I am trying to measure the rf power of a homemade 120GHz transmitter (diode tripler). I have a Tektronix 492b spectrum analyzer with diplexer. I want a ballpark rf level by working backwards from the 492 display which is -40dBm. The issue I have is documentation indicates either a conversion loss around 20 or 30dB, depending on which chapter you read! (Old doc Tek 70-03731-01) 1.7 says 30 dB up to 60GHz using 10th harmonic and page 1.20 says 30dB for 120 to 240GHz which can't be the same as page 1.7 because of a much higher 23rd harmonic is used. I would like to get to +/-3 dB accuracy.
Anybody have experience with conversions at 120GHz that knows more or better.