 0Þ, and since
g3 6¼ 0 & g2 6¼ 0 ) C11 6¼ 0; n3 ¼6 0 & n1 6¼ 0 ) C34 6¼ 0; g3 6¼ 0 & g1 6¼ 0 ) C12

6¼ 0; n3 6¼ 0 & n2 6¼ 0 ) C33 ¼6 0
equilibrium points of RFID TAGs with double rectangular spiral antenna system
consists in adding to coordinates ½X1 VLcalc1 X2 VLcalc2 arbitrarily small increments of
exponential ½x1 vLcalc1 x2 vLcalc2 ekt , and retaining the first order terms in
X1 VLcalc1 X2 VLcalc2 . The system of four homogeneous equations leads to a polyno-
mial characteristics equation in the eigenvalues k. The polynomial characteristics
equations accept by set the below voltages and voltages derivative respect to time
into two RFID TAGs with double rectangular spiral antenna system equations.
RFID TAGs with double rectangular spiral antenna system fixed values with
arbitrarily small increments of exponential form ½x1 vLcalc1 x2 vLcalc2 ekt are: i = 0
(first fixed point), i = 1 (second fixed point), i = 2 (third fixed point), etc.,
ðiÞ ðiÞ
X1 ðtÞ ¼ X1 þ x1 ekt ; VLcalc1 ðtÞ ¼ VLcalc1 þ vLcalc1 ekt
ðiÞ ðiÞ
X2 ðtÞ ¼ X2 þ x2 ekt ; VLcalc2 ðtÞ ¼ VLcalc2 þ vLcalc2 ekt
We choose the above expressions for our X1 ðtÞ; VLcalc1 ðtÞ and X2 ðtÞ; VLcalc2 ðtÞ as
small displacement ½x1 vLcalc1 x2 vLcalc2 from the system fixed points at time t = 0.
ðiÞ ðiÞ
X1 ðt ¼ 0Þ ¼ X1 þ x1 ; VLcalc1 ðt ¼ 0Þ ¼ VLcalc1 þ vLcalc1
ðiÞ ðiÞ
X2 ðt ¼ 0Þ ¼ X2 þ x2 ; VLcalc2 ðt ¼ 0Þ ¼ VLcalc2 þ vLcalc2
For k\0; t [ 0 the selected fixed point is stable otherwise k [ 0; t [ 0 is

k\0; t [ 0 otherwise go away from the selected fixed point exponentially. k is
the eigenvalue parameter which establish if the fixed point is stable or Unstable,
additionally his absolute value (jkj) establish the speed of flow toward or away from
the selected fixed point (Table 8.2).
Table 8.2 RFID TAGs with double rectangular spiral antennas variables for different k and t
values
k<0 k>0
t=0 ðiÞ ðiÞ
X1 ðt ¼ 0Þ ¼ X1 þ x1 X1 ðt ¼ 0Þ ¼ X1 þ x1
ðiÞ ðiÞ
ðiÞ ðiÞ
X2 ðt ¼ 0Þ ¼ X2 þ x2 X2 ðt ¼ 0Þ ¼ X2 þ x2
ðiÞ ðiÞ
ðiÞ ðiÞ
t>0 X1 ðtÞ ¼ X1 þ x1 ejkjt X1 ðtÞ ¼ X1 þ x1 ejkjt
ðiÞ ðiÞ
ðiÞ ðiÞ
X2 ðtÞ ¼ X2 þ x1 ejkjt X2 ðtÞ ¼ X2 þ x1 ejkjt
ðiÞ ðiÞ
ðiÞ
t>0 X1 ðt ! 1Þ ¼ X1 X1 ðt ! 1; k [ 0Þ x1 ejkjt
t!∞
ðiÞ VLcalc1 ðt ! 1; k [ 0Þ vLcalc1 ejkjt
X2 ðt ! 1Þ ¼ X2
ðiÞ X2 ðt ! 1; k [ 0Þ x2 ejkjt
ðiÞ VLcalc2 ðt ! 1; k [ 0Þ vLcalc2 ejkjt
The speeds of flow toward or away from the selected fixed point for RFID TAGs
with double rectangular spiral antenna system voltages and voltages derivatives
respect to time are as follow:
h i
ðiÞ ðiÞ
dX1 ðtÞ X1 ðt þ DtÞ X1 ðtÞ X1 þ x1 ekðt þ DtÞ X1 þ x1 ekt
¼ lim ¼ lim
dt Dt!0 Dt Dt!0 Dt

x1 ekt ekDt 1 ekDt 1 þ kDt
¼ lim ! k x1 ekt
Dt!0 Dt
dVLcalc1 ðtÞ dVLcalc2 ðtÞ dX2 ðtÞ

¼ k vLcalc1 ekt ; ¼ k vLcalc2 ekt ; ¼ k x2 ekt
dt dt dt
First we take the RFID TAGs with double rectangular spiral antenna system
dV dV
voltages VLcalc1 ; VLcalc2 differential equations: X1 ¼ Ldt calc1
; X2 ¼ Ldt
calc2
and add-
ing coordinates ½X1 VLcalc1 X2 VLcalc2 arbitrarily small increments of exponential
terms ½x1 vLcalc1 x2 vLcalc2 ekt and retaining the first order terms in x1 vLcalc1 x2 vLcalc2 .
ðiÞ ði¼0Þ
k vLcalc1 ekt ¼ X1 þ x1 ekt ; X1 ¼ 0 ) k vLcalc1 þ x1 ¼ 0
ðiÞ ði¼0Þ
k vLcalc2 e kt
¼ X2 þ x2 e kt
; X2 ¼ 0 ) k vLcalc2 þ x2 ¼ 0
Second we take the RFID TAGs with double rectangular spiral antenna system’s
voltages derivatives X1 ; X2 differential equations:
dX1 dX2
¼ C12 VLcalc1 þ C11 X1 ; ¼ C34 VLcalc2 þ C33 X2
dt dt
Adding coordinates ½X1 VLcalc1 X2 VLcalc2 arbitrarily small increments of expo-

nential terms ½x1 vLcalc1 x2 vLcalc2 ekt and retaining the first order terms in
x1 vLcalc1 x2 vLcalc2 .
ðiÞ ðiÞ
k x1 ekt ¼ C12 ½VLcalc1 þ vLcalc1 ekt þ C11 ½X1 þ x1 ekt
ðiÞ ðiÞ
k x1 ekt ¼ C12 VLcalc1 þ C11 X1 þ C12 vLcalc1 ekt þ C11 x1 ekt
ðiÞ ðiÞ
At fixed points C12 VLcalc1 þ C11 X1 ¼ 0
ðiÞ ðiÞ
C12 VLcalc1 þ C11 X1 ¼ 0 ) k x1 þ C11 x1 þ C12 vLcalc1 ¼ 0
ðiÞ ðiÞ
k x2 ekt ¼ C34 VLcalc2 þ vLcalc2 ekt þ C33 ½X2 þ x2 ekt
ðiÞ ðiÞ
k x2 ekt ¼ C34 VLcalc2 þ C33 X2 þ C34 vLcalc2 ekt þ C33 x2 ekt
ðiÞ ðiÞ
At fixed points C34 VLcalc2 þ C33 X2 ¼ 0
ðiÞ ðiÞ
C34 VLcalc2 þ C33 X2 ¼ 0 ) k x2 þ C33 x2 þ C34 vLcalc2 ¼ 0
Double rectangular spiral coils system arbitrarily small increments equations:
k vLcalc1 þ x1 ¼ 0; k vLcalc2 þ x2 ¼ 0; k x1 þ C11 x1 þ C12 vLcalc1 ¼ 0

k x2 þ C33 x2 þ C34 vLcalc2 ¼ 0
We define the following expressions:
ðiÞ ðiÞ
X1 ðt D1 Þ ¼ X1 þ x1 ekðtD1 Þ ; VLcalc1 ðt s1 Þ ¼ VLcalc1 þ vLcalc1 ekðts1 Þ
ðiÞ ðiÞ
X2 ðt D2 Þ ¼ X2 þ x2 ekðtD2 Þ ; VLcalc2 ðt s2 Þ ¼ VLcalc2 þ vLcalc2 ekðts2 Þ
Then we get four delayed differential equations respect to coordinates

½X1 VLcalc1 X2 VLcalc2 arbitrarily small increments of exponential
½x1 vLcalc1 x2 vLcalc2 ekt .
k ekt x1 ¼ C11 ekðtD1 Þ x1 þ C12 ekðts1 Þ vLcalc1 ; k ekt vLcalc1 ¼ ekðtD1 Þ x1

k ekt x2 ¼ C33 ekðtD2 Þ x2 þ C34 ekðts2 Þ vLcalc2 ; k ekt vLcalc2 ¼ ekðtD2 Þ x2
ði¼0Þ ði¼0Þ ði¼0Þ ði¼0Þ

In the equilibrium fixed point X1 ¼ 0; VLcalc1 ¼ 0; X2 ¼ 0; VLcalc2 ¼ 0.
The small increments Jacobian of our RFID TAGs with double rectangular spiral
antenna system is as bellow:
!11 ¼ k þ C11 ekD1 ; !12 ¼ C12 eks1 ; !13 ¼ 0; !14 ¼ 0; !21 ¼ ekD1 ;
!22 ¼ k; !23 ¼ 0; !24 ¼ 0
!31 ¼ 0; !32 ¼ 0; !33 ¼ k þ C33 ekD2 ; !34 ¼ C34 eks2 ;
!41 ¼ 0; !42 ¼ 0; !43 ¼ ekD2 ; !44 ¼ k
0 1
0 1 x1 0 1
!11 . . . !14 B C !11 ... !14
B . C B vLcalc1 C B .
B . . .
.. C B C .. .. C
@ . .. AB C ¼ 0; A k I ¼ B@ .. .
C
B x1 C
@ A
!41 !44 !41 !44
vLcalc2
P
2 P2
k½ si þ Dj
Dðk; s1 ; s2 ; D1 ; D2 Þ ¼ k4 þ C12 C34 e i¼1 j¼1
P
2 P
2
k½s2 þ Dj k½s1 þ Dj
þ k fC11 C34 e j¼1
þ C33 C12 e j¼1
g
P
2
k Dj
kðD2 þ s2 Þ kðD1 þ s1 Þ
þ k fC34 e
2
C12 e þ C11 C33 e j¼1
g
k3 fC33 ekD2 þ C11 ekD1 g
We have three stability cases: s1 ¼ s2 ¼ s & D1 ¼ D2 ¼ 0 Or s1 ¼ s2 ¼

they are positive parameters. There are other possible simple stability cases:
s1 ¼ s; s2 ¼ 0; D1 ¼ D2 ¼ 0 or s1 ¼ 0; s2 ¼ s; D1 ¼ D2 ¼ 0
s1 ¼ s2 ¼ 0; D1 ¼ D; D2 ¼ 0 or s1 ¼ s2 ¼ 0; D1 ¼ 0; D2 ¼ D
of value of the time delays s; D; sD for the general characteristic equation
Dðk; s=D=sD Þ. If we choose s parameter then Dðk; sÞ ¼ Pn ðk; sÞ þ Qm ðk; sÞ eks .
P
The expression for Pn ðk; sÞ :Pn ðk; sÞ ¼ nk¼0 Pk ðsÞ kk ¼ P0 ðsÞ þ P1 ðsÞ k þ
P2 ðsÞ k þ P3 ðsÞ k þ . . .. . .
2 3
Pm
The expression for Qm ðk; sÞ is Qm ðk; sÞ ¼ qk ðsÞ kk ¼ q0 ðsÞ þ
k¼0
q1 ðsÞ k þ q2 ðsÞ k2 þ . . .. . .
The case we analyze is when there is delay in RFID TAGs first and second
rectangular spiral antennas voltages (s1 ¼ s2 ¼ s) and no delay in RFID TAGs first
and second rectangular spiral antennas voltages derivatives. The general charac-
teristic equation D(k, s) is as follow:
Dðk; sÞ ¼ k4 k3 ðC33 þ C11 Þ þ k2 C11 C33 þ fC12 C34 eks

þ k ðC11 C34 þ C12 C33 Þ k2 ðC34 þ C12 Þg eks
Under Taylor series approximation: eks 1 k s þ 12 k2 s2

The Maclaurin series is a Taylor series expansion of a eks function about zero
(0). We get the following general characteristic equation D(k, s) under Taylor series
Dðk; sÞ ¼ k4 k3 ½C33 þ C11 þ k2 C11 C33 þ fC12 C34 þ k ½C11 C34

1
þ C12 C33 C12 C34 s þ k2 ½ C12 C34 s2 C34 C12 g eks
2
Pn
The expression for Pn ðk; sÞ: Pn ðk; sÞ ¼ k¼0 Pk ðsÞ k
k
X
n
Pn ðk; sÞ ¼ Pk ðsÞ k ¼ P0 ðsÞ þ P1 ðsÞ k þ P2 ðsÞ k þ P3 ðsÞ k þ P4 ðsÞ k
k 2 3 4
k¼0
¼ k k3 ½C33 þ C11 þ k2 C11 C33

4
P0 ðsÞ ¼ 0; P1 ðsÞ ¼ 0; P2 ðsÞ ¼ C11 C33 ; P3 ðsÞ ¼ ½C33 þ C11 ; P4 ðsÞ ¼ 1

Pm
The expression for Qm ðk; sÞ: Qm ðk; sÞ ¼ k¼0
qk ðsÞ kk ¼ q0 ðsÞ þ q1 ðsÞ k
þ q2 ðsÞ k2
X
m
Qm ðk; sÞ ¼ qk ðsÞ kk ¼ C12 C34 þ k ½C11 C34 þ C12 C33
k¼0
1
C12 C34 s þ k2 ½ C12 C34 s2 C34 C12 ; q0 ðsÞ ¼ C12 C34
2
1
q1 ðsÞ ¼ C11 C34 þ C12 C33 C12 C34 s; q2 ðsÞ ¼ C12 C34 s2 C34 C12
2
The homogeneous system for X1 VLcalc1 X2 VLcalc2 leads to a characteristic equa-

tion for the eigenvalue k having the form Pðk; sÞ þ Qðk; sÞ eks ¼ 0; PðkÞ ¼
P
4 P
2
aj k j ; QðkÞ ¼ cj k j and the coefficients faj ðqi ; qk ; sÞ; cj ðqi ; qk ; sÞg 2 R
j¼0 j¼0
depend on qi ; qk and delay s, qi ; qk are any double rectangular spiral coils antennas
system’s parameters, other parameters keep as a constant [5, 6].
a0 ¼ 0; a1 ¼ 0; a2 ¼ C11 C33 ; a3 ¼ ½C33 þ C11 ; a4 ¼ 1

c0 ¼ C12 C34 ; c1 ¼ C11 C34 þ C12 C33 C12 C34 s;
1
c2 ¼ C12 C34 s2 C34 C12
2
The designation of the variation arguments ðqi ; qk Þ will subsequently be omitted

from P, Q, aj, cj. The coefficients aj, cj are continuous, and differentiable functions
of their arguments, and direct substitution shows that a0 + c0 6¼ 0 for 8 qi ; qk 2 R þ ,
i.e. k = 0 is not a Pðk; sÞ þ Qðk; sÞ eks ¼ 0. We assume that Pn ðk; sÞ and
Qm ðk; sÞ can’t have common imaginary roots. That is for any real number x:
pn ðk ¼ i x; sÞ þ Qm ðk ¼ i x; sÞ 6¼ 0; pn ðk ¼ i x; sÞ
¼ x4 þ i x3 ðC33 þ C11 Þ x2 C11 C33
Qm ðk ¼ i x; sÞ ¼ C12 C34 þ i x ½C11 C34 þ C12 C33 C12 C34 s
1
x2 ½ C12 C34 s2 C34 C12
2
1
¼ x4 x2 ½ C12 C34 s2 C34 C12 þ C11 C33
2
þ C12 C34 þ i x3 ðC33 þ C11 Þ
þ i x ½C11 C34 þ C12 C33 C12 C34 s 6¼ 0
jPði x; sÞj2 ¼ x8 þ x6 fðC33 þ C11 Þ2 2 C11 C33 g þ x4 C211 C233

jQði x; sÞj2 ¼ C212 C234 þ x2 f½C11 C34 þ C12 C33 C12 C34 s 2
1
2 C12 C34 ½ C12 C34 s2 C34 C12 g
2
4 1
þ x ½ C12 C34 s2 C34 C12 2
2

1
þ x4 fC211 C233 ½ C12 C34 s2 C34 C12 2 g
2
x2 f½C11 C34 þ C12 C33 C12 C34 s 2
1
2 C12 C34 ½ C12 C34 s2 C34 C12 g C212 C234
2
N0 ¼ C212 C234 ; N2 ¼ ½C11 C34 þ C12 C33 C12 C34 s 2

1
þ 2 C12 C34 ½ C12 C34 s2 C34 C12
2
1
N4 ¼ C11 C33 ½ C12 C34 s2 C34 C12 2 ;
2 2
2
N6 ¼ ðC33 þ C11 Þ2 2 C11 C33 ; N8 ¼ 1

X
4
¼ N0 þ N2 x2 þ N4 x4 þ N6 x6 þ N8 x8 ¼ N2k x2k
k¼0
P4
2k
¼ 0 and its roots are given by
solving the above polynomial. Furthermore PR ði x; sÞ ¼ x4 x2 C11 C33
PI ði x; sÞ ¼ x3 ðC33 þ C11 Þ; QR ði x; sÞ
1
¼ C12 C34 x2 ½ C12 C34 s2 C34 C12
2
Hence
sin hðsÞ ¼
jQði x; sÞj2
cos hðsÞ ¼
jQði x; sÞj2
fx4 x2 C11 C33 g x ½C11 C34 þ C12 C33 C12 C34 s

þ x3 ðC33 þ C11 Þ fC12 C34 x2 ½12 C12 C34 s2 C34 C12 g
sin hðsÞ ¼
C212 C234 þ x2 f½C11 C34 þ C12 C33 C12 C34 s 2
2 C12 C34 ½12 C12 C34 s2 C34 C12 g þ x4 ½12 C12 C34 s2 C34 C12 2
fx4 x2 C11 C33 g fC12 C34 x2 ½12 C12 C34 s2 C34 C12 g
þ x4 ðC33 þ C11 Þ ½C11 C34 þ C12 C33 C12 C34 s
cos hðsÞ ¼ 2
C12 C234 þ x2 f½C11 C34 þ C12 C33 C12 C34 s 2 2 C12 C34
½12 C12 C34 s2 C34 C12 g þ x4 ½12 C12 C34 s2 C34 C12 2
These are continuous and differentiable in s based on Lemma 1.1. Hence we use
Theorem 1.2 and this prove the Theorem 1.3.
Our RFID TAGs with double rectangular spiral system for x1 vLcalc1 x2 vLcalc2 leads
to a characteristic equation for the eigenvalue k having the form
PðkÞ þ QðkÞ eks ¼ 0; The case s1 ¼ s2 ¼ s; D1 ¼ D2 ¼ 0.
Dðk; s1 ¼ s2 ¼ s; D1 ¼ D2 ¼ 0Þ ¼ k4 k3 ðC33 þ C11 Þ þ k2 C11 C33


Maclaurin series is a Taylor series expansion of a eks function about zero (0).
We get the following general characteristic equation D(k, s) under Taylor series
Dðk; sÞ ¼ k4 k3 ½C33 þ C11 þ k2 C11 C33

þ fC12 C34 þ k ½C11 C34 þ C12 C33 C12 C34 s
1
þ k2 ½ C12 C34 s2 C34 C12 g eks
2

P
4 P
2
Additionally Pn ðk; sÞ ! PðkÞ; Qm ðk; sÞ ! QðkÞ then PðkÞ ¼ aj k j ; QðkÞ ¼ cj k j
j¼0 j¼0
Pk ¼ k k ½C33 þ C11 þ k C11 C33

4 3 2
1
Qk ¼ C12 C34 þ k ½C11 C34 þ C12 C33 C12 C34 s þ k2 ½ C12 C34 s2
2
C34 C12
n; m 2 N0 ; n [ m; aj ; cj : R þ 0 ! R. They are continuous and differentiable

function of s such that a0 þ c0 6¼ 0. In the following “_” denotes complex and
conjugate. Functions PðkÞ; QðkÞ are analytic functions in k and differentiable in s.
The coefficients: faj ðC1 ; R1 ; double rectangular spiral antennas parametrsÞ and
cj ðC1 ; R1 ; s; double rectangular spiral antennas parametrsÞg 2 R depend on RFID
TAGs with double rectangular spiral antennas system’s C1, R1,s values and
antennas parameters.
a0 ¼ 0; a1 ¼ 0; a2 ¼ C11 C33 ; a3 ¼ ½C33 þ C11 ; a4 ¼ 1; c0 ¼ C12 C34 ;

c1 ¼ C11 C34 þ C12 C33 C12 C34 s
1
c2 ¼ C12 C34 s2 C34 C12
2
Unless strictly necessary, the designation of the variation arguments.

ðR1 ; C1 ; s; double rectangular spiral antennas parametrs) will subsequently be omit-
ted from P, Q, aj, cj. The coefficients aj, cj are continuous, differentiable functions of
their arguments and direct substitution shows that a0 þ c0 6¼ 0; C12 C34 6¼ 0.
g1 n1 ½Lcalc1
1
þ ðRdc1 þ Rdc2 Þ
R1 Lcalc1 ½Lcalc2 þ
1 ðRdc1 þ Rdc2 Þ
R1 Lcalc2
¼ qffiffiffiffiffiffiffiffiffiffi qffiffiffiffiffiffiffiffiffiffi 6¼ 0
g3 n3 C2 ½1 þ Lcalc2 þ 2 K Lcalc2 ½1 þ Lcalc1 þ 2 K Lcalc1
1 Lcalc1 Lcalc1 Lcalc2 Lcalc2
8 C1 ; double rectangular spiral antennas parametrs 2 R þ i.e. k ¼ 0 is not a root

of characteristic equation. Furthermore PðkÞ; QðkÞ are analytic function of k for
which the following requirements of the analysis (see kuang 1993, Sect. 3.4) can
also be verified in the present case.
imaginary roots. This condition was verified numerically in the entire (R1, C1,
double rectangular spiral antennas parameters) domain of interest.
fC12 C34 þ k ½C11 C34 þ C12 C33 C12 C34 s

QðkÞ þ k2 12 C12 C34 s2 C34 C12 g
j j¼j j
PðkÞ k4 k3 ½C33 þ C11 þ k2 C11 C33

1
þ x4 fC211 C233 ½ C12 C34 s2 C34 C12 2 g
2
x2 f½C11 C34 þ C12 C33 C12 C34 s 2
1
2 C12 C34 ½ C12 C34 s2 C34 C12 g C212 C234
2
Has at most a finite number of zeros. Indeed, this is a polynomial in x (degree

in x8 ).
(d) Each positive root x(R1,C1,s, double rectangular spiral antennas parameters)
of FðxÞ ¼ 0 is continuous and differentiable with respect to R1 ; C1 ; s; double
rectangular spiral antennas parameters. This condition can only be assessed
numerically.
of characteristic equation situated on the imaginary axis of the complex k-plane,
where by increasing the parameters R1 ; C1 ; s; double rectangular spiral antennas
parameters, Rek may, at the crossing Change its sign from (−) to (+), i.e. from a
stable focus Eð0Þ ðX1 ; VLcalc1 ; X2 ; VLcalc2 Þ ¼ ð0; 0; 0; 0Þ to an unstable one, or vice
versa. This feature may be further assessed by examining the sign of the partial
derivatives with respect to C1 ; R1 ; s and double rectangular spiral coils antennas
parameters [2, 3].
@Rek
^1 ðC1 Þ ¼ ð Þ ; R1 ; s; double rectangular spiral antennas parametrs ¼ const
@C1 k¼ix
@Rek
^1 ðR1 Þ ¼ ð Þ ; C1 ; s; double rectangular spiral antennas parametrs ¼ const
@R1 k¼ix
@Rek
^1 ðLcalc1 Þ ¼ ð Þ ; C1 ; R1 ; s ¼ const
@Lcalc1 k¼ix
@Rek
^1 ðLcalc2 Þ ¼ ð Þ ; C1 ; R1 ; s ¼ const
@Lcalc2 k¼ix
@Rek
^1 ðsÞ ¼ ð Þ ; C1 ; R1 ; double rectangular spiral antennas parametrs
@s k¼ix
¼ const where x 2 R þ :
For the case s1 ¼ s2 ¼ s & D1 ¼ D2 ¼ 0 we get the following results:
PR ði x; sÞ ¼ x4 x2 C11 C33 ; PI ði x; sÞ ¼ x3 ðC33 þ C11 Þ

1
QR ði x; sÞ ¼ C12 C34 x2 ½ C12 C34 s2 C34 C12 ;
2
N0 ¼ C212 C234 ; N2 ¼ ½C11 C34 þ C12 C33 C12 C34 s 2
1
þ 2 C12 C34 ½ C12 C34 s2 C34 C12
2
1
N4 ¼ C211 C233 ½ C12 C34 s2 C34 C12 2 ;
2
2
N6 ¼ ðC33 þ C11 Þ 2 C11 C33 ; N8 ¼ 1
Fðx; sÞ ¼ jPði x; sÞj2 jQði x; sÞj2 ¼ N0 þ N2 x2 þ N4 x4 þ N6 x6
X
4
þ N8 x8 ¼ N2k x2k
k¼0
P4
2k
¼ 0 when writing PðkÞ ¼
PR ðkÞ þ i PI ðkÞ and QðkÞ ¼ QR ðkÞ þ i QI ðkÞ, and inserting k ¼ i x into double
rectangular spiral coils antennas system’s characteristic equation, x must satisfy the
following:
jQði xÞj2
jQði xÞj2
Where jQði xÞj2 6¼ 0 in view of requirement (a) above, ðg; hÞ 2 R.

Note: FðxÞ is dependent of s. Now it is important to notice that if s 62 I(assume that
I
R þ 0 is the set where xðsÞ is a positive root of FðxÞ and for s 62 I, xðsÞ is not
define. Then for all s in I xðsÞ is satisfies that Fðx; sÞ ¼ 0). Then there are no
any s 2 I, where xðsÞ is a positive solution of Fðx; sÞ ¼ 0, we can define the angle
hðsÞ 2 ½0; 2 p as the solution of the below equations:
sin hðsÞ ¼
jQði xÞj2
cos hðsÞ ¼
jQði xÞj2
þ n2p
given by sn ðsÞ ¼ hðsÞxðsÞ ; n 2 N0 ; s 2 I. Let us introduce the functions I ! R;
following, the subscripts k; x; R1 ; C1 and RFID TAGs with double rectangular
spiral antennas parameters (Aavg1 ; Bavg2 ; A01 ; B02 ; Nc1 ; Nc2 ; g1 ; g2 ; . . .) indicate the
corresponding partial derivatives. Let us first concentrate on, ^ðxÞ remember in
kðAavg1 ; Bavg2 ; A01 ; B02 ; Nc1 ; Nc2 ; g1 ; g2 ; . . .Þ; xðAavg1 ; Bavg2 ; A01 ; B02 ; Nc1 ; Nc2 ; g1 ; g2 ; . . .Þ,
and keeping all parameters except one (x) and s. The derivation closely follows that
in reference [BK]. Differentiating RFID TAGs with double rectangular spiral
antennas characteristic equation PðkÞ þ QðkÞ eks ¼ 0 with respect to specific
parameter (x), and inverting the derivative, for convenience, one calculates:
Remark: x ¼ Aavg1 ; Bavg2 ; A01 ; B02 ; Nc1 ; Nc2 ; g1 ; g2 ; . . .

ð Þ ¼
Where Pk ¼ @P
@k ; . . . etc., Substituting k ¼ i x, and bearing Pði xÞ ¼
Pði xÞ, Qði xÞ ¼ Qði xÞ then i Pk ði xÞ ¼ Px ði xÞ and i Qk ði xÞ ¼
Qx ði xÞ that on the surface jPði xÞj2 ¼ jQði xÞj2 , one obtains

ð Þjk¼ix ¼ ð Þ

Q ¼ QR þ i QI
Px ¼ PRx þ i PIx ; Qx ¼ QRx þ i QIx ; Px ¼ PRx þ i PIx ; Qx ¼ QRx þ i QIx
P2 ¼ P2R þ P2I . When (x) can be any RFID TAGs with double rectangular spiral
antennas parameters R1, C1, And time delay s etc. Where for convenience, we have
dropped the arguments ði x; xÞ, and where
Fx ¼ 2 ½ðPRx PR þ PIx PI Þ ðQRx QR þ QIx QI Þ ; Fx

¼ 2 ½ðPRx PR þ PIx PI Þ ðQRx QR þ QIx QI Þ
xx ¼ Fx =Fx . We define U and V:

PRx ¼ 2 x ½2 x2 C11 C33 ; PIx ¼ 3 x2 ðC33 þ C11 Þ;

PRs ¼ 0; PIs ¼ 0;
QRs ¼ x2 C12 C34 s; QIs ¼ x C12 C34
PRx PR ¼ 2 x3 ½2 x4 3 x2 C11 C33 þ C211 C233 ;
PIx PI ¼ 3 x5 ðC33 þ C11 Þ2 ; xs ¼ Fs =Fx
PIx PI ¼ 3 x5 ðC33 þ C11 Þ2 ; xs ¼ Fs =Fx ;
1
QRx ¼ 2 x ½ C12 C34 s2 C34 C12
2
QIx ¼ C11 C34 þ C12 C33 C12 C34 s;
QIx QI ¼ x ½C11 C34 þ C12 C33 C12 C34 s 2
1
QRx QR ¼ 2 x ½ C12 C34 s2 C34 C12 ½C12 C34 x2
2
1
ð C12 C34 s C34 C12 Þ
2
2
Fs ¼ 2 ½ðPRs PR þ PIs PI Þ ðQRs QR þ QIs QI Þ ;

PR PIx ¼ 3 x4 ðx2 C11 C33 Þ ðC33 þ C11 Þ
1
Fs ¼ 2 x2 C12 C34 ½C11 C34 þ C12 C33 s x2 ð C12 C34 s2 C34 C12 Þ
2
PI PRx ¼ 2 x4 ðC33 þ C11 Þ ð2 x2 C11 C33 Þ; V ¼ ðPR PIs PI PRs Þ

ðQR QIs QI QRs Þ
1
QR QIx ¼ ½C12 C34 x2 ð C12 C34 s2 C34 C12 Þ
2
½C11 C34 þ C12 C33 C12 C34 s
QI QRx ¼ 2 x2 ðC11 C34 þ C12 C33 C12 C34 sÞ
1
ð C12 C34 s2 C34 C12 Þ
2
PR PIs ¼ 0; PI PRs ¼ 0; QR QIs ¼ x C12 C34 ½C12 C34 x2
1
ð C12 C34 s2 C34 C12 Þ
2
QI QRs ¼ x3 C12 C34 s ½C11 C34 þ C12 C33 C12 C34 s ; Fðx; sÞ ¼ 0

@s þ Fs ¼ 0; s 2 I ) @s ¼ Fx
Fs
@Rek 2 ½U þ s jPj2 þ i Fx @x Fs
^1 ðsÞ ¼ ð Þk¼ix ; ^1 ðsÞ ¼ Ref g; ¼ xs ¼
@s Fs þ i 2 ½V þ x jPj 2 @s F x
@Rek
signf^1 ðsÞg ¼ signfð Þ g; signf^1 ðsÞg
@s k¼ix
@x U @x
@s þ V
¼ signfFx g signfs þxþ g
@s jPj2

in a RFID TAGs with double rectangular spiral antennas system, about the equi-
librium point Eð0Þ ðX1 ; VLcalc1 ; X2 ; VLcalc2 Þ as a result of a variation of delay
parameter s. The analysis consists in identifying the roots of our system charac-
teristic equation situated on the imaginary axis of the complex k-plane where by
increasing the delay parameter s, Re k may at the crossing, change its sign from −
to +, i.e. from a stable focus E(*) to an unstable one, or vice versa. This feature may
be further assessed by examining the sign of the partial derivatives with respect to s,
^1 ðsÞ ¼ ð@ Re
@s Þk¼ix
k
@Rek
^1 ðsÞ ¼ ð Þ ; C1 ; R1 ; RFID TAGs with double rectangular
@s k¼ix
spiral antennas parameters ¼ const where x 2 R þ :
For our stability switching analysis we choose each of our system’s rectangular
spiral antenna on a substrate, width is 300 lm and lr = 450. The antenna is
constructed from silver ointment which his resistance is bigger than pure silver by
50 %. Track width is 20 lm, gap between tracks 20 lm, track depth is 20 lm up 10
100 lm. The requested rectangular spiral antenna inductance is 2.66 mH and
parasitic resistance less than 10 X. We need to find the possible number of rect-
angular spiral antenna’s turns (Nc1 and Nc2 for the first and second rectangular
spiral antenna respectively). Since each RFID antenna substrate permeability is 450
(lr = 450), our RFID antennas permeability is an average value between air per-
meability and magnet. We consider that the TAGs permeability is 100–300 (lr) and
possible TAGs dimension: 5 mm 5 mm, 6 mm 6 mm, 7 mm 7 mm, 8 mm
8 mm and permeability 100, 200, 300 (lr). Tables 8.3, 8.4, and 8.5 describes the
analysis for l = lr ∙ l0 (lr = 100, 200, 300). Table 8.3 is for l = lr ∙ l0; lr = 100; l
= 125.66 ∙ (1e−6) H/m. Table 8.4 is for l = lr ∙ l0; lr = 200; l = 251.32 ∙ (1e−6)
H/m and Table 8.5 is for l = lr ∙ l0; lr = 300; l = 376.98 ∙ (1e−6) H/m.
Results: The most close inductance analysis to 2.66 mH is 2.7 mH and subcases.
Result Table 8.3: Lcalc = 2.7 mH, Nc = 60, (A0 = 7 mm) (B0 = 7 mm); l = lr
∙ l0; lr = 100; l = 125.66 ∙ (1e−6) H/m. The DC resistance of rectangular spiral
RFID antenna: RDC ¼ rS LT
¼ rpa
LT
2 . LT—total length of the wire. r—Conductivity of
the wire (ʊ/m). S—Cross section area p a2 . a—radius of the wire.
LT LT LT
RDC ¼ ¼ ; RDC ¼
r S r p a2 r p a2
2 fðA0 þ B0 Þ ð1 þ NC Þ ðw þ gÞ ½4 NC 3 g
¼
r p a2
Table 8.3 Rectangular spiral antenna Lcalc as a function of different number of turns for (Nc) for l
= lr ∙ l0; lr = 100; l = 125.66∙(1e−6) H/m
Nc Lcalc (5 mm 5 Lcalc (6 mm 6 Lcalc (7 mm 7 Lcalc (8 mm 8
mm)—[H] mm)—[H] mm)—[H] mm)—[H]
10 1.08 10−4 1.37 10−4 1.67 10−4 1.98 10−4
20 3.383 10−4 4.38 10−4 5.419 10−4 6.48 10−4
30 6.212 10−4 8.25 10−4 0.001 0.0013
60 0.0014 0.002 0.0027 = 2.7 mH 0.0035
100 0.001 0.0024 0.0041 0.0058
150 −0.0014 NaN 0.0021 0.0051
120 1.3598 10−4 0.0018 0.0038 0.0061
170 −0.0040 −0.0012 2.5454 10−4 0.0033
200 −0.0106 −0.0062 −0.0023 NaN
220 −0.0171 −0.0115 −0.0063 −0.0020
250 −0.0306 −0.023 −0.0158 −0.0092
= lr ∙ l0; lr = 200; l = 251.32 ∙ (1e−6)H/m
Nc Lcalc (5 mm Lcalc (6 mm Lcalc (7 mm Lcalc (8 mm
5 mm)—[H] 6 mm)—[H] 7 mm)—[H] 8 mm)—[H]
10 2.17 10−4 2.75 10−4 3.35 10−4 3.97 10−4
20 6.767 10−4 8.76 10−4 0.0011 0.0013
30 0.0012 0.001 0.0021 0.0025
60 0.0027 = 2.7 mH 0.004 0.0055 0.0069
100 0.0020 0.0049 0.0081 0.0116
150 −0.0028 NaN 0.0042 0.0102
120 2.719 10−4 0.0035 0.0077 0.0122
170 −0.008 −0.0025 5.09 10−4 0.0066
200 −0.0212 −0.0123 −0.0046 NaN
220 −0.0342 −0.0229 −0.0127 −0.0039
250 −0.0612 −0.046 −0.0316 −0.0184
= lr ∙ l0; lr = 300; l = 376.98 ∙ (1e−6)H/m
Nc Lcalc (5 mm Lcalc (6 mm Lcalc (7 mm Lcalc (8 mm
5 mm)—[H] 6 mm)—[H] 7 mm)—[H] 8 mm)—[H]
10 3.25 10−4 4.13 10−4 5.03 10−4 5.95 10−4
20 0.001 0.0013 0.0016 0.0019
30 0.0019 0.0025 0.0031 0.0038
60 0.0041 0.0061 0.0082 0.0104
100 0.003 0.0073 0.0122 0.0173
150 −0.004 NaN 0.0062 0.0153
120 4.079 10−4 0.0053 0.0115 0.0183
170 −0.0119 −0.0037 7.63 10−4 0.0099
200 −0.0317 −0.0185 −0.0069 NaN
220 −0.0513 −0.0344 −0.0190 −0.0059
250 −0.0918 −0.069 −0.0474 −0.0276
A0 þ B0 ¼ 0:014 m ; Nc ¼ 60; w þ g ¼ 40 106 ) LT ¼ 1:689 m. Cross sec-

tion area S ¼ 20 lm 20 lm ¼ 400 1012 m2 . Conductivity of silver
r ¼ 6:1 107 ðf=mÞ. Conductivity has SI units of Siemens per meter
(S/m).rsilver @ 20 C ¼ 6:3 107 ðS/mÞ.
The track depth (x) does not influence our total inductance, and then we can take
it as a variable and find his minimum value for RDC < 10 Ω. S ¼ 20 lm x.
1:689
\10 ) x [ 1:3405 104 m ¼ 134:05 lm;
6:3 107 20 106 x
RDC@t¼20 lm ¼ 67 X
Actually the track is a mixture of silver then the conductivity is half of silver
conductivity.
rsilver @ 20 C 6:3 107 ðS/mÞ

¼ ¼ 3:15 107 ðS/mÞ:
2 2
1:6890
\10 ) x [ 2:681 104 m ¼ 268:1 lm;
3:15 107 20 106 x
RDC@t¼20 lm ¼ 134 X
Result Table 8.4: Lcalc = 2.7 mH, Nc = 60, (A0 = 5 mm) (B0 = 5 mm); l = lr
∙ l0; lr = 200; l = 251.32 ∙ (1e−6) H/m. The DC resistance of rectangular spiral
RFID antenna: RDC ¼ rSLT
¼ rpa
LT
2 . LT—total length of the wire. r—conductivity of
the wire (ʊ/m). S—Cross section area p a2 . a—radius of the wire.
LT LT LT
RDC ¼ ¼ ; RDC ¼
r S r p a2 r p a2
2 fðA0 þ B0 Þ ð1 þ NC Þ ðw þ gÞ ½4 NC 3 g
¼
r p a2
A0 þ B0 ¼ 0:01 m ; Nc ¼ 60 ; w þ g ¼ 40 106 ) LT ¼ 1:201 m
Cross section area S ¼ 20 lm 20 lm ¼ 400 1012 m2 . Conductivity of Silver

r ¼ 6:1 107 ðf=mÞ. Conductivity has SI units of siemens per meter (S/m).
rsilver @ 20 C ¼ 6:3 107 ðS/mÞ. The track depth (x) does not influence our total
inductance, and then we can take it as a variable and find his minimum value for
RDC < 10 X. S ¼ 20 lm x.
1:201
\10 ) x [ 9:5317 105 m ¼ 95:317 lm;
6:3 107 20 106 x
RDC@t¼20 lm ¼ 47:65 X
conductivity.
rsilver @ 20 C 6:3 107 ðS/mÞ

¼ ¼ 3:15 107 ðS/mÞ:
2 2
1:201
\10 ) x [ 19:06 105 m ¼ 190:6 lm;
3:15 107 20 106 x
RDC@t¼20 lm ¼ 95:30 X
Follows 8.3 and 8.4 tables results we choose for Lcalc-1 = 2.7 mH
Rdc1 ¼ 134 X; Lcalc-2 = 2.7 mH; Rdc2 ¼ 95:30 X. Typical other values for
our system R1 = 100 kX, C1 = 23 pF, K = 0.6 (RFID IC is represented as parallel
circuit of capacitor C1 and R1). g1 ¼ 371:21; g2 ¼ 3:395 105 ; g3 ¼ 7:36 1011
n1 ¼ 371:21; n2 ¼ 3:395 105 ; n3 ¼ 7:36 1011 ; C21 ¼ C43 ¼ 1

3:395 105 371:21
C11 ¼ ¼ 4:6128 105 ; C12 ¼ ¼ 5:0436 1012
7:36 1011 7:36 1011
3:395 105 371:21
C33 ¼ ¼ 4:6128 105 ; C34 ¼ ¼ 5:0436 1012
7:36 1011 7:36 1011
C13 ¼ C14 ¼ C22 ¼ C23 ¼ C24 ¼ C31 ¼ C32 ¼ C41 ¼ C42 ¼ C44 ¼ 0
Then we get the expression for Fðx; sÞ typical RFID TAGs with double rect-
angular spiral antenna parameters values.

1
þ x4 fC211 C233 ½ C12 C34 s2 C34 C12 2 g
2
x2 f½C11 C34 þ C12 C33 C12 C34 s 2
1
2 C12 C34 ½ C12 C34 s2 C34 C12 g C212 C234
2

¼ x8 þ x6 f8:5112 1011 2 2:1278 1011 g
1
þ x4 f4:5275 1022 ½ 2:5438 1025 s2
2
þ 5:0436 10 þ 5:0436 1012 2 g
12
x2 f½2:3265 1018 þ 2:3265 1018 2:5438 1025 s 2

1
2 2:5438 1025 ½ 2:5438 1025 s2
2
þ 5:0436 1012 þ 5:0436 1012 g 6:4709 1050
We find those x, s values which fulfill Fðx; sÞ ¼ 0. We ignore negative,

complex, and imaginary values of x for specific s values. s 2 ½0:001. . .10 (s) and it
can be express by 3D function Fðx; sÞ ¼ 0.

þ x4 f4:5275 1022 ½1:2719 1025 s2 þ 1:0087 1013 2 g
x2 f½4:653 1018 2:5438 1025 s 2
5:0876 1025 ½1:2719 1025 s2 þ 10:0872 1012 g 6:4709 1050
P4
Hence Fðx; sÞ ¼ 0 implies k¼0 N2k x2k ¼ 0
Nj ! Phij (j = 0, 2, 4, 6, 8). Running MATLAB script for s values

(s 2 ½0:001. . .10 ). Phij(j = 1,3,5,7) = 0
MATLAB script: Tau = 0.001; Phi0 = -6.4709e50; Phi1 = 0; Phi2 =
(4.653e18-2.5438e25 * Tau).^2-5.0876e25 * (1.2719e25 * Tau * Tau +
10.0872e12); Phi3 = 0; Phi4 = 4.5275e22-(1.2719e25 * Tau * Tau + 1.0087e13).
^2; Phi5 = 0; Phi6 = 4.2556e11; Phi8 = 1; Phi7 = 0; p = [Phi8 Phi7 Phi6 Phi5 Phi4
Phi3 Phi2 Phi1 Phi0]; r = roots(p) (Tables 8.6, and 8.7).
We can summary our xi(s) results for xi(s) > 0 and real number (ignore com-
plex, negative and imaginary values). We exclude from our table (Table 8.12) the
high and real xi(s) values (1.0e+009*, 1.0e+010*, 1.0e+011*, …) and add results
for s = 15 s and s = 20 s (Table 8.12). Next figure describes the RFID TAGs with
double loop rectangular spiral antennas system, x as a function of delay parameter s
(case s1 ¼ s2 ¼ s; D1 ¼ D2 ¼ 0) (Tables 8.8, 8.9, 8.10, 8.11, 8.14).
RFID TAGs with double loop rectangular spiral antennas system, x as a func-
tion of delay parameter s (Case s1 ¼ s2 ¼ s; D1 ¼ D2 ¼ 0) (Fig. 8.5).
Table 8.6 RFID TAGs with double rectangular spiral antennas systemroots xi(s)
s s = 0.001 s s = 0.01 s s = 0.1 s
x1 1.0e+009* 1.0e+010* 1.0e+011*
x2 −3.5664 −3.5664 −3.5664
x3 −0.0000 + 3.5664i 0.0000 + 3.5664i −0.0000 + 3.5664i
x4 −0.0000 − 3.5664i 0.0000 − 3.5664i −0.0000 − 3.5664i
x5 3.5664 3.5664 3.5664
x6 −0.0000 + 0.0000i 0.0000 + 0.0000i −0.0000 + 0.0000i
x7 −0.0000 − 0.0000i 0.0000 − 0.0000i −0.0000 − 0.0000i
x8 0.0000 + 0.0000i -0.0000 + 0.0000i 0.0000 + 0.0000i
x9 0.0000 − 0.0000i −0.0000 − 0.0000i 0.0000 − 0.0000i
Table 8.7 RFID TAGs with double rectangular spiral antennas system roots xi(s)
s s=1s s=2s s=3s
x1 1.0e+012* 1.0e+012* 1.0e+013*
x2 −3.5664 −7.1327 −1.0699
x3 0 + 3.5664i −0.0000 + 7.1327i −0.0000 + 1.0699i
x4 0 − 3.5664i −0.0000 − 7.1327i −0.0000 − 1.0699i
x5 3.5664 7.1327 1.0699
x6 −0.0000 + 0.0000i 0.0000 + 0.0000i −0.0000 + 0.0000i
x7 -0.0000-0.0000i 0.0000-0.0000i −0.0000 − 0.0000i
x8 0.0000 + 0.0000i −0.0000 + 0.0000i 0.0000 + 0.0000i
x9 0.0000 − 0.0000i −0.0000 − 0.0000i 0.0000 − 0.0000i
s s=4s s=5s s=6s
x1 1.0e+013* 1.0e+013* 1.0e+013*
x2 −1.4265 −1.7832 −2.1398
x3 0 + 1.4265i 0.0000 + 1.7832i 0 + 2.1398i
x4 0 − 1.4265i 0.0000 − 1.7832i 0 − 2.1398i
x5 1.4265 1.7832 2.1398
x6 −0.0000 + 0.0000i 0.0000 + 0.0000i −0.0000 + 0.0000i
x7 −0.0000− 0.0000i 0.0000 − 0.0000i −0.0000 − 0.0000i
x8 0.0000 + 0.0000i −0.0000 + 0.0000i 0.0000 + 0.0000i
x9 0.0000 − 0.0000i −0.0000 − 0.0000i 0.0000 − 0.0000i
s s=7s s=8s s=9s
x1 1.0e+013* 1.0e+013* 1.0e+013*
x2 −2.4965 −2.8531 −3.2097
x3 0.0000 + 2.4965i −0.0000 + 2.8531i −0.0000 + 3.2097i
x4 0.0000 − 2.4965i −0.0000 − 2.8531i −0.0000 − 3.2097i
x5 2.4965 2.8531 3.2097
x6 −0.0000 + 0.0000i 0.0000 + 0.0000i −0.0000 + 0.0000i
x7 −0.0000 − 0.0000i 0.0000 − 0.0000i −0.0000 − 0.0000i
x8 0.0000 + 0.0000i −0.0000 + 0.0000i 0.0000 + 0.0000i
x9 0.0000 − 0.0000i −0.0000 − 0.0000i 0.0000 − 0.0000i
Table 8.10 RFID TAGs with s s=0s s = 10 s

double rectangular spiral
antennas system roots xi(s) x1 1.0e+006* 1.0e+013*
x2 −3.4542 −3.5664
x3 3.4542 −0.0000 + 3.5664i
x4 0.0000 + 2.6095i −0.0000 − 3.5664i
x5 0.0000−2.6095i 3.5664
x6 −0.1553 + 1.6727i −0.0000 + 0.0000i
x7 −0.1553 − 1.6727i −0.0000 − 0.0000i
x8 0.1553 + 1.6727i 0.0000 + 0.0000i
x9 0.1553 - 1.6727i 0.0000 − 0.0000i
Table 8.11 RFID TAGs with s s = 15 s s = 20 s

antennas system roots xi(s) x1 1.0e+013* 1.0e+013*
x2 −5.3496 −7.1327
x3 0 + 5.3496i 0.0000 + 7.1327i
x4 0 − 5.3496i 0.0000 − 7.1327i
x5 5.3496 7.1327
x6 0.0000 + 0.0000i −0.0000 + 0.0000i
x7 0.0000 − 0.0000i −0.0000 − 0.0000i
x8 −0.0000 + 0.0000i 0.0000 + 0.0000i
x9 −0.0000 − 0.0000i 0.0000 − 0.0000i
Table 8.12 RFID TAGs with s(s) x sin(x ∙ s) cos(x ∙ s)

antennas system positive and 0 3.4542 −1.43e−018 −9.98e−014
real roots xi(s) values and sin 0.001…1 3.5664
(x ∙ s), cos(x ∙ s) values 2 7.1327 −5.86e−016 4.14e−015
3 1.0699 −1.11e−015 1.44e−015
4 1.4265 −3.65e−016 9.77e−016
5 1.7832 −1.50e−016 6.51e−016
6 2.1398 −7.23e−017 4.59e−016
7 2.4965 −3.90e−017 3.39e−016
8 2.8531 −2.29e−017 2.60e−016
9 3.2097 −1.42e−017 2.06e−016
10 3.5664 −9.37e−018 1.67e−016
15 5.3496 −1.85e−018 7.43e−017
20 7.1327 −5.85e−019 4.18e−017
Hint: e − x = 10−x
MATLAB script: plot([0 0.001 0.01 0.1 1 2 3 4 5 6 7 8 9 10 15 20], [3.4542

3.5664 3.5664 3.5664 3.5664 7.1327 1.0699 1.4265 1.7832 2.1398 2.4965 2.8531
3.2097 3.5664 5.3496 7.1327], ‘-or’).
RFID TAGs with double loop rectangular spiral antennas system F(x,s) function
(Fig. 8.6).
MATLAB script:
[w,t] = meshgrid(1:0.1:8,0:0.1:20); f = w.^8 + w.^6 * 4.2556e11 + w.^4. *
(4.5275e22 − (1.2719e25. * t.^2 + 1.0087e13).^2) − w.^2. * ((4.653e18 −
2.5438e25. * t).^2 − 5.0876e25. * (1.2719e25. * t.^2 + 10.0872e12)) − 6.4709e50;
set(gcf, ‘renderer’, ‘painters’); meshc(f);%x ! w,s ! t.
P
Table 8.13 Single, two, and three turn rectangular planar coils L0, M+, M−, and M
846
Single turn rectangular planar coil (N = 4) Two turn rectangular planar coil (N = 8) Three turn rectangular planar coil (N = 12)
N
X 4
X 8
P 12
P
L0 ¼ Li Li ¼ L1 þ L2 Li Li
i¼1 i¼1
i¼1 i¼1
i ¼ 1; 2; . . .; N þ L3 þ L4
Mþ 0 2 ðM1;5 þ M2;6 2 ðM1;9 þ M1;5 þ M5;9
þ M3;7 þ M4;8 Þ þ M2;6 þ M2;10 þ M6;10
þ M3;7 þ M3;11 þ M11;7
þ M4;8 þ M4;12 þ M8;12 Þ
M 2 ðM1;3 þ M2;4 Þ 2 ðM1;7 þ M1;3 þ M5;7 2 ðM1;11 þ M1;7 þ M1;3
þ M5;3 þ M2;8 þ M2;4 þ M5;11 þ M5;7 þ M5;3
þ M6;8 þ M6;4 Þ þ M9;11 þ M9;7 þ M9;3
þ M2;12 þ M2;8 þ M2;4
þ M6;12 þ M6;8 þ M6;4
þ M10;12 þ M10;8 þ M10;4 Þ
P X
M M ¼ M 2 ðM1;5 þ M2;6 2 ðM1;9 þ M1;5 þ M5;9
¼ 2 ðM1;3 þ M2;4 Þ þ M3;7 þ M4;8 Þ þ M2;6 þ M2;10 þ M6;10
2 ðM1;7 þ M1;3 þ M5;7 þ M3;7 þ M3;11 þ M11;7
þ M5;3 þ M2;8 þ M2;4 þ M4;8 þ M4;12 þ M8;12 Þ
þ M6;8 þ M6;4 Þ 2 ðM1;11 þ M1;7 þ M1;3
þ M5;11 þ M5;7 þ M5;3
þ M9;11 þ M9;7 þ M9;3
þ M2;12 þ M2;8 þ M2;4
þ M6;12 þ M6;8 þ M6;4
þ M10;12 þ M10;8 þ M10;4 Þ
8 Antennas System Stability Analysis
Table 8.14 Single, two turn, and three turn number of term contributing to M+ and M−
Rectangular planar N+: number of term contributing to M+ N−: number of terms
coil type (number of positive mutual inductance terms) contributing to M−
Single turn 0 4
rectangular planar
coil (n = 1, Zs = 4)
Two turn 8 16
rectangular planar
coil (n = 2, Zs = 8)
Three turn 24 36
rectangular planar
coil (n = 3,
Zs = 12)
Fig. 8.5 RFID TAGs with

double loop rectangular spiral
antenna system x as a
function of delay parameter s
Then we get the expression for Fðx; sÞ for typical RFID shifted gate parameters
values.

þ x4 f383:17 1020 ½7:8 1024 s2 þ 7:9 1012 2 g
x2 f½34:94 1017 15:6 1024 s 2
31:2 1024 ½7:8 1024 s2 þ 7:9 1012 g 243:39 1048

complex, and imaginary values of x for specific s values. s 2 ½0:001. . .10 and we
can be express by 3D function Fðx; sÞ ¼ 0. Since it is a very complex function
We recommend to solve it numerically rather than analytic.
Fig. 8.6 RFID TAGs double rectangular spiral antennas F(x,s) function
double rectangular spiral coils antennas system. Since it is a very complex function
we recommend to solve it numerically rather than analytic.
s ¼ xh þþ ðsÞ
x þ ðsÞ if only x þ is feasible. Additionally when all double rectangular spiral coils
antennas system’s parameters are known and the stability switch due to various
time delay values s is describe in the following expression:
signf^1 ðsÞg ¼ signfFx ðxðsÞ; sÞg signfs xs ðxðsÞÞ þ xðsÞ

þ g
jPðxðsÞÞj2
Remark: we know Fðx; sÞ ¼ 0 implies it roots xi ðsÞ and finding those delays
values s which xi is feasible. There are s values which xi are complex or imaginary
number, then unable to analyse stability [5, 6].
8.3 Single-Turn Square Planar Straight Thin Film Inductors Antenna … 849
8.3 Single-Turn Square Planar Straight Thin Film

Inductors Antenna System Stability Optimization
Under Microstrip Delayed in Time
We have a system of single turn square planar straight thin film inductors antenna
(four segments). The system is constructed from four straight thin film inductors
which are connected in a single turn square structure. The straight thin film inductors
are connected by microstrip lines (A, B, and C). The single turn square planar
straight thin film inductors antenna system is connected to transceiver module
through two microstrip lines (D and E) [85]. Index (i) stands for the first (i = 1),
second (i = 2), third (i = 3), and fourth (i = 4) straight thin film inductors. wi is the
width of straight thin film inductor (i) in cm, zi is the thickness of straight thin film
inductor (i) in cm, and li is the length of straight thin film conductor (inductor) in cm.
The calculated inductance of straight thin film inductor (i) is as follow (Li is the
segment inductance in lH):
2 li w i þ zi
Li ¼ 0:002 li fln½ þ 0:50049 þ g½lH ; i ¼ 1; 2; 3; 4
wi þ zi 3 li
Remark: we assume that the magnetic permeability of the conductor material is 1

and the four straight thin film inductors are not identical wi 6¼ wj ; li 6¼ lj zi 6¼ zj
ðj 6¼ i; j ¼ 1; 2; 3; 4; i ¼ 1; 2; 3; 4Þ (Fig. 8.7).
N-turn planar rectangular coil structure: If we have N-turn planar rectangular coil
structure the total inductance of this coil is equal to the sum of the self-inductance
P
of each of the straight segment ( Ni¼1 Li ; i ¼ 1; 2; . . .; N) plus all the mutual
inductances between the segments. The mutual inductance between segment (k) and
(j) has a component Mk,j caused by the current flowing in segment (k), and a
component Mj,k caused by the current flowing in segment (j). Since the frequency
and phase in both segments are identical, the total mutual inductance linking them
equals Mk;j þ Mj;k . An analogous relationship exists between segment pairs 2-6, 3-7,
4-8, etc., in each of these pairs, current flow is in the same direction in both
segments and all mutual inductances are positive. The mutual inductance between
segment 1 and 7, on the other hand, has a component M1,7 caused by the current in
segment 1, and a component M7,1 caused by the current in segment 7. It can be
implemented in the same manner to other segments. The total mutual inductance
linking these two segments equals M1;7 þ M7;1 but is negative because current flow
in segment 1 is opposite in direction to current flow in segment 7
(k ¼ 1; 2; . . .; N; j ¼ 1; 2; . . .; N; k 6¼ j; k; j 2 N þ ). An analogous relationship exists
between segment pairs 1-3, 5-7, 5-3, 2-8, 2-4, 6-8, and 6-4 (case of two-turn
rectangular planar coil). Current magnitude is identical in all segments, with the
result that Mj,k = Mk,j. The total inductance LT for the case of two turn coil is as
follow:
Fig. 8.7 Single turn square planar straight thin film inductors antenna system
X
N ¼8
LT ¼ Li þ 2 ðM1;5 þ M2;6 þ M3;7 þ M4;8 Þ
i¼1
2 ðM1;7 þ M1;3 þ M5;7 þ M5;3 þ M2;8 þ M2;4 þ M6;8 þ M6;4 Þ
P PN¼8
We define the mutual inductances term as M, (L0 ; L0 ¼ i¼1 Li ).
X
M ¼ 2 ðM1;5 þ M2;6 þ M3;7 þ M4;8 Þ
2 ðM1;7 þ M1;3 þ M5;7 þ M5;3 þ M2;8 þ M2;4 þ M6;8 þ M6;4 Þ
Where LT is the totalPinductances, L0 is the sum of the self-inductances of all the

straight segments, and M is the sum of all the mutual inductances, both positive
and negative.
M þ ¼ 2 ðM1;5 þ M2;6 þ M3;7 þ M4;8 Þ; M

¼ 2 ðM1;7 þ M1;3 þ M5;7 þ M5;3 þ M2;8 þ M2;4 þ M6;8 þ M6;4 Þ
P
General case, N-turn planar rectangular coil structure LT ¼ Ni¼1 Li þ M þ
M where M+ is the sum of the positive mutual inductances and M- is the sum of
the negative mutual inductances. The mutual inductance between two parallel
conductors is a function of the length of the conductors and of the geometric mean
distance between them. The figure and Table 8.13 describe single-turn rectangular
planar coil (N = 4), two-turn rectangular planar coil (n = 8), and three-turn rect-
angular planar coil (n = 12), sum of the self-inductances (L0), sum of the positive
mutual inductances (M+), sum of P the negative mutual inductances (M−) and the
sum of all the mutual inductances M (Table 8.15, Fig. 8.8).
We define (n) as the number of full turns and (Zs) as the total number of
segments. The number of term contributing to M+ is N+ and the number of terms
contributing to M− is N−.
N þ ¼ 4 n ðn 1Þ þ 2 n ðZs 4 nÞ
N ¼ 4 n4 þ 2 n ðZs 4 nÞ þ ðZs 4 n 2Þ ðZs 4 n 1Þ ðZs 4 nÞ=3
Table 8.15 Single turn square planar straight thin film inductors antenna equivalent circuit
VAi 8 i ¼ 1; 2; . . .; 14 expressions
Node voltage Expression
VA1 IRa Ra
VA2 dIL0
IRa Ra La dt
1
VA3 dIL0
IRa Ra ðLa þ L01 Þ dt
1
VA4 ðVA4 VA5 Þ dIL0

1
IL01 R1
VA5 dIL0
1
IL01 R1
VA6 dIL0 dIL0 ðts1 Þ
1
IL01 R1 L02 1
dt
VA7 ðVA7 VA8 Þ dIL0 dIL0 ðts1 Þ
1
IL01 R1 L02 1
dt IL01 ðt s1 Þ R2
VA8 dIL0 dIL0 ðts1 Þ
1
IL01 R1 L02 1
dt IL01 ðt s1 Þ R2
P3
VA9 dIL0 ðt sÞ
i¼1 i
P3 P2
IRb Rb þ ðLb þ L04 Þ 1
dt þ IL01 ðt i¼1 si Þ R4 þ IL01 ðt i¼1 si Þ R3
P3
VA10 ðVA10 VA11 Þ dIL0 ðt sÞ
i¼1 i
P3
dt þ IL01 ðt i¼1 si Þ R4
P3
VA11 dIL0 ðt sÞ
i¼1 i
P3
dt þ IL01 ðt i¼1 si Þ R4
P3
VA12 dIL0 ðt sÞ
i¼1 i
IRb Rb þ ðLb þ L04 Þ 1 dt
P3
VA13 dIL0 ðt sÞ
i¼1 i
IRb Rb þ Lb 1
dt
VA14 IRb Rb
Fig. 8.8 Square planar straight thin film inductor antenna N=4, N=8, and N=12
Table 8.16 Single turn square planar straight thin film inductors antenna equivalent circuit’s
remarks 1.0–1.6 expressions
Remark no. Expression
1.0 P3 P
3
d IL0 ðt
2
sÞ d 2 IL0 ðtÞ k si
i¼1 i
1
dt2 ¼ 1
dt2 e i¼1
1.1 Pi1 P
j1
P
4 d 2 IL0 ðt s Þ
k¼1 k
d 2 IL0 ðtÞ P
4 k sk
L01 1
dt2 ¼ 1
dt2 L0j e k¼1
i¼2 j¼2
1.2 Pi1 P
j1
P
4
qi li
dIL0 ðt s Þ
k¼1 k
dIL0 ðtÞ P
4
qj lj
k sk
wi zi ¼ wj zj e
1 1 k¼1
dt dt
i¼2 j¼2
1.3 P3 P
3
dIL0 ðt sÞ dIL0 ðtÞ k si
i¼1 i
1
dt ¼ 1
dt e i¼1
1.4 P3 P
3
d 3 IL0 ðt sÞ d 3 IL0 ðtÞ k si
i¼1 i
1
dt3 ¼ 1
dt3 e i¼1
1.5 Pi1 P
j1
P
4 d 3 IL0 ðt s Þ
k¼1 k
d 3 IL0 ðtÞ P
4 k sk
L01 1
dt3 ¼ 1
dt3 L0j e k¼1
i¼2 j¼2
1.6 Pi1 P
j1
P
4
qi li
d 2 IL0 ðt s Þ
k¼1 k
d 2 IL0 ðtÞ P
4
qj lj
k sk
wi zi 1
dt2 ¼ 1
dt2 wj zj e k¼1
i¼2 j¼2
Mutual inductance between two parallel conductors (same length):

The mutual inductance Mj,k between two parallel conductors (j and k segments
lj ; lk ; j 6¼ k; lk ¼ lj ¼ l) is a function of the length of the conductors and of the
Geometric Mean Distance (GMD) between them (Mj;k ¼ 2 l Q). Mj;k is the
mutual inductance in (nH), l is the conductor length in (cm), and Q is the mutual
inductance parameter, calculated from the equation (Fig. 8.9):
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
l l2 ðGMDÞ2 GMD
Q ¼ lnf þ 1þ g 1 þ þ
GMD ðGMDÞ2 l2 l
Fig. 8.9 Two parallel conductors (same length)
In this equation, l is the length corresponding to the subscript of Q, and GMD is

the Geometric Mean Distance (GMD) between the two conductors (lj ; lk ), which is
approximately equal to the distance dj;k ðGMD ’ dj;k ¼ dÞ between the track cen-
ters. The exact value of the GMD may be calculated from the equation:
1 1 1 1
ln GMD ¼ ln dj;k f d
þ d
þ d
þ d
12 ð wj;k Þ2 60 ð wj;k Þ4 168 ð wj;k Þ6 360 ð wj;k Þ8
1
þ d
þ . . .g
660 ð wj;k Þ10
where (w) is the track width wj ¼ wk ¼ w.
1 1 1 1
GMD ¼ exp½ln dj;k f d
þ d
þ d
þ d
12 ð wj;k Þ2 60 ð wj;k Þ4 168 ð wj;k Þ6 360 ð wj;k Þ8
1
þ d
þ . . .g
660 ð wj;k Þ10
We can represent our single turn square planar straight thin film inductors
antenna system equivalent circuit. Microstrip lines (A, B, and C) are represented as
delay lines s1, s2, and s3 respectively. We neglect microstrip lines D and E parasitic
effects (no delays). The sum of all the mutual P
Inductances, both positive and negative is marked as (SigmaM ¼ M)
inductance. Transceiver module is represented as an equivalent circuit of mixer,
with input and output impedances of the mixer board. The mixer itself has a
common gate input. The input is dominated by parasitic impedances of the package
and PCB stray, such as shunt capacitance of the device gates along with the bond
wire inductances and resistance which is inversely proportional to the mixer current
setting. In similar way as for the inputs, the impedance at the outputs can be
modeled by a resistance with a shunt PCB stray capacitance and bond wires
inductances. The resistance Ri ¼ qAi li i of a straight thin film (strip) number
(i; i ¼ 1; 2; 3; 4) is expected to depend on the DC resistivity qi of the straight thin
film strip material and the strip cross section Ai ¼ wi zi ; Ri ¼ wqii lzii . The dimension
of the straight thin film strip that affects the inductance most strongly is the length li .
The width wi has much weaker influence, and straight thin film strip thickness can
be neglected completely for the limit (zi wi ). This is in contrast to the resistance
Ri ¼ wqii lzii , which is inversely proportional to the straight thin film strip cross section
wi zi and depends on the material properties via its resistivity qi [85] (Fig. 8.10).
Inductance of a single turn square planar coil (straight thin film inductors sys-
tem) calculation:
2 li w i þ zi
Li ¼ 0:002 li fln½ þ 0:50049 þ g½lH ; i ¼ 1; 2; 3; 4
w i þ zi 3 li
Fig. 8.10 Single turn square planar straight thin film inductors antenna system equivalent circuit
2 l1 w1 þ z1
L1 ¼ 0:002 l1 fln½ þ 0:50049 þ g;
w1 þ z1 3 l1
2 l2 w2 þ z2
L2 ¼ 0:002 l2 fln½ þ 0:50049 þ g
w2 þ z2 3 l2
2 l3 w3 þ z3
L3 ¼ 0:002 l3 fln½ þ 0:50049 þ g;
w3 þ z3 3 l3
2 l4 w4 þ z4
L4 ¼ 0:002 l4 fln½ þ 0:50049 þ g
w4 þ z4 3 l4
The derivations are producing the following calculation results (Grover):
X
N ¼4 X X
N ¼4
LT ¼ Li þ M¼ Li þ M þ M
i¼1 i¼1
Since the currents in parallel straight thin film segments flow in opposite
directions, there is no positive mutual inductance in this segment coil; that is
M þ ¼ 0. The negative mutual inductance is equal to the sum of M1;3 ; M3;1 and
M2;4 ; M4;2 Or, since M1;3 equals M3;1 and M2;4 equals M4;2 ,
M ¼ 2 ðM1;3 þ M2;4 Þ.
dj;k ; d1;3 ¼ d3;1 ; w1 ¼ w3 ¼ w13
1 1 1
GMD1;3 ¼ exp½ln d1;3 f d1;3 2
þ d1;3 4
þ d
1;3 6
þ
12 ðw13 Þ 60 ðw13 Þ 168 ðw13 Þ
1 1
d1;3 8
þ d
1;3 10
þ . . .g
360 ðw13 Þ 660 ðw13 Þ
This expression and that for l1 , when substitute into Q ¼ . . ., yield a mutual
inductance parameter Q1 :
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
l1 l21 ðGMD1;3 Þ2 GMD1;3
Q1 ¼ lnf þ 1þ 2
g 1þ þ
GMD1;3 ðGMD1;3 Þ l21 l1
Now, using Mj;k ¼ 2 l Q and the fact that l1 equals l3 we can write
M1;3 ¼ 2 l1 Q1
M1;3 ¼ 2 l1 Q1
¼ 2 l1 ½lnf þ 1þ g 1 þ þ
GMD1;3 ðGMD1;3 Þ2 l21 l1
Fig. 8.11 Two parallel conductors (not the same length)
However, because l2 does not equal l4 (l2 [ l4 ; l24 ¼ l2 l4 ), we use two par-
allel (no equal length) filament geometry calculation: Two segments of lengths l2
and l4 , respectively, are separated by a Geometric Mean Distance (GMD2-4). In this
case, 2 M2;4 ¼ ðM2 þ M4 Þ M24 and the individual M terms are calculated using
equation M ¼ 2 l Q and the length corresponding to the
Subscript; that is, M2 ¼ 2 l2 Q2 where Q2 is the mutual inductance parameter
GMD
Q for l2 2;4 . We consider w2 ¼ w4 ¼ w24 (Fig. 8.11).
2 M2;4 ¼ ðM2 þ M4 Þ M24 ; M2 ¼ 2 l2 Q2 ; M4 ¼ 2 l4 Q4 ; M24

¼ 2 ðl2 l4 Þ Q24
Q2 ¼ lnf þ 1þ g 1 þ þ
Q4 ¼ lnf þ 1þ g 1 þ þ
1 1 1
GMD2;4 = exp½ln d2;4 f d2;4 2
þ d2;4 4
þ d
2;4 6
12 ðw24 Þ 60 ðw24 Þ 168 ðw24 Þ
1 1
þ d2;4 8
þ d
2;4 10
þ . . .g
360 ðw24 Þ 660 ðw24 Þ
Since l2 equals l1 and the GMD2-4 remains constant, Q2 must equal Q1 (Q2 = Q1)
rffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
l2
as calculated from Q2 ¼ lnfGMD l2
2;4
þ 1 þ ðGMD2 Þ2 g . . .. It follows that
2;4
M2 ¼ M1;3 .
rffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
l2
To obtain Q4 and Q2-4, however, Q4 ¼ l4
lnfGMD 2;4
þ 1 þ ðGMD4 Þ2 g . . . must
2;4
be solved for a GMD2;4 .

X
M ¼ M ¼ f2 ðM1;3 þ M2;4 Þg ¼ 2 M1;3 2 M2;4 ¼ 2 M1;3
½ðM2 þ M4 Þ M24
X
M ¼ 4 l1 Q1 2 ðl2 Q2 þ l4 Q4 Þ þ 2 ðl2 l4 Þ Q24 ; M þ
X
¼0) M\0
X l1 l21 ðGMD1;3 Þ2 GMD1;3
M ¼ 4 l1 flnf þ 1þ 2
g 1þ þ g
GMD1;3 ðGMD1;3 Þ l21 l1
2 ðl2 flnf þ 1þ g 1 þ þ g
þ l4 flnf þ 1þ g 1 þ þ gÞ
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
l4 l24
þ 2 ðl2 l4 Þ flnf þ 1þ g . . .g
GMD2;4 ðGMD2;4 Þ2
Result discussion: Negative mutual inductance (M ) results from coupling

between two conductors having current vectors in opposite directions. In electronic
circuits, negative mutual inductance is usually much smaller in magnitude than
overall inductance that it can be neglected with little effect. It is not so true in
microelectronic circuits. P
The sum of all the mutual inductances both positive and negative ( M) is a
negative value and must be deduce from the sum of the self-inductances in our system
for getting the total inductance of single turn square planar straight thin film inductors
antenna system. Each straight thin film inductor segment is connected to the other by
microstrip line. To analyze our system stability, we represent our equivalent circuit as
a four inductors in series (with series resistance Ri ¼ qAi li i of a straight thin film (strip)
number i; i ¼ 1; 2; 3; 4) and deduce from each segment P inductance expression the
quarter value of the sum of all mutual inductances ( M) (Fig. 8.12). .
1 X 1 X X 1 X
L0i ¼ Li þ M; Li ! Li þ M 8 i ¼ 1; 2; 3; 4; M\0; Li [ j Mj
4 4 4
1 X 1 X 1 X 1 X
L01 ¼ L1 þ M; L02 ¼ L2 þ M; L03 ¼ L3 þ M; L04 ¼ L4 þ M
4 4 4 4
Tau1 ¼ s1 ; Tau2 ¼ s2 ; Tau3 ¼ s3 ; Vsi ! e 8 i ¼ 1; 2; 3; Vs1 ! e; Vs2 ! e; Vs3 ! e

Fig. 8.12 Single turn square planar straight thin film inductors antenna system equivalent circuit
with delay lines
X
2
IL01 ¼ IR1 ; IL02 ¼ IR1 ; IL03 ¼ IR3 ; IL04 ¼ IR4 ; IL02 ðtÞ ¼ IL01 ðt s1 Þ; IL03 ðtÞ ¼ IL02 ðt s2 Þ ¼ IL01 ðt si Þ
i¼1
X
3 X
3
VA1 VA
IL04 ðtÞ ¼ IL03 ðt s3 Þ ¼ IL02 ðt si Þ ¼ IL01 ðt si Þ; IL04 ¼ ILb ; IL01 ¼ ILa ; IRa ¼ ; IRb ¼ 14
i¼2 i¼1
Ra Rb
d
I Cp ¼ Cp ðVA14 VA1 Þ; ILb ¼ ICp þ IRb ; ICp ¼ ILa þ IRa ; VA4 VA5 ; VA7 VA8 ; VA10 VA11
dt
dILa dIL dIL0

VA1 VA2 ¼ La ; VA13 VA14 ¼ Lb b ; VA2 VA3 ¼ L01 1 ;
dt dt dt
0
dIL02
VA5 VA6 ¼ L2
dt
dI L 0 dIL0
VA8 VA9 ¼ L03 3 ; VA12 VA13 ¼ L04 4 ;
dt dt
VA3 VA4 VA6 VA7 VA VA10
IR1 ¼ ; IR 2 ¼ ; IR 3 ¼ 9
R1 R2 R3
VA11 VA12 X4 X4
IR4 ¼ ; VCp þ VLa þ VLb þ VL0i þ VRi ¼ 0
R4 i¼1 i¼1
VA3 VA5 VA VA8

VA4 VA5 ; VA4 ! VA5 ; IR1 ¼ ; VA7 VA8 ; VA7 ! VA8 ; IR2 ¼ 6
R1 R2
VA9 VA11 VA3 VA5
VA10 VA11 ; VA10 ! VA11 ; IR3 ¼ ; IR1 ¼ IL01 ; IR1 ! IL01 ; IL01 ¼
R3 R1
VA6 VA8 VA9 VA11
IR2 ¼ IL02 ; IR2 ! IL02 ; IL02 ¼ ; IR3 ¼ IL03 ; IR3 ! IL03 ; IL03 ¼
R2 R3
VA VA12 X2
IR4 ¼ IL04 ; IR4 ! IL04 ; IL04 ¼ 11 ; IL02 ðtÞ ! IL01 ðt s1 Þ; IL03 ðtÞ ! IL01 ðt si Þ
R4 i¼1
X
3
IL04 ðtÞ ! IL01 ðt si Þ
i¼1
System equation Version 1.0:
X
3
VA1 VA VA VA 5
ILa ¼ IL01 ; ILb ¼ IL04 ¼ IL01 ðt si Þ; IRa ¼ ; IRb ¼ 14 ; IL01 ¼ 3 ;
i¼1
Ra Rb R1
VA VA8 VA VA11 VA VA12 d
IL02 ¼ 6 ; IL03 ¼ 9 ; IL04 ¼ 11 ; ICp ¼ Cp ðVA14 VA1 Þ;
R2 R3 R4 dt
X 3 dIL0
IL01 ðt si Þ ¼ ICp þ IRb ; ICp ¼ IL01 þ IRa ; VA1 VA2 ¼ La 1 ;
i¼1
dt
P
3
dIL01 ðt si Þ
dIL01
VA13 VA14 ¼ Lb i¼1
; VA2 VA3 ¼ L01
dt dt
P
2
si Þ dIL01 ðt
dI ðt s1 Þ
L01
VA5 VA6 ¼ L02 0
; VA8 VA9 ¼ L3 i¼1
;
dt dt
P3
dIL0 ðt i¼1 si Þ X4 X4
VA12 VA13 ¼ L04 1 ; VCp þ VLa þ VLb þ VL0i þ VRi ¼ 0
dt i¼1 i¼1
&&&
VA1 ¼ IRa Ra ; VA14 ¼ IRb Rb ; VA3 VA5 ¼ IL01 R1 ; VA6 VA8 ¼ IL02 R2 ;
VA9 VA11 ¼ IL03 R3
d d
VA11 VA12 ¼ IL04 R4 ; ICp ¼ Cp ðVA14 VA1 Þ ¼ Cp ðIRb Rb IRa Ra Þ
dt dt
dIR dIR dIL0 dIL0
ICp ¼ Cp Rb b Cp Ra a ; VA1 VA2 ¼ La 1 ) VA2 ¼ VA1 La 1
dt dt dt dt
dIL01 dILb dILb dIL0
VA2 ¼ IRa Ra La ; VA13 VA14 ¼ Lb ) VA13 ¼ VA14 þ Lb ¼ VA14 þ Lb 4
dt dt dt dt
P
3
dIL01 ðt si Þ
dIL01 dIL01
VA13 ¼ IRb Rb þ Lb i¼1
; VA2 VA3 ¼ L01 ) VA3 ¼ VA2 L01
dt dt dt
dIL01 dIL01 dIL01
V A3 ¼ I Ra R a L a L01 ¼ IRa Ra ðLa þ L01 Þ
dt dt dt
P
3 P
3
dIL01 ðt si Þ dIL01 ðt si Þ
VA12 VA13 ¼ L04 i¼1
) VA12 ¼ VA13 þ L04 i¼1
dt dt
P
3 P
3 P
3
dIL01 ðt si Þ dIL01 ðt si Þ dIL01 ðt si Þ
VA12 ¼ IRb Rb þ Lb i¼1
þ L04 i¼1
¼ IRb Rb þ ðLb þ L04 Þ i¼1
dt dt dt
VA3 VA5
IL01 ¼ ) VA3 VA5 ¼ IL01 R1 ;
R1
dIL01
VA5 ¼ VA3 IL01 R1 ¼ IRa Ra ðLa þ L01 Þ IL01 R1
dt
dIL02 dIL01 ðt s1 Þ dIL01
VA5 VA6 ¼ L02 ¼ L02 ; VA6 ¼ VA5 L02
dt dt dt
dIL01 dIL01 ðt s1 Þ
VA6 ¼ IRa Ra ðLa þ L01 Þ IL01 R1 L02
dt dt
VA VA8
IL02 ¼ 6 ) VA6 VA8 ¼ IL02 R2 ;
R2
VA8 ¼ VA6 IL02 R2 ¼ VA6 IL01 ðt s1 Þ R2
dIL01 dIL01 ðt s1 Þ
VA8 ¼ IRa Ra ðLa þ L01 Þ IL01 R1 L02 IL01 ðt s1 Þ R2
dt dt
VA11 VA12 X
3
IL04 ¼ ) VA11 VA12 ¼ IL04 R4 ¼ IL01 ðt si Þ R 4 ;
R4 i¼1
X
3
VA11 ¼ VA12 þ IL01 ðt si Þ R 4
i¼1
P
3
dIL01 ðt si Þ X
3
VA11 ¼ IRb Rb þ ðLb þ L04 Þ i¼1
þ IL01 ðt s i Þ R4
dt i¼1
VA9 VA11
IL03 ¼ ) VA9 VA11 ¼ IL03 R3 ;
R3
X
2
VA 9 ¼ VA11 þ IL03 R3 ¼ VA11 þ IL01 ðt si Þ R 3
i¼1
P
3
dIL01 ðt si Þ X
3 X
2
VA9 ¼ IRb Rb þ ðLb þ L04 Þ i¼1
þ IL01 ðt si Þ R4 þ IL01 ðt si Þ R 3
dt i¼1 i¼1
X
4 X
4 Z
1
VCp þ VLa þ VLb þ VL01 þ VRi ¼ 0; VCp ¼ VA14 VA1 ¼ ICp dt;
i¼1 i¼1
Cp
dILa
VLa ¼ VA1 VA2 ¼ La
dt
dIL X4
VLb ¼ VA13 VA14 ¼ Lb b ; VL0 ¼ VL01 þ VL02 þ VL03 þ VL04 ;
dt i¼1 i
dIL01
VL01 ¼ VA2 VA3 ¼ L01
dt
P
2
dIL01 ðt si Þ
dIL01 ðt s1 Þ
VL02 ¼ VA5 VA6 ¼ L02 ; VL03 ¼ VA8 VA9 ¼ L03 i¼1
dt dt
P
3
dIL01 ðt si Þ X4
VL04 ¼ VA12 VA13 ¼ L04 i¼1
; VRi ¼ VR1 þ VR2 þ VR3 þ VR4 ;
dt i¼1
VR1 ¼ VA3 VA4 ¼ IL01 R1
q1 l1 q l2
VR1 ¼ IL01 ; VR2 ¼ VA6 VA7 ¼ R2 IL01 ðt s1 Þ ¼ IL01 ðt s1 Þ 2
w1 z1 w 2 z2
X2 X2
q l3
VR3 ¼ VA9 VA10 ¼ R3 IL03 ¼ R3 IL01 ðt si Þ ¼ IL01 ðt si Þ 3
i¼1 i¼1
w3 z3
X
3 X
3
q4 l4
VR4 ¼ VA11 VA12 ¼ R4 IL04 ¼ R4 IL01 ðt si Þ ¼ IL01 ðt si Þ
i¼1 i¼1
w4 z4
X
4
q 1 l1 q l2 X2
q l3
VRi ¼ IL01 þ IL01 ðt s1 Þ 2 þ IL01 ðt si Þ 3
i¼1
w 1 z1 w 2 z2 i¼1
w3 z3
X
3
q4 l4
þ IL01 ðt si Þ
i¼1
w4 z4
X
4
q1 l1 X4
qi li X
i1 X4 dIL0
VRi ¼ IL01 þ IL01 ðt sk Þ; VL01 ¼ L01 1
i¼1
w1 z1 w zi
i¼2 i k¼1 i¼1
dt
iP
1
X
4 dIL01 ðt sk Þ
L0i
k¼1
þ
i¼2
dt
P
2 P
3
X dIL01 ðt si Þ dIL01 ðt si Þ
4 dIL01 dIL01 ðt s1 Þ
VL01 ¼ L01 þ L02 þ L03 i¼1
þ L04 i¼1
i¼1
dt dt dt dt
P
3
dIL01 ðt si Þ
dIL01 dIL04 i¼1
ILa ¼ IL01 ; VLa ¼ La ; ILb ¼ IL04 ) VLb ¼ Lb ¼ Lb
dt dt dt
X
4 X
4
VCp þ VLa þ VLb þ VL0i þ VRi ¼ 0
i¼1 i¼1
P
3 P
i1
Z dIL01 ðt si Þ X4 dIL01 ðt sk Þ
1 dIL0 dIL 0
I Cp dt þ La 1 þ Lb i¼1
þ L01 1 þ L0i k¼1
Cp dt dt dt i¼2
dt
q1 l1 X4
qi li X
i1
þ IL01 þ IL01 ðt sk Þ ¼ 0
w1 z1 w zi
i¼2 i k¼1
We derivate the above equation and get the following expression: (Table 8.17)
P
3 P
i1
2 d 2 IL01 ðt si Þ 2 X
4 d 2 IL01 ðt sk Þ
1 d I L0 d IL01
I C p þ La 2 1 þ L b i¼1
þ L01 þ L01 k¼1
Cp dt dt2 dt2 i¼2
dt2
iP
1
X dIL01 ðt sk Þ
q1 l1 dIL01 4
qi li k¼1
þ þ ¼0
w1 z1 dt i¼2
w i zi dt
P
3 P
i1
d 2 IL01 ðt si Þ X d 2 IL01 ðt sk Þ
d 2 IL0 d 2 0
IL
4
ICp ¼ Cp fLa 2 1 þ Lb i¼1
þ L01 2 1 þ L0i k¼1
dt dt2 dt i¼2
dt2
ðÞ
P
i1
X dIL01 ðt sk Þ
q1 l1 dI 4
qi li
L01 k¼1
þ þ g
w1 z1 dt w zi
i¼2 i
dt
Table 8.17 Single turn Case no. s1 s2 s3

square planar straight thin
film inductors antenna s1, s2, 1 s 0 0
and s3 options 2 0 s 0
3 0 0 s
4 s s 0
5 0 s s
6 s 0 s
7 s s s
X
3 X
3
IL01 ðt si Þ ¼ ICp þ IRb ) IRb ¼ IL01 ðt si Þ ICp ; ICp ¼ IL01 þ IRa ) IRa ¼ ICp IL01
i¼1 i¼1
d d X3
I C p ¼ Cp ðIRb Rb IRa Ra Þ ¼ Cp ð½IL01 ðt si Þ ICp Rb ½ICp IL01 Ra Þ
dt dt i¼1
P
3
dIL01 ðt si Þ
i¼1 dIL01 dICp
ðÞ ICp ¼ Cp ðRb þ Ra ½Rb þ Ra Þ
dt dt dt
Derivative equation (*):
P
3 iP
1
d 3 IL01 ðt si Þ X d 3 IL01 ðt sk Þ
dICp d 3 IL01 d 3 IL01 4
þ L01 L0i
i¼1 k¼1
¼ Cp fLa 3 þ Lb þ
dt dt dt3 dt3 i¼2
dt3
iP
1
2 X d 2 IL01 ðt sk Þ
q1 l1 d IL01 4
qi li k¼1
þ þ g
w1 z1 dt2 i¼2
wi zi dt2
Expression (**):
P
3 P
i1
d I L01 d I L01
Cp fLa þ Lb i¼1
þ L01 þ L0i k¼1
dt2 dt2 dt2 i¼2
dt2
iP
1 P3
X dIL01 ðt sk Þ dIL01 ðt si Þ
q1 l1 dIL01 4
qi li k¼1 i¼1 dIL01
þ þ g ¼ Cp ðRb þ Ra
w1 z1 dt w zi
i¼2 i
dt dt dt
P
3 iP
1
d IL01 d IL01
þ Cp fLa þ Lb i¼1
þ L01 þ L01 k¼1
dt3 dt3 dt3 i¼2
dt3
iP
1
2 0 X d 2 IL01 ðt sk Þ
q l1 d I L
4
qi li k¼1
þ 1 21 þ g ½Rb þ Ra Þ
w1 z1 dt w zi
i¼2 i
dt2
Remark 1.0
P
3
P
d 2 IL01 ðt si Þ X
3 kðt
3
si Þ
i¼1
; IL01 ðt si Þ ¼ IL0 þ iL01 e i¼1 ; IL01 ðtÞ ¼ IL0 þ iL01 ekt
dt2 i¼1
1 1
P
3
P P
3
d 2 IL01 ðt si Þ 3
d 2 IL01 ðt si Þ
d 2 IL01 ðtÞ i¼1
kðt si Þ
i¼1
¼ iL01 k2 ekt ; ¼ iL01 k2 e i¼1 ;
dt2 dt2 dt2
P
3
d 2 IL01 ðtÞ k si
¼ e i¼1
dt2
Remark 1.1
iP
1 P
2
X d 2 IL0i ðt sk Þ d 2 IL01 ðt sk Þ
4 d IL01 ðt s1 Þ
2
L0i k¼1
¼ L02 þ L03 k¼1
þ L04
i¼2
dt2 dt2 dt2
P 3
d 2 IL01 ðt sk Þ
k¼1

dt2
P
2
X
2 kðt sk Þ
IL01 ðt s1 Þ ¼ IL0 þ iL01 e kðts1 Þ
; IL01 ðt sk Þ ¼ IL0 þ iL01 e k¼1 ;
1 1
k¼1
P
3
X
3 kðt sk Þ
IL01 ðt sk Þ ¼ IL0 þ iL01 e k¼1
1
k¼1
d 2 IL01 ðtÞ d 2 IL01 ðt s1 Þ
IL01 ðtÞ ¼ IL0 þ iL01 ekt ; ¼ iL01 k2 ekt ; ¼ iL01 k2 ekt eks1
1 dt2 dt2
P
2
sk Þ P
d 2 IL01 ðt 2
d IL01 ðt s1 Þ d IL01 ðtÞ ks
2 2
k¼1
k sk
2
¼ 2
e 1
; 2
¼ iL01 k e e k¼1
2 kt
dt dt dt
P2 P
3
d 2 IL01 ðt sk Þ P
2
d 2 0
I ðt sk Þ P3
k¼1 d IL1 ðtÞ k sk

2 0 L 1
k¼1 d 2 IL01 ðtÞ k sk
¼ e k¼1 ; ¼ e k¼1
dt2 dt2 dt2 dt2
iP
1
d 2 IL0i ðt sk Þ P
2 P
3
X
4 d 2 IL01 ðtÞ d 2 IL01 ðtÞ k sk d 2 IL01 ðtÞ k sk
L0i k¼1
¼ L02 e ks1
þ L03 e k¼1 þ L04 e k¼1
i¼2
dt2 dt2 dt2 dt2
P
i1
P P
X d 2 IL0i ðt sk Þ 2 3
4 d 2 IL01 ðtÞ k sk k sk
L0i k¼1
¼ ½L02 e ks1
þ L03 e k¼1 þ L04 e k¼1
i¼2
dt2 dt2
P
i1
P
X d 2 IL0i ðt sk Þ j1
4 d 2 IL01 ðtÞ X 4 k sk
L0i k¼1
¼ 0
Lj e k¼1
i¼2
dt2 dt 2
j¼2
Remark 1.2
iP
1
X dIL01 ðt sk Þ
4
q i li k¼1 dIL01 ðt s1 Þ dIL01 ðtÞ ks
; ¼ e 1
i¼2
w i zi dt dt dt
iP
1
X dIL01 ðt sk Þ
4
qi li k¼1 q2 l2 dIL01 ðt s1 Þ q l3
¼ þ 3
i¼2
wi zi dt w 2 z2 dt w3 z3
P2 P3
dIL01 ðt sk Þ dIL01 ðt sk Þ
k¼1 q 4 l4 k¼1
þ
dt w 4 z4 dt
dIL01 ðt s1 Þ
IL01 ðt s1 Þ ¼ IL0 þ iL01 ekðts1 Þ ; ¼ iL01 k ekt eks1 ; IL01 ðtÞ
1 dt
dIL0 ðtÞ
¼ IL0 þ iL01 ekt ; 1 ¼ iL01 k ekt
1 dt
P
3
P2 P
2
dIL01 ðt sk Þ P
3
dIL01 ðt k¼1 sk Þ dIL01 ðtÞ k sk
k¼1 dIL01 ðtÞ k sk
¼ e k¼1 ; ¼ e k¼1
dt dt dt dt
Pi1
X4
qi li dIL01 ðt k¼1 sk Þ q2 l2 dIL01 ðtÞ ks1
¼ e
w zi
i¼2 i
dt w 2 z2 dt
P2 P 3
q3 l3 dIL01 ðtÞ k sk q4 l4 dIL01 ðtÞ k sk

þ e k¼1 þ e k¼1
w 3 z3 dt w4 z4 dt
Pi1
X4
qi li dIL01 ðt k¼1 sk Þ dIL01 ðtÞ q2 l2 ks1
¼ ½ e
w zi
i¼2 i
dt dt w2 z2
P2 P3
q3 l3 k sk q4 l4 k sk
þ e k¼1 þ e k¼1
w 3 z3 w 4 z4
Pi1 P
j1
X4
qi li dIL01 ðt k¼1 sk Þ dIL01 ðtÞ X 4
qj lj k sk
¼ e k¼1
w zi
i¼2 i
dt dt w zj
j¼2 j
Remark 1.3
P3 P
3
dIL01 ðt i¼1 si Þ dIL01 ðtÞ k si
¼ e i¼1
dt dt
Remark 1.4
P P3
d 3 IL01 ðt 3i¼1 si Þ d 3 IL01 ðtÞ k si
¼ e i¼1
dt3 dt3
Remark 1.5
Pi1 Pj1
X
4 d 3 IL01 ðt k¼1 sk Þ d 3 IL01 ðtÞ X 4 k sk
L0i ¼ L0
j e k¼1
i¼2
dt3 dt3 j¼2
Pi1 P2
X
4 d 3 IL01 ðt k¼1 sk Þ d 3 IL01 ðt s1 Þ d 3 IL01 ðt k¼1 sk Þ
L0i ¼ L02 þ L03 þ L04
dt3 dt3 dt3
i¼2
P
d 3 IL01 ðt 3k¼1 sk Þ

dt3
P2 P
2
d 3 IL01 ðt s1 Þ d 3 IL01 ðtÞ ks1
d 3 IL01 ðt k¼1 sk Þ d 3 IL01 ðtÞ k sk
¼ e ; ¼ e k¼1
dt3 dt3 dt3 dt3
P P3
d 3 IL01 ðt 3k¼1 sk Þ d 3 IL01 ðtÞ k sk
¼ e k¼1
dt3 dt3
Pi1 P
2
X
4 d 3 IL01 ðt k¼1 sk Þ d 3 IL01 ðtÞ d 3 IL01 ðtÞ k sk
L0i ¼ L02 e ks1
þ L03 e k¼1
i¼2
dt3 dt3 dt3
P3
d 3 0
I L ðtÞ k sk
þ L04 1
e k¼1
dt3
P P2
X
4 d 3 IL01 ðt i1k¼1 sk Þ
k sk
L0i 3
¼ ½L02 eks1 þ L03 e k¼1
i¼2
dt
P3
0
k sk d 3 IL01 ðtÞ
þ L4 e k¼1
dt3
Remark 1.6
Pi1 P
j1
X4
qi li d 2 IL01 ðt k¼1 sk Þ d 2 IL01 ðtÞ X 4
qj lj k sk
¼ e k¼1
w zi
i¼2 i
dt2 dt2 w zj
j¼2 j
iP
1
X d 2 IL01 ðt sk Þ
4
qi li k¼1 q2 l2 d 2 IL01 ðt s1 Þ q l3
¼ þ 3
i¼2
w i z i dt 2 w 2 z 2 dt 2 w 3 z3
P2 P3
d 2 IL01 ðt sk Þ d 2 IL01 ðt sk Þ
k¼1 q4 l4 k¼1
þ
dt2 w4 z4 dt2
P
2
d 2 IL01 ðt sk Þ
d IL01 ðt s1 Þ
2
d IL01 ðtÞ
2
eks1 ;
k¼1
¼
dt2 dt2 dt2
P
3
P
2
d 2 IL01 ðt sk Þ P3
d IL01 ðtÞ k sk
2
k¼1 d 2 IL01 ðtÞ k sk
¼ e k¼1 ; ¼ e k¼1
dt2 dt2 dt2
Pi1 P2
X4
qi li d 2 IL01 ðt k¼1 sk Þ q2 l2 d 2 IL01 ðtÞ ks1 q3 l3 d 2 IL01 ðtÞ k sk
¼ e þ e k¼1
w zi
i¼2 i
dt2 w 2 z2 dt2 w 3 z3 dt2
P 3
q4 l4 d 2 IL01 ðtÞ k k¼1 sk
þ e
w4 z 4 dt2
Pi1 P
2
X4
qi li d 2 IL01 ðt k¼1 sk Þ d 2 IL01 ðtÞ q2 l2 ks1 q3 l3 k k¼1 sk
¼ ½ e þ e
w zi
i¼2 i
dt2 dt2 w2 z 2 w 3 z3
P 3
q4 l4 k k¼1 sk
þ e
w4 z 4
We can summary remarks 1.0–1.6 in Table 8.16.
Expression (**):
P3 P
j1
d 2 IL01 d 2 IL01 ðtÞ k si d 2 0
I L d 2 0
I L ðtÞ X4 k sk
Cp fLa 2 þ Lb 2
e i¼1 þ L01 2 1 þ 1
2
L0j e k¼1
dt dt dt dt j¼2
P
j1 P3
q1 l1 dIL01 dIL01 ðtÞ X 4

qj lj k sk dIL01 ðtÞ k si dIL01
þ þ e k¼1 g ¼ Cp ðRb e i¼1 þ Ra
w1 z1 dt dt w zj
j¼2 j
dt dt
P3 P
j1
d 3 IL01 d 3 IL01 ðtÞ k si 0
d 3 IL01 d 3 IL01 ðtÞ X 4
0
k sk
þ Cp fLa 3 þ Lb e i¼1 þ L 1 þ L j e k¼1
dt dt3 dt3 dt3 j¼2
P
j1
2
q1 l1 d IL01 d 2 IL01 ðtÞ X 4
qj lj k sk
þ þ e k¼1 g ½Rb þ Ra Þ
w1 z1 dt2 dt2 w zj
j¼2 j
We define for simplicity the following functions:
P
3 P
j1
k si X
4 k sk
n1 ¼ n1 ðk; si ; i ¼ 1; 2; 3Þ ¼ e i¼1 ; n2 ¼ n2 ðk; L0j ; j ¼ 2; 3; 4; sk ; k ¼ 1; 2; 3Þ ¼ L0j e k¼1
j¼2
P
j1
X
4
qj lj k sk
n3 ¼ n3 ðk; sk ; k ¼ 1; 2; 3; qj ; lj ; wj ; zj ; j ¼ 1; 2; 3; 4Þ ¼ e k¼1
j¼2
wj zj
d 2 IL01 ðtÞ d 2 IL01 ðtÞ d 2 IL01 ðtÞ d 2 IL01 ðtÞ q l1 dIL01 ðtÞ dIL01 ðtÞ
Cp fLa þ Lb n1 þ L01 þ n2 þ 1 þ n3 g
dt 2 dt 2 dt 2 dt 2 w1 z1 dt dt
dIL0 ðtÞ dIL01 ðtÞ d 3 IL01 ðtÞ d 3 IL01 ðtÞ d 3 IL01 ðtÞ
¼ Cp ðRb 1 n1 þ Ra þ Cp fLa þ Lb n1 þ L01
dt dt dt3 dt3 dt3
d 3 IL01 ðtÞ q1 l1 d 2 IL01 ðtÞ d 2 IL01 ðtÞ
þ n2 þ þ n3 g ½Rb þ Ra Þ
dt3 w1 z1 dt2 dt2
We define new variables: X1 ¼ X1 ðtÞ; X2 ¼ X2 ðtÞ
dIL01 ðtÞ dX1 ðtÞ d IL01 ðtÞ dX2 ðtÞ d IL01 ðtÞ
2 3
X1 ðtÞ ¼ ; X2 ðtÞ ¼ ¼ 2
; ¼ ; IL01 ¼ IL01 ðtÞ
dt dt dt dt dt3
q1 l1
Cp fLa X2 þ Lb X2 n1 þ L01 X2 þ X2 n2 þ X 1 þ X 1 n3 g
w1 z1
¼ Cp ðRb X1 n1 þ X1 Ra
dX2 dX2 dX2 dX2 q l1
þ Cp fLa þ Lb n1 þ L01 þ n2 þ 1 X2 þ X2 n3 g ½Rb þ Ra Þ
dt dt dt dt w1 z1
dX2 ½wq11 lz11 þ n3 þ Rb n1 þ Ra

¼ X1
dt fLa þ Lb n1 þ L01 þ n2 g ½Rb þ Ra Cp
½La þ Lb n1 þ L01 þ n2 þ Cp ½Rb þ Ra ðwq11 lz11 þ n3 Þ
X2
fLa þ Lb n1 þ L01 þ n2 g ½Rb þ Ra Cp
q1 l1
½ þ n3 þ Rb n1 þ Ra
We define new functions: w1 ¼ fLa þ Lwb1n
z1
0
1 þ L þ n2 g½Rb þ Ra Cp 1
½La þ Lb n1 þ L01 þ n2 þ Cp ½Rb þ Ra ðwq11 lz11 þ n3 Þ

w2 ¼
fLa þ Lb n1 þ L01 þ n2 g ½Rb þ Ra Cp
dIL01 dX1 dX2

¼ X1 ; ¼ X2 ; ¼ w1 X1 w2 X2
dt dt dt
Case 1.0: No delays, s1 ¼ s2 ¼ s3 ¼ 0
n1 ðsi ¼ 0Þ ¼ n1 ðk; si ¼ 0; i ¼ 1; 2; 3Þ ¼ 1; n2 ðsk ¼ 0Þ ¼ n2 ðk; L0j ; j ¼ 2; 3; 4;

X
4
sk ¼ 0; k ¼ 1; 2; 3Þ ¼ L0j
j¼2
X
4
qj lj
n3 ðsk ¼ 0Þ ¼ n3 ðk; sk ¼ 0; k ¼ 1; 2; 3; qj ; lj ; wj ; zj ; j ¼ 1; 2; 3; 4Þ ¼
j¼2
wj zj
P4qj lj
½j¼1 wj zj þ Rb þ Ra
w1 ðsi;k ¼ 0Þ ¼ P ; w2 ðsi;k ¼ 0Þ
fLa þ Lb þ 4j¼1 L0j g ½Rb þ Ra Cp
P P q l
½La þ Lb þ 4j¼1 L0j þ Cp ½Rb þ Ra 4j¼1 wjj zjj
¼ P
fLa þ Lb þ 4j¼1 L0j g ½Rb þ Ra Cp
dIL0

At fixed point: dt1 ¼ 0; dX dt ¼ 0; dt ¼ 0; X1 ¼ 0; X2 ¼ 0
1 dX2
equilibrium points of Single turn square planar straight thin film inductors antenna
system consists in adding to its coordinated [X1 X2 IL01 ] arbitrarily small increments
of exponential terms ½x1 x2 iL01 ekt , and retaining the first order terms in x1 x2 iL01 . The
system of eight homogeneous equations leads to a polynomial characteristic
equation in the eigenvaluek. The polynomial characteristic equation accepts by set
the Single turn square planar straight thin film inductors antenna system equations.
The Single turn square planar straight thin film inductors antenna system fixed
values with arbitrarily small increments of exponential form ½x1 x2 iL01 ekt are; i = 0
(first fixed point), i = 1 (second fixed point), i = 2 (third fixed point), etc., [2–4].
ðiÞ ðiÞ ðiÞ

X1 ðtÞ ¼ X1 þ x1 ekt ; X2 ðtÞ ¼ X2 þ x2 ekt ; IL01 ðtÞ ¼ IL0 þ iL01 ekt
1
dX1 ðtÞ dX2 ðtÞ dIL0 ðtÞ

¼ x1 k ekt ; ¼ x2 k ekt ; 1 ¼ iL01 k ekt
dt dt dt
dIL01 ðiÞ ðiÞ
dt
dX1 ðiÞ ðiÞ
¼ X2 ) x1 k ekt ¼ X2 þ x2 ekt ; X2 ¼ 0; x1 k þ x2 ¼ 0
dt
dX2
¼ w1 ðsi;k ¼ 0Þ X1 w2 ðsi;k ¼ 0Þ X2
dt
ðiÞ ðiÞ
x2 k ekt ¼ w1 ðsi;k ¼ 0Þ ½X1 þ x1 ekt w2 ðsi;k ¼ 0Þ ½X2 þ x2 ekt
ðiÞ ðiÞ
x2 k ekt ¼ w1 ðsi;k ¼ 0Þ X1 w2 ðsi;k ¼ 0Þ X2 w1 ðsi;k ¼ 0Þ x1 ekt
w2 ðsi;k ¼ 0Þ x2 ekt
ðiÞ ðiÞ
At fixed points: w1 ðsi;k ¼ 0Þ X1 w2 ðsi;k ¼ 0Þ X2 ¼ 0
x2 k w1 ðsi;k ¼ 0Þ x1 w2 ðsi;k ¼ 0Þ x2 ¼ 0
We can summary our single turn square planar straight thin film inductors
antenna system arbitrarily small increments equations:
x1 k þ x2 ¼ 0; x2 k w1 ðsi;k ¼ 0Þ x1 w2 ðsi;k ¼ 0Þ x2 ¼ 0; iL01 k þ x1

¼0
0 10 x 1
k 1 0 1
@ w1 ðsi;k ¼ 0Þ k w2 ðsi;k ¼ 0Þ 0 AB C
@ x2 A ¼ 0; A k I
1 0 k iL01
0 1
k 1 0
¼ @ w1 ðsi;k ¼ 0Þ k w2 ðsi;k ¼ 0Þ 0 A
1 0 k
0 1
k 1 0
B C
detðA k IÞ ¼ 0; det@ w1 ðsi;k ¼ 0Þ k w2 ðsi;k ¼ 0Þ 0 A ¼ 0
1 0 k

k w2 ðsi;k ¼ 0Þ 0 w1 ðsi;k ¼ 0Þ 0
detðA k IÞ ¼ k det det
0 k 1 k
detðA k IÞ ¼ k ðk þ w2 ðsi;k ¼ 0ÞÞ k w1 ðsi;k ¼ 0Þ k ¼ k ½ðk þ w2 ðsi;k ¼ 0ÞÞ k þ w1 ðsi;k ¼ 0Þ
detðA k IÞ ¼ 0 ) k1 ¼ 0; k2 þ w2 ðsi;k ¼ 0Þ k þ w1 ðsi;k ¼ 0Þ ¼ 0
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
w2 ðsi;k ¼ 0Þ ½w2 ðsi;k ¼ 0Þ 2 4 w1 ðsi;k ¼ 0Þ
k2;3 ¼ ;
2
w1 ðsi;k ¼ 0Þ [ 0; w2 ðsi;k ¼ 0Þ [ 0
rffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
w ðs ¼0Þ
w2 ðsi;k ¼ 0Þ ½w2 ðsi;k ¼ 0Þ 2 4 ½w 1ðs i;k¼0Þ 2 ½w2 ðsi;k ¼ 0Þ 2
2 i;k
k2;3 ¼
2
rffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
w ðs ¼0Þ
w2 ðsi;k ¼ 0Þ w2 ðsi;k ¼ 0Þ 1 4 ½w 1ðs i;k¼0Þ 2
2 i;k
k2;3 ¼ ;
2 sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
1 w ðsi;k ¼ 0Þ
k2;3 ¼ w2 ðsi;k ¼ 0Þ f1 1 4 1 g
2 ½w2 ðsi;k ¼ 0Þ 2
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
w ðsi;k ¼ 0Þ 1
1ðw1 ; w2 Þ ¼ 1 1 4 1 ; k ¼ w2 ðsi;k ¼ 0Þ 1ðw1 ; w2 Þ
2 2;3
½w2 ðsi;k ¼ 0Þ 2
The sign of 1ðw1 ; w2 Þ establish the sign of eigenvalues k2 and k3

(w2 ðsi;k ¼ 0Þ [ 0)
w ðs ¼0Þ w ðs ¼0Þ
Case a: 1 4 ½w 1ðs i;k¼0Þ 2 [ 0 ) 0\ ½w 1ðs i;k¼0Þ 2 \ 14
2 i;k 2 i;k
w ðsi;k ¼ 0Þ w ðsi;k ¼ 0Þ
0\f 1 4 1 2
g\1; 1 [0
½w2 ðsi;k ¼ 0Þ ½w2 ðsi;k ¼ 0Þ 2
) 1ðw1 ; w2 Þ\0; k2 \0; k3 \0
w1 ðsi;k ¼ 0Þ w1 ðsi;k ¼ 0Þ 1
Case b: 1 4 2
¼0) 2
¼
½w2 ðsi;k ¼ 0Þ ½w2 ðsi;k ¼ 0Þ 4
1
1ðw1 ; w2 Þ ¼ 1; k2;3 ¼ w2 ðsi;k ¼ 0Þ; k2 \0; k3 \0
2
w1 ðsi;k ¼ 0Þ w1 ðsi;k ¼ 0Þ 1
Case c: 1 4 2
\0 ) 2
[
½w2 ðsi;k ¼ 0Þ ½w2 ðsi;k ¼ 0Þ 4
w ðsi;k ¼ 0Þ
1ðw1 ; w2 Þ ¼ a b j; 1ðw1 ; w2 Þ ¼ 1 1 4 1
½w2 ðsi;k ¼ 0Þ 2
w1 ðsi;k ¼ 0Þ pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi

iðw1 ; w2 Þ ¼ 1 4 2
; iðw1 ; w2 Þ\0; iðw1 ; w2 Þ ¼ j jiðw1 ; w2 Þj
½w2 ðsi;k ¼ 0Þ
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
1ðw1 ; w2 Þ ¼ 1 j jiðw1 ; w2 Þj; a ¼ 1; b ¼ jiðw1 ; w2 Þj
1 1
k2;3 ¼ w2 ðsi;k ¼ 0Þ 1ðw1 ; w2 Þ ¼ w2 ðsi;k ¼ 0Þ ða b jÞ; w2 ðsi;k ¼ 0Þ [ 0
2 2
We get three eigenvalues for our single turn square planar straight thin film
inductors antenna system. The first eigenvalue (k1) is equal to zero and other
eigenvalues (k2, k3) can be real numbers (k2 ; k3 2 R) and negative (cases a
and b) or complex conjugate numbers (k2;3 ¼ 12 w2 ðsi;k ¼ 0Þ ða b jÞ;
w2 ðsi;k ¼ 0Þ [ 0).
If the first eigenvalue (k1 = 0) and two other eigenvalues (k2 ; k3 2 R) are negative
and real numbers then our single turn square planar straight thin film inductors
antenna system fixed point is attracting line. If the first eigenvalue (k1 = 0) and two
other eigenvalues (k2 ; k3 2 C) are complex conjugate numbers with negative real
part (a ¼ 1; a\0) then our single turn square planar straight thin film inductors
antenna system fixed point is attracting stable spiral node line [2–4].
The next cases we analyze are when at least one of the delay parameters is real
and positive value. We analyze our single turn square planar straight thin film
inductors antenna system for the following cases:
Remark: It is reader exercise to analyze single turn square planar straight thin film
inductors antenna system stability for cases 1 to 7 (at least one delay parameter
s1 ; s2 ; s3 is positive and real number). The stability analysis is done by using geo-
metric stability switch criteria in delay differential systems (E. Beretta and Y. Kuang).
It is a practical guideline that combines graphical information with analytical work to
effectively study the local stability of single turn square planar straight thin film
inductors antenna system model involving delay dependent parameters. The stability
of single turn square planar straight thin film inductors antenna system steady state
is determined by the graphs of some functions of s which can be expressed explicitly
[5, 6].
switching resulting from the increase of value of the time delay s for our single turn
square planar straight thin film inductors antenna system characteristic equation.
Dðk; s1 ; s2 ; s3 Þ ¼ detðA k IÞ; Dðk; s1 ; s2 ; s3 Þ ¼ 0; Dðk; sÞ ¼ Pn ðk; sÞ þ Qm ðk; sÞ eks

X n X
m
k¼0 k¼0
pk ðÞ; qk ðÞ : R þ 0 ! R are continuous and differentiable functions of s.
assume:
(II) lim supfjQm ðk; sÞ=Pn ðk; sÞj : jkj ! 1; Rek 0g\1 for any s.
whenever it exists.
8.4 Helix Antennas System Stability Analysis Under Parameters Variation 873
8.4 Helix Antennas System Stability Analysis Under

The helix antenna is a type of antenna which uses curved segments. The main
issues regarding helix antenna are helix geometry and parameters, wire connection
concepts, and the Perfect Electric Conductor (PEC) ground plane. A helical antenna
is an antenna consisting of a conducting wire wound in the form of a helix. Helical
antennas are mounted over a ground plane. The feed line is connected between the
bottom of the helix and the ground plane.
Helical antennas can operate in one of two principal modes—normal mode or
axial mode. In the normal mode or broad side helix, the dimensions of the helix are
small compared with the wavelength. The antenna acts similarly to an electrically
short dipole or monopole, and the radiation pattern has a maximum radiation at
right angles to the helix axis. Manly used for compact antennas for portable and
mobile two way radios, and for UHF broadcasting antennas. Additionally the
normal mode helical antenna (NMHA) is used for applications such as mobile and
satellite communication, RFID and medical devices. In the axial mode or end-fire
helix, the dimensions of the helix are comparable to a wavelength. The antenna
functions as a directional antenna radiating a beam off the ends of the helix, along
the antenna’s axis. It radiates circularly polarized radio waves and used for satellite
communications. The normal mode helical antenna (NMHA) lumped-element
equivalent circuit is represented the input impedance of a normal mode helical
antenna. The normal mode helical antenna (NMHA) is modeled as the combination
of a wire dipole antenna and distributed inductors. The equivalent circuit is divided
into two parts: a five element circuit for the equivalent wire antenna with a different
radius, and a three element sub-circuit representing the effects of the loops. The
geometry of a typical helix antenna is shown in the below figure. There are some
important parameters which characterize our helix antenna. The helix wire of
radius, a, and uniformly wound with a constant pitch, S. The diameter of the
imagined cylinder over which the axis of the helical conductor is wrapped is D and
the radius is R (D = 2 • R). The number of turns is N and the half axial length of the
antenna is h = N • S. We can modeled the helix antenna as a series of loops and
linear conductors when the physical dimensions of the helix are much smaller than
the wavelength [130–132]. We can represent each helix antenna turn as two radi-
ating components, one an axial wire segment of length S, and the other a loop of
diameter D. The wire and loop model can be representing as a short wire segments
connected by lumped elements representing the inductance of the loop, where the
loops are functioned as inductors (each one inductance L). Actually the helix
structure model is a wire segment with a length of S and one inductive element per
turn. We define new radius for the helix antenna wire segment (a0 ; a ! a0 ) since we
need to keep the correct capacitance of the helix turn (a0 ¼ S ðalÞ½ l ). The length of
S
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
helix antenna one turns, l where l ¼ ðp DÞ2 þ S2 . The induced magnetic field of
a single turn can be represented by an inductance Lseg , which includes the
Fig. 8.13 Geometries of

Helical antenna
self-inductance Lself of one turn and the mutual inductance Mseg coupled from two
adjacent turns Lseg ¼ Lself þ 2 Mseg (Fig. 8.13).
The Lself expression integrates the effects of pitch angle (a), l0 is the of free
space.
8R p l0 R4
Lself ¼ l0 R ½lnð Þ 2 cosðaÞ; Mseg ¼ pffiffiffi 3
a 2 ðR2 þ S2 Þ2
pffiffiffi
8R p l0 R4 2
Lseg ¼ l0 R ½lnðð Þ 2 cosðaÞ þ 3
a ðR2 þ S2 Þ2
The normal mode helical antenna (NMHA) equivalent circuit is divided to two
parts: one modeling the equivalent wire antennas (five elements circuit) and the
other modeling the inductive loops (three elements circuit) (Fig. 8.14).
We define two resonance frequencies for our helix antenna equivalent circuit
(eight elements) f01 ðx01 Þ; f02 ðx02 Þ. Capacitor C0 is the antenna capacitance at fre-
quency below the first resonance frequency f01 ðx01 Þ. We select L0 to resonate with
C0 at the second resonant frequency f02 ðx02 Þ. Inductor L1 is chosen to resonate with
C1 at frequency f02 ðx02 Þ. We estimate C1, L1 and R1 at the first resonant frequency
f01 ðx01 Þ, the reactance of the antenna vanishes and the resistance is R0.
Fig. 8.14 Helix antenna

equivalent circuit with eight
frequency independent
elements
p e0 h 1 lnðah0 Þ 1 1 1 x01 1
C0 ¼ ; L0 ¼ ¼ ; A ¼ x01 L0 ¼ ½
lnðah0 Þ 1 x202 C0 x202 p e0 h x01 C0 C0 x202 x01
lnðah0 Þ 1 x01 1 A x01 R20 þ A2 x01 1
A¼ ½ 2 ; C1 ¼ 2 ð Þ; L1 ¼ ð 2 Þ
p e0 h x02 x01 R0 þ A2 x01 x02
2 2 A x02 x01
p ffiffi
ffi
R2 þ A2 Lseg R 8R p l0 R4 2
R1 ¼ 0 ; Lunit ¼ ¼ l0 ½lnð Þ 2 cosðaÞ þ 3
R0 S S a ðR2 þ S2 Þ2
The axial wave number for the helical structure, k is k ¼ vxp ¼ k02 þ c2 , where
vp is the axial velocity of the wave in the helical structure, which is less than the
light velocity, k0 is a free space wave number, and c is the radial wave number. We
define factor F, which integrate k, h parameters [130, 131].
1 sin2 ðk hÞ h h Lseg
F ¼ 1þ ð 1Þ; L2 ¼ Lunit ¼
ð1 p2 Þ 1
2kh sinð2khÞ F F S
2kh
Another factor is H which is frequency dependent. The expression for L3 is

dependent on operational wavelength and function of H. A wavelength of 4 l N is
considered for our helix antenna (k0 ¼ 4lN
2p
).
p2 l2 N 2 k02
ð 2 4c2 l2 N 2
Þ
4ð1 þ c2 Þ def
k 1þ
H ¼ Hðc; k0 ; l; N0 Þ ¼ 0 p2
;
ðp4 l2 N 2 k02 Þ
2
L
2 h=½Lunit p2 c2 2 h=½ Sseg p2 c2
C3 ¼ ¼
H H
The value of H varies significantly at low frequencies and changes little when k0
is sufficiently large near the resonance. A relationship between k0 and c can be
obtaining from the following expression:
c D 2 I0 ð 2 Þ K0 ð 2 Þ
cD cD
2 Lunit l2 N 2 1
ð Þ cD tan2
ðaÞ ¼ 1; L3 ¼ ðH Þ
k0 D I1 ð 2 Þ K1 ð 2 Þ
cD h 1þ
4l2 N 2 c2
def
p2
where I0 ; K0 ; I1 , and K1 are modified Bessel functions.

