Digital Signal Processing
Dr. Rohan Katti
Assistant Professor
Department of ECE
Bennett University
Module 4 (Contact hours: 10)
IIR Filter design:
▪ Analog Filter Design
▪ Design of IIR filter from Analog Filter
▪ IIR Filter design Using Impulse Invariance Method and Bilinear transformation Method
▪ Design of Butterworth and Chebyshev IIR Filters
▪ Frequency Transformation in Digital Domain
▪ Finite word length effect in digital filters
Dept. of ECE, Bennett University EECE313L - DSP
Introduction
▪ The specification of a digital filter is desired frequency response, Hd(ejw ).
▪ The desired impulse response, hd(n) of the digital filter can be obtained by
taking inverse Fourier transform of Hd(ejw).
▪ Now, the hd(n) will be an infinite duration discrete time signal defined for all
values of n in the range –∞ to +∞ .
▪ The filters designed by considering all the infinite samples of impulse response
are called IIR (Infinite Impulse Response) filters.
Dept. of ECE, Bennett University EECE313L - DSP
Introduction
▪ In digital domain, the processing of infinite samples of impulse response is
practically not possible.
▪ Hence direct design of IIR filter is not possible. Therefore, the IIR filters are
designed via analog filters.
▪ In design of IIR filter, the specification of an IIR filter is transformed to specification
of an analog filter and an analog filter with transfer function, H(s) is designed to
satisfy the specification.
▪ Then the analog filter is transformed to digital filter with transfer function, H(z).
Dept. of ECE, Bennett University EECE313L - DSP
Analog Digital
Time t, T Sample n, N
Frequency domain Fourier transform (ω) Frequency domain: DTFT, DFT
Transfer function Laplace transform, S Plane, H(S) Z transform, Z plane, H(Z)
S = 𝜎 + 𝑗𝜔 Z =𝑟𝑒 𝜃 , where 𝜃 = 𝑗𝜔
Plane S plane Z plane
Definition
Introduction
▪ An ideal filter will have an amplitude response that is unity (or at a fixed gain) for
the frequencies of interest (called the pass band) and zero everywhere else
(called the stop band).
▪ The frequency at which the response changes from passband to stopband is
referred to as the cutoff frequency.
▪ The idealized filters defined above, unfortunately, cannot be easily built. The
transition from pass band to stop band will not be instantaneous, but instead
there will be a transition region.
▪ The cutoff frequency (Fc) is the frequency at which the filter
response leaves the error band (or the −3 dB point for a
Butterworth response filter).
▪ The stop band frequency (Fs) is the frequency at which the
minimum attenuation in the stopband is reached.
▪ The pass band ripple (Amax) is the variation (error band) in the
pass band response.
▪ The minimum stop band attenuation (Amin) defines the minimum
signal attenuation within the stop band.
▪ The steepness of the filter is defined as the order (M) of the filter.
M is also the number of poles in the transfer function. A pole is a
root of the denominator of the transfer function.
Not all filters will have all these features. For instance, all-pole configurations (i.e. no zeros in the transfer function)
will not have ripple in the stop band. Butterworth and Bessel filters are examples of all-pole filters with no ripple in
the pass band.
Dept. of ECE, Bennett University EECE313L - DSP
Pole-zero locations of analog filters
Dept. of ECE, Bennett University EECE313L - DSP
Standard Responses
▪ The impulse response of a filter, in the time domain, is proportional to the
bandwidth of the filter in the frequency domain.
▪ The narrower the impulse, the wider the bandwidth of the filter.
Dept. of ECE, Bennett University EECE313L - DSP
Standard Responses
Standard Response Characteristic
Butterworth ▪ It has no ripple in the pass band or the stop band, and because of this is sometimes
called a maximally flat filter.
▪ The Butterworth filter achieves its flatness at the expense of a relatively wide
transition region from pass band to stop band.
Chebyshev ▪ It has a smaller transition region than the same order Butterworth filter, at the
(all-pole filters) expense of ripples in its pass band.
▪ The number of cycles of ripple in the pass band is equal to the order of the filter
Inverse Chebyshev ▪ Chebyshev response can be defined that has a monotonic pass band and ripple in the stop
(both poles, zeros) band.
▪ The inverse Chebyshev has better pass band performance than even the Butterworth.
▪ It is also better than the Chebyshev, except very near the cutoff frequency.
▪ In the transition band, the inverse Chebyshev has the steepest rolloff.
Deep into the stop band, both the Butterworth and Chebyshev will have better attenuation
In terms of transient performance, the inverse Chebyshev lies midway between the Butterworth and the Chebyshev.
