I did notice I made two unintentional errors as I was speaking. Brain freeze. It happens to us older folks :-)
1:15- I meant to say, “Two tones, not harmonically related”
8:06-the difference between the peak and noise floor in that particular test scenario is 50db. I said 30db. With the CORRECT input attenuation , you’ll find the same 50db difference between peak and noise floor across various spans, as shown in this video; 5khz, 270mhz and 1 ghz. The slight variation in attenuation (48db vs. book spec of 50db) is mostly due to coax connections. And that variation across all 3 of my spec analyzers can easily be factored and accounted for with the manufacturers’ coax specs for db loss at a particular frequency, Velocity of Propagation, etc. Generally anything down in the HAM bands will have minimal effect on loss due to the lower frequencies being used; especially when using such short jumper coaxes across the bench. If a UHF radio was being tested, signal loss, VOP, etc. would need to be closely monitored to achieve an accurate measurement result.
The above procedures are taken directly from the HP8920A user’s manual.
I also use my Rohde and Schwarz CMU 200 spectrum analyzer to verify compliance with 14 CFR Part 95 for the adjacent band harmonic attenuation requirements. On that spectrum analyzer, I have my input attenuation set at 50db, which allows me to see the 65db attenuation difference between the 27mhz band and the 54mhz band.