APPLICATION OF THE LAPLACE TRANSFORM TO CIRCUIT ANALYSIS

LEARNING GOALS
Laplace circuit solutions Showing the usefulness of the Laplace transform Circuit Element Models Transforming circuits into the Laplace domain Analysis Techniques All standard analysis techniques, KVL, KCL, node, loop analysis, Thevenins theorem are applicable Transfer Function The concept is revisited and given a formal meaning Pole-Zero Plots/Bode Plots Establishing the connection between them Steady State Response AC analysis revisited

LAPLACE CIRCUIT SOLUTIONS

We compare a conventional approach to solve differential equations with a technique using the Laplace transform

di (t ) dt di VS ( s ) = RI ( s ) + LL i = iC + i p Initial conditions dt di KVL : v S (t ) = Ri ( t ) + L (t ) are automatically di P dt = sI ( s ) i (0) = sI ( s ) included L a dt Complementary equation r 1 1 I ( s) = diC t = RI ( s ) + LsI ( s ) (t ) = 0 iC (t ) = K C e t i RiC (t ) + L s ( R + Ls ) s dt 1/ L K K2 Only algebra c I ( s) = R = 1+ t t RK C e + LK C (e ) = 0 = u s ( R / L + s ) s s + R / L is needed L l 1 Particular solution for this case a K1 = sI ( s ) |s =0 = No need to R v S = 1 = RK p i p (t ) = K p r 1 search for R Use boundary conditions particular K 2 = ( s + R / L) I ( s ) | s = R / L = t 1 R or complei (t ) = + K C e L v (t) = 0 for t < 0 i( 0 ) = 0 S R R mentary t 1 R L ; t > 0 t i ( t ) = 1 e 1 solutions 1 e L ; t > 0 i (t ) = 2 R R v S (t ) = Ri (t ) + L

Comple mentary

Take Laplace transform of the equation

LEARNING BY DOING

Find v (t ), t > 0

v vS R

In the Laplace domain the differential equation is now an algebraic equation

C
vS

dv dt

1 s 1 1 / RC = V ( s) = s ( RCs + 1) s ( s + 1 / RC ) RCsV ( s ) + V ( s ) =
Use partial fractions to determine inverse

Model using KCL

dv v v S + =0 R dt

V ( s) =

1 / RC K K2 = 1+ s ( s + 1 / RC ) s s + 1 / RC

RC

dv + v = vS dt

K1 = sV ( s ) |s =0 = 1

dv RCL + V ( s ) = VS ( s ) dt

K 2 = ( s + 1 / RC )V ( s ) |s = 1 / RC = 1

dv L = sV ( s ) v (0) = sV ( s ) dt v S (t ) = 0, t < 0 v (0) = 0 Initial condition v S = u( t ) VS ( s ) = 1 s

v (t ) = 1 e

t RC ,

t0

given in implicit form

3

CIRCUIT ELEMENT MODELS

The method used so far follows the steps: 1. Write the differential equation model 2. Use Laplace transform to convert the model to an algebraic form For a more efficient approach: 1. Develop s-domain models for circuit elements 2. Draw the Laplace equivalent circuit keeping the interconnections and replacing the elements by their s-domain models 3. Analyze the Laplace equivalent circuit. All usual circuit tools are applicable and all equations are algebraic.

Independent sources v S ( t ) VS ( s )

Resistor

i S (t ) I S ( s)
Dependent sources v D (t ) = AiC (t ) VD ( s ) = AI C ( s )

i D (t ) = BvC (t ) I D ( s ) = BVC ( s ) ...

v (t ) = Ri (t )

V ( s ) = RI ( s )

Capacitor: Model 1

Source transformation

I eq

v (0) = s = Cv (0) 1 Cs

I ( s) t L i ( x )dx = s 0

1t v (0) 1 v (t ) = i ( x )dx + v (0) V ( s ) = I ( s) + Cs C0 Cs s

Capacitor: Model 2 Impedance in series with voltage source

1t v (t ) = i ( x )dx + v (0) C0

I ( s ) = CsV ( s ) Cv (0)
Impedance in parallel with current source
5

Inductor Models

di L = sI ( s ) i (0) dt

v (t ) = L

di (t ) dt

V ( s ) = L( sI ( s ) i (0)) I ( s) = V ( s ) i (0) + Ls s

V ( s ) = LsI ( s ) Li (0)

I ( s) =

V ( s ) i (0) + Ls s
6

Mutual Inductance

v1 (t ) = L1

di1 di (t ) + M 2 (t ) dt dt di di v2 (t ) = M 1 (t ) + L2 2 (t ) dt dt

V1 ( s ) = L1sI1 ( s ) L1i1 (0) + MsI 2 ( s ) Mi2 (0) V2 ( s ) = MsI1 ( s ) Mi1 (0) + LsI 2 ( s ) Li2 (0)

Combine into a single source in the primary Single source in the secondary

LEARNING BY DOING

Determine the model in the s-domain and the expression for the voltage across the inductor

Steady state for t<0

i ( 0) = 1 A

Inductor with initial current

KVL : 1 = (1 + s ) I ( s )

Ohm' s Law

V ( s ) = 1 I ( s )
Equivalent circuit in s-domain

V ( s) =

1 s +1

ANALYSIS TECHNIQUES
All the analysis techniques are applicable in the s-domain LEARNING EXAMPLE Draw the s-domain equivalent and find the voltage in both s-domain and time domain One needs to determine the initial voltage across the capacitor

I S ( s) =

3 s +1

i S (t ) = 0, t < 0

v o (0) = 0

RC = (10 103 )(25 106 ) = 0.25 120 K K = 1 + 2 ( s + 4)( s + 1) s + 4 s + 1 K1 = ( s + 4)Vo ( s ) |s = 4 = 40 K 2 = ( s + 1)Vo ( s ) |s =1 = 40

vo (t ) = 40 e t e 4 t u(t )
9

Vo ( s ) = 1 Vo ( s ) = R || I S ( s) Cs R 3 Cs I ( s ) = 1 / C 3 10 Vo ( s ) = 1 S s + 1 / RC s +1 R+ Cs

LEARNING EXAMPLE

Write the loop equations in the s-domain Do not increase number of loops

Loop 1 VA ( s) R1I1 ( s ) + v1 (0) v2 (0) + L1i1 (0) = s s

1 1 I1 ( s ) + ( I1 ( s ) I 2 ( s )) + L1s ( I1 ( s ) I 2 ( s )) C1s C2 s Loop 2 L1i1 (0) v 2 ( 0) L2 i2 (0) + VB ( s ) = s

L1s ( I 2 ( s ) I1 ( s )) +

1 ( I 2 ( s ) I1 ( s )) + ( L2 s + R2 ) I 2 ( s ) 10 C2 s

LEARNING EXAMPLE

Write the node equations in the s-domain

Node V1 I A ( s)

i (0) i1 (0) C1v1 (0) + 2 = s s 1 1 1 G1 + + + C1s V1 ( s ) + C1s V2 ( s ) L s L1s L2 s 2 Node V2 I B ( s ) C 2v2 (0) + C1v1 (0)

Do not increase number of nodes

i2 (0) = s 1 1 G2 + C 2 s + C1s + V2 ( s ) C1s + V1 ( s ) L2 s L2 s

11

LEARNING EXAMPLE Find vo (t ) using node analysis, loop analysis, superposit ion,

source transforma tion, Thevenin' s and Norton' s theorem.

