i have a signal with many frequencies. how to remove a particular frequency and reconstruct the signal.

I heard about doing fft and then ifft but don't know how to implement. for example, the frequencies are close to each other as ( 0.1282 0.5128 0.8974 1.1538) now, how can I remove only 0.5128 Hz frequency and reconstruct the signal.

I've attached my data file and frequency finding code.

Answer 1

[Star Strider]

Try this filter:

signal = load('datee.txt');                             % Signal Vector
signal = detrend(signal,0);
Fs = 10;                                                % Sampling Frequency (Hz)
Ts = 1/Fs;                                              % Sampling Interval (sec)
tv = linspace(0, 1, numel(signal))*Ts;                  % Time Vector
Fn = Fs/2;                                              % Nyquist Frequency (Hz)
Wp = [0.503 0.523]/Fn;                                  % Passband Frequency (Normalised)
Ws = [0.483 0.543]/Fn;                                  % Stopband Frequency (Normalised)
Rp =   1;                                               % Passband Ripple (dB)
Rs =  50;                                               % Stopband Ripple (dB)
[n,Ws] = cheb1ord(Wp,Ws,Rp,Rs);                         % Filter Order
[z,p,k] = cheby1(n,Rs,Ws,'stop');                       % Filter Design
[sossb,gsb] = zp2sos(z,p,k);                            % Convert To Second-Order-Section For Stability

figure(5)
freqz(sossb, 2^16, Fs)                                  % Filter Bode Plot

filtered_signal = filtfilt(sossb, gsb, signal);         % Filter Signal

figure(6)
subplot(2,1,1)
plot(tv, signal)
grid
subplot(2,1,2)
plot(tv, filtered_signal)
grid

This designs a narrow Chebyshev Type I bandstop filter. Experiment with the filter stopbands and passbands to get the result you want.

To use it with my previous code, append it to the end of my previous code, and substitute this assignment:

signal = load('datee.txt');                            % Signal Vector

with this one:

signal = filtered_signal;

[rohith bharadwaj]

thanks for the help. but can I get the code like, enter the frequency to remove: like scanf in C and then get the reconstructed signal by removing this frequency

[Star Strider]

You could use the inputdlg function to set the centre frequency, then add that to the passband and stopband offsets to create the appropriate passbands and stopbands.

Example —

ctrfreqc = inputdlg('Enter centre frequency (Hz): ', 'Bandstop Filter', [1 40]);
ctrfreq = str2double(ctrfreqc);

... CODE ...

Wp = (ctrfreq+[-0.02 0.02])/Fn;                         % Passband Frequency (Normalised)
Ws = (ctrfreq+[-0.01 0.01])/Fn;                         % Stopband Frequency (Normalised)

... CODE ...

I tested the code that I posted here. It should work with the rest of my code without change.

[rohith bharadwaj]

Thank you so much. THe code is working and I'm getting the output. Thank you :-)

[Star Strider]

My pleasure.

Since my Answer solved your problem, please Accept it!

Answer 2

[Image Analyst]

Find the index for that frequency, and set it to 0, then call ifft().

See attached demo (though it's for a 2-D image, not a 1-D signal though the concept is the same).

[rohith bharadwaj]

thanks for the help. but can I get the code like, enter the frequency to remove: like scanf in C and then get the reconstructed signal by removing this frequency