Reports until 13:08, Sunday 07 April 2019
H1 General
jeffrey.bartlett@LIGO.ORG - posted 13:08, Sunday 07 April 2019 - last comment - 16:57, Sunday 07 April 2019(48301)
OPS Day Mid Shift Summary 
   Busy morning.

   Lock loss just after "LSC Servo Board Split Voltage High" Message. 
   Had GRB alert while relocking the IFO 
   Rode through a mag5.7 EQ in Chile - SIESMON said to select EARTHQUAKE state. Did not change state and IFO continued to locking at NLN with no apparent issues. 

   Dropped out of Observing due to OPO SDF Diffs. ZM1 UL, LL and ZM2 UL, LL, UR, LR saturating. Looked like Squeezer was still locked. Called Keita. He cleared the Servo Filters on the SQZ_ASC Overview, relocked the SQZ_SHG, and requested SQUEEZING from the SQZ_MANAGER. SDF Diffs cleared, and the IFO is back into Observing. 

   Right now the IFO is locked in NLN with a range of 110.6Mpc. Winds are a no more than a Light Breeze. Am tracking 5 increasing Violin modes (ETM-Y mode2, mode4, mode8 & ETM-X mode6 & ITM-Y mode12). All are under 2.5 and growing slowly if at all. Will continue to monitor.
Comments related to this report
sheila.dwyer@LIGO.ORG - 16:57, Sunday 07 April 2019 (48304)
Link

About the squeezer issues:

  • From what Keita described on the phone, I thought that the squeezer laser mode hopping could have been the problem similar to what we saw on Friday, but when Keita looked at the SHG trans diode the normalized power was something like 0.78, which is about normal, so that might not have been the problem.  
  • It sounds like the squeezer guardian reacted to an SHG lockloss by going to the LOCK_SHG state, which is the behavoir I expected after the changes made on Friday.  However, it left the beam diverter open and didn't turn off the squeezer ASC.  Keita went to the SQZ_ASC screen from the overview, turned the SQZ_WFS master switch off, and cleared the history of the ANG and POS servos, which is why the verbal alarms about the OPO model saturations stopped.  On Friday I added a decorator in SQZ_MANAGER called the OPO_checker that check for OPO locklosses and returns a jump transition.  Just now I logged in and added to that decorator lines that will turn off the SQZ_ASC, and request that the OPO guardian go to DOWN, which I will close the beam diverter.  (I would like to move the beam diverter into the SQZ_MANAGER, but won't try to do that remotely.  
  • The SHG guardian must have relocked the SHG without a problem, but the SQZ_MANAGER guardian was stalled, so it wasn't requesting the rest of the squeezer loops to relock.  I just now logged in to look at ISC_LOCK, and there is no unstall_nodes decorator on the NLN state (this is on most other states in ISC_LOCK, it prevents this kind of situation).  I have now added that decorator to the ISC_LOCK guardian in NLN.  

I haven't loaded either of these changes, since we are in observing right now, but if there is any reason to come out of observing tonight I think that it should be OK to load both of them.