The impressive accuracy of the nanoVNA compared to much more expensive VNA's had me wondering how the output of the device looked compared to the FieldFox I have access to. ("You ended that sentence in a preposition, John" - voice of 3rd grade teacher, Mrs. Rogers, in my head).
Once the return to normalcy occurs (if ever - I believe "normal" will be re-defined) I'll have access to a real spectrum analyzer - for now, the SA function/software for the RSPduo will have to do. At the very least it offers a comparison, if not absolute values, of what comes out of each VNA with a given sweep range programmed.
A step attenuator was used to lower the output of the nanoVNA. The FieldFox has two selectable output levels with the lower one being chosen for this test at three different frequencies.
In the case of both the nVNA and the FieldFox, a certain amount of time (5 to 30 seconds) was required to paint the resulting square wave that covers the 2 MHz-wide frequency range programmed in as a sweep. The FF did this much faster than the nVNA and also painted a more uniform square wave with less noise above and below the swept range.
The FF painted this range with what appeared to be a single narrow pulse; the nVNA did it with numerous simultaneous pulses of widely varying amplitudes.
In each of the images below, the nVNA result is on the left; the FF is on the right:
|49 to 51 MHz|
|499 to 501 MHz|
|999 to 1001 MHz|