Somehow the GRMC/MZ error signal demodulation phase issue was fixed. Actually, part of the problem seemed to be the SHG - for some reason the gain was too low causing the output to have some uncontrolled slow sinusoidal oscillation of output green power as well as poor lock stability. I turned it up to the threshold of oscillation from high gain (separate to the kind of oscillation mentioned in the previous sentence, which is much slower) and then reduced it so that the output stops oscillating - the setting value changes from 0.08 -> 1.6 just by eye. Maybe a little bit change by UGF optimization but I don't know why the optimal gain changed suddenly. Anyway once again the green path seems ok and sends stable 25 mW to OPO.

IRMC still gives me problems. Once again, the system can find the lock threshold but refuses to hold lock very well. Even when it "locks" there are a lot of quite significant bumps and glitches in the reflected power that can be seen on the oscilloscope. I tried switching PZT high voltage drivers with IRPS, GRPS, GRMC but it didn't work. I tried using a variable ND filter at ~ ND0.4 (vs ND1 previously)to bring the base level of the IRMC reflection spectrum to about 9 V, close to the PD saturation. The gain was too high but turning it down, I could see that the IRMC locks... for about 1 minute and then the same problem comes back. I tried a weaker signal with about ND1.5 worth of attenuation but it didn't even find the threshold. The carrier and sideband peaks in the error signal both appear to be fluctuating in power as well - this does not seem to happen for SHG error signal which comes from the same modulator, nor does it happen for IRMC reflection spectrum which uses the same PZT. I might have to try taking the 18 dB RF amplifier in the error signal path. Also when I got home I noticed there was some DC offset on the error signal (fig 1) which I guess I should check out as well.