While still at LOWNOISE_ASC, PI Modes started ringing up.  

• Mode 26:  able to bring this down
• Mode 27 & 19:  Tried working on 27 to no avail.  Mode 19 was at 0 phase/gain (probably brought up by 27?)
• Then Mode 28, 20, & 22 also joined in.  (20 & 22 were 0 phase/gain).
• Then lock was lost.

Just making notes of locking this morning, since after the PI Mode lockloss this morning. (it's been a month since I've worked on this machine!)  :)

Lock#1

• Had to FIND IR by hand.
• Talked with Terra about PI Modes
  • Determined MODE 27 was the problem for last lockloss & it looks like we are addressing at the right frequency (it just rung up and brought up other modes before lockloss.).
  • Next time, will monitor ETMY PI damp out channel (bottom right of PI Overview), to make sure it's not saturating (thousands are OK, but millions is bad....if millions, lower the gain!)
• Went to DC READOUT & hung out there for a while
  • PI Modes:  Were quiet & fine (spent long enough time down that they must have come down on their own).
  • VIOLINS:  Looked at violins, but they didn't look huge (all rms outputs on medm screen were low 3's or below & ~500Hz peak on nuc3 was below ref), so did not actively change anything.
  • ROLL:  ITMY Roll was a bit high.  
    • +60dg -> +90phase (+60dg + +30dg)
    • Accidentally put in huge gain & this broke lock :(

Lock#2:  went to DC read out (made it on its own)

• Checked resonances and only addressed ITMy Roll again (damped it down with +90deg & gain up to 120).
• After a while, violins came down.  Lock broke at early step of INCREASE POWER (or maybe it was due to effect from step before:  REDUCE_RF_MODULATION_DEPTH).

Lock#3:  went to DC read out (made it on its own)

• Lockloss during the REDUCE_RF_MODULATION_DEPTH (could clearly see the AS spot moving & PRMI striptool signals getting noisy), so we'll stop at DC READOUT for current Lock #4 and see what the issue is here (Jenne & Keita are currently here).

We confirmed that the problematic peak is at 18040 Hz and that Mode27 has the right BP and PLL Set Freq settings for damping that. I've also (mostly) confirmed that 18040 Hz is still ETMY. I've done this by looking at the frequency drifts over a recent lock, though all locks have been short so there's still some uncertainty. 

I've attached a spectrum of the four 18kHz peaks we've had problems with in the past and what Mode they should belong to (I will be able to further verify this with a longer lock/strong temperature change during a lock). Please update BandPasses and PLL set frequencies to match these observed frequencies - this should be checked somewhat regularly for all problematic PI modes (you can use the PI DTT button from the PI medm main screen to generate a current spectrum).