Craig Jenne Sheila

We were seeing a huge 120 Hz peak in pretty much all of our IFO witnesses, including the DARM spectrum, REFL A LF, REFL B LF, POP LF, REFL A 9, REFLAIR A+B LF, etc...
This caused REFL A LF to saturate the ADC and report negative values, even though the dark offsets had not changed.
Bad REFL A LF values caused the guardian to freeze at CARM_5_PICOMETERS where the requested CARM offset was too large (we weren't actually going to zero CARM offset and the guardian knew it wasn't close enough and called us on it).

I went out and unplugged the SR785 and Agilent from our PSL racks.  This apparently immediately corrected our 120 Hz saturation problem, and the guardian proceeded to complete the CARM_5_PICOMETERS state.

However, I was unable to reliably reproduce the problem by plugging the spectrum analyzers back in.

The SR785 was plugged into the I-MON of the REFL A 9 demod board, and the I-MON of the REFL B 9 demod board.
The Agilent was plugged into the IMC Servo Board EXC A, TEST1, EXC, and TEST2.
It is not clear which one was causing the initial 120 Hz peak, since we could not reproduce the issue.  The 120 Hz saturations would come and go seemingly independent of my plug combinations, but was never as large as it was initially.

I left the spectrum analyzers unplugged from the walls, and also unplugged Robert's shaker contraption and the fancy remote-viewable PSL oscilloscope.  If the 120 Hz saturations return we'll know it wasn't because of these items.