Assume all initial conditions are zero

V1 ( s ) KCL @ V1 1 2+ 12 s V1 ( s ) 4 s + V1 ( s ) Vo ( s ) = 0 s + 1 s s s KCL@Vo Could have

Node Analysis

Vo ( s ) Vo ( s ) V1 ( s ) + = 02 1 2 s (1 + s 2 )V1 ( s ) s 2Vo ( s ) =

used voltage divider here

4 s + 12 2s s (1 + s 2 ) 2 sV1 ( s ) + (1 + 2 s )Vo ( s ) = 0 Vo ( s ) = 8( s + 3) (1 + s ) 2

12

Loop Analysis

Loop 1 I1 ( s ) = Loop 2 1 12 s ( I 2 ( s ) I1 ( s )) + I 2 ( s ) + 2 I 2 ( s ) = s s 4( s + 3) I 2 ( s) = ( s + 1) 2 Vo ( s ) = 2 I 2 ( s ) = 8( s + 3) ( s + 1) 2 4 s

13

Source Superposition

Applying current source

' I2

Current divider

Vo' ( s ) = 2

s 4 1 2+ +2 s s

Applying voltage source

Voltage divider

8( s + 3) Vo ( s ) = V o ( s ) + V o ( s ) = ( s + 1) 2
' "

Vo" ( s ) =

12 2 1 2+ + s s s

14

Source Transformation

Combine the sources and use current divider

Vo ( s ) = 2

s 4 12 + 2 1 s + + 2 s s s

Vo ( s ) =

8( s + 3) ( s + 1) 2

The resistance is redundant

15

Using Thevenins Theorem Reduce this part

ZTh

s2 + 1 1 = +s= s s

Voltage divider

VOC ( s ) =

12 4 4 s + 12 +s = s s s

Vo ( s ) =

2 4 s + 12 s2 + 1 s 2+ s 8( s + 3) ( s + 1) 2
16

Only independent sources

Vo ( s ) =

Using Nortons Theorem Reduce this part

ZTh = s
Current division

Vo ( s ) = 2 4 12 / s 4 s + 12 + = s s s2

I SC ( s ) =

4 s + 12 s 1 s2 s+ +2 s 8( s + 3) Vo ( s ) = ( s + 1) 2
17

LEARNING EXAMPLE Determine the voltage vo (t ). Assume all initial conditions to be zero Selecting the analysis technique: . Three loops, three non-reference nodes . One voltage source between non-reference nodes - supernode . One current source. One loop current known or supermesh . If v_2 is known, v_o can be obtained with a voltage divider

Transforming the circuit to s-domain

Doing the algebra : V1 ( s ) = V2 ( s ) 12 / s I ( s ) = V2 ( s ) / 2 + 6 / s

(1 / 2)( s + 1)(V2 ( s ) 12 / s ) 2(V2 ( s ) / 2 + 6 / s ) + V2 ( s ) /( s + 1) = 0 12 s V ( s) V ( s) V ( s) KCL@ supernode : 1 + 1 2 I ( s ) + 2 =0 2 2/ s s +1 V ( s) Controlling variable : I ( s ) = 1 2 1 Voltage divider : V0 ( s ) = V2 ( s ) s +1 Supernode constraint : V2 ( s ) V1 ( s ) = V2 ( s ) = 12( s + 1)( s + 3) s ( s 2 + 4 s + 5) 12( s + 3) s ( s 2 + 4 s + 5)
18

Vo ( s ) =

Continued ...

Compute Vo (s ) using Thevenin' s theorem

-keep dependent source and controlling variable in the same sub-circuit -Make sub-circuit to be reduced as simple as possible -Try to leave a simple voltage divider after reduction to Thevenin equivalent

VOC 12 / s

12 / s +

VOC 12 / s VOC 12 / s + 2I '= 0 2 2/ s 12 VOC 12 / s VOC ( s ) = I '= s 2 I '+ (2 I ' ) /( 2 / s ) 2 I ' = 0 I '= 0

