A MATLAB-BASED TRANSMISSION LINE VIRTUAL TOOL:

FINITE DIFFERENCE TIME DOMAIN REFLECTOMETER

Levent Sevgi, aatay Uluk
Dou University, Electronics and Communication Engineering Department,
Zeamet Sok. No. 21, Acbadem / Kadky, 34722 Istanbul - Turkey


Abstract
This article introduces a simple virtual tool TDRMeter
1
for the investigation and
visualization of time domain pulsed voltage/current traveling along a terminated finite
length transmission line without and with faults somewhere between the source and
load. The package can be used as an education tool in various undergraduate
lectures to aid to teach electromagnetics as well as transmission lines.

Keywords Transmission lines, FDTD, pulsed voltage source, MATLAB, GUI,
simulation, visualization, engineering education, time domain reflectometer.

I. Introduction
A set of virtual tools has been introduced to the attention of the IEEE APM readers
[1-4] that can be used both for educational and engineering purposes. The
SNELL_GUI [1] may be used to visualize ray paths in two-dimensional (2D)
propagation environments along simple surface terrain profiles and with variable
atmospheric refractivity. The RAY_GUI [2] deals with the visualization of individual
rays and modes, and their field contributions inside the classical 2D parallel plate
non-penetrable waveguide. The SSSPE_GUI [3] simulates 2D groundwave
propagation in 2D along user-specified, non-flat terrain profiles, through specific
vertical refractivity profiles. The ANTEN_GUI [4], accommodated with beam steering
capabilities, was designed to plot 2D and 3D radiation patterns of a number of
selected and user-located isotropic radiators (i.e., arrays). These virtual tools, if used
properly, certainly increase teaching efficiency in classical undergraduate lectures,

1
TDRMeter can be downloaded from http://www3.dogus.edu.tr/lsevgi (or /culuisik) web site

2
such as Electromagnetic Field/Wave Theory, Antennas and Propagation,
Wireless Communication, Engineering electromagnetics, etc.

Another virtual tool, time-domain reflectometer (TDR) based the discretization of time
domain transmission line (TL) equations in terms of finite-difference approximation, is
introduced in this paper. The TDR is a remote sensing electrical measurement
device/system that has been used for many decades to determine the spatial location
and nature of various objects. Originally developed by Fellner-Feldegg [5] in the late
sixties TDR has been employed

in a variety of fields for determining the electrical
parameters of materials over a wide frequency bandwidth.

An early form of TDR,
dating from the 1930's, which most people are familiar with, is RADAR. Other forms
of the TDR are LIDAR, coax TDR, optical fiber OTDR and broadband impulse radars.

The TDR is used to locate and identify faults in all types of metallic paired cable. It
can locate major or minor cabling problems including; sheath faults, broken
conductors, water damage, loose connectors, crimps, cuts, smashed cables, shorted
conductors, system components, and a variety of other fault conditions. The basic
principle of the TDR is fairly simple. A pulse is generated and sent down the TL. If the
line is terminated by a load that is equal to the characteristic impedance of the line,
no reflected wave or echo will be returning to the generator or source, otherwise
reflection occurs. By measuring the time difference between the incident pulse and
the echo (reflected pulse), it is possible to determine the distance to the fault that
produced the echo pulse. Calculations must include the velocity factor of the cable.
Also, detailed analysis of the echo signal can reveal additional details of the faults or
reflecting objects.

The speed and accuracy of the TDR makes it today's preferred method of cable fault
location. Although today's instruments are more user friendly, a good understanding
of the basic principles and applications of a TDR is essential to successful
troubleshooting. The pulse generated by the TDR takes a certain amount of time and
thus distance to launch. This distance is known as the blind spot. The length of the
blind spot varies with the pulse width. The longer the pulse width, the larger the blind
spot. It is more difficult to locate a fault contained within the blind spot. If a fault is
suspected within the first section of the cable, it is advisable to add a length of cable
3
(jumper) between the TDR and the cable being tested. Any faults that may have been
hidden in the blind spot can now easily be located (only remember to take the length
of the jumper cable into account). It is best if the jumper cable has the same
characteristic impedance with the cable under test.

II. Time Domain Transmission Line (TL) equations
A transmission line (TL) is more than a set of long, parallel line; it is a distributed-
parameter physical system. The TLs are used to transmit electric energy and
communication signals from one point to the other. A basic TL connects a source to a
load. This may be a transmitter and an antenna, a television or a radio antenna and a
receiver, one port of a coupler and a power meter, etc. A typical two-wire TL is
sketched in Fig. 1a. A TL has two sets of constitutive parameters. The primary
parameters are listed in Table 1. A circuit equivalent of a TL is pictured in Fig. 1b.
The secondary parameters are the characteristic impedance Z
0
and complex
propagation constant. The real and imaginary parts of the propagation constant are
loss and phase variation per unit length, respectively.

