Jomo Kenyatta University of Agriculture and Technology

BSc Electrical and Electronic Engineering

and

BSc Electrical and Computer Engineering

EEE 2405: ANALOGUE FILTERS

February 1, 2015
EEE 2405: ANALOGUE FILTERS
Prerequisites
SMA 2271 Ordinary Differential Equations
EEE 2302 Analogue Electronics III
EEE 2303 Circuit and Network theory III

Purpose
The aim of this course is to enable the student to;

1. understand electrical filter technology

2. understand passive and active filters design

Learning Outcomes
At the end of this course, the student should be able to;

1. design passive and active filters

Course Outline
Analogue filters:
Survey of electrical filter technology, Approximation theory; ideal response, least square method.
Butterworth, Chebychev, Inverse chebychev, Guillenim algorithm, Thompson and Gaussian approx-
imation. Bode sensitivity, Magnitude of sensitivity in frequency domain, sensitivity comparison of
typical circuits.

Passive filters:
Outline of image parameter synthesis, One port and Two port synthesis of LC, RC and RL network,
ladder development with termination duality, transformations, pre-distortion and denormalisation.

Active filters:
Controlled sources, Operational amplifier as a controlled source and converter realization of second
order section, inductor replacement, effects of component imperfections, dynamic range, generalised
immitance (GIC)

Teaching Methodology
2 hour lectures and 1 hour tutorial per week, and at least five 3-hour laboratory sessions per semester
organized on a rotational basis.

Instructional Materials/Equipment
1. Electronics Lab

2. Overhead projector

i
Prescribed Text Books
1. Kendall L. Su, Kendall Ling-chiao Su, Inc ebrary (2002), Analog Filters, Springer, 2nd ed.

2. L. D. Paarmann (2001), Design and analysis of analog filters: a signal processing perspective,
Springer, illustrated Ed.

References
1. Terry L. E. Henderson (1999), Design of Analogue Filters, Deakin University.

2. M. E. Van Valkenburg (1995), Analog Filter Design, Oxford University Press, illustrated Ed.

ii
Analog Filters Lecture Notes by A. M. Muhia

1 ANALOG FILTERS

1.1 Introduction
An electrical filter is a system that can be used to modify or manipulate the frequency spectrum
of an electrical signal according to some prescribed requirements e.g amplify, attenuate, reject or
isolate a specific frequency component.

Circuit Design
Consider the circuit below

Figure 1: Two port network

If the circuit is operating in the sinusoidal steady state then the two voltages may be represented
as

V1 = |V1 | ∠θ1
V2 = |V2 | ∠θ2 (1)

The transfer function is defined as

output−quantity
T ransf er−f unction = input−quantity

The magnitude of the transfer function is given by

V2
H = (2)
V1
And the phase by
θ = θ2 − θ 1 (3)
Keeping V1 constant and setting θ1 = 0 to the reference phase, the variation of V2 and θ2 with
frequency constitute the frequency response of the circuit. If the circuit is known and the form of
the input, then the output can be determined. This is called circuit analysis.
If the input is known together with the output or the ratio H and θ2 can be determined, this is
called circuit design or synthesis
The concern of this study is that with prescribed variation of the magnitude, phase or related
quantities as a function of frequency, filters will be designed.

1
Analog Filters Lecture Notes by A. M. Muhia

1.2 Types of Filters

Filters are classified according to the function they are to perform in terms of range of frequencies
as

1. Low pass Filter

Its band extends from ω = 0 to ω = ωc 1
The ideal and practical responses of a lowpass filter are as shown in figure 2.

(a) Ideal low pass filter (b) Practical low pass filter

Figure 2: Lowpass filter response

2. High pass filter The frequency range from 0 to ωc is a stop band and from ωc to ∞ is pass
band. The ideal and practical responses of a highpass filter are as shown in figure 3

3. Band pass filter

In this case the frequency range from ω1 to ω2 are passed while all other frequencies outside
this band are stopped. The ideal and practical responses of a bandpass filter are as shown in
figure 4.

4. Band stop filter

The frequencies ω1 to ω2 are stopped while all other frequencies outside this band are passed.
The ideal and practical responses of a bandstop filter are as shown in figure 5.

It is impossible to accomplish ideal filters in which there is abrupt change from pass to stop and
stop to pass band.
To accommodate the realistic filter characteristics, the characteristic required are specified by dif-
ferent definitions of pass band and stop band

ˆ A pass band is one in which the attenuation is always less than a value designated by αmax

ˆ A stop band is one in which the attenuation is greater than a value designated as αmin

ˆ The bands between stop band and pass band are called transition bands.

