Daniel, Tony, Vicky

edit: see PSL lockloss trends for the previous lockloss, 81157. Both of this and the previous lockloss look PSL-related (ie upstream of IMC- which does see power fluctuations earlier than suggested by the IMC_TRANS channel). Might be worth trying whether the IMC can stay locked with the ISS_ON vs. ISS_OFF tests again, now that the PSL laser is not obviously mode hopping.

This lockloss maybe looks strange: see 20-ms trends, 1-second trends, 5-second trends.  It looks like a very fast lockloss (< 1ms) coincident with various FSS/ISS/TPD glitches - is this different than what we saw before? Here, AS port loses light within like < 1 ms (as witnessed on Keita's new lockloss monitor, AS_A, and LSC-REFL which sees some corresponding power increase). Lockloss looks within 1ms of changes in PSL-FSS TPD, FAST_MON, PC_MON and ISS AOM , SECONDLOOP, etc.

Weirdly IMC-TRANS_IN1_DQ (fast channel 16k) does not see power change until 5 ms later, which I don't understand?  I think like DARM loses lock (which should be why AS port and LSC-REFL both change inversely, right?), while the power on IMC-TRANS doesn't change, despite the PSL FSS and ISS loops all see glitches.

Daniel suggested maybe there could be some more analog filtering on IMC-TRANS slows down this channel (even though it is recorded at 16k?) - we're not sure why there is such a delay, and whether this IMC-TRANS channel a super reliable timing metric for what is happening. The ~5ms seems too fast for the storage time given IMC's ~8.8 kHz cavity pole (1/(2*pi*8.8kHz) ~ 18 microseconds).

Daniel helped add some fast PSL-ISS_SECONDLOOP channels to the Sheila's scopes, and I've added Keita's lockloss monitor channel too (81080, H1:PEM-CS_ASC_5_19_2K_OUT_DQ ), then saved this scope at /ligo/home/victoriaa.xu/ndscope/PSL/PSL_lockloss_search_fast_iss.yaml