Download as pdf or txt
Download as pdf or txt
You are on page 1of 23

Lecture Notes

Gen EE I
- Fall 2023 -

Lesson 9
1 Generalizing Resistive Circuits
In the last lecture we have learned that the analysis of circuits containing
active elements (inductors or capacitor) can be significantly simplified is a
Laplace frequency-domain representation of such elements is introduced.
Specifically, we have established that
dv
I=C ⇐⇒ I(s) = sCV (s) (Capacitors), (1)
dt
di
V =L ⇐⇒ V (s) = sLI(s) (Inductors), (2)
dt
and
V =R×I ⇐⇒ V (s) = RI(s) (Resistors), (3)
such that we may now define the relation Z(s) , V (s)/I(s), which is then
respectively given by
1
ZC (s) = (Capacitors), (4)
sC
ZL (s) = sL (Inductors), (5)
and
ZR (s) = R (Resistors). (6)
The quantity Z(s) is known as impedance, and a direct comparison of
equations (4) and (5) against (6) elucidates that the notion of impedance
generalizes that of resistance.

Impedance generalizes resistance, in the sense that Ohm’s Law


V = R · I generalizes to V (s) = Z(s) · I(s).

Just as impedance is the analog of resistance, admittance, defined


as Y (s) , I(s)/V (s), is the analog of conductance
.

1
Applications and Examples of Impedance in Linear Circuits
Let us now revise various concepts and circuits we have learned earlier under
the prism of the generalizing notion of impedance.

• Combination of Impedances
As can be readily inferred from the fact that impedances play the
same role in the generalized Ohm’s Law as a resistance plays in its
conventional variation, the combination of impedances in series and in
parallel follow similar rules.

Impedances in series add up,


while admittances in parallel add up.

Consider the circuits shown in Figure 1(a), and suppose we are inter-
ested in determining the current i drawn from the source V .
It can be immediately recognized that this circuit is of second-order
due to the presence of a capacitor and an inductor. We have learned
how to analyze such circuits both in the DC steady-state regime (rely-
ing on the converging behavior of capacitors and inductors at t → ∞),
as well as in the transient regime (using differential equations, their
Laplace transform or the state-space method).
R2

i
V R1 C L

R3
(a) Conventional representation.

R2

i
V R1 1/sC sL

R3
(b) Impedance representation.
Figure 1: A simple RLC circuit and its impedance-based representation

2
Next, consider the corresponding circuit obtained after converting all
components into impedance equivalents, as shown in Figure 1(b). The
sought-after current can then be easily calculated as
V
i= , (7)
Zeqv
where

Zeqv = ((R2 + sL)||(1/sC) + R3 )||R1 (8)


R1 R3 LCs2
+ R1 (R2 R3 C + L)s + R1 (R2 + R3 )
= 2
.
(R1 + R3 )LCs + (L + (R1 + R3 )R2 C)s + (R1 + R2 + R3 )

As already studied in previous lectures, the full transient behavior of


the current i(t) can be obtained by finding the Laplace pair of equa-
tion (8). Interesting, however, is that one can also obtain significant
information about the behavior of i(t), immediately from equation (8)
without requiring the often cumbersome Laplace inversion.
Specifically, the instantaneous values of i(t) at t = 0 (assuming no
energy was in the circuit before) and at t → ∞ (assuming a DC input)
can be easily calculate. To illustrate, consider that for t = 0, capacitors
behave as short-circuits and inductors as open-circuits, which referring
to equations (4) and (5) imply an equivalence with making s → ∞.
Doing so in equation (8) yields
R1 R3 (R1 + R3 )V
lim Zeqv (s) = =⇒ i t=0
= . (9)
s→∞ R1 + R3 R1 R3

Likewise, for t → ∞, capacitors behave as open-circuits and inductors


as short-circuits, which are equivalent to setting s = 0 in equations (4)
and (5) such that we have
R1 (R2 + R3 ) (R1 + R2 + R3 )V
Zeqv (0) = =⇒ i t→∞
= . (10)
R1 + R2 + R3 R1 (R2 + R3 )

In other words, although the above analysis does not capture the de-
tailed rate of variation of i(t), nor the its possible oscillatory behavior,
the equations are sufficient to (roughly) establish the range of varia-
tion of the i(t), and the fact that the source current decreases from
t = 0 to t → ∞.

Homework: Show that


R1 + R3 R1 + R2 + R3
> , for all R2 > 0.
R1 R3 R1 (R2 + R3 )

3
• OpAmp Circuits
Next, consider the circuit depicted in Figure 2, which shows a cascade
of two identical circuit blocks, one of which is highlighted by a dashed
box. We have seen that the circuit isolated by the dashed box is a dif-
ferentiating amplifier, such that we may expect that the cascaded
structure should yield a double differentiating amplifier.
Let us now see how the notion of impedance helps us establish that
fact easily. Straightforwardly we have
−R −R
Va (s) = VS (s) and Vout (s) = Va (s). (11)
1/sC 1/sC

Assuming i(0) = 0, the latter equation readily implies


d2 VS (t)
Vout (s) = R2 C 2 s2 VS (s) =⇒ Vout (t) = R2 C 2 . (12)
dt2

R
1/sC R
1/sC

VS i

Va
Vout

Figure 2: A double-differentiating amplifier.

• Transformation Theorems
At this point the reader can infer that the impact of generalizing Ohm’s
Law thanks to the definition of impedance is far more reaching then
illustrated so far. Indeed, the ability to account for the impact of
capacitors and inductors algebraically, immediately implicate the gen-
eralization of all transformation theorems studied so far, from resistive
circuits to higher-order circuits in general.
For instance, quite straightforwardly, a ∆ arrangement of impedances
can be transformed into an equivalent Y arrangement, as illustrated
in Figure 3, where
Z1 Z2 Z2 Z3 Z1 Z3
Za = , Zb = , Zc = .
Z1 + Z2 + Z3 Z1 + Z2 + Z3 Z1 + Z2 + Z3
(13)

4
n1

Za
n1 Z2
1
Z

Zb
c
Z
n3 Z3 n2 n3 n2
(a) A Delta of impedances. (b) A Y of impedances.
Figure 3: Illustration of ∆-Y transformation with impedances.

Similarly, a voltage source in series with a given impedance can be


transformed into a current source in parallel with the same impedance,
as illustrated in Figure 4

n1 n1
ZS

VS iS ZS

n2 n2
(a) Voltage-impedance. (b) Current-impedance.
Figure 4: Illustration of source transformation with impedances.

Finally, Thevénin’s and Norton’s theorems also have equivalents in


impedance circuits, such that the effect of any linear circuit with re-
sistors, capacitors and inductors over any give pair of terminals, can
be modeled by arrangements similar to those shown respectively in
Figures 4(a) and 4(b).

Clearly a significant gain in understanding linear circuits results from the


simplifying effect that the notion of impedance brings. That, combined
with the notion of circuit equivalent by means of Thevénin and Norton,
enables us to represent an linear circuits of arbitrary order by means
of a simple structure with a pair of terminals.

5
And given that each individual electric signal requires two terminals to be
meaningfully represented, a single pair of terminals is associated with a
single function, that is, an input or output. It is therefore common to refer
to a terminal and its function as a port of the circuit.
So far, this single-port representation of circuits has been sufficient, since
all networks analyzed until now are somewhat “unidirectional ”. A great ex-
ample are the OpAmp circuits briefly revised above, which are characterized
by a clear output/input relationship.
In the next session, we shall generalize our perspective of circuits and
their simplification to a two-port model, which is characterized by an
inherent bi-directivity.

Two-port Impedance Networks


Consider the two-port representation of a given circuit, in which both ports
are voltage-driven. We seek a characterization of the entire circuit in
terms of an equivalent voltage/current relationship. In other words, we
wish to write of the model an expression similar to that given by Ohm’s
Law, except that in matrix form, i.e,
V = Z · I, (14)
where V , [V1 V2 ]T and I , [I1 I2 ]T .

A two-port network impedance model


generalizes Ohm’s Impedance Law to matrix form.

As indicated by the notation, the matrix Z has the unit of impedance,


only in 2 × 2 dimensions, such that the elements in Z are referred to as
the impedance parameters of the two-port network. The question that
remains, however, is how to determine such values, which we shall study in
the sequel.

I1 Two-port I2
V1 Network V2

Figure 5: Voltage-Driven two-port network model of a circuit.


• Impedance Parameters
Let us expand equation (14), which yields
     
V1 Z11 Z12 I1
= · , (15)
V2 Z21 Z22 I2

6
or explicitly

V1 = Z11 I1 + Z12 I2 , (16a)


V2 = Z21 I1 + Z22 I2 . (16b)

It can be readily seen respectively form equations (16a) and (16b) that
V1 V1
Z11 = and Z12 = , (17)
I1 I2 =0 I2 I1 =0

while
V2 V2
Z21 = and Z22 = . (18)
I1 I2 =0 I2 I1 =0

Maintaining the convention followed so far, with the input on the


lefthand side port and the output on the righthand side port, we can
immediately recognize that the impedances Z11 and Z22 are the open-
circuit input and open-circuit output impedances, respectively.
In turn, since the other impedances involve both input and output
quantities, these are referred to as input-to-output and output-to-input
transfer impedances, respectively.

Impedance Parameters of Two-port Networks

V1
Open-circuit input: Z11 = ;
I1 I2 =0
V2
Open-circuit output: Z22 = ;
I2 I1 =0
V1
Open-circuit I/O transfer: Z12 = ;
I2 I1 =0
V2
Open-circuit O/I transfer: Z21 = .
I1 I2 =0

• Dependent-source Model for Impedance Two-port Networks


Given the equations above describing the impedance input/output re-
lationships found in a two-port network, let us consider whether a
slightly less abstract model can be build, using something familiar
such as dependent sources.
To begin with, consider then the dependent-source model of an impedance
two-port network depicted in Figure 6, where the constants α and β,
the impedances Z1 and Z2 , as well as the electric quantities X1 and
X2 are to be determined.
Let us first take the input port. There are four possibilities for the
dependent source αX2 , namely:

7
Z1 Z2
I1 I2
V1 V2
↵X2 X1

Figure 6: A generic dependent-source model of an impedance two-port net-


work, with internal parameters undetermined.

– Voltage-voltage Source
In this case, X2 = V2 and the output of the source would also be
the voltage αV2 , such that we would have
V1 − αV2 Z1 V1
I1 = =⇒ Z11 = =⇒ Z11 = indet.
Z1 V1 − αV2 I2 =0
(19)
– Voltage-current Source
In this case, X2 = V2 but the output of the source would be the
current αV2 , such that we would have
V1
I1 = αV2 =⇒ Z11 = =⇒ Z11 = indeterminate.
αV2 I2 =0
(20)
– Current-current Source
In this case, X2 = I2 and the output of the source would be the
current αI2 , such that we would have
V1
I1 = αI2 = 0 =⇒ Z11 = =⇒ Z11 → ∞. (21)
αI2 I2 =0

– Current-voltage Source
Finally, in this case X2 = I2 and the output of the source would
be the voltage αI2 , such that we would have
V1 − αI2 Z1 V 1
I1 = =⇒ Z11 = =⇒ Z11 = Z1 . (22)
Z1 V1 − αI2 I2 =0

8
Z11 Z22
I1 I2
V1 V2
↵I2 I1

Figure 7: A current-to-voltage dependent-source model of an impedance


two-port network, with gains undetermined.

From all the above, clearly the most suitable model is the latter.
Following the same arguments, and due to symmetry, it is evident
that dependent source at the output port is also a current-dependent
voltage source, with output voltage βI1 such that
V2 − βI1 Z2 V2
I2 = =⇒ Z22 = =⇒ Z22 = Z2 . (23)
Z2 V2 − αI1 I1 =0

It is left for us to determine the constant α and β. To this end,


reconsider the circuit of Figure 6, now with the impedances and types
of dependent sources determined, as shown in Figure 7.
Next, consider the input-to-output transfer impedance as given in
equation (17), which requires I1 = 0, such that

V1 − αI2 V1
I1 = =0 =⇒ α= . (24)
Z11 I2

In other words, we have directly


V1
Z12 = = α. (25)
I2 I1 =0

Proceeding similarly we also easily find


V2
Z21 = = β. (26)
I1 I2 =0

Using these results, we finally obtain the complete dependent-source-


based model of a general impedance two-port network, as depicted in
Figure 8.

9
Z11 Z22
I1 I2
V1 V2
Z12 I2 Z21 I1

Figure 8: A complete dependent-source model of an impedance two-port


network.

Notice that the model shown in Figure 8 is essentially a coupling


circuit, in the sense that both sides of the circuit are mostly inde-
pendent, expect that a voltage appearing in part of one is coupled
to the current form the other, and vice-versa. We show return to this
notion of coupling in the near future.

An impedance two-port network can be understood as a circuit


with coupled current-voltage dependent sources.

• T-equivalent Model for Impedance Two-port Networks


Let us now consider an interesting special case of impedance two-port
networks, in which the transfer impedances are identical, that is
Z12 = Z21 .
Two-port networks with this particular property are referred to as
reciprocal.
We shall see that reciprocal impedance two-port networks can
be represented also without dependent sources.
To this end, consider the circuit of Figure 8 and let us make Z12 =
Z21 = ZT . Then, from both meshes in the circuit we obtain

V1 − Z11 I1 − ZT I2 = 0 and V2 − Z22 I2 − ZT I1 = 0. (27)

Combining both equations and rewriting we have

VX = V1 − (Z11 − ZT )I1 = V2 − (Z22 − ZT )I2 . (28)

10
VX
Z11 ZT Z22 ZT

I1 I2
V1 ZX V2

Figure 9: A T model of an impedance two-port network.

From the latter equation we can draw the impedance-only model of


the two-port network shown in Figure 9, where the impedance ZX still
to be determined can be calculated directly by
VX V1 V2
= I1 +I2 =⇒ ZX = −(Z11 −ZT ) or ZX = −(Z22 −ZT ).
ZX I1 I2
(29)
Choosing without loss of generality the first of the latter two equivalent
equations, and noticing that it does not depend on I2 , we can replace
V1
= Z11 , to finally obtain
I1
ZX = ZT . (30)

Two-port Admittance Networks

In some cases, the input/output port relationship of a two-port net-


work is more convenient (or even only possible) to be given in terms
of admittance, as opposed to impedance, thus generalizing the re-
ciprocal of Ohm’s Impedance Law to matrix form, namely
     
I1 Y11 Y12 V1
I = Y · V or = · . (31)
I2 Y21 Y22 V2

By analogy with equations (17) and (18), we readily obtain the short-
circuit input and short-circuit output impedances
I1 I1
Y11 = and Y12 = , (32)
V1 V2 =0 V2 V1 =0

as well as the input-to-output and output-to-input transfer impedances


I2 I2
Y21 = and Y22 = . (33)
V1 V2 =0 V2 V1 =0

11
The latter can be summarized as follows.

Admittance Parameters of Two-port Networks

I1
Short-circuit input: Y11 = ;
V1 V2 =0
I2
Short-circuit output: Y22 = ;
V2 V1 =0
I1
Short-circuit I/O transfer: Y12 = ;
V2 V1 =0
I2
Short-circuit O/I transfer: Y21 = .
V1 V2 =0

We can obtain a dependent-source model for an admittance two-port


network by simply obtaining the “dual” of the circuit shown in Figure
8, namely,

I1 I2
V1 Y11 Y22 V2
Y12 V2 Y21 V1

Figure 10: A complete dependent-source model of an admittance two-port


network.

Analogously to the case of impedance two-port networks, the special


case of reciprocal networks allows for a simplified version of the lat-
ter model in which only admittances are required. The topology of
such model, however is in a Π-shape, and the circuit specifically can
be obtained following steps “dual” to those described above in the
derivation of the T-model for impedance two-port networks.

Homework: Obtain an admittance-only equivalent of the two-port


network model shown in Figure 10, for the reciprocal case where
YΠ , −Y12 = −Y21 .
Hint: Assume that the topology of the network is Π-shaped.

12
Two-port Hybrid Networks
While the Z-parameter (impedance) and Y-parameter (admittance) two-
port networks are direct generalization of Ohm’s Impedance Law v = Zi
and its reciprocal i = Y v to a matrix form, they also motivate more ab-
stract models to describe the relationship between input an output electric
quantities.
In general, any combinations of the quantities V1 , I1 , V2 and I2 , vectors of
two at a time, and any relationship between the latter can be considered.
For instance, grouping V = [V1 V2 ]T and I = [I1 I2 ]T , and focusing on
the relations V = ZI and I = YI, lead to the two-port network models
described above.
Grouping combinations of input/output currents and voltages, however,
leeds to varied relationships, which can be equivalent to impedances, ad-
mittances, or even unit-less “gains”, which therefore are generally referred
to as hybrid parameters.
In what follows, we briefly consider two of such variations of two-port net-
works that find application more often.

• H-parameter Two-port Networks


Consider a two-port network with the following input-to-output model:
     
V1 H11 H12 I1
= · . (34)
I2 H21 H22 V2

In this case we obviously have:

H-Parameters of Two-port Networks

V1
Short-circuit input impedance: H11 = ;
I1 V2 =0
I2
Open-circuit output admittance: H22 = ;
V2 I1 =0
V1
Open-circuit reverse voltage gain: H12 = ;
V2 I1 =0
I2
Short-circuit forward current gain: H21 = .
I1 V2 =0

• G-parameter Two-port Networks


Next, consider a two-port network with the following input-to-output
model:      
I1 G11 G12 V1
= · . (35)
V2 G21 G22 I2

13
In this case we obviously have:

G-Parameters of Two-port Networks

I1
Open-circuit input admittance: G11 = ;
V1 I2 =0
V2
Short-circuit output impedance: G22 = ;
I2 V1 =0
I1
Short-circuit reverse current gain: G12 = ;
I2 V1 =0
V2
Open-circuit forward voltage gain: G21 = .
V1 I2 =0

Two-port Transmission Networks


Similarly to hybrid two-port networks, if the relationship between currents
and voltages are mixed, but given exclusively in terms of input-to-output or
output-to-input direction, the corresponding network model can be seen as
to serve a transmission function.
Such networks are therefore characterized by transmission parameters,
as described below.

• ABCD-parameter Two-port Networks


Consider a two-port network with the following input-to-output model:
     
V1 A B V2
= · . (36)
I1 C D −I2

In this case we obviously have:

ABCD-Parameters of Two-port Networks

V1
Open-circuit reverse voltage gain: A = ;
V2 I2 =0
I1
Negative short-circuit reverse current gain: D = − ;
I2 V2 =0
V1
Negative short-circuit reverse transfer impedance: B = − ;
I2 V2 =0
I1
Open-circuit reverse transfer admittance: C = .
V2 I2 =0

14
• abcd-parameter Two-port Networks
Consider a two-port network with the following input-to-output model:
     
V2 a b V1
= · . (37)
I2 c d −I1

In this case we obviously have:

abcd-Parameters of Two-port Networks

V2
Open-circuit forward voltage gain: a = ;
V1 I1 =0
I2
Negative short-circuit forward current gain: d = − ;
I1 V1 =0
V2
Negative short-circuit forward transfer impedance: b = − ;
I1 V1 =0
I2
Open-circuit forward transfer admittance: c = .
V1 I1 =0

Relationship between Two-port Network Parameters


Obviously, since all the two-port network models described above can (in
principle1 ) be applied to any given circuit, offering different relationships
amongst the same quantities (namely, V1 , V2 , I1 and I2 ), it is to be expected
that we can transform any given model into another.

• Impedance-to-Admittance Two-port Network Transformation


To illustrate the latter, consider the example of converting an impedance
two-port network model into the corresponding admittance model.
Specifically, referring to equations (14) and (31), we with to transform

V =Z·I =⇒ I = Y · V, (38)

which reduces to finding Y, given that Z is known.


Now, since Z is a square matrix, it is in principle invertible. Assume
therefore that there exists a matrix Z−1 such that Z−1 ·Z = Z·Z−1 = I,
where I is the identity matrix. From the latter, we obviously have
Y = Z−1 , so that in order to find the elements of Y we must solve
     
Z11 Z12 Y11 Y12 1 0
· = , (39)
Z21 Z22 Y21 Y22 0 1
1
We shall soon see that some models cannot be applied to certain circuits.

15
or equivalently
     
Z11 Z21 0 0 Y11 1
 Z12 Z22 0 0   Y12   0 
·
  Y21  =  0
   . (40)
 0 0 Z11 Z21 
0 0 Z12 Z22 Y22 1

The linear system described by equation (40) can be easily solved by


Gaussian elimination, as seen multiple times before. Proceeding
that way yields the solution
 
Z22
−Z12 Z21
 Z11 Z22
   
1 0 0 0 Y11
−Z12

 0 1 0 0   Y12   Z Z −Z Z

 ·  =  11 22 12 21 

. (41)
 0 0 1 0   Y21   −Z21 
 Z11 Z22 −Z12 Z21 
0 0 0 1 Y22 Z11
Z11 Z22 −Z12 Z21

From the latter result, we may then write


 
−1 1 Z22 −Z12
Y=Z = . (42)
Z11 Z22 − Z12 Z21 −Z21 Z11

Notice how equation (42) generalizes the


reciprocal relation Y = 1/Z from scalar to matrices!

The fact that the denominator of all results on the righthand-side of


equation (41) are identical, as made more evident in equation (42),
should call the attention of the reader. In fact, this is no coincidence,
but rather a reflection of a general method to calculate the inverse
of (invertible) matrices. Namely, for any square matrix M, its inverse
is given by
adj(M)
M−1 = , (43)
det(M)
where adj(M) and det(M) respectively denote the adjoint and the
determinant of the matrix M .
The adjoint and determinant operations are defined for square ma-
trices M of arbitrary size, but for the particular case of 2×2 matrices
these quantities are particularly simple and can be stated verbally as
follows.

The adjoint of a 2 × 2 matrix is given by the


transpose of its diagonal, minus its anti-diagonal.

16
To exemplify:
     
Z22 0 0 Z12 Z22 −Z12
adj(Z) , − = . (44)
0 Z11 Z21 0 −Z21 Z11

The determinant of a 2 × 2 matrix is given by the


product of its diagonal, minus the
product of its anti-diagonal.

To exemplify:
det(Z) , Z11 Z22 − Z12 Z21 . (45)

From the discussion above we can immediately conclude that an impedance


two-port network such that det(Z) = 0 cannot be transformed into a
corresponding admittance equivalent!

• Admittance-to-Impedance Two-port Network Transformation


Straightforwardly, it is obvious that an admittance two-port network
can be converted to an impedance equivalent in a manner much sim-
ilar to the reverse case illustrated above. Specifically, provided that
det(Y) 6= 0 we obviously have
 
−1 adj(Y) 1 Y22 −Y12
Z=Y = = . (46)
det(Y) Y11 Y22 − Y12 Y21 −Y21 Y11

• H-G and ABCD-abcd Two-port Network Transformations


Referring to equations (34) and (35), we have

        H = G−1
V1 I1 I1 V1 and
= H· and = G· =⇒
I2 V2 V2 I2
G = H−1 .
(47)
Likewise, from equations (36) and (37) we have

        T = F−1
V1 V2 V2 V1 and
= T· and = F· =⇒
I1 −I2 I2 −I1
F = T−1 ,
(48)
where for convenience we have implicitly defined
   
A B a b
T, and F , . (49)
C D c d

17
It follows that
 
−1 adj(G) 1 G22 −G12
H=G = = , (50)
det(G) G11 G22 − G12 G21 −G21 G11
 
−1 adj(H) 1 H22 −H12
G=H = = , (51)
det(H) H11 H22 − H12 H21 −H21 H11
 
adj(F) 1 D −B
T = F−1 = = , (52)
det(F) AD − BC −C A
and finally
 
adj(T) 1 d −b
F = T−1 = = , (53)
det(T) ad − bc −c a

where all the above transformations are conditioned on det(H) 6= 0,


det(G) 6= 0, det(T) 6= 0 and det(F) 6= 0, respectively.

• Transformation of Non-reciprocal Two-port Networks


All the transformation of two-port networks described above have in
common the fact that each pair of models have a reciprocal relation-
ship between one another, as evidenced by the inverse relationship
between corresponding parameter matrices.
One can also, however, transform between two-port network models
that do not have such a reciprocal relationship. In this case, transfor-
mation is slightly less trivial, but still very easy to calculate by simple
association of the corresponding equations describing both two-port
models.
To illustrate, let us consider the transformation of an admittance
network to a forward transmission (abcd) network. To this end,
consider equations (31) and (37) side-by-side
           
I1 Y11 Y12 V1 V2 a b V1
= · and = · ,
I2 Y21 Y22 V2 I2 c d −I1
(54)
or explicitly
I1 = Y11 V1 + Y12 V2 V2 = aV1 − bI1
and (55)
I2 = Y21 V1 + Y22 V2 I2 = cV1 − dI1 .

Comparing these equations, it is evident that we must simply rearrange


the first two sets of equations so as to describe V2 and I2 as a function
of V1 and −I1 , such that the coefficients a, b, c and d can be found by
association.

18
The first (top-left) equation in (55) yields readily

−Y11 1
I1 = Y11 V1 + Y12 V2 =⇒ V2 = V1 + I1 , (56)
Y12 Y12
−Y11 −1
from which we easily obtain a = Y12 and b = Y12 .
Substituting the latter result into the second (bottom-left) equation
in (55) we obtain

Y22 − det(Y) Y22


I2 = Y21 V1 + (−Y11 V1 + I1 ) = V1 + I1 , (57)
Y12 Y12 Y12
− det(Y) −Y22
which finally yields c = Y12 and d = Y12 .
In summary, we have
" #
−1 Y11 1
F= . (58)
Y12 det(Y) Y22

Homework: Find the transformations between:

– Impedance parameters Z to reverse transmit parameters T;

– G-hybrid parameters G to forward transmit parameters F;

– Reverse transmit parameters T to H-hybrid parameters H.

Interconnection of Two-Networks
As mentioned earlier, two-port networks generalizes Ohm’s Impedance Law
to matrices. Alternatively, it can be said that a two-port network is a gen-
eralization of two-terminal circuit elements to two-terminal circuit
blocks.
As seen above, an immediate consequence of this generalization is that there
are six relationships amongst the various vectors containing voltages and
currents at the input and output of the two-port network. Notice that this
indeed can be seen as a generalization of two relationships between the
scalar electric quantities voltage and current – namely, voltage-to-current
and current-to-voltage – one can consider in conventional two-terminal
circuit elements such as resistors, capacitors and inductors.
Another consequence of the element-to-block generalization implied by the
two-port network model, however, is that two-port networks can be con-
nected not only in two, but three different ways. In the sequel, we shall
study these distinct types of connection between two-port networks.

19
• Two-port Networks in Series
Drawing a parallel with Kirchhoff’s Voltage Law, two-port networks
connected in series should be such that their voltages add up, and
such that the same current flow through both devices, where
both notions here are to be understood at input and output sepa-
rately. Mathematically we must therefore have

V1 = V1a + V1b and I1 = I1a = I1b , (59a)

and
V2 = V2a + V2b and I2 = I2a = I2b . (59b)

Equation (59) suggests the arrangement shown below in Figure 11.


Referring to equation (15), we have
         
V1a Z11a Z12a I1a V1b Z11b Z12b I1b
= · and = · ,
V2a Z21a Z22a I2a V2b Z21b Z22b I2b
(60)
which under the conditions of equation (59) readily yields
          
V1 V1a V1b Z11a Z12a Z11b Z12b I1
= + = + · ,
V2 V2a V2b Z21a Z22a Z21b Z22b I2
(61)
or simply
V = (Za + Zb ) · I =⇒ Z = Za + Zb . (62)

The impedance matrices of


two-port networks in series add up.

I1a I2a
V1a V2a
I1 I2

V1 V2

I1b I2b
V1b V2b

Figure 11: Two two-port networks connected in series.

20
• Two-port Networks in Parallel
Proceeding analogously to the above, let us now draw a parallel with
Kirchhoff’s Current Law, by which two-port networks connected in
parallel should be such that their currents add up, and such that
the both devices are subjected to the same voltage, again con-
sidering input and output separately. Mathematically we must
therefore have
I1 = I1a + I1b and V1 = V1a = V1b , (63a)
and
I2 = I2a + I2b and V2 = V2a = V2b . (63b)

Equation (63) implies the arrangement shown below in Figure 12. Re-
ferring to equation (31), we have
         
I1a Y11a Y12a V I1b Y11b Y12b V
= · 1a and = · 1b ,
I2a Y21a Y22a V2a I2b Y21b Y22b V2b
(64)
which under the conditions of equation (63) readily yields
           
I1 I1a I1b Y11a Y12a Y11b Y12b V1
= + = + · ,
I2 I2a I2b Y21a Y22a Y21b Y22b V2
(65)
or simply
I = (Ya + Yb ) · V =⇒ Y = Ya + Yb . (66)

The admittance matrices of


two-port networks in parallel add up.

I1 I1a I2a I2
V1a V2a

V1 V2

I1b I2b
V1b V2b

Figure 12: Two two-port networks connected in parallel.

21
• Two-port Networks in Cascade
Finally, let us consider another possible arrangement of two-port net-
works in which the output port of one is directly connected to the
input port of the other, that is, such that
V1 = V1a , V2a = V1b , and V2b = V2 , (67a)
and
I1 = I1a , I2a = −I1b , and I2b = I2 . (67b)
The diagram of this connection is shown in Figure 13 below. Despite
the pictorial resemblance of a serial connection between conventional
elements, by force of the mathematical relationship between the volt-
ages and currents shown above, which differ from those of the serial
connection studied earlier. This kind of arrangement is therefore fun-
damentally different from the serial connection and is referred to
as a cascade.
Invoking equation (36), we have
         
V1a Aa B a V2a V1b Ab Bb V2b
= · , and = · ,
I1a Ca Da −I2a I1b Cb D b −I2b
(68)
where, by force of equation (67)
           
V2a V1b V1 V1a V2b V2
= , as well as = and = .
−I2a I1b I1 I1a I2b I2
(69)
Combining these expressions yields
       
V1 Aa Ba Ab B b V2
= · · , (70)
I1 Ca D a Cb D b −I2
or simply    
Aa B a Ab Bb
T= · . (71)
Ca D a Cb D b
Due to symmetry, a similar relationship is found for abcd-parameters.

The transmission matrices of


two-port networks in cascade multiply.

I1 = I1a I2a = I1b I2b = I2


V1 = V1a V2a = V1b V2b = V2

Figure 13: Two two-port networks connected in cascade.


22
Mandatory Reading
- Alexander and Sadiku [1]: Ch. 9† (Sec. 9.5 to 9.7) and Ch. 18
- Irwin and Nelms [2]: Ch. 8† (Sec. 8.5) and Chapter 16
• Laplace Impedance and their Combinations
• Kirschhoff’s Laws with Impedance
• OpAmp Circuits with Impedance
• Impedance Two-port Networks
• Admittance Two-port Networks
• Transmission Two-port Networks
• Parameter Conversion of Two-port Networks
• Connections of Two-port Networks
• Examples in Learning Assessment sections
• Problems

† Replace jω by s in all equations. We will study the notion of


frequency under the Fourier Transform in the second semester.

References
[1] C. K. Alexander and M. M. O. Sadiku, Fundamentals of Electric Circuits,
3rd ed. Mc-Graw-Hill, 2007.

[2] J. D. Irwin and R. M. Nelms, Basic Engineering Circuit Analysis,


10th ed. John Wiley & Sons, 2011.

23

You might also like