On Tuesday, I tried running the HAM7 high voltage CPS electronics again. This time I was able to get all 6 sensors to the noise floor I expected. I replace another ethernet cable and one of the in-air triax cables, and I had to use one of the test cards the RichM used at MIT to get there though. After that I was able to turn the isolation loops on let the ISI run for a while when we were trying to lock. My initial measurement looked like the low frequency isolation was pretty bad, so I reverted to the old electronics that afternoon.
Attached first 3 images compare the performance with the high voltage cps (refs) and the nominal cps (live traces). The high frequency performance is not really any better, and in RX and Y the below .3hz performance seems much worse. Z seems to be worse pretty much everywhere below 10hz. Maybe one of these sensors has a funny calibration or something? I will probably try again next week and see if I can measure that with some ISI tfs.
Fourth image is calibrated spectra for each individual hv cps. The H1 still had a high noise floor for this test. Another thing to try to fix next time.