TITLE: 02/20 Day Shift: 1530-0030 UTC (0730-1630 PST), all times posted in UTC
STATE of H1: Relocking
OUTGOING OPERATOR: Ryan C
CURRENT ENVIRONMENT:
SEI_ENV state: CALM
Wind: 4mph Gusts, 2mph 3min avg
Primary useism: 0.03 μm/s
Secondary useism: 0.42 μm/s
QUICK SUMMARY: Just lost lock from a nearby earthquake. Barely saw it on the picket fence before we lost lock, but our control signals were moving. For the 8 hour lock that just ended, 4 hours ago there was a step down in the range and then less stable range. DARM looks to have more noise in the 80-200Hz area, my screenshot doesn't show it completely. Violin mode 6 was slowly ringing up overnight.
TJ, Camilla. No extra noise in the channels that usually show our low frequency non-stationary noise 82728, see plot. Comparing DARM before and after, there is very subtle changes <100Hz. TJ found the summary pages show more glitches at 60Hz and ~48Hz.
Additionally, we see the line at 46.09Hz or 46.1Hz grow, see plot. Georgia noted this line in 2019 47447 and Evan piont us to O4aH1lines list where this apears to be the PR3 roll mode.
Running the range comparison scripts for a few different times and spans around the range step. There looks to be a slight bit more noise all the way below 100Hz, and the 60Hz is very slightly higher.
The range step happened at 330am almost exactly and since the 60Hz line got worse, I'm wondering if there is something that turned on or updated right then.
TITLE: 02/20 Eve Shift: 0030-0600 UTC (1630-2200 PST), all times posted in UTC
STATE of H1: Lock Acquisition
INCOMING OPERATOR: Ryan C
SHIFT SUMMARY:
IFO is LOCKING at LOCKING_ALS (Once again, we lose lock in the last few mins of shift...)
IFO is in NLN and OBSERVING as of 03:07 UTC.
Overall very calm shift in which we seem to have improved squeeze locking, improving range to what it was before last weekend.
We had one Lockloss seemingly caused by oscillations in PRCL in the seconds before the losing lock - alog 82916. According to the first OLG PRCL measurements from alog 82920, Elenna found that the PRCL2 gain was low by about 30-40%. Following this, she made a change to up the gain from 1 to 1.4 (alog 82917). Accepted SDF attached.
While 1.4 was a bit too high and caused a PRCL ring-up and a LL at LOWNOISE_LENGTH_CONTROL (where the gain switches on), the next setting of 1.2 worked! We were able to fully automatically re-lock and get to NLN and OBSERVING. Before I went into OBSERVING, I took another OLG PRCL measurement, which is the second measurement in alog 82920.
Other than this, the infamous IY Mode 5_6 Violin has been ringing up, visible in the top right screen of the attached screenshot, which shows mode 6 as slowly increasing since Lock. New settings may be needed for this.
Just as I was about to submit, we had a LL, though there wasn't the characteristic PRCL ring-up from the last few LLs. It also doesn't look environmentally caused since wind is low, there are no EQs and microseism is high, but unchanged mostly from the beginning of the day. Currently experiencing known ALS lock issues.
LOG:
None
TJ, Ibrahim, Sheila, Elenna
Measured PRCL OLG at 2 different times during NLN - both attached.
First (done by TJ) at 3hrs into NLN - Screenshot 1.
Second at 15 minutes into seperate NLN. This one was done after a PRCL-related lockloss (alog 82916) at which point Elenna changed the PRCL2 Gain from 1 to 1.2 (alog 82917) - Screenshot 2.
Just a note that we are trying for a UGF of about 30 Hz here. Right after lock, this is clearly a bit too high, but hopefully with the 20% gain boost after thermalization it will settle closer to 30 Hz.
