fft of higher data

[Dost]

I have about 40,000 data point of the data that is sampled at a rate of 1000 hz.I want to do FFT on this data.I used the FFT power spectrum VI,But I found that the frequency range that it should give theortically is not correct i expect the frequecny range of fs/2= 500. but it give me 0.5, then the frequeccy interval should be fs/4000 i.e 1000/4000 =.25 but  it does not give it like this.will any body here on this form will guide me how to do FFT with this data point and how to set the correct frquency range

Dost

[altenbach]

Could you attach your code so we can have a look?

Which VI are you using (express, waveform measure, etc.)? Does you waveform have time information or are you feeding it a plain array? 

[Dost]

I have used two VI in order to check my result.these are :Powe spectrum.VI  and FFT power spectrum separately.I have my data in the text file so first I take off my require block of the data from my text file and then pass this to the both VIs.I also used Nysquit VI in order to shift my data DC central spectrum.As u can seen in the vi attached as gif file that I also have used bundle before fedding this in to wave form .I have attached the out put in out put.gif.looking forward for ur suggestion

dost

[Dost]

I have used two VI in order to check my result.these are Power spectrum.VI  and FFT power spectrum separately.I have my data in the text file so first I take off my require block of the data from my text file and then pass this to the both VIs.I also used Nysquit VI in order to shift my data DC central spectrum.As u can seen in the vi attached as gif file that I also have used bundle before fedding this in to wave form .I have attached the out put in out put.gif.looking forward for ur suggestion.Please note that my data set is about 2200 in this example.

dost

[altenbach]

I don't see anything reasonable in your code where you set the dx for the display of the various outputs. You need to set df and dt.

For an idea, have a look at figure 25 in the following link: http://zone.ni.com/devzone/cda/tut/p/id/4541

Also, what is your LabVIEW version? If you built your data as proper waveforms or dynamic data with timing information (instead of plain arrays), this will be handled automatically by the express VIs.

Notice also that your loop spins at full bore, recalculating the same old data "millions" of times per second. It would be beneficial to place a small delay inside the loop or even use an event structure so the calculation only happens when the index changes.

[Dost]

can u please tell  me where i can get raw data vi

[altenbach]

You don't need it. Substitute your real data instead. 🙂

[Dost]

Please see the attached files I have used as per  the figure 25,I used del T= .001 becaz I sampled the data at  a rate of 1000 hz.I have store the data in the text file for about 60 second saving one every milli seconds.it gave ne a data set of about than more 50 thousands, and using this value of N i found Del F 0 .0182 ( any comment) .can u checke my out put gif and tell me what is FFT analysis produced by power spectrum I think it is given me a amplitude at zero frequency, is it two side power spectrum or single side spectrum.In fact i expect to see more peaks but i coudl nt find any comments

[altenbach]

You get a large component at zero frequency because your data has a DC offset!

I don't do well with pictures. Why don't you run your VI, make current values default, and attach the VI?

[Dost]

Thank you very much for your help.here is attached Vi ,Please tell me what do u mean offset Dc value is it noise?

morever power spectrum VI gives me two side spectrum or onde sided according to help it calcultes by taking the product of fft and its conguate.what i need to do if want to use PSD.I really appreciate your kind help.