622 (ESS)
1 Filter Approximation Concepts
How do you translate filter specifications into a mathematical expression which can be synthesized ? • Approximation Techniques
Why an ideal Brick Wall Filter can not be implemented ? • Causality: Ideal filter is non-causal • Rationality: No rational transfer function of finite degree (n) can have such abrupt transition
|H(jω)| h(t)
Passband Stopband
ω t -ωc ωc -Tc Tc 2 Filter Approximation Concepts
Practical Implementations are given via window specs.
|H(jω)|
1
Amax
Amin
ω ωc ωs
Amax = Ap is the maximum attenuation in the passband
Amin = As is the minimum attenuation in the stopband
ωs-ωc is the Transition Width 3 Approximation Types of Lowpass Filter
1 Definitions Ap
As
Ripple = 1-Ap wc ws Stopband attenuation =As Filter specs Maximally flat (Butterworth) Passband (cutoff
frequency) = wc
Stopband frequency = ws
Equal-ripple (Chebyshev) Elliptic
4 Approximation of the Ideal Lowpass Filter
Since the ideal LPF is unrealizable, we will accept a small error in the passband, a non-zero transition band, and a finite stopband attenuation 1 퐻 푗휔 2 = 1 + 퐾 푗휔 2 |H(jω)| • 퐻 푗휔 : filter’s transfer function • 퐾 푗휔 : Characteristic function (deviation of 푇 푗휔 from unity)
For 0 ≤ 휔 ≤ 휔푐 → 0 ≤ 퐾 푗휔 ≤ 1 For 휔 > 휔푐 → 퐾 푗휔 increases very fast Passband Stopband
ω ωc 5 Maximally Flat Approximation (Butterworth) Stephen Butterworth showed in 1930 that the gain of an nth order maximally flat magnitude filter is given by
1 퐻 푗휔 2 = 퐻 푗휔 퐻 −푗휔 = 1 + 휀2휔2푛
퐾 푗휔 ≅ 0 in the passband in a maximally flat sense
푑푘 퐾 푗휔 2 2 푘 = 0 for 푘 = 1,2, … , 2푛 − 1 푑 휔 휔=0
The corresponding pole locations (for 휀 = 1) can be determined as follows
1 1 2푛 퐻 푠 2 = = → 푠 = − −1 −푛 = 푒푗휋 2푘−1+푛 푠 2푛 푛 2푛 푝 1+ 1+ −1 푠 푗 휋 2푘−1+푛 푗 푠푝 = 푒 2 푛 푘 = 1, … , 2푛 6 Pole Locations: Maximally flat (ε=1)
The poles are located on the unity circle at equispaced angles 휋 2푘−1+푛 푠 = 푒푗휃푘 where 휃 = 푘 = 1,2, … , 2푛 푝 푘 2 푛 The real and imaginary parts are 2푘−1 휋 2푘−1 휋 푅푒 푆 = − sin Im 푆 = cos 푝 푛 2 푝 푛 2 Poles in the LHP are associated with 퐻 푠 and poles in the RHP are associated with 퐻 −s
휽풌 ퟎ° ±ퟏퟑퟓ° ±ퟏퟐퟎ°, ퟏퟖퟎ° ±ퟏퟏퟐ. ퟓ°, ±ퟏퟓퟕ. ퟓ° ±ퟏퟎퟖ°, ±ퟏퟒퟒ°, ퟏퟖퟎ° Magnitude Response of Butterworth Filter
8 Design Example
Design a 1-KHz maximally flat lowpass filter with: • Attenuation at 10 kHz ≥2000
Normalized prototype: 휔푐 = 1 푟푎푑 푠 , 휔푠 = 10 푟푎푑 푠 2 2 1 1 2푛 6 퐻 푗휔푠 = 2푛 ≤ → 10 ≥ 4 × 10 or 푛 ≥ 3.3 1+휔푠 2000
Choose 푛 = 4 Pole locations 푗휃푘 푠푝 = 푒 where 휃푘 = ±112.5° , ±157.5° 푠푝1,2 = −0.383 ± 푗0.924 푠푝3,4 = −0.924 ± 푗0.383 Normalized transfer function 1 퐻 푠 = 푠2 + 0.765푠 + 1 푠2 + 1.848푠 + 1
퐻1 퐻2 푄 = 1.306, 휔푐 = 1 푄 = 0.541, 휔푐 = 1 9
Design Example
푠 Denormalized transfer function 푠 = 휔푐 1.5585 × 1015 퐻 푠 = 푠2 + 4.8 × 103 푠 + 3.95 × 107 푠2 + 1.16 × 104 푠 + 3.95 × 107
퐻1 퐻2 3 3 푄 = 1.306 , 휔푐 = 2휋10 푄 = 0.541 , 휔푐 = 2휋10
