Adapted from notes by

ECE 5317-6351 Prof. Jeffery T. Williams

Microwave Engineering
Fall 2019
Prof. David R. Jackson
Dept. of ECE

Notes 16
Network Analysis

1
 Multiport Networks
A general circuit can be represented by a multi-port network, where the “ports” are
defined as access terminals at which we can define voltages and currents.

Note:
Equal and opposite currents are assumed on
the two wires of a port.
Examples:

1) One-port network
+ +
R
- -

2) Two-port network

+
+ + +

- - - -

2
 Multiport Networks (cont.)

3) N-port network
+ -


+ +
- -


+ 
- 
  
+ -

Note: Passive sign convention is used at the ports.

3
 Multiport Networks (cont.)
+ -


+ +
N-port network - -


+ 
- 
To represent multi-port networks we use:   
 Z (impedance) parameters + -
Not easily
 Y (admittance) parameters
measurable at
 ABCD parameters high frequency

 S (scattering) parameters Measurable at high frequency

4
 Two-Port Networks
Consider a 2-port linear network:

I1 I2
+ +
V1 1 2 V2
- -

In terms of Z-parameters, we have (from superposition):

V1 Z11 I1 + Z12 I 2
= Impedance (Z) matrix
V2 Z 21 I1 + Z 22 I 2
=

Therefore
V1   Z11 Z12   I1 
=V   Z    ⇒ [V ]
= [ Z ][ I ]
 2   21 Z 22   I 2 
5
 Elements of Z-Matrix: Z-Parameters
V1 Z11 I1 + Z12 I 2
=
V2 Z 21 I1 + Z 22 I 2
=

Port 2 open circuited Port 1 open circuited

V V1
Z11 = 1 Vi Z12 =
I1 I 2 =0
Z ij = I2 I1 =0
Ij
Ik 0 k ≠ j
=

V V2
Z 21 = 2 Z 22 =
I1 I 2 =0
I2 I1 =0
I1 I2
+ +
V1 1 2 V2
- -

6
 Z-Parameters (cont.)
N-port network


Vi
Z ij =
Ij 
Ik 0 k ≠ j
=
+ 
- 
  

We inject a current into port j and measure the voltage (with an ideal
voltmeter) at port i. All ports are open-circuited except j.

7
 Summary of Z Parameters
Summary of Z Parameters

[V ] = [ Z ][ I ] Vi
Z ij =
Ij
Ik 0 k ≠ j
=

+ - N-port network

 
+ +
- -

 
+  + 
-  - 
     
+ -
8
 Admittance (Y) Parameters
Consider a 2-port linear network:

I1 I2
+ +
V1 1 2 V2
- -

I1 Y11V1 + Y12V2
= Admittance matrix

I 2 Y21V1 + Y22V2
=

or

 I1  Y11 Y12  V1 

= I  Y Y  V = ⇒ [I ] [Y ][V ]
 2   21 22   2 

Ii
Yij =
Vj
Vk 0 k ≠ j
=
9
 Y-Parameters (cont.)
N-port network


[ I ] = [Y ][V ] +-

Ii 
Yij = 
Vj
Vk 0 k ≠ j
=

  

We apply a voltage across port j and measure the current (with an

ideal current meter) at port i. All ports are short-circuited except j.

10
 Summary of Y Parameters
Summary of Y Parameters

[ I ] = [Y ][V ] Ii
Yij =
Vj
Vk 0 k ≠ j
=

+ - N-port network

 
+
-
+ +-
-

 
+  
-  
     
+ -
11
Relation Between Z and Y Parameters
Relation between [Z] and [Y] matrices:

[V ] = [ Z ][ I ]
[ I ] = [Y ][V ]
Hence:
[V ] = [ Z ] ([Y ][V ])
= ([ Z ][Y ] ) [V ]

It follows that

[ Z ][=
Y ] [U
= ] Identity Matrix

Therefore [Y ] = [ Z ] −1

12
 Reciprocal Networks
If a network does not contain non-reciprocal devices or materials*
(i.e. ferrites, or active devices), then the network is “reciprocal”, which
means that the Z and Y matrices are symmetric.

Z ij Z ji
⇒= (Yij Y ji )
=
Note:
⇒ [ Z ] and [Y ] are symmetric matrices The inverse of a
symmetric matrix is
(proof omitted) symmetric.

* A reciprocal material is one that has symmetric permittivity and permeability matrices. A
reciprocal device is one that is made from reciprocal materials.

Example of a nonreciprocal material: a biased ferrite

(This is very useful for making isolators and circulators.)

13
 Reciprocal Materials
ε ⋅E
D= µ⋅Η
B=

 Dx  ε xx ε xy ε xz   Ex   Bx   µ xx µ xy µ xz   H x 
 D  = ε 
ε yy ε yz   E y  B  = µ µ yy

µ yz   H y 
 y   yx  y   yx
 Dz  ε zx ε zy ε zz   Ez   Bz   µ zx µ zy µ zz   H z 

Reciprocal:=ε ij ε=
ji , µij µ ji

 α jγ 0
µ µ 0  − j γ
Ferrite: = α 0  µ is not symmetric!

 0 0 1  (not a reciprocal material)

14
 Reciprocal Networks (cont.)
We can show that the equivalent circuits for reciprocal 2-port networks are:

T-equivalent
“T network”

Pi-equivalent
“Π network”

15
 ABCD-Parameters
There are defined only for 2-port networks.

I 2′
V1   A B  V2 
 I  = C D   I ′ 
 1   2
I 2′ = − I 2

V1 I1
=A = C
=
V2 I ′ 0= V2 I ′ 0
2 2

V1 I1
=B = D
=
I 2′ V 0= I 2′ V 0
2 2

16
 Cascaded Networks
Port 1 Port 2

I 2′

− A B −

V1  V1A  A V2 

A
=
 I  = A
 ABCD   A  A nice property of the ABCD
 1   I1   I 2′  matrix is that it is easy to use
B with cascaded networks: you
A V 1  simply multiply the ABCD
=  ABCD   B 
matrices together.
 I1 
A V2B 
B
=  ABCD   ABCD   B 
 I 2′ 

V1  AB V2 
I  =  ABCD   
 1  I 2′ 
17
 Scattering Parameters

 At high frequencies, Z, Y, and ABCD parameters are difficult

(if not impossible) to measure.
 V and I are not always uniquely defined (e.g., microstrip, waveguides).
 Even if defined, V and I are extremely difficult to measure (particularly I).
 Required open and short-circuit conditions are often difficult to achieve.

 Scattering (S) parameters are often the best

representation for multi-port networks at high frequency.

Note: We can always convert from S parameters to Z, Y, or ABCD parameters.

18
 Scattering Parameters (cont.)
A Vector Network Analyzer (VNA) is usually used to measure S parameters.

Port 1
Port 2

Keysight (formerly Agilent) VNA shown performing a measurement.

19
 Scattering Parameters (cont.)
For scattering parameters, we think in terms of incident and
reflected waves on transmission lines connected to a device.

Local coordinates

 The a coefficients represent incident waves.

 The b coefficients represent reflected waves.

20
 Scattering Parameters (cont.)

Local coordinates

On each transmission line:

Vi ( zi ) = Vi 0+ e −γ i zi + Vi 0− e + γ i zi = Vi + ( zi ) + Vi − ( zi )
Vi + ( zi ) Vi − ( zi )
I i ( zi )
= − i = 1, 2
Z 0i Z 0i

We define normalized voltage wave functions:

Incoming wave function ≡ ai ( zi ) ≡ Vi ( zi ) +

Z 0i
Outgoing wave function ≡ bi ( zi ) ≡ Vi ( zi )
−
Z 0i
21
 Scattering Parameters (cont.)

Why are the wave functions (a and b) defined as they are?

 The normalization makes power calculation easy (see next slide).

 The [S] matrix is unitary (discussed later).

22
 Scattering Parameters (cont.)
Power Calculations

i = 1, 2
2
i ( 0)
+
1 1 V
Pi + ( 0 ) Re Vi ( 0 ) I i ( 0 )
 
*
+ +
= (assuming lossless lines, so Z0 is real)
2   2 Z
0i

Recall: ai ( 0 ) ≡ Vi + ( 0 ) Z 0i Similarly,
2
1 Vi ( 0 )
−
1 2 1 2
⇒ Pi ( 0 ) =
+
ai ( 0 ) Pi ( 0 ) =
= −
bi ( 0 )
2 2 Z 0i 2

23
 A One-Port Network

V1− ( 0 )
ΓL = +
V1 ( 0 )
V1− ( 0 ) Z 01
=
V1+ ( 0 ) Z 01
b1 ( 0 )
=
a1 ( 0 ) b1 ( 0 ) = S11a1 ( 0 ) Definition of S11 for a one-port

For a one-port
S11 = Γ L network, S11 is the
same as ΓL.
Recall:
Incoming wave function ≡ ai ( zi ) ≡ Vi ( zi )
+
Z 0i
Outgoing wave function ≡ bi ( zi ) ≡ Vi ( zi )
−
Z 0i
24
 A Two-Port Network

From linearity:

b1 ( 0 ) S11a1 ( 0 ) + S12 a2 ( 0 )
=
Scattering matrix
b2 ( 0 ) S 21a1 ( 0 ) + S 22 a2 ( 0 )
=

or

 b1 ( 0 )   S11 S12   a1 ( 0 ) 
=
     =  ⇒ [b ] [ S ][ a ]
b
 2 ( 0 )   S 21 S 22   a2 ( 0 ) 

25
 A Two-Port Network (cont.)
b1 ( 0 ) S11a1 ( 0 ) + S12a2 ( 0 )
=
b2 ( 0 ) S21a1 ( 0 ) + S22a2 ( 0 )
=

b1 ( 0 ) Output is
S11 = matched input reflection coef.
a1 ( 0 ) a =0
2 w/ output matched
b1 ( 0 ) Input is
S12 = reverse transmission coef.
a2 ( 0 ) a = 0 matched
1 w/ input matched
b2 ( 0 ) Output is
S21 = forward transmission coef.
a1 ( 0 ) a =0 matched
w/ output matched
2

b2 ( 0 ) Input is
output reflection coef.
S22 = matched
a2 ( 0 ) a = 0 w/ input matched
1

26
 A Two-Port Network (cont.)
1
S11 1 2 
S11 Matched or semi-infinite b1 ( 0 )
S11 =
a1 ( 0 ) a =0
S12 1 2

S12  1 2 b1 ( 0 )
S12 =
Matched or semi-infinite a2 ( 0 ) a = 0
1

b2 ( 0 )
1 S 21 S21 =
a1 ( 0 ) a =0
S21 1 2  2

Matched or semi-infinite b2 ( 0 )
S22 =
a2 ( 0 ) a = 0
1 1

S22
 1 2
Matched or semi-infinite S 22
27
 Three-Port Network
Illustration of a three-port network

2
1
3

b1 ( 0 ) = S11a1 ( 0 ) + S12 a2 ( 0 ) + S13a3 ( 0 )

b2 ( 0 ) = S21a1 ( 0 ) + S22 a2 ( 0 ) + S23a3 ( 0 )
b3 ( 0 ) = S31a1 ( 0 ) + S23a2 ( 0 ) + S33a3 ( 0 )

To Illustrate:
b2 ( 0 )
b2 ( 0 ) =S21a1 ( 0 ) + S22a2 ( 0 ) + S23a3 ( 0 ) ⇒ S21 = when a2 =a3 =0
a1 ( 0 )
28
 Three-Port Network (cont.)

Illustration of S21:

Semi-infinite

1 2 S 21 •••

1
3 •••

b2 ( 0 )
S 21= when a= a= 0
a1 ( 0 )
2 3

29
 N-Port Network
For a general multiport network:
All ports except port j are semi-infinite (or with
bi ( 0 ) matched load at ports) with no incident wave.
Sij =
a j ( 0)
k 0 k≠ j
a= Semi - infinite N-port network

•••
••• aj Port j

•••

•••
•••

Port i bi •••
•••

We send in an incident wave on port j and measure the outgoing wave on

port i, when all lines except j are semi-infinite (or terminated in a matched
load), and thus there is an incident wave only on port j.
30
 Summary of S Parameters
Summary of S Parameters

[b] = [ S ][ a ] bi ( 0 )
Sij =
a j ( 0)
k 0 k≠ j
a=

N-port network Semi - infinite

•••
aN bN
••• ••• aj Port j
a1 am
•••

b1 bm
a2

•••
•••

•••

b2 Port i bi
••• •••
a3 b3
•••

31
Scattering Parameters with Waveguides
A microwave system may have waveguides entering a device. In this case,
the transmission lines are TEN models for the waveguides.

TE10 TE10
Waveguide Device Waveguide

TEN

Local coordinates

32
 Shift in Reference Planes
(illustrated for a two-port)
z1 = −l1 z2 = −l2
l1 l2

i, j = 1 or 2

bi ( zi ) bi ( 0 ) e +γ i zi bi ( 0 ) e −γ i li
=Sij′ ⇒ Sij′
= ⇒ Sij′
=
aj (zj ) a j ( 0) e a j ( 0) e
−γ j z j +γ j l j

k 0 k≠ j
a= 0 k≠ j
ak = 0 k≠ j
ak =

− γ i li −γ j l j
Sij′ = Sij e e
33
 Properties of the S Matrix
For reciprocal networks (networks containing only reciprocal
materials), the S-matrix is symmetric:

=Sij S ji i≠ j

⇒ [ S ] = [ S ]T

Example of a nonreciprocal material: a biased ferrite

34
 Properties of the S Matrix (cont.)
 For lossless networks, the S-matrix is unitary†.

⇒ [=
S ] [ S ] [=
T *
S ] [ S ] [U ]
* T
Identity matrix

Note :
[ ][ ] [ ]
If A B = U
Hence,
then [ B ][ A] = [U ]
−1
[ S ] = [S ]
T −1*
= S * 
Alternate notation:

[S ] T*
= [S ]
−1
[=
S ] [S ] ≡ [S ]
† H T*

Therefore, [ S ] = [ S ]
† −1
(a “unitary” matrix)
(“Hermetian conjugate” or “Hermetian transpose”)

†A proof is in the Pozar book.

35
 Properties of the S Matrix (cont.)
Start with the first part of the unitary equation:

[ S ] [ S ] = [U ]
T *

N-port network

N N
Take (i , j )
=
⇒
element ∑ T
ik
k 1=
*
S S =
kj
k 1
∑ Ski Skj* = δ ij

1 ; i = j
δ ij = 
0; i ≠ j
36
 Properties of the S Matrix (cont.)
N

∑
k =1
Ski Skj* = δ ij

Interpretation: The inner product of columns i and j is zero unless i = j.

 S11 S12 S13 

[ S ] =  S21 S23 
The rows also form orthogonal sets
S22 (this follows from starting with the
 second part of the unitary equation).
 S31 S23 S33 

S1 vector S3 vector

δ ij
Si ⋅ S *j =

Physical interpretation: All of the power outgoing on the ports is equal to

all of the power incident on the ports.
37
 Comment on Normalization
Note: If all lines entering the network have the same characteristic impedance, then

bi ( 0 ) Vi − ( 0 ) In this case, there is no difference

Sij =
= between normalized and
a j ( 0 ) V j+ ( 0 ) + unnormalized scattering parameters.
k 0 k≠ j
V=

In general (different characteristic impedances):

Vi − ( 0 )
Sij ≡ + “unnormalized” scattering parameters
V j ( 0) +
k 0 k≠ j
V=

bi ( 0 )
Sij ≡ “normalized” scattering parameters
a j ( 0)
k 0 k≠ j
a=

Note:
The unitary property of the scattering matrix requires normalized parameters.
We use normalized parameters in these notes.
38
 Note on Z0
Important Note:
The S parameters depend on Z0.
(The Z and Y parameters do not.)

Z0 Zs Z0

Example: The device is a section of transmission line.

=S11 0=
when Z 0 Z s
S11 ≠ 0 when Z 0 ≠ Z s ( in general )

39
 Scattering Parameters (cont.)
A general formula for converting from Z parameters to S parameters is:
−1
 1  1 
[ ]  [ ] [ ] [ ]
S = Z − U U + [ ]
Z
 Z0  Z0 

Z 0 ([U ] + [ S ]) ( ([U ] − [ S ]) )
−1
[Z ] =
(This assumes all transmission lines are identical with characteristic impedance Z0.)

Z0

Z0 Z0
N − port
Z0 Network Z0 Note:
The derivation is in
the Pozar book.
Z0

40
 Example
Find the S parameters for a series impedance Z.

Note that two different coordinate systems are being used here!

41
 Example (cont.)
Semi-infinite

S11 Calculation:

b1 ( 0 ) V1− ( 0 ) Z in − Z 0 ( Z + Z0 ) − Z0
S11
= = = =
a1 ( 0 ) a =0 V1 ( 0 ) Z in + Z 0
+
( Z + Z0 ) + Z0
2 a2 = 0

By symmetry:
Z
⇒ S11 = S 22 = S11
Z + 2Z0

42
 Example (cont.)
S21 Calculation:
Semi-infinite

2 ( z2 ) 2 ( z2 ) V2− ( 0 ) e + jk z 2 z2
−
V= V=
b2 ( 0 )
S21 =
a1 ( 0 ) a =0
 Z0 
( 0) V= ( ) ( )
2
−
Voltage divider: V2= 0 V 0  
V2− ( 0 ) 2 1
 Z + Z 0 
= +
V1 ( 0 ) a =0
V1 ( 0 ) V1+ ( 0 )(1 + S11 )
Input voltage: =
2

 Z0 
V ( 0 ) V ( 0 )(1 + S11 ) 
From first equation: = 2
−
1
+

 Z + Z 0 

V2− ( 0 )  Z0 
S21= = ( 11 ) 
1 + S 
V1 ( 0 )
+
 Z + Z0 
43
 Example (cont.)
Semi-infinite

Hence
 Z   Z   Z0   2Z + 2Z0   Z0   Z + Z0   Z0 
(1 + S11 )  0  =
S21 =  1 +  =
    2
=   
 Z + Z0   Z + 2Z0   Z + Z0   Z + 2Z0   Z + Z0   Z + 2Z0   Z + Z0 

We then have:

2 Z0
S21 = S12 = S21
Z + 2 Z0

44
 Example
Find the S parameters for a length L of transmission line.

Note that three different coordinate systems are being used here!

(The subscript “s” denotes the middle section.)

45
 Example (cont.)
Z − Z0s
S11 Calculation: ΓL = 0
Z0 + Z0s

+ +
+
- - -

Semi-infinite

b1 Z in a =0 − Z 0
S11 =
= 2
= S22 (by symmetry )
a1 a =0 Z in a =0 + Z 0
2 2

=Z in a =0 Z=
( Z 0 + jZ 0 s tan β s L ) Z0s
(1 + Γ e
L
− j 2 βs L
)
2
0s
( Z 0 s + jZ 0 tan β s L ) (1 − Γ e
L
− j 2 βs L
)

46
 Example (cont.)

+ +
+
- - -

Semi-infinite

Hence

Z0s
( Z 0 + jZ 0 s tan β s L )
− Z0
S= S=
( 0s
Z + jZ 0 tan β s L )
11 22

Z0s
( Z 0 + jZ 0 s tan β s L )
+ Z0
( Z 0 s + jZ 0 tan β s L )

Note: If Z 0 s =Z 0 ⇒ Z in a =0 =Z 0 ⇒ S11 =S22 =0

2

47
 Example (cont.)
Z0 − Z0s
S21 Calculation: ΓL =
Z0 + Z0s

+ +
+
- - -

Semi-infinite

b2 V2− ( 0 ) Z0 V2− ( 0 )
S21 =
= = +
a1 a =0 V1+ ( 0 )
2
Z 0 a =0 V1 ( 0 ) V − =0
2 2

V1 ( 0 ) V1+ ( 0 )(1 + S11 )

Total voltage at port 1: =

Hence, for the denominator of the S21 equation we have

V1 ( 0 ) We now try to put the

V1 ( 0 ) =
+
numerator of the S21 equation
1 + S11 in terms of V1 (0).

48
 Example (cont.)

+ +
+
Vs ( z )
- - -

Semi-infinite

( 0) V2=
V2−= ( 0) Vs=
( 0) Vs+ ( 0)(1 + Γ L )
Next, we need to get Vs+(0) in terms of V1(0):

Vs ( z ) Vs+ ( 0 ) e − jβ s z (1 + Γ L e + j 2 β s z )
) Vs ( − L=) Vs+ ( 0) e + jβs L (1 + Γ Le − j 2 βs L )
⇒ V1 ( 0=
V1 ( 0 ) Hence, from the top equation we have
⇒V +
( 0) =
s + jβ L
e s
(1 + Γ e L
− j 2 βs L
) V1 ( 0 )
=V ( 0) −
(1 + Γ L )
2
e + jβ s L
(1 + Γ e
L
− j 2 βs L
)
49
 Example (cont.)

+ +
+
- - -

Semi-infinite

V1 ( 0 )
=V2− ( 0 ) (1 + Γ L )
e + jβ s L (1 + Γ L e − j 2 β s L )

Therefore, we have V1 ( 0 )
V1+ ( 0 ) =
1 + S11
V2− ( 0 ) (1 + S11 )(1 + Γ L ) e− jβ L s

=S 21 =
V1+ ( 0 ) a =0 1 + Γ L e− j 2 βs L
2

so

S 21
(=
1 + S11 )(1 + Γ L ) e − j β L s

S12 (the S matrix is symmetric)

− j 2 βs L
1 + Γ Le
50
 Example (cont.)
Special cases:
a) Z 0 s = Z 0

Z 0 s = Z 0 ⇒ S11= S 22 = 0, ΓL = 0
− jβs L
S 21
= S=
12 e

 0 e − jβ s L 
a) [ S ] =  − jβ s L 
λg e 0 
b) L =
2
λg 2π λg
L= ⇒ βs L = =π
2 λg 2  0 −1
b) [S ] =  
⇒ Z in a2 = 0
= Z0 ⇒ S11 = S 22 = 0
 − 1 0 
e− jβs L =
−1, e − j 2 β s L =
+1 ⇒ −1
S 21 =
51
 Example
Find the S parameters for a step-impedance discontinuity.

S11 Calculation:

Z 02 − Z 01
S11 =
Z 02 + Z 01
Z 01 − Z 02
S22 = = − S11
Z 02 + Z 01

S21 Calculation:
V2− ( 0 )
b2 ( 0 ) Z 02
=S21 =
a1 ( 0 ) a =0 V1+ ( 0 )
2

Z 01 a2 = 0
52
 Example (cont.)
Semi-infinite
Because of continuity of the voltage +
across the junction, we have: +
-
-

V2− ( 0 ) 2 ( 0 ) a =0
= V= 1 ( 0)
V= V1+ ( 0 )(1 + S11 )
a2 = 0 2 a2 = 0

V2− ( 0 ) V1+ ( 0 )(1 + S11 ) Z 02 − Z 01

1 + S11 =+
1
Z 02 Z 02 Z 02 + Z 01
S 21 =
V1+ ( 0 ) V1+ ( 0 ) 2 Z 02
=
Z Z Z 02 + Z 01
=01 a2 0=01 a2 0

so
Z 01 Z 01Z 02
S21= (1 + S11 ) Hence S= S= 2
Z 02 21 12
Z 01 + Z 02
53
 Example
 −j −j
 0 2 2
  2
−j 1
 S50Ω  =  0 0  S50Ω
 2 
  3
−j 0 0 
 2 
These are the S parameters assuming
50 Ω lines entering the device.
Not unitary → lossy
(For example, column 2 doted with the conjugate of column 3 is not zero.)

1) Find the input impedance looking into port 1 when ports 2 and 3 are
terminated in 50 [Ω] loads.
2) Find the input impedance looking into port 1 when port 2 is terminated in a
75 [Ω] load and port 3 is terminated in a 50 [Ω] load.

54
 Example (cont.)
1) Zin if ports 2 and 3 are terminated in 50 [Ω]: a=
2 a=
3 0

b1 = S11a1 + S12 a2 + S13 a3  −j −j

 0 2 2
 b1     a1 
b1 b  =  − j 0 0  a 
⇒ Γin1= = S11= 0 ⇒ Z in1= Z 01  2  2   2
a1  b3     a3 
−j 0 0 
 2 
Z=
in1 50 [Ω]

2
1
3

55
 Example (cont.)
2) Zin if port 2 is terminated in 75 [Ω] and port 3 in 50 [Ω]:

a2 75 − 50 1  −j −j
Γ2 = = =  0 2 2
b2 75 + 50 5  b1     a1 
b  =  − j 0 0  a 
 2  2   2
a3 50 − 50  b3     a3 
Γ3 = = = 0
b3 50 + 50 −j 0 0 
 2 

2
1
3

56
 Example (cont.)
 −j −j
 0 2 2
2
 b1     a1  1
b  =  − j 0 0  a  3
 2  2   2
 b3     a3 
−j 0 0 
 2 
a  a 
b2 / a1 =
S21 + S22  2  + S23  3 
b1 a a  a1   a1 
⇒ Γin1= = S11 + S12 2 + S13 3
a1 a1 a1
Γ b   − j  1  − j  1
= S12 ( 2 S 21 )
S12  2 2  =Γ =
    =
−
 a1   2  5  2  10

a2 = Γ 2b2

 1 + Γin1 
Z in1 50  =
Hence =  44.55 [Ω]
 1 − Γin 2 
57
 Example
Find Γin for the general two-port system shown below.

Z0 Z0
Z − Z0  S11 S12  Z − Z0
Γin = in Z in ZL ΓL = L
Z in + Z 0 S S 22  Z L − Z0
V1+ ( 0 )  21 V2+ ( 0 )
V1− ( 0 ) V2− ( 0 )
Two-port device

Assume : V1+ ( 0 ) = 1V

so Γ=
in V1− (=
0 ) S11 (1) + S12V2+ ( 0 )

We also have:
V2+ ( 0 ) = Γ LV2− ( 0 )
Γ L ( S21 (1) + S22V2+ ( 0 ) )
V2+ ( 0 ) =
−
V2= ( 0 ) S21 (1) + S22V2+ ( 0 )
Solve for V2+ ( 0 )

Γ L S 21
V2+ ( 0 ) =
1 − S 22 Γ L 58
 Example (cont.)
Z0 Z0
 S11 S12 
Z in S ZL
 21 S 22 
V1+ ( 0 ) V2+ ( 0 )
V1− ( 0 ) V2− ( 0 )
Two-port device

Hence

Γin = S11 + S12V2+ ( 0 )

 Γ S 
= S11 + S12  L 21 
 1 − S 22 Γ L 

so

 Γ L S12 S 21  Z − Z0
Γin = S11 +   ΓL = L
 1 − S Γ
22 L 
Z L − Z0

59
 Transfer (T) Matrix
For cascaded 2-port networks:

1 2 1 2

A B

T Matrix:  1 − S 22 
S S 21 
 a1  T11 T12   b2  [T ] =  21 
 S11 S S 
 b  = T T   a  S S12 − 11 22 
S 21 
 1   21 22   2   21
b  (derivation omitted)
= [T ]  2   −T21 1 
 a2   T T22 
[ S ] =  22 2 
T12 T12 
T11 − 
 T22 T22 
60
 Transfer (T) Matrix (cont.)
Cascading property:

 a1A   b2A 
 A    A
= T  A (definition of T matrix)
 b1   a2 
 a1B 
= T   B 
A

 b1 
 b B

= T  T   B 
A B 2

 a2 
Conclusion:
The T matrices can
a  A
b 
B
be multiplied
     
  
1 A B 2
so that =
A
T T  
B together, just as for
b 1 a 
2
ABCD matrices.

T AB 
61
Conversion Between Parameters
(Two-Ports)

62
 Example
Derive Sij from the Z parameters for a 2 port network.
(The result is given inside row 1, column 2, of the previous table.)

S11 Calculation:

1 2
Z11 − Z 21 Z 22 − Z 21

S11 Z0 Z 21 Z0

Semi-infinite

Z in − Z 0
S11 =
Γ in1 =
Z in + Z 0
, ( Z11 − Z 21 ) +
Z in = (( Z 21 ) || ( Z 22 − Z 21 ) + Z 0  )
63
 Example (cont.)
1 2
Z11 − Z 21 Z 22 − Z 21

S11 Z0 Z 21 Z0

Semi-infinite

Z in = ( Z11 − Z 21 ) + (( Z) || ( Z 22 − Z 21 ) + Z 0  )
21

Z 21 ( Z 22 − Z 21 + Z 0 )
= ( Z11 − Z 21 ) +
Z 21 + ( ( Z 22 − Z 21 ) + Z 0 )
( Z − Z 21 )( Z 22 + Z 0 ) + Z 21 ( Z 22 − Z 21 + Z 0 )
= 11
Z 22 + Z 0
2
Z11Z 22 + Z11Z 0 − Z 21Z 22 − Z 21Z 0 + Z 21Z 22 − Z 21 + Z 21Z 0
=
Z 22 + Z 0
2
Z11Z 22 + Z11Z 0 − Z 21
=
Z 22 + Z 0

64
 Example (cont.)
1 2
Z11 − Z 21 Z 22 − Z 21

S11 Z0 Z 21 Z0

Semi-infinite

From the last slide:

Z11 ( Z 0 + Z 22 ) − Z 21
2

Z in =
Z 22 + Z 0
so

=S11
Z in − Z 0 Z in ( Z 0 + Z 22 ) − Z 0 ( Z 0 + Z 22 )
= =
(Z (Z
11 0 + Z 22 ) − Z 21
2
) − Z 0 ( Z 0 + Z 22 )
Z in + Z 0 Z in ( Z 0 + Z 22 ) + Z 0 ( Z 0 + Z 22 ) (Z (Z
11 0 + Z 22 ) − Z 21
2
) + Z 0 ( Z 0 + Z 22 )

We next simplify this.

65
 Example (cont.)
1 2
Z11 − Z 21 Z 22 − Z 21

S11 Z0 Z 21 Z0

Semi-infinite

Z11 ( Z 0 + Z 22 ) − Z 21
2
− Z 0 ( Z 0 + Z 22 )
S11 =
Z11 ( Z 0 + Z 22 ) − Z 21
2
+ Z 0 ( Z 0 + Z 22 )
2
Z11Z 0 + Z11Z 22 − Z 21 − Z 02 − Z 0 Z 22
= 2
Z11Z 0 + Z11Z 22 − Z 21 + Z 02 + Z 0 Z 22

=
( Z 0 + Z 22 )( Z11 − Z 0 ) − Z 21
2

( Z 0 + Z 22 )( Z11 + Z 0 ) − Z 212

66
 Example (cont.)
1 2
Z11 − Z 21 Z 22 − Z 21

S11 Z0 Z 21 Z0

Semi-infinite

Hence

S11 =
( Z 0 + Z 22 )( Z11 − Z 0 ) − Z 21
2

This agrees with the table.

( Z 0 + Z 22 )( Z11 + Z 0 ) − Z 212

To get S22, simply let Z11 → Z22 in the previous result.

Hence, we have:

S 22 =
( Z 0 + Z11 )( Z 22 − Z 0 ) − Z 21
2

( Z 0 + Z11 )( Z 22 + Z 0 ) − Z 212
This agrees with the table.
67
 Example (cont.)
S21 Calculation:

V1 1 Vc 2 V2
Z11 − Z 21 Z 22 − Z 21
+ +
S11 Z0 V1 ( 0 ) Z 21 V2 ( 0 ) Z0
- -

Semi-infinite

Assume V1+ ( 0 ) = 1 [V]

V1 ( 0 ) = 1 + S11

⇒ S 21= V2− ( 0 )= V2 ( 0 )

Use voltage divider equation twice to get V2(0): V1 ( 0 ) → Vc → V2 ( 0 )

68
 Example (cont.)
V1 ( 0 ) = 1 + S11
V1 1 Vc 2 V2
Z11 − Z 21 Z 22 − Z 21

S11 Z0 Z0
Z 21

Semi-infinite

Use voltage divider equation twice:

 ( Z 21 ) || ( Z 22 − Z 21 ) + Z 0  
Vc = V1 ( 0 )  
 ( Z11 − Z 21 ) + ( Z 21 ) || ( Z 22 − Z 21 ) + Z 0  
  

 Z0 
V2 ( 0 ) = Vc  
 ( Z 22 − Z 21 ) + Z 0 
69
 Example (cont.)
V1 ( 0 ) = 1 + S11
1 2
Z11 − Z 21 Z 22 − Z 21

S11 Z0 Z 21 Z0

Semi-infinite
Hence

 ( Z 21 ) || ( Z 22 − Z 21 ) + Z 0   Z0 
S 21= V2 ( 0 )= (1 + S11 )    
 ( Z11 − Z 21 ) + ( Z 21 ) || ( Z 22 − Z 21 ) + Z 0   ( Z 22 − Z 21 ) + Z 0
    

where

1 + S11 =+
1
( Z 0 + Z 22 )( Z11 − Z 0 ) − Z 21
2

( Z 0 + Z 22 )( Z11 + Z 0 ) − Z 212

70
 Example (cont.)
Our result:
 ( Z 21 ) || ( Z 22 − Z 21 ) + Z 0   Z0 
S 21= (1 + S11 )    
 ( Z11 − Z 21 ) + ( Z 21 ) || ( Z 22 − Z 21 ) + Z 0   ( Z 22 − Z 21 ) + Z 0
    

where

1 + S11 =+
1
( Z 0 + Z 22 )( Z11 − Z 0 ) − Z 21
2

( Z 0 + Z 22 )( Z11 + Z 0 ) − Z 212

After simplifying, we should get the result in the table:

(You are welcome to check it!)

2Z12 Z 0
S= S=
21 21
( Z 0 + Z 22 )( Z11 + Z 0 ) − Z 212
This is the result in the table.
71
 Example (cont.)
Different approach: Use the formula on slide 40:
−1
 1  1 
[ S ] = [ Z ] − [U ]   [U ] + [ Z ] 
 Z0  Z0 

so
(  Z  − [U ]) ([U ] +  Z  )
−1
[S ] =

where
1
 Z  ≡ [ Z ]
Z0
Hence, for a two-port, we have
−1
 S11 S12    Z11 − 1 Z12     Z11 + 1 Z12  
S  =       

S 22    Z 21 Z 22 − 1    Z 21 Z 22 + 1 
 21
72
 Example (cont.)
−1
 S11 S12    Z11 − 1 Z12     Z11 + 1 Z12  
S  =       

S 22    Z 21 Z 22 − 1    Z 21 Z 22 + 1 
 21

Hence,

 S11 S12    Z11 − 1 Z12     Z 22 + 1 − Z12  1 

  =      
 S 21 S 22    Z 21 Z 22 − 1    − Z 21 Z11 + 1 ∆ 

where

=
∆ determinant
= (Z 11 + 1)( Z 22 + 1) − Z 21Z12

This gives us directly the components of the S matrix.

73
 Example (cont.)
Examining the components, we have:

 S11 S12    Z11 − 1 Z12     Z 22 + 1 − Z12  1 

  =      
 S 21 S 22    Z 21 Z 22 − 1    − Z 21 Z11 + 1 ∆ 

1
S11
=
∆
( (
Z11 − 1)( Z 22 + 1) − Z12 Z 21 )
1
(
S12 = − ( Z11 − 1) Z12 + Z12 ( Z11 + 1)
∆
)
1
(
S21 = − ( Z 22 − 1) Z 21 + Z 21 ( Z 22 + 1)
∆
)
1
S22
=
∆
( (
Z11 + 1)( Z 22 − 1) − Z12 Z 21 )
74
 Example (cont.)

Simplifying the terms S12 and S21, we have:

1 This agrees with the table.

S11
=
∆
( (
Z11 − 1)( Z 22 + 1) − Z12 Z 21 )
1
S12 = ( 2 Z12 )
∆
1
S21 = ( 2 Z 21 )
∆
1
S22
=
∆
((Z11 + 1)( Z 22 − 1) − Z12 Z 21 )

75
 Signal-Flow Graph
This is a way to graphically represent the S parameters.
Please see the Pozar book for more details.

The wave amplitudes are

represented as nodes in the
single-flow graph.

b1 S11a1 + S12 a2
=
b2 S 21a1 + S 22 a2
=
(wave amplitudes
evaluated at zi = 0)

Rule: The value at each node is the sum of the values coming into the
node from the various other nodes.
76