Remark: Helix antenna is modeled as a series of loops and linear conductors and
each turn can be resolved into two radiating components: first is the axial wire
segment of length S and second it a loop of diameter D. We define two parameters
which related to wavelength, Sk ðkÞ ¼ Sk ; Ck ðkÞ ¼ p2R
k . We define the axial ratio of
ðkÞ
the equivalent wire and loop model AR ¼ ½C2SkðkÞ 2.
k
AR ¼ 2RSk
2 p2 . The typical normal mode helical antennas, the axial ratio is much
greater than one (AR 1 ) 2RSk 2 p2 1 ) S k 2 R p ). The helical antenna

2 2
wire and inductor model is presented in the below figure (Fig. 8.15).
One application of Helix antenna is RFID antennas. We use it for identification
and tracking of objects using radio waves. RFID tags employ helical antennas
embedded in a dielectric material. The antenna is designed to resonate at around
Fig. 8.15 Helical antenna wire and inductor model

specific RFID frequency. RFID IC can be represented as a parallel equivalent

circuit of capacitor and resistor in parallel. The complete RFID tag with helical
antenna equivalent circuit is describe in the below figure [85].
Remark: Helical antenna is connected to RFID IC through microstrip lines. We
neglect in our analysis the parasitic effects of microstrip lines (delay in time)
(Fig. 8.16).
ICrfic þ IRrfic þ IC0 ¼ 0 ðKCL @ A2 Þ; IC0 ¼ IL0 ; IC1 þ IR1 þ IL1 ¼ IL0 ðKCL @ A4 Þ
VA1 VA2 dðVA1 VA2 Þ
IC1 þ IR1 þ IL1 ¼ IC3 þ IL2 ðKCL @ A5 Þ; IL3 ¼ IC3 ; IRrfid ¼ ; ICrfid ¼ Crfid
Rrfid dt
dIL2 dIL3 dðVA6 VA5 Þ
VL2 ¼ VA1 VA5 ¼ L2 ; VL3 ¼ VA1 VA6 ¼ L3 ; IC3 ¼ C3
dt dt dt
V A5 V A4 dIL1 dðVA5 VA4 Þ
I R1 ¼ ; VL1 ¼ VA5 VA4 ¼ L1 ; IC1 ¼ C1
R1 dt dt
dIL0 dðVA3 VA2 Þ
VL0 ¼ VA4 VA3 ¼ L0 ; IC0 ¼ C0
dt dt
Fig. 8.16 Helix antenna equivalent circuit

VA1 VA2 dðVA1 VA2 Þ dIR

IRrfid ¼ ) VA1 VA2 ¼ Rrfid IRrfid ; ICrfid ¼ Crfid ¼ Crfid Rrfid rfid
Rrfid dt dt
VA5 VA4 dIL1 dIR1
IR1 ¼ ) VA5 VA4 ¼ IR1 R1 ; L1 ¼ VA5 VA4 ¼ IR1 R1 ; IC1 ¼ C1 R1
R1 dt dt
dðVA5 VA4 Þ d dIL1 d 2 IL1 d 2 IL1
IC1 ¼ C1 ¼ C1 ðL1 Þ ¼ C1 L1 2 ; IC1 ¼ C1 L1 2
dt dt dt dt dt
dIR1 d 2 IL1 d 2 IL1 R1 dIR1
R1 ¼ L1 2 ) 2 ¼ :
dt dt dt L1 dt
P
KVL : VAkl ¼ VAk VAl ; VAkl ¼ 0; k 6¼ l; k ¼ 1; 2; . . .; 6; l ¼ 1; 2; . . .; 6
VA12 þ VA23 þ VA34 þ VA45 þ VA51 ¼ 0; VA51 ¼ VA56 VA61 ;

Z
1
VA12 ¼ VA1 VA2 ¼ ICrfid dt
Crfid
Z
1 dIL
VA23 ¼ VA2 VA3 ¼ IC0 dt; VA34 ¼ VA3 VA4 ¼ L0 0 ;
C0 dt
Z
1
VA45 ¼ VA4 VA5 ¼ IC1 dt
C1
dIL dIL
VA45 ¼ VA4 VA5 ¼ L1 1 ¼ R1 IR1 ; VA51 ¼ VA5 VA1 ¼ L2 2
dt dt
X Z Z
1 1 dIL
VAkl ¼ 0 ) ICrfid dt IC0 dt L0 0
Crfid C0 dt
Z
1 dIL2
IC1 dt L2 ¼0
C1 dt
d X 1 1 d 2 IL 1 d 2 IL
ð VAkl Þ ¼ 0 ) ICrfid I C0 L 0 2 0 I C1 L 2 2 2 ¼ 0
dt Crfid C0 dt C1 dt
Z
dIL2 1 dIL3
VA51 ¼ VA56 VA61 ) L2 ¼ IC3 dt þ L3 ;
dt C3 dt
d 2 I L2 1 d 2 IL
L2 ¼ I C 3 þ L3 2 3
dt2 C3 dt
System differential equations (Ver.1):
dIR1 d 2 IL dIR
IC1 ¼ C1 R1 ; IC1 ¼ C1 L1 2 1 ; ICrfid ¼ Crfid Rrfid rfid
dt dt dt
2 2 2
d IL 1 d IL d IL R1 dIR1 dIL1 R1
L2 2 2 ¼ IC3 þ L3 2 3 ; 2 1 ¼ ; ¼ IR1
dt C3 dt dt L1 dt dt L1
1 1 d 2 IL 1 d 2 IL
ICrfid IC0 L0 2 0 IC1 L2 2 2 ¼ 0
Crfid C0 dt C1 dt
System differential equations (Ver.2): IC0 ¼ IL0 ; IL3 ¼ IC3 ; IL3 ! IC3
IC1 ¼ IL0 IR1 IL1 ; ICrfid ¼ IRrfid IC0 ¼ IRrfid IL0 ; IC0 ! IL0
dIR1 d 2 IL dIR
IL0 IR1 IL1 ¼ C1 R1 ; IL0 IR1 IL1 ¼ C1 L1 2 1 ; ICrfid ¼ Crfid Rrfid rfid
dt dt dt
d 2 IL2 1 d 2 IC3 dIL1 R1
L2 ¼ IC3 þ L3 2 ; ¼ IR1
dt2 C3 dt dt L1
1 1 d 2 IL 1 d 2 IL
ðIRrfid IL0 Þ IL0 L0 2 0 ðIL0 IR1 IL1 Þ L2 2 2 ¼ 0
Crfid C0 dt C1 dt
dIR1 1 1 1 d 2 IL
¼ I L0 IR 1 I L1 ; 2 1
dt C1 R1 C1 R1 C1 R1 dt
1 1 1
¼ IL IR1 IL
C1 L1 0 C1 L1 C1 L1 1
d 2 IL2 1 d 2 IC dIL R1 dIR 1

L2 ¼ IC3 þ L3 2 3 ; 1 ¼ IR1 ; rfid ¼ ICrfid
dt 2 C3 dt dt L1 dt Crfid Rrfid
1 1 1 1 d 2 IL 1 1 d 2 IL
IRrfid ð þ þ Þ IL0 L0 2 0 þ IR1 þ IL1 L2 2 2 ¼ 0
Crfid Crfid C1 C0 dt C1 C1 dt
d 2 IL1 dIL1 d 2 IL2 dIL2

We define new variables: dY1
dt ¼ dt2 ; Y1 ¼ dt ; dYdt2 ¼ dt2 ; Y2 ¼ dt
dY3 d 2 IC3 dIC3 dY4 d 2 IL0 dIL

¼ ; Y 3 ¼ ; ¼ 2 ; Y4 ¼ 0
dt dt2 dt dt dt dt
dIR1 1 1 1 dY1 1 1 1
¼ IL IR1 IL ; ¼ IL IR1 IL
dt C1 R1 0 C1 R1 C1 R1 1 dt C1 L1 0 C1 L1 C1 L1 1
dIRrfid 1 dIC
¼ ICrfid ; 3 ¼ Y3
dt Crfid Rrfid dt
dY2 1 dY3 dIL1 R1 dIL dIL dIL

L2 ¼ IC3 þ L3 ; ¼ IR1 ; 1 ¼ Y1 ; 2 ¼ Y2 ; 0 ¼ Y4
dt C3 dt dt L1 dt dt dt
1 1 1 1 dY4 1 1 dY2
IRrfid ð þ þ Þ IL0 L0 þ IR1 þ IL L2 ¼0
Crfid Crfid C1 C0 dt C1 C1 1 dt
dIR1 dIRrfid dIC3

At fixed points: dt ¼ 0; dYdt1 ¼ 0; dt ¼ 0; dt ¼ 0; dYdt2 ¼ 0; dYdt3 ¼ 0
dIL1 dIL2 dIL dY4

¼ 0; ¼ 0; 0 ¼ 0; ¼ 0; IL0 ¼ IL1
dt dt dt dt
dIR1 dY1 dIR dIC
¼ 0 ) IC 1 ¼ 0; ¼ 0 ) IC 1 ¼ 0; rfid ¼ 0 ) IC rfid ¼ 0; 3 ¼ 0 ) Y3 ¼ 0
dt dt dt dt
dY2 dY3 dIL1 dIL1
¼ 0; ¼ 0 ) IC3 ¼ 0; ¼ 0 ) IR1 ¼ 0; ¼ 0 ) Y1 ¼ 0
dt dt dt dt
dIL2 dIL
¼ 0 ) Y2 ¼ 0; 0 ¼ 0 ) Y4 ¼ 0; IC rfid ¼ 0 ) IRrfid þ IL0 ¼ 0
dt dt
dY4 dY2 1 1 1 1 1
¼ 0; ¼0) I ð þ þ Þ I þ I ¼ 0; IL0 ¼ 0
dt dt Crfid Rrfid Crfid C1 C0 L0 C1 L1
1 1
IL0 ¼ IL1 ) IRrfid ¼ Crfid ð þ Þ IL0 ; IRrfid þ IL0 ¼ 0
Crfid C0
1 1
) ½1 Crfid ð þ Þ IL0 ¼ 0
Crfid C0
Our circuit fixed point:
EðÞ ðY1 ; Y2 ; Y3 ; Y4 ; IL0 ; IL1 ; IL2 ; IC 1 ; IC 3 ; IR1 ; IC rfid ; IRrfid Þ

¼ ð0; 0; 0; 0; 0; 0; 0; 0; 0; 0; 0; 0Þ
equilibrium points of helix antenna system consists in adding to its coordinated
[Y1 Y2 Y3 Y4 IL0 IL1 IL2 IC1 IC3 IR1 ICrfid IRrfid ] arbitrarily small increments of exponential
terms ½y1 y2 y3 y4 iL0 iL1 iL2 iC1 iC3 iR1 iCrfid iRrfid ekt , and retaining the first order terms in
y1 y2 y3 y4 iL0 iL1 iL2 iC1 iC3 iR1 iCrfid iRrfid . The system of nine homogeneous equations leads
to a polynomial characteristic equation in the eigenvalue k. The polynomial char-
acteristic equation accepts by set the helix antenna system equations. The helix
antenna system fixed values with arbitrarily small increments of exponential form
½y1 y2 y3 y4 iL0 iL1 iL2 iC1 iC3 iR1 iCrfid iRrfid ekt are; i = 0 (first fixed point), i = 1 (second
fixed point), i = 2 (third fixed point), etc., [2–4].
IL0 ðtÞ ¼ IL0 þ iL0 ekt ; IL1 ðtÞ ¼ IL1 þ iL1 ekt ; IC1 ðtÞ ¼ IC1 þ iC1 ekt ; IC3 ðtÞ ¼ IC3 þ iC3 ekt
ðiÞ ðiÞ ðiÞ
IR1 ðtÞ ¼ IR1 þ iR1 ekt ; ICrfid ðtÞ ¼ ICrfid þ iCrfid ekt ; IRrfid ðtÞ ¼ IRrfid þ iRrfid ekt
dY1 ðtÞ dY2 ðtÞ dY3 ðtÞ dY4 ðtÞ

¼ y1 k ekt ; ¼ y2 k ekt ; ¼ y3 k ekt ; ¼ y4 k ekt
dt dt dt dt
dIR1 ðtÞ dIR ðtÞ dIC ðtÞ dIL ðtÞ
¼ iR1 k ekt ; rfid ¼ iRrfid k ekt ; 3 ¼ iC3 k ekt ; 1 ¼ iL1 k ekt
dt dt dt dt
dIL0 ðtÞ ðiÞ kt dIL2 ðtÞ
¼ iL0 k e ; IL2 ðtÞ ¼ IL2 þ iL2 e ;
kt
¼ iL2 k e kt
dt dt
&&&
dIR1 1 1 1
¼ IL IR1 IL
dt C1 R1 0 C1 R1 C1 R1 1
1 ðiÞ 1 ðiÞ 1 ðiÞ
iR1 k ekt ¼ ½I þ iL0 ekt ½I þ iR1 ekt ½I þ iL1 ekt
C1 R1 L0 C1 R1 R1 C1 R1 L 1
1 ðiÞ 1 ðiÞ 1 ðiÞ 1
iR1 k ekt ¼ I I I þ iL ekt
C1 R1 L0 C1 R1 R1 C1 R1 L1 C1 R1 0
1 1
iR ekt iL ekt
C1 R1 1 C1 R1 1
ðiÞ ðiÞ ðiÞ

At fixed points: 1
C1 R1 IL0 C11R1 IR1 C11R1 IL1
1 1 1
iR1 k iR1 þ i L0 iL ¼ 0
C1 R1 C1 R1 C1 R1 1
dY1 1 1 1
¼ IL IR1 IL
dt C1 L1 0 C1 L1 C1 L1 1
1 ðiÞ 1 ðiÞ 1 ðiÞ
y1 k ekt ¼ ½I þ iL0 ekt ½I þ iR1 ekt ½I þ iL1 ekt
C1 L 1 L 0 C1 L1 R1 C1 L1 L1
1 ðiÞ 1 ðiÞ 1 ðiÞ 1
y1 k ekt ¼ I I I þ iL ekt
C1 L1 L0 C1 L1 R1 C1 L1 L1 C1 L1 0
1 1
iR ekt iL ekt
C1 L1 1 C1 L 1 1
ðiÞ ðiÞ ðiÞ

At fixed points: 1
C1 L1 IL0 C11L1 IR1 C11L1 IL1 ¼ 0
1 1 1
y1 k þ i L0 iR1 iL ¼ 0
C1 L1 C1 L1 C1 L1 1
dIRrfid 1 1 ðiÞ
¼ ICrfid ; iRrfid k ekt ¼ ½I þ iCrfid ekt
dt Crfid Rrfid Crfid Rrfid Crfid
1 ðiÞ 1
iRrfid k ekt ¼ I þ iCrfid ekt
Crfid Rrfid Crfid Crfid Rrfid
ðiÞ
At fixed points: 1
Crfid Rrfid ICrfid ¼ 0; iRrfid k þ 1
Crfid Rrfid iCrfid ¼ 0
dIC3 ðiÞ ðiÞ

¼ Y3 ) iC3 k ekt ¼ Y3 þ y3 ekt ; Y3 ¼ 0; iC3 k þ y3 ¼ 0
dt
dY2 1 dY3 1 ðiÞ
L2 ¼ I C 3 þ L3 ; L2 y2 k ekt ¼ ½I þ iC3 ekt þ L3 y3 k ekt
dt C3 dt C3 C3
1 ðiÞ 1 1 ðiÞ
L2 y2 k ekt ¼ I C3 þ iC3 ekt þ L3 y3 k ekt ; IC3 ¼ 0
C3 C3 C3
1 1
L2 y2 k L3 y3 k ¼ iC3 ; L2 L3 ¼ L23 ; ðy2 y3 Þ L23 k ¼ i C3
C3 C3
dIL R1 R1 ðiÞ
y2 y3 ! e; iC3 ! e; 1 ¼ IR1 ; iL1 k ekt ¼ ðIR1 þ iR1 ekt Þ
dt L1 L1
R1 ðiÞ R1 R1 ðiÞ R1
iL1 k ekt ¼ I þ iR ekt ; IR1 ¼ 0 ) iL1 k þ iR ¼ 0
L1 R1 L1 1 L1 L1 1
dIL1 ðiÞ ðiÞ

¼ Y1 ; iL1 k ekt ¼ Y1 þ y1 ekt ; Y1 ¼ 0; iL1 k þ y1 ¼ 0
dt
dIL2 ðiÞ ðiÞ
dt
dIL0 ðiÞ ðiÞ
dt
1 1 1 1 dY4 1 1 dY2
IRrfid ð þ þ Þ IL0 L0 þ IR1 þ IL L2 ¼0
Crfid Crfid C1 C0 dt C1 C1 1 dt
1 ðiÞ 1 1 1 ðiÞ
ðIRrfid þ iRrfid ekt Þ ð þ þ Þ ðIL0 þ iL0 ekt Þ
Crfid Crfid C1 C0
1 ðiÞ 1 ðiÞ
L0 y4 k ekt þ ðI þ iR1 ekt Þ þ ðI þ iL1 ekt Þ L2 y2 k ekt ¼ 0
C1 R1 C1 L 1
1 1 1 ðiÞ 1 ðiÞ 1 ðiÞ 1 ðiÞ 1
ð þ þ ÞI þ I þ I I iR ekt
Crfid C1 C0 L0 C1 R1 C1 L1 Crfid Rrfid Crfid rfid
1 1 1 1
ð þ þ Þ iL0 ekt L0 y4 k ekt þ iR ekt
Crfid C1 C0 C1 1
1
þ iL ekt L2 y2 k ekt ¼ 0
C1 1

At fixed points: ðC1rfid þ 1
C1 þ C0 Þ
1
I L0 þ 1
C1 IR 1 þ 1
C1 IL1 C1rfid IRrfid ¼ 0
1 1 1 1 1 1
ðL2 y2 þ L0 y4 Þ k iR ð þ þ Þ i L0 þ iR þ iL
Crfid rfid Crfid C1 C0 C1 1 C1 1
¼0
L2 L0 ¼ L20
1 1 1 1 1 1
) ðy2 þ y4 Þ L20 k iRrfid ð þ þ Þ iL0 þ iR1 þ
Crfid Crfid C1 C0 C1 C1
i L1
¼0
1 1 1 1 1
y2 y4 ) y2 þ y4 2 y2 ) y2 k iR ð þ þ Þ iL0
2 L20 Crfid rfid 2 L20 Crfid C1 C0
1 1
þ iR þ iL ¼ 0
2 L20 C1 1 2 L20 C1 1
1 1 1 1 1
y2 y4 ) y2 þ y4 2 y4 ) y4 k iR ð þ þ Þ iL0
2 Crfid L20 rfid 2 L20 Crfid C1 C0
1 1
þ iR þ iL ¼ 0
2 C1 L20 1 2 C1 L20 1
We can summary our helix antenna system arbitrarily small increments

equations:
1 1 1 1 1 1
iR1 k iR þ iL iL ¼ 0; y1 k þ iL iR iL ¼ 0
C1 R1 1 C1 R1 0 C1 R1 1 C1 L1 0 C1 L1 1 C1 L1 1
1 R1
iRrfid kþ iCrfid ¼ 0; iC3 k þ y3 ¼ 0; iL1 k þ iR ¼ 0
Crfid Rrfid L1 1
iL1 k þ y1 ¼ 0; iL2 k þ y2 ¼ 0; iL0 k þ y4 ¼ 0
1 1 1 1 1 1 1
y2 k iR ð þ þ Þ iL0 þ iR þ iL ¼ 0
2 L20 Crfid rfid 2 L20 Crfid C1 C0 2 L20 C1 1 2 L20 C1 1
1 1 1 1 1 1 1
y4 k iR ð þ þ Þ iL0 þ iR þ iL ¼ 0
2 Crfid L20 rfid 2 L20 Crfid C1 C0 2 C1 L20 1 2 C1 L20 1
0 1
iR1
B y C
B 1C
B C
B iRrfid C 0 1
0 1 B B C
C i11 i12 !
N11 . . . N19 B i C3 C B : : C
B .. .. .. C B B
C B
C C iCrfid
@ . . . A B i L1 C þ B C ¼0
B C @ : : A y3
N91 N99 B i L2 C
B C i91 i92
B iL C
B 0C
B C
@ y2 A
y4
!
iCrfid
Assumption: !e
y3
1 1 1
N11 ¼ k ; N12 ¼ N13 ¼ N14 ¼ 0; N15 ¼ ; N16 ¼ 0; N17 ¼ ; N18 ¼ N19 ¼ 0
C 1 R1 C 1 R1 C 1 R1
1 1 1
N21 ¼ ; N22 ¼ k; N23 ¼ N24 ¼ 0; N25 ¼ ; N26 ¼ 0; N27 ¼ ; N28 ¼ N29 ¼ 0
C 1 R1 C1 L1 C1 L1
N31 ¼ N32 ¼ 0; N33 ¼ k; N34 ¼ . . . ¼ N39 ¼ 0; N41 ¼ N42 ¼ N43 ¼ 0; N44 ¼ k
R1
N45 ¼ . . . ¼ N49 ¼ 0; N51 ¼ ; N52 ¼ N53 ¼ N54 ¼ 0; N55 ¼ k; N56 ¼ . . . ¼ N59 ¼ 0
L1
N61 ¼ . . . ¼ N65 ¼ 0; N66 ¼ k; N67 ¼ 0; N68 ¼ 1; N69 ¼ 0; N71 ¼ . . . ¼ N76 ¼ 0; N77 ¼ k
1 1 1
N78 ¼ 0; N79 ¼ 1; N81 ¼ ; N82 ¼ 0; N83 ¼ ; N84 ¼ 0; N85 ¼
2 L20 C1 2 L20 Crfid 2 L20 C1
1 1 1 1 1
N86 ¼ 0; N87 ¼ ð þ þ Þ; N88 ¼ k; N89 ¼ 0; N91 ¼
2 L20 Crfid C1 C0 2 L20 C1
1 1
N92 ¼ 0; N93 ¼ ; N94 ¼ 0; N95 ¼ ; N96 ¼ 0
2 Crfid L20 2 C1 L20
1 1 1 1
N97 ¼ ð þ þ Þ; N98 ¼ 0; N99 ¼ k
2 L20 Crfid C1 C0
1
i11 ¼ i12 ¼ i21 ¼ i22 ¼ 0; i31 ¼ ; i32 ¼ 0; i41 ¼ 0; i42 ¼ 1;
Crfid Rrfid
i51 ¼ i52 ¼ 0; i61 ¼ i62 ¼ 0
i71 ¼ i72 ¼ 0; i81 ¼ i82 ¼ 0; i91 ¼ i92 ¼ 0
0 1 0 1
N11 . . . N19 N11 ... N19
B .. .. .. C B .. .. .. C
AkI ¼@ . . . A; detðA k IÞ ¼ 0; det@ . . . A¼0
N91 N99 N91 N99
Eigenvalues stability discussion: Our helix antenna system involving Na vari-

ables (Na > 2, Na = 9), the characteristic equation is of degree Na = 9
P
( 9k¼0 kk Pk ¼ 0) and must often be solved numerically. Expect in some partic-
ular cases, such an equation has (Na = 9) distinct roots that can be real or complex.
These values are the eigenvalues of the 9 9 Jacobian matrix (A). The general rule
is that the Steady State (SS) is stable if there is no eigenvalue with positive real
part. It is sufficient that one eigenvalue is positive for the steady state to be unstable.
Our 9-variables (Y1 Y2 Y4 IL0 IL1 IL2 IC3 IR1 IRrfid ) helix antenna system has nine eigen-
values. The type of behavior can be characterized as a function of the position of
these eigenvalues in the Re/Im plane. Five non-degenerated cases can be distin-
guished: (1) the nine eigenvalues are real and negative (stable steady state), (2) the
nine eigenvalues are real, eight of them are negative (unstable steady state), (3) and
eigenvalues are real and negative (stable steady state), two cases can be
distinguished depending on the relative value of the real part of the complex
eigenvalues and of the real one, (5) two eigenvalues are complex conjugates with a
negative real part and at least one eigenvalue is positive (unstable steady state).
Exercises
1. We have a system of three N-turn multilayer circular antennas which are

integrated with RFID IC for complete RFID transponder. The first (i = 1) and
second (i = 2) N-turn multilayer circular antennas are connected in parallel
through microstrip lines 1 and 2. They are connected to switch S1 which can be
connected to third N turn multilayer circular antenna (S1 in position a) or to
capacitor Cx and microstrip line 3 (S1 in position b). An N-turn multilayer
circular coil antennas system is influenced by electromagnetic interferences,
which affect their stability behavior but we neglect it in our analysis.
Additionally, microstrip lines in the system have a parasitic effects, a delay
lines Dl1 ; Dl2 and Dl3 respectively. We need to inspect our system performance
under microstrip lines (1, 2, and 3) parasitic affects.
Every N-turn multilayer circular coil antenna has a parasitic DC resistance

which need to be calculated. Index (i) indicated the first N-turn multilayer
circular coil antenna (i = 1) or second N-turn multilayer circular coil antenna
(i = 2) or third N-turn multilayer circular coil antenna (i = 3). We define RFID’s
N-turn multilayer coil antenna parameters, ai —Average radius of the coil in

cm, Ni —number of turns, bi —winding thickness in cm, Si —wire cross section
area, mi —radius of the wire and hi —winding height in cm. Integrating all those
parameters gives the equation for N-turn multilayer circular coil antenna
2
i Ni Þ
inductance calculation (Lcalci ¼ 6a0:31ða i þ 9hi þ 10bi
½lH ). The length of N turn is
lNi ¼ 2 p ai Ni (Assumption: ai bi ; ai þ bi ai ). We consider system
three N-turn multilayer circuit coil antennas and not identical.
ak 6¼ al ; Nk 6¼ Nl ; hk 6¼ hl ; bk 6¼ bl ; k 6¼ l; k ¼ 1; 2; 3; l ¼ 1; 2; 3
ak ; Nk ; hk ; bk 2 R þ ; al ; Nl ; hl ; bl 2 R þ
The DC resistance of the N-turn multilayer circular coil antenna:
i Ni 2ai Ni
l
RDCi ¼ riNSi i ¼ 2pa
ri Si ¼ ri m2 . lNi —Total length of the wire, ri —Conductivity
i
of the wire (S/m), Si —wire cross section area (p m2i ), mi —Radius of the wire.
1:1 Write system differential equations for cases: S1 in position (a) and S1 in
position (b).
1:2 Find system fixed points for the cases: S1 in position (a) and S1 in position
(b).
1:3 Discuss stability and stability switching for the simple case Dl1 ¼ 0, Dl2 ¼
0; Dl3 ¼ 0 Under variation of circuit parameters (switch S1 in position (a).
1:4 Discuss stability and stability switching for the cases: Dl 2 R þ
(1) Dl1 ¼ Dl ; Dl2 ¼ 0; Dl3 ¼ 0 (2) Dl1 ¼ 0; Dl2 ¼ Dl ; Dl3 ¼ Dl .
(3) Dl1 ¼ Dl3 ¼ Dl ; Dl2 ¼ 0 (4) Dl1 ¼ Dl ; Dl2 ¼ C Dl1 ; Dl3 ¼ 0;
C 2 Rþ
Under variation of Dl and C parameters (switch S1 in position (a)).
1:5 Return (1.3) and (1.4) for the case, switch S1 in position (b).
Remark: The delay is on the current that flow through microstrip line
IðtÞ ! Iðt Dl Þ, we consider that VDl ! e(neglect the voltage on
microstrip line). Take care in your analysis and calculation the mutual
inductances
P between each two N-turn multilayer circular antennas in our
system ( M ¼ M þ M ).
2. We have a system of three N-turn multilayer circular antennas in series, which
are integrated with RFID IC for complete RFID transponder system.
Additionally, there is a capacitors bridge (CA1, CA2, CA3, and CA4) which is
connected to our RFID transponder system and balance the circuit currents
flow. An N-turn multilayer circular coil antennas system is influenced by
electromagnetic interferences which affect their stability behavior but we
neglect it in our analysis. Additionally, microstrip lines in the system have a
parasitic effects, a delay lines Dl1 ; Dl2 and Dl3 respectively. We need to inspect
our system performance under microstrip lines (1, 2, and 3) parasitic affects.
Every N-turn multilayer circular coil antenna has a parasitic DC resistance
which needs to be calculated. Index (i) indicated the first N-turn multilayer
circular coil antenna (i = 1) or second N-turn multilayer circular coil antenna
Exercises 887
(i = 2) or third N-turn multilayer circular coil antenna (i = 3). We define RFID’s

N-turn multilayer coil antenna parameters, ai —Average radius of the coil in
cm, Ni —number of turns, bi —Winding thickness in cm, Si —wire cross section
area, mi —radius of the wire and hi —winding height in cm. Integrating all those
parameters give the equation for N-turn multilayer circular coil antenna
2
i Ni Þ
inductance calculation (Lcalci ¼ 6a0:31ða
i þ 9hi þ 10bi
lNi ¼ 2 p ai Ni (Assumption: ai bi ; ai þ bi ai ). We consider System
three N-turn multilayer circuit coil antennas and not identical
ak 6¼ al ; Nk 6¼ Nl ; hk 6¼ hl ; bk 6¼ bl ; k 6¼ l; k ¼ 1; 2; 3; l ¼ 1; 2; 3
ak ; Nk ; hk ; bk 2 R þ ; al ; Nl ; hl ; bl 2 R þ
RDCi ¼ ¼ ¼ :
ri Si ri Si ri m2i
lNi —Total length of the wire, ri —Conductivity of the wire (S/m), Si —wire
cross section area (p m2i ), mi —Radius of the wire.

2:2 Discuss system stability and stability switching for the simple case under
circuit parameters variation (Dl1 ¼ 0; Dl2 ¼ 0; Dl3 ¼ 0).
2:3 Discuss stability and stability switching for the cases: Dl 2 R þ

(1) Dl1 ¼ Dl ; Dl2 ¼ 0; Dl3 ¼ 0 (2) Dl1 ¼ 0; Dl2 ¼ Dl ; Dl3 ¼ Dl .
p ffiffiffiffiffiffi
(3) Dl1 ¼ Dl3 ¼ Dl ; Dl2 ¼ 0 (4) Dl1 ¼ Dl ; Dl2 ¼ ð1 þ C2 Þ Dl1 ;
3
Dl3 ¼ 0; C 2 R þ
Under variation of Dl and C parameters.
2:4 Bridge’s capacitor CA3 is disconnected. How it influences circuit behavior?
Find system differential equations, fixed points and discuss stability and
stability switching under parameters variations.
2:5 Bridge’s capacitor CA2 is shortened. How it influences circuit behavior?
Find system differential equations, fixed points and discuss stability and
stability switching under parameters variations.
inductances
P between each two N-turn multilayer circular antennas in our
system ( M ¼ M þ M ).
3. We have a system of two N-turn multilayer circular antennas which can be
integrated with RFID IC for complete RFID transponder system. The first (i = 1)
and second (i = 2) N-turn multilayer circular antennas are connected by two
microstrip lines (1 and 2) and matching network. The matching network can be
Pi-type or T-type. Switch S1 has two positions: first position (a), RFID IC is
connected to first (i = 1) N-turn multilayer circular antenna through matching
network and microstrip line 1. The second (i = 2) N-turn multilayer circular
antenna is disconnected. Second position (b), RFID IC is connected to two N-turn
multilayer circular antennas. An N-turn multilayer circular coil antennas system is
influenced by electromagnetic interferences which affect their stability behavior
but we neglect it in our analysis. Additionally, microstrip lines in the system have
parasitic effects, a delay lines Dl1 ; Dl2 respectively. We need to inspect our system
performance under microstrip lines (1 and 2) parasitic affects. Every N-turn
multilayer circular coil antenna has a parasitic DC resistance which needs to be
calculated. Index (i) indicated the first N-turn multilayer circular coil antenna (i =
1) or second N-turn multilayer circular coil antenna (i = 2). We define RFID’s
N-turn multilayer coil antenna parameters, ai —Average radius of the coil in cm,
Ni —number of turns, bi —Winding thickness in cm, Si —wire cross section area,
mi —radius of the wire and hi —winding height in cm. Integrating all those
parameters give the equation for N-turn multilayer circular coil antenna induc-
2
i Ni Þ
tance calculation Lcalci (Lcalci ¼ 6a0:31ða
i þ 9hi þ 10bi
lNi ¼ 2 p ai Ni (Assumption: ai bi ; ai þ bi ai ). Two N-turn multilayer
circuit coil antennas are not identical.
a1 6¼ a2 ; N1 6¼ N2 ; h1 6¼ h2 ; b1 6¼ b2 ; i ¼ 1; 2; ai ; Ni ; hi ; bi 2 R þ
Exercises 889
RDCi ¼ ¼ ¼ :
ri Si ri Si ri m2i
lNi —Total length of the wire,

ri —Conductivity of the wire (S/m), Si —wire cross section area (p m2i ),
mi —Radius of the wire.
3:1 Write system differential equations for the cases: S1 in position (a) and S1 in
position (b). Matching network is T-type.
3:2 Find system fixed points for the cases: S1 in position (a) and S1 in position
(b). Matching network is T-type.
3:3 Discuss stability and stability switching for the simple case Dl1 ¼ 0; Dl2 ¼
0 under variation of circuit parameters. Switch S1 is in (a) position.
Matching network is T-type.
3:4 Return (3.1), (3.2) and (3.3) for the case we use Pi-type matching network.
3:5 How the circuit dynamic is changed if switch S1 is in (b) position? Return
(3.3) for the case that switch S1 is in position (b).
3:6 Return (3.3) for the case Dl1 ¼ Dl ; Dl2 ¼ C Dl ; C; Dl 2 R þ . Discuss
stability and stability switching for variation of C; Dl parameters.
inductances
P between two N-turn multilayer circular antennas in our system
( M ¼ M þ M ).
4. We have a system of double rectangular spiral antennas which are connected in
parallel through microstrip lines and discrete components. The antennas are
integrated with RFID IC to complete RFID transponder system. A double rect-

angular spiral antennas system influences by electromagnetic interferences which
effect there stability behavior. Additionally, microstrip lines which are connect
antennas in parallel and have parasitic effects, delays in time Dl1 ; Dl2 respec-
tively. We inspect our system performances under electromagnetic interferences
and microstrip lines parasitic effects. Every rectangular spiral antenna has a
parasitic DC resistance which needs to be calculated. Index (i) indicates first
rectangular spiral antenna (i = 1) or second rectangular spiral antenna (i = 2).
We define RFID’s coil dimensional parameters. A0i, B0i—Overal dimensions of
the coil. Aavgi, Bavgi—Average dimensions of the coil.
ti—track thickness. wi—track width. gi—gaps between tracks. Nci—number of
turns. di—equivalent diameter of the track. Average coil area Aci = Aavgi ∙
Bavgi. Pi—coil manufacturing technology parameter.
Integrating all those parameters gives the equations for rectangular spiral
antenna inductance calculation:
l0 X
4
Lcalci ¼ ð Xki X3 Þ NciPi ;
p k¼1;k6¼3
2 Aavgi Bavgi
X1i ¼ Aavgi lnð qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi Þ
di ðAavgi þ A2avgi þ B2avgi Þ
2 Aavgi Bavgi
X2i ¼ Bavgi lnð qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi Þ;

X4i ¼ ; di ¼ ; Aavgi ¼ A0i Nci ðgi þ wi Þ
4 p
the first turn l0i ¼ 2 ðA0i þ B0i Þ ðwi þ gi Þ. lk is the length of turn k + 1.
We define the following:
NX
ci 1

k¼1
NX
C 1 NX
C 1

k¼1 k¼1

Exercises 891
lTi lTi
RDCi ¼ ¼ :
ri Si ri p a2i


RDCi ¼
ri p w2i

ðD1 6¼ D2 ; s1 0Þ;ðs2 0; D1 ; D2 0Þ.
4:1 Write system differential equations and find fixed points for the following
cases: s; D; Dl 2 R þ
(1) s1 ¼ s; s2 ¼ 0; D1 ¼ D; D2 ¼ 0; Dl1 ¼ Dl2 ¼ 0.
(2) s1 ¼ s2 ¼ s; D1 ¼ D2 ¼ 0; Dl1 ¼ Dl2 ¼ Dl .
(3) s1 ¼ s2 ¼ 0; D1 ¼ D2 ¼ D; Dl1 ¼ Dl ; Dl2 ¼ 0.
(4) s1 ¼ s2 ¼ s; D1 ¼ D2 ¼ D; Dl1 ¼ 0; Dl2 ¼ Dl .
4:2 Discuss system stability and stability switching under variation of param-
eters s; D; Dl for all cases in (4.1).
4:3 We short inductor La, How it influences our system dynamics and stability.
Discuss stability and stability switching for different values of Ca1 and Ca2.
4:4 We short capacitor Ca1, How it influences our system dynamic and sta-
bility? Discuss stability and stability switching for different values of La.
bility? Discuss stability and stability switching for different values of Ca1.
IðtÞ ! Iðt Dl Þ, we consider that VDl ! e (neglect the voltage on
inductances
P between double rectangular spiral antennas in our system
( M ¼ M þ M ).
5. We have a system of double rectangular spiral antennas which are connected
through switch S1 to RFID IC. The antennas are integrated with RFID IC and
discrete components (Ca1 ; Ca2 ; La ; Lb ) to complete RFID transponder system.
A double rectangular spiral antennas system influences by electromagnetic
interferences which effect there stability behavior. Additionally, microstrip lines
which are connect antennas to RFID IC and have parasitic effects, delays in
time Dl1 ; Dl2 respectively. We inspect our system performances under elec-
tromagnetic interferences and microstrip lines parasitic effects. Every rectan-
gular spiral antenna has a parasitic DC resistance which needs to be calculated.
Index (i) indicates first rectangular spiral antenna (i = 1) or second rectangular
spiral antenna (i = 2).
Exercises 893
We define RFID’s coil dimensional parameters. A0i, B0i—Overal dimensions of

the coil. Aavgi, Bavgi—Average dimensions of the coil.
ti—track thickness. wi—track width. gi—gaps between tracks. Nci—number of
turns. di—equivalent diameter of the track. Average coil area Aci = Aavgi ∙
Bavgi. Pi—coil manufacturing technology parameter.
Integrating all those parameters gives the equations for rectangular Spiral
antenna inductance calculation:
l0 X
4
Lcalci ¼ ð Xki X3 Þ NciPi ;
p k¼1;k6¼3
2 Aavgi Bavgi
X2i ¼ Bavgi lnð qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi Þ;

X4i ¼ ; di ¼ ; Aavgi ¼ A0i Nci ðgi þ wi Þ
4 p
the first turn l0i ¼ 2 ðA0i þ B0i Þ ðwi þ gi Þ. lk is the length of turn k + 1.
We define the following:
NX
ci 1

k¼1
NX
C 1 NX
C 1

k¼1 k¼1
lTi lTi
RDCi ¼ ¼ :
ri Si ri p a2i


RDCi ¼
ri p w2i

ðD1 6¼ D2 ; s1 0Þ;ðs2 0; D1 ; D2 0Þ.
5:1 Write system differential equations and find fixed points for the following
cases: S1 in position (a), S1 in position (b) s; D; Dl 2 R þ and for below
subcases.
(1) s1 ¼ s; s2 ¼ 0; D1 ¼ D; D2 ¼ 0; Dl1 ¼ Dl2 ¼ 0.
(2) s1 ¼ s2 ¼ s; D1 ¼ D2 ¼ 0; Dl1 ¼ Dl2 ¼ Dl .
(3) s1 ¼ s2 ¼ 0; D1 ¼ D2 ¼ D; Dl1 ¼ Dl ; Dl2 ¼ 0.
(4) s1 ¼ s2 ¼ s; D1 ¼ D2 ¼ D; Dl1 ¼ 0; Dl2 ¼ Dl .
5:2 Discuss system stability and stability switching under variation of param-
eters s; D; Dl for all cases in (5.1).
5:3 We short inductor La, How it influences our system dynamics and stability.
Discuss stability and stability switching for different values of Ca1 and Ca2.
bility? Discuss stability and stability switching for different values of La.
5:5 We short inductor Lb, How it influences our system dynamic and stability?
Discuss stability and stability switching for different values of Ca1.
IðtÞ ! Iðt Dl Þ, we consider that VDl ! e (neglect the voltage on
inductances
P between double rectangular spiral antennas in our system
( M ¼ M þ M ).
6. We have system of almost two turn square planar straight thin film inductors
antenna (seven segments). The system is constructed from seven straight thin
film inductors which are connected in almost two turn square structure. The
straight thin film inductors are connected by microstrip lines (A, B, C… F). The
almost two turns square planar straight thin film inductors antenna system is
connected to transceiver module (represent as a transceiver mixer output
equivalent circuit) through two microstrip lines (G and H). Index (i) stands for
straight thin film inductor in place (i). wi is the width of straight thin film
inductor (i) in cm, zi is the thickness of straight thin film inductor (i) in cm, and
li is the length of straight thin film conductor (inductor) in cm. The calculated
inductance of straight thin film inductor (i) is as follow (Li is the segment
inductance in lH):
Exercises 895
2 li wi þ zi
Li ¼ 0:002 li fln½ þ 0:50049 þ g½lH ; i ¼ 1; 2; 3; 4; . . .; 7
w i þ zi 3 li
Remark: we assume that the magnetic permeability of the conductor material is

1 and the seven straight thin film inductors are not identical
wi 6¼ wj ; li 6¼ lj zi 6¼ zj ðj 6¼ i; j ¼ 1; 2; 3; 4; . . .; 7; i ¼ 1; 2; 3; 4; . . .; 7Þ:
Consider in your analysis the DC resistivity qi of the straight thin film strip
material and the strip cross section Ai ¼ wi zi ; Ri ¼ wqii lzii . The dimension of the
straight thin film strip that affects the inductance most strongly is the length li .
The width wi has much weaker influence, and straight thin film strip thickness
can be neglected completely for the limit (zi wi ). This is in contrast to the
resistance Ri ¼ wqii lzii , which is inversely proportional to the straight thin film strip
cross section wi zi and depends on the material properties via its resistivity qi .
The length of each straight thin film strip in our system is not the same to each
other. Microstrip line in our system is represented as a delay line and the delay
is on the current that flows through the microstrip line (s1 ; s2 ; . . .) respectively
(Vsi ! e; i ¼ 1; 2; . . .; A ! 1; B ! 2; . . .).
6:1 Find the expression of system total inductance (LT) which constructed from
the sum of the self-inductances
P of all straight segments and the sum of all
mutual inductances ( M), both negative and positive).
6:2 We short straight segment number 5, find the expression of system total
inductance (LT) which constructed from the sum of the self-inductances
P of
all straight segments and the sum of all mutual inductances ( M), both
negative and positive).
6:3 Find system differential equations and fixed points for (6.1) and (6.2).
6:4 Discuss stability and stability switching under variation of system param-
eters (simple case: no delays s1 ¼ s2 ¼ ¼ 0).
6:5 Discuss stability and stability switching under variation of delay parameter
s (s1 ¼ s2 ¼ ¼ s).
7. We have system of almost two turn square planar straight thin film inductors
antenna (six segments). The system is constructed from six straight thin film
inductors which are connected in almost two turn square structure. The straight
thin film inductors are connected by microstrip lines (A, B, C… E). The almost
two turns square planar straight thin film inductors antenna system is connected
to transceiver module (represent as a transceiver mixer output equivalent circuit)
through capacitor bridge (Ca ; Cb ; Cc ; Cd ) and two microstrip lines (F and G).
Index (i) stands for straight thin film inductor in place (i). wi is the width of
straight thin film inductor (i) in cm, zi is the thickness of straight thin film
inductor (i) in cm, and li is the length of straight thin film conductor (inductor) in
cm. The calculated inductance of straight thin film inductor (i) is as follow (Li is
the segment inductance in lH):
2 li wi þ zi
Li ¼ 0:002 li fln½ þ 0:50049 þ g ½lH ; i ¼ 1; 2; 3; 4; . . .; 7
wi þ zi 3 li
Remark: we assume that the magnetic permeability of the conductor material is

1 and the six straight thin film inductors are not identical
wi 6¼ wj ; li 6¼ lj zi 6¼ zj ðj 6¼ i; j ¼ 1; 2; 3; 4; . . .; 6; i ¼ 1; 2; 3; 4; . . .; 6Þ:
Consider in your analysis the DC resistivity qi of the straight thin film strip
material and the strip cross section Ai ¼ wi zi ; Ri ¼ wqii lzii . The dimension of the
straight thin film strip that affects the inductance most strongly is the length li . The
width wi has much weaker influence, and straight thin film strip thickness can be
neglected completely for the limit (zi wi ). This is in contrast to the resistance
Exercises 897
Ri ¼ wqii lzii , which is inversely proportional to the straight thin film strip cross
section wi zi and depends on the material properties via its resistivity qi . The
length of each straight thin film strip in our system is not the same to each other.
Microstrip line in our system is represented as a delay line and the delay is on
the current that flows through the microstrip line (s1 ; s2 ; . . .) respectively
(Vsi ! e; i ¼ 1; 2; . . .; A ! 1; B ! 2; . . .).
7:1 Find the expression of system total inductance (LT) which constructed from
the sum of the self-inductances
P of all straight segments and the sum of all
mutual inductances ( M), both negative and positive).
7:2 We short straight segment number 5, find the expression of system total
inductance (LT) which constructed from the sum of the self-inductances
P of
all straight segments and the sum of all mutual inductances ( M), both
negative and positive).
7:3 Find system differential equations and fixed points for (7.1) and (7.2).
7:4 Discuss stability and stability switching under variation of system param-
eters (simple case: no delays s1 ¼ s2 ¼ ¼ 0).
7:5 Discuss stability and stability switching under variation of delay parameter
s (s1 ¼ s2 ¼ ¼ s).
8. We have a system of two helical antennas which are connected to RFID IC
through capacitors network (Ca ; Ca1 ; Ca2 ). The dimensional parameters of
helical antennas are hi ; Si ; ai ; Ri ; Di ; i ¼ 1; 2 (hi ; Si ; ai ; Ri ; Di 2 R þ ). We define
the mathematical relationships between helical antennas parameters as follow:
R2 ¼ C1 R1 ; S2 ¼ S1 C2 þ S1 C22
pffiffiffiffiffi pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
h2 ¼ C3 h1 ; aa12 ¼ C1 þ C2 ; C1 ; C2 ; C3 2 R þ . The two helical antennas are
not identical. The length of helix antenna on turns, li where
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
li ¼ ðp Di Þ2 þ S2i ; i ¼ 1; 2. The induced magnetic field of a single turn can
be represented by an inductance LsegðiÞ , which includes the self-inductance
Lself ðiÞ of one turn and the mutual inductance MsegðiÞ coupled from two adjacent
turns LsegðiÞ ¼ Lself ðiÞ þ 2 MsegðiÞ . The Lself ðiÞ expression integrates the effects of
pitch angel (ai ), l0 is the permeability of free space.
Exercises 899
8Ri p l0 R4i
Lself ðiÞ ¼ l0 Ri ½lnð Þ 2 cosðai Þ; MsegðiÞ ¼ pffiffiffi 3 ; i ¼ 1; 2
ai 2 ðR2i þ S2i Þ2
pffiffiffi
8Ri p l0 R4i 2
LsegðiÞ ¼ l0 Ri ½lnð Þ 2 cosðai Þ þ 3 ; i ¼ 1; 2
ai ðR2 þ S2 Þ2
i i
parts: one modeling the equivalent wire antenna (five elements circuit) and the
other modeling the inductive loops (three elements circuit). In your system
analysis use the helical antenna equivalent circuit with eight frequency inde-
pendent elements.
Lself ð1Þ
8:1 Find the ratio Lself ð2Þ as a function of parameters C1, C2 and C3. Draw 3D
L ð1Þ
graph, Z-axis (Lself
self ð2Þ
), Y-axis (C1) and X-axis (C2) for the case C2 = C3.
Lself ð2Þ
8:2 Find the ratio Lself ð1Þ as a function of parameters C1, C2 and C3. Draw 3D
L ð2Þ
graph, Z-axis (Lself
self ð1Þ
), Y-axis (C2) and X-axis (C3) for the case C1 = C2.
8:4 Discuss stability and stability switching under variation of parameters C1,
C2 and C3 (C1 ; C2 ; C3 2 R þ ).
8:5 We have a simple case C ¼ C1 ¼ C2 ¼ C3 , discuss stability and stability
switching under variation of C parameter C 2 R þ .
Remark: Take care in your analysis and calculation
P the mutual inductances
between helical antennas in our system ( M ¼ M þ M ).
9. We have a system of two helical antennas which are connected to RFID IC
through capacitors and inductor network (La ; Ca1 ; Ca2 ) and selective switch S1
(positions a, b). The dimensional parameters of helical antennas are
hi ; Si ; ai ; Ri ; Di ; i ¼ 1; 2 (hi ; Si ; ai ; Ri ; Di 2 R þ ). We define the mathematical
pffiffiffiffiffiffi
relationships between helical antennas parameters as follow: R2 ¼ C1
pffiffiffiffiffiffi pffiffiffiffiffi
R1 ; S2 ¼ S1 C2 þ S1 C22 ; h2 ¼ C3 h1
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
p ffiffiffiffiffi
ffi
a2 ¼ C1 þ C2 ; C1 ; C2 ; C3 2 R þ . The two helical antennas are not identical.
a1
The length of helix antenna on turns, li where li ¼ ðp Di Þ2 þ S2i ; i ¼ 1; 2.
The induced magnetic field of a single turn can be represented by an inductance
LsegðiÞ , which includes the self-inductance Lself ðiÞ of one turn and the mutual
inductance MsegðiÞ coupled from two adjacent turns LsegðiÞ ¼ Lself ðiÞ þ 2 MsegðiÞ .
The Lself ðiÞ expression integrates the effects of pitch angel (ai ), l0 is the per-
meability of free space.
8Ri p l0 R4i
pffiffiffi
8Ri p l0 R4i 2
ai ðR2 þ S2 Þ2
i i
pendent elements.
9:1 Write system differential equations and find fixed points, S1 in a position.
9:2 Write system differential equations and find fixed points, S1 in b position.
C2 and C3 (C1 ; C2 ; C3 2 R þ ).
9:4 We disconnected capacitor Ca1, how it influences our system behavior?
Discuss stability and stability switching under variation of C parameter
pffiffiffiffi
(C1 ¼ C; C2 ¼ C2 ; C3 ¼ C 2 R þ ).
9:5 We short capacitor Ca2, how it influences our system behavior? Discuss
stability and stability switching for different values of C parameter
Exercises 901
p ffiffiffiffiffiffi pffiffiffiffi
(C1 ¼ C þ 1; C2 ¼ C2 ; C3 ¼ C 2 R þ ).
3
Remark: Take care in your analysis and calculation

