I pointed the GDS calibration pipeline to the filters file (H1GDS_1243387557.npz) that has been running on the test machine for about a week or so and restarted the GDS calibration pipeline around GPS time 1244312463.  These filters are documented in alog 49587. This restart will also pick up the new installation of gstlal-calibration-1.2.10, which is expected to fix the power mains line subtraction issues seen so far in the run.  The new configuration file is located in 

aligocalibration/trunk/Runs/O3/GDSFilters/H1GDS_1244311784.ini

I reset both PSL power watchdogs at 17:41 UTC (10:41 PDT).  This completes FAMIS 10714.

OpLev centering was done on both ETMS and ITMX. I'm experimenting with compensating the alignments of these to the degrees that they are driven by the loops during an extended NLN lock. I was able to do EX at NLN and get a good 0/0. Unlocked/Aligned I compensated EY with +2u rad in P and +4u rad in Y. ITMX was compensated with +3u rad in P and nothing in Y. Below is a shot of the positions in NLN.

PT-524 cold cathode vacuum gauge tripped at end-X for unknown reason. Bubba reports that Hanford fire dept was not at that location (their radios sometimes trip gauges). I reset the gauge and waiting for it to come back online.

Bypass will expire:

Tue Jun 11 12:50:55 PDT 2019

For channel(s):

    H0:FMC-CS_FIRE_PUMP_1

    H0:FMC-CS_FIRE_PUMP_2

TITLE: 06/11 Day Shift: 15:00-23:00 UTC (08:00-16:00 PST), all times posted in UTC
STATE of H1: Unknown
OUTGOING OPERATOR: Jim
CURRENT ENVIRONMENT:
    Wind: 14mph Gusts, 12mph 5min avg
    Primary useism: 0.04 μm/s
    Secondary useism: 0.12 μm/s
QUICK SUMMARY:

H1 was not in observing due to SUS computer issue overnight, so we were not in OBSERVING.  Maintenance Day is well under way.

TITLE: 06/11 Owl Shift: 07:00-15:00 UTC (00:00-08:00 PST), all times posted in UTC
STATE of H1: Unknown
INCOMING OPERATOR: Corey
SHIFT SUMMARY: It was fine. Now it's hosed.
LOG:
13:38 The OS on the SUS HAM2 computer crashed, now we're in the weird state where we are locked, but can't go to observe

Just got a bunch of SW watchdog tripped notifications and got kicked out of observe, but we are still locked. All of the fields for the large HAM2 suspensions on the CDS overview are white, as well as H1IOPSUSH2A. I've called Dave, he's looking into it.

TITLE: 06/11 Eve Shift: 23:00-07:00 UTC (16:00-00:00 PST), all times posted in UTC
STATE of H1: Observing at 115Mpc
INCOMING OPERATOR: Jim
SHIFT SUMMARY:
LOG:

• 23:20 UTC - Gerardo back from MX
• 00:11 UTC - Sundae back from optics lab
• 00:55 UTC - Sundae to optics lab
• 01:16 UTC - Sundae back from optics lab
• 01:28 UTC - airplane(?)
• 0207:55 UTC - SQZ out of lock, H1 out of Observe
• 02:09:20 UTC - SQZ relocked on it's own, Georgia put H1 back in Observe (my lunchtime)

The preliminary message from today's squeezing measurements on the homodyne is that it seems like we do not have any large extra losses or phase noise when we are measuring using the homodyne, using the PSL as a local oscillator.  I will continue and do these measurements for a few different nonlinear gains during tomorrow's maintenance window.

Earlier today I went to HAM6 while Jenne and Georgia were working on increasing the input power to make some squeezing measurements on the squeezer.  I mostly followed Haocun's procedure on the git wiki: Measurement procedure

Since I used the PSL LO, there was only about 0.75mW of total LO power available, so that is what I used.  The table below shows visibility measurements taken in both the positive and negative homodyne arms, the first set was after my first attempt, I then realized that I had messed up the balancing of the homodyne, and went to a couple of meetings.  After I came back, I got the second set of measurements.  In both cases the asymmetry between the two arms I think indicates that the balancing isn't that good, but since the shot noise spectrum looked flat I went ahead with the measurements anyway.  

I noticed that when I looked at the intensity noise on the LO, in either arm of the homodyne, there is a large whistling line.  It also shows up in the squeezing spectrum, although much smaller. 

