LabVIEW

cancel
Showing results for 
Search instead for 
Did you mean: 

Amplitude of Fundamental after FFT

HI all,

I have a tdms file and reading this using spectral measurement block I am trying to get a plot of the FFT of it -no averaging no window linear scale magnitude (peak) settings selected- but the initial acceleration over give me a spike of almost 0.9g and the fundamental on the fft is barely reaching 0.3g @14Hz, thus i am not really convinced that this is write. Should i mention as well that using diadem i am getting aproximately 0.3 and when iam trying to plot it in labview i get 0.18g and if i seperate the time span that the spike takes place and plot it I am getting round 0.3. So i don't get the difference between the two -diadem and labview- , I don't get why i get two different values within labview and I am not convinced that what I am getting is the correct value for the fundamental. Is there any factorisation that takes place or something??

I am attaching all the files I am using. Could you please tell me if I go wrong at something?

 

Thanks 

Download All
0 Kudos
Message 1 of 4
(3,323 Views)

to add up something more to this there is an excel file that I made for FFt that looks much better and close to what I expected...So there must be something wrong that I am doing in excel but I can't figure it out....Could anyone please help...

 

 

Thanks

0 Kudos
Message 2 of 4
(3,318 Views)

Hey Aris 

 

This is due to that fact that each time you are analysising slightly different windows of data. This means that infact all results are acurate but obviously with different data in you are going to get different data out. 

 

Hope this helps,

Matthew Trott
Applications Engineer
National Instruments UK
www.ni.com/ask
Message 3 of 4
(3,283 Views)

Aris,

 

When the frequency of the signal is is not an integer factor of the sampling frequency, the peak value in the spectrum will not fall exactly on one of the frequency bins.  So the amplitude of the largest spectral element is not the same as the amplitude of the fundamental component of the signal.

 

One indication that this is happening is that the spectral peak is not symmetrical and that the energy does not all fall into one bin. The image below is from the data in your Excel file after a standard FFT in LabVIEW.

 

Spectrum.png

 

The frequency has not been scaled to standard frequency units.  Notice that the maximum data point has an amplitude of 650. The two points to the left of the peak are both higher than the first point to the right.  About 10 points are significnatly higher than the baseline.  The actual value of the fundamental is some composite of those 10 points.

 

Lynn

Message 4 of 4
(3,266 Views)