Network Analysis and Synthesis

Chapter 2
Network transform representation
and analysis
 2.1 The transformed circuit
• When analyzing a network in time domain we
will be dealing with
– Derivation and
– Integration
• However, when transformed to complex
frequency domain these become
– Derivation -> multiplication by ‘s’
– Integration -> division by ‘s’
• Hence, it is easier to do network analysis in
complex frequency domain.
By Mr. Yosef B. (MSc.) 2
• The voltage current relationships of network
elements in time domain and complex
frequency domain are given as:

• Resistor

v(t ) Ri (t ) V ( s ) RI ( s )

By Mr. Yosef B. (MSc.) 3

• Inductor
– The time domain relation ships are
di (t )
v(t ) L
dt
t
1
i (t )  v( )d  i (0  )
L 0
– In frequency domain they become
V ( s ) sLI ( s )  Li (0  )
V ( s ) i (0  )
I ( s)  
sL s
By Mr. Yosef B. (MSc.) 4
• An inductor is represented in frequency

– An impedance sL in series with a voltage source

domain as

Used in mesh analysis.

– An admittance 1/sL in parallel with a current

or

source
Used in nodal analysis.

By Mr. Yosef B. (MSc.) 5

By Mr. Yosef B. (MSc.) 6
• Capacitor
– The time domain relation ships are
t
1
v(t )  i ( )d  v(0  )
C 0
dv(t )
i (t ) C
dt
– In frequency domain they become
I ( s ) v (0  )
V ( s)  
sC s
I ( s ) sCV ( s )  Cv(0  )
By Mr. Yosef B. (MSc.) 7
• A capacitor is represented in frequency

– An impedance 1/sC
domain as
in series with a voltage
source
Used in mesh analysis.

– An admittance sC in parallel with a current source

or

Used in nodal analysis.

By Mr. Yosef B. (MSc.) 8

By Mr. Yosef B. (MSc.) 9
 Example 1
• In the figure below, the switch is switched from
position 1 to 2 at t=0. Draw its transformed circuit and
write the transformed equations using mesh analysis.

By Mr. Yosef B. (MSc.) 10

• The transformed circuit is

By Mr. Yosef B. (MSc.) 11

• The transformed equations become

By Mr. Yosef B. (MSc.) 12

 Example 2

i(t). i (0 ) 2amp
• The switch is thrown to position 2 at t=0. Find

L

vC (0  ) 2V

By Mr. Yosef B. (MSc.) 13

• The transformed circuit is

By Mr. Yosef B. (MSc.) 14

• Writing the transformed equation
5 2  2
 2   3  s   I ( s )

• Solving for I(s)

s s  s

2s  3
I ( s) 
( s  2)( s  1)
1 1
I ( s)  
s  2 s 1
• Inverse transforming
 2t t
i (t ) e e
By Mr. Yosef B. (MSc.) 15
 Example 3
• At t=0, the switch is opened. Find the node
voltages v1 and v2
1
L h C 1f
2
G 1 mho V 1v

By Mr. Yosef B. (MSc.) 16

• The transformed circuit becomes

By Mr. Yosef B. (MSc.) 17

• The transformed equations become

• Solving these 2 equations

By Mr. Yosef B. (MSc.) 18

 2.2 System function
• The excitation , e(t), and response, r(t), of a linear
system are related by a linear differential equation.
• When transformed to complex frequency domain the
relationship between excitation and response is
algebraic one.

response are related by the system function H(s)

• When the system is initially inert, the excitation and

given by

R ( s ) H ( s ) E ( s )

By Mr. Yosef B. (MSc.) 19

• The system function may have many different
forms and may have special names. Such as:
– Driving point admittance
– Transfer impedance
– Voltage or current ratio transfer function
• This is because the excitation and response
may be taken from the same port or different
ports and the excitation and response can be
either voltage or current.
By Mr. Yosef B. (MSc.) 20
 Impedance
• Transfer impedance is when the excitation is a
current source and the response is a voltage.
V0 ( s )
H (s) 
I g ( s)
• When both the excitation and response is at
the same port it is called driving point
impedance.
1
sL
H ( s ) R  sC
1
 sL
By Mr. Yosef B. (MSc.) sC 21
 Admittance
• Transfer admittance is when the excitation is a
voltage source and the response is a current.
I 0 (s)
H (s) 
Vg ( s )

1
H (s) 
1
sL  R
sC

By Mr. Yosef B. (MSc.) 22

 Voltage ratio transfer function
• When the excitation is a voltage source and
the response is a voltage.
V0 ( s )
H (s) 
Vg ( s )

Z 2 (s)
H (s) 
Z1 ( s )  Z 2 ( s )

By Mr. Yosef B. (MSc.) 23

 Current ratio transfer function
• When the excitation is a current source and
the response is a current.
I 0 ( s)
H (s) 
I g ( s)
1
H ( s )  sL  R
1
sC 
sL  R
1
H ( s)  sC
1
R  sL 
sC
By Mr. Yosef B. (MSc.) 24
• Note that, the system function is a function of
the system elements only.
• It is obtained from the network by using the
standard circuit laws. Such as:
– Kirchhoffs law
– Nodal analysis
– Mesh analysis

By Mr. Yosef B. (MSc.) 25

 2.3 Poles and zeros of system
• We will discuss the relationship between the
poles and zeros of a system function and its
steady state sinusoidal response.

of positions of poles and zeros upon H(s) on

• In other words, we will investigate the effect

the jw axis.

By Mr. Yosef B. (MSc.) 26

 a system function we replace ‘s’ by ‘jw’.
• To find the steady-state sinusoidal response of

• Hence, the system function becomes

H ( jw) H ( s ) |s jw
H ( jw) M ( w)e j ( w)

 M(w) is the amplitude or magnitude response

Where

 φ(w) is the phase response

By Mr. Yosef B. (MSc.) 27

• The amplitude and phase response of a
system provide valuable information in the
analysis and design of transmission circuits.
• Consider the low pass filter
• Observe that

below wc
– It passes only frequency

almost linear till wc

– The phase response is

By Mr. Yosef B. (MSc.) 28

 less than wc , then the system will produce
• Hence, if all the significant harmonic terms are

minimum phase distortion.

• In the rest of this section, we will concentrate
on methods to obtain amplitude and phase
response curves.

By Mr. Yosef B. (MSc.) 29

 R-C network
• 1 1
V ( s)
H ( s)  2  sC  RC
V1 ( s ) R  1 s
1

• To obtain H(jw) we substitute s by jw.

sC RC

1
H ( jw )  RC
1
jw 
RC

By Mr. Yosef B. (MSc.) 30

• In polar form H(jw) becomes
1
H ( jw )  RC e  j tan  1 wRC
 M ( w)e j ( w )
1
 2 1  2
w  2 2 
 RC 
1
M ( w)  RC
1
 2 1  2
 w  2 2 
 R C 
 ( w)  tan  1 wRC 
By Mr. Yosef B. (MSc.) 31
• The amplitude is unity and the phase is zero
degrees at w=0.
• The amplitude and phase decrease
monotonically as we increase w.
• When w=1/RC, the amplitude is 0.707 and
phase is -450.
Half power point

• As w increases to infinity M(w) goes to zero and

the phase approaches -900.
By Mr. Yosef B. (MSc.) 32
 Amplitude and phase from pole-zero
diagram
• For the system function
A0 ( s  z0 )( s  z1 )
H ( s) 

• H(jw) can be written as

( s  p0 )( s  p1 )( s  p2 )

A0 ( jw  z0 )( jw  z1 )
H ( jw ) 
( jw  p0 )( jw  p1 )( jw  p2 )

a vector from zi or pj to the jw axis at w.

• Each one of the ( jw  zi ) or ( jw  p j ) represent

By Mr. Yosef B. (MSc.) 33

By Mr. Yosef B. (MSc.) 34
• In general,

By Mr. Yosef B. (MSc.) 35

 Example
4s
F ( s)  2
• For s  2s  2 find the magnitude and
phase for w=2.
• Solution
– First let us find the zeros and poles
4 jw
F ( jw) 
( jw  1  j )( jw  1  j )
– Zero at jw=0
– Poles at ( jw  1  j ) and ( jw  1  j )

By Mr. Yosef B. (MSc.) 36

By Mr. Yosef B. (MSc.) 37
• Magnitude
2 4
M ( j 2) 4 * 
2 * 10 5

• Phase
 ( j 2) 900  71.80  450  26.80

By Mr. Yosef B. (MSc.) 38

 Exercise

• Examine the property of F(s) around the poles

and zeroes.

By Mr. Yosef B. (MSc.) 39

 Bode plots
• In this section we turn our attention to semi
logarithmic plots of system function, called
Bode plots.
• In these plots we take the logarithm of the
amplitude and plot it on linear frequency

• For amplitude M(jw), if we express in terms

scale.

of decibel it becomes
20 log M ( jw)

By Mr. Yosef B. (MSc.) 40

• For system function
N (s)
H ( s) 
D( s)
| N ( jw) |
M ( jw) | H ( jw) |
| D( jw) |
• If we express the amplitude in terms of
decibels we have
20 log M ( jw) 20 log | N ( jw) |  20 log | D( jw) |

By Mr. Yosef B. (MSc.) 41

• In factored from both N(s) and D(s) are made up

Constant K
of 4 kinds of terms

A root at origin, s
1.

A simple real root, s-a

2.
3.
s 2  2s   2   2
4. A complex set of roots,
• To understand the nature of log-amplitude
plots, we only need to discuss the amplitude
response of these 4 terms.
• If the term is on the numerator it carries
positive sign, if on denominator negative sign.
By Mr. Yosef B. (MSc.) 42
 1. Constant K
• The dB gain or loss is
20 log K K 2
• K2 is either positive |K|>1 or negative |K|<1.
• The phase is either 00 for K>0, or 1800 for K<0.

By Mr. Yosef B. (MSc.) 43

 Single root at origin, s
• The loss or gain of a single root at origin is
20 log | jw |20 log w
• Thus the plot of magnitude in dB vs frequency
is a straight line with slope of 20 or -20.
• 20 when s is in the numerator.
• -20 when s is in the denominator.
• The phase is either 900 or -900.
• 900 when s is in the numerator.
• -900 when s is in the denominator.

By Mr. Yosef B. (MSc.) 44

By Mr. Yosef B. (MSc.) 45
 The factor s+α
• For convenience lets set α=1. Then the
magnitude is 1
20 log | jw  1 |20 log w  1 2 2

• The phase is
arg( jw  1) tan  1 w
• A straight line approximation can be obtained

factor jw+1.
by examining the asymptotic behavior of the

By Mr. Yosef B. (MSc.) 46

• For w<<1, the low frequency asymptote is
1
20 log w  1 20 log 1 0dB
2 2

• For w>>1, the high frequency asymptote is

1
20 log w  1 20 log w
2 2

• These 2 asymptotic approximations meet at w=1.

Which has a slope of 20 log w decibel/de cade

By Mr. Yosef B. (MSc.) 47

By Mr. Yosef B. (MSc.) 48
• Note that the maximum error is for w=1 or
for the non normalized one w=α.
• For the general case α different from 1, we
normalize the term by dividing by α.
• The low frequency asymptote is
1
w 2
 2
20 log  2  1 20 log 1 0dB
 
• The high frequency asymptote is
1
w2
 2
20 log  2  1 20 log w  20 log 
  By Mr. Yosef B. (MSc.) 49
 For complex conjugates
• For complex conjugates it is convenient to
adopt a standard symbol.
• We describe the pole (zero) in terms of
magnitude ω0 and angle θ measured from the
negative real axis.
• These parameters that describe the pole
(zero) are ω0, the undamped frequency of
oscillation, and ζ, the damping factor.

By Mr. Yosef B. (MSc.) 50

• If the pole (zero) pair is given as
p1, 2    j

• α and β are related to ω0 and ζ with

 0 cos  0
 0 sin  0 1   2

• Substituting these terms in the conjugate

equation ( s  p1 )( s  p2 )
 
( jw    j )( jw    j )  jw  0  j0 1   2 jw  0  j0 1   2 

  w2  2 jw 0  0
2


By Mr. Yosef B. (MSc.) 51

• For ω0=1 (for convenience), the magnitude of
conjugate pairs can be expressed as
 
1

20 log 1  w2  j 2w 20 log 1  w   4

2 2 2
w 2 2

• The phase is
2 2
 ( w) tan  1
1  w2

By Mr. Yosef B. (MSc.) 52

• The asymptotic behavior is
– For low frequency, w<<1
  20 log 1 0dB
1

20 log 1  w   4
2 2 2
w 2 2

– For high frequency, w>>1

  40 log w
1

20 log 1  w   4
2 2 2
w 2 2

which is a straight line with slope of 40dB/decade.

• These 2 asymptotes meet at w=1.

By Mr. Yosef B. (MSc.) 53

By Mr. Yosef B. (MSc.) 54
 Example
• Using Bode plot asymptotes, draw the
magnitude vs. frequency for the following
system function
0.1s
G ( s) 
 s   s 2
s 
  1  4
 3  1
 50   16 *10 10 

By Mr. Yosef B. (MSc.) 55

By Mr. Yosef B. (MSc.) 56
Actual plot

By Mr. Yosef B. (MSc.) 57