10 Design Example -Discussion
We can sharpen the transition in the previous example by increasing the quality factor of one of
the two cascaded filters (Q1=3.5 instead of 1.3)
To alleviate the peaking problem in the previous ′ response, we can reduce 휔푐2 휔푐2 = 0.6휔푐1
Passband ripples are now existing • They can be tolerated in some applications
The resulting response has steeper transition than Butterworth response 11 Equiripple Filter Approximation (Chebyshev I) This type has a steeper transition than Butterworth filters of the same order but at the expense of higher passband ripples |H(jω)|2
Magnitude response of this type is given by 1 1/(1+ε2) 1 퐻 푗휔 2 = 1+ 퐾 푗휔 2 ω -1 1 퐾 푗휔 = 휀퐶 휔 푛 2 |K(jω)| −1 퐶푛 휔 = cos 푛 cos 휔 휔 ≤ 1 ε2 ω Going back and forth in ±1 range for -1 1 휔 ≤ 1 ε-1|K(jω)| 푪풏 is called Chebyshev’s polynomial 1 ω -1 -1 0 1 Defining the passband area of 푲 풋흎 Equiripple Filter Approximation (Chebyshev I) Chebyshev’s Polynomial e푗푛휙+e−푗푛휙 퐶 휔 = cos 푛 cos−1 휔 = 푛 2 풏 푪풏 For the stopband (휔 > 1) 0 1 1 휔 −1 휙 = cos 휔 is complex 2 2휔2 − 1 Thus, 퐶 > 1 푛 3 3 4휔 − 3휔 Since 4 8휔4 − 8휔2 + 1 cos 푛휙 = cosh 푛푗휙 푛 2휔퐶 − 퐶 and 푛−1 푛−2 푗휙 = cosh 휔 Then −1 퐶푛 휔 = cosh 푛 cosh 휔 for 휔 > 1
13 Properties of the Chebyshev polynomials
1 Cn cos(ncos (w))1 w1
Faster response in the stopband In the passband, w<1, Cn is For n>3 and w>1 limited to 1, then, the ripple is determined by e
Butterwort h Chebyshev 2 1 n1 N( jw) C 2 2nw2 n1 23 w2 n1 2 2 n 1e C n w
2 2 The -3 dB frequency can be found as: e C n w3dB 1 w1 1 1 1 w3dB cosh cosh w1 n e
14 Pole Locations (Chebyshev Type I)
15 Magnitude Response (Chebyshev Type I)
16 Comparison of Design Steps for Maximally Flat and Chebyshev Cases
Step Maximally Flat Chebyshev Find n log 100.1훼푚푖푛 − 1 / 100.1훼푚푎푥 − 1 cosh−1 100.1훼푚푖푛 − 1 100.1훼푚푎푥 − 1 0.5 푛 = 푛 = −1 2 log 휔푠 2 cosh 휔푠 Round up to an integer Round up to an integer Find ε 100.1훼푚푎푥 − 1 1 2 100.1훼푚푎푥 − 1 1 2 180° Find pole If 푛 is odd 휃 = 0°, ±푘 Find 휃푘 as in Butterworth case 푘 푛 180° Find locations If 푛 is even 휃 = ±푘 푘 2푛 1 1 훼 = sinh−1 푛 휀 −1 푛 Radius = Ω0 = 휀 휔푝 Then, −휎 = Ω cos 휃 푘 0 푘 −휎푘 = sin 휃푘 sinh 훼 ±휔 = Ω sin 휃 푘 0 푘 ±휔푘 = cos 휃푘 cosh(훼)
17 Inverse Chebyshev Approximation (Chebyshev Type II) This type has • Steeper transition compared to Butterworth filters (but not as steep as type I) • No passband ripples |H(jω)|2 • Equal ripples in the stopband 1 Magnitude response of this type is given by 2 1 1/(1+ε-2) 퐻 푗휔 = ω 1+ 퐾 푗휔 2 -1 1 2 |K(jω)| 1 ∞ 퐾 푗휔 = 1 휀퐶푛 휔 ε-2 ω -1 1 1 −1 1 1 퐶 = cos 푛 cos ≤ 1 ε|K(jω)| 푛 휔 휔 휔 ∞ 휔 ≥ 1 1 Going back and forth in ±1 range for ω 0 휔 ≥ 1 -1 1 -1 푪 is Chebyshev’s polynomial 풏 ∞ Defining the stopband area of 푲 풋흎 Inverse Chebyshev Approximation (Chebyshev Type II) For the passband (휔 < 1)
1 1 퐶 = cosh 푛 cosh−1 for 휔 < 1 푛 휔 휔 1 1 푛 퐶 ≅ 2푛−1 for 휔 ≪ 1 푛 휔 휔 Attenuation 훼 1 훼 = 10 log 1 + 1 푑퐵 휀2퐶2 푛 휔 1 1 훼 = 10 log 1 + 훼 = 10 log 1 + 푚푎푥 2 2 1 푚푖푛 휀2 휀 퐶푛 휔푝
To find the required order for a certain filtering template 훼 10 1 2 2 1 −1 1 10 푚푖푛 −1 퐶푛 = cosh 푛 cosh = 훼 10 휔푝 휔푝 10 푚푎푥 −1
1 2 cosh−1 10훼푚푖푛 10 − 1 10훼푚푎푥 10 − 1 푛 = 1 cosh−1 19 휔푝 Pole/zero Locations (Inverse Chebyshev)
Pole/zero locations 2 2 1 1 휀 퐶푛 휔 퐻 푗휔 2 = = 1 2 2 1 1 + 휀−2퐶−2 1 + 휀 퐶푛 휔 푛 휔 We have imaginary zeros at ±휔푧,푘 where 2 1 퐶푛 = 0 휔푧,푘 푘휋 휔 = sec , 푘 = 1,3,5, … , 푛 푧,푘 2푛 If 푠푘 = 휎푘 + 푗휔푘 are the poles of Chebyshev filter
Then, 1 푝푘 = 훼푘 + 푗훽푘 = are the poles of inverse Chebyshev filter 푠푘
Magnitude and quality factor of imaginary poles 1 푝푘 = 푄푖퐶ℎ푒푏 = 푄퐶ℎ푒푏 푠푘 20 Pole/zero Locations (Inverse Chebyshev)
Poles of Chebyshev and inverse Chebyshev filters are reciprocal
Since the poles are on the radial line, they have the same pole Q
Imaginary zeros creates nulls in the stopband
n=2 n=3 n=4 n=5 21 Magnitude Response (Inverse Chebyshev)
22 Elliptic Filter Approximation
Elliptic filter Hjw • Equal ripple passband and stopband
• Nulls in the stopband w
• Sharpest transition band compared I to same-order Butterworth and m Chebyshev (Type I and II)
Re
Ellipse Pole/zero Locations (Elliptic)
Imaginary zeros creates nulls in the stopband
n=2 n=3 n=4 n=5
24 Magnitude Response (Elliptic)
25 Design Example
Design a lowpass filter with: • 휔푝 = 1 푅푝 = 1 푑퐵 • 휔푠 = 1.5 푅푠 = 40 푑퐵
Matlab function buttord, cheb1ord, cheb2ord, and ellipord are used to find the least order filters that meet the given specs. 26 Design Example
Magnitude response
Filter approximation meeting the same specification yield Order (Butterworth)> Order (Chebyshev)> Order (Elliptic) 27 Design Example
Pole/zero locations
Note that Chebyshev and Eliptic approximations needs high-Q poles 28 Design example
Filter order for 푅푝 = 1 and 푅푠 = 40
Eliptic filter always yields the least order. Is it always the best choice ? 29 Step response of the design example
• Inverse Chebyshev filter has the least overshoot and ringing
• Ringing and overshoots can be problematic in some applications
• The pulse deformation is due to the fact that the filter introduces different time delay to the different frequency components (Phase distortion) 30 Phase Distortion
• Consider a filter with a transfer function 퐻 푗휔 = 퐻 푗휔 푒푗휙 휔 • Let us apply two sine waves at different frequencies 푣푖푛 푡 = 퐴1 sin 휔1푡 + 퐴2 sin 휔2푡 • The filter output is 휙1 휙2 푣표푢푡 푡 = 퐴1 퐻 푗휔1 sin 휔1 푡 + + 퐴2 퐻 푗휔2 sin 휔2 푡 + 휔1 휔2
• Assuming that the difference between 퐻 푗휔1 and 퐻 푗휔2 is small, the shape of the time domain output signal will be preserved if the two signals are delays by the same amount of time 휙 휔 휙 휔 1 = 2 휔1 휔2 • This condition is satisfied for
휙 휔 = 푡0휔 푡0 = 푐표푛푠푡푎푛푡
• A filter with this characteristic is called “linear phase” 31 Linear Phase Filters
• For this type of filters: The magnitude of the signals is scaled equally, and they are delayed by the same amount of time v (t) Kv (t t ) out in 0
Filter Vout Vin jw(t0 ) Vout(s) Vin (s)Ke
• The filter transfer function is 퐻 푠 = 퐾푒−푗휔푡0 퐻 푗휔 = 퐾 휙 푗휔 = 푡0휔
휙 푗휔 • In this types of filters the phase delay 휏 = − , and the group delay 휏 = 푃퐷 휔 퐺퐷 푑휙 푗휔 − are constant and equal 휔
32 Linear Phase Filter Approximation
• For a typical lowpass filter 퐾 퐾 퐻 푠 = 2 = 2 3 1 + 푎1푠 + 푎2푠 + ⋯ 1 − 푎2휔 + ⋯ + 푗휔 푎1 − 푎3휔 + ⋯
Thus the phase shift is given by 3 −1 푎1 − 푎3휔 + ⋯ 휙 휔 = arg 퐻 푗휔 = − tan 2 1 − 푎2휔 + ⋯
푥3 푥5 • Using power-series expansion tan−1 푥 = 푥 − + − ⋯ 3 5 The condition for linear phase is satisfied if
휕 휕 푥3 푥5 tan−1 푥 = 푥 − + − ⋯ = 푐표푛푠푡푎푛푡 휕휔 휕휔 3 5
These are called Bessel polynomials, and the resulting networks are called Thomson filters 33 Linear Phase Filter Approximation
• Typically the phase behavior of these filters shows some deviation
휏퐺퐷 Ideal response • Errors are measured in time t0 • 휏퐺퐷 is the Group Delay 휕휙 휏 = − Actual response 퐺퐷 휕휔
w
34 Bessel (Thomson) Filter Approximation
• All poles
• Poles are relatively low Q
• Maximally flat group delay (Maximally linear phase response)
• Poor stopband attenuation
http://www.rfcafe.com/references/electrical/bessel-poles.htm
35 Bessel Filter Approximation
36 Comparison of various LPF Group delay
Ref: A. Zverev, Handbook of filter synthesis, Wiley, 1967.
37 Filter Design Conventional Procedure
Transform your filter specs into a normalized LPF
Filter order, zeros, poles and/or values for the passive elements can be obtained from tables or from a software package like FIESTA or Matlab
If you use biquadratic sections, you need poles and zeros matching
For ladder filters, the networks can be obtained from tables
Transform the normalized transfer function to your filter by using Filter transformation (LP to BP, HP, BR) Frequency transformation Impedance denormalization
You obtain the transfer function or your passive network
38 Frequency Transformations
Lowpass to Highpass n 1 H0 H0 H0p s then Hlp (s) Hhp (p) p n n i n a si 1 a pni i ai i i0 i0 p i0
H(s) H(p)
w 0.66 1 1.5 2 0.5 1 1.5 2 p 2 1 0.66 0.5
39 Lowpass to Highpass H H pn H (s) 0 H (p) 0 lp n hp n i ni ais aip i0 i0
N zeros at are translated to zero Poles are not the same!!! The main characteristics of the lowpass filter are maintained
for a highpass filter with cutoff frequency at w0, then
w0 H(p) s This transformation scheme p translates w=1 to p=w0 p 0.66 1 1.5 2
p/w0 1 40 Frequency Transformations
. The Lowpass to Highpass transformation can also be applied to the elements
Element impedance transforme d Resistor R R The resistors are not affected w L inductor sL 0 p L is transformed in a capacitor Ceq=1/w0L 1 p capacitor sC w C 0 C is transformed in an inductor Leq=1/w0C
Example: Design a 1KHz HP-filter from a LP prototype.
1.4142 -4 1 1 1.12x10
1.4142 1 1.12x10-4 1
41 Frequency Transformations
Lowpass to Bandpass transformation
n p2 1 H H p s then H (s) 0 H (p) 0 lp n bp 2n p i i ais bi p i0 i0 • n zeros at w=0 and n zeros at •even number of poles •The bandwidth of the BP is equal to the bandwidth of the LP
•In the p-domain 2 2 s s w w p jv 1 or v 1 2 2 2 2
v1 0.5 1.25 v2 0.5 1.25
w v -1 0 1 -1 0 1 Bandwidth=1
42 Frequency Transformations
Note that v2 v1 1 and v2 v1 1
1
w v -1 1 v1 v2
Lowpass prototype Bandpass filter
43 Frequency Transformations
2 2 General transformation 1 p w s 0 BW p 2 BW BW 2 1 v1 w0 2 2 2 BWw BWw 2 v w0 0 w0 2 2 2 BW BW 2 1 v2 w0 2 2 BW
w v -1 1 v1 w0 v2 Lowpass prototype Bandpass filter
44 Frequency Transformations
The Lowpass to Bandpass transformation can also be applied to the elements
Element impedance transforme d L BW 2 Resistor R R BW Lw0 L w 2L 1 inductor sL p 0 BW BW p BW 1 1 2 capacitor Cw0 sC C w 2C 1 p 0 BW BW p
C BW Note that for w=w0
for the inductor Zeq=0 for the capacitor Yeq=0 (Zeq=)
45 Frequency Transformations
• In general, for double-resistance terminated ladder filters
around w=w0 Rin Lossless
ladder R 1 H s L ww 0 R R R L RL in L 1 Rin
In the passband, the transfer function can be very well controlled
Low-sensitivy
46 Frequency Transformations
Lowpass to Bandreject transformation
1 s 2 2 1 p w0 BW p
• Lowpass to Highpass transformation (notch at w=0)
•Shifting the frequency to w0 and adjusting the bandwidth to BW . Bandpass transformation!!! BW
w p 1 1 w0
47 Transformation Methods • Transformation methods have been developed where a low pass filter can be converted to another type of filter by simply transforming the complex variable s.
• Matlab lp2lp, lp2hp, lp2bp, and lp2bs functions can be used to transform a low pass filter with normalized cutoff frequency, to another low-pass filter with any other specified frequency, or to a high pass filter, or to a band-pass filter, or to a band elimination filter, respectively.
48 LPF with normalized cutoff frequency, to another LPF with any other specified frequency • Use the MATLAB buttap and lp2lp functions to find the transfer function of a third-order Butterworth low-pass filter with cutoff frequency fc=2kHz.
% Design 3 pole Butterworth low-pass filter (wcn=1 rad/s) [z,p,k]=buttap(3); [b,a]=zp2tf(z,p,k); % Compute num, den coefficients of this filter (wcn=1rad/s) f=1000:1500/50:10000; % Define frequency range to plot w=2*pi*f; % Convert to rads/sec fc=2000; % Define actual cutoff frequency at 2 KHz wc=2*pi*fc; % Convert desired cutoff frequency to rads/sec [bn,an]=lp2lp(b,a,wc); % Compute num, den of filter with fc = 2 kHz Gsn=freqs(bn,an,w); % Compute transfer function of filter with fc = 2 kHz semilogx(w,abs(Gsn)); grid; xlabel('Radian Frequency w (rad/sec)') ylabel('Magnitude of Transfer Function') title('3-pole Butterworth low-pass filter with fc=2 kHz or wc = 12.57 kr/s')
49 LPF with normalized cutoff frequency, to another LPF with any other specified frequency 3-pole Butterworth low-pass filter with fc=2 kHz or wc = 12.57 kr/s 1
0.9
0.8
0.7
0.6
0.5
0.4
0.3 Magnitude of FunctionTransfer Magnitude 0.2
0.1
0 3 4 5 10 10 10 Radian Frequency w (rad/sec) 50 High-Pass Filter • Use the MATLAB commands cheb1ap and lp2hp to find the transfer function of a 3-pole Chebyshev high-pass analog filter with cutoff frequency fc = 5KHz.
% Design 3 pole Type 1 Chebyshev low-pass filter, wcn=1 rad/s [z,p,k]=cheb1ap(3,3); [b,a]=zp2tf(z,p,k); % Compute num, den coef. with wcn=1 rad/s f=1000:100:100000; % Define frequency range to plot fc=5000; % Define actual cutoff frequency at 5 KHz wc=2*pi*fc; % Convert desired cutoff frequency to rads/sec [bn,an]=lp2hp(b,a,wc); % Compute num, den of high-pass filter with fc =5KHz Gsn=freqs(bn,an,2*pi*f); % Compute and plot transfer function of filter with fc = 5 KHz semilogx(f,abs(Gsn)); grid; xlabel('Frequency (Hz)'); ylabel('Magnitude of Transfer Function') title('3-pole Type 1 Chebyshev high-pass filter with fc=5 KHz ')
51 High-Pass Filter 3-pole Chebyshev high-pass filter with fc=5 KHz 1
0.9
0.8
0.7
0.6
0.5
0.4
0.3 Magnitude of FunctionTransfer Magnitude 0.2
0.1
0 3 4 5 10 10 10 Frequency (Hz)
52 Band-Pass Filter • Use the MATLAB functions buttap and lp2bp to find the transfer function of a 3-pole Butterworth analog band-pass filter with the pass band frequency centered at fo = 4kHz , and bandwidth BW =2KHz.
[z,p,k]=buttap(3); % Design 3 pole Butterworth low-pass filter with wcn=1 rad/s [b,a]=zp2tf(z,p,k); % Compute numerator and denominator coefficients for wcn=1 rad/s f=100:100:100000; % Define frequency range to plot f0=4000; % Define centered frequency at 4 KHz W0=2*pi*f0; % Convert desired centered frequency to rads/s fbw=2000; % Define bandwidth Bw=2*pi*fbw; % Convert desired bandwidth to rads/s [bn,an]=lp2bp(b,a,W0,Bw); % Compute num, den of band-pass filter % Compute and plot the magnitude of the transfer function of the band-pass filter Gsn=freqs(bn,an,2*pi*f); semilogx(f,abs(Gsn)); grid; xlabel('Frequency f (Hz)'); ylabel('Magnitude of Transfer Function'); title('3-pole Butterworth band-pass filter with f0 = 4 KHz, BW = 2KHz') 53 Band-Pass Filter
3-pole Butterworth band-pass filter with f0 = 4 KHz, BW = 2KHz 1.4
1.2
1
0.8
0.6
0.4 Magnitude of FunctionTransfer Magnitude
0.2
0 2 3 4 5 10 10 10 10 Frequency f (Hz) 54 Band-Elimination (band-stop) Filter • Use the MATLAB functions buttap and lp2bs to find the transfer function of a 3-pole Butterworth band-elimination (band-stop) filter with the stop band frequency centered at fo = 5 kHz , and bandwidth BW = 2kHz.
[z,p,k]=buttap(3); % Design 3-pole Butterworth low-pass filter, wcn = 1 r/s [b,a]=zp2tf(z,p,k); % Compute num, den coefficients of this filter, wcn=1 r/s f=100:100:100000; % Define frequency range to plot f0=5000; % Define centered frequency at 5 kHz W0=2*pi*f0; % Convert centered frequency to r/s fbw=2000; % Define bandwidth Bw=2*pi*fbw; % Convert bandwidth to r/s % Compute numerator and denominator coefficients of desired band stop filter [bn,an]=lp2bs(b,a,W0,Bw); % Compute and plot magnitude of the transfer function of the band stop filter Gsn=freqs(bn,an,2*pi*f); semilogx(f,abs(Gsn)); grid; xlabel('Frequency in Hz'); ylabel('Magnitude of Transfer Function'); title('3-pole Butterworth band-elimination filter with f0=5 KHz, BW = 2 KHz') 55 Band-Elimination (band-stop) Filter 3-pole Butterworth band-elimination filter with f0=5 KHz, BW = 2 KHz 1
0.9
0.8
0.7
0.6
0.5
0.4
0.3 Magnitude of FunctionTransfer Magnitude 0.2
0.1
0 2 3 4 5 10 10 10 10 Frequency in Hz 56 How to find the minimum order to meet the filter specifications ? The following functions in Matlab can help you to find the minimum order required to meet the filter specifications:
• Buttord for butterworth • Cheb1ord for chebyshev • Ellipord for elliptic • Cheb2ord for inverse chebyshev
57 Calculating the order and cutoff frequency of a inverse chebyshev filter • Design a 4MHz Inverse Chebyshev approximation with Ap gain at passband corner. The stop band is 5.75MHz with -50dB gain at stop band.
