MATLAB: Butterworth filter cutoff frequency calculation

butterworth filter

Hi All, I am not a electrical engineer so my question may sound stupid. Sorry about that. I have a some noisy data and I want to filter my data using a butterworth filter of order 2. I am having trouble finding the cutoff frequency for the filter design. How do I calculate cutoff frequency? I am sampling my data every 10 seconds. Thnanks.

Best Answer

First, a Butterworth (or any) filter of order 2 is not likely to work as well as you want it to. You may find a Chebyshev Type II design more appropriate, especially with your extremely low sampling frequency.
Second, the best way to design a band-selective filter is to take the fft (link) of your signal to determine where the signal frequencies are and where the noise frequencies are. Use that information to design your filter. Note that if you have broadband noise, a frequency-selective filter will not eliminate all of the noise. You might want to eliminate baseline offset and low-frequency baseline variation as well, the reason I usually use a bandpass design.
The designfilt function can make filter design straightforward. Another acceptable way to design a Butterworth filter is this bandpass prototype:
Fs = 0.1; % Sampling Frequency (Hz)
Fn = Fs/2; % Nyquist Frequency
Wp = [0.002 0.010]/Fn; % Specify Bandpass Filter
Ws = [0.001 0.015]/Fn; % Stopband (normalised)
Rp = 10; % Passband Ripple (dB)
Rs = 30; % Stopband Ripple (dB)
[n,Wn] = buttord(Wp, Ws, Rp, Rs);
[z,p,k] = butter(n,Wn); % ZPK
[SOS,G] = zp2sos(z,p,k); % Convert To SOS for Stability
figure(2)
freqz(SOS, 2^16, Fs)
To design a Chebyshev Type II filter, replace buttord with cheb2ord and butter with cheby2. The arguments are slightly different, so keep that in mind. The rest of the code does not change.