Scribd
Upload
49 views11 pages

DT Iir Filters

This document discusses infinite impulse response (IIR) filters. IIR filters have feedback which allows poles other than at the origin, but this can cause stability issues. IIR filters also generally do not have linear phase. However, IIR filters have an advantage over finite impulse response (FIR) filters in that they can achieve good magnitude response with lower computational complexity. MATLAB provides commands for designing common IIR filter types including Butterworth, Chebyshev, and elliptic filters. The placement of poles and zeros affects the frequency response, and feedback is what allows a system to be stable or unstable.

Uploaded by

ahmd
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
49 views11 pages

DT Iir Filters

This document discusses infinite impulse response (IIR) filters. IIR filters have feedback which allows poles other than at the origin, but this can cause stability issues. IIR filters also generally do not have linear phase. However, IIR filters have an advantage over finite impulse response (FIR) filters in that they can achieve good magnitude response with lower computational complexity. MATLAB provides commands for designing common IIR filter types including Butterworth, Chebyshev, and elliptic filters. The placement of poles and zeros affects the frequency response, and feedback is what allows a system to be stable or unstable.

Uploaded by

ahmd
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 11

EECE 301

Signals & Systems


Prof. Mark Fowler
Note Set #30
D-T Systems: IIR Filters

1/11

IIR Filters (Recursive Filters)


IIR (Recursive) filters have two issues that constrain their use
1. They do not have linear phase (can get approx. linear phase w/ special designs)
linear phase is more crucial in certain areaslike digital comm & radar
(which involve pulses) or filtering images (which involves edges).
2. May not be inherently stable (feedback gives poles other than at origin)
This can be a serious issue when implementing IIR filters
IIR Filters have one main advantage over FIR filters:
can get good magn. resp. w/o high computational complexity

You can usually get


quite good filters even
with fairly low orders
(like 10 or so).
Feedback
Terms!

2/11

Complexity Comparison: IIR vs FIR


|H()| (dB)

0
-20
-40

Order 7 Ellipical IIR Design

-60
-80
0

0.1

0.2

0.3

0.4
0.5
0.6
Normalized DT Frequency /

0.7

0.8

0.9

0.7

0.8

0.9

|H()| (dB)

0
-20
-40

Order 54 firpm FIR Design

-60
-80
0

0.1

This IIR requires:


15 multiplies
13 additions

0.2

0.3

0.4
0.5
0.6
Normalized DT Frequency /

This FIR requires:


55 multiplies
54 additions

Almost 4x as much computation for the FIR filter!


3/11

MATLAB-Based IIR Design


MATLAB has several easy commands for IIR design, including:

butter, cheby1, cheby2, ellip


Sets Order

Chebyshev IIR

Sets Stopband Height

[b,a]=cheby2(7,60,0.7);
Sets Stopband Edge

b = [0.0692

0.3789

0.9728

1.5028

1.5028

0.9728

0.3789

0.0692]

a = [1.0000

1.2028

1.6599

1.0991

0.6240

0.2098

0.0473

0.0048]

-20

0.8

-40
-60

0.6

-80

0.4

-100
0

0.1

0.2

0.3
0.4
0.5
0.6
0.7
Normalized DT Frequency /

0.8

0.9

<H() radians

Imaginary Part

|H()| (dB)

0.2
0
-0.2

-2

-0.4

-4

-0.6

-6

-0.8

-8
-10

-1

0.1

0.2

0.3
0.4
0.5
0.6
0.7
Normalized DT Frequency /

0.8

0.9

-1

-0.5

0
Real Part

0.5

4/11

Sets Order

Elliptical IIR
Sets Passband Ripple
Sets Stopband Height

[b,a] = ellip(7,0.1,60,0.5);
Sets Passband Edge

b = [0.0338

0.1302

a = [1.0000 -0.8994

0.2821

0.4013

2.1386 -1.5364

0.4013

0.2821

0.1302

1.4793 -0.7327

0.0338]

0.3178 -0.0725]

|H()| (dB)

