spectrogram with PXIe-5646r VST

[Jdogg]

Hello I am using a PXIe-5646r vector signal transceiver to take spectral measurements. I am using it similarly to a spectrum analyzer, as in I’m displaying the FFT and a waterfall plot. 

My problem is the waterfall plot or spectrogram. I can get something that looks and behave similar the spectrum analyzer, using a intensity chart, however every time the chart updates I get a discontinuity of some sort have attached a figure to illustrate this issue. 

What I’m looking for is something similar to what I have seen in software defined radio (SDR) setups were you have say the FM frequency band  you have a waterfall display showing what frequencies are transmitting. I have also attached a screen grab of this. I haven’t been able to find a LabVIEW example of how they do this so I’m not sure where to start.

Also I have attached my attempts to use a STFT vi. To display the waterfall display using intensity charts. This is a modified version of the spectrogram and spectral measurements vi =. Any help or directions to go with this would be much appreciated

Thanks

Jeremy  

[PeanutButterOven]

Jdogg,

The VI you posted has a missing dependency/sub vi. So feel free to zip the example and any sub vis and re post. I have also attached a link to an example that uses the Gabor Transform to create the waterfall graph. See if that can help get you going in the right direction.

waterfall example

Mark B.

[Jdogg]

Hello Mark,

I apologize for that here are the sub Vi that I’m using. You may still not be able to run it because im using the advances signal processing toolkit and the RFSA toolkit.

I played around with the Gabor transform waterfall graph however my issue is I need to display essentially the history of the spectrogram. In other words the time frequency analysis as it evolves in time, like you would see if you set up a spectrogram display on a spectrum analyzer. The Gabor and STFT transforms don’t display this when using the intensity graph, and if I use the intensity chart with these transforms I get ringing or a discontinuity at the beginning and end of the sample as its added to the chart you can see this as the vertical lines in the (waterfall plot with intensity chart) screen shot that I attached earlier.

I have been playing around with adding averaging and padding but this could reduce the sensitivity and we are looking for some small signals.

thanks 

Jeremy

[dbell0971]

Hey Jeremy,

I am not 100% sure about the source of the discontinuity, but it could be related to how you are sampling the data.  It looks like you are taking 2 ms chunks of data for each capture, then trying to string them together (see the Configure Number of Samples VI) .  The problem with taking 2 ms captures is that the you will miss data in between captures.  There are two problems with that:

1)  Missed data-  I don't know anything about the signal of interest here, but the signal could appear and disappear during the down time of the VSA.

2)  Change in frequency content-  During that down time, the frequency content of your signal may have changed substantially.  If it has, it would likely appear as a rapid jump/discontinuity on your chart.

I would recommend performing a continuous acquisition as in the attached example, then add a parallel loop to perform the FFT's over time.  Again, not guaranteeing this will fix the problem, but likely worth testing.

Thanks,

David B

National Instruments

Applications Engineer