# How to properly measure signal power with FSW signal analyzer

I am using a R&S FSW Vector Signal Analyzer as RF capturing device (using the IQ Analyzer mode and transfering IQ data via SCPI).

In lack of a powermeter I also want to use it to measure the signal power. It sounds like a trivial thing to do but I do not find a good way to do it.

Since I have the device in "IQ Analyzer" mode for capturing data it would be great if I could do the measurement in this mode and read it out via SCPI.

If not, I could do Spectrum Analyzer mode and use ACP measurements. This gives the expected results but the setup is complicated and above all, I want the total signal power, not just a channel (i.e., integrate the entire frequency range or time domain power).

If I use "Time Domain Power" measurements, the result is as expected if I use CW. But the result is completely off (in my case >10dB less) if I use a wideband signal.

(I am not looking for a manual way to read off power from a generic signal analyzer, e.g. by placing a marker and multiplying by the number of ResBWs in my signal band etc. I am looking for an automated way to read out the total signal power in dBm via SCPI)

• I just realized the I/Q data I get back from the FSW seems to be scaled in voltage. Is it correct to just calculate the variance of the I/Q samples and convert to dBm, i.e. pwr = 10*log10(var(iq_data)/50*1000) ? (and I guess this requires that all the signal power falls into the analysis bandwidth of the analyzer) – divB Oct 14 '18 at 22:23