AnsweredAssumed Answered

Analyzer uncertainty under specific conditions

Question asked by daldrich on Jun 6, 2011
There is a requirement to measure the insertion loss of 2 adapters used to calibrate high frequency power sensors.  In the most recent case, it was an R-Band to 2.4mm.  The waveguide ends of the adapters were connected together and the other ends of those were male and female 2.4mm.  Our E8364C has option 014, if that makes any difference.  Anyway, I did a (2.4mm) full 2 port cal, nominalized the "Thru" with an S21 measurement and then inserted the adapter set up and took the insertion loss readings.... the frequency range was 26 to 40GHz and the calibration was done in a 10Hz bandwidth.

Using the Uncertainty Calculator, the Magnitude Accuracy uncertainty at 0 dBm in the 20 to 40GHz range is 0.142 dB or something close to that.  Is this really the number to use.  For me, it seems more like a repeatability issue as opposed to a magnitude accuracy issue.  What would you say the measurement uncertainty is in this situation?

Second condition is similar to the first except this involves the calibration of a power splitter..... say, 11667C.  After the 2 port cal, nominalize one leg of the splitter and compare it to the other leg.... there is a 6 dB insertion loss with each leg, but is that really the measurement I am making.  It seems again, to me, like a repeatability issue, measuring the same 6dB value.