Frequency domain mask and limit testing.

[trosier]

I thought this would not be this complicated but I am turning to the forums for a reason. I am trying to perform frequency domain mask testing on a trace read from a spectrum analyzer. I have the trace data along with the upper and lower mask limits. One problem I am having is that the spectrum analyzer trace has frequency and power arrays and the graph is plotting them accordingly. When this happens along with plotting the mask limits which obviously start at 0 the waveforms are not lined up. If anyone has any ideas about how to compensate for that offset I would appreciate it. The second part and for me the most frustrating is trying to use the Mask and Limit Testing VI. I thought this would make this testing pretty simple but I cannot get it to work correctly. It seems that it is always using the time domain which for this testing does not work. Are there any examples out there describing how this VI would work using the frequency domain and how to configure it? I want to use this VI because I do not want to re-invent the wheel for something NI is already providing for me. Any help with this will be much appreciated.

Thanks in advance.

[Starla_T.]

Hi trosier,

You post a good question.  I haven't personally used the Mask and Limit Testing VI with the Frequency Domain.  I do know that the Limit Test Express VI will accept signals in time or frequency domain.  There is also a good tutorial and example on how to use these VIs here.   Do you have the sound and vibration toolkit?

Best Regards,

Starla T. 

[trosier]

I do not have the sound and vibration tool kit nor do I believe I need them being in the Satellite/RF test industry. This is the second time I had to perform spectral mask testing and each time it is suggested to use the before mentioned VIs that come with Labview and I can never get them to work. Every example I find is using the time domain. One of the biggest problems I am seeing especially with the Mask and Limit Testing express VI is that it is not plotting the frequency points correctly or the power points. For some reason the output of the tested signals is omitting some of the points I am giving it when configuring the mask manually. When configuring the mask it looks correct on the output example but when it is run the mask come out differently. I have attached a .png of the output I am trying to obtain. Do not pay attention to the labels on the graph as that is the easy part to fix after I have the VI running the way I want. The X axis is frequency and the Y is power. This is done by running a cluster of arrays into an XY Graph. Each array is a two dimensional array that contains x and y data. On another note if I use the Limit Specification and Limit Testing VIs explicitly the arrays are treated as two plots each. So for each two dimensional array there are two plots on the output instead on one. I have been all over the place trying to make this work and I cannot find a solution other than just to perform this VI development manually. That works I just thought from looking through these VIs provided by NI that this task could be done in a more out of the box kind of way.

[Online_Courses]

Hi trosier,

Have you tried converting the signal to the time domain and then inputing it into the Mask and Limit Testing VI?  As far as I know, the Mask and Limit Testing VI will only accept time domain.  There is a Specral Measurement toolkit for LabVIEW.  I've personally never need to use it since I don't do much specral analysis, but it might be worth looking into if you're looking for VIs with more functionality for specral analysis.  

Best Regards,

Starla T. 

[Dennis_Knutson]

I don't think it matters at all whether it is frequency or time domain, The mask is no more than a set of x and y points. What I don't understand is talk of an XY graph. You cannot input an XY graph to the Limit Test function. The input is a single waveform data type - either time or frequency domain. The Express VI uses dynamic data and if you wire your XY graph cluster, I have no idea how it would be converted. Just another example of how evil dynamic data can be. If you convert the XY graph to a normal graph with a regular df component, I suspect you will have much better luck.

[trosier]

All, thanks for the replies. Being that I can do this myself I cannot justfy buying a toolkit for Labview. What I have is working I was just trying to explore other options that came with Labview and they are not obviously designed to do what I need right out of the box as far as RF analysis is concerned. Again thanks for all the replies.