0
1

-50

0.8

-100

0.6

0.1

0.2

0.3
0.4
0.5
0.6
0.7
Normalized DT Frequency /

0.8

0.9

<H() radians

Imaginary Part

0.4

-150

0.2
0
-0.2

-2

-0.4

-4

-0.6

-6

-0.8

-8

-1
-1

-10

0.1

0.2

0.3
0.4
0.5
0.6
0.7
Normalized DT Frequency /

0.8

0.9

-0.5

0
Real Part

0.5

5/11

Butterworth IIR

Sets Order
Sets Passband Edge

[b,a] = butter(7,0.5);
b = [0.0166

0.1160

a = [1.0000 -0.0000

0.3479

0.5798

0.5798

0.9200 -0.0000

0.3479

0.1927 -0.0000

0.1160

0.0166]

0.0077 -0.0000]

|H()| (dB)

0
1

-50

0.8
0.6

-100

0.1

0.2

0.3
0.4
0.5
0.6
0.7
Normalized DT Frequency /

0.8

0.9

<H() radians

Imaginary Part

0.4

-150

0.2
0
-0.2
-0.4

-5

-0.6
-0.8

-10

-1
-1

-15

0.1

0.2

0.3
0.4
0.5
0.6
0.7
Normalized DT Frequency /

0.8

0.9

-0.5

0
Real Part

0.5

6/11

Visualizing IIR Freq Resp.


Im{z}
0.3
0.8
Re{z}

Closer the poles are to the UC the


more pronounced are the peaks

7/11

Effect of Poles & Zeros on Frequency Response of DT filters


Imz

Note: Including a
pole or zero at the
origin

z
Imz
z

Placing a zero
on UC

doesnt change
the magnitude but
does change the
phase

Imz
z

makes |H| = 0

at angle where
zero is placed

Imz

Placing more
zeros/poles

gives sharper
transitions.

Imz
z

Figure from B.P. Lathi, Signal Processing and Linear Systems


8/11

Effect of Zero at Origin

Effect of Pole at Origin

H1(z) has an extra zero at the origin:

H1(z) has an extra pole at the origin:

H1 ( z ) zH ( z )

H1 ( z ) z H ( z )

H 1 ( ) e j H ( )

H 1 ( ) e j H ( )

H 1 ( ) e j H ( )

H 1 ( ) e j H ( )

e j H ( )
H ()
H1 () e j H ()

e j H ( )
No Effect on
Magnitude

H ()
H1 () e j H ()
H ()

H ()
Effect on Phase

9/11

Effect of Feedback
We know two important things:
Feedback is what can give us poles other than at the origin
Any pole outside the UC causes the system to be unstable
Thus. It is feedback that can cause a system to be stable.

y[n ] 2a cos( ) y[n 1] a 2 y[n 2] x[n ] x[n 1] For a>0 and 0


z z 1
1 z 1
H ( z)
2
1
2 2
1 2a cos( ) z a z
z 2a cos( ) z a 2
Has roots (poles!) at:

z ae j

Im{z}

a
a

Re{z}

So if a is increased to be 1
or larger then the system is
unstable!
10/11

Effect of Changing Value of a (with = /4):


6

As a increases spike becomes:


Higher,
Sharper,
Closer to .

a = 0.9

|H ( )|

a = 0.75
a = 0.5

a = 0.25

0
-1 -0.9 -0.8 -0.7 -0.6 -0.5 -0.4 -0.3 -0.2 -0.1 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Normalized DT Frequency /

As increases spike:
Moves up in freq.,
Becomes Higher.

Effect of Changing Value of (with a = 0.75):


8

= 3/4
= /2

|H()|

WHY???

= /4
4
2
0
-1 -0.9 -0.8 -0.7 -0.6 -0.5 -0.4 -0.3 -0.2 -0.1 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Normalized DT Frequency /

11/11

You might also like

pFad - Phonifier reborn

Pfad - The Proxy pFad of © 2024 Garber Painting. All rights reserved.

Note: This service is not intended for secure transactions such as banking, social media, email, or purchasing. Use at your own risk. We assume no liability whatsoever for broken pages.


Alternative Proxies:

Alternative Proxy

pFad Proxy

pFad v3 Proxy

pFad v4 Proxy