There is a direct analogy between a uniform TL and a uniform plane wave. The field
distributions for the uniform plane wave and for the uniform TL are both known as
transverse electromagnetic (TEM) waves, where both electric and magnetic fields are
perpendicular to the direction of propagation. Because of this analogy TEM wave
approach may be used in analyzing TL, which is the EM field approach. Another
approach is the circuit (or network) approach where voltages and currents in terms of
two independent variables, time and distance, can be used to establish sets of
differential equations by using the model in Fig. 1b.

For a uniform TL (along x-direction) having primary parameters R [/m], L [H/m], C
[F/m] and G [S/m] per unit length, the coupled time domain TL equations of the
voltage and current in differential forms may be written as
0 ) , (
) , ( ) , (
= +

t x Ri
t
t x i
L
x
t x v
(1a)
0 ) , (
) , ( ) , (
= +

t x Gv
t
t x v
C
x
t x i
(1b)
4
(here, v(x,t), i(x,t) are the space- and time-dependent voltage and current,
respectively) while Maxwell equations for a uniform plane wave propagating along
the +x direction inside a homogeneous medium, and having only
y
E and
z
H
components can be written as
0
) , (
) , (
=

t
t x H
x
t x E
z
y
(2a)
0 ) , (
) , (
) , (
= +

t x E
t
t x E
x
t x H
y
y
z
(2b)
In (2), [F/m], [H/m], and [S/m] are permittivity, permeability and conductivity of
the medium, respectively. Inspecting equations (1) and (2) shows the quantities that
are analogous as listed in Table 2, therefore classical 1D FDTD method can easily be
applied [6].

2A) FDTD representations: Lossless case
The representations in (1) and (2) are identical if the TL is infinite, homogeneous and
loss-free (i.e., R=0, G=0), and if the electromagnetic medium of the plane wave is
lossless (=0). In other words, one can obtain EM equations (2) from transmission
line equations (1) by just replacing magnetic and electric fields with current and
voltage, and permittivity and permeability with unit length conductance and unit
length inductance, respectively.

The 1D FDTD equations of (1) in the well-known leap-frog scheme is given as [7]
( ) ) 1 ( ) ( ) ( ) (
0
1

=

k H k H
x
t
k E k E
n
z
n
z
n
y
n
y

(3a)
( ) ) 1 ( ) ( ) ( ) (
0
1

=

k E k E
x
t
k H k H
n
y
n
y
n
z
n
z

(3b)
which connect the nodes of electric fields to the two adjacent nodes of magnetic
fields (or, vice versa). Here the integers, k and n, respectively, represent spatial (x)
and time (t) indices, so that physical space and time values are specified via
) ( ) , ( ) , ( k E n k E t x E
n
y y y
= = (4a)
) ( ) , ( ) , ( k H n k H t x H
n
z z z
= = (4b)
(by using x k x = and t n t = ).

5
Using the analogy mentioned above iterative FDTD equations of the loss-free TL may
be written directly as
( ) ) 1 ( ) ( ) ( ) (
1

=

k i k i
x C
t
k v k v
n n n n
(5a)
( ) ) 1 ( ) ( ) ( ) (
1

=

k v k v
x L
t
k i k i
n n n n
(5b)
The leap-frog scheme of (5) is given in Fig. 3. As in 1D FDTD simulations, the
stability and the discretization conditions are given as
c x t / ,
LC
c
1
= (6a)
10 /
min
x (6b)
Since, the characteristic impedance of a loss-free TL is C L Z /
0
= equations (5) can
also be written as
( ) ) 1 ( ) ( ) ( ) (
0
1
=

k i k i Z k v k v
n n n n
(7a)
( ) ) 1 ( ) (
1
) ( ) (
0
1
=

k v k v
Z
k i k i
n n n n
(7b)
which represent traveling of voltage and current pulses along an infinite (i.e.,
reflection-free or unterminated) TL.

2B) FDTD representations: Lossy case
Following the same strategy the 1D FDTD representations of a lossy TL can be given
as [6]
( )
x
k i k i
G
t
C
k v
G
t
C
G
t
C
k v
n n
n n

(
(
(
(

(
(
(
(

=

) 1 ( ) (
2
1
) (
2
2
) (
1
(8a)
( )
x
k v k v
R
t
L
k i
R
t
L
R
t
L
k i
n n
n n

(
(
(
(

(
(
(
(

=

) 1 ( ) (
2
1
) (
2
2
) (
1
(8b)
These equations directly reduce to (5) if R=0 and G=0. It should be noted that, if the
number of nodes along the TL is N, the voltage and current calculations in (5) and/or
(8) starts from the first node (k=1) and ends on the last node k=N). On the other
hand, each node uses the two neighboring nodes and first and last node have only
one neighboring nodes. If not specified by the user manually, the computer programs
6
assume zero voltages for the unavailable nodes and this correspond to open-circuit
(OC) termination.

2C) FDTD representations: Fault modeling along a TL
Inhomogeneouity of a TL can be introduced in various different ways; a series of
different transmission lines may be assumed cascaded [6], or any/all of primary
parameters of the line assumed to be different at specific nodes. The first approach
requires the satisfaction of the continuity equations at the interface since iterative
equations based on the differential operators become invalid. If the second approach
is used, then one needs to replace the primary parameters R, L, C, and G in (8) with
R, L, C, and G where
f
R R R + = ,
f
G G G + = ,
f
L L L + = , and
f
C C C + = ,
where
f
R ,
f
G ,
f
L , and
f
C being the unit-length parameters of the fault.

2D) FDTD representations: Terminated lines
The TL representation given in Fig. 1a includes a voltage source v
s
(t) with internal
resistance R
s
at the left end and a load resistor R
L
at the right. Its symmetrical circuit
model is given in Fig. 2. The source-excited and load terminated TL uses the same
equations (8) for all the nodes except the first and last ones. For the first node,
Thevenin source as shown in Fig. 1a can be replaced by its Norton equivalent as
given in Fig. 4a and Kirchhoffs current law (KCL)
s
n n
n
s
n
s
R
v
t
v
x
C i
R
v
) 1 ( ) 1 (
2
) 1 ( +

+ = (9)
can be used to obtain the discrete
{ }
( )
2
) 1 (
2
1
2
1
) 1 (
2
1
2
2
1
2
) 1 (
1
1

(
(
(
(

(
(
(
(

=
n
s
n
s n
s
s
n
s
s
n
v v
i R
t
x C
R
v
t
x C
R
t
x C
R
v (10a)
form that can be used for the first (k=1) node. If similar procedure is followed for the
last (k=N) node (see Fig. 4b)
) 1 (
2
1
2
1
) (
2
1
2
2
1
2
) (
1

(
(
(
(

+
(
(
(
(

=

N i R
t
x C
R
N v
t
x C
R
t
x C
R
N v
n
L
L
n
L
L
n
(10b)
is obtained and can be used to represent load termination.

7
2E) FDTD representations: Complex termination
In practice, the TLs are mostly terminated by complex loads. Examples of parallel RC
(e.g., see Fig. 4b) and serial RL terminations are given below.

Parallel R
L
C
L
termination
If the KCL is written for the last node
t
N v
C
t
N v
z C
R
z G
N v N i
n
L
n
L
n n

+
|
|
.
|

\
|
+

=
) ( ) (
2
1
2
) ( ) 1 ( (11a)
and, is discretized in terms of FDTD principles the iterative form of this complex loads
(parallel RC) is derived as

) 1 (
1
2 2
1
2
2
1
) (
1
2 2
1
2
2
1
2 2
1
2
2
) (
1

(
(
(
(
(

|
|
.
|

\
|
+

+
+
+
(
(
(
(
(

|
|
.
|

\
|
+

+
|
|
.
|

\
|
+

+
=

N i
R t
x G C x C
N v
R
x G
t
C x C
R
x G
t
C x C
N v
n
L
L
n
L
L
L
L
n
(11b)
It should be noted that choosing R
L
=0 for the parallel RC load must reduce to SC
termination, unfortunately, (11b) does NOT automatically yields this result. Therefore,
(11b) must be replaced with
) ( ) (
1
N v N v
n n
= (11c)
if R
L
=0 is selected for the parallel RC termination.

