Power amplifiers characterization automated test bench

Under construction...

IMD bench setup

E4406A VSA + 2 x HP8657B + HP6632B + GPIB card + homemade control SW in Python

IMD test bench

RF sources output level correction

The good old (and relatively cheap) HP 8657B signal generators do not have an extremely accurate RF output level; they are specified for ±1 dB from -127 dBm to 3.5 dBm and can output nominally up to 17 dBm with unspecified accuracy.
As usual for HP specs, they are normally much better than that but a little characterization and SW correction can greatly improve their accuracy.

The following pictures show the RF output power measured by the E4406A at the output of the combiner: first picture show the raw, uncorrected output: HP8657B #1 RF output not corrected Mean losses through the combiner are about 4 dB for this path and there is about 0.5 dB of maximum deviation due to the generator output uncertainty.
After SW correction the maximum deviation is about ±0.06 dB; cannot be much better than that since the 8657B output level resolution is 0.1 dB;. HP8657B #1 RF output after correction

Same for the 2nd generator: HP8657B #1 RF output not corrected HP8657B #1 RF output after correction