dsp5-iir 1
DESCRIPTION
lectureTRANSCRIPT
25-Mar-15
1
CH 5
IIR
ME-4722
Digital Signal Processing
Chapter 5
Design of Digital IIR
Filters(Elective)
Spring 2015, SZABIST, Karachi
CH 5
IIR
Instructor:
Engr. Humera Rafique
Assistant Professor (Mechatronics)
Office: FR-404 (100 Campus )
Course Support
Official: ZABdesk
25-Mar-15 HR DSP S15
2
25-Mar-15
2
CH 5
IIR Chapter Contents
Bilinear Transformation: IIR Digital Filter DesignBilinear Transformation: IIR Digital Filter DesignBilinear Transformation: IIR Digital Filter DesignBilinear Transformation: IIR Digital Filter Design• Introduction• Design summary• Implementation of BLT:
1. 1st order lowpass IIR filter2. 1st order highpass IIR filter3. 2nd order banpass IIR filter (Peaking)4. 2nd order bandstop IIR filter (Notching)5. Higher order lowpass/highpass* filters (Butterworth)6. Higher order bandpass/bandstop* filters (Butterworth)
25-Mar-15 HR DSP S15
3
CH 5
IIR
Digital IIR Filter Design:Digital IIR Filter Design:Digital IIR Filter Design:Digital IIR Filter Design:Bilinear TransformationBilinear TransformationBilinear TransformationBilinear Transformation
25-Mar-15 HR DSP S15
4
25-Mar-15
3
CH 5
IIR Bilinear Transformation• Simplest & effective method of IIR filter design
• Instead of designing a digital filter directly, this algorithm maps a digital filter into an equivalent analog filter e.g.,
* Butterworth* Chebychev* Elliptical etc
• Designed analog filter is then mapped back into the desired digital filter
25-Mar-15 HR DSP S15
5
CH 5
IIR Bilinear Transformation
25-Mar-15 HR DSP S15
6
25-Mar-15
4
CH 5
IIR Bilinear Transformation
4565789 :;<=><?@A: B C2D:
:E<=>5F89<?7 8?89G6 :;<=><?@A: Ω C 6"B%
I C : J C : <KL
M N CO P NQO
O R NQO
Ω C tanB
2Bilinear transform
• The z-plane design of the digital filter is replaced by s-plane of an equivalent analog filter, using suitable
mapping relationships:
Bilinear transform
• Because of non-linear relationship of analog & digital frequencies, this is called ‘pre-warping
transformation’
25-Mar-15 HR DSP S15
7
CH 5
IIR Bilinear Transformation
Design Method SummaryDesign Method SummaryDesign Method SummaryDesign Method Summary
1. Given: Magnitude response specification of a desired digital filter2. Transform these specifications using suitable bilinear transformation relationship into an analog filter’s
specification3. Using an analog filter design technique, design an analog filter Ha(s)4. Use bilinear transformation and map analog filter back into desired digital filter H(z)
Z;8?I:<; :>?@75G?: [ J C [\(I)]E ^_(`)
C [\ :(J)
a86?57>4< ;<IbG?I<: [ B C [\(Ω)]c ^d(L)
C [\ 6(B)
25-Mar-15 HR DSP S15
8
25-Mar-15
5
CH 5
IIR
Design Method SummaryDesign Method SummaryDesign Method SummaryDesign Method Summary
5. To design a stable and causal digital filter a causal and stable analog filter is required. Bilinear
transformation maps left hand side of the s-plane to ‘inside of the unit circle’ in z-plane:
25-Mar-15 HR DSP S15
9Bilinear Transformation
CH 5
IIR
• All analog filter design methods give rise to stable and causal transfer functions Ha"s%
• This property guarantees that the digital filter H"z% obtained will also be stable and causal
