ECEN620: Network Theory

Broadband Circuit Design

Fall 2022

Lecture 2: Linear Circuit Analysis Review

Sam Palermo
Analog & Mixed-Signal Center
Texas A&M University
Agenda
• Transfer Functions
• Mason’s Rule
• Second-Order Systems
• Review Material
• Laplace Transform
• Passive Circuit s-Domain Models
• Transfer Functions
• Sinusoidal Steady-State Response
• Poles & Zeros
• Bode Plots 2
Transfer Function

L vo t  Vo s 
H s   
L vi t  Vi s 

• The transfer function H(s) of a network is the

ratio of the Laplace transform of the output and
input signals when the initial conditions are zero
• This is also the Laplace transform of the
network’s impulse response
3
Sinusoidal Steady-State Response

If input vi t  is sinusoidal

vi t   A cost   

The steady - state output will be

v ss t   H  j  A cost    H  j 

4
Mason’s Rule
[Franklin]

• Mason’s Rule is useful to find the transfer function of

complex networks
• For Mason’s Rule, you need to find the following
• The direct (forward) path(s) from the input(s) to output
• The system loops
• The loops that do not touch the forward path(s)
• Loops that don’t touch, i.e. share elements or nodes 5
Mason’s Rule
[Franklin]

Y s  1
G s   
U s   G 
i
i i

Gi = path gain of the ith forward path

 = the system determinant = 1 – (all individual loop gains) + (gain
products of all possible two loops that do not touch) – (gain products of
all possible three loops that do not touch) + …
i = ith forward path determinant = value of  for that part of the block
diagram that does not touch the ith forward path
6
Mason’s Rule Example 1
Y s  1
[Franklin] G s   
U s   G 
i
i i

System Determinant : Note, all loops touch

 a a a  a a a
  1    1  22  33   0  1  1  22  33
 s s s  s s s
Forward Path Determinants
Note, all loops touch the forward paths
1  1  0  1
Forward Path Gains Loop Gains
2  1 0  1
b1 a1
G1  1236  l1  232   3  1  0  1
s s
b2 a2 System Transfer Function
G2  12346  2
l 2  2342  
s s2 b1 b2 b3
 
b3 a3 Y s  s s2 s3 b1s 2  b2 s  b3
 
G3  123456  l3  23452   U s  a1 a 2 a3 s 3  a1s 2  a 2 s  a3
s3 s3 1  2  3
s s s
7
 Mason’s Rule Example 2
Y s  1
[Franklin] G s   
U s   G 
i
i i

Forward Path Gains System Determinant : Note, 2 loops don' t touch

G1  1236  H 4   1  H 1 H 5  H 2 H 6  H 3 H 7  H 4 H 7 H 6 H 5   H 1 H 5 H 3 H 7 
G2  12456  H1 H 2 H 3 Forward Path Determinants

Note, l 2 does not touch G1

Loop Gains
1  1  H 2 H 6
l1  242  H 1 H 5 does not touch l3 
2  1 0  1
l 2  454  H 2 H 6
System Transfer Function
l3  565  H 3 H 7 does not touch l1 
Y s  H 4 1  H 2 H 6   H1 H 2 H 3

l 4  236542  H 4 H 7 H 6 H 5 U s  1  H1 H 5  H 2 H 6  H 3 H 7  H 4 H 7 H 6 H 5   H 1 H 5 H 3 H 7
8
Second-Order Systems
Y s  1
G s   
U s   G 
i
i i

Forward Path Gain

System Determinant : Note, all loops touch
2
      2  2
G1  k1  0   0 0  0  0 
 s    1     0  1  
 sQ  s   sQ  s 
 
Loop Gains

0 Forward Path Determinant System Transfer Function

l1  
sQ Note, all loops touch the forward path 2
 
k1  0 
2
1  1  0  1 Vo 2 s   s  k102
 
l 2   0  H s    
 s 
Vi s   0   0  s 2  s 0   2
2

1   0
sQ  s  Q
9
Second-Order Systems:
Real or Complex Poles?

k102
H s 
0
s2  s  02
Q
2
0  
2 poles p1 ,p2     0   02
2Q  2Q 

2 real poles if Q  0.5

2 complex conjugate poles if Q  0.5

10
Bode Plots
• Technique to plot the Magnitude (squared) and
Phase response of a transfer function
• Magnitude is plotted in Decibels (dB), which is a power
ratio unit
dB

H  j  10 log10 H  j 
2 2
dB  20 log
10  H  j   dB
• Phase is typically plotted in degrees

ImH  j  
1 
H  j   tan  
 ReH  j  

11
Second-Order Systems – Real Poles (1)
104 104 2 poles : p1  1, p2  1000
H s   2 
s  1001s  1000 s  1s  1000 
Note, Q  0.032

-20dB/dec.
-45/dec.

-40dB/dec.

-45/dec.

• If poles are spaced by more than 2 decades, there

are 2 distinct regions of -45/dec phase slope
12
Second-Order Systems – Real Poles (2)
100 100 2 poles : p1  1, p2  10
H s   2 
s  11s  10 s  1s  10 Note, Q  0.287

-20dB/dec. -45/dec.

-40dB/dec.
-90/dec.

-45/dec.

• If poles are spaced by less than 2 decades, there is a region

of -90/dec phase slope
• Watch out for system stability!
13
 Second-Order Systems – Complex Poles

k102
H s  
0
s2  s  02
Q
What is the low frequency magnitude?
H  j 0   k1
What is the high frequency magnitude?
k102
H  j    -40dB/dec. slope at high frequencies
   2

What happens in the middle, particularly near 0?

k102
H  j 0    k1Q
02
 02  j  02
Q
Note, if Q  1 then the magnitude exceeds the low frequency value, i.e. frequency peaking occurs!

14
Frequency Peaking w/ Complex Poles
Where is the peak frequency? For k1=1 and 0=1

 
 
d H  j 
2
d  2
k1 04 
  0
d d  2

 02
  
2 2  0
 
 
 Q  
   

1
 pk  0 1   0 for large Q
2Q 2

At  pk , the peak value is

k1Q
T pk   k1Q for large Q
1
1
4Q 2

• Note, phase always

crosses -90 at 0
[Schaumann] 15
Next Time
• PLL System Analysis

16
Review Material
• The following material reviews Laplace
transforms, transfer functions, sinusoidal
steady-state response, and Bode plots

• Please review this material, as it is

fundamental for the analysis of the
broadband circuits covered it the class

17
References
• Continuous & Discrete Signal & System Analysis,
3rd Ed., C. McGillem and G. Cooper, Saunders
College Publishing, 1991.

• Feedback Control of Dynamic Systems, 3rd Ed., G.

Franklin, J. Powell, and A. Emami-Naeini,
Addison-Wesley, 1994.

• Design of Analog Filters, R. Schaumann and M.

Van Valkenburg, Oxford University Press, 2001.

18
Laplace Transform
• Laplace transforms are useful for solving
differential equations
• One-Sided Laplace Transform

Lxt   X s   xt e  st dt

0

where s is a complex variable

s    j
Note, j   1 and  is the angular frequency (rad/s)

• s has units of inverse seconds (s-1)

19
Laplace Transform of Signals

[McGillem]

20
Laplace Transform of Operations

[McGillem]

21
Resistor s-Domain Equivalent Circuit
vt   Rit 
Time-domain Representation: 1
i t   vt 
R

1
Complex Frequency V s   RI s  I s   V s 
R
Representation:

22
Capacitor s-Domain Equivalent Circuit
t
1
vt   i  d  v0 

C
Time-domain Representation: 0
dv
i t   C
dt

V s  
1 1
I s   v0  I s   sCV s   Cv0
Complex Frequency sC s
Representation:

23
Inductor s-Domain Equivalent Circuit
di
vt   L
dt
Time-domain Representation: t
1
i t   v d  i0 

L
0

1 1
Complex Frequency V s   sLs   Li 0 I s   V s   i0
sL s
Representation:

24
s-Domain Impedance w/o I.C.

V s   I s R 1 V s   I s sL
V s   I s 
SC
Z s   R Z s   sL
1
Z s  
sC

25
Transfer Function

L vo t  Vo s 
H s   
L vi t  Vi s 

• The transfer function H(s) of a network is the

ratio of the Laplace transform of the output and
input signals when the initial conditions are zero
• This is also the Laplace transform of the
network’s impulse response
26
RC Transfer Function

1
ZC 1
Vo s   Vin s   sC Vin s   Vin s 
Z R  ZC 1 1  sRC
R
sC

AC Transfer Function, H(S)

Vo s  1
H s   
Vin s  1  sRC

27
Laplace Transform Circuit Example
Given vo 0  0

Convert to
Laplace Domain
Vo s  1 1 10 5
H s     
Vin s  1  sRC 1  s s  10 5
10 5
 10 5  10 5 

Vo s   H s Vi s    
 s  10 5  2
  
 s  10 5 
2 

with partial fraction expansion

1 1
s  
1 5
10
Vo s   2  2  2
 
s  10 5 s 2  10 5 2 s 2  10 5 2 
with inverse Laplace Transform

v o t  
1 105 t
2
e
1
2
1
2
1
2
5
 cos10 5 t  sin 10 5 t  e 10 t 
1

sin 10 5 t  45 
2
28
Laplace Transform Circuit Example
We can decompose the output into it' s transient and steady - state response
1
v o t   e 10 t 
2
5 1
 
sin 10 5 t  45  vtr t   v ss t 
2
1
v tr t   e 10 t
5

2
v ss t  
1
 
sin 10 5 t  45
2

• Note that the transient response decays very quickly!

29
Sinusoidal Steady-State Response

If input vi t  is sinusoidal

vi t   A cost   

The steady - state output will be

v ss t   H  j  A cost    H  j 

30
RC Circuit Sinusoidal Steady-State Response

Vo s  1 s  j 1
H s     H  j  
Vin s  1  sRC 1  jRC

Output Magnitude
   1
H  j   H  j H  j   
* 1

1
 H  j  
 1  jRC  1  jRC  1  RC 2

Output Phase
 ImH  j    ImNum    ImDen  
H  j   tan 1    tan 1    tan 1  
 Re  H  j     Re  Num    Re  Den  
H  j    tan 1 RC 
where Num  Numerator and Den  Denominator of H  j 

0  RC 
H  j   tan 1    tan 1     tan RC 
1
1  1 

31
RC Circuit Sinusoidal Steady-State
Response Example

1
H s  
s
1
10 5
with s  j  j10 5


H j10 5   1
1 j


H j10 5
 1
2

1
2

 
H j10 5   tan 1 1  45

v ss t  
1

sin 10 5 t  45 
2
32
Complex Numbers Properties
[Silva]

Numerical Example
1  j1010  j10
100  j101000  j10
1  j1010  j10 
12  10 2 10 2  10 2
 1.41 10 3
100  j101000  j10 100 2  10 2 1000 2  10 2


1  j1010  j10  tan 1  10   tan 1  10   tan 1  10   tan 1  10   123
       
100  j101000  j10  1 10
   100   1000 
33
 Poles & Zeros
H s   A
s  z1 s  z2 ...s  zm 
s  p1 s  p2 ...s  pn 
• Poles are the roots of the denominator (p1, p2, … pn) where H(s)∞
• Zeros are the roots of the numerator (z1, z2, … zm) where H(s)0
105
Example 1 : H s   100s  15
s  105 Example 3 : H s  
s 2  50 s  1500
s  105  0
s  15  0
p1  s  105 rad / s z1  s  15rad / s
s s 2  50 s  1500  0
Example 2 : H s  
s  105
z1  s  0rad / s  50  2500  6000
p1, 2  s1, 2   25  j 29.6rad / s
2
s  105  0

p1  s  105 rad / s
34
Bode Plots
• Technique to plot the Magnitude (squared) and
Phase response of a transfer function
• Magnitude is plotted in Decibels (dB), which is a power
ratio unit
dB

H  j  10 log10 H  j 
2 2
dB  20 log
10  H  j   dB
• Phase is typically plotted in degrees

ImH  j  
1 
H  j   tan  
 ReH  j  

35
RC Bode Plot Example

Vo s  1 1 1
H s     
Vin s  1  sRC 1  s10 5 1  j10 5

1 1
H s    , where p1  10 5 rad/s
1  j10 5 j
1
p1

Magnitude Squared (dB):

20 log10 H  j   20 log10
1 

 20 log10 1  20 log10  1  10 5 
2 


1  10 5 
2  

Phase: PhaseH  j    tan 1 10 5 

36
 RC Bode Plot Example
Magnitude:

20 log10 H  j   20 log10
1 

 20 log10 1  20 log10  1  10 5 
2 


1  10 5 
2  

Phase: 
PhaseH  j    tan 1 10 5 
 (rad/s) |H(j)| |H(j)|2 20log10|H(j)| (dB) Phase (H(j)) ()
103 0.9999 0.9999 ~0 ~0
~20log10 (1) 104 0.995 0.990 -0.043 -5.71
= 0dB 5x104 0.894 0.800 -0.969 -26.6
105 0.707 0.500 -3.01 -45.0 -45/dec
5x105 0.196 0.039 -14.2 -78.7
106 0.100 0.010 -20.0 -84.3
~-20log10 (10-5) 107 10-2 10-4 -40.0 -89.4
= -20dB/dec
108 10-3 10-6 -60.0 -89.9

37
RC Bode Plot Example
Max Error = 3.01dB

-20dB/dec

Max Error = 5.71

-45/dec

38
Transient Response

 = 103 rad/s = -p1/100  = 105 rad/s = -p1  = 106 rad/s = 10*p1

vo t   1 1 vo t   0.1
vo t  

2
Phase Shift  0 Phase Shift  84.3
Phase Shift  -45

39
Bode Plot Algorithm - Magnitude
1. Where is a good starting point?
a. Calculate DC value of |H(j)|
b. If not a reasonable value, I like to calculate |H(j)| at  equal
to the lowest value of p1/10 or z1/10
2. Where to end?
a. Calculate |H(j)| as ∞
3. Where are the poles and zeros?
a. Beginning at each pole frequency, the magnitude will decrease
with a slope of -20dB/dec
b. Beginning at each zero frequency, the magnitude will increase
with a slope of +20dB/dec
4. Note, the above algorithm is only valid for real poles
and zeros. We will discuss complex poles later.

40
 Bode Plot Algorithm - Magnitude
104 s  1 101  s 
H s    
s  10s  100 1  s 1  s  20 log10 H  j   20 log10
10 1   2

  
 10  100  1  10 
1 2
1  10 
2 2

DC Magnitude  10  20dB
 
20 log10 10   20 log10 1   2  20 log10  1  101    20 log10  1  10 2  

2

 
2


HF Magnitude  0  dB

z1  1, p1  10, p2  100

+20dB/dec. -20dB/dec.

-20dB/dec.

+20dB/dec.
-20dB/dec.

41
Bode Plot Algorithm - Phase
1. Calculate low frequency value of Phase(H(j))
a. An negative sign introduces -180 phase shift
b. A DC pole introduces -90 phase shift
c. A DC zero introduces +90 phase shift
2. Where are the poles and zeros?
a. For negative poles: 1 dec. before the pole freq., the phase will
decrease with a slope of -45/dec. until 1 dec. after the pole
freq., for a total phase shift of -90
b. For zeros poles: 1 dec. before the zero freq., the phase will
increase with a slope of +45/dec. until 1 dec. after the zero
freq., for a total phase shift of +90
c. Note, if you have positive poles or zeros, the phase change
polarity is inverted
3. Note, the above algorithm is only valid for real poles
and zeros. We will discuss complex poles later. 42
 Bode Plot Algorithm - Phase

104 s  1 101  s        
H s     H  j   180  tan 1    tan 1    tan 1  
s  10s  100 1  s 1  s  1  10   100 
  
 10  100 
LF Phase   180
z1  1, p1  10, p2  100

+45/dec.
-90/dec.

+45/dec. -45/dec.
-45/dec. -45/dec.

43