# bearing frequency detection

I am working on bearing faults detection. I am doing the following steps:
signal->mean zero->high pass->hilbert->y->fft->cepstrum of y.

my code is:

    analy = hilbert(Signal);
y = abs(analy);
N = 2*2048;T=N/Fs;
sig_f = abs(fft(y(1:N)',N));
sig_n = sig_f/(norm(sig_f));
Signal_fft_hilbert = sig_n;
freq_s = (0:N-1)/T;

figure(3)
plot(freq_s(2:250),Signal_fft_hilbert(2:250))
title ('Average FFT after Hilbert enveloping');
xlabel('Frequency in Hz')

figure(4)
c2=cceps(y(2:250));
plot(freq_s(2:250),c2)


The problem is I am not able to extract the fundamental frequency out of the spectrum or cepstrum. my bearing frequency is 42 Hz.I am not getting prominent peaks in this region. Please suggest something.

• The spectrum appears to have a fundamental frequency and third and fifth harmonics -- all of which are below 200Hz. Are you sure the bearing frequency is 42Hz? There are other frequencies related to it (electromotores.com/PDF/InfoT%C3%A9cnica/EASA/…: FTF, BSF, BPFO, BPFI that may be relevant). – Peter K. Mar 3 '14 at 12:56
• well yes the bearing frequency is 42 Hz.and I am taking the cepstrum after fft. Is that correct? Or i need to take it after hilbert transform? – user3217416 Mar 4 '14 at 1:32
• Oh I get your doubt. It is the fundamental ball frequency. – user3217416 Mar 4 '14 at 7:31