I "= 6 / s I SC 2 I " I "2 I " /( 2 / s ) = 0 6( s + 3) Z = VOC ( s ) = 2 TH I SC = I SC ( s ) s + 3 s Vo ( s ) = 1 1+ s + 2 s+3 12 s

19

Continued

Computing the inverse Laplace transform

Analysis in the s-domain has established that the Laplace transform of the output voltage is

12( s + 3) s 2 + 4 s + 5 = ( s + 2 j1)( s + 2 + j1) = ( s + 2) 2 + 1 Vo ( s ) = s ( s 2 + 4 s + 5) * 12( s + 3) Ko K1 K1 = + + Vo ( s ) = s ( s + 2 j1)( s + 2 + j1) s ( s + 2 j1) ( s + 2 + j1) K o = sVo ( s ) |s =0 = 36 5

* K1 K1 + 2 | K1 | e t cos( t + K1 ) u(t ) ( s + j ) ( s + + j )

12(1 + j1) 12 245 One can also use = K = ( s + 2 j1)V ( s ) | = 1 o s = 2 + j1 (2 + j1)( j 2) 5153.43(290) quadratic factors... = 3.79 198.43 = 3.79161.57 36 C o C1 ( s + 2) C2 12( s + 3) vo (t ) = + 7.59e 2 t cos(t + 161.57 u(t ) Vo ( s ) = + = + 5 (s + 2)2 + 1 s ( s + 2)2 + 1 ( s + 2)2 + 1 s

C1 ( s + ) C2 + e t [C1 cos t + C 2 sin t ]u(t ) 2 2 2 2 (s + ) + (s + ) + 12( s + 3) = C o (( s + 2) 2 + 1) + s[C1 ( s + 2) + C 2 ] s = 2 12 = C o 2C 2 C 2 = 36 / 10 6 = 12 / 5

C o = sVo ( s ) |s =0 = 36 / 5 C1 = 36 / 5
20

Equating coefficien ts of s 2 : 0 = C o + C1

12 36 vo (t ) = (1 e 2 t cos t ) e 2 t sin t u(t ) 5 5

LEARNING EXTENSION

Find io (t ) using node equations

supernode Assume zero initial conditions Implicit circuit transformation to s-domain

Vo + VS

VS

Vo

* K1 K1 + 2 | K1 | e t cos( t + K1 ) u(t ) ( s + j ) ( s + + j )

KCL at supernode

1 15 I o ( s) | 1 K1 = s + j s= + j 4 4 4

1 15 1 6 + j 4 4 = 15 15 4 2j 4

6.33 66.72 2 V ( s) V ( s) K1 = = 6.53 156.72 Cs (Vo ( s ) + VS ( s )) + o + o =0 0.9790 s Ls 2 12 V ( s) VS ( s ) = , I o ( s ) = o t 15 2 s io (t ) = 13.06e 4 cos t 156.72 4 Doing the algebra 1 6s 1 6s = I o ( s) = 2 2 s + 0.5 s + 1 1 15 + s + 4 16 * K1 1 6s K1 I o ( s) = = + 1 15 1 15 1 15 1 15 s + j s + + j s + j s + + j 4 4 4 4 4 4 4 4 21

LEARNING EXTENSION

Find vo (t ) using loop equations

I 2 ( s) =

K0 K1 K2 + + s s + 0.27 s + 3.73

supermesh

K 0 = sI 2 ( s ) |s =0 = 2

constraint due to source 2 = I 2 I1 s KVL on supermesh 1 12 I1 + 2 I1 + sI 2 + 2 I 2 = 0 s s

Solve for I2

K1 = ( s + 0.27) I 2 ( s ) |s =0.27 = K 2 = ( s + 3.73) I 2 ( s ) |s = 3.73 =

16(0.27) + 2 = 2.48 (0.27)(0.27 + 3.73)

16(3.73) + 2 = 4.47 (3.73)(3.73 + 0.27)

i2 (t ) = 2 + 2.48e 0.27 t 4.47e 3.73t u( t ) vo ( t ) = 2i2 ( t )

