Numerical values of harmonics using Spectral Measurement with DAQ Assistant (USB 6009)

[LG_NI]

Hello.

I used the Spectral Measurement to get the FFT of a signal coming from the DAQ Assistant (USB 6009). The graph itself is fine, but the numerical values of the harmonics do not correspond or do not make any sense.

What could I use to get the numerical values of the harmonics?

Any support or comments will be appreciated.

Thanks…

[Bob_Schor]

You've provided no useful data -- no code (a VI would be nice), no picture of your Front Panel showing the graph and numerical value of the harmonics.  How many samples, and at what sampling rate, did you take?  What was the input to the FFT routine?  How do you expect us to give you useful help if the only data that we know is that you are using a USB 6009 (so what?) and are using the DAQ Assistant (which suggests you are very new to LabVIEW and might not have taken the Tutorials or gotten any class instruction)?

Provide more information so we can do more than complain that you haven't told us enough for us to help you.

Bob (Designated Curmudgeon) Schor

[LG_NI]

Yes sorry about that, I was in a rush. I am a beginner in LabView and I have studied the tutorials but not enough, however I am using the LabView book (by John Essick) which has been really helpful.

Please review the attached pictures (block diagram and front panel)

This is the information:

-DAQ Assistant configured for input voltage and differential connection, using the USB 6009.

-External function generator for the main signal: square/sine waveform, 50HZ, 10VPP.

The graphs show the values of the harmonics, but I am trying to get the numerical values of them.

Thank you for your support.

[Hooovahh]

http://forums.ni.com/t5/LabVIEW/input-terminal-configuration/m-p/3199107#M927339

The 6009 is a very slow cheap DAQ device.  There is no hardware sampling, so the time between samples will be very inconsistent.  On top of that you have a random 900ms wait in there which will mess with your software sampling.  I'm not certain you will get the data at the rate you need consistently enough to find harmonics.  

Actually looking at your graphs it doesn't look too terrible on the raw sampling graph.

[GerdW]

Hi,

well, the AI on the USB6009 support hardware sampling…

The images look quite ok: you get 100 samples of a ~50Hz square signal, taken at 1kHz sample rate.

 This results in a FFT signal from 0 to 500Hz, with main frequency at 50Hz and a lot of harmonics…

What do you think is wrong here (apart of Hooovahhs comments)?

I am trying to get the numerical values of them.

This would be so much easier if you would get rid of all those ExpressVIs: all they do is hiding the underlying datatype from you!

Connecting a scalar numeric indicator to a DDT is (close to) nonsense…

[Hooovahh]

---

@GerdW wrote:

well, the AI on the USB6009 support hardware sampling…

---

Does it?  I've never used one myself but I was under the impression that it didn't support any hardware timing or finite sampling requiring a hardware timer.

Now that I read up I think I was confused, and the AI supports buffered reads, but the AO does not support buffered outputs.  Sorry for the confusion.