CHA PTER 8

Circuit Representations of
Two-Port RF/Microwave
Networks

8.1 INTRODUCTION
8.2 LOW-FREQUENCY PARAMETERS
8.3 HIGH-FREQUENCY PARAMETERS
8.4 FORMULATION OF THE S-PARAMETERS
8.5 PROPERTIES OF S-PARAMETERS
8.6 SHIFTING REFERENCE PLANES
8.7 TRANSMISSION MATRIX
8.8 GENERALIZED SCATTERING PARAMETERS
8.9 SIGNAL FLOW GRAPHS
8.10 SUMMARY

8.1 INTRODUCTION
RF/microwavedevices, circuits, and components can be classified as one-, two-, three-,
or N-port networks. A majority of circuits under analysis are two-port networks.
Therefore, we will focus primarily on two-port network characterization and will study
its representation in terms of a set of parameters that can be cast into a matrix format.
DEFINITION-A TWO-PORT NETWORK: A network that has only two access
ports, one for input or excitation and one for output or response.
The description of two-port networks from a circuit viewpoint can best be
achieved both at low and high frequencies through the use of network parameters
These parameters are discussed in the upcoming sections.

8.2 LOW-FREQUENCY PARAMETERS

"To characterize a linear network at low frequencies, several different sets of param
eters are available, where one maybe selected to fit the application to obtain optimum
results.
287
288 Chapter 8 Representations of Two-Port RF/Microwave Networks
CIrcuit

four variables of interest: V1

Voltages and currents at each port hrovide us with of four variables, but only
out of a set
21 and l2. There are six wavs of pidkne two unique parameters. Tnese are
TOur combinations (or sets) will vield nontrivial andnetwork with terminal voltages
aed , h-, and ABCD-parameters, A two-port
sets of parameters are defined as To1oWs.
and currents isshown in Figure 8.1. These four

FIGURE 8.1 Ablock diagram

of a two-port network.

Two-Port V2
Network

Port 2
Port 1

B.2.1 Impedance or Z-Parameters

(8.1a)
V1 = Z11ij + Z122
(8.1b)

Or in matrix form:

(8.2)

(8.3 a)

(8.3b)

(8.4)
|21 z2
Y-Parameters
B.2.2 Admittance or
form as:
Similarly, we can write the Y-parameters in matrix
(8.5)
where andv are defined as before and y as followS:

(8.6)
21 Yz2
 Section 8.2 Low-Frequency Parameters 289

8.2.3 Hybrid or h-Parameters

(8.7)

8.2.4 Transmission or ABCD-Parameters

(8.8)

EXAMPLE 8.1
Find the [ABCD]matrix for a series impedance element (Z) as shown in Figure 8.2.

FIGURE 8.2 A series element ig

(Example 8.1).
V2
V1

Solution:
can be written:
Using KVL and KCL, the following
VË =Vy-Ziy=Av,-Bi,

is=-h-0-i,= Cv- Di,

given by: N.
Thus, the (ABCD] matrix is
AB
|CDI

EXAMPLE 8.2
element (Y) as shown in Figure &3.
the [ABCD]matrix for ashunt
Find

Ashunt element
FIGURE 8.3
(Example 8.2).
V2
V1
290 Chapter 8 Circuit Representations of Two-Port RE/Microwave Networks

Solution:
Using KVL and KCL, the following can be written:
V|= y= V2 +0=Av- Bi,
i=Yy-i, = Cvy- Di,
Thus, the (ABCD]matrix is given by:

EXAMPLE 8.3
a shunt ele
rma the (ABCD] matrix fora circuit consisting of aseries element (Z) and
ment (Y) as shown in Figure &.4.

FIGURE 8.4 A series-shunt

circuit (Example 8.3). o+

Vi

Solution:
The [ABCD]matrix for the whole circuit, which is a cascade of a series and a shunt
element, is a multiplication of the two matrices as follows:
A B_4, B|A, B
C D) C DC, Da
Thus, the [ABCD]matrix is given by:

EXAMPLE 8.4
Find the [ABCD] matrix for a transformer as shown in Figure &.5.
Solution:
Using the transformer voltage and current rule, which states that if the voltage is
stepped down then in order to preserve the power flow the current must be propor
tionately stepped up, we have:
 Section 8.2 Low-Frequency Parameters 291

FIGURE 8.5 A transformer

n:1
(Example 8.4). i2

V1

V1= nVy=Avy- Bi,

1,
72=CVy- Di,
Thus the [ABCD] matrix is given by:
AB
1
|C D n

EXAMPLE 8.5
transmission line of length (l) and characteristic
Find the [ABCD)matrix for alossless
impedance (Z) as shown in Figure 8.6.

FIGURE 8.6 Lossless

transmission line
(Example 8.5).
V2
V1

Propagation, we
Solution:
Chapter 7, Fundamental Concepts in Wave lags
Using results obtained in has the same magnitude as the input but
voltage or current
know that the output following can be written:
eP, Thus, the
behind in phase by
jv, sinße
e ’ VË=yet= vy Cosßl +
i=-ih et =-in cosßl-ji sinße
line, we can
considered to be matched
to the transmission
Because the load end is
write:
Vy=-Zi,
292
Chapter 8 Circuit Representations of Two-Port RF/Microwave Networks

V|=(cosße)v, - (j , sinße)iz =Avy- Bin

i| =(Y,sinBl)v, -(cosBe)i = Cvy- Din
Thus, the (ABCD] matrix can be written as:
=
cosße jZ,sin pe
iY, sin pe cosBe

8.3 HIGH-FREQUENCY PARAMETERS

following consider
e toe that Z-, Y-, h-, and ABCD-parameters are based on the
ations at each of the network ports:
Net voltage (v) and net current ()
" Short and open circuit terminations
Simple observations at high RF/microwave frequencies reveal the folowing
" Shorts and open circuit terminations are difficult to implement over a broad
Tange of frequencies and, thus,cannot be used to characterize networks.
At high RF/microwave frequencies, the net voltage (or net current) is a combina
tion of two or more voltage (or current) traveling waves.
Based on these observations, the Z-, Y-, h-, and ABCD-parameters cannot be
accurately measured at these higher frequencies; therefore, we have to use the concept
of propagating or traveling waves to define the network parameters.
The network representation of a two-port network at high RF/microwave fre
quencies is called scattering parameters" (orS-parameters" for short).
When cascading networks, a variation of S-parameters called chain scattering
parameters (or T-parameters) are used to simplify the analysis.
These two types of high-frequency parameters are very popular and are prima
rily used at the high RF/microwave frequencies.

8.4 FORMULATION OF THE S-PARAMETERS

The high frequency S- and T-parameters are used to çharacterize high RF/microwave
Two-port networksXor -port networks, in general) (These parameters are based on
the concept of traveling waves and provide a complete characterization of any two
port network-under analysis or test at high RF/microwave frequencies.)
In view of the lËinearity of the electromagnetic field equations and the linearity
displaved by most microwave components and networks, the "scattered waves" (i.e..
the reflected and transmitted wave amplitudes) are linearly related to the incident
wave amplitudé} The matrixdescribing this linear relationship iscalled the "scattering
matrix," or (S1)
 CWhle the lower-frequency network
defined in terms of net (or total) voltage andDarameters(such as Z-or Y-matrices, etc.) are
currents at the ports these concepls are not
practical at high RF/microwave frequencies.where it is found that any set of
to be meaningful, must be defined in terms of a parameters,
combination of traveling waves.
(lo characterize a two-port network that has identical characteristic impedances
(Zo) at both the input and output ports. let us consider the incident and reflected volt
age waves at each port,)as shown in Figure 8.7.
FIGURE 8.7 A two-port with
incident and reflected waves at
each port. V
Two-Pot
Zo
Network V2
V

Klo define/the S-parameters accurately, we will consider a voltage phasor [Vi]

terminals of a twO-port net
incident on and avoltage phasor fVlreflected from the
worki 1,2) as shown in Figure 8.7.)) linear relationship
defined to describe the
(The scattering matrix, [S], is now matrix [V] and the refiected or transmit
between the incident voltage wave
phasor
two ports as follows:
ted wave phasor matrix [V at any of the

can write:
Or, in matrix form we
1
(8.9)

or,

where

V2

(8.10)
and
294
Chapter 8 Circuit Representations of Two-Port RE/Microwave Networks

CIhis linear relationship is expressed in terms of a ratio of two phasors that are
umbers with the magnitude of the ratio alwavs less than or equal to J Each
specific element of the |S] matrix is defined as:

Ihputreflection coefficient when

Output port is terminated in a matched
load. (8.11)
Iv; =0

S = Forward transmission coefficient when

V Outputport is terminated in a matched load. (8.12)
lvj=o
S2 = Reverse transmission coefficient when
V inputport isterminated in amatched load. (8.13)
lv;=0
TouT Output reflection coefficient when
(8.14)
2 input port is terminated in a matched load.
lvf= 0
S-parameters, as defined above, have many advantages at high RF/microwave fre
quencies that can be briefly stated as follows:
*S-parameters provide a complete characterization of a network)as seen at its
two ports.
" S-parameters make the use of short or open(as prescribed at lower frequencies)
completely unnecessary ,at higher frequencieslt is a known fact that the imped
ance of a short or an open varies with frequency which is one reason why they
are not useful for device characterization at high RF/microwave frequencies.
Furthermore, the presence of a short or open in a circuit can cause strong reflec
tions (because IIl=1), which usually lead to oscillations or damage to the tran
sistor circuitry.
"{S-parameters require the use of matched loads for termination jand because the
Toads absorb all the incident energy, the possibility of serious reflections back to
the device or source eliminated.
(S-paramneters can be converted to low-frequency parameters)as illustrated in the
next example.

EXAMPLE 8.6
Given the (ABCD] matrix for a two-port network, derive its (S] matrix (see
Solution:
Figure 8.8).
Toobtain S11, we terminate port 2 in a matched load and find the input
refection coef
ficient (rIN) as follows:
S =TN = ZIN-Z
ZNtZ
 Section 8.4 Formulation of the S-Parameters 295

FIGURE 8.8 Network for

Example 8.6. i2
Two-Port
Zo V1 Network V2 Zo
[S]. (ABCD] ...
ZIN
Where

ZÊN =Vyli
VË =Avy- Bi,
iË=Cvy- Di,
Vy=-Zi,
Substituting for v and i in terms of [ABCD], we have:
V1 Av, - Bi, AZ,- B
Ij CVy - Di, CZ, -D
Now, substituting for ZN, we can write S1Ë as:
A+BY.
o-CZ,-D
S1 = TIN
where Ais given by:
A=A + BY, + CZ,+ D
Similarly, Si2,S21, and S2 can be found as follows:
2(AD - BC)
S
S1 =
-A+ BY, -CZ, +D
Sg = Tour =

Ingeneral, using the same technique as demonstrated in Example .6, any set of
network parameters can be converted into another set of parameters, Appendix H.
Conversion Among Two-Port Network Parameters, shows the conversion relation
between the z-, y-, h-, ABCD-, and the S-parameters.
The conversion among the three transistor configurations is an important rela
tion that becomes useful in many practicaldesign situations, Appendix I,Conversion
and CC).
among the Y-Parameters of a Transistor (Three Configurations: CE, CB, common
shows the conversion relation between the y-parameters of atransistor in
other
emitter. common-base, and common-collector configurations. If parameters
296
Chapter 8 Circuit Representations of Two-Port RF/Microwave Networks

Eatameters (e.g., S-parameters) are needed, then Appendix H

can be used
effectively to convert Y- to S-parameters.

8.5 PROPERTIESOF S-PARAMETERS

properties and inter
Ine -parameters of an N-port network, in general, have certainconsidering.
Telattonships among the parameters themselves that are worth In the fol
IoWing iscussion, due to their popularity and freguent use, we limit our discussion
reciprocal or loss
OyotWO-port networks. Depending on whether the network is
less, the S-parameters will have different properties, as discussed next.
8.5.1 Reciprocal Networks
A reciprOcal network is defined to be a network that satisfies the reciprocity theorem,
which is defined as follows:
DEFINITION-RECIPROCITY THEOREM: Isatheorem sating that theinterchange
OJ electromotive force at one point(e.g.. in branch k, v) in a passive linear net
Work,with the current produced at any other pointe.g., branch m, im results n
ne same current,(in branch k, i) when the same electromotive force is applied in
the new location (branch m, vm); that is,
(8.15)

Zkm = Zmk
As observed, this theorem applies only to passive networks having inear bilat
eral impedances. Networks that satisfy this condition include all passive networks that
contain linear passive elements including resistors, capacitors, inductors, and trans
formers except independent or dependent sources, nonlinear elements, and/or active
solid-state devices such as diodes, transistors, and so on.
It can be shown that for all reciprocal networks, the [S] matrix is symmetrical:
Sn = S1 (8.16a)
Generalizing the previous equation, it can be shown that for an N-port network:
Sij= S for i*j (8.16b)
where
i=1,...N
j=1,...N

ASpecial Case: ASymmetrical Reciprocal Network. A special case of a

reciprocal network is a symmetrical network. These networks have identical size and
arrangement for corresponding electrical elements in reference to a plane or line of
symmetry.
 Section 8.5 Properties of S-Parameters 297

Due to symmetry of the network topology and by observation, the input imped
ance obtained by looking into the input port is equal to the impedance looking into
the output port. The equalityof input and output impedances leads to the
equay
input and output reflection coefficients in addition to equality of S12 and s21 as
required by the reciprocity theorcm stated earlier. Therefore, for symmetrical net
works, we can always write:
S11 =S2 (8.17 a)
S12 = S1 (8.17b)
Or,in general for any symmetrical passive N-port network, we can write:
SH= Sj
Sj= Sji (8.18b)
Where ij, and
i=1,....,N
j=1,...N.

8.5.2 Lossless Networks

(For a lossless passive network (i.e., one containing no resistive elements), the power
entering the circuit will always be equal to the power leaving the network, i.c., the
power is conserved§This condition will impose a number of restrictions on the S
parameters that give rise to the unity and zero properties as follow:

The Unity Property of [S] Matrix. (This property states that for a passive
lossless N-port network, the sum of the products of each term of any one row (or any
one column) multiplied by its own complex conjugate is unity, i.e., )
N

i-1
2SS, = 1, j=12,../ (8.19)

{whereiand j are row and column numbers respectively.)

For a two-port network, Equation 8.19 yields two equations:
(SnS, +S,,Sy =1 (8.20 a)
S2Sp+SS = 1) (8.20b)
Furthermore, if the lossless network is also reciprocal (i.e., S12 = S1), these two
follows:
equations are greatly simplified as
S2= S21 (8.21 a)
ISjl=1S2l (8.21b)
IS,+ISz1=1 (8.21 c)
 RF/Microwave
Networks
298 of Two-Port
Chapter8 Circuit
Representations

pasei.i
Matrix. This property states that for a
The Zero Property of [S] of the products of each term of any row (or any
column
lossless N-port network. the sum row (
conjugate of the corresponding terms of any other
multiplied by the complex
column) is zero:

SSk; = 0 for i#j, &ij =

1,2,.,N (8.22)
k=1
respectively.
(where i andj are row and column numbers,
simplifies into two equations:
For a two-port network, this equation
(8.23 a)
Si281 +S2Sz = 0) (8.23b)
Furthermore, if the lossless network is also reciprocal (i.e., S12 = S1), then the
previous two equationssimplify into:
S12= S21
S1Sz + S1Sg = 0 (8.24a)
(8.24b)
NOTE: Amatrix satisfying the zero and unity property is called a unitary matrix.
Analysis of Reciprocal Lossless Netvworks. From the zero and unity prop
erties of the S-matrix, the S-parameters of areciprocal lossless network are con
strained by Equations 8.20, 8.21, and 8.24 as follows:
S1 = S12 (8.25 a)
IS11l = 1Syl (8.25b)
IS;+ 1S1= 1 (8.25 c)
SiS + S, Sp = 0 (8.25d)
If we let:

and

Then, Equations 8.25c and 8.25d give:

ISz1l =(1-IS,)2
(8.26a)
+ =0
which yields:
 Properties of S-Parameters 299
Section 8.5

+e,0 je1-921)in j(®21-Op)

11+ 2= 2, - t t 2nn
or

+ For n = 0,1,2.,... (8.26b)

2

Equations 8.26a and 8.26b provide the magnitude and phase of S2 (or S12) in
terms of magnitude and phase of S and So.Therefore, from a measurement knowl
edge of S1and S2,we can completely describe and specify a reciprocal lossless two
port network. This use of S-parameters in specifying a reciprocal lossless two-port net
work shows its usefulness and versatility. The following will illustrate the concept of
S-parameters further.

EXAMPLE 8.7
What are the S-parameters of a series element (Z) as shown in Figure 8.9?

FIGURE 8.9 Circuit for

Example 8.7.
a) A series element

Zo ZiN
b) Circuit for S11-parameter calcula
tion

E,6

Solution:
Because this is a reciprocal and symmetrical network, we have:
S11 = S2,
S12= S21
So,we only need to find S11 and S1.
NOTE: This is not a lossless network because Z = R +jX has a loss component!

S1| = (8.27)

vË= 0
300 Chapter 8 RE/Microwave Networks
Circuit Representations of Two-Port
coefficient when the out-
According to Equation 8.277, S1 is the input reflection
put ismatched; that is,
S1 =TN =(ZN-Zy/(ZIN +Z),
Where ZN=Z+ Z, thus we have:
(8.28)
S11= Z/(Z+ 2Z,)
b.
(8.29)
Sy1 =
lv;=0
rrom Equation 8.29, we can see that So: is the voltage gain (or loss) when the
(E;) at port 1, the voltage
Output is matched. Thus by applving a source voltage
gain is found as follows:
I= E/(Z, +ZIN)

Because the load is matched, V2 = 0.Thus, we have:

(8.30 a)
(8.30b)
V,= Vj+V= V{1+s;) = ZylV= ZN(1 +Si)
Dividing Equation 8.30a by 8.30b, we have:
Sz1 = Vy/V =Z,(1 + S11)/ZIN S= 2ZI(Z+2Z,) (8.31)
Therefore, the whole S-matrix can be written as:
2Z,
S =
Z+22, Z+ 22% (8.32)
2Z Z
Z+22, Z+ 22
OBSERVATION: For a"series Z" network, from Equations 8.28 and 8.31, we
can see that:

S1=1-Su

EXAMPLE 8.8
What are the S-parameters ofa shunt element (Y) as shown in Figure 8.10?
Solution:
Similar to Example 8.1, this isareciprocal and symmetrical network, thus:
S11 = S2,
and
S12 = S21
so we only need to find S11 and S.
 Section 8.5 Properties of S-Parameters 301

FIGURE 8.10 Circuit for

Example 8.8.
a) A shunt element

b) Circuit for S11-parameter calcula ZiN

tion

lv, = 0
Sy =(ZN-Z,V(ZIN +Z,)
ZIN =Zo +(1/YI|Z,)
Substituting for ZIN in S11, we obtain:
S11 =-Z YI(2+Z,) (8.33)
V2
b. S1 =
V
Iv,*=0
By applying a source voltage E to port 1, we obtain:
I= E/(Zo + ZIN)
Because port 2 is terminated in a matched load (i.e., V =0), we can write:
(8.34 a)

V=V+ V= V(1 +Sit) =ZN ’ VË=ZIEll(1 +Si) (8.34b)

obtain:
Dividing Equation 8.34a by 8.34b, we
S = V/VË =(1/YIZ)(1 +Si)ZIN ’ S1 =2/(2 +Z,Y)
(8.35)
written as:
Therefore, the whole S-matrix can be
-Z,Y 2

S =
|2+ Z,Y 2+Z, (8.36)
2 -Z,Y
 Networks
RFIMicrowave

302 of Two-Port
Chapter 8
Representations

Circuit
Equations 8.33 and 8.35, we can
network, from
Fora"shunt y"
OBSERVATION:
see that:
S1 = 1 + S|

8.6 SHIFTING REFERENCE PLANES waves that are incident. on,

and phase of traveling
Therefore, the loocation of
The S-parameters relate amplitudefrom a network terminal.
transmitted through, or reflected precisely to calculate or measure the exact phase
the reference plane must be known
a
of the S-parameters.
which the reference plane at port 1has moved
Consider a two-port network
in moved a distance ,
Similarly, the reference plane at port 2 has
tance tË to port 1. at each new port (i.e.,
8.11. The phasorsfor voltage waves
to port 2' as shown in Figure
1' and 2')can now be written as:
, i=1,2
(8.37 a)
(8.37b)
V= Vje
corresponding to the reference plane
where 6;= Be; (i =1,2) is the electrical length
shift at each port.

FIGURE 8.11 A two-port

network with new reference
A
Two-Port
planes. Network
[s]
Port 1' Port1 Port 2 Port 2
(New Reference (New Reference
Plane) Plane)

Inverting Equation 8.37, we can write:

i= 1,2 (8.38 a)
i=1,2 (8.38b)
Upon substitution of Equations 8.38 in:

and further mathematical manipulation, we obtain [S'), which is the shifted

S-parameters as:
-j20, -j(8, +0,)
Sie Spe (8.39)
S1e +8,) -j28,
Spe
 Section 8.7 Transmission Matrix 303

or conversely
j20,
(8.40)
S'ze*,)

To summarize this analysis, we note

that:
S'; =S;e -j20, i=1,2 (8.41)

S'y=SjeJ0,+9,) i#j, i= 1,2 (8.42)

Equation 8.41 shows that the phase of S; is shifted by twice the electrical length
because the incident wave travels twice over this length upon reflection. On the other
hand, at port i(i= 1,2), Equation 8,42 shows that S.. (i+ i) is shifted by the sum of the
electrical lengths because the incident wave must páss through both lengths in order to
travel from one shifted port to the other.

3.7 TRANSMISSION MATRIX

The following discussion in general applies to a cascade of N-port networks. For the
sake of simplicity, however, we limit our analysis to twO-port networksonly. When cas
cading a number of two-port networks in series, amore useful network representation
is needed to facilitate the calculation of the overall network parameters.
This new representation should relate the output quantities in terms of inputquan
complete
tities. Using such a representation will enable us to obtain a description of the
network.
cascade bysimply multiplying together the matrices describing each
as the ABCDmatrix) is
/Atlow frequencies, the transmission matrix (also known
independent variables and
defined in terms of the net input voltage andcurren as the
dependent variables.
output net voltage and current as the transmission matrix T] is
(At high RF and microwave frequencies, however, the as the independent vari
waves
expressed in terms of the input incident and reflected dependent variables)
the
ables and the output incident and reflected waves as frequencies, the transmission matrix
Using the latter definition at RF/microwave multistage circuits\such as filters
formulation becomes very usefül when dealing with
or infinitely long periodic structures such as those used in circuits for
amplifiers,etc.)
traveling wave tubes, etc. scattering parameters or scattering
Fhe transmission matrix (also called chain
parameters) for a two-port network, as
shown in Figure8.12, is defined as:
transfer

FIGURE 8.12 A two-port

network.
v
ATwo-Port
Network V2
(T]
304 Chapter8 RF/Microwave Networks
Circurt Representations of Two-Port

(8.43)

derived using the above

Ihe relationship between S- and T: parameters can be
basic definition as follows:
S2
|Sz1 S1 (8.44)
S11
S12
SySz2
Sz1 Sz1
derived
ne reverse relationship expressing [S]in terms of [T]matrix can also be
with the following result:
T21 T,;12
S Si2 T.1 T2-Ti (8.45)
S21 S2 1 T12
For acascade connection of twO-port networks, as shown in Figure 8.13, the
overall Tmatrix can be obtained as follows:

(8.46a)

=
T11 12 (8.46b)

v v
Network Network
[T]
V2 V
FIGURE 8.13 Cascade of two-port networks.

But we note that:

(8.47 a)
(8.47b)
Therefore, combining Equations 8.46 and 8.47, yields:
 Generalized Scattering Parameters 305
Section 8.8

(8.48)

Thus, the total T-matrix is the multiplication of the two T-matrices:

(8.49)
[7lhor-[T][T]

8.8 GENERALIZED SCATTERING PARAMETERS

assumption that all ports have
The scattering mnatrix defined earlier was based on the
2). Even though this is the case in
the same characteristic impedances (usually Z, = 50where this may not apply andeacn
many practical situations, however, there
are cases
impedance (see Figure &.14). Thus, a need to
port has a nonidentical characteristic
generalize the scattering parameters arises.
a2
FIGURE 8.14 Normalized
incident and reflected waves.
A Two-Port
Zo2
Zoy Network
b2
b

matrix to
to modify our ordinary definition for the (S]
In this case, we have port's charac
characteristic imnpedances at each port. Taking each waves as
include the different need to define two normalized
voltage
account, we
teristic impedance into
follows: (8.50)
a, = V7Zoi , i=1,2
(8.51)
b, =V7Zois i= 1,2 ,b; repre
represents the normalized incident voltage
where "i" is the port number, a; from the port and Zo: is the character
reflected voltage wave
sents the normalized Zot is a real number for lossless lines).
port (NOTE: written as:
istic impedance at thevoltage and current at each port can now be
Thus, the total (8.52)
V+V=Zo(, + b,)
V,= (8.53)
V/Zo;-V,/2o; =(a,-b,)Zoi
1, =lf-;= port can now be
expressed in terms of
power delivered to the i
The average net different Zoi at each
port:
further concern about
a, and b;with no (8.54)
Rela-b+(a,b,-a,b;)\/2
P,= RelVv2 =
306 Networks
Chaptér 8 Circuit Representations of Two-Port
RF/Microwave

the
yields the expression for
Noting that (a,b, - a,b,)
net real power: term ispurely imaginary
(8.55)
P,= (la' -lby2
it is clearly showing that the net power
rauaton is meaningful because less the normalized
delivered to each port is equal to the normalized incident power
reflected power. terms of the normalized volt
in
w8eralized [S]matrix can nOw be defined
age waves as follows:
(8.56)
[b] =[S]a]
Wnere each element of the generalized ISl matrix is now defined as.

S11 = =IN reflection coefficient at port 1 with (8.57)

04\a2 port 2matched

transmission coefficient from port 2

(8.58)
to port 1 with port 1 matched

S1 = b transmission coefficient from port 1

la , =0 to port 2 with port 2 matched
(8.59)

Sy= b2 =IoUT reflection coefficient at port 2 with

la, =0 port 1 matched (8.60)

Clearly, thisdefinition isvery similar to the earlier one for the (S] matrix except
that V and V; are replaced by 4; and b;, respectively. Alternately, each element can
be expressed as a general equation by:
for i,j,k = 1,2 and k #j (8.61 a)

A Special Case. Consider a network, with identical characteristic impedances

at all ports, having a known (S] matrix. If transmission lines of unequal characteristic
impedances (Zo) are connected to each port, the new [S] matrix for the entire net
work with the help of Equation 8.6la, can now be written as:

(Si) new for ij,k = 1,2 and k j (8.61b)

lv=0
we can simplify Equation 8.61b to yield:

(Sy)new =(Spold for ij = 12 (8.61 c)

 Section 8.9 Signal Flow Graphs 307

Equation 8.6lc shows how the S-parameters of a network with equal characteris
tic impedances at each portcan be convertedto a network connected to transmission
lines with unequal characteristic impcdances.
In the next section,we willdiscuss the subject of signal flow graphs, whereby any
complex circuit can be analyzed in terms of a simple diagram that can yield the rela
tion between desired variables.