2
Analog Filters Lecture Notes by A. M. Muhia

(a) Ideal highpass filter (b) Practical highpass filter

Figure 3: Highpass filter response

(a) Ideal bandpass filter (b) Practical bandpass filter

Figure 4: Bandpass filter response

? In the diagram, the pass band extends from ω0 to ωp

? The range of frequencies between ωp and ωs is the transition band

? All frequencies greater than ωs constitute the stop band

? The design problem then becomes ; Given the quantities αmax , αmin , ωp , and ωs determine
the attenuation specification that satisfies the four requirements.
1
ωc is the cutoff frequency of the filter

3
Analog Filters Lecture Notes by A. M. Muhia

(a) Ideal bandstop filter (b) Practical bandstop filter

Figure 5: Bandstop filter response

1.3 Filter Design

Filter design is based on the accurate selection of an appropriate input/output relationship which
will satisfy the required specification of the filter.

Filter Transfer function

A suitable filter transfer function must satisfy the following filter characteristics for it to be imple-
mentable
1. Time Invariant
Since the designed analogue filter will be implemented with fixed value resistors, capacitors,
inductors, etc, then the implementation will be time invariant. the transfer function H (s)
must also be time invariant.

2. Causal
The filter obtained from the function has its output dependent on the present input and past
inputs only.

3. Real
The impulse response h (t) of an analogue filter made of resistors, capacitors, inductors, etc is
inherently real. From the properties of Fourier transform we have

Hjω = H ∗ (−jω)

Re H (jω) = Re H (−jω)
and

Im H (jω) = Im H (−jω)

4
Analog Filters Lecture Notes by A. M. Muhia

(a) Lowpass filter (b) Highpass filter

(c) Bandpass filter (d) Bandstop filter

Figure 6: Filter specifications in terms of attenuation

hence

H (jω) = H (−jω)
and
∠H (jω) = −∠H (−jω)

4. Rational transfer function

It is required that H (s) be a rational transfer function. This will ensure that the filter will be
realizable in a common network topology. A rational polynomial is one which can be presented
by a polynomial in s over a polynomial in s
PM
k=0 bk sk
H(s) = PN (1)
k
k=0 ak s

5
Analog Filters Lecture Notes by A. M. Muhia

where N > M

5. Real transfer function coefficient

When equation (4) is expanded as a product of the 1st and 2nd order terms it becomes

k M
Q 1 Q M2 2

k=1 (s + σk ) k=1 s + αk s + βk
H (s) = QN1 QN2 (2)
2
k=1 (s + µk ) k=1 (s + εk s + λk )

where

2M2 + M1 = M

and

2N2 + N1 = N

bm
K=
an
Equation (5) can be expanded such that

N1  N2 
ρk s2 + k s + σk
 
X τk s + σk X
H (s) = + (3)
s + µk s2 + k s + λk
k=1 k=1

If any of the first order coefficients in equation (5) are complex then there will be complex
coefficients in equation (6). Therefore the inverse Laplace transform of H (s) denoted by µ (t)
is the sum of the inverse Laplace transform of the individual terms in (6). Hence H (s) must
be rational transfer functions with real coefficients.

6. Number of Zeros NOT greater than number of Poles

From equation (4) the number of finite poles in H (s) is equal to the order of the denominator
polynomial N while the number of zeros is equal to the order of the numerator polynomial M.
Practically only finite poles and zeros are considered which ensures that the number of poles
are greater than the number of zeros since the transfer function is rational.

7. Finite order
Since the order of the transfer function is equal to the number of independent storage elements
in the analog realization, then the transfer function H (s) must be finite.

8. Stable
H (s) must be stable to be practically useful. H (s) is stable if all the poles are in the left half
of the S-plane.
Summary
2
A given function of ω H (jω) that has a corresponding H (s) that conforms to the above
constraints iff

(a) The function is a polynomial of ω over a polynomial in ω with only real coefficients

6
Analog Filters Lecture Notes by A. M. Muhia

(b) There are only even powers of ω in these polynomials

(c) The order of the numerator is less than the order of the denominator
(d) The denominator has no real roots
(e) If the numerator has real roots they occur with even order.

Example 1.1
Consider the function
1
X (ω) =
ω2 + 1
since there are no real roots of ω 2 + 1, then it satisfies the analog filter theorem hence X (ω) is an
2
appropriate H (jω) function and therefore
1 1
X (s) =  2 =
s 1 − s2
+1
j
1
=
(1 − s) (1 + s)
1
X (s) = H (s) H (−s) =
(1 − s) (1 + s)
1
H (s) =
1+s
1
H (−s) =
1−s
Example 1.2
4
X (ω) =
ω4 + 5ω 2 + 4
4
=
(ω 2 + 1) (ω 2 + 4)
4
X (s) =  2 !   !
s s 2
+1 +4
j j
4
=
(1 − s2 ) (4 − s2 )
4
=
(1 − s) (1 + s) (2 − s) (2 + s)
2
H (s) =
(1 + s) (2 + s)
2
H (−s) =
(1 − s) (2 − s)