P the mutual inductances
between helical antennas in our system ( M ¼ M þ M ).
10. We have a system of two helical antennas which are connected to two RFID
ICs through capacitors and inductor network (La ; Ca1 ; Ca2 ; Ca3 ) and selective
switch S1 (positions a, b). The dimensional parameters of helical antennas are
hi ; Si ; ai ; Ri ; Di ; i ¼ 1; 2 (hi ; Si ; ai ; Ri ; Di 2 R þ ). We define the mathematical
ffiffiffiffiffiffiffiffiffi
pp ffiffiffiffiffiffi
relationships between helical antennas parameters as follow: R2 ¼ 3 C1
pffiffiffiffiffiffi p ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
pffiffiffiffiffiffi
R1 ; S2 ¼ S1 C2 þ S1 C22 ; h2 ¼ C3 h1 aa12 ¼ C3 þ C2 ; C1 ; C2 ; C3 2 R þ .
The two helical antennas are not identical. The length of helix antenna on turns,
li where li ¼ ðp Di Þ2 þ S2i ; i ¼ 1; 2. The induced magnetic field of a single
turn can be represented by an inductance LsegðiÞ , which includes the
self-inductance Lself ðiÞ of one turn and the mutual inductance MsegðiÞ coupled
from two adjacent turns LsegðiÞ ¼ Lself ðiÞ þ 2 MsegðiÞ . The Lself ðiÞ expression
integrates the effects of pitch angel (ai ), l0 is the permeability of free space.
8Ri p l0 R4i
pffiffiffi
8Ri p l0 R4i 2
ai ðR2 þ S2 Þ2
i i
pendent elements.
10:1 Write system differential equations and find fixed points, S1 in a position.
10:2 Write system differential equations and find fixed points, S1 in b position.
C2 and C3 (C1 ; C2 ; C3 2 R þ ).
10:4 We disconnected capacitor Ca3, how it influences our system behavior?
Discuss stability and stability switching under variation of C parameter
p ffiffiffiffiffiffi pffiffiffiffi
(C1 ¼ C2 ; C2 ¼ C2 ; C3 ¼ C 2 R þ ).
3
10:5 We short capacitor Ca2, how it influences our system behavior? Discuss
stability and stability switching for different values of C parameter
pffiffiffiffiffiffiffiffiffiffiffi p ffiffiffiffiffiffi pffiffiffiffi
(C1 ¼ C þ 1; C2 ¼ C2 ; C3 ¼ C 2 R þ ).
3
Remark: Take care in your analysis and calculation P the mutual induc-
tances between helical antennas in our system ( M ¼ M þ M ).
Chapter 9
Microwave RF Antennas and Circuits
Bifurcation Behavior, Investigation,
Comparison and Conclusion
Microwave RF antennas are an integral part of every RF or microwave system. An

antenna is an electrical device which converts electric power into radio waves, and
vice versa. In many wireless applications antennas are required by radio receiver or
transmitter to couple its electrical connection to the electromagnetic field. When we
inspect system stability which includes radio waves, we inspect electromagnetic
waves which carry signals through the space (or air) at the speed of light with
almost no transmission loss. There are mainly two category antennas, the first is
omnidirectional antenna which receives and/or radiate in all directions. The second
is directional antenna which radiates in a particular direction or pattern. Antennas
are characterized by a number of parameters, radiation pattern and the resulting
gain. Antenna’s gain is dependent on its power in the horizontal directions, and
antenna’s power gain takes into account the antenna’s efficiency (figure of merit).
The physical size of an antenna is a practical issue, particularly at lower frequen-
cies. Resonant antennas mainly use a linear conductor or pair of such elements.
When we implement RF or microwave antennas in higher frequency system (UHF,
microwave), there is no essential need for a smaller physical size. Another
important antenna parameter is the frequency range or bandwidth over which an
antenna functions. The antenna bandwidth can be wide or narrow like in resonant
antennas. In every RF system which includes antenna, we need to choose the
suitable matching network between the transceiver and the antenna. Matching
network is the practical circuit which is corresponding to maximize the power
transfer or minimize signal reflection from the load RF antenna to the transceiver
unit. In wireless application we can differentiate RF and microwave antennas. Radio
spectrum antennas cover radio waves, microwaves and terahertz radiations. Optical
spectrum covers infrared, visible, UV, X-rays and gamma radiations. Radio waves
antennas range from 3 kHz to 300 GHz. Hence RF starts from much lower than the
microwave starting range. Microwave antennas are mainly for EM waves above
1 GHz in frequency. RF and microwave antenna ranges are different in operation
range and applications are concerned. Microwave range starts from 300 MHz to
300 GHz and most microwave applications range up to 100 GHz.
DOI 10.1007/978-3-319-45427-6_9
904 9 Microwave RF Antennas and Circuits Bifurcation …
We can characterize microwave antenna as a high antenna gain and directivity,

large bandwidth, travel by line of sight, microwaves penetrate ionosphere with less
attenuation and less distortion, and in applications of 1–10 GHz range microwave
noise level is very low and hence very low signal can be easily detected at receiver.
The mainly application for RF antennas are mobile, AM/FM radio, television and
the mainly applications for microwave antennas are radar, satellite and space
communication. Antennas are used over much broader frequency ranges and are
achieved using further techniques. The adjustment of an antenna matching network
can allow for any antenna to be matched at any frequency. The loop antennas have
a very narrow bandwidth and are tuned using a parallel capacitance which is
adjusted according to the receiver tuning. The complex impedance of antennas is
related to the electrical length of the antenna at the wavelength in use. The impe-
dance of an antenna can be matched to the feed line and radio by adjusting the
impedance of the feed line. The antenna feed line is as an impedance transformer
and the impedance is adjusted at the load with an antenna tuner, a balun, a matching
transformer, matching networks composed of inductors and capacitors (T-type,
Pi-type, and L-type), or matching sections. We choose wide range of RFID
antennas to fit an equally wide range of tags, readers, and systems. This includes
UHF antennas, patch antennas, and linear or circular polarized antennas.
Each RFID antenna has different strengths, and each fits specific types of RFID
systems. In RFID system, tags are attached to all items that are to be tracked. These
tags are made from tag chip (RFID IC), that is connected to an antenna that can be
built into many and wide variety of industrial asset tags. The tag chip contains
memory which stores the product EPC and other variable information so that it can
be read and tracked by RFID reader anywhere. In our analysis we represent tag chip
as a parallel resistor (Rrfid) and capacitor (Crfid). An RFID reader is a network
connected device (fixed or mobile) with an antenna that sends power as well as data
and commands to the tags. The RFID reader acts like an access point for RFID
tagged items. An RFID tags are comprised of an integrated circuit (RFID IC)
attached to an antenna that has been printed, etched, stamped or vapor-deposited
onto a mount which is often a paper substrate or PolyEthylene Terephthalate (PET).
We inspect RF and microwave systems which involve, RF and microwave devices
(RFID transponders, RF transistors, RF diodes, MMICs, Reflection Type Phase
Shifter (RTPS), cylindrical RF network antennas, Tunnel Diode (TD), microwave
field effect transistor (FETs), Impact Ionization Avalanche Transit Time (IMPATT),
PIN diodes, Small Signal (SS) amplifiers, matching networks, Power Amplifiers
(PAs), RF oscillators, RF filters, and antenna systems) as a dynamical system where
a fixed rule describes the time dependent of specific RF circuit voltage in a geo-
metrical space. Examples include the mathematical models that describe circuit
with RF and microwave devices. At any time a dynamical system has a state given
by a set of real numbers (a vector) which can be represented by a specific voltage in
an appropriate state space (a geometrical manifold). Small changes in the state of
the RF and microwave system create small changes in the numbers. The evolution
rule of the RF and microwave dynamical system is a fixed rule that describes what
future states follow from the current state. The rule is deterministic, for a given time
9 Microwave RF Antennas and Circuits Bifurcation … 905
interval only one future state follows from the current state. A RF and microwave
dynamical system is a phase (or state) space endowed with a family of smooth
evolution functions that for any element of time (t), map a point of the phase space
back into the phase space. RF and microwave systems can be described by num-
bers. The state vector is a numerical description of the current configuration of a
system. For example, RFID circuit with RFID IC and antenna (inductance element)
can be described using some numbers: its voltages (V1 ; V2 ; V3 ; . . .) and currents
(I1 ; I2 ; I3 ; . . .). Once we know these numbers V1 ; V2 ; V3 ; . . . and I1 ; I2 ; I3 ; . . . the
voltages and currents trajectories are completely determined. The group of numbers
(V1 ; V2 ; V3 ; . . ., I1 ; I2 ; I3 ; . . .) is a vector which completely describes the state of our
RF and microwave system and hence is called the state vector system. There are
two main behaviors which are related to RF and microwave systems: (1) the system
gravitates toward a fixed point, or (2) the system blowup. There are additional cases
of oscillators related to RF and microwave system. We assume fi ; i ¼ 1; 2; 3; . . . are
differentiable with continuous derivatives. The vectors Vi are the state of the
microwave and RF dynamical system, and the functions fi ; i ¼ 1; 2; 3; . . . tell us
how the system moves. In special circumstances, however, the system does not
move. The system can be stuck (we will say fixed) in a special state; we call these
states fixed points of the dynamical RF and microwave system. Not all fixed points
are the same. We call some stable and others unstable. Is the specific fixed point
Vi ; Ii stable or unstable? The answer is, neither. To see that it is not stable, consider
any points Vi0 ; Ii0 near (but not equal to) Vi ; Ii (i = 1, 2, 3…). At t ! 1 the RF
system never approaches (Vi ; Ii ). Further, Vi ; Ii are not unstable. To be unstable,
points near Vi ; Ii must be sent “far” away from Vi ; Ii . Clearly, if we start at certain
distance from Vi ; Ii the system does not get any farther away. Stable fixed points
give excellent information about the fate of a dynamical system.
In our analysis we investigate RF and microwave circuits bifurcation and dy-
namical behavior. Bifurcation behavior in our RF system is the study of changes in
the qualitative or topological structure of RF system, vector fields, and the solutions
of a family of differential equations. A bifurcation occurs when a small smooth
change made to the parameter values (the bifurcation parameters) of a RF and
microwave system causes a sudden topological change in behavior. Bifurcations
occur in both our continuous RF systems (ODEs, DDEs, and PDEs) and discrete
systems (described by maps). We can inspect in our microwave and RF system two
principal bifurcation classes: local bifurcation, which our RF system can be ana-
lyzed through changes in the local stability properties of equilibria, periodic orbits
or other invariant sets as parameters cross through critical thresholds. Global
bifurcations, which often occur in RF and microwave system happened when larger
invariant sets of the system collide with each other, or with equilibria of the system.
They cannot be detected only by stability analysis (fixed points). A local bifurcation
occurs when a system parameter change causes the stability of an equilibrium (or
fixed point) to change. In continuous system, this corresponds to the real part of an
eigenvalue of an equilibrium passing through zero. Global bifurcations occur when
larger invariant sets, such as periodic orbits, collide with equilibria. This causes
changes in the topology of the trajectories in the phase space which cannot be
confined to a small neighborhood (local bifurcation). The changes in system
topology extend out to an arbitrarily large distance. In our analysis we pay attention
to the co-dimension of a bifurcation which is the number of parameters which must
be varied for the bifurcation to occur. The co-dimension of the parameter set for
which the bifurcation occurs within the full space of RF system parameters. The
simple case for stability analysis is when there are no time delay elements in our RF
and microwave system (s1 ¼ 0; s2 ¼ 0; s3 ¼ 0; . . .; D1 ¼ 0; D2 ¼ 0; D3 ¼ 0; . . .;
Dli ¼ 0; i ¼ 1; 2:3; . . .).
If our RF and microwave system involving Na variables (Na > 2), the charac-
P a k
teristic equation is of degree Na ( Nk¼0 k Pk ¼ 0) and must often be solved
numerically. Expect in some particular cases, such an equation has Na distinct roots
that can be real or complex. These values are the eigenvalues of the Na Na
Jacobian matrix (A). The general rule is that the Steady State (SS) is stable if there
is no eigenvalue with positive real part. It is sufficient that one eigenvalue is positive
for the steady state to be unstable. Our Na-variables RF and microwave system has
Na eigenvalues. The type of behavior can be characterized as a function of the
position of these eigenvalues in the Re/Im plane. Five non-degenerated cases can be
distinguished: (1) the Na eigenvalues are real and negative (stable steady state),
(2) the Na eigenvalues are real, Na − 1 of them are negative (unstable steady state),
(3) and (4) two eigenvalues are complex conjugates with a negative real part and the
other eigenvalues are real and negative (stable steady state), two cases can be
distinguished depending on the relative value of the real part of the complex
eigenvalues and of the real one, (5) two eigenvalues are complex conjugates with a
negative real part and at least one eigenvalue is positive (unstable steady state). The
next case is when there are delay elements in our RF and microwave system and we
can’t neglect them. In that case, our RF and microwave system stability analysis is
related to two main cases: first case, RF or microwave system is characterized by a
set of voltages (V1 ; V2 ; V3 ; :. . .) or/and currents (I1 ; I2 ; I3 ; . . .). Due to electromag-
netic interferences there are differences in time delays with respect to system
voltages and current variables
V1 ðtÞ ! V1 ðt s1 Þ; V2 ðtÞ ! V2 ðt s2 Þ; Vi ðtÞ ! Vi ðt si Þ; I1 ðtÞ ! I1 ðt s1 Þ

I2 ðtÞ ! I2 ðt s2 Þ; Ii ðtÞ ! Ii ðt si Þ; i ¼ 1; 2; 3; . . .; sk 6¼ sl ; k 6¼ l; si 2 R þ :
Sometimes the delay in time is related to circuit voltages derivatives:
dV1 ðtÞ dV1 ðt D1 Þ dV2 ðtÞ dV2 ðt D2 Þ

! ; ! ;
dt dt dt dt
dVi ðtÞ dVi ðt Di Þ
! ; i ¼ 1; 2; 3; . . .
dt dt
We assume Dk ¼ 6 Dl ; k 6¼ l; Di 2 R þ . The stability of a given steady state is

determined by the graphs of some function of s1 ; s2 ; s3 ; . . . or/and D1 ; D2 ; D3 ; . . .
which can be expressed explicitly and thus can be easily depicted by software. f
need to look at one such function and locate the zeros. The stability switching is due
to different values of delay parameters s1 ; s2 ; s3 ; . . . or/and D1 ; D2 ; D3 ; . . .. Second
case, RF and microwave circuits include microstrip lines. Microstrip lines have
parasitic effects, a delay in time Dli ; i ¼ 1; 2; 3; . . .. The delays are on the current
that flow through microstrip lines IðtÞ ! Iðt Dli Þ; i ¼ 1; 2; 3; . . .. We consider
VDli ! e (neglect the voltage on microstrip lines). We inspect the stability behavior
and stability switching under variation of delay parameters Dli ; i ¼ 1; 2; 3; . . .. In
our RF and microwave systems which include antennas (inductances elements), we
take care in our analysis
P and calculation the mutual inductances between antennas
within the system ( M ¼ M þ M ).
If we minimize our RF and microwave system to specific case where there is
only one time delay parameter s then the general geometric criterion: The occur-
rence of any possible stability switching resulting from the increase of value of the
time delay s for our RF and microwave system characteristic equation.
Dðk; s1 ; s2 ; s3 ; . . .Þ ¼ detðA k IÞ; Dðk; s1 ; s2 ; s3 ; . . .Þ ¼ 0;

X
n X
m
k¼0 k¼0
pk ðÞ; qk ðÞ: R þ 0 ! R are continuous and differentiable functions of s.
assume (Pn ðk; sÞ ! P; Qm ðk; sÞ ! Q):
(II) lim supfjQm ðk; sÞ=Pn ðk; sÞj : jkj ! 1; Rek 0g\1 for any s.
whenever it exists.
of characteristic equation situated on the imaginary axis of the complex k—plane,
where by increasing the parameters of RF and microwave system, Re k may, at the
crossing Change its sign from (−) to (+), i.e. from a stable focus Eð0Þ to an unstable
the partial derivatives with respect to s and RF/Microwave system parameters
[2, 3].

1 @Rek
^ ðsÞ ¼ ;
@s k¼ix
Other RF and microwave parameters ¼ const where x 2 R þ :
PN
2k
¼ 0 ; N 2 Z þ when writing
PðkÞ ¼ PR ðkÞ þ i PI ðkÞ and QðkÞ ¼ QR ðkÞ þ i QI ðkÞ, and inserting k ¼ i x into
RF system’s characteristic equation, x must satisfy the following:
jQði xÞj2
jQði xÞj2
where jQði xÞj2 6¼ 0 in view of requirement (a) above, ðg; hÞ 2 R. Furthermore, it

follows above sin x s and cos x s equations that, by squaring and adding the
sides, x must be a positive root of FðxÞ ¼ jPði xÞj2 jQði xÞj2 ¼ 0. Note: FðxÞ
is dependent of s. Now it is important to notice that if s 62 I (assume that I R þ 0 is
the set where xðsÞ is a positive root of FðxÞ and for s 62 I , xðsÞ is not define. Then
for all s in I, xðsÞ is satisfies that Fðx; sÞ ¼ 0). Then there are no positive xðsÞ
solutions for Fðx; sÞ ¼ 0, and we cannot have stability switches. For any s 2 I ,
where xðsÞ is a positive solution of Fðx; sÞ ¼ 0, we can define the angle hðsÞ 2
½0; 2 p as the solution of the below equations:
sin hðsÞ ¼
jQði xÞj2
cos hðsÞ ¼
jQði xÞj2
þ n2p
given by sn ðsÞ ¼ hðsÞxðsÞ ; n 2 N0 ; s 2 I. Let us introduce the functions I ! R ;
following, the subscripts k; x; R1 ; C1 and RF microwave parameters indicate the
corresponding partial derivatives. Let us first concentrate on, ^ðxÞ remember in
k ðparametersÞ; x ðparametersÞ, and keeping all parameters except one (x) and s.
The derivation closely follows that in reference [BK]. Differentiating RF system
characteristic equation PðkÞ þ QðkÞ eks ¼ 0 with respect to specific parameter
(x), and inverting the derivative, for convenience, one calculates:
Remark: x ¼ RF system specific parameter

1
@k Pk ðk; xÞ Qðk; xÞ þ Qk ðk; xÞ Pðk; xÞ s Pðk; xÞ Qðk; xÞ
¼
where Pk ¼ @P
@k ; :. . . etc., Substituting k ¼ i x, and bearing Pði xÞ ¼ Pði xÞ,
Qði xÞ ¼ Qði xÞ then i Pk ði xÞ ¼ Px ði xÞ and i Qk ði xÞ ¼ Qx ði xÞ
that on the surface jPði xÞj2 ¼ jQði xÞj2 , one obtains
1 !
jk¼ix ¼
Upon separating into real and imaginary parts, with P ¼ PR þ i PI ; Q ¼

QR þ i QI Px ¼ PRx þ i PIx ; Qx ¼ QRx þ i QIx ; Px ¼ PRx þ i PIx ; Qx ¼
QRx þ i QIx P2 ¼ P2R þ P2I . When (x) can be any RF system parameters and time
delay s etc., where for convenience, we have dropped the arguments ði x; xÞ, and
where Fx ¼ 2 ½ðPRx PR þ PIx PI Þ ðQRx QR þ QIx QI Þ; Fx ¼ 2 ½ðPRx
PR þ PIx PI Þ ðQRx QR þ QIx QI Þ xx ¼ Fx =Fx . We define U and V: U ¼
ðPR PIx PI PRx Þ ðQR QIx QI QRx Þ V ¼ ðPR PIx PI PRx Þ ðQR
QIx QI QRx Þ. We choose our specific parameter as time delay x = s.
@s þ Fs ¼ 0; s 2 I ) @s ¼ Fx
Fs

@Rek
^1 ðsÞ ¼ ;
@s k¼ix
( )
1 2 ½U þ s jPj2 þ i Fx
^ ðsÞ ¼ Re ;
@x Fs
¼ xs ¼
@s Fx

@Rek
signf^1 ðsÞg ¼ sign ;
@s k¼ix
( )
1 @x U @x
@s þ V
signf^ ðsÞg ¼ signfFx g sign s þxþ
@s jPj2

in RF system, about the equilibrium point as a result of a variation of delay
parameters. The analysis consists in identifying the roots of our system charac-
teristic equation situated on the imaginary axis of the complex k-plane where by
increasing the delay parameter s, Re k may at the crossing, change its sign from
− to +, i.e. from a stable focus E(*) to an unstable one, or vice versa. This feature
tos, ^1 ðsÞ ¼ ð@ Re
@s Þk¼ix :
k
Among microwave elements there are discrete circuit, packages

diodes/transistors which are mounted in coax and waveguide assemblies, mi-
crowave integrated circuit and monolithic microwave integrated circuit. Stability
analysis is also done for microwave transmission lines and full optimization. Many
microwave circulators involve Reflection Type Phase Shifter (RTPS). Microstrip
transmission lines are integral parts in RF circulator. A special type of antenna
system is a cylindrical RF network antenna for coupled plasma sources which
include copper legs. Many RF circuits include Tunnel Diode (TD) which is a p-n
junction device that exhibits negative resistance. Microwave oscillators integrate
Tunnel Diode (TD) elements and stability switching analysis is done. In many
applications there are microwave semiconductor amplifiers. Bipolar transistor,
FETs and IMPATT amplifiers are inspected when we integrate them in RF and
microwave systems. Devices internal parameters influence the functionality of those
circuit and RF modules. Internal parameters variation and circuit microstrip lines
parasitic effects are inspected for best performance. There are many types of
amplifiers, among them zero frequency amplifiers (DC amplifiers), audio amplifiers,
RF amplifiers, and they come in three basic flavours: Common Base
(CB) amplifiers, Common Collector (CC) amplifiers, and Common Emitter
(CE) amplifiers. It is very important to design the right matching network which
matches between the amplifier output and the load. Good matching avoids reflec-
tions and enhances the functionality of our RF system. Bias—T circuit is a very
important element of any RF and microwave system which there is a need to
combine RF and DC signals or to separate combined signal to RF and DC signals.
Stability analysis of Bias—T circuit is done when integrating it in RF system to get
the best performances. Power Amplifiers (PAs) are analysed for best performances
and stability analysis is done. There are different types of amplifiers which clas-
sified according their circuit configurations and method of operation. There are two
basic amplifier classes groups. The first are the classically controlled conduction
angle amplifiers forming the more common amplifier classes (A, B, AB, and C).
The second set amplifiers are the newer so-called “switching” amplifier classes (D,
E, F, G, S, and T). We analyse the stability of these amplifiers by inspecting the
equivalent circuit differential equations. BJT transistor is replaced by large signal
model and more enhance model is Gummel-Poon model. We analyse the stability
of wideband LNAs with negative feedback under circuit’s parameters variation.
A microwave oscillator is an active device to generate power and a resonator to
control the frequency of the microwave signal. Important issues in oscillators are
frequency stability, frequency tuning, and phase noise. Phase shift resonator circuit
is widely used and stability analysis is done by considering BJT Small Signal
(SS) equivalent circuit model. Closed loop functioning oscillator can be viewed as
feedback system. There are three types of transistor LC oscillators, Colpitts,
Hartley, and Clapp. Colpitt’s oscillator circuit stability analysis is done by criterion
of Liapunov. The target of analogy and RF filtering is to modify the magnitude and
phase of signal frequency components. Many analogy or Radio Frequency
(RF) circuits perform filtering on the signal passing through them. A diplexer is a
passive device that implements frequency domain multiplexing. We analyse BPF
diplexer circuit stability by using geometric stability switch criteria in delay dif-
ferential systems. There are lattice crystal filter, half lattice and cascaded half lattice
filters. A tunable BPF employing varactor diodes is ideal for many diverse wireless
applications. BPF (varactor diodes) circuit involving N variables and stability
behaviour is inspected. An antenna is a conductor or group of conductors used for
radiating electromagnetic energy into space or collecting electromagnetic energy
from space. There are many types of antennas and the operation at microwave
frequencies is inspected for the best performances. N—turn multilayer circular coil
antennas with RFID IC is investigated and stability optimization under delayed
electromagnetic interference and parasitic effects is analysed. Double rectangular
spiral antennas are constructed from two antennas and they are connected in series
with microstrip lines and RFID IC. A system of single turn square planar straight
thin film inductors antenna is constructed from four straight thin film inductors
which are connected in a single turn square structure. A Helical antenna is an
antenna consisting of a conducting wire wound in the form of a helix. Helix antenna
system stability is inspected under parameters variation.
Appendix A
RFID LF TAG 125 kHz/134 kHz
Design and Analysis
A.1 LF TAG 125 kHz/134 kHz Design and Analysis
We have RFID Antenna system on a substrate, width 300 lm and lr = 450. The
antenna is constructed from silver ointment which his resistance is bigger than pure
silver by 50 %. Track width is 20 lm, gap between tracks 20 lm, track depth is
20 lm up to 100 lm. The requested antenna inductance is 2.66 mH and parasitic
resistance less than 10 X. We need to find the number of rectangular spiral
antenna’s turn (Nc). Since the RFID antenna substrate permeability is 450 (lr =
450), RFID antenna permeability is average value between air permeability and the
magnet. We consider that the TAG permeability is 100–300 (lr). Possible TAG
dimensions: 5 mm 5 mm, 6 mm 6 mm, 7 mm 7 mm, 8 mm 8 mm and
permeability 100, 200, 300 (lr) (Fig. A.1).
Part A: Analysis for l = l0 and l = lr l0 (lr = 450).
H H
l0 ¼ 4 p 107 1:2566 106 ;
m m
l
lr ¼ ) l ¼ lr l0 jlr ¼450 ¼ 450 1:2566 106 ¼ 565:47 106
l0
ðt þ wÞ ð20 lm þ 20 lmÞ
t ¼ g ¼ 20 lm ) d ¼ 2 ¼2 ¼ 25:47 lm; t ¼ 20 lm
P 3:14
d—Equivalent diameter of the track
Aavg ¼ A0 Nc ðg þ wÞ ¼ 0:005 Nc ð20 lm þ 20 lmÞ ¼ 0:005 Nc 40 lm

¼ 0:005 Nc 40 106
Bavg ¼ B0 Nc ðg þ wÞ ¼ 0:005 Nc ð20 lm þ 20 lmÞ ¼ 0:005 Nc 40 lm
¼ 0:005 Nc 40 106

DOI 10.1007/978-3-319-45427-6
914 Appendix A: RFID LF TAG 125 kHz/134 kHz Design and Analysis
5 mm
0.3 mm
5 mm
Fig. A.1 RFID antenna system constructed from silver ointment
Fig. A.2 RFID rectangular spiral antenna overall parameters

the coil. t—Track thickness, w—Track width, g—Gap between tracks. Nc—
Number of turns, d—Equivalent diameter of the track. Average coil area;
Ac = Aavg Bavg. Integrating all those parameters give the equations for induc-
tance calculation (Fig. A.2):
2 3 2 3
6 2 Aavg Bavg 7 6 2 Aavg Bavg 7
X1 ¼ Aavg ln4 qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi5; X2 ¼ Bavg ln4 qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi5
d ðAavg þ Aavg þ Bavg
2 2 d ðBavg þ Aavg þ Bavg
2 2
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi ðAavg þ Bavg Þ

X3 ¼ 2 ½Aavg þ Bavg A2avg þ B2avg ; X4 ¼
4
Appendix A: RFID LF TAG 125 kHz/134 kHz Design and Analysis 915
Table A.1 RFID coil manufacturing technology (wired, etched, printed)

Wired 1.8–1.9
Etched 1.75–1.85
Printed 1.7–1.8

Lcalc ¼ lp0 ðX1 þ X2 X3 þ X4 Þ Ncp ; L1 = Lcalc
inductance calculation (Table A.1).
We integrate the Lcalc value inside the differential equations which characterize
the RFID system with the Coil inductance. Nc ! z; A0 ! x; B0 ! y
MATLAB Script:
Z = 10; x = 0.005 – z * 40 * (1e−6); y = 0.005 – z * 40 * (1e−6); x1 = x *
log (2 * x * y/(25.47 * (1e−6) * (x + sqrt(x * x + y * y)))); x2 = y * log(2 * x * y/
(25.47 * (1e−6) * (y + sqrt(x * x + y * y)))); x3 = 2 * (x + y− sqrt(x * x + y * y));
x4 = (x + y)/4; l = ((1.2566 * (1e−6)/3.14) * (x1 + x2 − x3 + x4) * power(z,1.8))
The results in Table A.2:
Result: the most close inductance to our request 2.66 mH is 2.8 mH (Nc = 30,
lr = 450).
Rectangular spiral RFID antenna length calculation & resistance
calculate the total length (Fig. A.3).
the coil. w—Track width, g—Gap between tracks. Nc—Number of turns. L0 is the
length of the first turn L0 ¼ 2 ðA0 þ B0 Þ ðw þ gÞ. Lk is the length of turn k + 1.
X
N c 1
k¼1
Table A.2 RFID rectangular spiral Lcalc for different values of Nc and l
Nc l = l0; lr = 1 Lcalc (5 mm 5 mm) − [H] Lcalc (8 mm 8 mm) − [H]
10 l0 = 1.25.66 (1e−6)H/m 1.085 lH 1.98 lH
20 l0 3.38 lH 6.48 lH
30 l0 6.21 lH 12.5 lH
60 l0 1.35 lH 34.6 lH
100 l0 10.025 lH 55.7 lH
150 l0 −1.37 1e−5H (N/A) 50.7 lH
120 l0 1.35 lH 61 lH
170 l0 −3.97 1e−5H (N/A) 33 lH
200 l0 −1.0582 1e−4H (N/A) NaN
220 l0 −1.7096 1e−4H (N/A) −1.9 1e−5H (N/A)
250 l0 −3.059 1e−4H (N/A) −9.19 1e−5H (N/A)
Nc l = lr l0; lr = 450 Lcalc (5 mm 5 mm) − [H] Lcalc (8 mm 8) − [H]
10 565.47 (1e−6)H/m 0.48 mH 0.89 mH
20 565.47 (1e−6)H/m 1.5 mH 2.9 mH
30 565.47 (1e−6)H/m 2.8 mH 5.7 mH
60 565.47 (1e−6)H/m 6.1 mH 15.6 mH
100 565.47 (1e−6)H/m 4.5 mH 26 mH
150 565.47 (1e−6)H/m −0.0062 N/A 22.8 mH
120 565.47 (1e−6)H/m 0.611 mH 27.5 mH
170 565.47 (1e−6)H/m −0.0179 N/A 14.9 mH
200 565.47 (1e−6)H/m −0.0476 N/A NaN
220 565.47 (1e−6)H/m −0.0769 N/A −0.0088 N/A
250 565.47 (1e−6)H/m −0.1377 N/A −0.0414 N/A

X
Nc 1
k¼1
NX
C 1
k¼1
NX
C 1
¼2 fðA0 þ B0 Þ 4 k ðw þ gÞg

k¼1
NX
C 1
¼ 2 ðA0 þ B0 Þ ðNC 1Þ 2 ½4 k ðw þ gÞ

k¼1
NX
C 1 NX
C 1
k¼1 k¼1
X
N c 1
k¼1
LT ¼ L0 þ 2 ðNC 1Þ ½A0 þ B0 4 ðw þ gÞ ¼ L0 þ 2 ðA0 þ B0 Þ

ðw þ gÞ þ 2 ðNC 1Þ
½A0 þ B0 4 ðw þ gÞ ¼ 2 ðA0 þ B0 Þ ðw þ gÞ þ 2 ðNC 1Þ ðA0 þ B0 Þ
8 ðNC 1Þ ðw þ gÞ
¼ L0 þ 2 ðA0 þ B0 Þ NC
ðw þ gÞ ð8 NC 7Þ
Final result:
LT ¼ L0 þ 2 ðA0 þ B0 Þ NC ðw þ gÞ ð8 NC 7Þ
LT LT
RDC ¼ ¼
r S r p a2
LT—total length of the wire. r—Conductivity of the wire (m X/m).

S—Cross section area p a2 . a—radius of the wire.
RDC ¼ ¼ ¼
rS rpa 2 r p a2
6
A0 þ B0 ¼ 0:01 m; NC ¼ 30; w þ g ¼ 40 10 ) LT ¼ 0:6106 m ¼ 61:06 cm
Cross section area
S ¼ 20 lm 20 lm ¼ 400 1012 m2
Conductivity of Silver r ¼ 6:1 107 ðf mÞ. Conductivity has SI units of

Siemens per meter (S/m).
rsilver@20 C ¼ 6:3 107 ðS/mÞ:
it as a variable and find his minimum value for RDC < 10 X. S ¼ 20 lm x.
0:6106
\10 ) x [ 4:846 105 m ¼ 48:46 lm; RDC@t¼20 lm
6:3 107 20 106 x
¼ 24:23 X
conductivity.
rsilver@20 C 6:3 107 ðS=mÞ

¼ ¼ 3:15 107 ðS=mÞ;
2 2
RDC@t¼20 lm ¼ 48:46 X:
0:6106
\10 ) x [ 9:6921 105 m ¼ 96:92 lm
3:15 107 20 106 x
Conclusion: In case of pure silver track. The track depth needs to be bigger than
48.46 lm to meet inductance resistance less than 10 X. In case of mixture of silver
then track depth needs to be bigger than 96.92 lm.
Part B: Analysis for l = lr l0 (lr = 100, 200, 300) (Table A.3).
Results: The most close inductance analysis to 2.66 mH is 2.7 mH and subcases
Result B.1: Lcalc = 2.7 mH, Nc = 60, (A0 = 7 mm) (B0 = 7 mm); l = lr
l0; lr = 100
l = 125.66 (1e−6)H/m.
LT LT
RDC ¼ ¼
r S r p a2

RDC ¼ ¼ ¼
r S r p a2 r p a2
6
A0 þ B0 ¼ 0:014 m; NC ¼ 60; w þ g ¼ 40 10 ) LT ¼ 1:689 m
Cross section area S ¼ 20 lm 20 lm ¼ 400 1012 m2

Conductivity of Silver r ¼ 6:1 107 ðf=mÞ. Conductivity has SI units of
rsilver@20 C ¼ 6:3 107 ðS=mÞ:
1:689
\10 ) x [ 1:3405 104 m ¼ 134:05lm;
6:3 107 20 106 x
RDC@t¼20 lm ¼ 67 X
conductivity.
rsilver@20 C 6:3 107 ðS=mÞ 1:6890

¼ ¼ 3:15 107 ðS=mÞ: \10
2 2 3:15 107 20 106 x
) x [ 2:681 104 m ¼ 268:1 lm; RDC@t¼20 lm ¼ 134 X
Result B.2: Lcalc= 2.7 mH, Nc = 60, (A0 = 5 mm) (B0 = 5 mm); l = lr l0;
lr = 200
l = 251.32 (1e−6)H/m.
Table A.3 RFID rectangular spiral Lcalc for different values of Nc and l and tag overall dimension
920
Nc l = lr l0; lr = 100 Lcalc Lcalc Lcalc (7 mm 7 mm) − [H] Lcalc

l = 125.66(1e−6)H/m (5 mm 5 mm) − [H] (6 mm 6 mm) − [H] (8 mm 8 mm) − [H]
10 125.66 (1e−6)H/m 1.085 10−4 1.377 10−4 1.677 10−4 1.986 10−4
20 125.66 (1e−6)H/m 3.3837 10−4 4.384 10−4 5.4196 10−4 6.4831 10−4
30 125.66 (1e−6)H/m 6.2129 10−4 8.2571 10−4 0.001 0.0013
60 125.66 (1e−6)H/m 0.0014 0.002 0.0027 = 2.7 mH 0.0035
100 125.66 (1e−6)H/m 0.001 0.0024 0.0041 0.0058
150 125.66 (1e−6)H/m −0.0014 NaN 0.0021 0.0051
120 125.66 (1e−6)H/m 1.3598 10−4 0.0018 0.0038 0.0061
170 125.66 (1e−6)H/m −0.0040 −0.0012 2.5454 10−4 0.0033
200 125.66 (1e−6)H/m −0.0106 −0.0062 −0.0023 NaN
220 125.66 (1e−6)H/m −0.0171 −0.0115 −0.0063 −0.0020
250 125.66 (1e−6)H/m −0.0306 −0.023 −0.0158 −0.0092
Nc l = lr l0; lr = 200 Lcalc Lcalc Lcalc (7 7 mm) − [H] Lcalc (8 8 mm) − [H]
l = 251.32(1e−6)H/m (5 mm 5 mm) − [H] (6 mm 6 mm) − [H]
10 251.32 (1e−6)H/m 2.17 10−4 2.7539 10−4 3.3553 10−4 3.9719 10−4
20 251.32 (1e−6)H/m 6.7674 10−4 8.7693 10−4 0.0011 0.0013
30 251.32 (1e−6)H/m 0.0012 0.0017 0.0021 0.0025
60 251.32 (1e−6)H/m 0.0027 = 2.7 mH 0.004 0.0055 0.0069
100 251.32 (1e−6)H/m 0.0020 0.0049 0.0081 0.0116
150 251.32 (1e−6)H/m −0.0028 NaN 0.0042 0.0102
120 251.32 (1e−6)H/m 2.7196 10−4 0.0035 0.0077 0.0122
170 251.32 (1e−6)H/m −0.008 −0.0025 5.09 10−4 0.0066
200 251.32 (1e−6)H/m −0.0212 −0.0123 −0.0046 NaN
220 251.32 (1e−6)H/m −0.0342 −0.0229 −0.0127 −0.0039
250 251.32 (1e−6)H/m −0.0612 −0.046 −0.0316 −0.0184
Appendix A: RFID LF TAG 125 kHz/134 kHz Design and Analysis
Table A.3 (continued)
Nc l = lr l0; lr = 300 Lcalc Lcalc Lcalc Lcalc
l = 376.98 (1e−6) (5 mm 5 mm) − [H] (6 mm 6 mm) − [H] (7 mm 7 mm) − [H] (8 mm 8 mm) − [H]
H/m
10 376.98 (1e−6)H/m 3.2562 10−4 4.1309 10−4 5.0329 10−4 5.9579 10−4
20 376.98 (1e−6)H/m 0.001 0.0013 0.0016 0.0019
30 376.98 (1e−6)H/m 0.0019 0.0025 0.0031 0.0038
60 376.98 (1e−6)H/m 0.0041 0.0061 0.0082 0.0104
100 376.98 (1e−6)H/m 0.003 0.0073 0.0122 0.0173
150 376.98 (1e−6)H/m −0.0041 NaN 0.0062 0.0153
120 376.98 (1e−6)H/m 4.0794 10−4 0.0053 0.0115 0.0183
170 376.98 (1e−6)H/m −0.0119 −0.0037 7.6362 10−4 0.0099
200 376.98 (1e−6)H/m −0.0317 −0.0185 −0.0069 NaN
220 376.98 (1e−6)H/m −0.0513 −0.0344 −0.0190 −0.0059
250 376.98 (1e−6)H/m −0.0918 −0.069 −0.0474 −0.0276
921
LT LT
RDC ¼ ¼
r S r p a2

RDC ¼ ¼ ¼
r S r p a2 r p a2
A0 þ B0 ¼ 0:01 m; NC ¼ 60; w þ g ¼ 40 106 ) LT ¼ 1:201 m
Cross section area
S ¼ 20 lm 20 lm ¼ 400 1012 m2

rsilver@20 C ¼ 6:3 107 ðS=mÞ:
1:201
\10 ) x [ 9:5317 105 m ¼ 95:317lm; RDC@t¼20 lm
6:3 107 20 106 x
¼ 47:65 X
conductivity.
rsilver@20 C 6:3 107 ðS/mÞ

¼ ¼ 3:15 107 ðS/mÞ:
2 2
1:201
\10 ) x [ 19:06 105 m ¼ 190:6 lm;
3:15 10 20 106 x
7
RDC@t¼20 lm ¼ 95:30 X
A.2 LF TAG 125 kHz/134 kHz Two Rectangular

Spiral Antennas Design Analysis
We have RFID Antenna system on a substrate, width 300 lm and lr = 100… 450.
The construction is made from two rectangular spiral antennas, the first rectangular
spiral antenna is on the other rectangular spiral antenna (Symmetric or antisymmetric
structure, mirror picture). We consider the symmetric structure for our calculation.
Each antenna is constructed from silver ointment which his resistance is bigger than
pure silver by 50 %. The track width is 20 lm, gap between tracks 20 lm, track
depth is 20 lm up to 100 lm. The requested antennas total inductance is 2.66 mH
and parasitic resistance less than 10 X. Both rectangular spiral antennas are in series.
We need to find the number of rectangular spiral antenna’s turn (Nc). We consider
that the first and second spiral antennas have the same number of turns
(Nc1 = Nc2 = Nc). Since the RFID antenna substrate permeability is between 100
and 450 (lr = 100… 450), RFID antenna permeability is an average value between
air permeability and the magnet. We consider that the TAG permeability is 100–300
(lr). Possible Antennas dimensions: 2 mm 2 mm, 3 mm 3 mm, 4 mm 4
mm, and permeability 100, 200, 300, 450 (lr). We neglect micro strip parasitic
resistance (Fig. A.4).
Double rectangular spiral antennas can be represented as a two inductors in
series (Lcalc-1 and Lcalc-2), parasitic resistances (RDC-1 and RDC-2) and micro strip
(neglect parasitic resistance). The rectangular spiral antennas in series are connected
in parallel to RFID TAG IC. The Equivalent Circuit of Passive RFID TAG with
double rectangular antennas is Capacitor (C1) and Resistor (R1) in parallel with
double rectangular antennas in the series (Fig. A.5).
Lcalc-1 and Lcalc-2 are mostly formed by traces on planar PCB. 2 Lm element
represents the mutual inductance between Lcalc-1 and Lcalc-2. Since two inductors
(Lcalc-1, Lcalc-2) are in series and there is a mutual inductance between Lcalc-1 and
Lcalc-2, the total antenna inductance LT: LT = Lcalc-1 + Lcalc-2 + 2 Lm and
Lm ¼ K Lcalc1 Lcalc2 . Lm is the mutual inductance between Lcalc-1 and Lcalc-2.
A01 A02
Micro-strip
B02 B02
w1 g1 RFID IC w2 g2
Fig. A.4 RFID rectangular spiral antennas system

Fig. A.5 Double rectangular spiral antennas in series with RFID TAG IC equivalent circuit
K is the coupling coefficient of two inductors 0 K 1. We do our analysis for

K = 0.3, 0.5, 0.7. We consider for simplicity that two rectangular spiral antennas
are identical (same parameter values).
Part A: Analysis for l = l0 and l = lr l0 (lr = 450).
H H l
l0 ¼ 4 p 107 1:2566 106 ; lr ¼
m m l0
) l ¼ lr l0 jlr ¼450 ¼ 450 1:2566 106 ¼ 565:47 106
ðt þ wÞ ð20lm þ 20lmÞ
t ¼ g ¼ 20lm ) d ¼ 2 ¼2 ¼ 25:47lm; t ¼ 20 lm
P 3:14
d—Equivalent diameter of the track (Fig. A.6)
Aavg ¼ A0 Nc ðg þ wÞ ¼ 0:005 Nc ð20 lm þ 20 lmÞ ¼ 0:005 Nc 40 lm ¼ 0:005 Nc 40 106

Bavg ¼ A0 Nc ðg þ wÞ ¼ 0:005 Nc ð20 lm þ 20 lmÞ ¼ 0:005 Nc 40 lm ¼ 0:005 Nc 40 106


the coil. t—Track thickness, w—Track width, g—Gap between tracks. Nc—
Number of turns, d—Equivalent diameter of the track. Average coil area;
Ac = Aavg Bavg. Integrating all those parameters give the equations for induc-
tance calculation:
2 3 2 3
6 2 Aavg Bavg 7 6 2 Aavg Bavg 7
X1 ¼ Aavg ln4 qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi5; X2 ¼ Bavg ln4 qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi5
d ðAavg þ A2avg þ B2avg d ðBavg þ A2avg þ B2avg
h qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffii ðAavg þ Bavg Þ
X3 ¼ 2 Aavg þ Bavg A2avg þ B2avg ; X4 ¼
4

hl i
Lcalc ¼ 0
ðX1 þ X2 X3 þ X4 Þ Ncp ; L1 ¼ Lcalc
p

inductance calculation (Table A.4).
We integrate the Lcalc value inside the differential equations which characterize
the RFID system with the Coil inductance. Nc ! z; A0 ! x; B0 ! y
MATLABScript: z = 10; x = 0.005 – z * 40 * (1e−6); y = 0.005 – z * 40 *
(1e−6); x1 = x * log(2 * x * y/(25.47 * (1e−6) * (x + sqrt(x * x + y * y))));
x2 = y * log(2 * x * y/(25.47 * (1e−6) * (y + sqrt(x * x + y * y)))); x3 = 2 *
(x + y − sqrt(x * x + y * y)); x4 = (x + y)/4; l = ((1.2566 * (1e−6)/3.14) *
(x1 + x2 − x3 + x4) * power(z,1.8))
The results in Tables A.5, A.6, A.7, A.8, A.9, A.10, A.11, A.12 and A.13).
Rectangular spiral RFID antenna length calculation & resistance
calculate the total length (Fig. A.7).
Table A.4 RFID coil manufacturing technology (wired, etched, printed)

Wired 1.8–1.9
Etched 1.75–1.85
Printed 1.7–1.8
Table A.5 RFID rectangular spiral Lcalc for different values of Nc and l and tag overall dimension
Nc l = l0; lr = 1 Lcalc (2 mm 2 mm) − [H] Lcalc (3 mm 3 mm) − [H] Lcalc (4 mm 4 mm) − [H]
10 l0 = 1.2566 (1e−6)H/m 2.922 10−7 5.3857 10−7 8.048 10−7
20 l0 7.0241 10−7 1.5222 10−6 2.425 10−6
30 l0 8.532 10−7 2.45 10−6 4.266 10−6
60 l0 −5.916 10−7 2.009 10−6 7.351 10−6
−7 −6
100 l0 −1.767 10 −6.6295 10 NaN
150 l0 −9.167 10−5 −6.3052 10−5 −3.6674 10−5
120 l0 −3.8527 10−5 −2.1249 10−5 −6.573 10−5
170 l0 −1.45 10−4 −1.074 10−4 −7.2119 10−5
200 l0 −2.577 10−4 −2.047 10−4 −1.538 10−4
220 l0 −3.58 10−4 −2.932 10−4 −2.3072 10−4
250 l0 −5.5179 10−4 −4.673 10−4 −3.8522 10−4
Nc l = lr l0; lr = 450 Lcalc (2 mm 2 mm) − [H] Lcalc (3 mm 3) − [H] Lcalc (4 mm 4 mm) − [H]
10 565.47 (1e−6)H/m 1.3149 10−4 2.4236 10−4 3.621 10−4
20 565.47 (1e−6)H/m 3.1608 10−4 6.85 10−4 0.0011
30 565.47 (1e−6)H/m 3.8394 10−4 0.0011 0.0019
60 565.47 (1e−6)H/m −2.6625 10−4 9.04 10−4 0.0033
100 565.47 (1e−6)H/m −0.008 −0.003 NaN
150 565.47 (1e−6)H/m −0.0413 −0.0284 −0.0165
120 565.47 (1e−6)H/m −0.0173 −0.0096 −0.003
170 565.47 (1e−6)H/m −0.0653 −0.0484 −0.0325
200 565.47 (1e−6)H/m −0.116 −0.0921 −0.0692
220 565.47 (1e−6)H/m −0.1611 −0.132 −0.1038
250 565.47 (1e−6)H/m −0.2483 −0.2103 −0.1734
Table A.6 RFID rectangular spiral Lcalc, Lm, and LT for K = 0.3 (2 mm 2 mm)
Nc Lcalc-1 = Lcalc-2 = Lcalc, Lm ¼ K Lcalc1 Lcalc2 LT = Lcalc-1 + Lcalc-2
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
l = lr l0 lr = 450 Lm ¼ K Lcalc Lcalc ¼ K Lcalc + 2 Lm = 2 Lcalc + 2 Lm
(2 mm 2 mm) − [H] (K = 0.3) (K = 0.3)
10 1.3149 10−4 3.9447 10−5 3.4187 10−4
−4 −5
20 3.1608 10 9.4824 10 8.218 10−4
−4 −4
30 3.8394 10 1.1518 10 9.9824 10−4
Nc Lcalc-1 = Lcalc-2 Lm ¼ K Lcalc1 Lcalc2 LT = Lcalc-1 + Lcalc-2
= Lcalc, l = lr l0 Lm ¼ K Lcalc Lcalc ¼ K Lcalc + 2 Lm = 2 Lcalc + 2 Lm
lr = 450 (K = 0.5) (K = 0.5)
(2 mm 2 mm) − [H]
10 1.3149 10−4 6.574 10−5 3.9447 10−4
20 3.1608 10 10−4 1.58 10−4 9.482 10−4
−4 −4
30 3.8394 10 1.9197 10 0.0012
(2 mm 2 mm) − [H] (K = 0.7) (K = 0.7)
10 1.3149 10−4 9.2043 10−5 4.47 10−4
20 3.1608 10−4 2.212 10−4 0.0011
30 3.8394 10−4 2.6876 10−4 0.0013
lr = 450 (K = 0.3) (K = 0.3)
(3 mm 3 mm) − [H]
10 2.4236 10−4 7.2708 10−5 6.3014 10−4
−4 −4
20 6.85 10 2.055 10 0.0018
30 0.0011 0.0033 0.0029 = 2.9 mH
60 9.04 10−4 2.712 10−4 0.0024 = 2.4 mH
= Lcalc, l= lr l0 Lm ¼ K Lcalc Lcalc ¼ K Lcalc + 2 Lm = 2 Lcalc + 2 Lm
lr = 450 (K = 0.5) (K = 0.5)
(3 mm 3 mm) − [H]
10 2.4236 10−4 1.2118 10−4 7.27 10−4
−4 −4
20 6.85 10 3.425 10 0.0021 = 2.1 mH
30 0.0011 5.5 10−4 0.0033
−4 −4
60 9.04 10 4.52 10 0.0027 = 2.7 mH

the coil. w—Track width, g—Gap between tracks. Nc—Number of turns. L0 is the
length of the first turn L0 ¼ 2 ðA0 þ B0 Þ ðw þ gÞ. Lk is the length of turn k + 1.
lr = 450 (K = 0.7) (K = 0.7)
(3 mm 3 mm) − [H]
10 2.4236 10−4 1.6965 10−4 8.24 10−4
20 6.85 10−4 4.795 10−4 0.0023 = 2.3 mH
−4
30 0.0011 7.7 10 0.0037
60 9.04 10−4 6.328 10−4 0.0031
Table A.12 RFID rectangular spiral Lcalc, Lm, and LT for K = 0.3, 0.5 (4 mm 4 mm)
lr = 450 (K = 0.3) (K = 0.3)
(4 mm 4 mm) − [H]
10 3.621 10−4 1.0863 10−4 9.4146 10−4
20 0.0011 0.0033 0.0029 = 2.9 mH
30 0.0019 5.7 10−4 0.0049
60 0.0033 9.9 10−4 0.0086
(4 mm 4 mm) − [H] (K = 0.5) (K = 0.5)
10 3.621 10−4 1.8105 10−4 0.0011
20 0.0011 5.5 10−4 0.0033
30 0.0019 9.5 10−4 0.0057
60 0.0033 0.0017 0.0099
lr = 450 (K = 0.7) (K = 0.7)
(4 mm 4 mm) − [H]
10 3.621 10−4 2.5347 10−4 0.0012
20 0.0011 7.7 10−4 0.0037
30 0.0019 0.0013 0.0065
60 0.0033 0.0023 0.0112
X
N c 1
k¼1
X
N c 1
k¼1
NX
C 1
k¼1
NX
C 1
¼2 fðA0 þ B0 Þ 4 k ðw þ gÞg ¼ 2 ðA0 þ B0 Þ ðNC 1Þ

k¼1
NX
C 1
2 ½4 k ðw þ gÞ
k¼1
NX
C 1 NX
C 1
k¼1 k¼1
X
N c 1
k¼1
LT ¼ L0 þ 2 ðNC 1Þ ½A0 þ B0 4 ðw þ gÞ ¼ L0 þ 2 ðA0 þ B0 Þ

ðw þ gÞ þ 2 ðNC 1Þ
½A0 þ B0 4 ðw þ gÞ ¼ 2 ðA0 þ B0 Þ ðw þ gÞ
þ 2 ðNC 1Þ ðA0 þ B0 Þ 8 ðNC 1Þ ðw þ gÞ
¼ L0 þ 2 ðA0 þ B0 Þ NC ðw þ gÞ ð8 NC 7Þ
Final result:
LT ¼ L0 þ 2 ðA0 þ B0 Þ NC ðw þ gÞ ð8 NC 7Þ
LT LT
RDC ¼ ¼
r S r p a2

RDC ¼ ¼ ¼
r S r p a2 r p a2
A0 þ B0 ¼ TBD; NC ¼ TBD; w þ g ¼ 40 106 ) LT ¼ TBD
To be define—value which need to be chosen according analysis results.

Cross section area
S ¼ 20 lm 20 lm ¼ 400 1012 m2

Part B: Analysis for l = lr l0 (lr = 100, 200, 300) (Tables A.14, A.15, A.16,
A.17, A.18, A.19, A.20, A.21, A.22, A.23, A.24, A.25, A.26, A.27, A.28, A.29,
A.30, A.31, A.32, A.33, A.34 and A.35).
LT LT
RDC ¼ ¼
r S r p a2

RDC ¼ ¼ ¼
r S r p a2 r p a2
A0 þ B0 ¼ TBD; NC ¼ TBD; w þ g ¼ 40 106 ) LT ¼ TBD
Cross section area
S ¼ 20 lm 20 lm ¼ 400 1012 m2
Table A.14 RFID rectangular spiral Lcalc for different values of Nc and l and tag overall
932
Nc l = lr l0; lr = 100 l = 125.66 (1e-6)H/m Lcalc (2 mm 2 mm) − [H] Lcalc (3 mm 3 mm) − [H] Lcalc (4 mm 4 mm) − [H]
10 125.66 (1e−6)H/m 2.922 10−5 5.3857 10−5 8.0487 10−5
20 125.66 (1e−6)H/m 7.0241 10−5 1.5222 10−4 2.425 10−4
−5 −4
30 125.66 (1e−6)H/m 8.532 10 2.4523 10 4.266 10−4
−5 −4
60 125.66 (1e−6)H/m −5.9168 10 2.009 10 7.351 10−4
−4
100 125.66 (1e−6)H/m −0.0018 −6.629 10 NaN
150 125.66 (1e−6)H/m −0.0092 −0.0063 −0.0037
120 125.66 (1e−6)H/m −0.0039 −0.0021 −6.573 10−4
170 125.66 (1e−6)H/m −0.0145 −0.0107 −0.0072
200 125.66 (1e−6)H/m −0.0258 −0.0205 −0.0154
220 125.66 (1e−6)H/m −0.0358 −0.0293 −0.0231
250 125.66 (1e−6)H/m −0.0552 −0.0467 −0.0385
Nc l = lr l0; lr = 200 l = 251.32 (1e−6)H/m Lcalc (2 mm 2 mm) − [H] Lcalc (3 mm 3 mm) − [H] Lcalc (4 mm 4 mm) − [H]
−5 −4
10 251.32 (1e−6)H/m 5.8439 10 1.0771 10 1.6097 10−4
−4 −4
20 251.32 (1e−6)H/m 1.4048 10 3.0445 10 4.8501 10−4
−4 −4
30 251.32 (1e−6)H/m 1.7064 10 4.9045 10 8.532 10−4
−4 −4
60 251.32 (1e−6)H/m −1.1834 10 4.0179 10 0.0015
100 251.32 (1e−6)H/m −0.0035 −0.0013 NaN
150 251.32 (1e−6)H/m −0.0183 −0.0126 −0.0073
120 251.32 (1e−6)H/m −0.0077 −0.0042 −0.0013
170 251.32 (1e−6)H/m −0.029 −0.0215 −0.0144
200 251.32 (1e−6)H/m −0.0516 −0.0409 −0.0308
220 251.32 (1e−6)H/m −0.0716 −0.0587 −0.0461
250 251.32 (1e−6)H/m −0.1104 −0.0935 −0.077
(continued)
Nc l = lr l0; lr = 300 Lcalc (2 mm 2 mm) − [H] Lcalc (3 mm 3 mm) − [H] Lcalc (4 mm 4 mm) − [H]
l = 376.98 (1e−6)H/m
10 376.98 (1e−6)H/m 8.7659 10−5 1.6157 10−4 2.4146 10−4
−4 −4
20 376.98 (1e−6)H/m 2.1072 10 4.5667 10 7.2751 10−4
−4 −4
30 376.98 (1e−6)H/m 2.5596 10 7.3568 10 0.0013
60 376.98 (1e−6)H/m −1.775 10−4 6.0269 10−4 0.0022
100 376.98 (1e−6)H/m −0.0053 −0.002 NaN
150 376.98 (1e−6)H/m −0.0275 −0.0189 −0.011
120 376.98 (1e−6)H/m −0.0116 −0.0064 −0.002
170 376.98 (1e−6)H/m −0.0435 −0.0322 −0.0216
200 376.98 (1e−6)H/m −0.0773 −0.0614 −0.0462
220 376.98 (1e−6)H/m −0.1074 −0.088 −0.0692
250 376.98 (1e−6)H/m −0.1655 −0.1402 −0.1156
933
lr = 100 (K = 0.3) (K = 0.3)
(2 mm 2 mm) − [H]
10 2.922 10−5 8.766 10−6 7.5972 10−5
20 7.0241 10−5 2.1072 10−5 1.8263 10−4
−5 −5
30 8.532 10 2.5596 10 2.2183 10−4
lr = 100 (K = 0.5) (K = 0.5)
(2 mm 2 mm) − [H]
10 2.922 10−5 1.461 10−5 8.766 10−5
−5 −5
20 7.0241 10 3.5121 10 2.1072 10−4
−5 −5
30 8.532 10 4.266 10 2.5596 10−4
lr = 100 (K = 0.7) (K = 0.7)
(2 mm 2 mm) − [H]
10 2.922 10−5 2.0454 10−5 9.9348 10−5
−5 −5
20 7.0241 10 4.9169 10 2.3882 10−4
−5 −5
30 8.532 10 5.9724 10 2.9009 10−4
lr = 100 (K = 0.3) (K = 0.3)
(3 mm 3 mm) − [H]
10 5.3857 10−5 1.6157 10−5 1.4003 10−4
−4 −5
20 1.5222 10 4.5666 10 3.9577 10−4
−4 −5
30 2.4523 10 7.3569 10 6.376 10−4
−4 −5
60 2.009 10 6.027 10 5.2234 10−4
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
Nc Lcalc-1 = Lcalc-2 = Lcalc, Lm ¼ K Lcalc1 Lcalc2 LT = Lcalc-1 + Lcalc-2 + 2 Lm
l = lr l0 Lm ¼ K Lcalc Lcalc ¼ K Lcalc (K = 0.5) = 2 Lcalc + 2 Lm (K = 0.5)
lr = 100 (3 mm 3 mm) − [H]
10 5.3857 10−5 2.6929 10−5 1.6157 10−4
20 1.5222 10−4 7.611 10−5 4.5666 10−4
−4 −4
30 2.4523 10 1.2262 10 7.35 10−4
−4
60 2.009 10 1.0045 6.027 10−4
Table A.20 RFID rectangular spiral Lcalc, Lm, and LT for K = 0.7 (3 mm 3 mm) and K = 0.3,
0.5 (4 mm 4 mm)
Nc Lcalc-1 = Lcalc-2 = Lcalc, Lm ¼ K Lcalc1 Lcalc2 LT = Lcalc-1 + Lcalc-2 + 2 Lm =
l = lr l0 Lm ¼ K Lcalc Lcalc ¼ K Lcalc 2 Lcalc + 2 Lm (K = 0.7)
lr = 100 (K = 0.7)
(3 mm 3 mm) − [H]
10 5.3857 10−5 3.77 10−5 1.8311 10−4
−4 −4
20 1.5222 10 1.0655 10 5.1755 10−4
−4 −4
30 2.4523 10 1.7166 10 8.3378 10−4
−4 −4
60 2.009 10 1.4063 10 6.8306 10−4
l = lr l0 lr = 100 Lm ¼ K Lcalc Lcalc ¼ K Lcalc 2 Lcalc + 2 Lm (K = 0.3)
(4 mm 4 mm) − [H] (K = 0.3)
10 8.0487 10−5 2.4146 10−5 2.0927 10−4
−4 −5
20 2.425 10 7.275 10 6.305 10−4
−4 −4
30 4.266 10 1.2798 10 0.0011
60 7.351 10−4 2.2053 10−4 0.0019
(4 mm 4 mm) − [H] − [H] (K = 0.5)
10 8.0487 10−5 4.0244 10−5 2.4146 10−4
−4 −4
20 2.425 10 1.2125 10 7.275 10−4
−4 −4
30 4.266 10 2.133 10 0.0013
60 7.351 10−4 3.6755 10−4 0.0022 = 2.2 mH
Table A.21 RFID rectangular spiral Lcalc, Lm, and LT for K = 0.7 (4 mm 4 mm) and K = 0.3,
0.5 (2 mm 2 mm)
Nc Lcalc-1 = Lcalc-2 = Lcalc, l = lr l0 Lm ¼ K Lcalc1 Lcalc2 LT = Lcalc-1 + Lcalc-2
lr = 100 (4 mm 4 mm) − [H] Lm ¼ K Lcalc Lcalc ¼ K Lcalc + 2 Lm = 2 Lcalc +
(K = 0.7) 2 Lm (K = 0.7)
10 8.0487 10−5 5.6341 10−5 2.7366 10−4
−4 −4
20 2.425 10 1.6975 10 8.245 10−4
−4 −4
30 4.266 10 2.9862 10 0.0015
60 7.351 10−4 5.1457 10−4 0.0025 = 2.5 mH
Nc Lcalc-1 = Lcalc-2 = Lcalc, l = lr l0 Lm ¼ K Lcalc1 Lcalc2 LT = Lcalc-1
lr = 200 (2 mm 2 mm) − [H] Lm ¼ K Lcalc Lcalc ¼ K Lcalc + Lcalc-2 + 2 Lm =
(K = 0.3) 2 Lcalc + 2 Lm (K = 0.3)
10 5.8439 10−5 1.7532 10−5 1.5194 10−4
(continued)

Nc Lcalc-1 = Lcalc-2 = Lcalc, l = lr l0 Lm ¼ K Lcalc1 Lcalc2 LT = Lcalc-1
lr = 200 (2 mm 2 mm) − [H] Lm ¼ K Lcalc Lcalc ¼ K Lcalc + Lcalc-2 + 2 Lm =
(K = 0.3) 2 Lcalc + 2 Lm (K = 0.3)
20 1.4048 10−4 4.2144 10−5 3.6525 10−4
−4 −5
30 1.7064 10 5.1192 10 4.4366 10−4
Nc Lcalc-1 = Lcalc-2 = Lcalc, l = lr l0 Lm ¼ K Lcalc1 Lcalc2 LT = Lcalc-1 + Lcalc-2
lr = 200 (2 mm 2 mm) − [H] Lm ¼ K Lcalc Lcalc ¼ K Lcalc + 2 Lm = 2 Lcalc + 2 Lm
(K = 0.5) (K = 0.5)
10 5.8439 10−5 2.922 10−5 1.7532 10−4
20 1.4048 10−4 7.024 10−5 4.2144 10−4
30 1.7064 10−4 8.532 10−5 5.1192 10−4
(2 mm 2 mm) − [H] (K = 0.7)
10 5.8439 10−5 4.0907 10−5 1.9869 10−4
20 1.4048 10−4 9.8336 10−5 4.7763 10−4
−4 −5
30 1.7064 10 1.1945 10 5.8018 10−4
(3 mm 3 mm) − [H] (K = 0.3)
10 1.0771 10−4 3.2313 10−5 2.8005 10−4
−4 −5
20 3.0445 10 9.1335 10 7.9157 10−4
−4 −4
30 4.9045 10 1.4714 10 0.0013
60 4.0179 10−4 1.2054 10−4 0.001
(3 mm 3 mm) − [H] (K = 0.5)
10 1.0771 10−4 5.3855 10−5 3.2313 10−4
−4 −4
20 3.0445 10 1.5223 10 9.1335 10−4
−4 −4
30 4.9045 10 2.4523 10 0.0015
60 4.0179 10−4 2.009 10−4 0.0012
Table A.25 RFID rectangular spiral Lcalc, Lm, and LT for K = 0.7 (3 mm 3 mm), K = 0.3, 0.5
(4 mm 4 mm)
Nc Lcalc-1 = Lcalc-2 = Lcalc, l = lr l0 Lm ¼ K Lcalc1 Lcalc2 LT = Lcalc-1 + Lcalc-2 + 2 Lm =
lr = 200 (3 mm 3 mm) − [H] Lm ¼ K Lcalc Lcalc ¼ K Lcalc 2 Lcalc + 2 Lm (K = 0.7)
(K = 0.7)
10 1.0771 10−4 7.5397 10−5 3.6621 10−4
−4 −4
20 3.0445 10 2.1312 10 0.001
30 4.9045 10−4 3.4332 10−4 0.0017
60 4.0179 10−4 2.8125 10−4 0.0014
(4 mm 4 mm) − [H] (K = 0.3)
10 1.6097 10−4 4.8291 10−5 4.1852 10−4
−4 −4
20 4.8501 10 1.455 10 0.0013
30 8.532 10−4 2.5596 10−4 0.0022 = 2.2 mH
60 0.0015 4.5 10−4 0.0039
(4 mm 4 mm) − [H] (K = 0.5)
10 1.6097 10−4 8.0485 10−5 4.8291 10−4
−4 −4
20 4.8501 10 2.425 10 0.0015
30 8.532 10−4 4.266 10−4 0.0026 = 2.6 mH
60 0.0015 7.5 10−4 0.0045
(4 mm 4 mm) − [H] (K = 0.7) (K = 0.7)
10 1.6097 10−4 1.1268 10−4 5.473 10−4
−4 −4
20 4.8501 10 3.3951 10 0.0016
30 8.532 10−4 5.9724 10−4 0.0029 = 2.9 mH
60 0.0015 0.001 0.0051
(2 mm 2 mm) − [H] (K = 0.3) (K = 0.3)
10 8.7659 10−5 2.6298 10−5 2.2791 10−4
−4 −5
20 2.1072 10 6.3216 10 5.4787 10−4
30 2.5596 10−4 7.6788 10−5 6.655 10−4
(2 mm 2 mm) − [H] (K = 0.5) (K = 0.5)
10 8.7659 10−5 4.383 10−5 2.6298 10−4
−4 −4
20 2.1072 10 1.0536 10 6.3216 10−4
−4 −4
30 2.5596 10 1.2798 10 7.6788 10−4
(2 mm 2 mm) − [H] (K = 0.7) (K = 0.7)
10 8.7659 10−5 6.1361 10−5 2.9804 10−4
−4 −4
20 2.1072 10 1.475 10 7.1645 10−4
−4 −4
30 2.5596 10 1.7917 10 8.7026 10−4
(3 mm 3 mm) − [H] (K = 0.3) (K = 0.3)
10 1.6157 10−4 4.84 10−5 4.2008 10−4
−4 −4
20 4.5667 10 1.37 10 0.0012
30 7.3568 10−4 2.207 10−4 0.0019
60 6.0269 10−4 1.8081 10−4 0.0016
(3 mm 3 mm) − [H] (K = 0.5) (K = 0.5)
10 1.6157 10−4 8.0785 10−5 4.847 10−4
−4 −4
20 4.5667 10 2.2834 10 0.0014
30 7.3568 10−4 3.6784 10−4 0.0022 = 2.2 mH
60 6.0269 10−4 3.0135 10−4 0.0018
(3 mm 3 mm) − [H] (K = 0.7) (K = 0.7)
10 1.6157 10−4 1.131 10−4 5.4934 10−4
−4 −4
20 4.5667 10 3.1967 10 0.0016
30 7.3568 10−4 5.1498 10−4 0.0025 = 2.5 mH
60 6.0269 10−4 4.2188 10−4 0.002
(4 mm 4 mm) − [H] (K = 0.3) (K = 0.3)
10 2.4146 10−4 7.2438 10−5 6.278 10−4
−4 −4
20 7.2751 10 2.1825 10 0.0019
30 0.0013 3.9 10−4 0.0034
60 0.0022 6.6 10−4 0.0057
l = lr l0 lr = 300 Lm ¼ K Lcalc Lcalc ¼ K Lcalc + 2 Lm = 2 Lcalc +
(4 mm 4 mm) − [H] (K = 0.5) 2 Lm (K = 0.5)
10 2.4146 10−4 1.2073 10−4 7.2438 10−4
−4 −4
20 7.2751 10 3.6376 10 0.0022 = 2.2 mH
30 0.0013 6.5 10−4 0.0039
60 0.0022 0.0011 0.0066
Nc Lcalc-1 = Lcalc-2 = Lcalc, Lm ¼ K Lcalc1 Lcalc2 LT=Lcalc-1 + Lcalc-2
(4 mm 4 mm) − [H] (K = 0.7) (K = 0.7)
10 2.4146 10−4 1.6902 10−4 8.2096 10−4
−4 −4
20 7.2751 10 5.0926 10 0.0025 = 2.5 mH
30 0.0013 9.1 10−4 0.0044
60 0.0022 0.0015 0.0075

Appendix B
RF Amplifiers Basic and Advance Topics
and Design Methods
B.1 Amplifier Design Concepts and Matching Guidelines
An amplifier is an active device that has the ability to amplify voltage, current and
amplify both voltage and current. There are some types of amplifiers. Amplifiers
types: zero frequency amplifiers (DC amplifiers), low frequency amplifiers (Audio
amplifiers), and high frequency amplifiers (RF amplifiers). Power is P = V I,
when current (I) or voltage (V) is raising and will create amplification. The main
desirable specification when designing amplifier: High P1dB (high input power that
cause the UUT gain to drop by 1dB from small signal value). P1dB (IP1dB,
OP1dB), low noise (example LNA), and maximum efficiency (Efficiency (P.E
%) = [output signal power]/[power supply power]). High Gain (S21), b2 is output
signal power, a1 is input signal power S21 = b2/a1 for a2 = 0 (no input signal at
amplifier output). Good return loss (RL). S11 (IRL) ! e and (ORL) ! 1. Return
Loss = Reflection Loss. Amplifiers come in three flavors: Common Base
(CB) amplifiers, Common Collector (CC) amplifiers, and Common Emitter (CE)
amplifiers. It depends whether the base, collector or emitter is common to both the
input and output of the amplifier. Common Base (CB) amplifier, input signal
inserted at emitter (E) and output signal taken from the collector (C). The CB
amplifier can operate as a voltage amplifier for low input impedance circuits.
Fig. B.1 Common base (CB)

amplifier Rfin (E) Rfout (C)
CB Amplifier
(B)
VIN (t) IE VRE

0 degree phase shift

DOI 10.1007/978-3-319-45427-6
942 Appendix B: RF Amplifiers Basic and Advance Topics and Design Methods
Cin, Cout are DC

block capacitors
Fig. B.2 Common base (CB) amplifier schematic
CB amplifier can be found at the 50 X antenna input of the radio receiver (Figs. B.1
and B.2).
There is a JFET’s CB amplifier circuit which can be used in receiver’s IF unit.
C2, C3, R2 and RFC (RF choke) are for decoupling. C4, C6 are RF decoupling. C5 is
for flatter frequency response throughout its pass band. T1 is for impedance
matching (Fig. B.3).
Fig. B.3 jfet common gate amplifier

Appendix B: RF Amplifiers Basic and Advance Topics and Design Methods 943
There is a JFET’s CB amplifier circuit which can be used in receiver’s IF unit.

C2, C3, R2 and RFC (RF choke) are for decoupling. C5 is for flatter frequency
response throughout its pass band. T1 is for impedance matching. The most popular
amplifier circuit is Common Emitter (CE). The CE amplifier has a greater current
gain and voltage gain combination than any other type. CE amplifier make excellent
Rfin (B) Rfout (C)

CE Amplifier
(E)
Fig. B.4 Common Emitter (CE) amplifier
Fig. B.5 Common emitter (CE) amplifier schematic

Fig. B.6 Input to output Input to

180 deg & 0GHz
phase (deg) versus frequency Output
(GHz) phase 6 GHz & 0 deg
(deg)
Frequency (GHz)
power amplifier. The Common Emitter (CE) amplifier, input signal inserted at base
(B) and output signal taken from the collector (C). The CE amplifier’s output
voltage is shifted by *180° in phase compared to CE amplifier’s input signal
(Figs. B.4 and B.5).
At RF frequencies there is an effect of “positive feedback”, which creates
amplifier instability and oscillation. “positive feedback” is when there is internal
feedback capacitance between transistor’s collector and its base. It cause to unde-
sired CE oscillations. Inter feedback capacitance can be as high as 25 pF or higher.
At a specific frequency this capacitance will send an in phase signal back into the
base input from the collector’s output. This back in phase signal creates “oscillator”.
Transistor’s internal resistance and capacitance along with other phase delays yield a
powerful phase shift to normally out of phase 180° feedback signal. Only phase
delays that are at a total 360° (0°) will bring amplifier instability and oscillations.
A phase = g(frequency) for a typical CE amplifier unmatched (Fig. B.6).
The Common Collector (CC) amplifier (emitter follower amplifier) has the input
signal inserted into the base, and output signal from emitter. The CC amplifier has
current and power gain, voltage gain less than one (GV < 1). The CC amplifier’s
used as a buffer or active impedance matching circuit. The CC amplifier has high
input impedance and low output impedance. There is no phase inversion between
CC amplifier’s input and output (Figs. B.7 and B.8).
Rfin (B) Rfout (E)

CC Amplifier
(C)
VIN (t) IE VRE

Fig. B.7 Common collector (CC) amplifier
VIN (t) IE VRE

Fig. B.8 Common Collector (CC) amplifier schematic
When an amplifier’s output impedance matches the load impedance, maximum

power is transferred to the load (RS = RL). When an amplifier’s output impedance
matches the load impedance (RS = RL) all reflections are eliminated. When an
amplifier’s output impedance (RS) unmatched the load impedance (RL), there are
reflections and less than maximum power is transferred to the load RL. The
amplifier’s efficiency in its DC case is defined by (Fig. B.9 and Table B.1).

Rload
EFFð%Þ ¼ 100 %:
Rload þ Rsource
2 3
2 3 Point of max ðDCÞ
Point of max 6 amplifier efficiency 7
6 7
4 Power transfer 5 6¼ 6 EFFð%Þ ! 100 % 7
6 7
RL ¼ RS 4 RL RS ðRS !eÞ 5
RL ! 1
In order to develop maximum power the Zout of the amplifier must be complex
conjugate of the Zin of the load. The low efficiency (EFF%) level can be increased if
the load has a higher input resistance, thus dropping more power across the load
Fig. B.9 Amplifier’s output impedance matches the load impedance
Table B.1 Amplifier’s conditions and power across the load

Condition Voltage or current on Power across the load Current or voltage drawn
the load from Amplifier’s power
supply
RL ! e VOUT ¼ VL ! e pL ¼ 0 Load DC current = Vcc/Rs
RL ¼ 0 VL ¼ 0 POUT ¼ PL ! e (maximum DC power)
(Short
load)
RL ! ∞ IL ! e pL = 0 Load DC voltage = Vcc
(No load) IL ¼ 0 but lload = 0
RL = RS
IL ¼
Vcc
j POUT ¼ PL ! e Load DC voltage = Vcc/2
(max RS þ RL RS ¼RL V2cc
Load DC current = Vcc/
power Vcc Vcc PL ¼ I2L RL ¼ (Rsource + Rload)
transfer) ¼ ¼ 4 RL
2 RS 2 RL
(Zload Zout). The total output power across load will be less in this condition
(Zload Zout) than if Zload = Zout (pure resistive). The transfer of maximum power
from the source to the load will not maximize efficiency (EEF%). Maximum power
transfer only occurs when the source impedance equals the load impedance
Zload = Zout (pure resistive). Any impedance mismatches will end in a loss of
power, Mismatch Loss (ML). ML(dB) = Mismatch Loss. VSWR = Voltage Stand
Fig. B.10 Amplifier’s Zout

and load
VSWR1 2
½1ðVSWR þ 1Þ
Wave Ratio (dimensionless units). ML ¼ 10 log10 ; Mismatch Loss
(ML) (Fig. B.10).
2 n o 32
1 þ jC j
1
41ðn 1jCj
o Þ5

1 þ jCj
½ VSWR1 2
1ðVSWR þ 1Þ
1jCj
þ1
ML ¼ 10 log10 ¼ 10 log10
1 þ jC j
VSWR¼ 1jCj
2 n o 32
2jCj
41ðn 1jCj
oÞ5
2
1jCj ½1jCj2
ML = 10 log10 ¼ ML ¼ 10 log10
Good amplifier match (C = 0), no reflections then VSWR = 1; ML = 10 log

(1) = 0. Bad amplifier match (C = 1), all power is reflected then VSWR ! ∞;
10 log(0) = −∞.
Amplifier matching allows: amplifier maximum power transfer and attenuation
of harmonics to be achieved between stages. RF matching network able and
facilitate impedance matching and filtering of signal, coupling between RF stages.
The amplifier matching networks types are L matching network, T matching net-
work, and PI matching network. In design of microwave matching network the
device parasitic and the effect of length on RF circuit matching are very important.
L matching network: LC matching topology which especially for narrowband
impedance matching. The L network name is due to its L shape and can furnish low
pass filtering to decrease harmonic output. Two stages (amplifier, and load) with no
matching network. Two stages (amplifier, load) with simple low pass L network
between stages. Low pass L network can matches a higher output impedance source
(Zhigh) to a lower input impedance load (Zlow). Low pass L network can match also
a lower output impedance source (Zlow) to a higher input impedance load (Zhigh)
(Fig. B.11).
Fig. B.11 Amplifier (source) Poor

and load poor match Amplifier
(source)
Load
Zlow Zhigh
Perfect
match
Fig. B.12 High to low impedance matching L network
Figure B.12 is a high to low (Zhigh ! Zlow) impedance matching L network

between two amplifiers.
Figure B.13 is a low to high (Zlow ! Zhigh) impedance matching L network
between two amplifiers.
A T matching network is a popular impedance matching network circuit. It can
furnish almost any impedance matching level between two stages and we can
selectable loaded Q (Fig. B.14).
A PI matching network can be applicable to many matching applications all
types. We can alter the ratio between capacitors C1 and C2 in the next figure so the
output impedance of the load can be matched to the source impedance. PI network
also decreasing the harmonic output (Figs. B.15 and B.16).
PI network’s topology is a low pass filter. It has a small resonant band pass like
S21 gain pick and excellent return loss S11 at specific frequency (Fig. B.17).
It is very important in any microwave matching circuit first that device parasitic
are part of any active or passive component and second the effect that length has a
Perfect
match
Fig. B.13 Low to high impedance matching L network

Fig. B.14 Impedance matching T network
Fig. B.15 PI impedance matching network
Fig. B.16 PI impedance matching network

Low pass
response
Resonant
Peak (S21)
Fig. B.17 PI network’s topology as a low pass filter
RF circuit matching. There are vital components parasitic effects. Effect that length:
PCB trace and pad reactance (j X elements), and strong influence of circuit length.
In low frequency we can ignore effect of length. In low frequency we not consider
PCB parasitic effects and distance between each matching component and
source/load. At higher RF frequencies the distance between components and cir-
cuits is very critical. The distance affects the moving wave’s wavelengths and the
expected performance of the matching network. It is very important where the
circuit and individual components are located on PCB. Any mismatched line that is
a significant portion of a wavelength, the impedance will vary along that line. The
variations on a mismatched line are due to the standing wave (SWR). The reflected
RF wave is bouncing off of the mismatch load and interacting with forward wave.
The interaction between reflected and forward waves, creating fixed peaks and
valley of voltage and current. The fixed peaks and valleys of voltage and current are
created at every half wavelength (k/2) along the trace. The distance between each
circuit element is varied; this distance will completely destroy any predicted RF
match. Figure B.18 describes the voltage and current standing waves on a mis-
matched transmission line.
First the match is calculated with zero micro-strip length and second the cal-
culated match will degrade with micro-strip length. We must take into consideration
the micro-strips length effects in our impedance matching calculation (Fig. B.19).
Fig. B.18 Current and volt- Amplitude Current Voltage

age signal amplitudes versus
wavelength
Wavelength
λ/2
Fig. B.19 Amplifier and load, matching network with microstrips
B.2 Amplifier Distortion and Noise Products
There are two unavoidable and undesirable elements of any electronic circuits: dis-
tortion and noise. Distortion can deform the carrier and its sidebands at the transmitter
and receiver. Distortion causes to spectral regrowth and adjacent channel interface.
Additionally distortion causes faulty, distorted replica of the original baseband sig-
nal, and increasing the Bit Error Rate (BER). Noise degrades all important BER of
the entire system. Distortion forms frequencies inter modulation products. Distortion
frequency inter-modulation is cause by internal nonlinear mixing of any signal with
one or more other signals, mixing with other signals, and modulated or unmodulated
waveform that is altered is shape or amplitude from the original signal (improper
circuit response). The distortion types are frequency distortion, amplitude and phase
distortion, inter-modulation distortion, second order inter modulation distortion,
harmonic distortion and noise. Frequency distortion happened when passive or active
circuit increases or decreases the amplitude of particular frequencies differently than
the other frequencies. Frequency distortion is a common problem wide band IF or RF
amplifiers. The frequency limitations of amplifiers have many causes. The reasons
for frequency limitation of amplitudes are active device’s transit time, negative
Fig. B.20 Input signal and resultant output signal

effects of junction capacitance, reactive nature of the transistor’s matching, filtering

and coupling, and decoupling networks (highly frequency dependent and act like a
filter, band-pass, high-pass, etc.,). The frequency distortion due to harmonic is pre-
sented in Fig. B.20.
The amplitude distortion is a form of nonlinear distortion (nonlinear behaviour).
It is produced by the incorrect biasing of an amplifier. It cause to either saturation or
cut-off of the transistor and generates harmonics and IMD products. Overdriving
the input of the amplifier (overload distortion) will create this same effect (flat
topping). Overdriving the input of the amplifier is not depending on amplifier’s
bias. The harmonics and IMDs generated will produce interference to other ser-
vices. The harmonics and IMDs generated will produce interference to adjacent
channels. It will increase the system BER (Bit Error Rate) in a digital data radio.
The voice band device will have an output signal with a harsh, coarse output. The
amplitude distortion due to incorrect biasing is described in Fig. B.21.
The amplitude distortion due to clipping is described in Fig. B.22.
In CE amplifier during the amplification process of the signal waveform, some
form of amplifier distortion has occurred. CE amplifier’s amplification may not be
Fig. B.21 The amplitude distortion due to incorrect biasing
Fig. B.22 The amplitude distortion due to clipping

taking place over the whole signal cycle due to incorrect biasing. If CE amplifier’s
input signal is too large, it causes the amplifier to be limited by the supply voltage
(clipping). CE amplifier’s amplification may not be linear over the entire frequency
range of inputs. CE amplifier’s multiplication factor is called the Beta (b) value of
the transistor. Common emitter or even common source type transistor circuits
work fine for small AC input signals. CE amplifiers suffer from one major disad-
vantages: The bias Q-point of a bipolar amplifier depends on the same (b) value
which may vary from transistors of the same type, Q-point for one transistor is not
necessarily the same as the Q-point for another transistor of the same type due to the
inherent manufacturing tolerances. If the CE amplifiers suffer from one major
disadvantage the amplifier may not be linear, amplitude distortion will result, and a
carful choice of the transistor and biasing components can minimize the effect of
amplifier distortion. The CE amplifier’s amplitude distortion is presented in
Fig. B.23.
Phase distortion or delay distortion occurs in a nonlinear transistor amplifier
when there is a time delay between the input signal and its appearance at the output
(Fig. B.24).
The phase change between the input and the output is zero at the fundamental
frequency. The resultant phase angle delay will be the different between the
Fig. B.23 CE amplifier’s amplitude distortion
Fig. B.24 Amplifier’s input signal X(t) and output signal Y(t + Dt)
Fig. B.25 Amplifier’s phase distortion due to delay
harmonic and fundamental. The time delay (D) will depend on the construction of
the amplifier and will increase progressively with frequency within the bandwidth
of the amplifier. Any practical amplifier will have a combination of both “fre-
quency” and “phase” distortion together with amplitude distortion. Most applica-
tions such as in audio amplifiers or power amplifiers, unless the distortion is
excessive or severe it will not generally affect the operation of the system.
Figure B.25 describes the phase distortion due to delay.
The Intermodulation Distortion (ID), quite similar to the amplitude distortion. ID
is produced when frequencies not harmonically related to the fundamental. Inter
Modulation Distortion (IMD) products can be formed by mixing together of the
carrier with interferers, harmonic, IMD products from other stages, other channels,
or sideband, producing various spurious response. IMD products are in band and
can swamp the desired signal, creating severs interference. When neighbouring
transmitted signal arrives at a PA’s stage, mix together with the transmitter’s carrier,
causing IMDs to be created. ID is produced when two or more frequencies mix in
any nonlinear device. It causes numerous sum and different combinations of the
original fundamental frequencies (second order products: f1 + f2, f1 − f2). It causes
intermodulation products (m f1 þ n; mf1 f2 n f2 ), n and m are whole numbers.
Third order ID products, which would be 2 f1 þ f2 , 2 f1 f2 , 2 f2 þ f1 , 2 f2 f1
can be most damaging of the higher or lower IMDs. The second order IMD
products would usually be too far from the receivers or transmitter’s band pass to
create many problems (Fig. B.26).
Third Order Intercept Point (TOIP, IP3): Third order spurious products will be
created within nonlinearity of a device (linear amplifier, active filter, and mixer).
Output IP3 point can never actually reached, since the amplifier will go into sat-
uration before this amplitude is ever truly attained. The value of the IP3 must be
measured only when it is in its linear operating range (DUT is not in compression).
Typically amplifier’s third order intercept point (IP3) is located approximately 10 to
15 dB above its P1 dB compression point. The output IP3 (OIP3) for a BJT
Fig. B.26 Amplifier’s Inter modulation distortion (IMD)
amplifier approximated by OIP3 ¼ 10 log½VCE IC 5, VCE is transistor’s col-

lector to emitter voltage [V]. IC is transistor’s collector current [mA]. The higher the
bias level, the higher will be the IP3 of the amplifier. IC is the easiest bias parameter
to increase for high IP3 (Figs. B.27 and B.28).
Fig. B.27 Amplifier’s output versus input and intercept point

Fig. B.28 Amplifier’s output versus input and intercept point saturation region
Harmonic Distortion (HD): HD occurs when an RF fundamental sine wave (fr) is

distorted due to nonlinearity within a circuit. It is generating harmonically related
frequencies (2 fr ; 3 fr ; . . .; n fr ). Interference to receivers tuned to megahertz, or
even gigahertz, away from the transmitter’s output frequency (Fig. B.29).
The dominant cause of transmitted harmonics is overdriving a poorly filtered
power amplifier. The cause to an extreme case of distortion resulting in the sine
wave carrier actually is changing into rough square wave. These non- perfect square
waves contain: fundamental frequency, odd harmonics, even harmonics. No active
stage can be completely linear and there are number of harmonics being produced
within all amplifiers.
Amplifier Noise: There are two principal classifications of noise, circuited
generated and externally generated. Both of them limit the possible sensitivity and
gain of the receiver. Amplifier’s noises are unavoidable, but can be minimized.
Circuit noise creates a randomly changing and wide frequency ranging voltage.
Circuit generated noises: White noise created by a component’s electrons randomly
moving around due to thermal energy, and shot noise caused by electrons randomly
moving across a semiconductor junction and into the collector drain of a transistor.
External noise: It is produced by atmospheric upheavals like lighting and space
Fig. B.29 Tx and Rx system

Fig. B.30 Amplifier’s source

resistance (RS) which gener-
ate noise
noise caused by sunspots and solar flares. Cosmic noise is created by interfering
signals from stars. Noise is generated in all stages in amplifier and radio receivers.
The first stages, operating at the lowest signal levels, which are of main concern,
particularly where low signals from aerials, microphones, etc., are not be amplified.
Figure B.30 describes the source resistance (RS) which generated noise.
Thermal noise is generated because there is no such a device as a perfect
amplifier. Thermal noise is the resistance generates a noise voltage (Et) as follow:
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
Et ¼ 2 K T B RS ; K—Boltzmann’s constant, T—Absolute temperature, B—
Bandwidth in hertz. At normal temperature (
17 C), this is simplifies to Et ¼
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
1:6 1020 B RS resistance’s noise is the lowest noise which can be achieved
at the amplifier input. The practical equivalent noise at the input is always higher
than this resistance’s noise. The noise whilst is generated by an amplifier system.
The degree of noise is evaluated by referring it to the amplifier input. It is con-
sidered as equivalent noise at the input as though it were being generated at that
point. The equivalent noise voltage (En) is calculated by dividing the noise mea-
sured at the amplifier output (Eno) by the gain of the amplifier (AV), i.e, En = Eno/
AV. Noise in any system is dependent on bandwidth of the system and this must be
specified when defining noise performance. It is common practice to define noise
for a 1 Hz bandwidth. Noise voltage might be specified in Nano volts per square
pffiffiffiffiffiffi
root of Hertz (nV= Hz). Over a limited bandwidth, noise power can be considered
to be proportional to bandwidth and the noise voltage is proportional to the square
root of bandwidth. If noise voltage is defined for a 1 Hz bandwidth system, noise
can be determined by multiplying by the square root of the system bandwidth. The
level of noise generated by an amplifier system generally varies over a wide
spectrum and for a wide bandwidth. Noise performance must be defined by plotting
pffiffiffiffiffiffi
noise (say in nV= Hz) against frequency. In solid state amplifiers, noise is often
resolved into two components at the amplifier input. An equivalent noise voltage
generator (Vn) and an equivalent noise current generator (In). The first component
(Vn) is independent of the value of source resistance (Rs). The second component
(In) develops a noise voltage across Rs and equal to In Rs . The noise voltage it
develops is directly proportional to the value of Rs (Fig. B.31).
Fig. B.31 Equivalent noise voltage and noise current at amplifier input
The Vn can be separated from the voltage developed by In and also the thermal
noise (Et) by short circuiting the input terminals. One method of defining the noise
performance of an amplifier is the noise figure (F). This can be defined as the ratio
of equivalent noise power developed at the input to that generated by thermal noise
in the source resistance (Rs). Noise figure is often expressed in decibel form and a
perfect amplifier would have a Noise Figure (NF) of 0 dB, if such a device were
possible. To establish Noise Figure (NF), the voltage gain (Av) of the amplifier is
measured and the noise voltage output (Eno) is measured at a known bandwidth (B).
For the second measurement, the amplifier input must be terminated in a resistance
(Rs) equal to the normal source resistance. Noise figure is calculated as follows:
( )
Eno
F ¼ 20 log pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi dB:
Av 1:6 1020 B RS
The noise figure formula assumes a high impedance input to the amplifier. The
effective value of Rs as far as the calculation is concerned, is the parallel result of
the source resistance and input resistance of the amplifier. If the source is a
transmission line and it is terminated in its characteristic impedance (Z0), then Rs
should be substituted by a value Z0 divided by 2 (Fig. B.32).
Small Signal (SS) amplifiers always bias in their linear region. Small Signal
(SS) amplifier needed to increase tiny signal levels to proper levels required for a
transmitter’s final power amplifier (PA). A microwave receivers, first RF amplifier
(class A SS), high gain type. The discreet RF amplifier design topics are choice of
active device, input and output impedance matching network, bias circuit, and
Fig. B.32 Source input, Rs and amplifier

physical layout. Typical transistor has not a 50 X resistive Zin and Zout and its
reactance will vary over frequency, A þ i B ! B ¼ gðf Þ then A þ i gðf Þ.
Matching network must be use to match the device. Using LC components, the
match be perfect for narrow band of frequencies. SS amplifier has also inductive or
capacitive parts when perfect match is 50 þ i 0.
The matching process is as follow: first to match the active device to the sys-
tem’s resistive impedance and second to cancel the innate reactive elements within
the transistor, permit a perfect 50 þ i 0 match with no reactance (capacitive and
inductive). This is calling conjugate matching. RF matching network: Take
S-parameter two port file, which represent transistor and initially ignore any effects
the added DC biasing network which may have on the active device in the final
physical design. This assumption is valid only if small amounts of RF feedback are
produced by the high values of Rf (RB) (feedback resistor) in an amplifier’s bias
network. Low value resistor for Rf (RB) employs heavy RF feedback. When we
choose low value resistor for RF then device’s S-parameter file calculation for the
matching networks may no longer be completely valid for the transistor. It is
accurate only when the bias network employs high resistance values within the bias
network (Fig. B.33).
S-parameter files (*.S2P) contain only RF parameters for few frequencies
( 20). It is possible that interest frequency may falls between two published
values. For accuracy we take mean value between two closest frequencies within
Fig. B.33 Collector feedback bias for BJT transistor circuit

the file. Example: S-parameters are given in a certain *.S2P file for 3 GHz and
4 GHz. Our design requires a centred frequency at 3.5 GHz. We take the mean
value of each S-parameter at 3 and 4 GHz. To compute S12 at 3.6 GHz we use the
following formula:
S12 MAGð@3 GHzÞ þ S12 MAGð@4 GHzÞ

¼ S12 MAGð@3:5 GHzÞ
2
S12 hð@3 GHzÞ þ S12 hð@4 GHzÞ
¼ S12 hð@3:5 GHzÞ
2
Filename: 21bfg425.001; BFG425WFieldC1; V1 = 7.884E001V, V2 = 2.000E

+000V, I1 = 1.297E−005A, I2 = 1.000E−003A (Table B.2).
When we design a linear amplifier the amplifier need to be stable for our fre-
quency and bias of interest, overall impedance variations, and very wide region of
frequencies both low and high. We use the following K formula (Table B.3):
1 þ ðjDs j2 jS11 j2 jS22 j2 Þ

K¼ ; Ds ¼ S11 S22 S12 S21
2 jS21 j jS12 j
1 þ ðjDs j2 jS11 j2 jS22 j2 Þ

K¼ ; Ds ¼ S11 S22 S12 S21
2 jS21 j jS12 j
Example: We need to calculate whether transistor is stable at 1.5 GHz, with

VCE = 10 V and IC = 6 mA. The S-parameters at that frequency and bias point are
Table B.2 S parameters for 3 and 4 GHz (Mag, Ang)

S11 S21 S12 S22
Freq (GHz) Mag Ang Mag Ang Mag Ang Mag Ang
3 0.63 −148 2.19 60.35 0.105 2.75 0.607 −77.07
4 0.591 167.91 1.76 30.28 0.104 −11.14 0.472 −97.41
Table B.3 Amplifier’s K parameter values and outcome

K values Outcome Important!
K>1 Active device stable for all Zins and Zouts Zin and Zout can be not exact
presented its ports match
K<1 Device is potentially unstable Zin and Zout must be very
cautiously selected
found to be (transistor or S-parameter text file). The following calculation gives

K > 1 @ 1.5 GHz with transistor bias conditions.
S11 ¼ 0:195 \ 167:6 ; S22 ¼ 0:508 \ 32 ; S12 ¼ 0:139 \ 61:2 ; S12 ¼ 2:5 \ 62:4
Ds ¼ S11 S22 S12 S21 ¼ ½0:195 \ 167:6 ½0:508 \ 32
½0:139 \ 61:2 ½2:5 \ 62:4 ¼ 0:25 \ 61:4
1 þ ðjDs j2 jS11 j2 jS22 j2 Þ 1 þ ðj0:25j2 j0:195j2 j0:508j2 Þ
K¼ ¼ ¼ 1:1
2 jS21 j jS12 j 2 j2:5j j0:139j
The maximum available gain (sometimes called MAG and sometimes called
GMAX) of a device is only defined where K is greater than one. Algebraically, this
is because the term under the square root becomes negative for values of K less than
1. If K < 1 then maximum available gain is infinite and infinite gain means
oscillator. GMAX is calculated from stability factor K and the forward and reverse
transmission coefficients (S21, S22).

pffiffiffiffiffiffiffiffiffiffiffiffiffiffi S21
GMAX ¼ ðK K 2 1Þ for K [ 1:
S12

If K = 1 then GMAX ¼ S21
S12 for K ¼ 1 and available gain is undefined when K
is less than one. That is when the square root of (K2 − 1) becomes imaginary.
pffiffiffiffiffiffiffiffiffiffiffiffiffiffi jS21 j pffiffiffiffiffiffiffiffiffiffiffiffiffiffi ðK þ K 2 1Þ jS21 j
GMAX ¼ ðK K 1Þ
2 ¼ ðK K 1Þ 2 pffiffiffiffiffiffiffiffiffiffiffiffiffiffi
jS12 j ðK þ K 2 1Þ jS12 j
" #
pffiffiffiffiffiffiffiffiffiffiffiffiffiffi jS21 j 1 jS21 j
GMAX ¼ ðK K 1Þ
2 ¼ pffiffiffiffiffiffiffiffiffiffiffiffiffiffi
jS12 j ðK þ K 1Þ jS12 j
2
" #
1 jS21 j pffiffiffiffiffiffiffiffiffi
1 jS21 j
GMAX ¼ pffiffiffiffiffiffiffiffiffiffiffiffiffiffi K 1) K 2 1K ¼ 2 K S
ðK þ K 2 1Þ jS12 j j 12 j
h i
pffiffiffiffiffiffiffiffiffi jS j
ðK þ K 2 1Þ 21
jS21 j pffiffiffiffiffiffiffiffiffi
jS12 j jS j ðK þ K 2 1Þ
GMAX log ¼ 10 log10 ¼ 10 log10 12 þ 10 log10
pffiffiffiffiffiffiffiffiffi
jS j jS j ðK þ K 1Þ
2
GMAX log ¼ 10 log1021 10 log1012 þ 10 log10
S11 ¼ 0:195 \ 167:6 ; S22 ¼ 0:508 \ 32 ; S12 ¼ 0:139 \ 61:2 ; S12 ¼ 2:5 \ 62:4
Ds ¼ S11 S22 S12 S21 ¼ ½0:195 \ 167:6 ½0:508 \ 32
½0:139 \ 61:2 ½2:5 \ 62:4 ¼ 0:25 \ 61:4
1 þ ðjDs j2 jS11 j2 jS22 j2 Þ 1 þ ðj0:25j2 j0:195j2 j0:508j2 Þ
K ¼ ¼ ¼ 1:1
2 jS21 j jS12 j 2 j2:5j j0:139j
jS j
½ 21 pffiffiffiffiffiffiffiffiffi j2:5j pffiffiffiffiffiffiffiffiffiffi
jS j ðK þ K 2 1Þ ½ ð1:1 1:12 1Þ
GMAX log ¼ 10 log1012 þ 10 log10 ¼ 10 log10
j0:139j
þ 10 log10
¼ 10:63 dB
SS (Small Signal) amplifier stability: A typical amplifier must be unconditionally

stable across all frequencies and input/output impedances. An amplifier may
oscillate at his band when gain is higher than one. Unstable transistor causes to shift
the bias point of the stage and it increases internal device dissipation and possibly
causing its destruction. The display an oscillation is seen in the frequency domain
on a spectrum analyser. It is distinguished by low voltage and low current spurs.
Spurs which begin to shift frequency are viewing instabilities in amplifier.
Instability spurs must be eliminated by stabilizing the circuit. Stability of an
amplifier stage is dependent on transistor’s temperature and bias, signal level, Hfe
spread b(frequency), active device’s positive internal feedback mechanism,
excessively high gain outside of the desired bandwidth, external positive feedback
caused by support components, PCB layout, and RF shield’s box mode (RF
shielding resonances). Additionally the stability of an amplifier stage is dependent
on low frequency gain of a normal amplifier, transistor’s possible instabilities when
Fig. B.34 Amplifier possess a load that no longer appears as 50 X

presented with anything other than 50 X termination, RF coupling inductor which

at low frequencies presents a true RF choke response over a higher band of limited
frequencies. In low frequency the RF choke will begin to look more like a piece of
straight, low impedance wire (Zchoke ¼ RL þ j x L) then amplifier now possess a
load that no longer appears as 50 X. This can cause oscillations in a conditionally
stable amplifier. The solution is to add a high value low frequency choke in series
with low value RF inductor (Fig. B.34).
Frequency decreases cause to lack of 50 X termination because the amplifier’s
matching circuit is good only over specific frequency band. The induction adopted
for decoupling of the low impedance power supply becomes close to short circuit as
frequency decreases (f ðxÞ ! e).
Another way to ensure no amplifier’s low frequency oscillations is to employ a
50 X resistor at the DC end of the bias circuit. The 50 X termination resistance at
low frequencies, the distributed RFC would have little effect. Capacitor CB helps to
shunt low frequency RF to ground, further decreasing the disruptive low frequency
RF gain. Another configuration is maintained strong decoupling from the power
supply at low frequencies. The circuit uses both a low and high frequency choke to
sustain high impedance into the power supply (Figs. B.35 and B.36).
Fig. B.35 Amplifier circuit with termination low frequencies into 50 X to prevent instability
Fig. B.36 Amplifier circuit with possible location of a shunt or series stabilization resistor
B.3 Small Signal (SS) Amplifier Design

& Matching Network
We get the SS amplifier gain and stability by scalar approximation. In the scalar
approximation only the magnitude of the S-parameters are employed and phase
angles are not employed. Gtu is the transducer unilateral gain in dB. It is an
amplifier’s power gain into an unmatched 50 X load, a worse case gain value is
Gtu ¼ 10 log½jS21j2 . Mismatch Loss (ML) is (ML ¼ a p) at transistor’s input in
decibels: a pin ¼ 10 logð1 ½S11 2 Þ. Mismatch Loss (ML ¼ a p) at transis-
tor’s output in decibels: a pout ¼ 10logð1 ½S22 2 Þ. The Mismatch Loss
(ML) for unmatched transistor is a Ptotal ¼ apin þ a pout . The Maximum
Available Gain (MAG) is calculated by MAG ¼ Gtu þ a Ptotal .
MAG ¼ Gtu þ a Ptotal ¼ 10 log½jS21j2 10 logð1 ½S11 2 Þ 10 logð1 ½S22 2 Þ:

n o
jS21 j2
ð1S2 Þð1S2 Þ
MAG ¼ 10 log10 11 22
:
The Maximum Stable Gain (MSG) is
jS21 j
MSG ¼ 10 logð Þ:
jS12 j
If MAG<MSG then the transistor is stable and if MAG>MSG then the transistor
is unstable. Example: we have transistor with the following S-parameters:
S11 ¼ 0:195 \ 167:6 ) S11 ¼ 0:195; S22 ¼ 0:508 \ 32 ) S22 ¼ 0:508
S12 ¼ 0:139 \ 61:2 ) S12 ¼ 0:139; S21 ¼ 2:5 \ 62:4 ) S21 ¼ 2:5:
ðj2:5j2 Þ ð10:1952 Þ
Gtu ¼ 10 log10 ¼ 7:96 dB; aPin ¼ 10 log10 ¼ 0:168 dB
ð10:508Þ2
aPout ¼ 10 log10 ¼ 1:29 dB
aPtotal ¼ aPin þ aPout ¼ 0:168 dB þ 1:29 dB ¼ 1:46 dB
MAG ¼ Gtu þ aPtotal ¼ Gtu þ aPin þ aPout ¼ 7:96 dB þ 1:46 dB ¼ 9:42dB
ðj2:5j=j0:139jÞ
MSG ¼ 10 log10 ¼ 12:55 dB; MAG\MSG ) transistor stable
The most amplifiers matching networks are type L (LC), type T, and type PI. We
need to design the amplifier stage’s matching network. Our target is that the
amplifier’s impedance is exactly matching the independences of the circuit. If
ZS ¼ ZL than maximum power is transferred from the source to load (no power
reflections) (Fig. B.37).
The amplifier and load ZS ¼ Rs þ j Xs ; ZL ¼ RL j XL , only if Rs ¼ RL and
Xs ¼ XL then there is a perfect match. There are three popular matching networks L,
PI, T (Fig. B.38).
Fig. B.37 Amplifier and load system
Fig. B.38 Popular matching network (L type, T type, and PI type)

Fig. B.39 Source load system—matched
We get perfect match when
1 1 1
Xs ¼ XL ) x L ¼ ; x2 ¼ ) x ¼ pffiffiffiffiffiffiffiffiffiffi
xC LC LC
1
f ¼ pffiffiffiffiffiffiffiffiffiffi :
2p LC
There is only one frequency which will be perfectly matched from source to load
(Fig. B.39).
And unmatched system: (Fig. B.40).
Example A: RS \RL ; f ¼ 1:5 GHz. If Zin ¼ Rs then there is a perfect match
otherwise Zin 6¼ Rs and there is unperfected match (Fig. B.41).
Fig. B.40 Source load system—unmatched

Fig. B.41 Source load system with L network, unmatched
1 ZC ZLoad
XL ¼ j x L; XC ¼ ; Zin ¼ ZL þ ZC k ZLoad ¼ ZL þ
jxC ZC þ ZLoad
jxC RL
1
RL
Zin ¼ j x L þ 1 ¼ j x Lþ
jxC þ RL
1 þ j x C RL
RL ð1 j x C RL Þ
Zin ¼ j x L þ ¼jxL
1 þ j x C RL ð1 j x C RL Þ
RL ð1 j x C RL Þ
þ
1 þ x2 C2 R2L
j x C R2L RL
Zin ¼ j x L þ
1 þ x2 C2 R2L 1 þ x2 C2 R2L

C R2L RL
¼jx L þ
1 þ x C RL
2 2 2 1 þ x C2 R2L
2
RL C R2L
ð1Þ ) ¼ RS ; ð2Þ ) L ¼ 0;
1 þ x2 C RL
2 2 1 þ x2 C 2 R2L
RL
ð1Þ ) ¼ 1 þ x2 C 2 R2L
RS
ð1Þ ) x2 C 2 R2L

RL 1 1 1 RL
¼ 1)C ¼ 2
¼ 1
RS RL RS x2 x2 R2L x2 R2L RS
s
ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
1 RL 1
ð1Þ ) C ¼ 1 ¼
x RL RS 2 p 1:5 GHz 58
rffiffiffiffiffiffiffiffiffiffiffiffiffiffi
58
1 ¼ 3:56 1012 ¼ 3:56 pF:
12
C R2L C R2L
ð2Þ ) L ¼ 0 ) L ¼
1 þ x2 C2 R2L 1 þ x2 C2 R2L
rffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
h iffi
xRL RS 1 RL
1 RL 2
¼ h i
1 þ x2 xR
1
L
RRLS 1 R2L
h iffi rffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
h iffi
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
ffi
RL
x
RL
RS 1 RL
x
RL
RS 1
RS RL
ð2Þ ) L ¼ h i ¼ RL ¼ 1
1 þ RRLS 1 RS
x RS
rffiffiffiffiffiffiffiffiffiffiffiffiffiffi
12 58
¼ 1 ¼ 2:48 109 ¼ 2:48 nH
2 p 1:5 GHz 12
Example B: RS [ RL ; f ¼ 1:5 GHz and for match Zin ¼ Rs (Fig. B.42).

1
1
ðj x L þ RL Þ
Zin ¼ ZC
ðZL þ RL Þ ¼
ðj x L þ RL Þ ¼ jxC
j x C 1
jxC þ ðj x L þ RL Þ
j x L þ RL
¼
j x C þ RL x 2 C L þ 1
RL þ j x L RL þ j x L
Zin ¼ ¼
½1 x2 C L þ j x C RL ½1 x2 C L þ j x C RL
½1 x2 C L j x C RL

½1 x2 C L j x C RL
ðRL þ j x LÞ f½1 x2 C L j x C RL g
Zin ¼
½1 x2 C L2 þ x2 C 2 R2L
RL ½1 x2 C L
¼
½1 x2 C L x2 C 2 R2L
x C R2L x L ½1 x2 C L
j þj
½1 x C L x C RL
2 2 2 2 ½1 x2 C L þ x2 C 2 R2L
x 2 L C RL
þ
½1 x C L2 þ x2 C 2 R2L
2
RL fx L x3 C L2 x C R2L g
Zin ¼ 2
þj
½1 x2 C L x2 C2 R2L ½1 x2 C L2 x2 C2 R2L
RL x fL x2 C L2 C R2L g
Zin ¼ þj
½1 x2 C L2 þ x2 C2 R2L ½1 x2 C L2 þ x2 C 2 R2L
RL
Zin ¼ RS ) ð1Þ ¼ RS ; ð2ÞL x2 C L2 C R2L ¼ 0
½1 x2 C L2 þ x2 C2 R2L
Fig. B.42 Source load system with L network unmatched
L
ð2ÞL x2 C L2 C R2L ¼ 0 ) C ¼ ; ð2Þ ) ð1Þ )
x2 L2 R2L
RL
ð1Þ h i h i2 ¼ RS
1 x2 x2 LL2 R2 L2 þ x2 x2 LL2 R2 RS
L L
2
RL x2 L2 x 2 L 2 R2
ð2Þ ¼ 1 2 2 þ )
RS x L R 2
½x2 L2 R2 2
" #2
12 ð2 p 1:5 109 Þ2 L2 ð2 p 1:5 109 Þ2 L2 122
ð2Þ 0:2 ¼ ¼ 1 2
þh i2
58 ð2 p 1:5 109 Þ L2 122 ð2 p 1:5 109 Þ2 L2 122
NSolve½fð1 ((2 3:14Þ^ 2

ð1:5 1;000;000;000Þ^ 2 x xÞ=ðð2 3:14Þ^ 2 ð1:5 10;000;000;000Þ^ 2 x x þ 144ÞÞ
^
2 þ ðð2 3:14Þ^ 2 ð1:5 1;000;000;000Þ^ 2 x x 144Þ=ðð2 3:14Þ^ 2
ð1:5 10;000;000;000Þ^ 2 x x þ 144Þ^ 2gh12=58; fxg
ffx ! 2.49413 109 g; fx ! 2.49413 1019 gg
Two solutions: L1 ¼ 2:49 109 ; L2 ¼ 2:49 109 ) L ¼ 2:49 nH

L 2:49 nH
ð2Þ C ¼ 2 2 )C¼
x L RL 2
ð2 p 1:5 109 Þ2 ½2:49 nH2 122
¼ 3:587 1012 ¼ 3:587 pF
EDU (2.49 0:000000001)/((2 3.14 1.5 2.49).^ 2 þ 144Þ

ans ¼
3:5870e012
When two different, but pure resistances must be matched the L-technique is
applied to perform the task. If reactance XL or XC must be cancelled, we use two
methods: first the absorption uses to reactance of the impedance matching network
itself to be absorb the undesired load and/or source reactance. Second the reso-
nance, which is utilized to resonant out the stray reactance of the device or circuit to
be matched at our desired frequency.
First absorption method: this is accomplished by positioning the matching
inductor in series with any load or source inductive reactance. In this way, the load
or source’s XL becomes a part of the matching inductor. The same outcome can be
attained by positioning a matching capacitor in parallel with any load or source XC.
Thus we are combining the two values into one larger value. This allows the
internal stray reactance of both devices to contribute the matching network. This
internal reactance is being subtracted from the calculated values of the LC matching
components. The transistors own stray reactance is now becoming an additive part
of the matching network. This absorption method is only useful if the stray internal
reactance of the device is less than the calculated reactance required for a proper
match. Figure B.43 describes the circuit that requires the addition of components to
absorb reactance.
LS þ L1 ! L01 is a new inductance value for matching L network. CS þ C1 ! C10
is a new capacitor value for matching L network. The absorption methods are by
using Z match network. The below figure describes the absorption methods by
using Z matching network (Figs. B.44, B.45 and B.46).
Second absorption (resonance) method: Resonance technique is utilized to res-
onate out the stray reactance of the device or circuit to be matched at our desired
frequency (f). It is done with a reactance that is equal in value, but opposite sign and
then continuing on as if the matching problem were a completely resistive one
(R þ j 0). This will make the internal stray reactance of the two devices or circuits
Fig. B.43 Amplifier (source) and load, absorb reactance methods

Fig. B.44 Amplifier (source) and load Z match network
Fig. B.45 Amplifier (source) and load L – network
Fig. B.46 Amplifier (source) and load matching flow chart

Fig. B.47 Amplifier (source) and load resonance impedance matching
Fig. B.48 Amplifier (source) and load canceling the load’s stray reactance
disappear and allowing only the pure resistances to be easily dealt with. We need to
design a matching network which employing the second method. Figure B.47
describes the resonance impedance matching.
We need to resonate out 1.5 pF (Cstray) of stray capacitance within the load. It is
done by employing a shunt inductor with value of L ¼ j2pf j12 C (Fig. B.48).
stray
1
Cstray kL ) ZT ¼ j x Lk
j x Cstray
j x L jxC1 stray jxL
) ZT ¼ ¼
j x Lþ 1
jxCstray
1 x2 L Cstray
xL
ZT ¼ j ; Resonate ) ZT ! 1ðdisconnected element)
1 x2 L Cstray
1e
) ZT ! 1 ) 1 - x2 L Cstray ! e ) L ! 2 ;e¼0
x Cstray
1 1
)L¼ 2 ¼
x Cstray j2 p f j2 Cstray
Fig. B.49 Combine shunt inductor with L1 (L—network inductor)
Since ZT ! ∞, the internal stray capacitance can be considered as no longer

existing within the load. Since the source is purely resistive (Zs ¼ Rs þ j 0) and the
load is also pure resistive (ZL ¼ RL þ j 0) then we utilize the basic resistive lumped
matching techniques. We need to design an L-network to match source to load. The
Z match network is L-network, we can combining both of the inductors (L1 & L),
with a single inductor (Fig. B.49).
Three elements impedance matching (PI or T) networks are used in many narrow
band applications. The narrow band is due to the higher loaded Q over what the
L-network possesses. PI and T networks also permit any Q to be selected. We
always consider that PI & T Q’s L-network Q’s. The Q desired for a particular
application is calculated with the following formula,
Fig. B.50 PI matching network between source and load

fc
When utilizing high Q inductor Q = f2 f 1
. Q is the loaded quality factor of the
circuit. fc is the center frequency of the circuit. f2 is the upper frequency, need to
pass with little loss. f1 is the lower frequency, need to pass with little loss, we use PI
network to match two different pure resistances (Fig. B.50).
We consider PI network as two L networks attached back to back. There is a
virtual resistor “R” in the center, which is used only as an aid in designing these
networks. The virtual “R” will not be in the final design. We choose (Fig. B.51)
XP1 ¼ 2:68 X; XS1 ¼ 2:55 X; XS2 ¼ 5:7 X; XP2 ¼ 5:8 X; Q ¼ 10
RL is pure resistance of the load. RS is pure resistance of the source (amplifier).

Rh = max (RS, RL).
00 00
R 1 RH RH RH
¼ 2 ) 00 00 ¼ Q þ 1 ) Q ¼ 00 00 1 ) Q ¼ 00 00 1
2 2
RH Q þ1 R R R
R 58
RH ¼ maxðRs ; Rl Þ; 00 R00 ¼ 2
H
¼ Q¼10 ¼ 0:57 X
Q þ 1 RH¼58 102 þ 1
RL RL 58
RL kXP2 ) Q ¼ ) XP2 ¼ ¼ ¼ 5:8 X
XP2 Q 10
XS2
Q ¼ 00 00 ) XS2 ¼ Q 00 R00 ¼ 10 0:57 ¼ 5:7 X
R
RS RS RS RS
XP1 ¼ ; Q1 ¼ 00 00 1; XP1 ¼ ¼ qffiffiffiffiffiffiffiffiffiffiffiffiffiffi ¼ 2:68 X
Q1 R Q1 RS
100 R00
rffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi rffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
RS 12 XS1
Q1 ¼ 00 00 1 ¼ 1 ¼ 4:48; Q1 ¼ 00 00 ) XS1
R 5:7 R
¼ Q1 00 R00 ¼ 4:48 0:57 ¼ 2:55
We need to convert the reactance calculated to L and C values.
Fig. B.51 Virtual resistor and two L networks to design a PI network

X
2
XS
XS1 þ XS2 ¼ XSi ; XS ¼ x L ¼ 2 p f L )
i¼1
2 p f
1 1 1
XP ¼ ¼ )C¼
x C 2 p f C 2 p f XP
To match two stages with a PI network, while canceling reactance and matching
resistances, first we convert the load/source to/from parallel or series equivalences
and second we make it easier to absorb any reactance. Any PI network can be
transformed to an equivalent T network. This is also known as the Wye-Delta
transformation, which is the terminology used in power distribution and electrical
engineering. The PI is equivalent to the Delta and the T is equivalent to the Wye (or
star) form. The PI network and T network topologies are described in Fig. B.52.
The impedances of the PI network (Za, Zb, Zc) can be found from the impe-
dances of the T-network with the following equations:
Za ¼ ½ðZ1 Z2 Þ þ ðZ1 Z3 Þ þ ðZ2 Z3 Þ=Z2 ; Zb ¼ ½ðZ1 Z2 Þ þ ðZ1 Z3 Þ þ ðZ2 Z3 Þ=Z1

Zc ¼ ½ðZ1 Z2 Þ þ ðZ1 Z3 Þ þ ðZ2 Z3 Þ=Z3 :
The common numerator in all these expressions can prove useful in reducing the
amount of computation necessary. The impedances of the T-network (Z1, Z2, Z3)
can be found from the impedances of the equivalent PI-network with the following
equations. The next expression describes the PI network to T network transfor-
mation. There is a common denominator in these expressions.
Z1 ¼ ðZa Zc Þ=ðZa þ Zb þ Zc Þ; Z2 ¼ ðZb Zc Þ=ðZa þ Zb þ Zc Þ;

Z3 ¼ ðZa Zb Þ=ðZa þ Zb þ Zc Þ
There is a case where all the impedances are creative (i.e. they are all in the form
j X). In that case, the (−1) factors from squaring j j on the top cancel the (−1)
Fig. B.52 PI network and T network

from bringing the j in the denominator up top. In that case also all T-network
impedances are reactive. The below equations describe the situation when all PI
network impedances are reactive.
Za Zc Zb Zc Za Zb
Z1 ¼ ; Z2 ¼ ; Z3 ¼
Za Zb Zc Za Zb Zc Za Zb Zc
Za ¼ j a; Zb ¼ j b; Zc ¼ j c
Za Zc j a j c j2 a c a c
Z1 ¼ ¼ ¼ ¼j
Za þ Zb þ Zc j a þ j b þ j c j ða þ b þ cÞ ða þ b þ cÞ
Zb Zc j a j c j2 b c b c
Z2 ¼ ¼ ¼ ¼j
Za Zc j a j b j2 a b a b
Z1 ¼ ¼ ¼ ¼j
SS (Small Signal) amplifier synthesis of PI and T networks to transform resis-

tances and create phase shifts. We assume that the desired port impedances are
purely resistive (i.e. real). The design of T or PI network with purely reactive
components is both to produce a desired phase shift (beta) and transform the
impedances with the following equations. T and PI networks design (R1, R2, b).
j R1 cos b R1 R2 j R2 cos b R1 R2
Z1 ¼ ; Z2 ¼
sin b sin b
j R1 R2 j R1 R2 sin b
Z3 ¼ ; Za ¼ pffiffiffiffiffiffiffiffiffiffiffiffiffiffi
sin b R2 cos b R1 R2
j R1 R2 sin b pffiffiffiffiffiffiffiffiffiffiffiffiffiffi
Zb ¼ pffiffiffiffiffiffiffiffiffiffiffiffiffiffi ; Zc ¼ j R1 R2 sin b
R1 cos b R1 R2
The beta (b) is the phase lag passing through the network from either port 1 to
port 2 or vice versa. If beta (b) is 0 or p, these expressions break down, except if
R1 = R2. To transform resistive impedances without any phase shift, we have to use
Fig. B.53 Source amplifier and load, T-network

a transformer. In many practical applications, the load or generator impedances may

be reactive (i.e. Z (port 1) and Z (port 2) are some general R + j X). This can be
accommodated by absorbing the external reactive impedance into the network,
reducing or increasing the series or shunt impedance as required. If T-network is
required to connect between two impedances: 50 + j 0 and 100 – j 20 with 45
degrees of phase shift (b) (Fig. B.53).
Zs ¼ 50 þ j 0; Zl ¼ 100 j 20; R1 ¼ 50; R2 ¼ 100; b ¼ 45

j 50 0:707 50 100
Cos b ¼ sin b ¼ 0:707; Z1 ¼ ¼ j 50
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi 0:707
j 100 0:707 50 100 j 50 100
Z2 ¼ ¼ 0; Z3 ¼ ¼ j 100
0:707 0:707
What happened if R1 = R2? (Two equal resistive impedances), T-network.

j R1 cos b R1 R2 j R ½cos b 1
Z1 ¼ jR1 ¼R2 ¼R ¼
sin b sin b
j R2 cos b R1 R2 j R ½cos b 1
Z2 ¼ jR1 ¼R2 ¼R ¼
sin b sin b
j R1 R2 j R
Z3 ¼ ¼
sin b sin b
What happened if R1 = R2? (Two equal resistive impedances), PI-network.
j R1 R2 sin b R sin b
Za ¼ pffiffiffiffiffiffiffiffiffiffiffiffiffiffi jR1 ¼R2 ¼R ¼ j
R2 cos b R1 R2 ½cos b 1
j R1 R2 sin b R sin b
Zb ¼ pffiffiffiffiffiffiffiffiffiffiffiffiffiffi j ¼j
R1 cos b R1 R2 R1 ¼R2 ¼R ½cos b 1
Zc ¼ j R1 R2 sin b
The certain high frequency microwave applications, we use distributed matching

elements. It may be a lower cost and higher performance alternative to using
lumped parts. The method limitation: inability to easily create series capacitors. We
should employ a shunt distributed capacitors when matching impedances in our
microwave designs. What happened when the series input impedance of the device
is inductive? And we would like to tune it out. We use conjugate series capacitance
to cancel the transistor’s series input inductance (Fig. B.54).
Fig. B.54 Conjugate series capacitance to cancel transistor’s series input impedance
Hint: We would like to get away from using a lump series capacitor.
Solution: We convert the series input impedance of the device into equivalent
parallel input impedance. The equivalent parallel input impedance circuit permits us
to exploit a shunt distributed element. The shunt distributed element resonates out
the input impedance of the device. Series input impedance: Xs series reactance (X)
and Rs series resistance (X). Parallel input inductance: Xp equivalent parallel
reactance (X) and Xp equivalent parallel resistance (X). We interest in amplifier’s
series input impedance circuit to parallel input impedance circuit conversion. We
need to get the expressions for Xp and Rp as functions of Xs, Rs and developing the
mathematical connections between those circuit’s parameters (Fig. B.55).
Fig. B.55 Series input impedance and parallel input impedance

j XP RP j XP RP j XP þ RP
j XS þ RS ¼ ) j XS þ RS ¼
j XP þ RP j XP þ RP j XP þ RP
X RP þ j XP RP
2 2
) j XS þ RS ¼ P ) j XS þ RS
XP2 þ R2P
X 2 RP XP R2P
¼ 2P þ j
XP þ R2P XP2 þ R2P
XP R2P XP R2P
ð1Þ XS ¼ 2 ) XP2 þ R2P ¼ ;
XP þ RP 2 XS
XP2 RP X 2 RP
ð2Þ RS ¼ ) XP2 þ R2P ¼ P
XP þ RP
2 2 RS
XP RP XP RP
2 2
ð1Þ & ð2Þ ) ¼
XS RS
RP XP RP RS 2 X 2 RP
) ¼ ) XP ¼ ; XP þ R2P ¼ P
XS RS XS RS
h i2
RP R R R 2
RP RS
RP
XP2 RP RS XS P P S
XP2 þ R2P
¼ ) þ RP ¼
2
¼
RS XS RS RS XS

RP RS 2 RP RP RS 2
þ R2P ¼
XS RS XS
2

R RP R2 R2 RP R2S
) R2P þ R2P S2 ¼ R2P S2 ) R2P 1 þ S2 ¼ R2P
XS RS XS XS RS XS2

R2 RP R2S R2 RP R2S R2 RS
R2P 1 þ S2 ¼ RP 2 ) 1 þ S2 ¼ 2 ) 1 þ S2 ¼ RP 2
XS RS XS XS RS XS XS XS
R2S RS X2
1þ 2
¼ RP 2 ) RP RS ¼ XS2 þ R2S ) RP ¼ S þ RS
XS XS RS
Conclusion: We could design the distributed matching circuit as we would the

lumped type. We simply substitute the equivalent distributed components
(microwave as equivalent components).
RP RS XS2
ðAÞXP ¼ ; ðBÞRP ¼ þ RS
XS RS
We substitute the equivalent distributed components by micro strip as equivalent

components. For small (and large) signal devices, we implement quarter wave line
transformer. We use a distributed transformer to match a 50 X resistive source to an
Fig. B.56 Distributed transformer to match a 50 X resistive source
unequal resistive load. The distributed transformer to match a 50 X resistive source

to an unequal resistive load is as follow (Fig. B.56):
The matching can be accomplished as follow: calculate the input/output impe-
dances of the device to be matched, it is series impedance, or R þ j XðR j XÞ.
Otherwise we obtain these values from the data sheet. We convert series R þ j
XðR j XÞ to parallel as required. Whether we elect to utilize parallel or series will
depend on the following. It would be easier, with micro strip, to resonate out the
reactance in series or to resonate out the parallel circuit equivalent. If a distributed
part must be used for this purpose, a shunt capacitor is always desire. We calculate
the required micro strip width and length, at the frequency of interest. Simulate a
lumped value that will cancel out the reactive component of the device being
matched. We make the input or output R + j 0. Lumped microwave capacitors
and inductors can also be utilized if the micro strip part is unrealizable. We match
the real (resistive) part of the transistor’s input or output by employing the mi-
crostrip transformer. The microstrip transformer section is placed between the two
mismatched impedances.
Example: 50 X for the system’s transmission line impedance, and 20 X for the
transistor’s input resistance. The transformer segment will be long ¼ k4 Vp
Vp is the propagation velocity. As wide as a microstrip transmission line would
be with an impedance of
Z ¼ R1 R2 ; R1 ¼ 20X; R2 ¼ 50 X ) Z ¼ 20 50 ¼ 31:6 X:
For other requirements we need to use equation to plug in different microstrip

widths to obtain the desired impedance. The characteristic impedance of the
microstrip is (X) and Z0 is the microstrip’s characteristic impedance. W is the width
of the microstrip conductor. h is the thickness of the substrate between the ground
plane and the microstrip conductor.
377
Z0 ¼ pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
pffiffiffiffiffiffi
W
h þ 1 Er þ Er
W and h use the same units and Er is the dielectric constant of the board material.
The dielectric constant of the medium does not have a unit because it is a ratio.
377
pffiffiffiffiffiffi
W
h þ 1 Er þ Er
A 50 X microstrip is utilized in microwave circuits to prevent reflections and

mismatch losses between physically separated components. A calculated nominal
width will prevent the line from being either inductive or capacitive at any point
along its length.
Our target: Source’s output impedance matched to the microstrip. Microstrip
matched to the input impedance of the load.
Results: No standing or reflected waves and no power dissipated at heat, except
in the actual resistance of the copper and dielectric as I2 R loss. In microstrip, the
dielectric constant (Er) of the PCB’s substrate material will not be the sole Er. There
is a flux leakage into the air above the PCB, combined with the flux penetrating into
the dielectric. Actual effective dielectric constant Eff, is true dielectric constant that
the microstrip see. There is some value between the surrounding air and the true
dielectric constant of the PCB. To avoid RF field leakage from microstrips,
transmission lines should be isolated by at least two or more line widths and create
spaces between traces and circuits for decreasing any mutual coupling effects.
We decrease any impedance bumps at high microwave frequencies by keeping
microstrip always be run as short and as straight as possible. Microstrip’s angle
(turn) should using a mitered or slow turn bend (Fig. B.57).
We need to find the actual wavelength of the signal which it is being slowed
down by the PCB’s substrate material. We calculate the microstrip’s velocity of
propagation (Vp): First find the effective dielectric constant Eff of the
microstrip. The signal will be partly in the dielectric and partly in the air above the
microstrip. This affects the propagation velocity through this combination of the
two dielectric mediums. We need to find the actual wavelength of the signal in
microstrip line. Eff is the effective dielectric constant that the microstrip sees. Er is
the actual dielectric constant of the PCB’s substrate material. h is the thickness of
the substrate material between the top conductor and the bottom ground plane of the
microstrip. W is the width of the top conductor of the microstrip (same units as h).
Fig. B.57 Miter bend and

curve bend in microstrip line
Vp is the fraction of the speed of light as compared to light in a vacuum. kvac is the
wavelength of the signal of interest in a perfect vacuum. f is the frequency of the
signal of interest, GHz. The actual wavelength of the signal in microstrip line is as
follows:
8 9
>
< >
=
Er þ 1 ðEr 1Þ 1 1
Eeff ¼ þ qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi &VP ¼ pffiffiffiffiffiffiffi
2 >
: 2 1 þ ð12h >
; Eeff
W Þ
1
) VP ¼ sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi

; kvac ¼ 11;800=f
Er þ 1 ðEr 1Þ
2 þ 2 pffiffiffiffiffiffiffiffiffiffiffiffi
1 ffi
1þð 12h
W Þ
kvac 11;800
k ¼ VP kvac ¼ sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi

¼ sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi

Er þ 1 ðEr 1Þ Er þ 1 ðEr 1Þ

2 þ 2 pffiffiffiffiffiffiffiffiffiffiffiffi
1 ffi f 2 þ 2 pffiffiffiffiffiffiffiffiffiffiffiffi
1 ffi
1þð
12h
W Þ 1þð 12h
W Þ
The functional connection between microstrip characteristics impedance and

PCB's parameters (k,f,h,W) is Z0 ¼ nðW; h; k; f Þ.
vffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
uE þ 1 (Er 1)
11;800 u r 1
k ¼ rffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi ) t þf qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffig
2 2
f Er 2þ 1 þ f(Er21) pffiffiffiffiffiffiffiffiffiffiffiffi 1 ffig 1 þ ð12h W Þ
1þð
12h
W Þ
11;800 Er þ 1 (Er 1) 1 11;8002

¼ þf qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffig ¼ 2
kf 2 2 1 þ ð12hÞ k f2
W
Er 1 Er 1 1 1 11;8002
) þ þ qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi ¼ 2
2 2 2 1 þ ð12hÞ 2 1 þ ð12hÞ k f2
W W
1 1 1 1 11;8002
Er ½1 þ qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi þ ½1 qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi ¼ 2
2 1 þ ð12h 2 k f2
W Þ 1 þ ð12h W Þ
1 1 11;8002 1 1
) Er ½1 þ qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi ¼ 2 ½1 qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
2 1þð W Þ k f 2 2 1 þ ð12h W Þ
12h
1 1 11;8002 1 1
Er ½1 þ qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi ¼ 2 ½1 qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
2 1þð W Þ k f 2 2 1 þ ð12h W Þ
12h
2118002
k2 f 2
½1 pffiffiffiffiffiffiffiffiffiffiffiffi
1 ffi
1þð
12h
Þ
) Er ¼ W
½1 þ pffiffiffiffiffiffiffiffiffiffiffiffi
1 ffi
1 þ ð12h
W Þ
377
pffiffiffiffiffiffi
W
h þ 1 Er þ Er
377
¼ vffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
u v u
ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi

u u211;8002
u211;8002 1pffiffiffiffiffiffiffiffiffiffi u p ffiffiffiffiffiffiffiffiffiffi
u
1 1
W 1
W þ u
k2 f 2 k2 f 2
u 1 þ ð12hÞ
u
1 þ ð12hÞ
W
h þ1 u t
t 1 þ pffiffiffiffiffiffiffiffiffiffi
1
1 þ pffiffiffiffiffiffiffiffiffiffi
1
12h1þð Þ 12h
1þð Þ
W W
Question: We have a system of amplifier and load. The source resistance

(Rs) value is lower than load resistance (Rl) value. Rs = 15 X and our load
resistance can’t be lower than Г0 [X] value and maximum load resistance variation
can be D [X]. We use L—matching network which have fix inductor L1 and fix
capacitor C1. We have trim capacitor in series to L—Network’s C1 capacitor, sign it
as Cx. What is the variation gap of Cx trim capacitor, if we want to match our
amplifier to load over all possible load values? Find Cx = n(Rs, Г0, D) and possible
maximum and minimum Cx values. f = 2 GHz (Fig. B.58).
Answer: First we need to get Zin expression. Capacitors C1 and Cx are in series
and the total Capacitance is C1−x. C1−x = C1Cx/(C1 + Cx); Zc1−x = 1/(j xC1−x).
C1 CX 1 1
C1X ¼ ; ZC1X ¼ ¼ n o
C1 þ CX j x C1X j x C1 CX
C1 þ CX
C1 þ CX C1 þ CX
¼ ¼ j
j x C1 CX x C1 CX
ZC1X RL
ZC1X k Zload ¼ ZC1X k RL ; Zload ¼ RL ; ZC1X k RL ¼
ZC1X þ RL
C1 þ CX
j xC 1 CX
RL
¼ C1 þ CX
j xC 1 CX
þ RL
j ðC1 þ CX Þ RL
ZC1X k RL ¼
j ðC1 þ CX Þ þ RL x C1 CX
j ðC1 þ CX Þ RL
¼
RL x C1 CX j ðC1 þ CX Þ
Fig. B.58 Unmatched system Rs < Rl
j ðC1 þ CX Þ RL RL x C1 CX þ j ðC1 þ CX Þ
ZC1X k RL ¼
RL x C1 CX j ðC1 þ CX Þ RL x C1 CX þ j ðC1 þ CX Þ
ðC1 þ CX Þ2 RL j ðC1 þ CX Þ R2L x C1 CX
ZC1X k RL ¼
R2L x2 C12 CX2 þ ðC1 þ CX Þ2
ðC1 þ CX Þ2 RL
ZC1X k Zload ¼ ZC1X k RL ¼
ðC1 þ CX Þ R2L x C1 CX
j
RS ½perfect match ¼ Zin ¼ RS ; Zin ¼ ZL þ ZC1X k Zload ; ZL ¼ j x L

ðC1 þ CX Þ2 RL
Zin ¼ ZL þ ZC1X k Zload ¼ j x L þ
ðC1 þ CX Þ R2L x C1 CX
j
ðC1 þ CX Þ2 RL
Zin ¼ ZL þ ZC1X k Zload ¼
ðC1 þ CX Þ R2L C1 CX
þ j x fL g
For perfect match we need: RS = Zin; Zin = RS.

ðC1 þ CX Þ2 RL
ð1Þ ¼ RS
ð2ÞL ¼0
We know that
RL C0 & RL C0 þ D
RL C0 & RL C0 þ D ) C0 RL C0 þ D; RL 2 ½C0 ; C0 þ D
ðC1 þ CX Þ2 RL
ð1Þ RS ¼ ) R2L x2 C12 CX2 RS
þ ðC1 þ CX Þ2 RS ¼ ðC1 þ CX Þ2 RL
R2L x2 C12 CX2 RS ¼ ðC1 þ CX Þ2 ðRL RS Þ ) ðC1 þ CX Þ2
R2L x2 C12 CX2 RS
¼
RL RS
ð1Þ ) ð2Þ ) L ¼0
) L ¼0
R2L x2 C12 CX2 RS
R2L x2 C12 CX2 þ RL RS
L¼
R2L x2 C12 CX2 RS
R2L x2 C12 CX2 þ RL RS
ðC1 þ CX Þ R2L C1 CX ðRL RS Þ
)L¼ 2
RL x C12 CX2 ðRL RS Þ þ R2L x2 C12 CX2 RS
2
R2L x2 C12 CX2 ðRL RS Þ L þ R2L x2 C12 CX2 RS L

¼ ðC1 þ CX Þ R2L C1 CX ðRL RS Þ

¼ C1 R2L C1 CX ðRL RS Þ þ CX R2L C1 CX ðRL RS Þ
¼ C12 R2L CX ðRL RS Þ þ CX2 R2L C1 ðRL RS Þ
CX2 R2L C1 ðRL RS Þ R2L x2 C12 CX2 ðRL RS Þ L
R2L x2 C12 CX2 RS L þ C12 R2L CX ðRL RS Þ ¼ 0
CX2 fR2L C1 ðRL RS Þ R2L x2 C12 ðRL RS Þ L
R2L x2 C12 RS Lg þ C12 R2L CX ðRL RS Þ ¼ 0
CX2 fR2L C1 RL R2L C1 RS R2L x2 C12 RL L þ R2L x2 C12 RS L

R2L x2 C12 RS Lg þ C12 R2L CX ðRL RS Þ ¼ 0
CX2 fR2L C1 RL R2L C1 RS R2L x2 C12 RL Lg þ C12 R2L CX ðRL RS Þ ¼ 0
CX2 R2L C1 fRL RS x2 C1 RL Lg þ C12 R2L CX ðRL RS Þ ¼ 0
CX ½CX R2L C1 fRL RS x2 C1 RL Lg þ C12 R2L ðRL RS Þ ¼ 0
ð1Þ CX ¼ 0; ð2ÞCX R2L C1 fRL RS x2 C1 RL Lg þ C12 R2L ðRL RS Þ ¼ 0
ð2ÞCX R2L C1 fRL RS x2 C1 RL Lg þ C12 R2L ðRL RS Þ ¼ 0 )
C12 R2L ðRS RL Þ C1 ðRS RL Þ
CX ¼ ) CX ¼
R2L C1 fRL RS x2 C1 RL Lg RL RS x2 C1 RL L
We need to get RL as a function of Cx and additional components.
C1 ðRS RL Þ
CX ¼ ) C X R L C X R S C X x2 C 1 R L L
R L R S x2 C 1 R L L
¼ C1 RS C1 RL
CX RL CX x2 C1 RL L þ C1 RL ¼ C1 RS þ CX RS
RL fCX CX x2 C1 L þ C1 g ¼ RS ðC1 þ CX Þ
RS ðC1 þ CX Þ
)RL ¼
CX CX x2 C1 L þ C1
RS ðC1 þ CX Þ RS ðC1 þ CX Þ
RL C0 ) RL ¼ C0 ) C0
CX CX x2 C1 L þ C1 CX f1 x2 C1 Lg þ C1
C0 ) C0 0
CX f1 x2 C1 Lg þ C1 CX f1 x2 C1 Lg þ C1
In case (=) exist we get
RS ðC1 þ CX Þ
C0 ¼ 0
CX f1 x2 C1 Lg þ C1
C0 ¼ 0 ) ¼ C0
CX f1 x C1 Lg þ C1
2 CX f1 x2 C1 Lg þ C1
RS ðC1 þ CX Þ ¼ CX f1 x2 C1 Lg C0 þ C1 C0
RS C1 þ RS CX ¼ CX f1 x2 C1 Lg C0 þ C1 C0
RS C1 C1 C0 ¼ CX f1 x2 C1 Lg C0 RS CX
ð1Þ
CX fC0 x2 C1 L C0 RS g ¼ RS C1 C1 C0 ) CX
C1 ðRS C0 Þ
¼
C0 ½1 x2 C1 L RS
The other option for case (=) is when upper limit is Г0 + D, Г0 ! Г0 + D then
ð1Þ C1 ðRS C0 Þ C1 ðRS C0 DÞ

CX ¼ j !
C0 ½1 x2 C1 L RS C0 !C0 þ D ðC0 þ DÞ ½1 x2 C1 L RS
ð2Þ C1 ðRS C0 DÞ
CX ¼
ðC0 þ DÞ ½1 x2 C1 L RS
ð2Þ ð1Þ
Since Г0 > 0 & D > 0 ! CX \CX
ð2Þ ð1Þ
Additionally there is no way that CX ! 1 or CX ! 1
ð1Þ C1 ðRS C0 Þ
CX ¼ ) C0 ½1 x2 C1 L RS 6¼ 0
C0 ½1 x2 C1 L RS
RS
) 1 x2 C1 L 6¼
C0
RS
x C1 L 6¼ 1
2
C0
rffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
1 RS 1 RS
) x 6¼
2
ð1 Þ ) x 6¼ pffiffiffiffiffiffiffiffiffiffiffiffi ð1 Þ
C1 L C0 C1 L C0
rffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi rffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
1 RS 1 RS
) 2 p f 6¼ pffiffiffiffiffiffiffiffiffiffiffiffi ð1 Þ ) f 6¼ pffiffiffiffiffiffiffiffiffiffiffiffi ð1 Þ;
C1 L C0 2 p C1 L C0
RS RS
1 [0 ) 1[ ) C0 [ RS
C0 C0
ð2Þ C1 ðRS C0 DÞ
CX ¼ ) ðC0 þ DÞ ½1 x2 C1 L RS 6¼ 0
ðC0 þ DÞ ½1 x2 C1 L RS
RS
1 x2 C1 L 6¼
C0 þ D
RS 1 RS
) x2 C1 L 6¼ 1 ) x2 6¼ ð1 Þ
C0 þ D C1 L C0 þ D
rffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi rffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
1 RS 1 RS
x 6¼ pffiffiffiffiffiffiffiffiffiffiffiffi 1 ) 2 p f 6¼ pffiffiffiffiffiffiffiffiffiffiffiffi 1
C1 L C0 þ D C1 L C0 þ D
rffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
1 RS RS
f 6¼ pffiffiffiffiffiffiffiffiffiffiffiffi 1 )1 [0
2 p C1 L C0 þ D C0 þ D
RS
) 1[ ) C0 þ D [ RS
C0 þ D
Discussion No.1: check the values interval for lower limit Г0.
C0 0 ) C0 [ 0
RS ðC1 þ CX Þ CX f1 x2 C1 Lg C0 C1 C0
[0
Case No.1.A:
RS ðC1 þ CX Þ CX f1 x2 C1 Lg C0 C1 C0 [ 0
& CX f1 x2 C1 Lg þ C1 [ 0
RS C1 þ RS CX CX f1 x2 C1 Lg C0 C1 C0 [ 0
& C1 [ CX fx2 C1 L 1g
CX ½RS f1 x2 C1 Lg C0 [ C1 C0 RS C1
& C1 [ CX fx2 C1 L 1g
CX ½RS f1 x2 C1 Lg C0 [ C1 ½C0 RS
& C1 [ CX fx2 C1 L 1g
We consider RS f1 x2 C1 Lg C0 [ 0 and x2 C1 L 1 [ 0
Then
C1 ½C0 RS C1
CX [ & CX \ 2
RS f1 x C1 Lg C0
2 x C1 L 1
C1 C1 ½C0 RS
[ CX [
x2 C1 L 1 RS f1 x2 C1 Lg C0
C1 C1 ½1 CRS0
) 2 [ CX [ RS
x C1 L 1 C f1 x C1 Lg
2
0
We already know that Г0 > RS then Cx always has positive value.
RS RS C1 C1 ½1 CRS0
\1 ) 1 [ ½1 [ 0; 2 [ 0 & RS [0
C0 C0 x C1 L 1 C0 f1 x C1 Lg
2
pffiffiffiffiffiffiffiffiffiffiffiffi pffiffiffiffiffiffiffiffiffiffiffiffi
x2 C1 L 1 [ 0 ) ðx C1 L 1Þ ðx C1 L þ 1Þ [ 0
pffiffiffiffiffiffiffiffiffiffiffiffi
) x C1 L 1 [ 0
pffiffiffiffiffiffiffiffiffiffiffiffi pffiffiffiffiffiffiffiffiffiffiffiffi
x C1 L 1 [ 0 ) x C1 L [ 1
1 1 1
) x [ pffiffiffiffiffiffiffiffiffiffiffiffi ) 2 p f [ pffiffiffiffiffiffiffiffiffiffiffiffi ) f [ pffiffiffiffiffiffiffiffiffiffiffiffi
C1 L C1 L 2 p C1 L
RS
f1 x2 C1 Lg [ 0
C0
RS 1 RS
) ð 1Þ þ x2 C1 L [ 0 ) ð 1Þ þ x2 [ 0
C0 C1 L C0
1 RS 1 RS
ð 1Þ þ x [ 0 ) x
2 2
ð1 Þ [ 0
C1 L C0 C1 L C0
1 RS 1 RS
x2 ð1 Þ [ 0 ) ½x pffiffiffiffiffiffiffiffiffiffiffiffi ð1 Þ
C1 L C0 C1 L C0
1 RS
½x þ pffiffiffiffiffiffiffiffiffiffiffiffi ð1 Þ [ 0
C1 L C0
1 RS 1 RS
x pffiffiffiffiffiffiffiffiffiffiffiffi ð1 Þ [ 0 ) x [ pffiffiffiffiffiffiffiffiffiffiffiffi ð1 Þ
C1 L C0 C1 L C0
1 RS
) 2 p f [ pffiffiffiffiffiffiffiffiffiffiffiffi ð1 Þ
C1 L C0
1 RS 1 RS
2 p f [ pffiffiffiffiffiffiffiffiffiffiffiffi ð1 Þ ) f [ pffiffiffiffiffiffiffiffiffiffiffiffi ð1 Þ
C1 L C0 2 p C1 L C0
We have two conditions:

1 1 RS
ff [ pffiffiffiffiffiffiffiffiffiffiffiffig [ ff [ pffiffiffiffiffiffiffiffiffiffiffiffi ð1 Þgjqffiffiffiffiffiffiffiffiffiffi
2 p C1 L 2 p C1 L C0 R
ð1CS Þ\0
0
1
¼ ff [ pffiffiffiffiffiffiffiffiffiffiffiffig
2 p C1 L
Case No.1.B:
C0 0 ) C0 [ 0
RS ðC1 þ CX Þ CX f1 x2 C1 Lg C0 C1 C0
[0
RS ðC1 þ CX Þ CX f1 x2 C1 Lg C0 C1 C0 \0 &
CX f1 x2 C1 Lg þ C1 \0
RS C1 þ RS CX CX f1 x2 C1 Lg C0 C1 C0 \0 &
C1 \CX fx2 C1 L 1g
CX ½RS f1 x2 C1 Lg C0 \C1 C0 RS C1 & C1 \CX fx2 C1 L 1g
CX ½RS f1 x2 C1 Lg C0 \C1 ½C0 RS & C1 \CX fx2 C1 L 1g
We consider RS f1 x2 C1 Lg C0 [ 0 and x2 C1 L 1 [ 0
Then
C1 ½C0 RS C1
CX \ & CX [ 2
RS f1 x2 C1 Lg C0 x C1 L 1
Then
C1 ½1 CRS0 C1
CX \ RS & CX [
C0 f1 x2 C1 Lg x2 C1 L 1
C1 C1 ½1 CRS0
\CX \
x2 C1 L 1 C0 f1 x C1 Lg
RS 2
Finally for Case No.1 we have possible two options:
C1 C1 ½1 CRS0
ð1Þ [ CX [
x2 C1 L 1 RS
C0 f1 x2 C1 Lg
C1 C1 ½1 CRS0
ð2Þ 2 \CX \ RS
x C1 L 1 C f1 x C1 Lg
2
0
We need to decide which option is feasible for us. We define X as the difference
between out limits bands (UL/LL).
C1 C1 ½1 CRS0
X¼
x2 C1 L 1 CRS f1 x2 C1 Lg
0
and need to find if X > 0 or X < 0 or X = 0.
C1 C1 ½1 CRS0
X¼
x2 C1 L 1 CRS f1 x2 C1 Lg
h 0
i
C1 CRS0 f1 x2 C1 Lg C1 ½1 CRS0 ½x2 C1 L 1
¼ h i
½x2 C1 L 1 CRS0 f1 x2 C1 Lg
C1 CRS0 C1 f1 x2 C1 Lg ½C1 C1 CRS0 ½x2 C1 L 1
X¼ h i
C1 CRS0 C1 f1 x2 C1 Lg fC1 ½x2 C1 L 1 C1 CRS0 ½x2 C1 L 1g

X¼ h i
C1 CRS0 C1 f1 x2 C1 Lg C1 ½x2 C1 L 1 þ C1 CRS0 ½x2 C1 L 1
X¼ h i
C1 CRS0 C1 f1 x2 C1 Lg þ C1 ½1 x2 C1 L þ C1 CRS0 x2 C1 L C1 CRS0
X¼ h i
C1 CRS0 x2 C1 L R
X¼ h i ;C1 S x2 C1 L [ 0
½x C1 L 1 C0 f1 x C1 Lg
2 RS 2 C 0
h i
If ½x2 C1 L 1 RS
C0 f1 x2 C1 Lg [ 0 then X [ 0
h i
If ½x2 C1 L 1 CRS0 f1 x2 C1 Lg \0 then X\0
If X [ 0 then
C1 C1 ½1 CRS0
[ CX [
x2 C1 L 1 RS
C0 f1 x2 C1 Lg
If X\0 then
C1 C1 ½1 CRS0
\CX \
x2 C1 L 1 C0 f1 x C1 Lg
RS 2
Discussion No.2: check the values interval for upper limit Г0 + D.
C0 þ D; C0 D 0
RS ðC1 þ CX Þ
) ðC0 þ DÞ 0
RS ðC1 þ CX Þ CX f1 x2 C1 Lg ðC0 þ DÞ C1 ðC0 þ DÞ
0
Appendix C
BJT Transistor Ebers-Moll Model
and MOSFET Model
A bipolar junction transistor (BJT or bipolar transistor) is a type of transistor that

relies on the contact of two types of semiconductor for its operation. BJTs can be
used as amplifiers, optoisolation circuits, switches, or in oscillators in many
industrial and commercial applications. BJTs can be found either as individual
discrete components, or in large numbers as parts of integrated circuits. The
operation of bipolar transistor involves both electron and holes. There are two kinds
of charge carriers which characteristic of the two kinds of doped semiconductor
material. Electrons are majority charge carriers in n-type semiconductors, whereas
holes are majority charge carriers in p-type semiconductors. Unipolar transistors
such as the field-effect transistors have only one kind of charge carrier. Charge flow
in a BJT is due to diffusion of charge carriers across a junction between two regions
of different charge concentrations. The regions of a BJT are called emitter, col-
lector, and base. A discrete transistor has three leads for connection to these regions.
Typically, the emitter region is heavily doped compared to the other two layers,
whereas the majority charge carrier concentrations in base and collector layers are
about the same. By design, most of the BJT collector current is due to the flow of
charges injected from a high-concentration emitter into the base where there are
minority carriers that diffuse toward the collector, and so BJTs are classified as
minority-carrier devices. There are two types of BJT transistors, PNP and NPN
based on the doping types of the three main terminal regions. An NPN transistor
comprises two semiconductor junctions that share a thin p-doped anode region, and
a PNP transistor comprises two semiconductor junctions that share a thin n-doped
cathode region. In an NPN transistor, when positive bias is applied to the base–
emitter junction, the equilibrium is disturbed between the thermally generated
carriers and the repelling electric field of the n-doped emitter depletion region. This
allows thermally excited electrons to inject from the emitter into the base region.
These electrons diffuse through the base from the region of high concentration near
the emitter towards the region of low concentration near the collector. The electrons
in the base are called minority carriers because the base is doped p-type, which
makes holes the majority carrier in the base.

DOI 10.1007/978-3-319-45427-6
994 Appendix C: BJT Transistor Ebers-Moll Model and MOSFET Model
The collector–emitter current can be viewed as being controlled by the base–

emitter current (current control), or by the base–emitter voltage (voltage control).
These views are related by the current–voltage relation of the base–emitter junction,
which is just the usual exponential current–voltage curve of a p-n junction (diode).
The Bipolar transistor exhibits a few delay characteristics when turning on and off.
Most transistors and especially power transistors, exhibit long base-storage times
that limit maximum frequency of operation in switching applications. One method
for reducing this storage time is by using a Baker clamp. The proportion of elec-
trons able to cross the base and reach the collector is a measure of the BJT effi-
ciency. The heavy doping of the emitter region and light doping of the base region
causes many more electrons to be injected from the emitter into the base than holes
to be injected from the base into the emitter. The common-emitter current gain is
represented by bF ðbf Þ or the h-parameter hFE. It is approximately the ratio of the
DC collector current to the DC base current in forward-active region. It is typically
greater than 50 for small-signal transistors but can be smaller in transistors designed
for high-power applications. Another important parameter is the common-base
current gain aF ðaf Þ. The common-base current gain is approximately the gain of
current from emitter to collector in the forward-active region. This ratio usually has
a value close to unity; between 0.98 and 0.998. It is less than unity due to re-
combination of charge carriers as they cross the base region.
IC IC IC IC =IE aF bF
aF ¼ ;b ¼ ;b ¼ ¼ ¼ ; aF ¼ :
IE F IB F IE IC 1 IC =IE 1 aF bF þ 1
Transistors can be thought of as two diodes (P–N junctions) sharing a common

region that minority carriers can move through. A PNP BJT will function like two
diodes that share an N-type cathode region, and the NPN like two diodes sharing a
P-type anode region. Connecting two diodes with wires will not make a transistor,
since minority carriers will not be able to get from one P–N junction to the other
through the wire. Both types of BJT function by letting a small current input to the
base control an amplified output from the collector. The result is that the transistor
makes a good switch that is controlled by its base input. The BJT also makes a good
amplifier, since it can multiply a weak input signal to about 100 times its original
strength. Networks of transistors are used to make powerful amplifiers with many
different applications. In the discussion below, focus is on the NPN bipolar tran-
sistor. In the NPN transistor in what is called active mode, the base–emitter voltage
VBE and collector–base voltage VCB are positive, forward biasing the emitter–base
junction and reverse-biasing the collector–base junction. In the active mode of
operation, electrons are injected from the forward biased n-type emitter region into
the p-type base where they diffuse as minority carriers to the reverse-biased n-type
Appendix C: BJT Transistor Ebers-Moll Model and MOSFET Model 995
Fig. C.1 Bipolar transistor shown as two back to back p-n junction
collector and are swept away by the electric field in the reverse-biased collector–
base junction. For a figure describing forward and reverse bias, see semiconductor
diodes. The bipolar junction transistor can be considered essentially as two p-n
junctions placed back to back, with the base p-type region being common to both
diodes. This can be viewed as two diodes having a common third terminal as shown
in the below figure. The two diodes are not in isolation, but are interdependent. This
means that the total current flowing in each diode is influenced by the conditions
prevailing in the other. In isolation, the two junctions would be characterized by the
normal diode equation with a suitable notation used to differentiate between the two
junctions as can be seen. When the two junctions are combined, to form a transistor,
the base region is shared internally by both diodes even though there is an external
connection to it (Fig. C.1).
In the forward active mode, aF of the emitter current reaches the collector. This
means that aF of the diode current passing through the base-emitter junction con-
tributes to the current flowing through the base-collector junction. Typically, aF has
a value of between 0.98 and 0.99. This is shown as the forward component of
current as it applies to the normal forward active mode of operation of the device.
This current is shown as a conventional current. It is equally possible to reverse the
biases on the junctions to operate the transistor in the “reverse active mode”. In this
case, aR ðar Þ times the collector current will contribute to the emitter current. For the
doping ratios normally used the transistor will be much less efficient in the reverse
mode and aR would typically be in the range 0.1–0.5. The Ebers-Moll transistor
model is an attempt to create an electrical model of the device as two diodes whose
currents are determined by the normal diode law but with additional transfer ratios
to quantify the interdependency of the junctions. Two dependent current sources are
used to indicate the interaction of the junctions. Figure C.2 describes NPN Bipolar
transistor Ebers Moll model.
Fig. C.2 NPN bipolar transistor Ebers moll model
Applying Kirchoff’s laws to the model gives the terminal current as:
IDE ¼ IE þ ar IDC ; IC þ IDC ¼ af IDE ; af Ise ¼ ar Isc ¼ Is ; IE ¼ IC þ IB
af ¼ 0:98 0:99 typically. ar ¼ 0:1 0:5 typically. Ise : reverse saturation

current of the base emitter diode. Isc : reverse saturation current of the base collector
VBC VBE
diode. IDC ¼ Isc ðe VT 1Þ; IDE ¼ Ise ðe VT 1Þ. VT – the thermal voltage VT ’
kT
q (approximately 26 mV at 300 K (*room temperature). IE is the transistor’s
emitter current. IC is the transistor’s collector current. IB is the transistor’s base
current. The base internal current is mainly by diffusion (see Fick’s law) and
VBE
JnðbaseÞ ¼ qDWn nbo e VT . W is the base width. Dn is the diffusion constant for electron
in the p type base.
VBE VBC
IDE ¼ IE þ ar IDC ) Ise ðe VT 1Þ ¼ IE þ ar Isc ðe VT 1Þ
VBE VBC
) IE ¼ Ise ðe VT 1Þ ar Isc ðe VT 1Þ
VBC VBE
IC þ IDC ¼ af IDE ) IC þ Isc ðe VT 1Þ ¼ af Ise ðe VT 1Þ
VBE VBC
) IC ¼ af Ise ðe VT 1Þ Isc ðe VT 1Þ
VBE VBC
IB ¼ IE IC ¼ ð1 af Þ Ise ðe VT 1Þ þ ð1 ar Þ Isc ðe VT 1Þ:
If we use the notation af Ise ¼ ar Isc ¼ Is ; Isc ¼ aIsr ; Ise ¼ aIsf the following Ebers
Moll equations:
Is VBE Is VBC Is VBE VBC

IE ¼ ðe VT 1Þ ar ðe VT 1Þ ) IE ¼ ðe VT 1Þ Is ðe VT 1Þ
af ar af
Is VBE I s
VBC VBE Is VBC
IC ¼ af ðe VT 1Þ ðe VT 1Þ ) IC ¼ Is ðe VT 1Þ ðe VT 1Þ
af ar ar
Is VBE I s
VBC 1 1 af 1 1 ar
IB ¼ ð1 af Þ ðe VT 1Þ þ ð1 ar Þ ðe VT 1Þ; ¼ ; ¼
af ar bf af br ar
Is VBE I s
VBC
IB ¼ ð1 af Þ ðe VT 1Þ þ ð1 ar Þ ðe VT 1Þ
af ar
Is VBE Is VBC
) IB ¼ ðe VT 1Þ þ ðe VT 1Þ
bf br
The expressions for VBE, VBC, and VCE are as follow:

VBC
VBE VBC VBE IE þ ar Isc ðe VT 1Þ
IE ¼ Ise ðe VT
1Þ ar Isc ðe VT
1Þ ) ðe VT
1Þ ¼
Ise
VBC
IE þ ar Isc ðe VT 1Þ VBC
IC ¼ af Ise ð Þ Isc ðe VT 1Þ ) IC
Ise
VBC VBC
¼ af IE þ af ar Isc ðe VT 1Þ Isc ðe VT 1Þ
VBC VBC I C af I E
IC ¼ af IE þ ðaf ar 1Þ Isc ðe VT 1Þ ) e VT ¼ þ1
ðaf ar 1Þ Isc

VBC I C af I E I C af I E
e VT ¼ þ 1 ) VBC ¼ VT ln þ1
ðaf ar 1Þ Isc ðaf ar 1Þ Isc
VBE
VBE VBC VBC af Ise ðe VT 1Þ IC
IC ¼ af Ise ðe VT
1Þ Isc ðe VT
1Þ ) ðe VT
1Þ ¼
Isc
VBE
VBE af Ise ðe VT 1Þ IC
IE ¼ Ise ðe VT
1Þ ar Isc ð Þ ) IE
Isc
VBE VBE
¼ Ise ðe VT 1Þ ar af Ise ðe VT 1Þ þ ar IC
VBE VBE I E ar I C
IE ¼ ð1 ar af Þ Ise ðe VT 1Þ þ ar IC ) e VT ¼ þ1
ð1 ar af Þ Ise

VBE I E ar I C I E ar I C
e VT ¼ þ 1 ) VBE ¼ VT ln þ1 ;
ð1 ar af Þ Ise ð1 ar af Þ Ise

ar I C I E
VBE ¼ VT ln þ1
ðar af 1Þ Ise
We can summery our intermediate results:

IC af IE ar I C I E
VBC ¼ VT ln ð Þ þ 1 ; VBE ¼ VT ln ð Þþ1
ðaf ar 1Þ Isc ðar af 1Þ Ise
VCE = VCB + VBE, but VCB = −VBC. Then VCE = VBE − VBC.
Remark: there is a use with capital and small letters in the Appendix compares to
book chapter 1, consider the terminology is the same.
Ie ¼ IE ; Ic ¼ IC ; Ib ¼ IB ; Vt ¼ VT ; Vbe ¼ VBE ; Vcb ¼ VCB ; Vce ¼ VCE

I C af I E
VCB ¼ VBC ¼ VT ln ð Þ þ 1 ; VCE ¼ VBE VBC
ðaf ar 1Þ Isc

ar I C I E
VCE ¼ VBE VBC ¼ VT ln ð Þþ1
ðar af 1Þ Ise

I C af I E
VT ln ð Þþ1
ðaf ar 1Þ Isc

VCE ¼ VT ln
ðar af 1Þ Ise

VT ln
ðaf ar 1Þ Isc

½ar IC IE þ ðar af 1Þ Ise ðaf ar 1Þ Isc

VCE ¼ VT ln
ðar af 1Þ Ise ½IC af IE þ ðaf ar 1Þ Isc

½ar IC IE þ ðar af 1Þ Ise Isc

VCE ¼ VT ln
½IC af IE þ ðaf ar 1Þ Isc Ise


VCE ¼ VT ln
Isc Isc Isc
þ VT lnð Þ; 1 ) lnð Þ ! e
Ise Ise Ise


VCE VT ln
Figure C.3 describes PNP Bipolar transistor Ebers Moll model.

VEB VCB
IDE ¼ IE þ ar IDC ) Ise ðe VT 1Þ ¼ IE þ ar Isc ðe VT 1Þ
VEB VCB
) IE ¼ Ise ðe VT 1Þ ar Isc ðe VT 1Þ
Fig. C.3 PNP Bipolar transistor Ebers Moll model
VCB VEB
IC þ IDC ¼ af IDE ) IC þ Isc ðe VT 1Þ ¼ af Ise ðe VT 1Þ
VEB VCB
) IC ¼ af Ise ðe VT 1Þ Isc ðe VT 1Þ
VEB VCB
IB ¼ IE IC ¼ ð1 af Þ Ise ðe VT 1Þ þ ð1 ar Þ Isc ðe VT 1Þ:
If we use the notation af Ise ¼ ar Isc ¼ Is ; Isc ¼ aIsr ; Ise ¼ aIsf the following Ebers
Moll equations:
Is VEB Is VCB Is VEB VCB

IE ¼ ðe VT 1Þ ar ðe VT 1Þ ) IE ¼ ðe VT 1Þ Is ðe VT 1Þ
af ar af
Is VEB I s
VCB VEB Is VCB
IC ¼ af ðe VT 1Þ ðe VT 1Þ ) IC ¼ Is ðe VT 1Þ ðe VT 1Þ
af ar ar
Is VEB I s
VCB 1 1 af 1 1 ar
IB ¼ ð1 af Þ ðe VT 1Þ þ ð1 ar Þ ðe VT 1Þ; ¼ ; ¼
af ar bf af br ar
Is VEB I s
VCB
IB ¼ ð1 af Þ ðe VT 1Þ þ ð1 ar Þ ðe VT 1Þ
af ar
Is VEB Is VCB
) IB ¼ ðe VT 1Þ þ ðe VT 1Þ
bf br
The expressions for VEB, VCB, and VEC are as follow:

VCB
VEB VCB VEB IE þ ar Isc ðe VT 1Þ
IE ¼ Ise ðe VT
1Þ ar Isc ðe VT
1Þ ) ðe VT
1Þ ¼
Ise
VCB
IE þ ar Isc ðe VT 1Þ VCB
IC ¼ af Ise ð Þ Isc ðe VT 1Þ ) IC
Ise
VCB VCB
¼ af IE þ af ar Isc ðe VT 1Þ Isc ðe VT 1Þ
VCB VCB I C af I E
IC ¼ af IE þ ðaf ar 1Þ Isc ðe VT 1Þ ) e VT ¼ þ1
ðaf ar 1Þ Isc
VCB I C af I E IC af IE
e VT ¼ þ 1 ) VCB ¼ VT ln½ þ 1
ðaf ar 1Þ Isc ðaf ar 1Þ Isc
VEB
VEB VCB VCB af Ise ðe VT 1Þ IC
IC ¼ af Ise ðe VT
1Þ Isc ðe VT
1Þ ) ðe VT
1Þ ¼
Isc
VEB
VEB af Ise ðe VT 1Þ IC
IE ¼ Ise ðe VT
1Þ ar Isc ð Þ ) IE
Isc
VEB VEB
¼ Ise ðe VT 1Þ ar af Ise ðe VT 1Þ þ ar IC
VEB VEB IE ar IC
IE ¼ ð1 ar af Þ Ise ðe VT 1Þ þ ar IC ) e VT ¼ þ1
ð1 ar af Þ Ise
VEB IE ar IC I E ar I C
e VT ¼ þ 1 ) VEB ¼ VT ln½ þ 1;
ð1 ar af Þ Ise ð1 ar af Þ Ise
ar I C I E
VEB ¼ VT ln½ þ 1
ðar af 1Þ Ise
We can summery our result regarding IC and IE:

VEB VCB VEB VCB
IC ¼ af Ise ðe VT 1Þ Isc ðe VT 1Þ; IE ¼ Ise ðe VT 1Þ ar Isc ðe VT 1Þ
VEB VCB
IB ¼ IE IC ¼ Ise ðe VT 1Þ ar Isc ðe VT 1Þ
VEB VCB
½af Ise ðe VT 1Þ Isc ðe VT 1Þ
VEB VCB
IB ¼ IE IC ¼ ð1 af ÞIse ðe VT 1Þ þ ð1 ar Þ Isc ðe VT 1Þ;
VBE ¼ VEB ; VBC ¼ VCB
VBE VBC
IC ¼ af Ise ðe VT
1Þ Isc ðe VT
1Þ;
VBE VBC
IE ¼ Ise ðe VT
1Þ ar Isc ðe VT
1Þ
VBE VBC
IB ¼ IE IC ¼ ð1 af ÞIse ðe VT
1Þ þ ð1 ar Þ Isc ðe VT
1Þ
VCE = VCB + VBE, but VCB = −VBC. Then VCE = VBE – VBC.

I C af I E ar IC IE
VCB ¼ VT ln þ 1 ; VEB ¼ VT ln þ1

I C af I E ar I C I E
VBC ¼ VT ln þ 1 ; VBE ¼ VT ln þ1

I C af I E ar I C I E
VCE ¼ VCB þ VBE ¼ VT ln þ 1 VT ln þ1

VCE ¼ VCB þ VBE ¼ VT ln
ðaf ar 1Þ Isc

VT ln
ðar af 1Þ Ise

IC af IE þ ðaf ar 1Þ Isc Ise
ar IC IE þ ðar af 1Þ Ise Isc


Ise Ise
þ VT ln ; Ise Isc ; ln !e
Isc Isc

VCE ¼ VCB þ VBE VT ln

VCENPN VT ln ; VCEPNP

VT ln
Summary of our BJT NPN and PNP transistors Ebers-Moll equations

(Table C.1):
There are three basic circuit configurations to connect bipolar junction transistor.
First: Common Base (CB), both the input and output share the base “in common”.
Second: Common Emitter (CE), both the input and output share the emitter “in
common”. Third: Common Collector (CC), both the input and output share the
collector “in common”. There are four bipolar junction transistor biasing modes.
Active biasing is useful for amplifiers (most common mode). Saturation biasing
mode is equivalent to an on state when transistor is used as a switch. Cutoff biasing
mode is equivalent to an off state when transistor is used as a switch. Inverted
biasing mode is rarely if ever used (Table C.2a, b ).
Table C.1 Summary of our BJT NPN and PNP transistors Ebers-Moll equations
BJT NPN transistor BJT PNP transistor
VBE VBC VBE VBC
IC IC ¼ af Ise ðe VT
1Þ Isc ðe VT
1Þ IC ¼ af Ise ðe VT
1Þ Isc ðe VT
1Þ
VBE VBC VBE VBC
IE IE ¼ Ise ðe 1Þ ar Isc ðe 1Þ
VT VT
IE ¼ Ise ðe VT
1Þ ar Isc ðe VT
1Þ
VBE
IB Is VBE
IB ¼ ð1 af Þ ðe VT 1Þ IB ¼ ð1 af ÞIse ðe VT
1Þ
af VBC
Is VBC þ ð1 ar Þ Isc ðe VT
1Þ
þ ð1 ar Þ ðe VT 1Þ
ar
n o n o
VCE ½a I I þ ða a 1ÞI I af IE þ ðaf ar 1ÞIsc
VCE VT ln ½ICr aC f IEE þ ðarf afr 1ÞIsesc VCE VT ln aCr IC I E þ ðar af 1ÞIse
Table C.2a Summary of NPN BJT transistor biasing mode

Biasing mode E-B junction bias C-B junction bias Applied voltages
(NPN) (NPN) (NPN) (NPN)
Saturation Forward Forward VE < VB > VC
Active (forward Forward Reverse VE < VB < VC
active)
Inverted (reverse Reverse Forward VE > VB > VC
active)
Cutoff Reverse Reverse VE > VB < VC
Table C.2b Summary of PNP BJT transistor biasing mode

Biasing mode (PNP) E-B junction bias C-B junction bias Applied voltages
(PNP) (PNP) (PNP)
Saturation Forward Forward VE > VB < VC
Active (forward Forward Reverse VE > VB > VC
active)
Inverted (reverse Reverse Forward VE < VB < VC
active)
Cutoff Reverse Reverse VE < VB > VC
The BJT transistor base current is much smaller that the emitter and collector
currents in forward active mode. If the collector of an NPN transistor was open
circuit, it would look like a diode. When forward biased, the circuit in the
base-emitter junction would consist of holes injected into the emitter from the base
and electrons injected into the base from the emitter. But since there are many more
electrons in the emitter than holes in the base, the vast majority of the current will be
due to electrons. When the reverse biased collector is added, It “sucks” the electrons
out of the base. Thus, the base-emitter current is due predominantly to hole current
(the smaller current component) while the collector-emitter current is due to elec-
trons (larger current component due to more electrons from the n+ emitter doping).
We define two BJT transistor performance parameters: emitter efficiency (c) and
base transport factor (aT). Emitter efficiency parameter characterizes how effective
the large hole current is controlled by the small electron current. Unity is best, zero is
worst. Base transport factor characterizes how much of the injected hole current is
lost to recombination in the base. Unity is best, zero is worst.
IEp IEp
c¼ ¼
IE IEp þ IEn
ICp
aT ¼ :
IEp
We define some equations in active mode, common base characteristics. ICBo is

defined as the collector current when the emitter is open circuit. It is the collector
base junction saturation current. IC is the fraction of emitter current making it across
the base + leakage current.
IC ¼ adc IE þ ICBo ;
where adc is the common base DC current gain.
ICp ¼ aT IEp ¼ c aT IE ; IC ¼ ICp þ ICn ¼ aT IEp þ ICn ¼ c aT IE þ ICn ; adc

¼ c aT
and ICBo ¼ ICn . We define some equations in active mode, common emitter
characteristics. ICEo is defined as the collector current when the base is open circuit.
IC is multiple of the base current making it across the base + leakage current.
IC ¼ bdc IB þ ICEo ; Where bdc is the common emitter DC current gain. ICEo is
defines as the collector current when the base is open circuit.
aF ¼ adc is common base current gain. IE ¼ aR IC ; aR 6¼ aDC . In inverse mode,
the emitter current is the fraction of the collector current “collected”.
adc ICBo
IE ¼ IC þ IB ; IC ¼ adc ðIC þ IB Þ þ ICBo ; IC ¼ IB þ
1 adc 1 adc
adc ICBo IC
bdc ¼ ; ICEo ¼ ;b ¼
1 adc 1 adc dc IB
We can break the BJT transistor up into a large signal analysis and a small signal
analysis and “linearize” the non-linear behavior of the Ebers-Moll model. Small
signal models are only useful for forward active mode and thus, are derived under
this condition. Saturation and cutoff are used for switches which involve very large
voltage/current swings from on to off states.
Small signal models are used to determine amplifier characteristics
(“Gain” = increase in the magnitude of a signal at the output of a circuit relative to
its magnitude at the input of the circuit). Just like when a diode voltage exceeds a
certain value, the non-linear behavior of the diode leads to distortion of the
current/voltage curves, if the inputs/outputs exceed certain limits, the full Ebers-Moll
model must be used. There are physical meanings of bf (bF) and br (bR). bF is the
current gain (IC/IB) of the device when it is operating with the emitter as the emitter
and the collector as the collector in the active mode. bR is the current gain of the
device when it is operating with the emitter as a collector and the collector as an
emitter in the reverse mode. The BJT device is made to have higher forward current
gain than reverse current gain. The terminals for emitter and collector are not
completely interchangeable due to different doping of the collector and emitter.
BJTtransistor modes of operation:
The Ebers-Moll BJT model is a good large signal, steady-state model of the
transistor and allows the state of conduction of the device to be easily determined
for different modes of operation of the device. The different modes of operation are
determined by the manner in which the junctions are biased. BJT NPN transistor
Ebers-Moll BJT model:
VBE VBC VBE
IC ¼ af Ise ðe VT 1Þ Isc ðe VT 1Þ; IE ¼ Ise ðe VT 1Þ
VBC
ar Isc ðe VT 1Þ
Is VBE Is VBC
IB ¼ ð1 af Þ ðe VT 1Þ þ ð1 ar Þ ðe VT 1Þ;
af ar
Is Is
af Ise ¼ ar Isc ¼ Is ; Isc ¼ ; Ise ¼
ar af
VBE VBC
IB ¼ ð1 af Þ Ise ðe VT 1Þ þ ð1 ar Þ Isc ðe VT 1Þ
(A) Forward Active mode:

VBE VBE VBE
B-E forward biased, VBE positive e VT 1; ðe VT 1Þ e VT . B-C reverse
VBC VBC
biased, VBC negative e VT 1; ðe VT 1Þ 1. Then from the Ebers-Moll model
equations we get the following results:
VBE VBE VBE
IE ’ Ise e VT þ ar Isc Ise e VT ; Ise e VT ar Isc ;
Relatively large.
VBE VBE VBE
IC ’ af Ise e VT þ Isc af Ise e VT ¼ af IE ; af Ise e VT Isc ;
Relatively large.
VBE VBE
IB ’ ð1 af Þ Ise e VT ð1 ar Þ Isc ð1 af Þ Ise e VT
VBE
¼ ð1 af Þ IE ; ð1 af Þ Ise e VT ð1 ar Þ Isc
(B) Reverse active mode: VBE VBE

B-E reverse biased, VBE negative e VT 1; ðe VT 1Þ 1. B-C forward biased,
VBC VBC VBC
VBC positive e VT 1; ðe VT 1Þ e VT . The transistor conducts in the opposite
direction. Then from the Ebers-Moll model equations we get the following results:
VBC VBC VBC
IE ’ Ise ar Isc e VT ar Isc e VT ; ar Isc e VT Ise ;
Moderately high.
VBC VBC VBC
IC ’ af Ise Isc e VT Isc e VT ; Isc e VT af Ise ; Moderate:
VBC VBC VBC
IB ’ ð1 af Þ Ise þ ð1 ar Þ Isc e VT ð1 ar Þ Isc e VT ; ð1 ar Þ Isc e VT
ð1 af Þ Ise
It is as high as 0.5 |IC|. This mode does not provide useful amplification but is
used, mainly, for current steering in switching circuits, e.g. TTL.
(C) Cut-off mode:
B-E is unbiased, VBE = 0 v. B-C is reverse biased, VBC negative.
VBE VBE VBC VBC
e VT ¼ 1; ðe VT 1Þ ! e ¼ 0; e VT 1; ðe VT 1Þ 1
IE ’ ar Isc ; Leakage current nA. IC ’ Isc ; Leakage current nA.

IB ’ ð1 ar Þ Isc .
This is equivalent to a very low conductance between collector and emitter, i.e.
open switch (Fig. C.4).
(D) Saturation mode: VBE VBE VBE
B-E is forward biased, VBE is positive e VT 1; ðe VT 1Þ e VT and both junc-
VBC VBC VBC
tions are forward biased. B-C is forward biased, VBCe VT 1; ðe VT 1Þ e VT .
We get the following currents expressions:
VBE VBC VBE VBC
IC af Ise e VT Isc e VT ; IE Ise e VT ar Isc e VT
VBE VBC
IB ð1 af Þ Ise e VT þ ð1 ar Þ Isc e VT
Fig. C.4 The cutoff mode of operation as equivalent to a leaky switch
In this case, with both junctions forward biased.
VBE 0:8 V; VBC 0:7 V; VCE ¼ VCB þ VBE ; VCB ¼ VBC ; VCE ¼ VBE VBC
¼ 0:1 V
There is a 0.1 V drop across the transistor from collector to emitter which is
quite low while a substantial current flows through the device. In this mode it can be
considered as having a very high conductivity and acts as a closed switch with a
finite resistance and conductivity (Fig. C.5).
BJT transistor avalanche breakdown region of operation:
An avalanche transistor is a bipolar junction transistor designed for operation in
the region of its collector-current/collector-to-emitter voltage characteristics beyond
the collector-to-emitter breakdown voltage, called avalanche breakdown region.
This region characterized by avalanche breakdown, a phenomenon similar to
Negative Differential Resistance (NDR). Operation in the avalanche breakdown
region is called avalanche-mode operation. It gives avalanche transistors the ability
Fig. C.5 Saturation mode of operation equivalent to a closed switch

to switch very high currents with less than nanosecond rise and fall times (transition
times). Transistors not specifically designed for the purpose can have reasonably
consistent avalanche properties.
Static avalanche regioncharacteristics: The static characteristic of an avalanche
transistor is IC-VCE. The static characteristic of an avalanche NPN transistor is the
same as PNP devices only changing sign to voltages and currents accordingly. The
avalanche breakdown multiplication is present only across the collector-base
junction. The first step of the calculation is to determine collector current as a sum
of various component currents through the collector since only those fluxes of
charge are subject to this phenomenon. Applying Kirchhoff’s current law (KCL) to
a bipolar junction transistor, implies the following relation which satisfied by the
collector current IC (IC = IE − IB) while for the same device working in the active
region. a ¼ af ; b ¼ bf ; IC ¼ b IB þ ðb þ 1Þ ICBo , IB is the base current. ICBo is the
collector-base reverse leakage current. IE is the emitter current. B is the common
emitter current gain of the transistor. Equating the two formulas for IC gives the
following result
IE ¼ ðb þ 1Þ IB þ ðb þ 1Þ ICBo and since a ¼ b þb 1; a is the common base
current gain of the transistor, then
a IE ¼ b IB þ b ICBo ¼ IC ICBo ) IC ¼ a IE þ ICBo :
When the avalanche effects in a transistor collector are considered, the collector
current IC is given by IC ¼ M ða IE þ ICBo Þ. M is miller’s avalanche multipli-
cation coefficient. It is the most important parameter in avalanche mode operation
M¼ 1
VCB n . BVCBo is the collector-base breakdown voltage. n is a constant
1ðBV Þ
CBo
depending on the semiconductor used for the construction of the transistor and
doping profile of the collector-base junction. VCB is the collector-base voltage.
Using Kirchhoff’s current law (KCL) for the bipolar junction transistor and the
expression for M, the resulting expression for IC is the following:
M a IB þ ICBo
IC ¼ ða IB þ ICBo Þ ) IC ¼
ð1 a MÞ 1 a ðBV VCB n
Þ
CBo
VCB ¼ VCE VBE ; VBE ¼ VBE ðIB Þ
where VBE is the base-emitter voltage.
a IB þ ICBo a IB þ ICBo
IC ¼ VBE ðIB Þ n
’ VCE n
1a ðVCEBV Þ 1 a ðBV CBo
Þ
CBo
Since VCE VBE . This is the expression of the parametric family of the col-
lector characteristics IC–VCE with parameter IB (IC) increases without limit if
VCE n pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi BVCBo

ð1 aÞ ¼ ð Þ ) VCE ¼ BVCEo ¼ BVCBo n ð1 aÞ ¼ p ffiffiffiffiffiffiffiffiffiffiffi ;
BVCBo n
bþ1
b 1
1a¼1ð Þ¼
bþ1 bþ1
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi BVCBo
b 1 ) VCE ¼ BVCEo ¼ BVCBo n ð1 aÞjb 1 ¼ p n
ffiffiffi :
b
where BVCEo is the collector-emitter breakdown voltage.