I increased the PRCL2 gain that is set in lownoise length control from 1.0 to 1.4 to increase the overall PRCL loop gain by 40%. We have been seeing locklosses with 11 Hz oscillations that are probably due to marginal stability in PRCL. I changed line 5577 of the ISC_LOCK guardian, saved, and loaded. Ibrahim will post an alog with more info and open loop gain plots.
This was too high and caused a 70 Hz ring up in PRCL. I put in a gain of 1.2 now.
Lockloss that matches the ones from over the weekend, where PRCL becomes unstable and oscillates at 11Hz in the seconds before the Lockloss.
H1 Back to OBSERVING 03:07 UTC
TITLE: 02/20 Day Shift: 1530-0030 UTC (0730-1630 PST), all times posted in UTC
STATE of H1: Observing at 153Mpc
INCOMING OPERATOR: Ibrahim
SHIFT SUMMARY: The shift started out with a lock loss just before commissioning time. We took this opportunity to move PR2 spot while out of lock to find the baffle edges. We then relocked and Sheila and Oli fixed and issue we had causing lock losses at transition from ETMX (alog82912). We went back to observing for a few hours before stepping out to tune the SQZ angle and run a quick PRCL OLG. Results on the latter will be posted by Ibrahim later after he takes another to compare.
LOG:
Start Time | System | Name | Location | Lazer_Haz | Task | Time End |
---|---|---|---|---|---|---|
19:34 | SAF | Laser Haz | LVEA | YES | LVEA is laser HAZARD!!! | 06:13 |
19:12 | FAC | Les Schwab | Xarm | n | Truck tire repair | 21:07 |
20:39 | FAC | Eric | EX mech | n | Serial numbers | 20:45 |
21:24 | - | Betsy | Opt Lab | n | Parts and more parts | 21:52 |
TITLE: 02/20 Eve Shift: 0030-0600 UTC (1630-2200 PST), all times posted in UTC
STATE of H1: Observing at 144Mpc
OUTGOING OPERATOR: TJ
CURRENT ENVIRONMENT:
SEI_ENV state: USEISM
Wind: 6mph Gusts, 4mph 3min avg
Primary useism: 0.03 μm/s
Secondary useism: 0.60 μm/s
QUICK SUMMARY:
IFO is in NLN and COMISSIONING
The plan is to optimize SQZ via SQZ angle adjustment followed by a PRCL Open Loop Gain Measurement since there is some evidence that PR is experiencing noise at certain problem frequencies. Then, we go back to OBSERVING
Back to observing after a lock loss, some commissioning time, and trying to fix the transition from etmx lock losses.
useism is still high and our range is a bit low at 145Mpc. If the range doens't improve with more thermalization, we will take it out for some tuning.
Oli, TJ, Sheila
There have been several locklosses over the last day from the LOWNOISE_ESD_ETMX state, which happened while the gain was ramping down ITMX darm control and ramping up ETMX control. This is similar to what Elenna was trying to avoid by adjusting ramp times in 81260 and 81195, which was also at a time when the microseism was high.
Oli and I found that the problem with some of our transitions today was that the power on the DCPDs was dropping too low during the initial transition, we lost lock when it got to less than 1mA, in one of the sucsesful transitions it was as low as 4mA. We edited the guardian to not turn off the darm boost (DARM1 FM1) before making the transition, and instead we are turning it off directly after transitioning control back to ETMX, before the other filter changes that happen in this state.
This is the boost that we thought was causing locklosses when ramping off, 81638 which motivated Erik's quadratic ramping change 82263 which was then reverted 82284 82277. Today Oli and I increased the ramp time on this filter from 10 to 30 seconds. We have make the guardian wait the full 30 seconds for this ramp time, so this is making us wait longer in this state.
The attached screenshot shows the transition with the boost on on the left, and off on the right, the wiggle in the DCPD sum is about 1 mA rather than 15mA.
Oli is thinking about adding a check for DCPD sum dropping low to the lockloss tool.
HAM7 & 8 look less noisy at high frequency, same with the BSCs especially ETMY.