For the nonlinear gain I measured 79.5 counts on the OPO IR diode without green, and locked the maximum amplification gave 139 counts, de-amplification gave 56.5 counts, so that is a nonlinear gain of 1.75 or x=0.244

The data is saved on a floppy disk, so plots can be produced.  I read off the display 3.5dB of squeezing and 4.7dB of anti squeezing.  With the total efficiency of 0.9 (escape efficiency= 0.98, VIP transmission= 0.96, visibility^2 = 0.96, quantum efficiency=0.99), and no phase noise, we would expect 3.6dB of squeezing and 4dB of anti squeezing, so this is actually a little bit too much anti squeezing.  

For the homodyne visibility measurements, I used the guardian state LOCKED_SEED_HIGH, but made some changes to thresholds, and added those thresholds to the sqzparams file.  I had trouble using the LO guardian for the homdyne, and since it is difficult to debug guardian code on the imacs, I just closed and opened the LO loop manually.  

During our commissioning window today I re-ran the noise budget template for frequency noise coupling to DARM, to check if anything changed when we increased the input power from 35W to 37W (and also increased the CO2X power from 0.7W to 0.8W last week).

The first attachment shows a REFL_A to DARM TF, comparable to the purple trace in Sheila's comment here. These transfer functions look the same which is good.

The second attachment shows DARM, REFL_B and REFL_A spectra during the noise budget injection, the traces currently used in the noise budget currently are shown in black (data taken in early May). Note squeezing was not being injected during these times. Today I ran two sets of frequency noise injections:

1. with REFL_A and REFL_B sensors both in the CARM loop (mint green and blue traces)
2. with REFL_A only in loop, so that REFL_B can be used as an out-of-loop frequency noise sensor (pink, and brown traces) - this is what we actually use in the noise budget.

The DARM BLRMS suggest that I caught a low-frequency glitch at the end of my injection with REFL_A only in loop, hence the extra noise below 80Hz in the pink trace. So I won't put this data in the noise budget but we should re-do this measurement, since the frequency noise coupling looks much better above ~300Hz now compared to what we measured in May.

WP8234

Jonathan set me up with systemd monitoring of the LVAlert listener that I have been running for the past month. This will replace the GraceDb polling that we used to get candidate events, GRBS, etc. We will continue to run the GraceDb polling in parallel for now, but it will be stopped in the near future.

For more information about this system see: https://cdswiki.ligo-wa.caltech.edu/wiki/ExternalAlertNotification

Operators, if the IFO is having trouble tonight, particularly due to PI modes (you could check a spectrum of H1:OMC-PI_DCPD_64KHZ_AHF_DQ versus the beginning of the lock), you can revert the PSL power back to 35W. 

I copy here the instructions that Sheila left in alog 49680, since they are the same:  "We are going to leave the input power request set to 37W, if there are any problems with this it can be undone by going to userapps/h1/guardian/lscparams.py and changing NLN power to 35 on line 17 and hit load on ISC_LOCK, and on LASER_POWER."

Summary: IFO running at 37 W PSL input.  ETM ring heaters increased to avoid PI ring ups (so far, successful). LSC FF measurements done. CAL sensing function measurements done. (Note: New feedforward filters are not installed, I've just taken measurements for the fits).

