Customer Update:
We realize that the spectrum analyser is being overdriven in the plot that shows the "mound" of broadband noise between 120 and 1250 MHz, and have considered whether the mound is composed of mixing products from within the analyser. What makes us believe the mound is really coming from the CC1125 and not being generated in the analyser is: The mound is exactly the same shape, width and height when measured on our Agilent analyser as in T.I.'s Rhode-Schwarz analyser. I would think that the R&S would be better than, or at least different from, our Agilent.
If the noise were band-limited to regions located away from the 902-928 MHz ISM band, we may be able to filter the noise out before amplification. Unfortunately, it appears that the noise continues through the ISM band.
Please advise on the availability of test data for a 1-watt design using the CC1125, as our development-cycle schedule is at a critical juncture.