The idea behind this question is for misalignment fault we say vibration signals across the misalignment are out of phase.

If the signals are just sinusoidal signals with one frequency calculating phase difference can be calculated via multiple ways like cross correlation,fft. But what if the signals has multiple frequencies and noise like real vibration signals how to get relative phase?

  • $\begingroup$ If they are both periodic signals (which is a Fourier Series of harmonically related sinusoids), the circular cross-correlation (and you would determine the period from the windowed auto-correlations of each signal) and look for the lag with the maximum positive correlation. If they're really periodic and the same period, this should be clean and easy. $\endgroup$ Feb 23 at 6:10


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Browse other questions tagged or ask your own question.