Figure 2.37 For well-spread pole and zero, the impedance pauses to R0 until the second zero frequency occurs.
Figure 2.38 Conditions exist in the transformed network examined at s = sz which yield a null in the output
response.
Transfer Functions 77
likely; it is a fixed element, independent of frequency. What if node 1 goes to ground via a transformed
short circuit? Indeed, in this case, the input signal can no longer reach the output. The network that
connects node 1 to ground is made of capacitor C1 in series with its ESR rC. The series impedance of
C1 and rC is simply:
1 1 srC C1
Z 1
s r C (2.130)
sC 1 sC 1
What is the condition for which Z 1 sz1 0? Solve for the root of (2.130) when 1 srC C1 0 and
you find that the zero occurs for an angular frequency equal to:
1 1
sz1 or ωz1 (2.131)
rC C 1 rC C1
No mathematics, no complex equations, just look at the figure.
A quick example will help you understand the concept. Look at Figure 2.39 in which a current
source IT – the excitation – ac-sweeps a simple network. For s different than sz, current ^i1 exists and
produces a response ^vout across R1. When s = sz, the series connection of rC and 1=sC forms a
transformed short circuit across the current source. The response ^vout disappears because all the
excitation current IT now flows in the shunt while ^i1 is a null.
Let’s continue the exploration in Figure 2.38. Node 1 is connected to node 2 via the parallel
arrangement of C2 and R2. To block the signal propagation, it would mean that Z2 impedance
evaluated at s = sz2 becomes infinite. Is it possible? The impedance of the paralleled connection of C2
and R2 is defined as:
1 R2
Z 2
s R2 jj (2.132)
sC2 1 sR2 C 2
R2 is fixed and cannot become infinite however, if we cancel the denominator, we have another zero
for which Z2 will become infinite:
1 sR2 C2 0 (2.133)
Figure 2.39 When the transformed circuit is examined at s = sz, the current in R1 is simply zero and the response
is nulled.
78 Linear Circuit Transfer Functions
implies that
1 1
sz2 or ωz2 (2.134)
R2 C2 R2 C 2
In this intermediate step, the pole of (2.132) becomes the zero of the transfer function. Let’s finish
since we are now at node 2. In parallel with the load, we see the series connection of L3 and its ESR
labeled rL. Can this series arrangement become a transformed short circuit for s = sz3? Let’s check:
Z 3
s sL3 r L 0 (2.135)
This is a low-entropy form that we obtained without a complex analysis of the circuit. This method
works fine when the network lends itself well to this kind of interpretation: looking at Figure 2.38, you
can easily identify inductive and capacitive associations that could become either a short or an open
circuit. It is usually easy with passive networks, without controlled sources. When inspection is not
possible, you need another method.
In this expression, α1 and β1 have the dimensions of ohms to form a time constant with C1. They
respectively represent the resistance offered by the capacitor terminals when the response is a null
(^vout 0 which occurs at the zero frequency) and when the excitation is set to zero for the pole (see
Figure 2.12).
When we calculate H0, we consider the circuit in static conditions, at a 0-Hz frequency. In this
mode, a capacitor offers a high impedance path and disappears from the analysis while an inductor is
assimilated as a short circuit. Removing a capacitor from a circuit can be seen as bringing its
capacitance to 0 F. Similarly, replacing an inductor by a strap means that the inductance has dropped
to 0 H. Capacitors and inductors can thus be put in different states depending on what needs to be
analyzed. The points below describe these different states and formalize the terminology: