[SOLVED] testing a iir filter in matlab

[lakshmikalyani]

hi
i am doing a project on filter design
i am stuck while testing the filter.
i want to know how to test a filer in matlab coding
my code is

---

Code dot - [expand]
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
clc;
clear all;
 
in=[100,200,300,470,500 550];
 
wp=420;
ws=490;
rp=3;
rs=80;
fs=1000;
w1=2*wp/fs;w2=2*ws/fs;
[n,wn]=cheb1ord(w1,w2,rp,rs)
[b,a]= cheby1(n,rp,wn)
[x,y]=freqz(b,a,fs);
plot(y,abs(x));

---

i want to give "in" as input to the filer that had designed. and check the output and analyze it
can u please help me
i know we can give a sine signal in simulink and realize the filter and fda tools but i am not getting the expected outputs
i want to do it in coding
can u help me out of it

 

[weetabixharry]

You have generated your transfer function coefficients, a and b. Therefore, you can use Matlab's filter function to apply the filter:

---

Code Matlab M - [expand]
1
out = filter(b, a, in);

---

 

[lakshmikalyani]

thanku weetabixharry

i did the same thing
infact my code is now

freq=0:10:10000;
t=0.001;
in=10*sin(freq*t);
subplot(2,2,1);
plot(freq,in);

wp=400;
ws=450;
rp=2;
rs=80;
fs=1000;
w1=2*wp/fs;w2=2*ws/fs;
[N,wn]=ellipord(w1,w2,rp,rs)
[b,a]=ellip(N,rp,rs,wn);
[x,y]=freqz(b,a,fs);
mag=20*log(abs(x));
subplot(2,2,2);
plot(y,mag);

out=filter(b,a,in);
subplot(2,2,3);
plot(freq,out);

but the problem is now i am unable to analyze the output
i want to know whether i gave the correct input and the o/p is correct or not
can u help me in analyzing the output

 

[weetabixharry]

First of all, whenever you post code on the forum, please use code tags so I can read it easily (this is rule number 1 in the forum rules). Also, please format and comment your code so it is easier to read.

Secondly, it looks like you have some confusion between frequency and time in your code. I have tried to tidy up your code and correct some errors, so now we have this:

---

Code Matlab M - [expand]
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
close all; clear all; clc;
 
% Generate input signal
fsamp = 1000; % (Hz)
dt = 1/fsamp; % (secs)
t = 0:dt:1;
fsine = 100;
in = 10*sin(fsine*t);
 
% Plot input signal
plot(t, in);
title('Input');
xlabel('Time (secs)');
ylabel('Amplitude');
grid on;
 
% Design filter
wp = 400;
ws = 450;
rp = 2;
rs = 80;
fs = 1000;
w1 = 2*wp/fs;w2=2*ws/fs;
[N,wn] = ellipord(w1,w2,rp,rs);
[b,a] = ellip(N,rp,rs,wn);
 
% Plot filter frequency response
[H,W] = freqz(b,a,fs);
mag = 20*log(abs(H));
figure();
plot(fsamp*W/(2*pi), mag);
title('Frequency Response');
xlabel('Frequency (Hz)');
ylabel('(dB)');
axis tight;
grid on;
 
% Apply filter and plot output
out = filter(b, a, in);
figure();
plot(t, out);
xlabel('Time (secs)');
ylabel('Amplitude');
grid on;

---

Okay, so now back to your question:

but the problem is now i am unable to analyze the output
i want to know whether i gave the correct input and the o/p is correct or not
can u help me in analyzing the output

What do you mean by "correct" input and output? In my opinion, the output will always be correct - it will be the filtered version of the input. Exactly what sort of analysis do you want to do?

 

[lakshmikalyani]

i just want to analyze the output by giving some input for example sine wave by coding in matlab

 

[weetabixharry]

i just want to analyze the output by giving some input for example sine wave by coding in matlab

Okay, well, you have designed your filter as a lowpass filter which should allow signals below 400Hz to pass, but remove all signals above 450Hz. Therefore, a simple experiment is to input the sum of two sine waves: one below 400Hz and one above 450Hz. The filter should then remove the higher frequency sine wave and allow only the lower frequency sine wave to pass. Therefore, the output should look like a nice clean sine wave.

Here is the code for this example:

---

Code Matlab M - [expand]
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
close all; clear all; clc;
 
% Generate input signal
fsamp = 1000; % (Hz)
dt = 1/fsamp; % (secs)
t = 0:dt:0.2;
fsine1 = 100; % (in passband: should remain in filter output)
fsine2 = 450; % (in stopband: should be removed by filter)
in = 10*sin(2*pi*fsine1*t) + 10*sin(2*pi*fsine2*t);
 
% Plot input signal
plot(t, in);
title('Input');
xlabel('Time (secs)');
ylabel('Amplitude');
grid on;
 
% Design filter
wp = 400; % Passband corner frequency
ws = 450; % Stopband corner frequency
rp = 2;   % Passband ripple (dB)
rs = 80;  % Stopband attenuation (dB)
w1 = 2*wp/fsamp;
w2 = 2*ws/fsamp;
[N,wn] = ellipord(w1,w2,rp,rs);
[b,a] = ellip(N,rp,rs,wn);
 
% Plot filter frequency response
[H,W] = freqz(b,a,fsamp);
mag = 20*log(abs(H));
figure();
plot(fsamp*W/(2*pi), mag);
title('Frequency Response');
xlabel('Frequency (Hz)');
ylabel('(dB)');
axis tight;
grid on;
 
% Apply filter and plot output
out = filter(b, a, in);
figure();
plot(t, out);
xlabel('Time (secs)');
ylabel('Amplitude');
grid on;

---

So here is the plot of the input:

And here is the plot of the output:

Clearly the filter has worked, because only the 100Hz sine wave remains. However, it is worth noting that the 100Hz sine is slightly attenuated (its amplitude is about 8.4, compared to 10 at the input). This is correct, because we can see in the frequency response graph that the amplitude response at 100Hz is a little below 0dB.

 

[lakshmikalyani]

thank you weetabixharry
i have one more doubt. during my observation of fitler frequency response the passband ripple is showing 2.3 db for rp=1db and
4.6db for rp=2.
i want to know how passband ripples are calculated