clear all; Fp = 4e6; Wp=2*pi*Fp; Fs=1.4375*Fp; Ws=2*pi*Fs; Fplot = 20*Fs; f = 1e6:Fplot/2e3:Fplot ; w = 2*pi*f; Ap = 1; As = 50; % Cheb2ord helps you find the order and wn (n and Wn) that %you can pass to cheby2 command. [n, Wn] = cheb2ord(Wp, Ws, Ap, As, 's'); [z, p, k] = cheby2(n, As, Wn, 'low', 's'); [num, den] = cheby2(n, As, Wn, 'low', 's'); bode(num, den) 58 Bode Plot
Bode Diagram 0
-50
-100
Magnitude (dB) Magnitude -150
-200 1260
1080
900
Phase (deg) Phase 720
540 5 6 7 8 9 10 10 10 10 10 10 10 Frequency (rad/sec)
59 References [1] S. T. Karris, “Signals and Systems with Matlab Computing and Simulink Modeling,” Fifth Edition. Orchard Publications [2] Matlab Help Files
60 Ladder Filters
.The ladder filter realization can be found in tables and/or can be obtained from FIESTA
.The elements must be transformed according to the frequency and impedance normalizations 0
-40
)
B
d (
-80
n
i
a g
-120
-160 4.0E+6 8.0E+6 1.2E+7 1.6E+7 2.0E+7 Frequency (Hz) 61 Sensitivity
y x y Definition S Y= transfer function and x = variable or element x y x
Some properties: ky y y x x Sx Skx Sx Skx Sx 1 1/ y y y yn y Sx S1/ x Sx Sx nSx y 1 y y y x2 S Sx Sx S Sx xn n x2 n n n yi yi n yi yiSx Si1 S yi Si1 i1 x x x n i1 yi i1 For a typical H(s) H i H(s) 0 i ai jw ai jw S j a jwi a j a jjw S i a j i i ai jw ai jw
62 Sensitivity
• Sensitivity is a measure of the change in the performance of the system due to a change in the nominal value of a certain element.
y x y y x Dy Dy y Dx Sx Sx or Sx y x y Dx y x
Normalized variations at the output are determined by the sensitivity function and the normalized variations of the parameter Example: If the senstivity function is 10, then variations of Dx/x=0.01(1%) produce Dy/y=0.1(10%)
For a good design, the sensitivity functions should be < 5. Effects of the partial positive feedback (negative resistors)?
63 Sensitivity
gm . For a typical amplifier Av g0
SAv 1, SAv 1 gm g0
•Sometimes the dc gain is enhanced by using a negative resistor
gm gm 1 Av g g g g For large dc gain g02=g0 0 02 0 1 02 g0 g0 g0Sg 1 SAv 1, SAv 0 gm g0 g g g 0 02 1 02 g0
The larger the gain improvment the larger the sensitivity!!!!
64 Properties of Stable Network Functions
• Typically the transfer function A(s) 1b sb s 2 .. presents the form of a ratio of two N(s) K 1 2 polynomials 2 B(s) 1a1sa 2s .. • For non-negative elements the coefficients are real and positive • The poles are located in the left side of the s-plane
1 H(s) • System is stable sa • BOUNDED OUTPUT FOR BOUNDED INPUT h(t)e at t0
65 Properties of Network Functions: Frequency Transformation
Most of the filter approximations are normalized to 1 rad/sec. Hence, it is necessary to denormalize the transfer function.
Using the frequency transformation p w For inductors and L n jp L jw capacitors: n 1 rad/sec is translated to 1/ C n rad/sec jp C jw n
66 Impedance denormalization
. Typically the network elements are normalized to 1 . Hence an impedance denormalization scheme must be used
Z n Z . or
R n R
L L n C C n . Note that the transfer function is invariant with the impedance denormalization (RC and LC products remain constant!!!!)
. In general both frequency and impedance denormalizations are used
67
• There is a number of conventional filter magnitude approximations
• The choice of a particular approximation is application dependent
• Besides the magnitude specifications, there exists also a phase (group delay) specification. For this the Thompson (Bessel) approximation is used
• There are a host of Filter approximation software programs, including Matlab, Filsyn, and Fiesta2 developed at TAMU
Acknowledgment: Thanks to my colleague Dr. Silva-Martínez for providing some of the material for this presentation
68