40m QIL Cryo_Lab CTN SUS_Lab TCS_Lab OMC_Lab CRIME_Lab FEA ENG_Labs OptContFac Mariner WBEEShop
  40m Log  Not logged in ELOG logo
Entry  Fri Jun 9 15:31:45 2017, gautam, Update, IMC, IMC wonkiness IMG_7409.JPG
    Reply  Fri Jun 9 17:45:21 2017, Gautam, Kaustubh, Update, IMC, IMC wonkiness 
       Reply  Fri Jun 9 19:18:10 2017, gautam, Update, IMC, IMC wonkiness MC2_UL_glitchy.pngMC2_glitch_fast.png
          Reply  Mon Jun 12 22:23:20 2017, rana, Update, IMC, IMC wonkiness 
          Reply  Tue Jun 13 08:40:32 2017, Steve, Update, IMC, IMC wonkiness happy_MC.pnglast_glitch.png
Message ID: 13055     Entry time: Fri Jun 9 15:31:45 2017     Reply to this: 13057
Author: gautam 
Type: Update 
Category: IMC 
Subject: IMC wonkiness 

I've been noticing some weird behaviour with the IMC over the last couple of days. In some lock stretches the WFS control signals ramp up to uncharacteristically huge values - at some point, the IMC loses lock, and doesn't re-acquire it (see Attachment #1). The fact that the IMC doesn't re-acquire lock indicates that there has been some kind of large alignment drift (this is also evident from looking at the (weak) flashes on the MCREFL camera while the IMC attempts to re-lock - I am asking Steve to restore the MC trans camera as well). These drifts don't seem to be correlated with anyone working near MC2.

The WFS servos haven't had their offsets/ DC alignments set in a while, so in order to check if these were to blame, I turned off the inputs to all the WFS servo filter modules (so no angular control of the IMC). I then tweaked the alignment manually. But the alignment seems to have drifted yet again, within a few minutes. Looking at the OSEM sensor signals, it looks like MC2 was the optic that drifted. Steve tells me no one was working near MC2 during this time. But the drift is gradual so this doesn't look like the infamous glitchy Satellite Box problem seen with MC1 in the recent past. The feedback signal to the NPRO / PCdrive look normal during this time, supporting the hypothesis that the problem is indeed related to angular alignment.

Once Steve restores the MC2 Trans cameras, I will hand-align the IMC again and see if the alignment holds for a few hours. If it does, I will reset all offsets for the WFS loops and see if they hold. In particular, the MC2 transmitted spot centering servo has a long time constant so could be something funny there.

*Another issue with the IMC autolocker I've noticed in the recent past: sometimes, the mcup script doesn't get run even though the MC catches a TEM00 mode. So the IMC servo remains in acquisition state (e.g. boosts and WFS servos don't get turned on). Looking at the autolocker log doesn't shed much light - the "saw a flash" log message gets printed, but while normally the mcup script gets run at this point, in these cases, the MC just remains in this weird state. 

Attachment 1: IMG_7409.JPG  705 kB  Uploaded Fri Jun 9 16:37:49 2017  | Show | Show all
ELOG V3.1.3-