For a direct comparison to Maggie's alog37867 I attached a 2-day plot of SHG transmitted and reflected power (+ all other relevant channels) stopping right before a temporary TEC cable was removed:

CH1 = SHG crystal temperature, kept at 35.9 deg C where the green conversion was optimized.

CH2 = Red transmitted power through the SHG (16Hz channel)

CH3 = Mephisto laser power (locked to PSL for the most part)

CH4 = Temporary temperature channel (sensor installed after Feb 8th 00:00:00 UTC). A resistor was taped on to the table right next to the SHG housing (Calibration: 1638.4 counts per Celsius).

CH5 = Green reflected power (16Hz channel)

CH6 = TTFSS locking status.

So far we don't see the fluctuation that LLO is seeing. It's not the most stable looking power trend because we couldn't live without opening and disturbing the table for long (+ PSL was not in its best behavior that week). Once the SHG is locked and undisturbed for long enough time we can start characterizing it more seriously.