Design Method SummaryDesign Method SummaryDesign Method SummaryDesign Method Summary
25-Mar-15 HR DSP S15
10Bilinear Transformation
25-Mar-15
6
CH 5
IIR
Implementation ofImplementation ofImplementation ofImplementation of
Bilinear TransformationBilinear TransformationBilinear TransformationBilinear Transformation
25-Mar-15 HR DSP S15
11
CH 5
IIR
� 1st order low pass filter
� 1st order high pass filter
� 2nd order band pass filter (peaking)
� 2nd order band stop filter (notching)
� Higher order (Butterworth) low and high pass filters
� Higher order (Butterworth) band pass and band stop filters
� Equalizers*
25-Mar-15 HR DSP S15
12Bilinear Transformation
25-Mar-15
7
CH 5
IIR
Implementation ofImplementation ofImplementation ofImplementation of
Bilinear TransformationBilinear TransformationBilinear TransformationBilinear Transformation
1111stststst Order Order Order Order LowpassLowpassLowpassLowpass Filter DesignFilter DesignFilter DesignFilter Design
25-Mar-15 HR DSP S15
13
CH 5
IIR
• A simple filter design of type digital low pass with Nth order can be represented as:
• 1st order:
1st order LPF Design
[ J Cg"J%
h"J%C
ij R ikJQk R ⋯ R imJQm
1 R 8kJQk R ⋯ R 8nJQn
[ J Cg"J%
h"J%C
ij R ikJQk
1 R 8kJQk
Parameters: Parameters: Parameters: Parameters: • ZAb<: 4565789 9Go b8II
• G;4<;: 1I7 • b;<I@;5i<4 :;<=><?@A @>7G:: : :p
• I8qb95?6 :;<=><?@A ∶ :s
25-Mar-15
8
CH 5
IIR
Design Problem:
• To determine the filter’s coefficients in terms of
• Range of frequencies pass through filter:
• Range of frequencies filtered through filter:
• Digital cutoff frequency:
ij, ik, 8k tp , :p , :E
0 v : v :p
:p v : v:E
2
Bp C2D:p
:E
25-Mar-15 HR DSP S15
151st order LPF Design
CH 5
IIR
Design Problem:
25-Mar-15 HR DSP S15
161st order LPF Design
25-Mar-15
9
CH 5
IIR
Design Problem:Ω C tan
B
2
25-Mar-15 HR DSP S15
17
CH 5
IIR
HertzianHertzianHertzianHertzian response parameters:response parameters:response parameters:response parameters:wp: b8II i8?4 685? 87 @>7G:: :;<=><?@Atp: 4x F<;I5G? G: wp:p: @>7G:: :;<=><?@A:E: I8qb95?6 :;<=><?@A
wp: b8II i8?4 685? 87 @>7G:: :;<=><?@Atp: 4x F<;I5G? G: wpBp: @>7G:: :;<=><?@A
wp: b8II i8?4 685? 87 @>7G:: :;<=><?@Atp: 4x F<;I5G? G: wpΩp: 8?89G6 @>7G:: :;<=><?@A
Desired digital Desired digital Desired digital Desired digital lowpasslowpasslowpasslowpass filter’s parameters:filter’s parameters:filter’s parameters:filter’s parameters:
Equivalent analog Equivalent analog Equivalent analog Equivalent analog lowpasslowpasslowpasslowpass filter’s parameters:filter’s parameters:filter’s parameters:filter’s parameters:
25-Mar-15 HR DSP S15
181st order LPF Design
25-Mar-15
10
CH 5
IIR
tp C P10 log wpy
wpy C 10Qz{/y|
5: ["0% y C 17}<? ["Bp%y C wp
y
Ωp C tanBp
2C tan
D:p
:E
Analog cutoff
in terms of
digital
prescribed
parameters
Bilinear Transformation:
wpy C
["Bp% y
["0% yC1
2
⟹ tp C P10 log["Bp%
y
["0% yC 34x
25-Mar-15 HR DSP S15
191st order LPF Design
CH 5
IIR
1st Order Low Pass Analog Filter Transfer Function:
[\ I C �
I R �
[\ Ω y C�y
Ωy R �y
Determining Filter Parameters:wp , :p i, 8
� Cwp
1 P wpyΩp
wp , :p �, Ωp Bp↔ ↔
digitalfilterspec
analogfilterconstant
analogfrequency
digitalfreq.
25-Mar-15 HR DSP S15
201st order LPF Design