FFT measurement from a spreadsheet file

[superfly]

At the moment I'm doing research to brainmeasurements. People in my unit did some measurements at a sample rate off 25 Hz. I'm reading the data that they collected by using a  "read a spreadsheetfile" . What I want too do now is to perform a FFT on the data that they collected. The data that I try to analyse is 1 signal at a time. The range that I'm interested in is from 0 till 1 Hz. So we are looking at the signal in the very low frequentie range. I tried some things like using a spectral measurements application but than I don't get a nice FFT out. I tried too make the data running by sending the data trough a for loop. This showed a total different view. But now we had also negative amplitudes in some frequenties. So looks like this it not the way.
I posted also some screenshots. In the file frontpanel 1 you see the original signal. In 2 you see a selection that was taken out the original signal. and in 3 you see the 2 different FFT's that I tried.

Can someone give me an idea about what I'm doing wrong??

[BasvanDijke]

Hi,

Could you post your data and some code?

Thanks,
Bas

[superfly]

This VI represent my problem.

Thanx allready for responding.

[johnsold]

You have a very large DC offset in your data which tends to mask the variations. Subtract it out before doing the spectral analysis.

Lynn

[superfly]

Hi,

Thank you for helping with that problem. But then I have another question. If you look at the X-scale, what does the numbres represent. Is it the time, is it the frequentie and why doesn't it continue after 0,5?

I ask this because if you look at the FFT you see a peak representing everytime after I geuss a few Hz, but the amplute is half the value off the peak before this. Maybe also a solution for this problem?

Thank you again.

[johnsold]

The X-axis in your data appears to be sample numbers (0, 1, 2, ...). The Spectral Measurements VI expects a waveform datatype as its input. The waveform datatype contains a time increment component (dt) as well as the array of Y values. If you feed it the proper dt, it should give the proper frequency values on the output. When I told it the sample rate was 25 Hz, the X-axis vales no range from 0 to 12.5, as expected. The first peak is just above 1 Hz.

Lynn

[superfly]

Thank you very much, you helped me a lot.

You're a star.