Wed Feb 19 10:05:46 2025 INFO: Fill completed in 5min 43secs
TCmins [-62C, -60C] OAT (+1C, 34F) DeltaTempTime 10:05:47
To get in Observing I had to accept some SDF diffs for SQZ, and PCALY. There was also still a PEM CS excitation point open as well. There was a notification about PCALY OFS servo malfunction so I looked at it and it was railed at -7.83, so I toggled it off and back on and it brought it back to a good value. I also did not receive a call, a voicemail just appeared.
09:21 UTC observing
H1:SQZ-LO_SERVO_IN1GAIN was left at -15 by accident, reverted to -12 and saved in sdf.
DriptaB, FranciscoL
SDF diffs for PCALY were incorrect. The timing of these changes match the h1iscey reboot done the same day (82902). Today around 19:00 UTC (almost three weeks later), we used EPICS values from the Pcal calibration update done in September (80220) to revert changes. Saved changes in OBSERVE and SAFE.
Daniel, Patrick, Jonathan, Erik, Fil, Ibrahim, Marc, Dave:
Starting around lunchtime and getting more frequent after 2pm the Timing system was showing errors with EY's fanout port_5 (the sixth port). This port sends timing to h1iscey's IO Chassis timing card.
At EY, Marc and I replaced the SFPs in the fanout port_5 and h1iscey's timing card. At this point we could not get port_5 to sync. We tried replacing the timing card itself, but no sync was possible using the new SFPs. Installing the original SFPs restored the sync, but the timing problem was still there. Moving to the unused port_7 (seventh port) of the fanout fixed the problem. We put the original timing card back into the IO Chassis, so at this point all the hardware was original and the fanout SFP had been moved from port_5 to port_6.
TITLE: 02/15 Eve Shift: 0030-0600 UTC (1630-2200 PST), all times posted in UTC
STATE of H1: Locking
INCOMING OPERATOR: Corey
SHIFT SUMMARY:
IFO is in NLN and OBSERVING as of 04:16 UTC.
EDIT: IFO is LOCKING at SDF_REVERT after a 05:59 UTC Lockloss
Overall calm shift with one lockloss due to an Earthquake (alog 82821).
Other:
LOG:
None
I took a look at the locklosses during the calibration measurements the past week. Looking at DARM right before the locklosses, both times a large feature grows around ~42 Hz right before the lockloss. Sat LL Thur LL
Thursday:
DARM_IN was dominated by the 42 Hz long oscillation and a ~505 short oscillation until the LL, DARM_OUT was dominated by the harmonic of the violins ~1020 Hz.
Saturday:
DARM_IN had a long and a short oscillation, the fund violin modes, ~510 Hz and ~7.5 Hz, DARM_OUT was dominated by the harmonic of the violins ~1020 Hz
I'm not sure how to/where to see exactly what frequencies the simulines were injecting during and before the lockloss.
Looking into what's going awry.
I pushed for a change of calibration sweep amplitudes on the Pcal and the PUM (which had been tested a couple of month's back) which was instilled into the calibration sweep wiki last week, labeled informatively as "settings_h1_20241005_lowerPcal_higherPUM.ini
".
Both of these sweeps were very near the end, where Pcal is driving at 7.68 Hz and PUM is driving at either 42.45 Hz or 43.6 Hz, which should clarifiy the source of the signals you are pointing out in this aLog.
The driving amplitude of the Pcal at 7.68 is about 20% lower than the injections that were being run the week before, deliberately done to reduce kicking the Pcal during ramping to reduce broad band coupling into DARM which would affect other measurement frequencies like the L1 which is driving at ~12 Hz at this time.
The driving amplitude of the PUM at ~42 Hz is unchanged from injections that had been running up until last week.
Not seeing any SUS stage saturating at lock losses. Presently unconvinced lock losses are related to new sweep parameters.
Both locklosses coincided with the ramping ON of the final DARM1_EXC at 1200 Hz
Tagging CAL properly