Avalanche Multiplication: The maximum reverse biasing voltage which may be
applied before breakdown between the collector and base terminals of the transistor,
under the condition that the emitter lead be open circuited, is represented by the
symbol BVCBo. This breakdown voltage is a characteristic of the transistor alone.
The breakdown may occur because of avalanche multiplication of the current ICO
that crosses the collector junction. As a result of this multiplication, the current
becomes M ICO , in which M is the factor by which the original ICO is multiplies by
the avalanche effect. It is possible to neglect leakage current, which does not flow
through the junction and is therefore not subject to avalanche multiplication. At a
high enough BVCBo, the multiplication factor M becomes nominally infinite and the
region of breakdown is then attained. The current rises abruptly, and large changes
in current accompany small changes in applied voltage. The avalanche multipli-
cation factor depends on the voltage VCB between transistor’s collector and base. If
a current IE is caused to flow across the emitter junction, then, neglecting the
avalanche effect, a fraction a IE , where a is the common-base current gain, reaches
the collector junction. If we take multiplication into account, IC has the magnitude
M a IE . In presence of avalanche multiplication, the transistor behaves as though
its common base current gain where M a. The maximum allowable collector to
emitter voltage depends not only upon the transistor, but also upon the circuit in
which it is used.
BJT transistor second breakdown avalanche mode: When the collector current
rises above the data sheet limit ICmax a new breakdown mechanism happened, the
second breakdown. This phenomenon is caused by excessive heating of some
points (hot spots) in the base-emitter region of the bipolar junction transistor, which
give rise to an exponentially increasing current through these points. This expo-
nential rise of current in turn gives rise to even more overheating, originating a
positive thermal feedback mechanism. While analyzing the IC-VCE static charac-
teristic, the presence of this phenomenon is seen as a sharp collector voltage drop
and a corresponding almost vertical rise of the collector current. While this phe-
nomenon is destructive for bipolar junction transistors working in the usual way, it
can be used to push up further the current and voltage limits of a device working in
avalanche mode by limiting its time duration. The switching speed of the device is
not negatively affected.
Small signal model of the BJT, base charging capacitance (diffusion capaci-
tance). In active mode when the emitter-base is forward biased, the capacitance of
the emitter-base junction is dominated by the diffusion capacitance (not depletion
capacitance). Recall for a diode we define the following: CDiffusion ¼ dQ
dv0 ¼ dt dv0 .
D dQD dt
D D
The sum up all minority carrier charges on either side of the junction.
Z1 v0
Z1 v0
pno ðeVT 1Þ eLP dX þ q A

X
npo ðeVT 1Þ eLn dX
D D X
QD ¼ q A
0 0
If we neglect charge injected from the base into the emitter due to p+ emitter in
R1 v0
D X
PNP then QD ¼ q A 0 pno ðeVT 1Þ e LP dX. Excess charge stored is due
almost entirely to the charge injected from the emitter. The BJT acts like a very
efficient “siphon”; As majority carriers from the emitter are injected into the base
and become “excess minority carriers”, the collector “siphons them” out of the
base. We can view the collector current as the amount of excess charge in the base
collected by the collector per unit time and we can express the charge due to the
excess hole concentration in the base as: QB ¼ ic sF or the excess charge in the
base depends on the magnitude of current flowing and the “forward” base transport
W2
time, sF , the average time the carriers spend in the base. sF ¼ 2D B
, W is the base
quasi-neutral region width. DBis the minority carrier diffusion coefficient. Thus, the
diffusion capacitance is
@QB W2 @ic IC
CB ¼ jQpoint ¼ ð Þ j ; C B ¼ sF ¼ s F gm :
@vBE 2 DB @vBE Qpoint VT
The upper operational frequencyof the transistor is limited by the forward base
transport time f 2ps
1
F
. It is the similarity to the diode diffusion capacitance.
jpno Lp þ npo Ln j q A
CDiffusion ¼ gd st ; st ¼ ;
IS
jpno Lp þ npo Ln j q A
CDiffusion ¼ gd
IS
st is the transit time. In active mode for small forward biases the depletion
capacitance of the base-emitter junction can contribute to the total capacitance.
CjEo
CjE ¼ qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
1 þ Vbi for Vemitterbase
EB
CjE zero bias depletion capacitance.

Vbi for emitter-base built in voltage for E-B junction. Thus, the emitter-base
capacitance is Cp = CB + CjE. In active mode when the collector-base is reverse
biased, the capacitance of the collector-base junction is dominated by the depletion

capacitance (not diffusion capacitance).
Clo
Cl ¼ qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi :
1 þ Vbi for collectorbase
VCB
Clo zero bias depletion capacitance.

Vbi for collector-base built in voltage for the B-C junction. In some integrated
BJTs (lateral BJTs in particular) the device has a capacitance to the substrate wafer
it is fabricated in. This results from a “buried” reverse biased junction. Thus, the
collector-substrate junction is reverse biased and the capacitance of the
collector-substrate junction is dominated by the depletion capacitance (not diffusion
capacitance).
Ccso
Ccs ¼ qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi :
1 þ Vbi for collectorsubstrate
Vcs
Ccs zero bias depletion capacitance.

Vbi for collector-substrate built in voltage for the C substrate junction.
Small signal model of the BJT, parasitic resistances:
rb base resistance between metal inter connect and B-E junction.
rc parasitic collector resistance.
rex emitter resistance due to polysilicon contact.
Complete BJT small signal model: (Fig. C.6).

What set the maximum limits of operation of the BJT circuit? Forward active
mode lies between saturation and cutoff. Thus, the maximum voltage extremes that
one can operate an amplifier over can easily be found by examining the boundaries
between forward active and cutoff and the boundaries between forward active and
saturation. Output signals that exceed the voltage range that would keep the tran-
sistor within its forward active mode will result in “clipping” of the signal leading
to distortion. The maximum voltage swing allowed without clipping depends on the
DC bias points.
MOSFET transistor model:
The basic static model of MOSFET transistor (Shichman and Hodges) is as
follow (Fig. C.7):
2

W VDS
IDS ¼ l0 Cox ðVGS VTH Þ VDS
Leff 2
1 W 2
IDSsat ¼ l0 Cox ðVGS VTH Þ
2 Leff
Fig. C.6 Complete BJT small signal model
Fig. C.7 MOSFET transistor

model graph
There is an empirical correction to these equations to account for the channel

length modulation:
W V2
IDS ¼ l0 Cox ½ðVGS VTH Þ VDS DS ½1 þ k VDS
Leff 2
1 W
IDSsat ¼ l0 Cox ðVGS VTH Þ2 ½1 þ k VDS
2 Leff
In the linear region:

2

W VDS
IDS ¼ KP ðVGS VTH Þ VDS ½1 þ k VDS
ðL 2 Xjl Þ 2
In the saturation region:
KP W
IDsat ¼ ðVGS VTH Þ2 ½1 þ k VDS
2 ðL 2 Xjl Þ
Xjl is the lateral diffusion parameter (Fig. C.8).

Threshold voltage (VTH): The threshold voltage changes with changes in
body-source voltage, VBS. The expression for threshold voltage
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi qffiffiffiffiffiffiffiffiffiffiffi
VTH ¼ VTO þ c ð 2 /p VBS 2 /p Þ
where VTO is the threshold voltage when the body-source voltage is zero, c is the
body effect parameter and Up is the surface inversion potential. If the bulk is
connected to the source (i.e. the MOSFET is acting as a 3 terminal device, the
threshold voltage is always equal to the value VTO). There is a depletion layer
which grows into the accumulation region and thus for a given VGS, cuts off the
channel. Need to add more VGS to re-establish the channel when we stacked
transistors in integrated circuits. If you connected bulk to source on each transistor
in an integrated circuit you would end up shorting many points in the circuit to
ground.
Fig. C.8 MOSFET transistor

structure and important
parameters
Complete DC model:
The model includes body-source and body-drain diodes. Equations used for the
diode model, for forward bias on the body-source/body-drain diodes (Fig. C.9):
VBS
h VBD i
IBS ¼ ISS ½e Vt 1 þ GMIN VBS ; IBD ¼ ISD e Vt 1 þ GMIN VBD
For the negative reverse bias on those diodes:
VBS VBD
IBS ¼ ISS þ GMIN VBS ; IBD ¼ ISD þ GMIN VBD
Vt Vt
MOSFET body diodes: The reverse bias terms are simply the first terms in a
power series expansion of the exponential term. The GMIN convergence resistance.
ISS and ISD are taken to be one constant in simulation.
DC MOSFET parameters: L = channel length, W = channel width, KP
(kp) = The trans-conductance parameter, VTO = Threshold voltage under zero bias
conditions, GAMMA (c) = Body effect parameter, PHI(Up) = surface inversion
potential, RS(RS) = source contact resistance, RD(RD) = Drain contact resistance,
LAMBDA(k) = channel length modulation parameter, XJ(Xjl) = lateral diffusion
parameter. IS(ISS, ISD) = reverse saturation current of body-drain/source diodes.
Large signal transient model: We add some capacitances to the DC model to
create the transient model to form the final transient model, as shown in Fig. C.10).
Capacitances: Static overlap capacitances between gate and drain (CGB0), gate
and source (CGS0), and gate and bulk (CGB0). These are fixed values, and are
specified per unit width. In saturation,
2
CGS ¼ C0x þ CGS0 W; CGD ¼ CGD0 W
3
Fig. C.9 MOSFET complete DC model

Fig. C.10 MOSFET large signal transient model
In saturation after pinch-off, it is assumed that altering the drain voltage does not
have any effect on stored charge in the channel and thus the only capacitance
between gate and drain is the overlap capacitance.
In the linear/triode region, in this region the following equations are used:
( 2 )
VGS VDS VTH
CGS ¼ C0x 1 þ CGS0 W
2 ðVGS VTH Þ VDS
( 2 )
VGS VTH
CGD ¼ C0x 1 þ CGD0 W
2 ðVGS VTH Þ VDS
As the device is moved further into the linear region, VGS becomes large
compared to (VDS–VTH) then the values of CGS and CGD become close to C0x/2
(plus the relevant overlap capacitance).
The body diode capacitances: The capacitances of the body diodes are given by
slightly modified expressions for junction capacitances of the diode model: The
C ð0Þ
expression for a PN diode capacitance: Cj ¼ pjffiffiffiffiffiffiffiVffi. The MOSFET equation is
1V
0
based on the following slightly modified equation:
Cj ð0Þ CjSW ð0Þ

Cj ¼ qffiffiffiffiffiffiffiffiffiffiffiffi þ qffiffiffiffiffiffiffiffiffiffiffiffi :
1 VV0 1 VV0
The junction capacitance is made up of two components. The main component, due
to Cj(0) is the normal junction capacitance. The second parameter is the perimeter
junction capacitance of the diffused source. The diffusion capacitance is zero in

reverse bias and the MOSFET must be operated with the bulk-drain and
bulk-sourceBulk-source diodes in reverse bias to stop large bulk currents flowing.
The additional parameters required for specifying the transient model in addition to
those required by the DC model are thus:
CGD0(CGD0) = Gate drain overlap capacitance per unit width of device.
CGS0(CGS0) = Gate source overlap capacitance per unit width of device.
CJ(Cj) = Zero bias depletion capacitance for body diodes.
CJSW(Cjsw) = Zero bias depletion perimeter capacitance for body diodes.
TOX(tox) = Oxide thickness (used for calculating Cox).
Bipolar transistor metrology and theory:
The interest topics regarding bipolar junction transistor (BJT) are operation, I-V
characteristics, current gain and output conductance. High level injection and heavy
doping induced band narrowing. SiGe transistor, transit time, and cutoff frequency
are important parameters. There are several bipolar transistor models which are
used (Ebers-Moll model, Small signal model, and charge control model). Each
model has its own areas of applications. The metal-oxide-semiconductor
(MOS) ICs have high density and low power advantages. The BJTs are preferred
in some high frequency and analog applications because of their high speed, low
noise, and high output power advantages such as in some cell phone amplifier
circuits. A small number of BJTs are integrated into a high density complementary
MOS (CMOS) chip integration of BJT and CMOS is known as the BiCMOS
technology. The term bipolar refers to the fact that both electrons and holes are
involved in the operation of a BJT. Minority carrier diffusion plays the leading role
as in the PN diode junction diode. A BJT is made of a heavily doped emitter, a
P-type base, and an N-type collector. This device is an NPN BJT, a PNP BJT would
have a P+ emitter, N-type base, and P-type collector. NPN transistor exhibit higher
trans conductance and speed than PNP transistors because the electron mobility is
larger than the hole mobility, BJTs are almost exclusively of the NPN type since
high performance is BJT’s competitive edge over MOSFETs (Fig. C.11).
When the base-emitter junction is forward biased, electrons are injected into the
more lightly doped base. They diffuse across the base to the reverse biased
base-collector junction which is the edge of the depletion layer and get swept into
the collector. This produces a collector current, IC. IC is independent of VCB as long
as VCB is a reverse bias or a small forward bias. IC is determined by the rate of
electron injection from the emitter into the base, determined by VBE. The rate of
qVBE
electron injection is proportional to e kT . The emitter is often connected to ground.
The emitter and collector are the equivalents of source and drain of a MOSFET
when the base is the equivalent of the gate. The ICcurve is usually plotted against
VCE. VCE = VCB + VBE, below VCE = 0.3 V the base-collector junction is strongly
forward biased and IC decreases. Because of the parasitic IR drops, it is difficult to
accurately ascertain the true base-emitter junction voltage. The easily measurable
Fig. C.11 NPN BJT transistor voltages connection
base current IB is commonly used as the variable parameter in lieu of VBE, IC is

proportional to IB (Fig. C.12).
Collector current: The collector current is the output current of a BJT transistor.
Applying the electron diffusion equation to the base region gives in Fig. C.13
d 2 n0 n0 pffiffiffiffiffiffiffiffiffiffiffiffiffiffi d 2 n0 n0
¼ 2 ; LB ¼ sB DB ; 2 ¼ :
dx 2 LB dx sB DB
Fig. C.12 NPN transistor structure, connections and graphs

Fig. C.13 NPN transistor structure with depletion layers
sB and DB are the recombination lifetime and the minority carrier (electron)
diffusion constant in the base, respectively. The boundary conditions are as follow:
qVBE qVBC
n0 ð0Þ ¼ nB0 ðe kT 1Þ; n0 ðWB Þ ¼ nB0 ðe kT 1Þ nB0 0:
n2
where nB0 ¼ NiB and NB is the base doping concentration. VBE is normally a
forward bias (positive value) and VBC is a reverse bias (negative value).
WB x
sinhð LB Þ
qVBE
We get the following expression for n0 ðxÞ ¼ nB0 ðe kT 1Þ W . Modern
sinhð L B Þ
B
BJTs have base widths of about 0.1 lm. This is much smaller than the typical
diffusion length of tens of microns. In the case of WB LB we get the n0 ðxÞ
n2 qVBE
expression: n0 ðxÞ ¼ n0 ð0Þ ð1 WxB Þ ¼ NiBB ðe kT 1Þ ð1 WxB Þ. niB is the intrinsic
carrier concentration of the base material. The subscript B, is added to ni because
the base may be made of a different semiconductor such as SiGe alloy, which has a
smaller band gap and therefore a larger ni than the emitter and collector material.
The minority carrier current is dominated by the diffusion current. The sign of IC is
positive and defined in the expression:

dn n0 ð0Þ DB n2iB

qVBE
IC ¼ AE q DB ¼ AE q DB ¼ AE q ðe kT 1Þ:
dx WB WB N B
AE is the area of the BJT specifically the emitter area. There is a similarity between
BJT transistor IC current and the PN diode IV relation. Both are proportional to
qV Dn2i dn0
ðekT 1Þ and to N . The only difference is that dx has produced the W1B term due to the
qVBE
linear n′ profile. We can condense the expression of IC to IC ¼ Is ðe kT 1Þ, where
qn2i qVBE n2
Is is the saturation current. IC ¼ AE GB ðe kT 1Þ and GB ¼ n2i DNBB WB
iB
n2i
¼ n2iB
p
DB WB , where p is the majority carrier concentration in the base. It is valid even
for no uniform base and high level injection condition if Gb is generalized to 1.

R W n2
GB ¼ 0 B n2i DpB dx, GB has the unusual dimension of s/cm4 and is known as the base
iB
Gummel number. In the special case of niB ¼ ni , DBis a constant, and pðxÞ ¼ NB ðxÞ
RW
which is low level injection. GB ¼ D1B 0 B NB ðxÞ dx ¼ D1B base dopant atoms
per unit area. The base Gummel number is basically proportional to the base dopant
density per area. The higher the base dopant density is, the lower the IC will be for a
given VBE. The concept of a Gummel number simplifies the IC model because it
contains all the subtleties of transistor design that affect IC; changing base material
through niB ðxÞ, non-constant DB, non-uniform base dopant concentration through
pðxÞ ¼ NB ðxÞ and even the high level injection condition, where p > NB. Although
many factors affect GB, GB can be easily determined from the Gummel plot. The
inverse slope of the straight line can be described as 60 mV per decade. The
extrapolated intercept of the straight line and VBE = 0 yields Is. GB is equal to AE
q n2i divided by the intercept (Fig. C.14).
The decrease in the slope of the curve at high IC is called the high level injection
effect. At large VBE, n′ can become larger than the base doping concentration NB,
n0 ¼ p0 NB . The condition of n0 ¼ p0 NB is called high level injection.
A consequence is that in the base
qVBE qVBE
n p ni e 2kT ; GB / ni e 2kT
qVBE qVBE
Yield to Ic / ni e 2kT . Therefore, at high VBE or high IC, Ic / e 2kT and the
inverse slope becomes 120 mV/decade. IKF, the knee current, is the current at
which the slope changes. It is a useful parameter in the BJT model for circuit
simulation. The IR drop in the parasitic resistance significantly increases VBE at
very high IC and further flattens the curve.
Base current: Whenever the base-emitter junction is forward biased, some holes
are injected from the P-type into the N+ emitter. These holes are provided by the
base current IB, IB is an undesirable but inevitable side effect of producing IC by
Fig. C.14 NPN transistor Ic

[A] versus VBE [volt]
forward biasing the BE junction. The analysis of IB, the base to emitter injection
current, is a perfect parallel of the IC analysis. The base current can be expressed as
ZWE
q n2i qVBE n2i p
IB ¼ AE ðe kT 1Þ; GE ¼ dx:
GE n2iE DE
0
GE is the emitter Gummel number. In case of uniform emitter, where niE ; NE

(emitter doping concentration) and DE are not functions of x (Fig. C.15).
DE n2iE qVBE
IB ¼ A E q ðe kT 1Þ:
WE NE
Current gain: The most important DC parameter of a BJT is its common emitter
current gain bF. Another current ratio, the common base current gain, is defined by
aF. bF IICB ; IC ¼ aF IE ; aF ¼ IICE ¼ IC IþC IB ¼ ICI=IC =I B
B þ1
¼ 1 þbFb . aF is typically very
F
close to unity, such as 0.99, because bF is large. aF ¼ 1 þbFb ;

F
bF ¼ 1a
aF
F
. I B is a load on the input signal source, an undesirable side effect of
forward biasing the BE junction. IB should be minimized (bF should be maxi-
D W N n2
mized). bF ¼ GGEB ¼ DBE WEB NEB n2iB . A typical good bF is 100. D and W cannot be
iE
changed very much. The most obvious way to achieve a high bF, is to use a large
NE and a small NB. A small NB, would introduce too large a base resistance, which
degrades the BJT’s ability to operate at high current and high frequencies. Typically
NB is around 1018 cm−3. An emitter is said to be efficient if the emitter current is
mostly the useful electron current injected into the base with little useless hole
current (the base current). The emitter efficiency is defined as
cE ¼ IE II
E
B
¼ IC IþC IB ¼ 1 þ G1B =GE . To raise bF, NE is typically made larger than
1020 cm−3. When NE is very large, n2iE becomes larger than n2i . This is called the
Eg
heavy doping effect. n2i ¼ NC NV ekT , heavy doping can modify the Si crystal
sufficient to reduce Eg and cause n2i to increase significantly. Therefore, the heavy
DEgE
doping effect is also known as band gap narrowing. n2iE ¼ n2i e kT , DEgE is the
Fig. C.15 NPN transistor structure electron flow and hole flow
narrowing of the emitter band gap relative to lightly doped Si and is negligible for
NE < 1018 cm−3, 50 meV at 1019 cm−3, 95 meV cm−3 at 1020 cm−3, and 140 meV
at 1021 cm−3. To further elevate bF, we can raise niB by using a base material that
has a smaller band gap than the emitter material. Si1g Geg is an excellent base
material candidate for an Si emitter. With η = 0.2, EgB is reduced by 0.1 eV. In a
SiGe BJT, the base is made of high quality P-type epitaxial SiGe. In practice, η is
graded such that η = 0 at the emitter end of the base and 0.2 at the drain end to
create a built in field that improves the speed of the BJT. Because the emitter and
base junction is made of two different semiconductors, the device is known as a
heterojunction bipolar transistor or HBT. HBTs made of InP emitter (Eg = 1.35 eV)
and InGaAs base (Eg = 0.68 eV) and GaAlAs emitter with GaAs base are other
examples of well-studied HBTs. The ternary semiconductors are used to achieve
lattice constant matching at the heterojunction. Whether the base material is SiGe or
plain Si, a high performance BJT would have a relatively thick (>100 nm) layer of
As doped N+ poly-Si film in the emitter. Arsenic is thermally driven into the “base”
by *20 nm and converts that single crystalline layer into a part of the N+ emitter.
This way, bF is larger due to the large WE, mostly made of the N+ poly-Si. This is
the poly-Silicon emitter technology. The simpler alternative, a deeper implanted or
diffused N+ emitter without the poly-Si film, is known to produce a higher density
of crystal defects in the thin base causing excessive emitters to collector leakage
current or even shorts in a small number of the BJTs. High speed circuits operate at
high IC, and low power circuits may operate at low IC. Current gain b, drops at both
high IC and at low IC. In Gummel plot the IC flattens at high VBE due to the high
level injection effect in the base. That IC curve arising from hole injection into the
emitter, does not flatten due to this effect because the emitter is very heavily doped,
and it is practically impossible to inject a higher density of holes than NE. Over a
wide mid-range of IC, IC and IB are parallel, indicating that the ratio IC/IB, i.e., bF is
a constant. Above 1mA, the slope of IC drops due to high level injection.
Consequently, the IC/IB ratio or bF decreases rapidly. This fall-off of current gain
unfortunately degrades the performance of BJTs at high current where the BJTs
speed is the highest. IB is the base emitter junction forward bias current. The
forward bias current slope decreases at low VBE or very low current due to the
Space Charge Region (SCR) current. As a result, the IC/IB ratio or bF decreases at
very low IC.
@IC
As in MOSFETs, a large output conductance, @V CE
, of BJTs is deleterious to the
voltage gain of circuits. The cause of the output conductance is base-width modu-
lation. The thick vertical line indicates the location of the base-collector junction.
With increasing Vce, the base-collector depletion region widens and the neutral base
width decreases. This leads to an increase in IC. If the curves IC–VCE are extrapo-
lated, they intercept the IC = 0 axis at approximately the same point. VA is defined as
early voltage. VA is a parameter that describes the flatness of the IC curves.
@IC 1
Specifically, the output resistance can be expressed as VA/IC: r0 ð@V CE
Þ ¼ VICA .
A large VA (large r0) is desirable for high voltage gains. A typical VA is 50 V. VA is
sensitive to the transistor design. We can except VA and r0 to increase, expect the
base width modulation to be a smaller fraction of the base width, if we increase the
base width, increase the base doping concentration NB or decrease the collector
doping concentration NC. Increasing the base width would reduce the sensitivity to
any given DWB. Increasing the base doping concentration NB would reduce the
depletion region thickness on the base side because the depletion region penetrates
less into the more heavily doped side of a PN junction. Decreasing the collector
doping concentration NC would tend to move the depletion region into the collector
and thus reduce the depletion region thickness on the base side, too. Both increasing
the base width and the base doping concentration NB would depress bF. Decreasing
the collector doping concentration NC is the most acceptable course of action. It is
also reduces the base-collector junction capacitance, which is a good thing.
Therefore, the collector doping is typically ten times lighter than the base doping.
The larger slopes at VCE > 3v are caused by impact ionization. The rise of IC due to
base-width modulation is known as the early effect. Model the collector current as a
function of the collector voltage: IC ¼ bF IB and differentiating with respect to VC
@IC @bF
while IB was held constant gave, @V C
¼ IB @V C
. The question is how can bF change
with VC, the collector depletion layer thickens as collector voltage is raised. The base
gets thinner and current gain raises. Bipolar transistor transit time and charge stor-
age: Static IV characteristics are only one part of the BJT theory. Another part is its
dynamic behavior or its speed. When the BE junction is forward biased, excess holes
are stored in the emitter, the base, and even the depletion layers. The sum of all
excess hole charges everywhere QF. QF is the stored excess carrier charge. If
QF = 1pC (Pico coulomb), there is +1 pC of excess hole charge and −1 pC of excess
electron charge stored in the BJT. The ratio of QF to IC is called the forward transit
time sF (sF QICF ). Ic and QF are related by a constant ratio sF. QF and therefore sF are
very difficult to predict accurately for a complex device structure. sF can be measured
experimentally and once sF is determined for a given BJT, equation sF QICF
becomes a powerful conceptual and mathematical tool giving QF as a function of IC,
and vice versa. sF sets a high frequency limit of BJT operation. The excess hole
W2
charge in the base QFB: QFB ¼ q AE n0 ð0Þ WB =2; QICFB sFB ¼ 2DBB . The base
transit time can be further reduced by building into the base a drift field that aids the
flow of electrons from the emitter to the collector. There are two ways of accom-
plishing this. The classical method is to use graded base doping (a large NB near the
EB junction), which gradually decreases toward the CB junction. Such a doping
gradient is automatically achieved if the base is produced by dopant diffusion. The
changing NB creates a dEv =dx and a dEc =dx. This means that there is a drift field.
Any electron injected into the base would drift toward the collector with a base
W2
transit time shorter than the diffusion transit time, 2DBB . In a SiGe BJT, P-type
epitaxial Si1g Geg is grown over the Si collector with a constant NB and η linearly
varying from about 0.2 at the collector end to 0 at the emitter end. A large dEc =dx can
be produced by the grading of EgB . These high speed BJTs are used in high fre-
quency communication circuits. Drift transistors can have a base transit time several
W2
times less than 2DBB , as short as 1psec. The total forward transit time, sF is known as
the emitter to collector transit time. sFB is only one portion of sF. The base transit
time typically contributes about half of sF. To reduce the transit (or storage) time in
the emitter and collector, the emitter and the depletion layers must be kept thin. sF
can be measured. sF starts to increase at a current density where the electron density
corresponding to the dopant density in the collector (n = NC) is insufficient to
support the collector current even if the dopant induced electrons move at the
saturation velocity. This intriguing condition of too few dopant atoms and too much
current is lead to a reversal of the sign of the charge density in the depletion region.
IC d}ðxÞ q
IC ¼ AE q n vsat ; q ¼ q NC q n ¼ q NC ; ¼ :
AE vsat dx es
when IC is small then qðq ¼ q NC Þ as expected from the PN junction analysis,

and the electric field in the depletion layer. The N+ collector is always present to
reduce the series resistance. No depletion layer is shown in the base for simplicity
because the base is much more heavily doped than the collector. As IC increases, q
decreases and d}ðxÞ
dx decreases. The electric field drops to zero in the very heavily
doped N+ collector as expected. Because of the base widening, sF increases as a
consequence. This is called the Kirk effect. Base widening can be reduced by
increasing NC and VCE. The Kirk effect limits the peak BJT operating speed.
Bipolar transistor small signal model: The equivalent circuit for the behavior of a
BJT in response to a small input signal (10 mV sinusoidal signal, superimposed on
the DC bias) is presented in Fig. C.16. BJTs are often operated in this manner in
analog circuits.
qVBE
If VBE is not close to zero, the “1” in IC ¼ Is ðe kT 1Þ is negligible; in that
case
qVBE qVBE
IC ¼ Is ðe kT 1Þ Is e kT :
Fig. C.16 Bipolar transistor small signal model

When a signal vBE is applied to the BE junction, a collector current gm vBE is

produced. gm, the trans-conductance, is
dIC d qVBE q qVBE kT kT

gm ¼ ðIs e kT Þ ¼ Is e kT ¼ IC = ; gm ¼ I C =
dVBE dVBE kT q q
At room temperature, gm ¼ IC =26 mV. The trans-conductance is determined by

the collector bias current, IC. The input node, the base, appears to the input drive
circuit as a parallel RC circuit. r1p ¼ dV dIB
BE
¼ b1 dV
dIC
BE
¼ gbm ; rp ¼ bgmF . QF is the excess
F F
carrier charge stored in the BJT. If QF = 1 pC, there is +1pC of excess holes and
−1pC of excess electrons in the BJT. All the excess hole charge, QF, is supplied by
the base current, IB. Therefore, the base presents this capacitance to the input drive
circuit: Cp ¼ dV
dQF
BE
¼ dVdBE ½sF IC ¼ sF gm . The capacitance CP may be called the
charge storage capacitance, known as the diffusion capacitance. There is one charge
component that is not proportional to IC and therefore cannot be included in QF.
That is the junction depletion layer charge. Therefore, a complete model of CP
should include the BE junction depletion layer capacitance, CdBE,
Cp ¼ sF gm þ CdBE . Once the parameters in the basic small signal model of the
BJT have been determined, one can use the small signal model to analyze circuits
with arbitrary signal source impedance network which composing resistors,
capacitors, and inductors, and additionally load impedance network. r0 is the in-
trinsic output resistance, VA/IC. Cl also arises from base width modulation; when
VBC varies, the base width varies; therefore, the base stored charge varies, thus
giving rise to Cl ¼ dQ dVCB . CdBC is the CB junction depletion layer capacitance.
FB
Model parameters are difficult to predict from theory with the accuracy required for
commercial circuit design. Therefore, the parameters are routinely determined
through comprehensive measurement of the BJT AC and DC characteristics.
Figure C.17 describes the small signal model which can be used to analyze a
BJT circuit by hand.
Cutoff frequency: We consider small signal model when the load is a short
circuit. The signal source is a current source ib, at a frequency f. The question is at
what frequency the AC current gain does b ic =ib fall to unity?
ib ib
vbe ¼ ¼ ; ic ¼ gm vbe
input admittance 1=rp þ j x Cp

ic gm 1
bðxÞ ¼ ¼
ib j1=rp þ j x Cp j j1=gm rp þ j x sF þ j x CdBE =gm j
1
¼
j1=bF þ j x sF þ j x CdBE k T=q IC j
Fig. C.17 Bipolar transistor small signal model which can be used to analyze a BJT circuit by
hand
Fig. C.18 Bipolar transistor small signal model for circuit simulation
Figure C.18 describes the small signal model for circuit simulation by computer.
At x = 0, i.e., DC b(x) = … equation reduces to bF as expected. As x increases,
b drops. By carefully analyzing the b(x) data, one can determine sF. If bF 1 so
that 1/bF is negligible, bðxÞ / x1 and b = 1 at fT , fT ¼ 2pðsF þ CdBE 1
kT=qIC Þ.
If we use a more complete small signal model, it can be shown that
fT ¼ 2p½sF þ ðCdBE þ CdBC ÞkT=ðqI
1
C Þ þ CdBC ðre þ rc Þ
. fT is the cutoff frequency and is com-
monly used to compare the speed of transistors. The above equations predict that fT
rises with increasing IC due to increasing gm, in agreement with the measured fT. At
very high IC, sF increases due to base widening (Kirk effect), and therefore, fT falls.
BJTs are often biased near the IC where fT peaks in order to obtain the best high
frequency performance. FT is the frequency of unity power gain. The frequency of
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
unity power gain, called the maximum oscillation frequency. fmax ¼ ð8prfbTCdBC Þ, it
is therefore important to reduce the base resistance, rb. While MOSFET scaling is
motivated by the need for high packing density and large Idsat, BJT scaling is often
motivated by the need for high fT and fmax. This involves the reduction of sF (thin
base, etc,) and the reduction of parasitic (CdBE ; CdBC ; rb ; re ; rc ). We interested in
BJT with poly-Si emitter, self-aligned base, and deep trench isolation. The base is
contacted through two small P+ regions created by boron diffusion from a P+
poly-Si film. The film also provides a low resistance electrical connection to the
base without introducing a large P+ junction area and junction capacitance. To
minimizing the base series resistance, the emitter opening is made very narrow. The
lightly doped epitaxial N-type collector is contacted through a heavily doped
sub-collector in order to minimize the collector series resistance. The substrate is
lightly doped to minimize the collector capacitance. Both the shallow trench and the
deep trench are filled with dielectrics (SiO2) and serve the function of electrical
isolation. The deep trench forms a rectangular moat that completely surrounds the
BJT. It isolates the collector of this transistor from the collectors of neighboring
transistors. The structure incorporates many improvements that have been devel-
oped over the past decades and have greatly reduced the device size from older BJT
design. BJT is a larger transistor than a MOSFET.
Bipolar transistor charge control model: The small signal model is ideal for
analyzing circuit response to small sinusoidal signals. If the signal is large, input is
step function IB switching from zero to 20 lA or by any IB(t) and then IC(t) is
produced. The response is analyzed with the charge control model which is a simple
extension of the charge storage concept. IC ¼ QsFF ) IC ðtÞ ¼ QsFFðtÞ, IC(t) becomes
known if we solve for QF ðtÞ. sF has to be characterized beforehand for the BJT being
used. IC is controlled by QF (charge control model). At DC condition IB ¼ bIC ¼ sFQbF ,
F F
the equation has a straightforward physical meaning. In order to sustain a constant
excess hole charge in the transistor, holes must be supplied to the transistor through
IB to replenish the holes that are lost to recombination. Therefore, DC IB is pro-
portional to QF. When holes are supplied by IB at the rate of QF =sF bF , the rate of
hole supply is exactly equal to the rate of hole loss to recombination and QF remains
at a constant value. In the case that IB is larger than QF =sF bF .
(IB [ QF =sF bF ), holes flow into the BJT at a higher rate than the rate of hole
loss and the stored hole charge QF increases with time (dQ dt ¼ IB ðtÞ sF bF ).
F QF
The presented equations together constitute the basic charge control model.
For any given IB(t), equation dQ dt ¼ IB ðtÞ sF bF can be solved for QF(t) ana-
F QF
lytically or by numerical integration. Once QF(t) is found, IC(t) becomes known

from equation IC ðtÞ ¼ QsFFðtÞ. Figure C.19 describes the charge control model. Excess
hole charge QF rises or falls at the rate of supply current IB minus loss (/QF ).
QF is the amount of charges in the vessel, and sFQbF is the rate of charge leakage.
F
IB is the rate of charges flowing into the vessel. The above figure is a basic version
of the charge control model. We can introduce the junction depletion layer
Fig. C.19 Bipolar transistor IB(t) QF/(τ F·βF)

charge controlmodel QF(t)
capacitances into equation dQ dt ¼ IB ðtÞ sF bF . Diverting part of IB to charge the

F QF
junction capacitances would produce an additional delay in IC(t).

Bipolar transistor model for large signal circuit simulation: The BJT model used
in circuit simulation can accurately represent the DC and dynamic currents of the
transistor in response to VBE(t) and VCE(t). A typical circuit simulation model or
compact model is made of the Ebers-Moll model when VBE and VBC are two
driving forces for IC and IB, plus additional enhancements for high level injection,
voltage dependent capacitances that accurately represent the charge storage in the
transistor, and parasitic resistances as shown. This BJT model is known as the
Gummel-Poon model. The two diodes represent the two IB terms due to VBE and
VBC. The capacitor labeled QF is voltage dependent such that the charge stored in it
is equal to the QF described in the bipolar transistor transit time and charge storage
discussion. QR is the counterpart of QF produced by a forward bias at the BC
junction. Inclusion of QR makes the dynamic response of the model accurate even
when VBC is sometimes forward biased. CBE and CBC are the junction depletion
layer capacitances. CCS is the collector to substrate capacitance (Fig. C.20).
qVBE qVBC VCB Is qVBC
IC ¼ Is0 ðe kT e kT Þ ð1 þ Þ ðe kT 1Þ:
VA bR
The 1 þ VVCBA factor is added to represent the early effect—IC increasing with
increasing VCB. Is0 differs from Is in that Is0 decreases at high VBE due to the high
level injection effect in accordance with equation
ZWB
n2i p
GB dx:
n2iB DB
0
Is qVBE Is qVBC qVBE

IB ¼ ðe kT 1Þ þ ðe kT 1Þ þ ISE ðenE kT 1Þ:
bF bR
ISE and nE parameters are determined from the measured BJT data as are all of
the several dozens of model parameters. We can summery the current appendix
discussion, that base emitter junction is usually forward biased while the
base-collector junction is reverse biased. VBE determines the rate of electron
Fig. C.20 Bipolar transistor BJT model used for circuit simulation
injection from the emitter into the base, and thus uniquely determines the collector
current, IC regardless of the reverse bias VCB.
ZWB
q n2i qVBE n2i p
IC ¼AE ðe kT 1Þ; GB dx
GB n2iB DB
0
GB is the base Gummel number, which represents all the subtleties of BJT
design that affects IC; base material, non-uniform base doping, non-uniform ma-
terial composition, and the high level injection effect. An undesirable but
unavoidable side effect of the application on VBE is a hole current flowing from the
base, mostly into the emitter. This base input current, IB, is related to IC by the
common emitter current gain bF (bF ¼ IICB GGEB ) where GE is the emitter Gummel
number. The common base current gain is aF IICE ¼ 1 þbFb . The Gummel plot
F
indicates that bF falls off in the high IC region due to high level injection in the base
and also in the low IC region due to excess base current. Base width modulation by
VCB results in a significant slope of the IC-VCEcurve in the active region. This is the
early effect. The slope, called the output conductance, limits the voltage gain that
can be produced with a BJT. The early effect can be suppressed with a lightly doped
collector. A heavily doped sub-collector is routinely used to reduce the collector
resistance. Due to the forward bias, VBE, a BJT stores a certain amount of excess
hole charge, which is equal but of opposite sign to the excess electron charge. Its
magnitude is called the excess carrier charge, QF. QF is linearly proportional to IC
(QF IC sF ).
sF is the forward transit time. If there were no excess carriers stored outside the
W2
base sF ¼ sFB ¼ 2DBB . sFB is the base transit time, sF [ sFB because excess carrier
storage in the emitter and in the depletion layer is also significant. All these regions
should be made small in order to minimize sF. Besides minimizing the base width,
WB, sFB may be reduced by building a drift field into the base with graded base
doping (also with graded Ge content in a SiGe base). sFB is significantly increased
at large IC due to base widening which known at the Kirk effect. In the Gummel
Poon model, both the DC and the dynamic (charge storage) currents are well
modeled. The early effect and high level injection effect are included. Simpler
models consisting of R, C, and current source are used for hand analysis of circuits.
The small signal models employ parameters such as trans-conductance gm ¼ dV dIC
BE
¼
IC = kT
q and input capacitance Cp ¼ dVBE ¼ dVBE ½sF IC ¼ sF gm and input resis-
dQF d
bF
tance rp ¼ dVdIB ¼ gm . The BJT’s unity gain cutoff frequency at which b falls to unity
BE
is fT. In order to raise device speed, device density, or current gain, a modern high
performance BJT usually employs poly-Si emitter, self-aligned poly-Si base con-
tacts, graded Si-Ge base, shallow oxide trench, and deep trench isolation. High
performance BJTs excel over MOSFETs in circuits requiring the highest device gm
and speed.
Gummel-Pooncharge control model: The Gummel-Poon model is a detailed
charge-controlled model of BJT dynamics, which has been adopted and elaborated
by others to explain transistor dynamics in greater detail than the terminal-based
models typically do. This model also includes the dependence of transistor b values
upon the DC current levels in the transistor, which are assumed current independent
in the Ebers-Moll model. A significant effect included in the Gummel-Poon model
is the DC current variation of the transistor bF and bR. When certain parameters are
omitted, the Gummel-Poon model reverts to the simpler Ebers-Moll model. The
basic circuit which describes the Gummel-Poon model is the large signal schematic.
It represents the physical transistor, a current-controlled output current sink, and
two diode structures including their capacitors (Fig. C.21).
Fig. C.21 Gummel-Poon large signal schematic of the BJT
We can derive from the Gummel-Poon large signal schematics of the bipolar
transistor the Small Signal (SS) schematics for high frequency simulations. This
mean, for a given operating point, the DC currents are calculated and the model is
linearized in this point. The schematic is a pure linear model. The AC Small Signal
(SS) schematic of bipolar transistor is described in Fig. C.22.
For simplicity we assume no voltage drops at RBB′, Rc, RE then VB′E′ = VBE,
VB′C′ = VBC, VC′E′ = VCE.
The Gummel-Poon BJT full model analysis: The Gummel-Poon model is a
compact model for bipolar junction transistors (BJT) which also takes into account
effects of low currents and at high level injection (Fig. C.23).
Remark: we consider no voltage drop on RBB′, Rc, RE (VEE0 ! e; VCC0 ! e;
VBB0 ! e).
Gmin is the minimum conductance which is automatically switched in parallel to
each PN junction.
Fig. C.22 AC small signal schematic of the bipolar transistor
Iss qVbe
Iss qVbc Iss qVbe qVbc
Icc ¼ ðenf kB T 1Þ; Iec ¼ ðenr kB T 1Þ; Ict ¼ Icc Iec ¼ ðenf kB T enr kB T Þ
qb qb qb
Iec Iss qVbc I cc I ss
qVbe
ID1 ¼ ¼ ðenr kB T 1Þ; ID3 ¼ ¼ ðenf kB T 1Þ

br qb br bf qb bf
qVbc qVbe
ID2 ¼ C4 Iss ðencl kB T 1Þ; ID4 ¼ C2 Iss ðenel kB T 1Þ
The model distinguishes four operating region: normal active region, inverse
region, saturated region, and off region.
Normal active region:
5 nf k B T 5 nr k B T
Vbe [ ; Vbc
q q
Is q Vbe qb Vbe 1 1
Ic ¼ ½expð Þ þ þ C4 Is þ ½ ð þ Þ Vbc Gmin
qb nf k B T br qb q b br
Fig. C.23 Gummel-Poon BJT full model
1 q Vbe 1 q Vbe
Ib ¼ Is ½ ½expð Þ 1 þ C2 Is ½expð Þ 1
bf nf k B T br nel kB T
Vbe Vbc
C4 Is þ ð þ Þ Gmin
bf br
5nf kB T
Inverse region: Vbe q ; Vbc [ 5nr qkB T
Is q Vbc qb q Vbc
Ic ¼ fexpð Þ þ ½expð Þ 1g
qb nr k B T br nr k B T
q Vbc Vbe 1 1
C4 Is ½expð Þ 1 þ ½ ð þ Þ Vbc Gmin
ncl kB qb q b br
Fig. C.24 BJT NPN Gummel-Poon static model
1 q Vbe 1 q Vbe
Ib ¼ Is ½ ½expð Þ 1 þ C2 Is ½expð Þ 1
bf nf k B T br nel kB T
Vbe Vbc
C4 Is þ ð þ Þ Gmin
bf br
5nf kB T
Saturated region: Vbe [ q ; Vbc [ 5nr qkB T
Is q Vbe q Vbc qb q Vbc

Ic ¼ f½expð Þ expð Þ ½expð Þ 1g
qb nf k B T nr k B T br nr k B T
q Vbc Vbe 1 1
C4 Is ½expð Þ 1 þ ½ ð þ Þ Vbc Gmin
ncl kB qb qb b r
1 q Vbe 1 q Vbc
Ib ¼ Is f ½expð Þ 1 þ ½expð Þ 1g
bf nf k B T br nr k B T
q Vbe q Vbc Vbe Vbc
þ C2 Is ½expð Þ 1 þ C4 Is ½expð Þ 1 þ ð þ Þ Gmin
nel kB t ncl kB T bf br
5nf kB T
Off region:Vbe q ; Vbc 5nr qkB T
Is Vbe 1 1
Ic ¼ þ C 4 Is þ ½ ð þ Þ Vbc Gmin ;
br qb q b br
bf þ br Vbe Vbc
Ib ¼ Is ½ ðC2 þ C4 Þ Is þ ð þ Þ Gmin
bf br bf br
By disconnecting all capacitor in Gummel-Poon BJT full model we get the NPN
Gummel-Poon static model (Fig. C.24).
References
1. Kawdungta, S., Phongcharoenpanich, C., & Torrungrueng, D. (2008). Novel design of

double loop antennas by using a shifted Gate for the LF–RFID system. Asian University,
Thailand: Faculty of engineering and Technology.
2. Kuznetsov, Y. A. (1995). Elements of applied bifurcation theory. Applied Mathematical
Sciences.
3. Hale, J. K. (2012). Dynamics and bifurcations. Texts in Applied Mathematics, 3.
4. Strogatz, S. H. (2014). Nonlinear dynamics and chaos. Boulder: Westview press.
5. Kuang, Y. (1993). Delay differential equations with applications in population dynamics.
Boston: Academic Press.
6. Beretta, E., & Kuang, Y. (2002). Geometric stability switch criteria in delay differential
systems with delay dependent parameters. SIAM Journal on Mathematical Analysis, 33,
1144–1165.
7. Aluf, O. (2008, May). RFID TAGs COIL’s dimensional parameters optimization as
excitable linear bifurcation systems. In IEEE COMCAS2008 Conference.
8. Aluf, O. (2011, November). RFID TAGs coil’s system stability optimization under delayed
electromagnetic interferences. In IEEE COMCAS2011 Conference.
9. Guckenheimer, J. (2002). Nonlinear oscillations, dynamical systems, and bifurcations of
vector fields. Applied Mathematical Sciences, 42.
10. Wiggins, S. (2003). Introduction to applied nonlinear dynamical systems and chaos. Text in
Applied Mathematics (Hardcover).
11. Haug, H., & Koch, S. W. (2008). Quantum theory of the optical and electronic properties
of semiconductors (5th ed.). Singapore: World Scientific.
12. Kuang, J., & Cong, Y. (2007). Stability of numerical methods for delay differential
equations. Amsterdam: Elsevier Science.
13. Balachandran, B., & Kalmár-Nagy, T., & Gilsinn, D. E. (2009, March 5). Delay differential
equations: Recent advances and new directions (1st ed.) (Hardcover). Heidelberg: Springer.
14. Sze, S. M., & Ng, K. K. (2006). Physics of semiconductor devices. New York: Wiley
Interscience (Hardcover—Oct 27).
15. Sah, C. T. Fundamentals of solid state electronics. Singapore: World Scientific.
16. Yang, W. Y., Cao, W., Chung, T.-S., & Morris, J. (2005). Applied numerical methods using
MATLAB. New York: Wiley.
17. Band, A., & Unguris, J. Optically isolated current-to-voltage converter for an electron
optics system. Electron physics group National Institute of Standards and technology
Gaithersburg, Maryland 20899.
18. Beretta, E., & Kuang, Y. (2002). Geometric stability switch criteria in delay differential
systems with delay dependent parameters. SIAM Journal on Mathematical Analysis, 33(5),
1144–1165 (Published electronically February 14, 2002).
19. Kuang, Y. (1993). Delay differential equations with applications in population dynamics.
Mathematics in Science and Engineering, 191.

DOI 10.1007/978-3-319-45427-6
1036 References
20. Flower, A. C., & Mcguinnless, M. J. (1982). A description of the Lorenz attractor at high
prandtl number. Physica 5D, 149–182.
21. Dullin, H. R., Schmidt, S., Richter, P. H., & Grossmann, S. K. (2005, April 12). Extended
phase diagram of the Lorenz model.
22. Sprott, J. C. Simplifications of the Lorenz attractor. Nonlinear Dynamics Psychology and
Life Sciences, 13(3), 271–278.
23. van der Schrier, G., & Mass, L. R. M. (2000). The diffusionless Lorenz equations;
Shil’nikov bifurcations and reduction to an explicit map. Physica D, 141, 19–36.
24. Marton, L. (1974). Advanced electronics and electron physics (Vol. 35). USA: Academic
Press Inc. Elsevier.
25. Cripps, S. C. (2006). RF power amplifiers for wireless communication. London: Artech
House microwave library.
26. Sayre, C. W. Complete wireless design. New York: McGraw-Hill companies.
27. Gopalsamy, K. Stability and oscillation in delay differential equations of population
dynamics. Berlin: Kluwer Academic publishers.
28. Kuznetsov, Y. (2010). Elements of applied bifurcation theory [Paperback]. NY, LLC:
Springer.
29. Butcher, E. A., Ma, H., Bueler, E., Averina, V., & Szabo, Z. (2004). Stability of linear time
periodic delay differential equations via Chebyshev polynomials. International Journal of
Numerical Methods Engineering, 59, 895–922.
30. Stepan, G., Insperger, T., & Szalai, R. (2005). Delay, parametric excitation, and the
nonlinear dynamics of cutting process. International Journal of Bifurcation and Chaos, 15
(9), 2783–2798.
31. Insperger, T., & Stepan, G. (2003). Stability of the damped Mathieu equation with time
delay. Journal of Dynamic Systems, Measurement, and Control, 125.
32. Garg, N. K., Mann, B. P., Kim, N. H., & Kurdi, M. H. (2007, March). Stability of a time
delayed system with parametric excitation. Journal of Dynamic Systems, Measurement, and
Control, 129.
33. Allan, W. S. (2005). Understanding microwaves. New York: Wiley.
34. Rogers, J. W. M., Plett, C., & Marsland, I. (2014). Radio frequency system architecture and
design. London: Artech house.
35. Linkhart, D. K. (2014). Microwave circulator design (2nd ed.). London: Artech house.
36. Tanaka, S., Shimomura, N., & Ohtake, K. (1965, March). Active circulators—The
realization of circulators using transistors. Proceedings of the IEEE, 260–267.
37. Dougherty, R. (1989, June). Circulate signals with active devices on monolithic chips.
Microwave and RF, 85–86, 89.
38. Looss, G., & Joseph, D. D. (1980). Elementary stability and bifurcation theory. Berlin:
Springer.
39. Guckenheimer, J., & Holmes, P. (1983). Nonlinear oscillations, dynamical systems, and
bifurcation of vector fields. Applied Mathematical Sciences, 42.
40. Perko, L. (1991). Differential equations and dynamics systems. Texts in Applied
Mathematics, 7.
41. Kuang, J., & Cong, Y. (2005). Stability of numerical methods for delay differential
equations. USA: Since press USA Inc.
42. Hollenstein, C., Guittienne, p., & Howling, A. A. Resonant RF network antennas for
large-area and large-volume inductively coupled plasma sources.
43. Granas, A., & Dugundji, J. (2003, July 24). Fixed point theory. Berlin: Springer.
44. Border, K. C. (1989, July 28). Fixed point theorems with application to economics and
game theory. Cambridge: Cambridge University Press (Rep sub edition).
45. Agarwal, R. P., Meehan, M., &O’Regan, D. (2009, March 19). Fixed point theory and
applications (1st ed.). Cambridge: Cambridge University Press.
46. Istratescu, V. I. (2001, November 30). Fixed point theory: An introduction. Berlin:
Springer.
References 1037
47. Geller, S. B., & Mantek, P. A. (1962, January–March). Tunnel diode large-signal
equivalent circuit study and the solutions of its nonlinear differential equations. Journal of
research of the National Bureau of Standards—C. Engineering and Instrumentation, 66C
(1).
48. Hines, M. E. (1960). High-frequency negative-resistance circuit principles for Esaki diode
applications. Bell System Technical Journal, 39, 477.
49. Cohen, S. (1960). Tunnel diode characterization. Electric Equipment & Engineering, 8,
102.
50. Lowry, H. R., Giorgis, J., Gottlieb, E., & Weischedel, R. C. (1961). Tunnel diode manual.
General Electric, 33–42.
51. Lefshetz, S. (1957). Differential equations: Geometric theory (p. 261). Geneva: Interscience
Publishers.
52. Hsia, P. S. (1952). A graphical analysis for nonlinear systems. Proceeding of IEEE, 99,
125.
53. Crisson, G. (1931). Negative impedances in the twin 21 type repeater. Bell System
Technical Journal, 10, 485.
54. Gautam, A. K., & Vishvakarma, B. R. (2006). Frequency agile microstrip antenna using
symmetrically loaded tunnel diodes. Indian Journal of Radio & Space physics, 35, 212–
216.
55. Srivastava, S., & Vishvakarma, B. R. (1999). Tunnel diode integrated rectangular patch
antenna. In Proceeding of the Radar Symposium India-99 (IRSI-99), Bangalore.
56. Srivastava, S., & Vishvakarma, B. R. (2003). Tunnel diode integrated rectangular
microstrip antenna for millimeter range. IEEE Transactions on Antennas Propagation
(USA), 51, 750.
57. Woo, C. F. (1964). Principles of tunnel diode circuits. New York: Wiley.
58. Sylvesten, G. P. (1962). Basic theory and applications of tunnel diode. Princeton, New
Jersey, USA: Van nostrand.
59. Soliman, F. A. S, & Kamh, S. A. (1993). Computer analysis for designing narrow band
tunnel diode amplifier circuit. Communication Faculty of Science University of Ankara
Series A2, A3, 42, 33–49.
60. Boric-Lubecke, O., Pan, D. S., & Itoh, T. (1995). Design and triggering of oscillators with a
series connection of tunneling diodes. Electronics and Engineering, 8(2), 271–286.
61. Brown, E. R., Soderstrom, J. R., Paker, C. D., Mahoney, L. J., Molvar, K. M., & McGill, T.
C. (1991). Oscillations up to 712 GHz in InAs/AISb resonant tunneling diodes. Applied
Physics Letters, 58(20), 20.
62. Menozzi, R., Piazzi, A., & Contini, F. (1996). Small-signal modeling for microwave FET
linear circuits based on a genetic algorithm. IEEE Transaction on Circuits and Systems—I:
Fundamental Theory and Applications, 43(10).
63. Jerinic, G., Fines, J., Cobb, M., & Schindler, M. (1985). Ka/Q band GaAs IMPATT
amplifier technology. International Journal of Infrared and Millimeter Waves, 6(2), 79–
130.
64. Mishra, L. P., & Mitra, M. (2015). Design and characterization of Ka-Band reflection-type
IMPATT amplifier. Intelligent Computing, Communication and Devices Advanced
Intelligent Systems and Computing, 308, 487–492.
65. Al-Attar, T., & Lee, T. H. (2005). Monolithic integrated millimeter wave IMPATT
transmitter in standard CMOS technology. IEEE Transaction on Microwave Theory and
Techniques, 53(11).
66. Gupta, M. S., & Lomax, R. J. (1971). A self-consistent large-signal analysis of read-type
IMPATT diode oscillator. IEEE Transactions on Electron Devices, ED-18(8), 544–550.
1038 References
67. Gupta, M. S., & Lomax, R. J. (1971). Injection locking in IMPATT diode oscillators. In
Proceedings of the Third Biennial Cornell Electrical Engineering Conference (pp. 215–
223). Ithaca, New York: School of Electrical Engineering, Cornell University. Reprinted in:
Avalanche Transit-Time Devices, G.I. Haddad, editor, Artech House, Dedham, Mass. 1973,
pp. 457–460.
68. Gupta, M. S., & Lomax, R. J. (1973). A current-excited large-signal analysis of IMPATT
devices and its circuit implications. IEEE Transactions on Electron Devices, ED-20(4),
395–399.
69. Gupta, M. S. (1973). Computer-aided characterization of IMPATT diodes. In Proceedings
of the Fourth Biennial Cornell Electrical Engineering Conference (pp. 349–358). Ithaca,
New York: School of Electrical Engineering, Cornell University.
70. Gupta, M. S. (1973). A small-signal and noise equivalent circuit for IMPATT diodes. IEEE
Transactions on Microwave Theory and Techniques, MTT-21(9), 591–594.
71. Gupta, M. S. (1973). Large-signal equivalent circuit for IMPATT diode characterization
and its application to amplifiers. IEEE Transactions on Microwave Theory and Techniques,
MTT-21(11), 689–694.
72. Gupta, M. S. Lomax, R. J., & Haddad, G. I. (1974). Noise considerations in self-mixing
IMPATT-diode oscillators for short-range doppler radar applications. IEEE Transactions
on Microwave Theory and Techniques, MTT-22(1), 37–43.
73. Gupta, M. S. (1975). A simple approximate method of estimating the effect of carrier
diffusion in IMPATT diodes. Solid-State Electronics, 18(4), 327–330.
74. Gupta, M. S. (1976). A nonlinear equivalent circuit for IMPATT diodes. Solid-State
Electronics, 19(1), 23–26.
75. Atherton, D. P. (1981). Stability of nonlinear system (Electronic & Electrical engineering
research studies). New York: Wiley.
76. Xue, D., Chen, Y. Q., & Atherton, D. P. (2009). Linear feedback control: Analysis and
design with MATLAB (advanced in design and control). Society for industrial and applied
mathematics (1st ed.).
77. Atherton, D. P. (1982). Nonlinear control engineering. London: Chapman & Hall (stu sub
edition).
78. Bar-yam, Y. (1997). Dynamics of complex systems (studies in nonlinearity) (1st ed.).
Boulder: Westview press.
79. Steeb, W. H. (2014). The nonlinear workbook: Chaos, fractals, cellular automata, generic
algorithms, gene expression programming, support vector machine, wavelets, hidden…
Java and symbolic C++ programs (6th ed.). Singapore: World scientific.
80. Sternberg, S. (2012). Curvature in mathematics and physics. New York: Dover
publications.
81. Sternberg, S. (2010). Dynamical systems. New York: Dover publications.
82. Scheinerman, E. R. (2012, January 18). Invitation to dynamical systems. New York: Dover
publications (Reprint edition).
83. Abraha, R. (1983). Dynamics the geometry of behavior part 2: Chaotic behavior (Visual
mathematics library). Aerial Press.
84. Abraham, R. (1982). Dynamics the geometry of behavior: Periodic behavior (Visual
mathematics library). Aerial Press.
85. Ahson, S., & Ilyas, M. (2008). RFID handbook, applications, technology, security, and
privacy. Boca Raton: CRC press, Taylor & Francis Group.
86. Aluf, O. (2014). Cylindrical RF network antennas for coupled plasma sources copper legs
delayed in time system stability. Transaction on networks and communications, 2(5),
116–146.
87. Cheban, D. N. (2015). Global attractors of non-autonomous dynamical and control systems
(2nd ed.). Singapore: World scientific.
88. Li, C., Wu, Y., & Ye, R. (2013). Recent advanced in applied nonlinear dynamics with
numerical analysis. Singapore: World scientific.
References 1039
89. Sidorov, D. (2014). Integral dynamical models. Singapore: World scientific.

90. Pozar, D. M. (2005). Microwave engineering (3rd ed.). New York: Wiley.
91. Collin, R. E. (1992). Foundation for microwave engineering (2nd ed.). New York:
McGraw-Hill.
92. Ludwig, R., & Bretchko, P. (2000). RF circuit design—Theory and applications. Upper
Saddle River: Prentice-Hall.
93. Gonzalez, G. (1997). Microwave transistor amplifiers—Analysis and design (2nd ed.).
Upper Saddle River: Prentice-Hall.
94. Vendelin, G. D., Pavio, A. M., & Rhode, U. L. (1990). Microwave circuit design—using
linear and nonlinear techniques. New York: Wiley.
95. Gilmore, R., & Besser, L. (2003). Practical RF circuit design for modem wireless System
(Vols. 1–2). Norwood: Artech House.
96. Iu, H. H. C., & Fitch, A. L. (2013). Development of memristor based circuit. Singapore:
World Scientific.
97. Ling, B. W. K., & Iu, H. H. C. (2008). Control of chaos in nonlinear circuits and systems.
Singapore: World Scientific.
98. Abarbanel, H. D. I., Rabinovich, M. I., & Sushchik, M. M. (1993). Introduction to
nonlinear dynamics for physicists. Singapore: World Scientific.
99. Kilic, R. (2010). A practical guide for studying chua’s circuits. Singapore: World
Scientific.
100. Qiying, W. (2015). Limit theorems for nonlinear cointegrating regression. Singapore:
World Scientific.
101. Tatarinova, T., & Schumitzky, A. (2015). Nonlinear mixture models. Singapore: World
Sientific.
102. Blake, L. M., & Long, M. W. (2009). Antennas: Fundametals, design, measurement. USA:
SciTech publishing Inc.
103. Milligon, T. A. (2005). Modern antenna design. New York: Wiley.
104. Hansen, R. C., & Collin, R. E. (2011). Small antenna handbook. New York: Wiley.
105. Volakis, J. L., Chen, C. C., & Fujimoto, K. (2010). Small antennas: Miniaturization
techniques & applications. New York: Mc Graw Hill education.
106. Kumar, N., & Grebennikov, A. (2015). Distributed power amplifiers for RF and microwave
communications. Norwood: Artech House.
107. Campos, R. S., & Lovisolo, L. (2015). RF positioning: Fundamentals, applications, and
tools. Norwood: Artech House.
108. Wallace, R., & Andreasson, K. (2015). Introduction to RF and microwave passive
components. Norwood: Artech House.
109. Gildenblat, G. (2010). Compact modeling: Principles, techniques and applications. Berlin:
Springer.
110. Diks, C. (1999). Nonlinear time series analysis. Singapore: World Scientific.
111. Letellier, C., & Gilmore, R. (2013). Topology and dynamics of chaos. Singapore: World
Scientific.
112. Wang, Q. (2015). Limit theorems for nonlinear cointegrating regression. Singapore: World
Scientific.
113. Sibani, P., & Jensen, H. J. (2013). Stochastic dynamics of complex systems. Singapore:
World Scientific.
114. Sprott, J. C. (2010). Elegant chaos. Singapore: World Scientific.
115. Elhadj, Z., & Sprott, J. C. (2011a). Robust chaos and its applications. Singapore: World
Scientific.
116. Elhadj, Z., & Sprott, J. C. (2011b). Frontiers in the study of chaotic dynamical systems with
open problems. Singapore: World Scientific.
117. Hoover, W. G. & Hoover, C. G. (2015). Simulation and control of chaotic nonequilibrium
systems. Singapore: World Scientific.
1040 References
118. Ivancevic, V. G., & Reid, D. J. (2014). Complexity and control. Singapore: World
Scientific.
119. Nicolis, G., & Nicolis, C. (2012). Foundations of complex systems. Singapore: World
Scientific.
120. Cencini, M., Cecconi, F., & Vulpiani, A. (2009). Chaos. Singapore: World Scientific.
121. Salleh, A., Abd aziz, M. Z. A., Misran, M. H., Othman, M. A., & Mohamad, N. R. (2013).
Design of wideband low noise amplifier using negative feedback topology for Motorola
application (Vol. 5, No. 1). ISSN: 2180-1843.
122. Duan, L., Huang, W., Ma, C., He, X., Jin, Y., & Ye, T. (2012). A single to differential low
noise amplifier with low differential output imbalance. Journal of Semiconductors, 33(3),
035002.
123. Kuang, Y., & Cushing, J. M. (1996). Global stability in a nonlinear difference delay
equation model of flour beetle population growth. Journal of Difference Equations and
Applications, 2, 31–37.
124. Elsayed, E. M., El-Dessoky, M. M., & Alotaibi, A. (2012). On the solutions of a general
system of difference equations, Hindawi Publishing Corporation. Discrete Dynamics in
Nature and Society, 2012(892571), 12.
125. Desoer, C. A. (1969). Basic circuit theory (1st ed.). New York: Mcgraw-Hill College.
126. Grover, F. W. (2004). Inductance calculations: Working formulas and tables. New York:
Dover publications.
127. Greenhouse, H. M. (1974). Design of planar rectangular microelectronic inductors. IEEE
Transactions on Parts, Hybrids, and Packaging, PHP-10(2).
128. Rodriguez, E. G. (2015). Reconfigurable transceiver architecture for multiband RF
frontends. Berlin: Springer.
129. Steinberg, K., Scheffler, M., & Dressel, M. (2010). Microwave inductance of thin metal
strips. Journal of Applied Physics, 108, 096102.
130. Liao, Y., Cai, K., Hubing, T. H., & Wang, X. (2014). Equivalent circuit of normal mode
helical antennas using frequency—Independent lumped elements. IEEE Transactions on
Antennas and Propagation, 62, 5885–5888.
131. Su, C., Ke, H., & Hubing, T. H. (2010). A simplified model for normal mode helical
antennas. Applied Computational Electromagnetics Society Journal, 25(1), 32–40.
132. Su, C., Ke, H., & Hubing, T. H. (2010). Corrections to a simplified model for normal mode
helical antennas. Applied Computational Electromagnetics Society Journal, 25(8), 722.
Index
A 811, 818, 825, 836, 848, 870, 874, 884,

Abrupt junction, 747 907, 930, 942
Absolute value, 25, 51, 116, 229, 469, 683, Antenna inductance, 15, 46, 786, 820, 913, 923
705, 738, 762, 795, 826 Anti-collision protocol, 3, 4
Absorption method, 970 Antisymmetric, 923
Absorptive filter, 674 Apparent, 302, 619, 713, 714
AC-cycle, 528, 529 Arbitrary scalar differential equation, 2, 14
AC-ground, 574, 585, 603, 611, 619, 632, 645, Argument, 36, 56, 66, 97, 141, 167, 189, 264,
746 367, 455, 479, 800, 804, 813, 831, 834, 909
AC signal, 280, 644 Assumption, 4, 169, 220, 231, 240, 245, 332,
Active circulator, 155, 171–173, 177, 178, 180, 345, 354, 417, 430, 435, 471, 558, 642,
185, 191, 194, 336, 338, 339, 347, 352, 682, 767, 785, 959
370, 377 Asymptotic stability, 614
Active RFID system, 2, 9, 46 Atmospheric upheavals, 956
Active tag, 1, 6 Attenuation, 411, 493, 673, 675, 904, 947
Alfa, 281 Attenuator, 406, 489, 493, 494, 498, 500
AlGaAs, 304 Attracting circle, 13
Algebraic manipulation, 165, 169 Attracting stable, 872
ALOHA, 3, 4 Attraction, 617, 630
AM/FM, 904 Audio power amplifier, 514
Amper, 305, 784 Auto-bias, 300, 301
Amplifier, 174, 300, 302, 336, 348, 405, 410, Autonomous system, 12, 159
417, 433, 454, 513, 527, 559, 601, 904, Avalanche, 279, 333, 904, 1006–1008
910, 941, 944, 945, 947, 953, 954, 958, Avalanche multiplication, 1008
964, 1010, 1015 Avalanche region, 333
Amplifier gain, 302, 411 Average, 3, 6, 15, 23, 106, 461, 785, 818, 839,
Analytical criteria, 23, 46, 157, 786, 820 913, 914, 923, 1009
Analytical function, 35, 63, 93, 96, 137, 139, Average analysis, 6
161, 164, 183, 185, 186, 208, 259, 262, Average radius, 785
362, 366, 367, 455, 456, 480, 482, 697, Axial ratio, 876
711, 800, 811, 833, 834, 873, 909 Axial velocity, 875
Analytical methods, 1 Axis, 42, 144, 171, 195, 220, 365, 488, 630,
Analyze, 1, 2, 11, 31, 61, 73, 181, 196, 219, 812, 817, 873, 1020
360, 424, 488, 696, 802, 872, 1023, 1025
Angle, 38, 68, 99, 188, 210, 370, 457, 513, B
559, 813, 874, 908, 910, 953, 964, 981 Baker clamp, 994
Antenna, 1, 14, 23, 36, 44–46, 55, 72, 105, Balance parameter, 358–360
156, 196, 204, 213, 573, 783, 792, 800, Balun, 904

DOI 10.1007/978-3-319-45427-6
1042 Index
Band attenuation, 673 70, 79, 85, 92, 96–98, 100, 114, 135–137,
Band Pass Filter (BPF), 674, 675, 683, 696, 139, 140, 141, 143, 148–153, 160, 161,
745, 752, 757, 761, 767, 911 163–165, 167, 171, 178, 181–183, 185,
Bandwidth, 3, 14, 338, 373, 573, 712, 785, 187, 188, 190, 195, 207–210, 212, 219,
903, 954, 957, 962 255–257, 259, 262, 263, 264, 266, 270,
Base current, 282, 584, 994, 996, 1002, 1003, 272, 273, 277, 356, 358, 360, 362, 365,
1007, 1008, 1018, 1019 366–370, 372, 389, 417, 430, 444, 454,
Base junction, 279, 280, 994, 1003, 1007, 457, 458, 459, 468, 478, 479, 483, 488,
1009, 1010 498, 536, 570, 572, 579, 614, 616, 624,
Base storage, 994 629, 638, 653, 683, 696, 704, 710, 711,
Base transport factor, 1003 736, 745, 761, 767, 797, 799, 800, 803,
Base width, 996, 1017, 1020, 1021, 1028 804, 806–808, 811–813, 816, 817, 820,
Battery, 1, 2, 44, 104 829, 830, 833–836, 870, 873, 881, 885,
Bias-T, 344, 355, 406, 471, 479, 484, 488 909, 910
BiCMOS, 1015 Charge carrier, 408, 489, 993, 994
Bi-cubic polynomial, 36, 164, 208 Charge control model, 1015, 1025, 1026, 1028
Bifurcation, 2, 14, 73, 101, 142, 164, 183, 208, Charge storage, 514, 1023, 1025, 1026
262, 372, 455, 480, 487, 811, 838, 905, 906 Charging capacitance, 1009
Binary Tree (BT), 3 Choke, 73, 156, 382, 493, 574, 631, 942, 943,
Bipolar transistor, 279–281, 299, 301, 910, 963
994, 995, 998, 1015, 1025, 1026, 1029 Circular coil, 784–786, 793–796
Birdcage coil, 196 Circular periodic, 12
Bit Error Rate (BER), 951 Circulator, 155, 171, 177, 178, 181, 191, 334,
BJT, 407, 408, 410, 423, 514, 517, 530, 559, 344, 355, 379, 910
574, 584–586, 619, 638, 645, 910, 993, Clap, 630, 910
1002, 1004, 1010, 1015, 1017, 1020, 1021 Class A, 514, 515, 574
Body effect parameter, 1012, 1013 Class AB, 514, 527
Boltzmann’s constant, 957 Class B, 513, 514, 559, 560
Bondwire, 71, 281, 854 Class C, 514, 528, 529, 536
Borderline, 22 Class D, 406
Branch, 173, 194, 305 Clipping, 952, 953, 1010
Breakdown, 156, 333, 747, 1006, 1008 Clockwise, 20
Breakdown voltage, 156, 747, 1007, 1008 Closed loop, 624, 629, 630
Bulk, 73, 515, 1012, 1013, 1015 CMOS, 103, 1015
Bulk-source, 1015 Coax cable, 461, 468
Burst switch, 104–106, 113, 116, 136, 142, 144 Co-dimension, 906
Bypass, 101, 222, 223, 225, 337, 618, 644 Coefficient, 36, 56, 59, 62, 66, 93, 98, 165,
Bypass capacitance, 101 208, 262, 362, 454, 456, 482, 612, 616,
804, 811, 830, 834, 1009
C Coil, 2, 8, 14, 24, 619, 620, 714, 784, 785, 793,
Capacitor, 2, 6, 155, 172, 197, 301, 335, 382, 795, 798, 808
460, 493, 574, 632, 697, 746, 842, 904, Collector current, 282, 528, 529, 993, 994, 996,
970, 980, 983, 1028, 1033 1007, 1015, 1022, 1027
Cascade, 711, 911 Collide, 905
CCNR, 336 Colpitts, 630, 910
Center, 19, 22, 422, 614, 617, 630, 643, 659, Common Base (CB), 405, 910, 941, 942, 1001
745, 853, 974 Common base forward short circuit current
Center manifold theorem, 617 gain, 282
Channel length modulation parameter, 1013 Common Collector (CC), 405, 910, 941, 944,
Chaos model, 1 945
Characteristic curve, 513 Common Emitter (CE), 405, 408, 910, 941,
Characteristic equation, 17, 23, 25, 28, 30–37, 943, 1001
39, 42, 45, 50, 53–57, 59, 60, 63–66, 68,
Index 1043
Complex, 19, 36, 67, 98, 103, 144, 167, 191, Decay oscillation spiral, 298
213, 262, 368, 456, 488, 527, 629, 696, Degenerate node, 22
710, 745, 797, 816, 834, 847, 884, 906, Delay, 2, 22, 24, 29, 33, 43, 45, 50, 55, 70,
909, 1021 101, 132, 157, 173, 196, 255, 360, 406,
Complex conjugate, 109, 366, 411, 615, 630, 461, 696, 793, 803, 817, 837, 906, 1026
696, 710, 745, 767, 872, 945 Delay dependent parameters, 2, 22, 46, 157,
Complex number, 19 696, 711, 819, 820, 872
Conductivity, 108, 785, 819, 839, 841, 918, Delay Differential Equation (DDE), 170
922, 1006 Delay parameter, 31, 57, 101, 142, 190, 229,
Conductor, 171, 279, 783, 784 372, 459, 479, 710, 838, 843, 907, 909
Conjugate, 833, 871, 884, 906, 959, 977 Depletion capacitance, 1009, 1010, 1015
Connectorize, 460 Derivative, 11, 18, 19, 21, 22, 25, 29, 38, 45,
Conservation, 619, 713, 714 68, 98, 115, 161, 165, 171, 180, 188, 204,
Continuous Wave (CW), 73 218, 230, 321, 348, 444, 483, 579, 653,
Converge, 18, 19, 21 794, 812, 819, 826, 835, 863, 907
Coordinates, 26, 52, 54, 82, 179, 180, 204, Detector, 1, 72, 74, 81, 94, 100–102, 489
252, 255, 324, 347, 349, 353, 794, 826, 828 Detector circuit, 101
Copper, 156, 196, 197, 202, 213, 335, 981 Determinant, 611, 613, 616, 726
Copper leg, 196, 213, 218, 910 Deterministic protocol, 3
Core, 619, 620, 713, 715 Deterministic TAG anti-collision, 4
Cos, 845 Device Under Test (DUT), 954
Cosmic noise, 957 Diagram, 14, 42, 102, 217, 574, 848
Counter, 20, 105, 410, 630 Diameter, 15, 23, 213, 784, 818, 873, 876, 914,
Counter clockwise, 20 925
Coupling coefficient, 46, 786, 820, 924 Dielectric constant, 980, 981
Cramer’s rule, 611, 726 Dielectric material, 222, 876
Critical threshold, 905 Differentiable, 29, 31, 35, 36, 43, 56, 59, 65,
Cross section, 73, 108, 785, 819, 841, 854, 97, 140, 186, 210, 365, 456, 482, 711, 804,
918, 919, 922, 931 817, 833, 872, 907
Crystal lattice, 674, 712 Differential equations, 2, 14, 16, 22, 45, 50, 77,
Current gain, 280, 405, 561, 943, 994, 1003, 113, 156, 177, 204, 285, 323, 443, 497,
1004, 1019, 1023, 1028 564, 607, 647, 679, 701, 721, 752, 760,
Current sink, 560, 1028 825, 879, 925
Current-voltage relation, 156, 994 Differential function, 456, 800
Curve, 336, 994, 1015, 1018, 1020, 1028 Diffusion, 156, 408, 993, 996, 1009
Curve bend, 981 Digital circuit, 72, 73, 513
Cutoff, 173, 339, 410, 1001 Diode, 72, 73, 101, 104, 155–157, 221, 230,
Cycle, 8, 19, 281, 515, 559, 953 334, 335, 347, 373, 410, 489, 490, 493,
Cylinder, 9 745, 749, 758, 765, 910, 995, 996, 1013,
Cylindrical RF network antennas, 156, 1014, 1026
202–204, 206, 210, 213, 217–219, 904 Diplexer, 673–675, 682, 689, 696, 704, 709,
711, 911
D Dipole antenna, 784, 873
dBm, 958, 964 Discrete system, 905
dB(s), 333, 410, 946, 964, 1009, 1017, 1018 Disjoint frequency band, 674
DC bias, 101, 156, 280, 335, 460, 574, 631, Displacement, 81, 203, 333
959, 1022 Dissipation balance, 8
DC resistance, 108, 785, 818, 819, 839, 841, Distortion, 408, 410, 513, 528, 559, 951–954,
918, 919, 922, 930, 931 1010
Decade bandwidth, 171, 172, 338 Distributed transformer, 979, 980
Decay, 10, 19, 617 Disturbance, 11, 528, 568
Decay exponentially, 19, 617 Disturbance grow, 10
Decaying oscillations, 19 Dopant, 1018, 1021, 1022
1044 Index
Double loop antennas, 44, 45, 48, 72 Etched, 8, 16, 904, 915
Drain, 172, 299, 301, 382, 574, 956, 1013, Euler’s formula, 162
1015, 1020 Excitable bifurcation system, 2
Drift region, 333, 334 Excitation, 45, 489
Dual band, 697, 704, 707, 710, 711 Exponential form, 25, 26, 28, 51, 81, 115, 161,
Duplexer, 674, 675 178, 204, 229, 252, 324, 347, 389, 430,
Dynamical behavior, 170, 905 468, 536, 624, 653, 704, 761, 794, 826,
Dynamic range, 408, 490, 712 869, 880
Exponential motion, 17
E External disturbance, 617
Ebers-Moll, 410, 514, 529, 995, 1003–1005,
1015, 1028 F
EFF, 945, 981 F20-FET-4 75, 171, 338, 190
Efficiency, 14, 333, 373, 513, 515, 529, 559, Faraday’s law, 620, 715
712, 945, 994, 1019 Far field, 783
Eigendirection, 19 Fast switch, 406, 489
Eigen solution, 17, 19 Feedback, 172, 280, 338, 382, 411, 573, 574,
Eigenvalue, 17–20, 56, 93, 164, 183, 229, 298, 579, 601, 602, 618, 630, 644, 944, 959, 962
347, 390, 422, 456, 498, 570, 579, 614, Feedback branch, 172, 338
617, 630, 659, 696, 710, 736, 745, 767, Feedback network, 602, 644
795, 797, 804, 826, 869, 871, 885, 906 Feedback resistor, 959
Eigenvector, 17, 18, 20 Feedback system, 617, 910
Electromagnetic, 1, 2, 42, 50, 222, 785, 818, Femtofarad, 305
911 Ferromagnetic material, 620, 715
Electromagnetic energy, 783 Fick’s law, 996
Electromagnetic interferences, 22, 24, 43–45, Field Effect Transistor (FET), 904
50, 72, 793 Film, 784, 849, 854, 855, 869, 872, 911,
Electron, 3, 279, 280, 299, 300, 333 1025
Electron injection, 1015, 1027 Filter, 105, 406, 529, 674, 675, 704, 709, 711,
Electrostatic, 1 712, 717, 733, 736, 738, 746, 911, 954
Emitter, 279, 280, 405, 408, 515, 528, 574, Five elements circuit, 874
584, 601, 631, 941, 944, 955, 993–995, Fixed point, 2, 10, 11, 13, 18–22, 25, 50, 51,
1001–1004, 1006–1008 81, 83, 115, 116, 161, 178, 179, 203, 229,
Emitter doping, 1002, 1019 286, 294, 309, 324, 329, 347, 354,
Empirical parameter, 221 429–431, 443, 467, 469, 498, 499, 525,
Encoder, 73 527, 535, 568, 579, 581, 611, 612, 624,
End of frame, 5 627, 636
Energy, 2–4, 6, 8, 19, 104, 113, 222, 223, 617, Flat topping, 952
619, 714, 911 Flux, 619, 620, 713, 714, 981
Energy consumption, 2–4, 6 Focus, 36, 70, 139, 164, 368, 488, 838, 994
Energy generation, 8 Forcing function, 9
Environment, 42, 72 Forcing signal, 169
EPC, 904 Forward active mode, 410, 995, 1002–1004,
Epitaxial layer, 299, 300 1010
Equilibrium, 25, 50, 70, 81, 115, 161, 178, Forward common emitter current gain, 282
203, 229, 291, 347, 388, 429, 444, 498, Free space, 783, 784, 874, 875
536, 614, 617, 624, 682, 735, 761, 816, Frequency, 1, 44, 104, 172, 221, 223, 280, 281,
828, 880, 993 301, 303, 305, 334, 335, 373, 405, 410,
Equivalent circuit, 2, 6, 7, 14, 15, 23, 73, 102, 490, 492, 528, 529, 601, 630, 631, 673,
164, 171, 299, 305, 383, 417, 461, 492, 674, 697, 745, 748, 874, 877, 904, 910,
498, 560, 579, 611, 712, 786, 852, 853, 941, 944, 950, 951, 954, 959, 962, 963,
874, 910, 1022 970, 974, 1021, 1024
Estimation, 8, 24, 915, 925 Frequency Division Duplex (FDD), 674
Index 1045
FSK encoder, 73 Heavily doped emitter, 993, 1015, 1020

F slots reader, 6 Heavy audio distortion, 528
Fully-OFF, 513 Helical, 784, 873–876, 911
Fully-ON, 513 Helix, 784, 873, 874, 876, 880, 883, 884,
911
G HEMT, 304
GaAlAs, 1020 Heterojunction Bipolar Transistor (HBT), 1020
GaAs, 299–301, 304, 1020 Hfe, 962
Gain, 280, 299, 301, 335, 405, 408, 410, 561, Higher harmonic, 410
573, 601, 783, 903, 904, 941, 948, High fidelity, 513
956–958, 962–964, 994, 1003, 1008, 1015, High frequency circuit, 747
1019–1021, 1024, 1028 High frequency model, 172, 173, 303, 305,
GAMMA, 903, 1013 338, 747
Gamma radiation, 903 High level injection, 406, 489, 514, 559, 1015,
Gap, 1, 15, 23, 106, 304, 818, 839, 913–915, 1018, 1020, 1026, 1027, 1029
923, 925, 927, 983, 1017, 1019 High-pass, 196, 461, 673, 674, 952
Gate, 44–46, 50–52, 55–57, 59, 61, 65–69, 71, High speed counter circuit, 221
98, 139, 172, 187, 299, 301, 338, 382, 1015 Hodges, 1010
Gate termination, 173, 338 Hole’s level, 489
GEC-Marconi, 171, 338 Hole current, 1002, 1003, 1019, 1027
Geometric criterion, 23, 46, 157, 360, 367, Hole mobility, 490, 1015
696, 711, 820, 872, 907 Homogeneous, 25, 35, 51, 56, 59, 62, 65, 81,
Geometric Mean Distance (GMD), 852, 856 93, 115, 135, 136, 178, 183, 185, 203, 208,
Geometric stability, 143, 170, 696, 710, 872, 229, 259, 324, 347, 362, 389, 454, 683,
911 704, 736, 761, 794, 800, 804, 826, 830,
GHz, 223, 279, 333–335, 573, 783, 903, 944, 869, 880
960, 982, 983 Homogeneous system, 35, 51, 56, 59, 62, 65,
Giacoletto, 409 93, 115, 135, 136, 178, 183, 185, 203, 208,
Global bifurcation, 905 229, 259, 324, 347, 362, 389, 454, 683,
GMAX, 961 704, 736, 761, 794, 800, 804, 826, 830,
Graph, 2, 22, 46, 157, 336, 527, 673, 696, 711, 869, 880
786, 820, 872, 906, 1011, 1016 Homomorphism, 9
Graphical information, 2, 22, 46, 157, 696, Hot spot, 1008
711, 819, 872 H-port, 674
Ground pad, 157 Hybrid, 155, 171, 408, 586, 619, 632, 645
Ground plane, 784, 873, 980, 981 Hybrid Pi model, 408, 586, 619, 632, 645
Ground terminal, 301 Hybrid realization, 171
Growing oscillator, 19, 72, 222, 298 Hyperabrupt junction, 747
Gummel number, 1018, 1019, 1027 Hyperbolic fixed point, 128
Gummel plot, 1018, 1020, 1027
Gummel-Poon model, 514, 559–561, 910, I
1026, 1028, 1029 Ideality factor, 221
Identity matrix, 17, 570
H I-layer, 156, 492
Half-duplex operation, 4 Imaginary axis, 36, 43, 67, 70, 98, 101, 139,
Half lattice, 711, 911 142, 160, 163, 165, 187, 190, 195, 209,
Harmonic distortion, 408, 951, 956 219, 263, 267, 365, 368, 372, 456, 459,
Harmonic oscillator, 601 482, 614, 643, 659, 817, 834, 838
Hartley oscillator, 631, 634, 638, 641, 644 Imaginary value, 41, 71, 102, 103, 191, 213,
Heat, 73, 333, 335, 981 372, 842, 843, 847
Heating, 513, 1008 Impact, 333, 1021
Heat sink, 333, 335 Impact ionization, 333, 904
1046 Index
IMPATT, 279, 333–336, 338, 339, 344, 347, K

348, 355, 366, 373, 374, 377, 379, 380, 910 Ka-band, 334
Impedance, 73, 101, 105, 302, 303, 335, 336, KCL, 384, 413, 425, 436, 495, 604, 621–623,
405, 406, 410, 411, 422, 513, 573, 574, 633, 634, 646–648, 699, 716, 722, 724,
673, 712, 853, 873, 904, 941–948, 950, 749, 752, 1007
958–960, 962, 963, 970, 972, 973, Kelvin, 282
975–978, 980, 981, 1023 K-formula, 960
Independent variable, 9 Kirchhoff, 305
Index, 467, 785, 818, 849 Kirchoff’s laws, 996
Inductance, 2, 8, 14–16, 18, 22, 23, 48, 73, 78, Kirk effect, 561, 1022, 1024, 1028
101, 105, 109, 196, 281, 302, 303, 630, K-parameter, 17
747, 784, 840, 841, 849, 851, 854, 857,
873, 914, 915, 918, 919, 923, 925, 931, L
940, 970, 977, 978 Label, 6, 14, 15, 22, 23, 31, 33, 39, 43
Inductance element, 46, 786, 820, 905 Ladder, 602, 711
Inductive Coupled Plasma (ICP), 196 Large signal amplifier, 406
Inferior rectifier, 489 Large signal transient model, 1013, 1014
Infinite, 161, 167, 784, 961, 1008 Lateral diffusion parameter, 1012, 1013
InGaAs, 1020 Lattice, 711, 715, 717, 721, 723, 734, 736–738,
Initial condition, 13, 18, 527, 617 743, 745, 911, 1020
Injection locked mode, 373 LC circuit, 222, 223, 631, 673
Inject RF signal, 172, 338 LC matching network, 574
InP, 299, 302, 1020 LC matching topology, 947
In-phase, 631 Leakage current, 747, 1003, 1005, 1007, 1008,
Intercept, 408, 954, 955, 1018, 1020 1020
Interference, 22, 24, 42, 43, 50, 72, 194, Leaky switch, 1006
784, 785, 793, 818, 819, 825, 906, 952, Leg, 156, 196, 197, 202, 205, 213, 218, 219,
954, 956 746
Interfering signal, 957 LF–tag, 913, 923
Inter Modulation Distortion (IMD), 675, 951, Light, 280, 333, 875, 903, 982, 994
954, 955 Limit, 2, 8, 9, 16, 24, 36, 66, 97, 138, 164, 186,
Internal resistance, 6, 14, 944 208, 262, 335, 367, 410, 456, 482, 491,
Interrogation signal, 3 514, 574, 811, 834, 854, 915, 925, 956,
Interrogator, 73 987, 988, 990, 991, 994, 1003, 1008, 1010,
Intrinsic material, 334, 406, 489, 1017 1021, 1028
Intrinsic output resistance, 1023 Limit cycle, 2, 9
Invariant, 9, 617, 905 Limiter circuit, 155, 156, 161, 171
Ionization, 333, 1021 Limit relation, 38, 68, 99, 140, 188, 210, 264,
Ionosphere, 904 370, 457, 483, 614, 836, 908
Isolation, 382, 460, 493, 573, 995, 1025, 1028 Linear-algebra, 18
Linear combination, 18, 20
J Linear hybrid Pi-model, 408, 411, 423, 424,
Jacobian, 28, 54, 85, 134, 181, 205, 255, 291, 435, 575, 611
294, 355, 452, 453, 459, 475, 477, 798, 829 Linear neutral equation, 160
Jacobian matrix, 11, 569, 614, 629, 696, 710, Lipunov, 333, 422, 434, 501, 558, 584
745, 767, 797, 884, 906 L–Matching network, 406, 411, 422, 440, 444,
JFET, 171, 299, 338, 942, 943 452–454, 456–458, 574, 947
JFET common gate amplifier, 942 LNA, 514, 573, 574, 578, 579, 910, 941
Junction, 73, 101, 156, 172, 173, 279, 281, L–Network, 411, 422, 444, 947, 948, 967, 969,
299, 300, 304, 335, 338, 490, 492, 529, 970, 974
746, 747, 952, 956, 993–995, 1002, 1004, Local bifurcation, 2, 14, 73, 905
1005, 1007–1009, 1014, 1015, 1020, 1021, Local stability, 2, 22, 25, 46, 51, 81, 157, 161,
1023, 1025, 1029 178, 203, 229, 324, 347, 388, 417, 430,
Index 1047
444, 468, 498, 536, 579, 624, 638, 652, Mixture, 841, 918, 919, 922
682, 696, 704, 711, 736, 761, 794, 820, Mobile, 818, 873, 904
826, 869, 872, 880, 905 MOD-FET, 299
Lower cut-off frequency, 461 Modulate, 73, 491, 951
Lower sideband ladde, 711 Monolithic, 155, 338, 910
Low pass response, 411, 673, 674, 947, 948 Monolithic Microwave Integrated Circuit
L–Shape, 411, 947 (MMIC), 155, 171, 338, 411, 904
MOSFET, 299, 304, 993, 1010–1015, 1020,
M 1025, 1028
Maclaurin series, 58, 61, 65, 135, 182, 185, Multilayer, 785, 786, 793–797, 800, 804, 811,
257, 366, 803, 807, 830, 833 812, 814, 816, 911
Magnetic circuit, 620, 715 Multiplication, 20, 243, 250, 953, 1007, 1008
Magnetic material, 620, 715 Multiplication coefficient, 1007
Majority carrier, 993, 1009, 1017 Multiplication factor, 1008
Manifold, 9, 167, 614, 617, 630, 643, 659, 904 Multi Slotted (MS) scheme, 3
Maps, 31, 33, 35, 38, 68, 99, 140, 188, 206, Multi Slotted with Assigned Slots (MAS), 3
208, 210, 264, 370, 457, 483, 813, 836, Multi Slotted with Selective sleep (MSS), 3
905, 908 Multistage, 373, 374, 573
Matching network, 73, 104, 105, 406, 411, Mutual inductance, 46, 786, 793, 820, 849,
417, 420–422, 430, 433, 435, 574, 575, 850, 852, 855–857, 874, 907, 923
675, 676, 697, 903, 904, 910, 947, 948, Mutual interaction, 169
950, 958, 959, 965, 970, 972 Mutually perpendicular, 19
Material composition, 1027
MATLAB, 22, 42, 46, 103, 157, 191, 213–215, N
217, 786, 820, 843, 845, 915, 925 N+ doped substrate, 156
Matrix, 11, 17, 77, 177, 346, 420, 452, 453, Narrow band, 422, 959, 973
499, 569, 570, 572, 613, 614, 617 N-Doped, 279, 299, 746, 993
Matrix formulation, 17 Negative Differential Resistance (NDR), 156,
Matrix system, 17 336, 1006
Maximum Available Gain (MAG), 961, 964 Negative resistance, 156, 221–224, 333–337,
Maximum oscillation frequency, 1025 355, 373, 380, 601, 618
Maximum Stable Gain (MSG), 964 Neural differential equation, 161
MESFETs, 171, 338 Noise current, 174, 957, 958
Metal structure, 222 Noise Factor (NF), 302
Meter, 784, 919, 922, 931, 940 Noise Figure (NF), 573, 958
Mho, 305 Noise voltage, 174, 957, 958
Microelectronic, 857 Noisy period, 105
Microphone, 957 Non-autonomous, 9
Microstrip, 155–157, 169, 171, 279, 335, 344, Nonlinear behavior, 410, 1003
374, 382, 383, 435, 461, 471, 786, 820, Nonlinear distortion, 952
849, 853, 857, 877, 907, 910, 911, 980–982 Nonlinear dynamic, 1, 299, 618
Microwave amplifier, 221, 222, 279 Nonlinearity, 675, 954, 956
Microwave bipolar transistor, 280 Nonlinear system, 15
Microwave cavity, 222, 223 Non-negative parameter, 12, 159
Microwave Integrated Circuit (MIC), 155, 910 Non-uniform, 1018, 1027
Microwave oscillator, 156, 221–223, 228, 333, Normally ON, 299
601, 910 Normal Mode Helical Antenna (NMHA), 873,
Millimeter, 280, 299, 302, 304, 783 876
Minor, 300, 471 NPN, 515, 517, 530, 559, 586, 618, 619, 632,
Mismatched line, 950 644, 645, 993–995, 1001, 1002, 1007, 1015
Mismatch load, 950 N–Region, 333, 334, 337, 489, 490, 746
Mismatch Loss (ML), 946, 964, 981 N-Turn multilayer, 784–786, 792–795, 797,
Miter bend, 981 800, 802, 806, 811–813, 816
1048 Index
Number index, 202, 218 263, 267, 370, 372, 457, 459, 482, 488,
Numerator, 242, 541, 611, 975 568, 715, 813, 816, 836, 908, 910
Numerically, 36, 66, 71, 97, 115, 138, 164, Passive, 1–3, 12, 44–46, 105, 155, 406, 674,
186, 187, 194, 208, 212, 262, 291, 367, 676, 911, 948, 951
456, 479, 482, 524, 614, 696, 745, 797, Passive tag, 1, 2
811, 834, 847, 884, 906 PCB, 950, 962, 981, 982
PCB stray, 854
O PDEs, 905
ODEs, 905 P-Doped, 279, 993
Ohmic region, 221 Perfect Electric Conductor (PEC), 873
Ointment, 839, 913, 914, 923 Perfect match, 109, 959, 965, 966, 984
One frame, 5 Periodic, 12, 13, 20, 159, 601
Open circuit stable, 336 Periodic forcing, 9
Operational frequency, 335, 1009 Periodic function, 9
Operator, 165, 169 Periodic orbit, 9, 905
Optical spectrum, 903 Periodic solution, 9
Optimal bandwidth, 373 Periodic transformation, 13, 159
Optimization, 2, 14, 22, 44, 73, 299, 305, 784, Permeability, 223, 784, 839, 849, 874, 913,
785, 818, 849, 910, 911 923
Orbit, 2, 9, 12, 14, 19, 73, 617, 630, 905 Permittivity, 223
Orientation, 12, 44 Perturbation, 12, 614, 616
Origin, 18–20, 613 Phase plan, 12, 18
Oscilloscope, 221 Phase portrait, 19, 527, 568
Overlapping, 1013, 1014 Phase response, 408
Oxide thickness, 1015 Phase shift, 171, 410, 602, 611, 617, 618, 910,
Oxide trench, 1028 944, 977
Phase shift network, 602, 603
P Phase shift resonator, 617
P+ top, 156 PHEMT SSM, 305
Pad reactance, 950 Photo detector, 406
Parameter, 2, 12–14, 16, 20, 25, 31, 37, 42, 46, Physical proximity, 620, 715
47, 55, 57, 60, 64, 66, 68, 69, 71–73, 81, Pi–matching network, 406, 422, 947, 949
92, 96, 98, 99, 101, 103, 116, 131, 133, PIN diode, 155, 406, 489, 492, 494, 498, 500,
137, 139, 141, 143, 163, 168, 171, 176, 904
181, 184, 187, 191, 194, 195, 206, 217, Pitch, 873, 874
219, 256, 264, 300, 319, 366, 370, 373, Planar PCB, 46, 820, 923
454, 457, 459, 475, 479, 481, 483, 488, Plasma, 155, 196, 910
574, 659, 683, 705, 733, 738, 785, 795, P-N junction, 156, 172, 221, 338, 910, 994,
809, 811, 812, 814, 818, 832–834, 852 995
Parameter function, 2, 14, 15, 22, 23, 36, 38, PNP, 515, 517, 586, 993, 994, 998, 999, 1001,
39, 46, 51, 70, 73, 86, 96, 138, 163, 169, 1007, 1009, 1015
188, 194, 203, 211, 219, 229, 261, 267, Polar coordinates, 13
302, 348, 454, 459, 478, 517, 784, 804, PolyEthylene Terephthalate (PET), 904
819, 834, 873, 908, 978, 1015, 1028 Polynomial, 25, 51, 57, 60, 66, 81, 94, 96, 97,
Parasitic delay, 73, 77, 194, 197, 218, 435, 115, 137, 138, 161, 167, 178, 183, 186,
784 203, 229, 263, 324, 347, 362, 364, 368,
Parasitic effect, 104, 105, 113, 155, 156, 171, 389, 417, 430, 444, 455, 468, 480, 482,
181, 194, 202, 218, 279, 406, 435, 471, 498, 536, 579, 616, 624, 638, 683, 704,
676, 697, 785, 786, 818, 853, 877, 907, 736, 794, 801, 810, 826, 834, 880
911, 950 Poly-Si base, 1028
Parasitic inductance, 73, 101, 490 Positive feedback, 405, 601–603, 618, 944
Partial derivative, 36, 38, 67, 68, 70, 99, 139, Positive root, 36, 37, 68, 94, 97, 99, 136, 139,
142, 144, 164, 169, 187, 188, 190, 210, 140, 164, 167, 183, 187, 188, 209, 260,
Index 1049
263, 264, 362, 368, 369, 455, 457, 483, Receiver, 72, 73, 101, 105, 155, 156, 173, 174,
813, 908 494, 573, 673, 674, 712, 783, 903, 904,
Power circuit, 105, 559, 1020 942, 943, 951, 954, 956, 958
Power gain, 405, 410, 574, 903, 944, 964 Recombination, 490, 491, 994, 1003, 1017,
Power limiter, 155, 156, 158–161, 163, 164, 1025
169, 171 Rectangular spiral antenna, 104, 105, 109, 784,
Pre-collection mode, 333 818–820, 824–827, 829, 833, 834,
Printed, 8, 16, 72, 904, 915, 925 836–838, 843, 911, 914, 923, 924
Printed coil, 8, 16 Rectifier, 489
Projection, 9, 167 Reference direction, 620, 714
Propagation, 980, 981 Reflection, 109, 155, 171, 334–336, 339, 377,
PRrx, 6 406, 411, 903, 910, 941, 945, 947, 965, 981
PRtx, 6 Reflection Type Phase Shifter (RTPS), 171,
PTrx, 6 194
PTtx, 6 Re/Im, 615, 630, 696, 710, 745, 767, 797, 884,
Pulse based half duplex, 4 906
Pulse Width Modulation (PWM), 513 Relative permeability, 213
Pure imaginary, 20, 43, 44, 195, 219, 299, 365, Relaxation oscillator, 601
817 Resistance noise, 957
Purely resistive, 461, 973, 976 Resistive component, 461
Push pull amplifier, 514, 527 Resistivity, 489, 490, 854
Resistor, 2, 6, 14, 18, 46, 155, 169, 172, 174,
Q 221, 301, 338, 382, 383, 489, 491, 493,
Q–Point, 411 494, 515, 528, 574, 602, 603, 618, 631,
QT protocol, 4 644, 676, 697, 745, 786, 811, 820, 877,
Quadratic solution, 18 904, 923, 959, 963, 964, 974, 1023
Quadratic term, 11, 568, 569 Resonance frequency, 14, 631, 874
Quarter wavelength long, 335 Resonant, 155, 196, 222, 223, 333, 334, 528,
Quarter wave line transformer, 979 529, 603, 631, 673, 674, 711, 712, 874,
Quartz crystal, 631, 673, 674, 712 903, 948, 970
Query, 3 Response mechanism, 4
Query Tree (QT), 4 Resultant output signal, 951
Reverse active mode, 995, 1005
R Reverse bias, 279, 280, 301, 333, 493, 494,
Radar, 73, 333–335, 674, 904 746, 747, 995, 1002, 1004, 1005, 1010,
Radial wave number, 875 1013, 1015, 1017, 1026
Radio frequency identification, 1 Reverse biased varactor, 674
Radius, 108, 784, 785, 819, 839, 873, 918, Reverse biasing, 280, 489, 493, 1008
919, 922, 930, 931 Reverse common emitter current gain, 282
Random Access Memory (RAM), 1 Reverse mode, 995
Randomly, 3, 956 Reverse saturation, 996, 1013
Reactance, 101, 280, 337, 339, 461, 490, 631, Reversing, 12
746, 874, 959, 970, 972, 974, 978, 980 RFC, 631, 942, 943, 963
Reader energy consumption, 6 RFID, 1–4, 6, 8, 9, 11–28, 31, 33, 35–38, 40,
Reader query, 4, 5 42, 44, 46, 48, 51, 52, 55, 57, 61, 65, 67,
Reader transmits energy, 4 68, 70–73, 78, 81, 85, 97, 98, 100, 101,
Real number, 30, 32, 34, 56, 59, 63, 94, 137, 104, 105, 108, 109, 113, 115, 117,
183, 206, 207, 213, 260, 298, 363, 455, 134–136, 138–140, 143, 144, 783–786,
480, 696, 710, 801, 804, 808, 831, 843, 792, 794–797, 812, 818–820, 824–827,
871, 872, 904 829, 833, 836–839, 841–843, 845, 848,
Receive mode, 4 873, 876, 904, 905, 911
1050 Index
RFin signal, 77, 177, 286, 290, 293, 301 Sleep commands, 3, 6
Room temperature, 282, 492, 996, 1023 Small displacement, 25, 51, 116, 162, 178,
Root, 30, 32, 34, 36, 37, 40, 42–44, 56, 59, 60, 324, 348, 795, 826
63, 64, 66, 68, 70, 72, 94, 96–99, 101, 103, Small disturbance, 10, 527, 568
136–138, 142, 144, 160–165, 171, 183, Small increment, 25, 26, 51, 52, 54, 82, 115,
184, 186, 187, 190, 191, 195, 206, 208, 117, 128, 161, 178–180, 203, 204, 229,
209, 212, 218, 260, 262, 263, 267, 362, 230, 254, 255, 324, 330, 347–349, 352,
365, 367–369, 372, 373, 455–457, 459, 353, 355, 389, 417, 420, 430, 432, 433,
480–483, 488, 490, 614, 629, 696, 697, 444, 459, 468, 476, 498, 536, 579, 624,
710, 745, 767, 797, 801, 804, 805, 808, 627–629, 638, 641, 642, 652, 653, 657,
810–812, 816, 817, 831, 832, 834, 838, 682, 683, 686, 704, 707, 736, 743, 745,
848, 872, 884, 906, 907, 957, 961 761, 765, 767, 794, 798, 826–828, 869,
Rotation, 20, 155, 171 870, 880, 883
Round strip, 197 Small signal amplifier, 407, 411
Routh-Hurwitz, 333, 422, 434, 501, 558, 584, 616 Small Signal Model (SSM), 338
Solar flare, 957
S Source, 980, 983, 1012, 1015, 1019
S/m, 785, 841, 919 Source contact resistance, 1013
S2P, 960 Space charge layer, 334
Saddle point, 298, 572 S–Parameters, 960, 965
Saturation current, 221, 1003 Spectrum, 222, 673, 957
Scalar, 161, 964 Speed, 51, 81, 179, 348, 499, 903, 1015, 1021,
Schottky diode, 73, 102 1024
Schottky junction, 338 Spice Gummel Poon (SGP) model, 559
Segment, 43, 72, 195, 219, 222, 855, 873, 876 Spiral, 14, 19, 22, 298, 818–820, 825–828,
Selective sleep scheme, 3 830, 833, 834, 839, 842, 844, 845, 848
Self-bias method, 300 Spiral antenna, 818, 819, 825, 827, 834, 916,
Semi-passive RFID, 44–46 923
Separation distance, 45 Spiral converge, 18
Sequential (SEQ) operational, 4 S–Port, 674
Shichman, 1010 Spurious, 493, 954
Shifted gate, 44, 45, 50–52, 54, 57, 65 Spurs, 962
Short analysis, 194, 218 Squaring, 167, 188, 264, 457
Shunt, 101, 156, 157, 574, 963, 977, 978 SSB, 674
Shunt PCB, 854 Stability, 2, 21–23, 28, 35, 38–40, 42, 44, 55,
Siemens, 840, 919, 931 70–73, 86, 99, 101, 102, 104, 115, 140,
SiGe, 1015, 1017, 1020 143, 144, 155, 157, 160, 163, 167, 170,
Sign, 812, 816, 835, 838, 907, 910 181, 190–192, 194, 205, 208, 212, 217,
Signal, 461, 489, 491, 493, 513, 515, 529, 559, 219, 223, 256, 259, 267, 279, 298, 319,
574, 602, 631, 674, 904, 941, 944, 951, 333, 356, 358, 360, 366, 372, 373, 383,
952, 958, 981, 1003, 1010, 1019, 1024 392, 406, 411, 422, 434, 435, 453, 457,
Signal feeding, 197 459, 476, 478, 487, 488, 493, 501, 514,
Signal source, 334, 1023 558, 573, 574, 584, 601, 603, 614, 619,
Silicon, 73, 299, 489, 490, 1020 632, 643, 645, 659, 660, 686, 696, 697,
Silver, 839, 841, 913, 918, 919, 931 710, 745, 784, 786, 799, 813, 816, 818,
Simulation, 820, 1013, 1018, 1026, 1029 820, 829, 838, 848, 857, 872, 903,
Simulation model, 514, 1026 905–907, 910, 911
Sin, 167 Stability analysis, 28, 35, 43, 55, 65, 72, 155,
Sine wave, 528, 956 181, 195, 205, 219, 256, 333, 405, 411,
Single Pole Single Throw (SPST, 493 435, 454, 459, 489, 559, 601, 611, 630,
Siphon, 1009 644, 673, 674, 696, 697, 711, 783, 799,
SI unit, 840, 841, 918, 922 829, 872, 906, 910
Skin depth, 490 Stability switch, 22, 42, 44, 46, 68, 71, 102,
Skin effect, 490 103, 157, 188, 191, 194, 195, 210, 212,
Index 1051
217, 220, 264, 360, 370, 372, 373, 483, Track, 818, 840, 913, 915, 918, 922, 940
488, 816, 817, 820, 848, 908 Track thickness, 8, 915, 925
Stable focus, 812, 910 Trajectory, 19, 527
Stable node, 19, 298, 572 Trans conductance, 1015
Stable spiral, 298, 573 Transducer unilateral gain (Gth), 964
Stable steady state, 630, 710, 797, 884 Transformation, 712, 722, 975
Standing Wave Ratio (SWR), 950 Transformer, 174, 515, 528, 712, 904, 977
Star, 422, 957 Transformer’s coil, 620, 714
Star node, 20, 22 Transistor, 629–631, 644, 910, 952, 953, 956,
Start of frame, 5 959, 994, 995, 1002, 1003, 1006, 1008,
Static avalanche region, 1007 1012, 1018, 1026, 1029
Stored energy, 461 Transition, 163, 190, 459, 909
Straight, 17, 849, 851, 854, 855, 857, 862, 869, Transit-time, 333
871, 872, 963, 1018 Transmission line, 155, 281, 339, 574, 950,
Straight line solution, 17 958, 980, 981
Stray capacitance, 972, 973 Trigger circuit, 221
Strip, 854 Trivial solution, 160
S-Type, 336 Tunable, 674, 746, 761
Subscript, 68, 165, 210, 457, 836, 1017 Tunable BPF, 745, 748, 761
Subspace, 617 Tunnel Diode (TD), 221–223, 228, 910
Substantial current, 1006 Tunneling, 221, 223
Suck, 1002 Turn, 489, 491, 619, 713, 847, 849, 857, 870,
Sunspot, 957 872, 876, 1008
Surface inversion potential, 1012 Turn ratio, 174
Surface mount, 460
Symmetric, 172, 923 U
UH, 915
T UHF television tuner, 221
TAG ID, 2–4, 6 Unforced oscillator, 9
Taylor expansion, 207, 212 Unforced system, 13
Taylor series expansion, 58, 61, 65, 182, 185, Uniformly asymptotically stable, 160
366, 807 Uniformly bounded, 160, 161
Terahertz, 903 Unperturbed system, 12
Terminology, 35, 65, 96, 138, 164, 185, 262, Unstable, 161, 179, 190, 203, 222, 229, 325,
338, 422, 481, 674, 975 411, 459, 498, 614, 616, 696, 710, 795,
Tetrode, 221 826, 906
Thermal voltage, 282, 561 Unstable multi-vibrator, 601
Thickness, 106, 156, 300, 490, 849, 854, 981, Unstable node, 19
1021 Unstable spiral, 22
Thin plate, 48, 825 Unstable steady state, 630, 696, 745, 906
Third Order Intercept Point (TOI), 408 Upper cut-off frequency, 461
Third time interval, 169 UV, 903
Three dimensional autonomous systems, 9
Three elements circuit, 874 V
Threshold, 46, 156, 820, 1012 Valley, 224, 950
Time delay, 29, 38, 43, 46, 72, 157, 194, 195, Valley voltage, 950
219, 222, 360, 410, 488, 825, 907 Van der Pol’s system, 8
Time slot, 4 Varactor, 173, 338, 746, 761, 911
T–Matching network, 383, 406, 411, 422, 676, Variable, 177, 194, 218, 282, 344, 389, 430,
948 444, 489, 498, 536, 579, 624, 653, 745,
T–Network, 976 840, 919
Topologically, 9 Variable parameters, 45, 73
Trace, 46, 923, 950 Variation argument, 208, 367, 804, 811
1052 Index
Varicap, 746 Wide, 1, 489, 903, 904, 956

Varying amplitude, 13 Wideband LNA, 573, 574
Vast majority, 1002 Wideband signal, 410
Vector Network Analyzer (VNA), 460 Wide intrinsic region, 489
Velocity, 280, 980, 981 Widening, 1022, 1024, 1028
Vertical Bipolar Intercompany model (VBIC), WiMAX, 818
559 Winding, 618
Vice versa, 674, 747, 816, 835, 903, 910, 1021 Winding height, 785
Video load resistance, 101 Winding thickness, 785
Video resistance, 73 Wire, 785, 819, 825, 841, 873, 918, 994
Voltage derivative, 222 Wire cross section area, 785
Voltage generator, 2, 46 Wired, 915, 926
Voltage source, 14, 337, 575, 585, 619 Wye-delta transformation, 422, 975
Voltage Standing Wave Ratio (VSWR), 573
X
W X–Band, 490
Wait time, 5 X-Ray, 903
Wake up, 72
Watchdog, 105 Z
Wavelength, 875, 876, 904, 950, 981 Zero frequency amplifiers, 910
Wave number, 875 Zeros, 36, 46, 56, 94, 136, 157, 171, 186, 260,
Weak signal, 73 263, 368, 455, 456, 786, 812, 820, 907
White noise, 956 Z–Match network, 970, 973

Mraacnaie PDF

Загружено:

Сведения о документе

Оригинальное название

Авторское право

Доступные форматы

Поделиться этим документом

Поделиться или встроить документ

Параметры публикации

Этот документ был вам полезен?

Это неприемлемый материал?

Авторское право:

Доступные форматы

Mraacnaie PDF

Загружено:

Авторское право:

Доступные форматы

Ofer Aluf

ISBN 978-3-319-45425-2 ISBN 978-3-319-45427-6 (eBook)

© Springer International Publishing Switzerland 2017

Printed on acid-free paper

This Springer imprint is published by Springer Nature

This book on microwave RF circuits: nonlinearity applications in engineering

several classes of delayed differential equations and Microwave RF Circuits and

(monolithic microwave integrated circuit), diodes, transistors, and microstrip

Chapter 7. Filter Systems Stability Analysis. In this chapter, ﬁlter systems in

Netanya, Israel Ofer Aluf

1 RFID Antennas Systems Descriptions and Analysis . . . . . . . . . . . . . . 1

3 Microwave Semiconductor Ampliﬁers Analysis. . . . . . . . . . . . . . .... 279

6.3 Hartley Oscillator Stability Analysis . . . . . . . . . . . . . . . . . . . . . . . . 630

Microwave RF antenna products are currently in a widely use in all aspects of

The input control lines can be connected to additional embedded system by

connected to additional RF systems and devices. The overall Microwave RF

The Microwave RF systems can be represented as RF input circuit and RF

Some of the Microwave/RF systems can be represented as an equation in

Microwave/RF systems equations can be turned to the dynamics and bifurca-

1.1 Active RFID TAGs System Analysis of Energy

features: Longest communication range of any TAG, the capability to perform an

Fig. 1.1 Reader TAG

Fig. 1.2 Reader TAGs system query and responds

Reader Wait time

Fig. 1.3 One frame reader query and wait time

TAG n, TAG n+1

No. of TAGs respond YES

Fig. 1.4 Flow chart for reader query and TAGs

One slot for a query F slots reader wait

Active RFID TAG can represent as a parallel Equivalent Circuit of Capacitor

Active RFID TAG

Fig. 1.6 Active RFID TAG system

Rs Active RFID Antenna

Active RFID's Equivalent circuit

Fig. 1.7 Active RFID TAG’s equivalent circuit

V ¼ Vc1 ¼ VL1 ¼ VR1

1 dV d2V 1 dV 1 dVsðtÞdt !eð0\e1Þ

Lcalc ¼ 0 ½X1 þ X2 X3 þ X4 Ncp

The RFID’s coil calculation inductance expression is the deﬁnition of limits,

Coil manufacturing technology P

“f” is a “T” periodic function of the independent variable t, and k ¼ Rs 1

The term k fs ðtÞ ¼ Rs V_ S ðtÞ is called the forcing function k ! 0 ) Rs

First examine the case of k ¼ 0 ) Rs C1 ! 1; C1 ¼ const; then Rs ! 1

Now we consider linear system: dV dt ¼ f ðV1 ; V2 Þ; dt ¼ gðV1 ; V2 Þ

And suppose that ðV1 ; V2 Þ is a ﬁxed point: f ðV1 ; V2 Þ ¼ 0; gðV1 ; V2 Þ ¼ 0

derive differential equations of U1 and U2. Let’s do the U1 equation ﬁrst: dU

Since V1 being constant.

Who’s dynamic can be analyzed by the general methods.

The basic Active RFID Forced Van Der Pol’s equation

In our case /ðVÞ ¼ 1; /ðVÞ [ 0 for jVj [ 1 and V_ S ðtÞ is T periodic

/ðVÞ ¼ 1 remain strictly positive as jVj ! 1 for unforced system, R11C1

Which “freezes” the unperturbed system and

This transformation is orientation reversing approximations the function n1; n2

Constant u0 is being determined by initial conditions.

Active RFID TAG system can be represented as Voltage source (internal

1.2 RFID TAG’s Dimensional Parameters Optimization

RFID Equivalent circuits of a Label can be represented as Parallel circuit of

C1 ¼ Cic þ Ccon þ Cc; R1 ¼ ðRic RpcÞ=ðRic þ RpcÞ:

The system shape is as nonlinear system equations:

dV1 dV2 1 1 dV3

d ¼ 2 ðt þ wÞ=p; Aavg ¼ a0 Nc ðg þ wÞ; Bavg ¼ b0 Nc ðg þ wÞ

a0, b0—Overall dimensions of the coil. Aavg, Bavg—Average dimensions of the

Fig. 1.9 RFID’s coil dimensional parameters

1 dV d2V 1 dV 1 dVsðtÞdt !eð0\e1Þ