OBW -20dB down

[Kevin.Shirey]

NI Only provides an OBW measurement for 99% power, not in dB. So how would I make a measurement of OBW -20dB down from the peak if I have IQ data? 

[Alex.W]

Hello @Kevin.Shirey,

Could you please upload a picture of the power spectrum of your signal so we can better understand what you are trying to measure?

Could you also upload your current code or a picture of your measurement code?

[Kevin.Shirey]

Certainly. I don't have any measurement code yet as that is the question I am asking. NI has 99% power, but nothing for 20dB OBW. What is the proper way, in agreement with FCC standard DA 00-705. IQ, IQ Carrier 905M, Center 915M, IQ Rate 33.554432M, clock 10M. Spectrum Zoom Power Center 915M, Span 10M, RBW 1024, dBm, pk. I attached images and the Spectrum data in XY.

[shawn.t]

Is this the standard that you're looking at? https://transition.fcc.gov/Bureaus/Engineering_Technology/Public_Notices/2000/da000705.txt

If so, I don't think any of the standard spec-an measurements will work out of the box for it. A good starting place might be to do exactly what it describes under "20 dB Bandwidth". Use NI-RFmx to read the spectrum using the settings listed in the standard. Take the spectrum data and use the array max function to return the peak. Search the array again for the first and last point within the -20 dB from peak bandwidth. Use those indices multiplied by dx plus x0 (all from the spectrum cluster) to get your start and stop frequencies. The difference is obviously the bandwidth.

Disclaimer: I am not familiar with the details of calculating an OBW measurement so I don't know if any additional filters should be applied to the data. This is just what makes sense to me.

[Kevin.Shirey]

Yes, correct standard with 20dB down Occupied Bandwidth.

Making the measurement on the Spectrum, no problem, that is working similar to how you described. The problem I am having is if I acquired the data in IQ vs Spectrum. I do not get the same answer in IQ that I do in Spectrum acquired data (IQ data is very far off). The same Spectrum acquired settings with respect to Window, RBW, Span are applied to the IQ data and it simply doesn't match causing the measurements invalid. I can see differences in the spectrum itself. I need to keep this in IQ to use this as a driver for those using USRPs and other measurements I am making from the IQ. 

I suppose the major difference in acquiring in Spectrum is using the peak hold method over the same time period as IQ. There isn't anything like that for IQ --> Spectrum that I am aware of.