Good evening gentlemen,
Thank you all for your contributions to this thread, and for your encouragements!
I already spend quite some time digging in the cal data structure. I think I begin to understand how they are used by the equipment, at least for part of them, most in the Rx/SA path. As you may guess, there is still a lot of work to do: this began this as a rainy winter project, continued as a shity spring projet but that was put on hold by a sunny summer!!! I'll continue this work when time permits, that is not in the short term...
With regard to spectrum analyzer calibration and measurement accuracy, as you know, HP8920 architecture is modular: i.e. each module (Input atten, input section, 1st miser, receiver, measurement, SA, etc.) are I beleived all calibrated separatelly and their calibration data is brought into the equipment either using a memory card (as confirmed here by some of you) or for some assemblies by some onboard eeprom.
If this assumption is correct, Spectrum Analyzer module is only one piece of the puzzle, and at the end of the day, overall measurement accuracy will depend on cal accuracy of all modules in the analyzer path (Atten/RF Input/1st mixer/ Part of Receiver). If any module calibration is not right, the result will not be right and tweaking SA module cal may not be the right thing to do... I still need to think a bit more about how to manage this the best...
Regards,
Jeff