Serial R
L
L
L
termination
A similar procedure may be applied for the complex termination with serial resistor and
inductor [6]. Adding the complex load directly to the serial arm of the last node of the
symmetrical model (see Fig. 4b) gives the probability of writing Kirchhoffs voltage law
(KVL). Following this procedure, using the central difference approximation, and
interpolating the current in time at the last node results in [6]
8
) (
2 2
1
2
2
1
) (
2 2
1
2
2
2 2
1
2
2
) (
2 / 1
1
N v
R
x R
t
L x L
N i
R
x R
t
L x L
R
x R
t
L x L
N i
n
L
L
n
L
L
L
L
n
+
+
(
(
(
(

|
.
|

\
|
+

+
+
(
(
(
(

|
.
|

\
|
+

+
|
.
|

\
|
+

+
=
(12a)
In this case ) (N v
n
can be calculated from ) (N i
n
and ) (
1
N i
n+
as
( )
x
N i N i
G
t
C
N v
G
t
C
G
t
C
N v
n n
n n

(
(
(
(

(
(
(
(

=

) 1 ( ) (
2
1
) (
2
2
) (
1
(12b)

Parallel R
L
L
L
C
L
termination
Adding an inductance to the load forms a kind of parallel resonance termination. In
this case writing the KCL at the last node results in
) (
) ( ) (
2
1
2
) ( ) 1 ( t i
t
N v
C
t
N v
x C
R
x G
N v N i
n
L
n
L
n
L
n n
+

+
|
|
.
|

\
|
+

= (13a)
which requires addition of the inductance current

0
0
) (
1
) ( i dt t v
L
t i
L
t
L
L
+ =

. (13b)
Assuming zero initial inductor current, using

=
2
) ( ) (
) ( ) (
1
2 / 1 2 / 1
2 / 1
0
0
N v N v
L
t
N v
L
t
dt t v
L
n n
L
n
n
k L
n
t
, (14)
and re-arranging (13a) yields
{ } sk N i
L
t
R
x G
t
C x C
N v
L
t
R
x G
t
C x C
L
t
R
x G
t
C x C
N v
n
L L
L
n
L L
L
L L
L
n

(
(
(
(
(

|
|
.
|

\
|
+ +

+
+
(
(
(
(
(

|
|
.
|

\
|
+ +

+
|
|
.
|

\
|
+ +

+
=
+
) 1 (
1
2 2
1
2
2
1
) (
1
2 2
1
2
2
1
2 2
1
2
2
) (
2 / 1 2 / 1
(15)
9
for the parallel RLC termination with
) (
2 / 1
2 / 1
N v
L
t
sk
k
n
k L

=

=
, (15b)
(sk) being the inductor current calculated from the time integral of the voltage.

Serial R
L
L
L
C
L
termination
Adding a serial contuctor to the load forms a kind of serial resonance termination.
Writing the KVL at the last node yields
) (
) ( ) (
2 2
) ( ) (
2 / 1 2 / 1
2 / 1 2 / 1
t v
t
N i
L
t
N i
x L x R
R N i N v
C
n
L
n
L
n n

+ |
.
|

\
|
+ =
+ +
+ +
. (16a)
0
0
) (
1
) ( v dt t i
C
t v
C
t
L
C
+ =

. (16b)
Again, assuming no stored capacitive energy at the load (i.e., 0
0
= v ), using

2
) ( ) (
) ( ) (
1
1 1
1
0
0
N i N i
C
t
N i
C
t
dt t i
C
n n
L
n
n
k L
n
t
L
(17)
and re-arranging (16a) results in
{ } sk N v
C
t
R
x R
t
L x L
N i
C
t
R
x R
t
L x L
C
t
R
x R
t
L x L
N i
n
L
L
L
n
L
L
L
L
L
L
n
+
(
(
(
(
(

|
|
.
|

\
|
+ +

+
+
(
(
(
(
(

|
|
.
|

\
|
+ +

+
|
|
.
|

\
|
+ +

+
=
+
+
) (
2 2 2
1
2
2
1
) (
2 2 2
1
2
2
2 2 2
1
2
2
) (
2 / 1
1
(18a)
for the serial RLC termination with
) (
1
1
N i
C
t
sk
k
n
k L

= . (18b)
It should be noted that, (18) requires calculation of the voltage at the last node via
( )
x
N i N i
G
t
C
N v
G
t
C
G
t
C
N v
n n
n n

(
(
(
(

(
(
(
(

=
+
) 1 ( ) (
2
1
) (
2
2
) (
2 / 1 2 / 1
. (19)


10
2F) Source specifications
Numerical dispersion determines the selection of discretization steps x (t is
determined by the Courant stability criteria) [8]. In practice, discretization values
between 10 /
min
= x and 150 /
min
= x depending whether the phase information is
required or not. When pulsed signals are used the highest frequency content of the
pulse should be taken into account. The ideal rectangular pulse has zero rise and fall
times, which means it has infinite number of sinuzoidals in the frequency domain.
Therefore, numerical dispersion effects are unavoidable for the rectangular pulse
source. On the other hand, the discretization parameters can be specified
accordingly for the Gaussian and/or trapezoidal pulses For example, the duration of
the Gaussian pulse may be specified automatically in terms of x. For this purpose
x is specified in terms of the length of the TL and the number of nodes first. Then,
the time step t is calculated from (6a). The bandwidth of the Gaussian pulse may
be specified from ) 10 /(
max
x c f B = = if the spatial extend of the pulse is assumed
to be 10 x . Finally, the two parameters of the Gaussian pulse ( ) { }
2
1 exp ) ( = t g
may be chosen accordingly as
2
3 . 3 B = and / 4 = . In this case 10 /
min

sampling is guaranteed automatically [8].

