ifft of power spectrum

[11212]

Hello,

 

I have power spectrum of spectral limited pulse (phase is flat - constant)
how to make ifft of such data.

 

I have tried all labview ifft *.vi , but all results are wrong.

 

I'm looking forward to hearing from you.

[Will_S.]

Could you please describe what you mean by the results are wrong? The following community example may be helpful:

http://decibel.ni.com/content/docs/DOC-6390

[11212]

I have a spectrum and i want to make inverse fourier transformation (phase is constant).

Here are my results that i achieved with labview:

http://img38.imageshack.us/img38/8164/labviewresults.jpg

I am converting spectrum data into power data using inverse fourier transformations, however results are wrong. They should look like these graphics:

http://img32.imageshack.us/img32/5629/dataandresult.jpg

As you can see power graphics do not match.

Thanks for your help.

[johnsold]

I do not see much difference.  The X-axis is zero at the center of the OriginPro and at the ends on the LabVIEW graph.  If you rescale so that both are set up the same way, the time domain signals will likely be quite similar. The time scales on the spectrumgraphs also differ.

The spectrum looks like an FM signal of some type.  What do you expect the time signal to be (on theoretical grounds, not based on what some other program produces)?

Please post the text file containing the data.

Lynn 

[11212]

Here is my data file

http://rapidshare.com/files/271262535/spectrumdata.TXT.html

I hope you will be able to achieve better results. Thanks in advance.

[11212]

Maybe this is correct ?

http://img166.imageshack.us/img166/3342/result.jpg

[Coq_Rouge]

Take a look here, at my last postings. It may help you. 

http://forums.ni.com/ni/board/message?board.id=170&message.id=435870#M435870

[11212]

Sorry, but it didn't help. If you can please take a look at my previous posts, maby they are working correct ly?

[Coq_Rouge]

I think you have some holes in your FFT knowledge ;). In order to get your spectrum correct you must provide information regarding the timing. In labview the dt value is often used. This is equal to 1/(sample frequency). If you take a look at my solution that is meant to be a sub VI. You must give this information in the control labeled "Sample period dt (sec)=1/f sample"

Good Luck

[11212]

I understand that i need df, but now i am looking only at ifft.vi result (that vi use only FFT(X) values and gives X values).