Dept. of ECE, Bennett University EECE313L - DSP
[Link]
[Link]
switched-capacitor-have-a-less-steep-roll-off-curve-ma
[Link]
urrent_Spectrum_for_Broken_Rotor_Bars_Fault_Diagnostics
Comparison between Butterworth and
Chebyshev filters
Butterworth Chebyshev
Magnitude response decreases Magnitude response exhibits ripples in
monotonically as frequency increases. passband or stop band
Wider transition band Narrower transition band
Poles lie on a circle Poles lie on an ellipse
Number of poles are more when Order of chebyshev filter is less than
compared to Butterworth filter that of Butterworth
Dept. of ECE, Bennett University EECE313L - DSP
Analog Butterworth low pass filter design
Analog Butterworth low pass filter design
▪ Find the order of the filter (N) from the given specifications.
▪ Round of the order to the next highest integer.
▪ Find the normalized transfer function H(s) for the obtained value of N.
▪ Calculate the value of cut off frequency wc.
▪ Find the transfer function Ha(s) for the above value of wc by substituting s = s/wc in H(s).
Dept. of ECE, Bennett University EECE313L - DSP
Butterworth Polynomials
Dept. of ECE, Bennett University EECE313L - DSP
Example
Design a Butterworth analog low pass filter with 1 dB passband ripple, passband edge
frequency is 2000π rad/sec, stop band edge frequency is 10,000π rad/sec and a
minimum stop band ripple of 40 dB.
Dept. of ECE, Bennett University EECE313L - DSP
Dept. of ECE, Bennett University EECE313L - DSP
1
The normalized transfer function is given by H(s) = 𝑆 2 +𝑆+1 . Use frequency transformation to find
a) Low pass filter with cut off frequency 10 rad/sec
b) High pass filter with cut off frequency 10 rad/sec
c) Bandpass filter with passband bandwidth 10 rad/sec and center frequency of 100 rad/sec.
d) Bandstop filter with stop band bandwidth of 2 rad/sec and center frequency of 10 rad/sec
1
The normalized transfer function is given by H(s) = 𝑆 2 +𝑆+1 . Use frequency transformation to find
a) Low pass filter with cut off frequency 10 rad/sec
b) High pass filter with cut off frequency 10 rad/sec
c) Bandpass filter with passband bandwidth 10 rad/sec and center frequency of 100 rad/sec.
d) Bandstop filter with stop band bandwidth of 2 rad/sec and center frequency of 10 rad/sec
Design of Digital IIR filters from analog
filters
Two methods for designing digital IIR filters from analog filter
✓ Impulse invariance method
✓ Bilinear transformation
Dept. of ECE, Bennett University EECE313L - DSP
Impulse Invariance method
Dept. of ECE, Bennett University EECE313L - DSP
Dept. of ECE, Bennett University EECE313L - DSP
Dept. of ECE, Bennett University EECE313L - DSP
Dept. of ECE, Bennett University EECE313L - DSP
Impulse Invariance method
Steps for Impulse Invariance method
1) Find the transfer function of the analog filter
2) Select the sampling period T
3) Express the transfer function of analog filter as sum of single poles
4) Analog to digital transformation is given by
Dept. of ECE, Bennett University EECE313L - DSP
Example
2
The transfer function of an analog filter is given by H(s) =
(𝑠+1)(𝑠+2)
Derive the transfer function of digital IIR filter using impulse invariance method.
Consider the sampling period to be 1 sec.
Dept. of ECE, Bennett University EECE313L - DSP
Example
2
The transfer function of an analog filter is given by H(s) =
(𝑠+1)(𝑠+2)
Derive the transfer function of digital IIR filter using impulse invariance method.
Consider the sampling period to be 1 sec.
Dept. of ECE, Bennett University EECE313L - DSP
Dept. of ECE, Bennett University EECE313L - DSP
Dept. of ECE, Bennett University EECE313L - DSP
Dept. of ECE, Bennett University EECE313L - DSP
Dept. of ECE, Bennett University EECE313L - DSP
Differences
•Bilinear transform warps the frequency axis so that no two analog frequencies
map to the same digital frequency — so no aliasing.
•Impulse invariance just samples the impulse response, and sampling always risks
aliasing unless pre-filtering or lowpass restriction is applied.
• Bilinear transformation method can be used for any filters
• IIM cannot be used for high pass and bandstop filters.
• IIM is many to one mapping whereas BLT is one to one mapping.
Steps in Bilinear transformation
1) Find the analog frequencies using the formula
2) Using the analog frequencies find H(s) of the analog filter
3) Substitute s = into the transfer function H(s) to get H(z).
Dept. of ECE, Bennett University EECE313L - DSP
Dept. of ECE, Bennett University EECE313L - DSP
Example
𝑠+0.1
Convert the analog filter with transfer function 𝐻 𝑠 = into a digital IIR filter using
(𝑠+0.1)2 +16
𝜋
bilinear transformation. The digital filter is said to have a resonant frequency 𝜔𝑟 =
2
In bilinear transformation, we replace s by
The relation between analog and digital frequency is given by
2 𝜔
Ω = tan( )
𝑇 2
𝜋 1−𝑧 −1
From the transfer function, Ω=4, 𝜔𝑟 = , thus, we get T = 0.5, thus substitute S by 4[ −1]
2 1+𝑧
in the given transfer function.