III. The Matlab-based TDRMeter Package
The TDRMeter package is a multi-purpose program designed in Matlab 6.5 (it is
strongly advised to the reader to review TL theory both in time and frequency
domains if he/she feels uncomfortable with equations given here. There are excellent
classical books and internet sources and there is no need to refer them. The authors
used [6] and [8], and references listed there, for this purpose).

The front panel of TDRMeter is given in Fig. 5. There are four input data blocks on
top of the panel. The user supplies the unit-length TL parameters on the left. The
mid-left block is reserved for the source parameters. The user may select one of
three different source types; Gaussian pulse, rectangular pulse and a trapezoidal
pulse, with pulse duration and rise/fall times (if the source is trapezoidal). Internal
source resistor is also supplied inside this block. The pulse length of the Gaussian
voltage source is automatically selected as explained in the previous section. The
third block on mid-right is used for the specification of the load. It may be a resistive
11
load, a parallel combination of a resistor and capacitor, or a serial combination of a
resistor and inductor. The user selects the load type from the pop-up menu and
supplies the numerical values. The last block at the right is used for the fault
definition. The user is allowed to change only unit-length parallel elements (C and G),
but this can be altered by the user easily by changing the codes in the M-file. The TL
length and the observation point are supplied at the right top of the front panel. All the
units should be given as specified on the front panel.

Two plots are used for the visualization purposes. The outcome of the FDTD
simulations can be given as either signal vs. TL length at each simulation time, or
signal vs. time at the specified observation point. The figure on the front panel is
reserved for the visualization of pulse propagation/reflections along the TL as the
time progresses. The user should start time simulations by pressing the Run button.
The second plot may be recalled by using Plot Sig. vs. Time command button.

The popup menu at the right top contains a key selection; the TL or the TD
Reflectometer. The default selection is the TL where user specifies all input
parameters, and with which he/she does simulations to visualize time domain TL
effects. The TD Reflectometer selection may be used to locate TL termination and
faults. It is designed in a way that the user specifies only the line and source
parameters; the load is selected automatically (also randomly), and a fault is
introduced at an arbitrary point along the TL so the user can find out its type and/or
numerical values (if possible) by only observing/analyzing the output plots, e.g.,
signal vs. time at the selected point on the TL.

The plot in Fig. 5 belongs to a 50, 0.5m loss-free homogeneous (uniform) TL
excited with a Gaussian pulse source having 10-internal resistor, and terminated by
a resistive load of 100 (the TL is fault-free, i.e., uniform). The simulation time step is
400. Since unit length inductance and capacitor values are 250nH/m, and 100pF/m,
respectively, the speed of voltage and current waves along the line is calculated from
the limiting case of (6a) to be v=2 m/s. The TL is divided into 100 nodes therefore
x=0.5/100=5mm. The time step t is calculated to be t=x/v=25ps. With this
choice the voltage (or current) pulse propagates one node at a time therefore it will
12
take 100t for the injected pulse to reach the load. The total of 400t will result in two
reflections from the load and two reflections from the source ends.

A typical signal vs. time plot observed at mid point of the TL with the same
parameters in Fig. 5 is plotted in Fig. 6. As observed, the total simulation time is
400t=1000ps=1ns. The first two pulses are the direct and load-reflected pulses.
The direct and reflected pulses are in phase since the value of the resistive load is
higher than the characteristic impedance of the TL -- remember the voltage reflection
coefficient defined as
0
0
Z Z
Z Z
V
V
L
L
L
+

= =
+

(20)
where
+
V and

V are the incident (traveling) pulse towards and reflected pulse from
the TL, respectively. The third pulse is the one with one-reflection from the load and
one-reflection from the source, and the last one is the pulse with two-reflections from
the load and one-reflection from the source. The TL is lossless in this example
therefore the comparison of the maximum amplitudes of pulse-1 (a) and pulse-2 (b)
yields the modulus of the reflection coefficient a b V V
L
/ / = =
+
, from where load
resistor can be calculated (if not known) with the help of the characteristic
impedance. The length of the TL and the observation point can also be predicted by
measuring the distances (time delays) between first, second and third pulses.

An example of time domain characteristics of the fault-free TL line for the parallel RC
termination (R
L
=10, C
L
=5pF) are given in Fig. 7. The internal resistor of the source
is R
s
=100. In Fig. 7a, a rectangular pulse having 400ps pulse length, traveling
along a 50 TL towards the load is shown. In Fig. 7b, pulse reflected from the load is
given (observe the exponential decay). Finally, Fig. 7c shows signal (voltage pulse)
vs. time at the selected observation point along the TL. Identification of each pulse
and measurement of the time delays among them certainly yields the TL length, the
observation distance from the source or load and the termination type.

Another example of time domain characteristics of the fault-free TL line for the serial
RL termination (R
L
=10, L
L
=100pF) are given in Fig. 8. Again, the internal resistor
of the source is R
s
=100. In Fig. 8a, a rectangular pulse having 400ps pulse length
13
reflected from the load of a 50 TL (at the 133 time step) is shown (observe
characteristic amplitude decrease in the DC portion of the pulse). In Fig. 8b, signal
vs. time is given. The internal resistor of the source is R
s
=100.

Finally, signal vs. time at 0.1m observation point on the SC terminated, 50 TL with
a G-type (G
f
=1 S/m) (i.e., imperfectness of the dielectric layer between the wires of
the TL) fault at 0.4m is given in Fig. 9. The source is a rectangular pulse with 250ps
pulse length. The internal resistor of the source is R
s
=50 (i.e. matched termination
at the source end). In this case, voltage pulse reflects not only at the load, but also
from the fault point in either direction. Pulses marked from 1 to 4 correspond to direct,
fault-reflected, load-reflected and fault-transmitted but load-reflected pulses,
respectively.

IV. The TD Reflectometer: Characteristic Examples
The TDRMeter package can be used to teach, visualize and understand time domain
pulse propagation and reflections from various terminations and faults
(discontinuities). Resistive terminations and faults can be measured/calculated by
analyzing signal vs. time plots under given sets of parameters. Distinguishing the
end- and fault-reflected pulses in time, measuring delays among them and marking
the maximum amplitudes are enough for this purpose.

Reactive components can also be measured using the TDR. As known, the inductor
initially appears as an open circuit to the fast rising edge of the TDR step pulse (i.e.
the high frequencies). Thus it initially gives a reflection coefficient of +1. Later in time,
the inductor appears as a short circuit to the flat top of the TDR step pulse (i.e. the
DC portion). Thus the final TDR value is a reflection coefficient of -1. The capacitor
performs exactly the opposite. Therefore, the L or C value may be determined from
=
0
Z L or
0
/ Z C = by measuring the exponential time constant () of the TDR
response.

Seven cases are included in the TDR tests and once the TDR option is selected from
the pop-up menu at top right of the front panel, the termination and fault blocks
disappear. The package generates a random number between 1 and 7 and the TL
termination and the location of the fault are applied automatically with randomly
14
selected element values. The user may do blind tests and distinguish problems along
the line after specifying the TL and source parameters. He/she can check his results
at the end by re-selecting TR Line option at the same pop-up menu, since termination
and fault data of the last TDR simulation re-appear.

A typical/characteristic example is discussed in this section to illustrate the effectives
and charm of the TDRMeter package. Fig. 10 belongs to this example. A 50, 0.5 m
long TL is excited with a trapezoidal pulse whose duration is 200ps, and rise and fall
times are 100ps. The source internal resistor is 50. Fig. 10a illustrates the front
panel of the TD Reflectometer during the time simulations. The signal vs. time at the
source point is plotted in Fig. 10b, from which one tries to distinguish termination
type, fault type and location. The user recalls termination and fault data by re-
selecting TR line option from the pop-up menu at top right (as explained above). Fig.
10c shows the front panel and all values that belong to Figs. 10a and b. As shown, a
1 k resistor is used ad the load. The fault is 0.35m away from the source and
C
f
=8pF/m.

Another example belongs to a fault and/or termination identification process. It is
well-known that fault and/or termination characteristics can be obtained by the
analysis of recorded time domain pulse characteristics; this is explained partially in
Fig. 6. If direct and reflected pulses can be discriminated in time it is easy to obtain
broad band frequency variation of the reflection coefficients. Various signal vs. time
variations at mid point of a 50 TL under various resistive terminations are given in
Fig. 11a. Here matched termination is used at the front end of the TL (R
s
=50) so
that only load- and/or fault reflected waves exist. As shown in the figure, the ratio of
the amplitudes of reflected and incident pulses yields the modulus of the voltage
reflection coefficient (||). Obviously, this approach does NOT work for complex
loads. In that case, the standard Fourier transform procedure should be applied; (i)
discriminate incident and reflected pulses in time, (ii) move to the frequency domain
by applying fast Fourier transformation (FFT), finally, ratio the reflected pulse to the
incident pulse in the frequency domain within the band of the applied source pulse.
An example of this procedure is given in Fig.11b as voltage reflection coefficient vs.
frequency. The solid and dashed lines represent the results of the procedure
15
explained here and the analytical exact solution obtained from (13), respectively.
Indistinguishable agreement shows the power of the procedure.

