We looked at the FSS EOM monitors on the oscilloscope today. Both the south and north path broadband FSS EOMs have large positive offsets from zero on the order of ~ few volts. Also, each are sawtoothing in a weird way.
It seems like the EOM is phase-wrapping its actuation, i.e. it's pushing so hard in one direction that it needs to go greater than 2π, but once it's done that it jumps back up to 0 phase around and just keeps going around and around, looking like a sawtooth function.
I feel like this sawtooth EOM monitor is indicative of saturation of the EOM, but am not sure how. awade posits that this might be how the EOM always actuates?
We can "control" the EOM monitor offset by messing around the Fast and Common gains. Increasing the Fast gain, in general, decreases the EOM offset. Increasing the Common gain increases the EOM offset, until it saturates fully and starts ringing.
The PZT control signal paths are glitchy: When we look at the the laser fast control signal coming from the boxes with the FSS loops off, we see 500 kHz underdamped oscillation glitches. The glitches happen at a rate of 60 Hz, which is indicative of the problem. It seems the the PZTs themselves are sourcing the glitches, could be mains causing some kicks to the PZT, which is then read back out through the PZT control voltage.
We have also noticed that the characteristic spikes in our transmission, reflection, and error signal occur at 60 Hz. Mains is probably the cause. Not sure how to get away from this one. These spikes are probably not noisy, but they could be triggering EOM ringing in our FSS loop.
We put a function generator into our South FSS box and looked at the output. The function generator was outting a signal at 1 kHz with an amplitude of 10 mV, but we put a -30 dB attenuator on to bump it down three orders of magnitude. Inputting signals as little as 10 µV would saturate the EOM path. This seems very small, and not great for our FSS box circuitry.