I 2 ( s) =

16 s + 2 16 s + 2 = s ( s 2 + 4 s + 1) s ( s + 0.27)( s + 3.73)

Determine inverse transform

22

TRANSIENT CIRCUIT ANALYSIS USING LAPLACE TRANSFORM For the study of transients, especially transients due to switching, it is important to determine initial conditions. For this determination, one relies on the properties: 1. Voltage across capacitors cannot change discontinuously 2. Current through inductors cannot change discontinuously LEARNING EXAMPLE Determine vo ( t ), t > 0

i L( 0 )

+ vC ( 0 )

Assume steady state for t<0 and determine voltage across capacitors and currents through inductors

For DC case capacitors are open circuit inductors are shortcircu it vC (0) = 1V , i L (0) = 1A
23

Circuit for t>0

Vo ( s ) =

2s + 7 2 s 2 + 3s + 2

Now determine the inverse transform

b 2 4ac < 0
Laplace Circuit for t>0

complex conjugate roots

* K1 K1 Vo ( s ) = + 3 7 3 7 s+ j s+ + j 4 4 4 4

3 7 = 2.14 76.5 K1 = s + j Vo ( s ) 4 4 3 7 s = + j
4 4 * K1 K1 + 2 | K1 | e t cos( t + K1 ) u(t ) ( s + j ) ( s + + j )

Use mesh analysis

4 ( s + 1) I1 sI 2 = + 1 s s

Solve for I2

vo (t ) = 4.28 cos(

2 1 sI1 + ( s + 1 + ) I 2 = 1 ( s + 1) s s I 2 ( s) = 2s 1 2 1 Vo ( s ) = I 2 ( s ) + s s 2 s 2 + 3s + 2

7 t 76.5) 4

24

LEARNING EXTENSION

Determine i1 (t ), t > 0

Initial current through inductor

i L (0 ) = i L ( 0 + ) = 1 A

I1 ( s ) =

s i1 (t ) = e 9 t u( t ) s+9

12 6
I1 ( s )

2s

1 s
Current divider
25

I1 ( s ) =

2s 1 2 s + 18 s

LEARNING EXTENSION

Determine vo (t ), t > 0

2s
+

8 V 3
+ Vo (s )

Determine initial current through inductor

Vo ( s ) =

i L ( 0)

2 12 8 + (voltage divider) 4 + 2s s 3

Use source superposition

Vo ( s ) =

i12V = 2 A

(8 s + 36) K1 K 2 + = s s+2 3s ( s + 2) 10 3
26

2 i4V = A 3 i L (0) = 4 A 3

K1 = sVo ( s ) |s =0 = 6

K 2 = ( s + 2)Vo ( s ) |s =2 = 8 v o ( t ) = 6 e 2 t u( t ) 3

TRANSFER FUNCTION
X (s )
System with all initial conditions set to zero

Y (s )

bn s n + ... + b1s + b0 H ( s) = am s m + ... + a1s + a0

For the impulse function

Y ( s) x (t ) = (t ) X ( s ) = 1 X ( s) H(s) can also be interpreted as the Laplace If the model for the system is a differential transform of the output when the input is an impulse and all initial conditions are zero equation H ( s) =
dny d n1 y dy bn n + bn1 n1 + ... + b1 + bo y dt dt dt dmx d m 1 x dx = am m + am 1 m 1 + ... + a1 + ao x dt dt dt If all initial conditions are zero
d k y k L k = s Y (s ) dt bn s nY ( s ) + ... + b1sY ( s ) + b0Y ( s )
= am s m X ( s ) + ... + a1sX ( s ) + a0 X ( s ) bn s n + ... + b1s + b0 Y ( s) = X ( s) m am s + ... + a1s + a0
27

The inverse transform of H(s) is also called the impulse response of the system If the impulse response is known then one can determine the response of the system to ANY other input

LEARNING EXAMPLE A network has impulse response h( t ) = e u(t)

Determine the response, vo (t ), for the input v i ( t ) = 10e 2 t u(t )

In the Laplace domain, Y(s)=H(s)X(s)

Vo ( s ) = H ( s )Vi ( s )

h(t ) = e t u(t ) vi ( t ) = 10e 2 t u(t )

H ( s) =

1 s +1 10 Vi ( s ) = s+2

Vo ( s ) =

K K 10 = 1 + 2 ( s + 1)( s + 2) s +1 s + 2

K1 = ( s + 1)Vo ( s ) |s = 1 = 10 K 2 = ( s + 2)Vo ( s ) |s = 2 = 10 vo (t ) = 10 e t e 2 t u(t )

28

Impulse response of first and second order systems

K First order system H ( s ) = s + 1

Normalized second order system
2 0 H ( s) = 2 2 s + 2 0 s + 0

h(t ) = Ke

Case 2 : < 1 : Underdampe d network poles : s1, 2 = 0 j0 1 2

h(t ) = Ke o t cos( o 1 2 t + )

poles : s1, 2 = 0 0 2 1

Case 1 : > 1 : Overdamped network h ( t ) = K 1e ( 0 + 0

2 1 ) t

+ K 2 e ( 0 0

2 1 ) t

Case 3 : = 1 : Critically damped network h(t ) = K1te o t + K 2e o t

29

LEARNING EXAMPLE

Determine the transfer function H ( s ) =

a) C = 8F

Vo ( s ) Vi ( s ) poles : s1, 2 = 0.25 j 0.25

Transform the circuit to the Laplace domain. All initial conditions set to zero

b) C = 16F c) C = 32F

poles : s1, 2 = 0.25

Vi (s )
Mesh analysis

poles : s1, 2 = 0.427, 0.073

Vi ( s ) = 2 I1 I 2

1 0 = I1 + 1 + s + I2 sC (1 / 2C ) Vo ( s ) = 2 s + (1 / 2) s + 1 / C

Vo ( s ) =

1 I 2 ( s) sC

30

LEARNING EXAMPLE

Determine the transfer function, the type of damping and the unit step response

V1 ( s ) VS ( s ) V1 ( s ) V1 ( s ) V1 ( s ) V0 ( s ) + + + =0 1 1 1 1 s 1 Vo ( s ) 32 = VS ( s ) s 2 + 1 s + 1 2 o 0 = 0.25 2 16
Transform the circuit to the Laplace domain. All initial conditions set to zero

2 o

=1
VS ( s ) = 1 s

Unit step response

Vo ( s ) = (1 / 32)
2

1 s s + 4 K o = sVo ( s ) |s =0 = 0.5 d s 2Vo ( s ) K11 = ds

Ko K11 K12 + + s s + 0.25 ( s + 0.25) 2

V+ = 0 V1 ( s ) Vo ( s ) + =0 1 V1 ( s ) = sVo ( s ) 1 s

K12 = ( s + 0.25) 2Vo ( s ) |s = 0.25 = 0.125

= 0.5
s = 0.25

vo (t ) = 0.5 (0.125t + 0.5)e 0.25 t u(t )

31

LEARNING EXTENSION

H ( s) =

s + 10 s2 + 4s + 8

Determine the pole-zero plot, the type of damping and the unit step response

zero : z = -10 poles :

1 s + 10 Y ( s) = H ( s) = s s ( s 2 + 4 s + 8) s 2 + 4 s + 8 = ( s + 2 j 2)( s + 2 + j 2)
* K1 K2 K2 Y ( s) = + + s s + 2 j2 s + 2 + j2 * K1 K1 + 2 | K1 | e t cos( t + K1 ) u(t ) ( s + j ) ( s + + j ) 10 K1 = sY ( s ) |s =0 = 8 8 + j2 K 2 = ( s + 2 j 2)Vo ( s ) |s = 2+ j 2 = (2 + j 2)( j 4)

s2 + 4s + 8 = 0

s1, 2 = 2 j 2
x

j j2

10

s + 4s + 8
2

2 o

2 o

2 2

K2 =

8.2514 = 0.73 211 2.83135 490

10 vo (t ) = + 1.46 cos(2t 211) 8

32

Second order networks: variation of poles with damping ratio Normalized second order system
2 0 H ( s) = 2 2 s + 2 0 s + 0

LEARNING EXAMPLE
2 o =

1 , LC

2 o =

R L 1 Cs

poles : s1, 2 = 0 0 2 1 Case 2 : < 1 : Underdampe d network poles : s1, 2 = 0 j 0 1 2

1 V ( s) LC = = Gv ( s ) = o 1 Vin ( s ) 1 + Ls + R R s2 + s + Cs LC L

= cos

Variation of poles. Use o = 2000

33

LEARNING EXAMPLE The Tacoma Narrows Bridge Revisited Previously the event was modeled as a resonance problem. More detailed studies show that a model with a wind-dependent damping ratio provides a better explanation

d 2 d 2 + 2 o + o = 0 dt dt 2

= 0.0046 0.00013U U = wind speed (mph)

Torsional Resonance Model

Conditions at failure wind speed twist

Problem: Develop a circuit that models this event

= 42mph = 12
34

time to collapse = 45min

model d 2
.. .

d 2 + 2 o + o = 0 dt dt 2
=0
2 = (2 o + o )

integrator

adder

2 + 2 o + o

..

= (0.001156 00013U ) 1.579

Using numerical values

vi

v1
Simulation building blocks

Vi ( s ) V ( s ) + =0 1 R Cs 1 V ( s ) = Vi ( s ) R C s

( )

v2

V1 V2 V + + =0 R1 R2 R f Rf Rf V = R V1 + R V2 1 2

d 2 2 dt

d dt

Simulation using dependent sources

35

Simulation results Wind speed=20mph initial torsion=1 degree

Wind speed=35mph initial torsion =1degree

36

POLE-ZERO PLOT/BODE PLOT CONNECTION Bode plots display magnitude and phase information of They show a cross section of G(s)

G ( s ) | s = j

s2 G ( s) = 2 s + 2s + 5 1 V ( s) LC = G ( s) = o 1 R Vin ( s ) s2 + s + LC L

If the poles get closer to imaginary axis the peaks and valleys are more pronounced

Cross section shown by Bode

37

Cross section

Due to symmetry show only positive frequencies

Front view Amplitude Bode plot

Uses log scales

38

STEADY STATE RESPONSE

Y ( s) = H ( s) X ( s)
Response when all initial conditions are zero

Laplace uses positive time functions. Even for sinusoids the response contains transitory terms EXAMPLE H ( s ) = 1 , X ( s ) =

s +1

s ( s2 + 2

x (t ) = [cos t ]u(t ))
If interested in the steady state response only, then dont determine residues associated with transient terms

* s K1 K2 K2 Y(s) = = + + (s +1)(s + j)(s j) s +1 s + j s j

y (t ) = Ke t + 2 | K 2 | cos( t + K 2 ) u( t )
transient Steady state response

If x (t ) = X M cos( o t + )u(t )
y ss (t ) =| X M | H ( j o ) | cos( o t + H ( j o ) + )

For the general case

X M cos t u(t ) =

1 XM X M jt XM e + e j t X ( s ) = + 2 2 s j o s + j o

K* 1 X M X M Kx x Y ( s) = H ( s) s j + s + j = s j + s + j + transient terms o o o o 2 1 y (t ) = 2 | K x | cos( o t + K 2 ) + transient terms K x = ( s j o )Y ( s ) |s = j o = X M H ( j o ) 2 39 y ss (t ) = X M | H ( j o ) | cos( o t + H ( j o ))

LEARNING EXAMPLE

Determine the steady state response

If x (t ) = X M cos( o t + )u(t )
y ss (t ) =| X M | H ( j o ) | cos( o t + H ( j o ) + )

o = 2, X M = 10
Transform the circuit to the Laplace domain. Assume all initial conditions are zero

s2 Vi ( s ) Vo ( s ) = 2 3s + 4 s + 4

s2 H ( s) = 2 3s + 4 s + 4

( j 2) 2 H ( j 2) = = 0.35445 3( j 2) 2 + 4( j 2) + 4 y s (t ) = 3.54 cos(2t + 45)V

KCL@V1 :

V1 Vi V1 V1 + + =0 2 2 2 +1 s 1 Voltage divider : Vo = V1 2 +1 s

40

LEARNING EXTENSION

Determine voss ( t ), t > 0

If x (t ) = X M cos( o t + )u(t )
y ss (t ) =| X M | H ( j o ) | cos( o t + H ( j o ) + ) Vo ( s ) = 2 VOC ( s ) 2 + ZTh ( s ) 2 1 Vi ( s ) s2 + s + 1 s + 1 2+ s +1

o = 2, X M = 12
Transform circuit to Laplace domain. Assume all initial conditions are zero Thevenin

Vo ( s ) =

Vi (s )

1 s

2 Vi ( s ) 2 s + 3s + 3 H (s ) 2 2 2 H ( j 2) = = = 4 + 6 j + 3 1 + 6 j 6.0899.46 Vo ( s ) = voss (t ) = 12 2 cos(2t 99.46) 6.08

1 1 VOC ( s ) = s Vi ( s ) = Vi ( s ) 1 s +1 1+ s 1 1 s2 + s + 1 ZTh ( s ) = s + || 1, ||= s + = s s +1 s +1

41

LEARNING BY APPLICATION

De-emphasize bass Pole-zero map for GvR (s )

Enhances bass to original level

The playback filter is the reciprocal

RIAA recording filter K (1 + s z1 )(1 + s z 2 ) (1 + s p ) z1 = 75s | KG ( s ) |s = j 2 1000 = 1 z 2 = 3180 s GvR ( s ) =

Gvp ( s ) =

1 Grp ( s )

Pole/zero of playback filter cancels pole/zero of recording filter

p = 318s
zeros : z1 = 13.33kr / s [2.12kHz ],

z1 = 313.46r / s[50 Hz ] pole : p = 3,1346r / s[500 Hz ]

42

LEARNING BY DESIGN

Filtering noise in a data transmission line Noise source is 100kHz SOLUTION: Insert a second order low-pass filter in the path. Should not affect data signal and should attenuate noise

Data bits at 1000bps

Proposed filter

V1 V = 0

R 1 3 R R R C C Vo ( s ) 1 2 3 1 2 = 1 Vdata ( s ) 1 1 1 s 2 + s + + + RC R C R C R2 R3C1C 2 1 1 2 1 3 1
2 s 2 + 2 o s + o

R1 = R2 = R3

o =

V1 Vdata V1 V V V + + 1 + 1 o =0 R3 R1 (1 / C1s ) R2 V1 Vo + =0 R2 (1 / C 2 s )

1 3 C2 , = R C1C 2 2 C1

Design equations Well below 100k above 1k Filter design criteria

>1

Below 100k
43

Selected pole location or filter

= 2 105
Select R = 40k, 0 = 25,000, = 2. Use design equations and determine C1 = 0.75nF , C 2 = 1.33nF

The filter eliminates noise but smooths data pulse 25kbps data transfer rate Filtered signal is useless REDESIGN!
44

noise

Circuit simulating the filter

New pole-zero selection

=1

Design equations

0 = 150,000 = =1=
3 C2 2 C1

1 40,000 C1C 2

Simulation for 25kbps

APPLICATION LAPLACE
45