The last example of this procedure is given in Fig.12a. Here, fault-free 0.5m-long 50
TL is terminated by a parallel resonance load (Z
L
=50, L
L
=10nH, C
L
=5pF) and
matched termination is used at the source (R
s
=50). The plot show signal vs. time
recorded at mid point of the TL where one can easily discriminate incident and load-
reflected pules. Applying the procedure explained above yields the voltage reflection
coefficient vs. frequency as given in Fig. 12b. The solid and dashed lines in the figure
represent the results of the procedure explained here and the analytical exact
solution obtained from (20), respectively. Indistinguishable agreement shows the
power of the procedure.

It should be noted that, one can still apply this procedure even when the pulses can
not be discriminated in time. In this case the TDRMeter simulations should be
repeated twice; first with along the TL with the unknown termination, then along the
TL with matched termination. The incident plus reflected pulses are recorded in the
first simulation, while only the incident pulse is recorded in the second one.
Subtracting the second from the first will yield the reflected-only pulse, and then the
standard procedure may be applied.

V. Concluding Remarks
A simple TL Matlab package is introduced in this article. The package is design to
visualize voltage pulse propagation and reflections from various discontinuities along
a TL. It may also be used as a TDR and the user may exercise tests of different type
randomly selected terminations and/or faults, and tries to find out what kind of
problem exists by analyzing time history of the simulated voltage pulses. The
package is a good virtual tool that can be used both in EM lectures as well as in
virtual EM labs.

The package can be modified easily to predict TL parameters such as the
characteristic impedance and the complex propagation constant from the analysis of
time-dependent traveling/reflecting pulses by using a known termination, which is left
to the reader.
16


References
[1] L. Sevgi, A Ray Shooting Visualization Matlab Package for 2D Groundwave
Propagation Simulations, IEEE Antennas and Propagation Magazine, Vol. 46, No
4, pp.140-145, Aug 2004
[2] L. B. Felsen, F. Akleman, L. Sevgi, "Wave Propagation Inside a Two-dimensional
Perfectly Conducting Parallel Plate Waveguide: Hybrid Ray-Mode Techniques
and Their Visualisations", IEEE Antennas and Propagation Magazine, Vol. 46,
No.6, pp.69-89, Dec 2004
[3] L. Sevgi, . Uluk, F. Akleman, "A Matlab-based Two-dimensional Parabolic
Equation Radiowave Propagation Package", IEEE Antennas and Propagation
Magazine, Vol. 47, No.4, Aug 2005
[4] L. Sevgi, . Uluk, "A Matlab-based Visualization Package for Planar Arrays of
Isotropic Radiators", IEEE Antennas and Propagation Magazine, Vol. 47, No. 1,
pp. 156-163, Feb 2005
[5] H. Fellner-Feldegg, The measurement of dielectrics in time domain, J. Phys.
Chem. Vol. 73, pp. 613623, 1969
[6] S. Gedney, EE699 FDTD Solution of the TR Line Equations, Lecture Notes, (visit
http://www.engr.uky.edu/~gedney)
[7] K. S. Yee, Numerical solution of initial boundary value problems involving
Maxwells equations, IEEE Transactions on Antennas and Propagat, V-14, no. 3,
pp. 302-307, May 1966
[8] L. SEVGI, Complex Electromagnetic Problems and Numerical Simulation
Approaches, IEEE Press John Wiley and Sons, NY 2003

Figure Captions

Figure 1: (a) finite length transmission line, the time domain voltage source v
s
(t) and
source resistor R
s
, terminated by a complex R
L
load, (b) its loss-free equivalent
circuit. L [H/m] and C [F/m] are unit length inductance and conductance, respectively.

Figure 2: A lossless but symmetrical equivalent circuit in terms of primary line
parameters (V
1
, V
2
, ... V
N
are line voltages).
17

Figure 3: A 1D FDTD leap-frog scheme, voltage and current nodes.

Figure 4a: Source injection at node 1 after Thevenin to Norton transformation.

Figure 4b: Load model at the last node (line is terminated by a parallel RC pair).

