Naoki, Vicky

At 25W, holding at REDUCE_RF45_MODULATION_DEPTH before MAX_POWER, we did a quick anti-sqz/sqz measurement to see if we could loosely infer sqz losses with the IFO relatively cold. We did not see clearly more squeezing here at lower power / colder ifo, as we have in the past, e.g. 66877. We weren't able to easily engage ASC here (it was before move_spots), but slider values are similar to when IFO was last locked and a little bit of walking alignment didn't make a big difference, so we left ASC off.

Looking at DARM (based on GDS), and the SQZ BLRMS, we can read the sqz levels:

• Anti-sqz:  DARM ~ +13.0 dB,   BLRMS 5,6~ 13 dB
• SQZ:     DARM ~ -3.9 dB,     BLRMS 5,6~ -4 dB

which are surprisingly consistent to what we have at full power, e.g. the recent sqz dataset in 71902 with IFO at 60W and thermalized in full lock.

Thoughts:
  -- I'm not sure what it means that now we have similar amounts of squeezing with the IFO at low and high power.
  -- For this test, we left PSAMS at 200V/200V and didn't try walking PSAMS, though in 66877 we saw better squeezing with a cold IFO and lower PSAMS.
  -- The slope of DARM at high ~kHz frequencies seems to be slightly different between 25W and 60W? From the dtt, compare red dashed = "60W no sqz" & cyan line = "25W FDS". Here, the lock sequence was at REDUCE_RF45_MODULATION_DEPTH, so before lownoise length control and laser noise suppression.

We then measured NLG ~ 11.05. In the sqz dataset 71902 we did not separately measure NLG, but given that the NLG has remained consistently accurate with the Feb 2023 NLG calibration, so we can re-affirm that trusting the calibration is probably okay, assuming opo temperature is well-tuned. For comparison, NLG calibration suggests NLG=11 (gen sqz ~ 14.75dB) with opo_trans=80uW.

• max_ir = 0.0081
• min_ir = 0.00022
• no_pump_ir = 7.1e-4
• dark = -2.5e-5