FFT (time x point)

[vitor22]

I´m trying to make a FFT of a data adquired from an accelerometer. Attached is the VI that I use to open the data saved from the accelerometer and the lvm (converted to txt) obtained from the accelerometer. One column of the data saved is time and the other is acceleration (g). I want to do a fft of the acceleration. The time spacing between the accelerations is not equal. Somebody can help me? Thanks.

[vitor22]

I forgot to say that I tried to use the "Unevenly Sampled Signal Spectrum VI" but the result are not right.

[vitor22]

Somebody?

[johnsold]

When you say the "results are not right" it is not very helpful.  How do you know the results are not right? Do you get an error message? What was it? What do you expect the results to be?

I think the biggest problem is that you have an offset of 513 on a signal with a peak to peak amplitude of 12.  When I subtract the mean before performing the spectral analysis, I get a result with peaks at 6, 56, and 68.  When I go back and look at the original spectrum, those peaks are there.  The DC component has an amplitude of ~260000 while the component at 6 has an amplitude of ~88.  Change the Y-axis scale to 100 to see these small components.

Lynn

[vitor22]

Thanks Johnsold for your answer. You are always helping me.
I was not seeing the small peaks. I did with the scale what you said and now I can see the peaks. Do you know why im getting this 260000 amplitude at 0Hz? The way Im doing (file attached teste  2.vi)using the "Unevenly Sampled power spectrum" it´s correct isn´t it?

Thanks

[johnsold]

I did not look closely at the data, but the connections to the Unevenly Sampled VI seem OK.  The result looks like a spectrum, although I have no idea how correct it is.

The reason for the 260000 is that it correspnds to 513^2, the offset in your signal.

Lynn