• 10:00am: EY ring heater increase by 0.5 W
  • guardian to RESET
  • Filter input to 0.7 W
  • guardian to FILTER_RH_INPUT
• 10:07am: lost lock
  • No sure why.  Sheila had just gone out to HAM6, also EQ warning (but no seismic signatures yet)
• 10:14am: EY ring heater lowered, EX increased.
  • Talked to Carl, they increase their ring heaters about 0.125W per 1W of PSL power increase, so (depending on absorption of our optics relative to theirs) we should have increased our ETM ring heaters symmetrically by 0.25W when increasing the PSL from 35.5W to 37.5W. 
  • So, both EX and EY ring heater guardians taken to RESET.
  • Both EX and EY ring heater filter inputs to 0.45 W (They were both at 0.2W before I touched them this morning)
  • Both EX and EY ring heater guardians to FILTER_RH_INPUT.
• Reacquisition: DRMI ASC not to a good place for SRM
  • When we transitioned BS to AS36, the POP buildups got worse.  So, I asked Patrick to stop at CHECK_AS_SHUTTERS, and I hand-adjusted SRM in pitch (checked yaw, no change needed).  We'll need to come back to this, but I wanted to get back to power-up for now.
• 11:17: Locked, at MAXIMUM_POWER (35W)
  • Change PSL request to 37W
  • PI at 10214.1 Hz a little high, but hopefully is transient and will go down now that PSL power increased.
• 11:21: SEI_CONF back to WINDY (small EQ during acquisition, so we had been in EARTHQUAKE)
• 11:24: 10214 Hz still going up a bit.  Brought EY ring heater down to 0.4W (guard to reset, then changed epics val, then back to filter_rh_input).
  • EX ring heater still low-ish at 0.35W; assume that the guardian means to be doing this?
  • 11:29: PI seems to be coming down!
• 11:35: Start moving spots, pitch
• 12:50: Still moving spots, working on yaw.  EY yaw had a big impact, but REFL WFS related ASCs started oscillating.  I turned the INP1 gain for both pitch and yaw down to -0.8 from -1.0, seems okay.
  • Also, SRC ASC (both AS_C and AS72) seem to move a lot more than I'd expect when I move the ETM yaw spots.  POP18 improves with ETM spot move, but then goes back down as SRC ASC converges.
• 12:56: Getting a bit hard to see some ADS lines because some other noise is a little higher (may need to tune ITMX A2L to match current spots).  Decreased ADS master gain by factor of 2 so I could turn up the line amplitudes by a factor of 2.
• 13:22: running A2L for ITMX, to match new spots.
  • Pit needed big change, but improved noise at low freq back to normal levels.
• 13:30: yaw offset on src1 improves POP18.  yaw offset on SRC2 basically no change in buildups.
• 13:35: pit offset on src1 doesn't really change pop18, but makes pop90 a little lower
  • But, seems to bring arm circulating powers down a bit.  Removing the pitch offset.
• 13:42: Lost lock; turned off LSC feedforward at the input, rather than raming down gain.  Ooops.
  • Change guardian's nominal power to 37W, loaded ISC_LOCK and LASER_PWR.
• Reacquisition:
  • DRMI ASC - SRC1 yaw offset turned off at same time as gain engaged, so even though the input switch wasn't on, src1 sent a big signal out and we lost the DRMI lock. Guardian changed to turn off that offset before gain is turned on.
  • DRMI ASC - switching to AS36 again brought pop18 down, but Sheila said she had success last week just letting the IFO lock as-is, so we're trying that.
    • Did okay, got to prep_asc_for_full_IFO, but then lost lock.  Invesitgating - haven't found anything obvious yet though.
  • DRMI ASC - this time things looked better after switching to AS36, but Patrick still touched SRM up a bit.  Yaw had basically no impact, but pitch improved the pop18 a bit.
• 15:15:  MICH and SRCL feedforward gains ramped down to zero.
• 15:17: INP1 gains down to -0.8 as they were last lock.  SRC1 yaw offset of 0.5 on.
• 15:37: Reduce RF9 modulation depth, laser noise suppression.  Manual-ed over SQZ since Sheila is working out there on it, and then went to NLN_CAL_MEAS for LSC feedforward measurements.
  • LSC FF measurements done. For ITM to DARM actuation, I only ran the PRCL template, so it should be used for all 3 of PRCL, SRCL, MICH when fitting.
• 15:52: LSC feedforward back on, CAL sensing measurement ongoing.
• Guardian changed for INP1 gain (in engage_asc_for_full_IFO) and SRC1 yaw offset turn on (at end of lownoise_ASC)
• 16:33: Back to NLN, checking SDFs while Sheila wraps up on SQZ inside

TITLE: 06/10 Day Shift: 15:00-23:00 UTC (08:00-16:00 PST), all times posted in UTC
STATE of H1: Commissioning
INCOMING OPERATOR: Cheryl
SHIFT SUMMARY: Commissioning for majority of the day. Jenne worked on 37 W. Sheila worked on squeezer measurements. Taking calibration sensing function measurement at end of shift.
LOG:

15:58 UTC Rick and Sundae to optics lab
Jeff B. working in receiving bay, opening rollup door
16:24 UTC Karen to mid Y
16:39 UTC Jeff B. done in receiving bay
16:59 UTC Changed to commissioning mode, Sheila to HAM6 area to start squeezer measurements, Jenne working on increasing power
17:08 UTC Lock loss
17:09 UTC Ed taking car to shipping and receiving
17:16 UTC Jason to optics lab
17:19 UTC Karen leaving mid Y
17:25 UTC Jason back
17:25 UTC SEI_CONF to EARTH_QUAKE
17:43 UTC Kyle to mid Y
18:00 UTC GRB alert (E335886), not locked, ignoring
18:09 UTC Gerardo back from HAM5
18:11 UTC Gerardo to mid X
18:53 UTC Gerardo back
19:06 UTC Kyle back
19:23 UTC Hugh to roof, staying on observation platform
19:28 UTC Hugh back
20:00 UTC GRB alert (E335892), in commissioning, ignoring
20:26 UTC Amber filling LN2 dewar from tank
20:31 UTC Niko and Sundae to laser lab
20:42 UTC Lock loss
21:56 UTC Sheila to HAM6
22:07 UTC Gerardo to mid X