Figure 5: The front panel of the TDRMeter package, Gaussian pulse traveling
towards the load (a 50 transmission line, R
s
=10, resistive load=500).

Figure 6: Signal vs. time at mid point along the line (line length=0.5m, Gaussian
pulse duration=250ps, a 50 transmission line, R
s
=10, resistive load=500).

Figure 7: The front panel of the TDRMeter package, parallel RC termination, (a)
rectangular pulse traveling towards the load (b) pulse reflected from the load (a 50
transmission line, R
s
=100, pulse width=400ps, R
L
=10, C
L
=5pF), (c) Signal vs.
time on the TL at 0.1m away from the source.

Figure 8: The front panel of the TDRMeter package, serial RL termination, (a)
rectangular pulse reflected from the load (b) Signal vs. time on the TL at 0.4m away
from the source (a 50, 0.5 m transmission line, rectangular pulse duration=400ps,
R
s
=100, R
L
=10, L
L
=100nH).

Figure 9: Signal vs. time at 0.1m observation point on the SC terminated, 50 TL
with a G-type fault at 0.4m (G
f
=1 S/m), rectangular pulse source with 250ps pulse
length (R
s
=50).

Figure 10a: The front panel of the TDRMeter, trapezoidal pulse reflected from the
end of the a 50, 0.5m TL and secondary reflections from the fault (pulse
length=200ps, rise time=fall time=100ps, R
s
=50)

Figure 10b: The TD Reflectometer; signal vs. time at the source point of a 50, 0.5m
TL (trapezoidal pulse duration=200ps, rise and fall times are 100ps, source internal
resistor R
s
=50.
18

Figure 10c: The front panel of the TDRMeter package with values of Figs. 10a and
b. The user re-selects TR Line option to make termination and fault values re-appear.

Figure 11a: Signal vs. time at mid point of a TL under various resistive termination
showing the incident and reflected pulses. The ratio of the amplitudes of reflected
and incident pulses yields the modulus of the voltage reflection coefficient (||).

Figure 11b: Voltage reflection coefficient vs. frequency obtained by the application of
off-line FFT as described in Sec. 4. Solid: TDRMeter simulation result, Dashed:
Analytical exact solution obtained from (13).

Figure 12a: Signal vs. time at mid point of a 50, 0.5m TL under parallel resonance
termination (Z
L
=50, L
L
=10nH, C
L
=5pF). Matched termination is used at the source
(R
s
=50). Pulse length=400ps).

Figure 12b: Voltage reflection coefficient vs. frequency obtained by the application of
off-line FFT as described in the text. Solid: TD simulation result, Dashed: Analytical
exact solution obtained from (20).



Table 1: Primary parameters of a transmission line

Parameter Symbol Unit
Unit length resistance R /m
Unit length conductivity G S/m
Unit length inductance L H/m
Unit length capacitance C F/m


Table 2: Analogy between uniform TL and uniform plane waves

TL Parameters EM Parameters
V [Volt] E [Volt/m]
I [Amp] H [Amp/m]
C [Farad/m] [Farad/m]
L [Henry/m] [Henry/m]
G [ Siemens/m] [ Siemens/m]
R [Ohm/m]
1
Figure 1:
V
s
R
s
R
L
Cx
V
s
R
s
R
L
Lx
V
1
V
2
V
3
V
N
I
1
I
2
(a)
(b)
Figure 2:
R
s
V
s
V
1 V
2
V
3
I
1
I
2
2
x C
Lx Lx
2
x C
2
x C
2
Figure 3:
k=m-1
k=m
k=m+1
voltages
k=m-1
k=m
k=m+1
currents
x
x
Figure 4a:
R
s
V
1 V
2
I
1
2
x C
Lz
2
x C
s
s
R
V
3
Figure 4b:
R
L
V
N-1
V
N
I
N-1
Lx
2
x C
2
x G
2
x C
2
x G
C
L
Figure 5:
4
Figure 6:
1
2
3
4
b
a
tx
Figure 7a:
5
Figure 7b:
Figure 7c:
6
Figure 8a:
Figure 8b:
7
Figure 9:
1
2
3
4
Figure 10a:
8
Figure 10b:
Figure 10c:
9
Figure 11a:
=
=
M Z
Z
L
1
50
0
=
=
0
50
0
L
Z
Z
=
=
350
50
0
L
Z
Z
=
=
30
50
0
L
Z
Z
1 =
4 / 3 =
1 =
4 / 1 =
Figure 11b:
pF C
R
L
L
10
50
=
=
= 50
0
Z
10
Figure 12a:
nH L
pF C
R
L
L
L
10
5
50
=
=
=
= 50
0
Z
Figure 12b: