General Characteristics of Filters 1-25

C1

G1

Vi C2 G2 Vo

FIGURE 1.10 First-order RC network.

for k ¼ 1, 2, . . . , K, then the loading effect produced by section k þ 1 on section k can be neglected
and hence

Vo YK
H(s) ¼ ¼ Hk (s)
Vi k
1

Evidently, a highly desirable property in active filter sections is that the magnitude of the input
impedance be large and=or that of the output impedance be small since in such a case the transfer
function of the cascade structure is equal to the product of the transfer functions of the individual
sections.
An arbitrary Nth-order transfer function obtained by using the Butterworth, Bessel, Chebyshev,
inverse-Chebyshev, or elliptic approximation can be expressed as

Y
K
a2k s2 þ a1k s þ a0k
H(s) ¼ H0 (s)
k¼1
s2 þ b1k s þ b0k

where
a
10 sþa00
for odd N
H0 (s) ¼ b10 sþb00
1 for even N

The first-order transfer function H0(s) for the case of an odd-order can be readily realized using the RC
network of Figure 1.10.

1.9.3 Biquads
From the above analysis, we note that all we need to be able to realize an arbitrary transfer function is a
circuit that realizes the biquadratic transfer function

a2 s2 þ a1 s þ a0 a2 ðs þ z1 Þðs þ z2 Þ
HBQ (s) ¼ ¼ (1:26)
s2 þ b1 s þ b0 ðs þ p1 Þðs þ p2 Þ

where zeros and poles occur in complex conjugate pairs, i.e., z2 ¼ z1* and p2 ¼ p1*. Such a circuit is
commonly referred to as a biquad.
1-26 Passive, Active, and Digital Filters

After some manipulation, the transfer function in Equation 1.26 can be expressed as

s2 þ ð2Re z1 Þs þ ðRe z1 Þ2 þðIm z1 Þ2

HBQ (s) ¼ K
s2 þ ð2Re p1 Þs þ ðRe p1 Þ2 þðIm p1 Þ2
s2 þ ðvz =Qz Þs þ v2z
¼K  
s2 þ vp =Qp s þ v2p

where
K ¼ a2
vz and vp are the zero and pole frequencies, respectively
Qz and Qp are the zero and pole quality factors (or Q factors for short), respectively

The formulas for the various parameters are as follows:

qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
vz ¼ ðRe z1 Þ2 þðIm z1 Þ2
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
vp ¼ ðRe p1 Þ2 þðIm p1 Þ2
vz
Qz ¼
2 Re z1
vp
Qp ¼
2 Re p1

The zero and pole frequencies are approximately equal to the frequencies of minimum gain and
maximum gain, respectively. The zero and pole Q factors have to do with the selectivity of the filter.
A high zero Q factor results in a deep notch in the amplitude response, whereas a high pole Q factor
results in a very peaky amplitude response.
The dc gain and the gain as v ! 1 in decibels are given by
!
v2z
M0 ¼ 20 logjHBQ (0)j ¼ 20 log K 2
vp

and
M1 ¼ 20 logj HBQ ( j1)j ¼ 20 log K

respectively.

1.9.4 Types of Basic Filter Sections

Depending on the values of the transfer function coefficients, five basic types of filter sections can be
identified, namely, low-pass, high-pass, bandpass, notch (sometimes referred to as bandreject), and
allpass. These sections can serve as building blocks for the design of filters that can satisfy arbitrary
specifications. They are actually sufficient for the design of all the standard types of filters, namely,
Butterworth, Chebyshev, inverse-Chebyshev, and elliptic filters.

1.9.4.1 Low-Pass Section

In a low-pass section, we have a2 ¼ a1 ¼ 0 and a0 ¼ Kv2p . Hence, the transfer function assumes the form

a0 Kv2p
HLP (s) ¼ ¼  
s2 þ b1 s þ b0 s2 þ vp =Qp s þ v2p

(see Figure 1.11a)

General Characteristics of Filters 1-27

1.9.4.2 High-Pass Section

In a high-pass section, we have a2 ¼ K and a1 ¼ a0 ¼ 0. Hence, the transfer function assumes the form

a2 s2 Ks2
HHP (s) ¼ ¼ 2  
s2 þ b1 s þ b0 s þ vp =Qp s þ v2p

(see Figure 1.11b)

1.9.4.3 Bandpass Section

In a bandpass section, we have a1 ¼ Kvp=Qp and a2 ¼ a0 ¼ 0. Hence the transfer function assumes the
form
 
a1 s K vp =Qp s
HBP (s) ¼ 2 ¼ 2  
s þ b1 s þ b0 s þ vp =Qp þ v2p

(see Figure 1.11c)

s plane

ωp M
|HLP ( jω)|

(a)
ωM

M
|HLP ( jω)|

2 K

(b)
ωM ω

M ωp
Qp
|HBP ( jω)|

K
1 √2

(c)
ωp ω

FIGURE 1.11 Basic second-order filter sections: (a) low-pass, (b) high-pass, (c) bandpass,
(continued)
1-28 Passive, Active, and Digital Filters

|HN( jω)| |HN( jω)| |HN( jω)|

ωp
M2 M
Qp
ωz
K 2 K K
ωp
K
√2
ω2z
K K
ω2p
≈ωM ωz ω ωz ≈ωM ω ωp = ω z ω
(d)

(e)

FIGURE 1.11 (continued) (d) notch, and (e) allpass.

1.9.4.4 Notch Section

In a notch section, we have a2 ¼ K, a1 ¼ 0, and a0 ¼ Kv2p . Hence, the transfer function assumes the form

 
a2 s2 þ a0 K s2 þ v2z
HN (s) ¼ 2 ¼  
s þ b1 s þ b0 s2 þ vp =Qp s þ v2p

(see Figure 1.11d)

1.9.4.5 Allpass Section

In an allpass section, we have a2 ¼ K, a1 ¼
Kwp=Qp, and a0 ¼ Kv2p . Hence the transfer function
assumes the form
General Characteristics of Filters 1-29

h   i
a2 s2 þ a1 s þ a0 K s
vp =Qp s þ vp
2 2
HAP (s) ¼ 2 ¼ 2  
s þ b 1 s þ b0 s þ vp =Qp s þ v2p

(see Figure 1.11e)

The design of active and switched-capacitor filters is treated in some detail in Section II.

References
1. A. Antoniou, Digital Filters: Analysis, Design, and Applications, 2nd ed. New York: McGraw-Hill,
1993.
2. R. J. Schwarz and B. Friedland, Linear Systems, New York: McGraw-Hill, 1965.
3. E. Kreyszig, Advanced Engineering Mathematics, 3rd ed. New York: Wiley, 1972.
4. R. Schaumann, M. S. Ghausi, and K. R. Laker, Design of Analog Filters, Englewood Cliffs, NJ: Prentice
Hall, 1990.
5. G. Szentirmai, FILSYN—A general purpose filter synthesis program, Proc. IEEE, 65, 1443–1458,
Oct. 1977.
6. A. S. Sedra and P. O. Brackett, Filter Theory and Design: Active and Passive, Portland, OR: Matrix,
1978.
7. J. K. Skwirzynski, Design Theory and Data for Electrical Filters, London: Van Nostrand, 1965.
8. R. Saal, Handbook of Filter Design, Backnang: AEG Telefunken, 1979.
9. A. I. Zverev, Handbook of Filter Synthesis, New York: Wiley, 1967.
10. E. Chirlian, LC Filters: Design, Testing, and Manufacturing, New York: Wiley, 1983.