0
$\begingroup$

I have data vibration in excel, the first row is time, and the second is vibration. I want to use fft to analyze the vibration. Because, i'm new to dsp, i don't know, what is sampling frecuency, etc. I searching and i found http://www.stem2.org/je/Excel_FFT_Instructions.pdf, but i still confused. Then, i searching again how to do fft in matlab using imported data from excel, and i found https://www.mathworks.com/matlabcentral/answers/329302-how-to-perform-fft-in-matlab-using-a-set-of-data-from-excel-sheet, again i don't know the sampling frequency and the other. How to do it? Thank you.

Thanks to @MattiP., i've been able to calculate fft using excel, now here is the graphic, how to know / analyse this?

enter image description here

$\endgroup$
  • 1
    $\begingroup$ The sampling frequency is the inverse of the time step in your data. For example, if the first data point is at $t=0.001$ and second is at $t=0.002$, the sampling frequency is $1000~\text{Hz}$. $\endgroup$ – Matti P. May 15 '18 at 7:58
  • $\begingroup$ @MattiP. , know i've been able to calculate fft, thanks for your previous comment! But the problem now is, how to analyse those spectrume? Is there any book/lecture note which contain about fft and explain briefly how to analyse it? Thank you! $\endgroup$ – Reza Habibi May 17 '18 at 3:20
  • 1
    $\begingroup$ The Fourier transform is a powerful tool, but in practical situations it can be tricky to analyse if you don't have much data. You should have a relatively high sampling frequency and a long measuring time. For the analysis, you have to carefully choose the frequency area that you are interested in. The parameters of the measuring itself impose limitations to the frequencies you will be able to measure. Sometimes you are only interested in only high frequencies, sometimes only low frequencies. I would also recommend plotting the frequencies on a logarithmic scale. $\endgroup$ – Matti P. May 17 '18 at 5:27
  • 1
    $\begingroup$ So basically you are just finding the largest peaks of the transform. That's it. The resulting graph you posted doesn't seem to have very large amplitude peaks, except maybe at the very lowest frequencies. $\endgroup$ – Matti P. May 17 '18 at 5:29

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.