travis.sadecki@LIGO.ORG - posted 17:40, Friday 10 March 2017 (34741)
Lockloss 1:40 UTC
Cause unknown.
Cause unknown.
TITLE: 03/11 Eve Shift: 00:00-08:00 UTC (16:00-00:00 PST), all times posted in UTC
STATE of H1: Observing at 61Mpc
OUTGOING OPERATOR: Jeff
CURRENT ENVIRONMENT:
Wind: 16mph Gusts, 12mph 5min avg
Primary useism: 0.04 μm/s
Secondary useism: 0.28 μm/s
QUICK SUMMARY: Locked for ~5.5 hours. Microseism and wind are both decreasing since they peaked about 6 hours ago.
TITLE: 03/10 Day Shift: 16:00-00:00 UTC (08:00-16:00 PST), all times posted in UTC STATE of H1: Observing at 61Mpc INCOMING OPERATOR: Travis SHIFT SUMMARY: One lock loss from NLN, no immediately apparent cause. Lost lock a couple of times before getting past FIND_IR. Ran a2l after reaching NLN and accepted ITMY TCS CO2 SDF differences (attached). Not certain what caused these. TCS guardian? PI mode 23 glitched a few times. LOG: 16:22 UTC Joe to mid Y 17:21 UTC Joe back 17:54 UTC Lock loss 18:28 UTC Corey to mid Y 18:58 UTC Back to observing (ran a2l and accepted TCS CS SDF differences) 19:08 UTC Corey back 21:12 UTC GRB 22:15 UTC Alfredo to mid Y 00:07 UTC Alfredo back
It seems like we should put some effort into our oplev damping loops.
It would be good to use the suspension model to look at our top mass damping loop shape while we have radiation pressure in the arms, as well as the oplev damping and see if we can get better damping out of the top mass. Next time we have some commisioning time, we could make some measurements before we switch to the low noise ASC to see if we can reshape these loops to have less noise.
FRS7556
During the recent upgrade of h1dmtlogin from Deb7 to Deb8 we lost the service which sends sysadmin an email when someone successfully ssh logs into the machine. Turning on this service was easy, most of the parts were already in place and it just needed a single line addition to the pam.d/sshd file. This service is now working, and I took the opportunity to document the install in the cdswiki.
Yesterday, Gerardo and I successfully silicate bonded the last ear to the final H1 ETM test mass. During the ~day-long process, ear #193 was bonded to the ETM16 S3 flat - no bubbles, and near perfect placement. This concluded the ear bonding task line items for H1 and L1 aLIGO. Knock on wood.
Starting CP3 fill. LLCV enabled. LLCV set to manual control. LLCV set to 50% open. Fill completed in 1205 seconds. LLCV set back to 18.0% open. Starting CP4 fill. LLCV enabled. LLCV set to manual control. LLCV set to 70% open. Fill completed in 1466 seconds. TC A did not register fill. LLCV set back to 40.0% open.
Increased each 1%.
CP3 now 19%.
CP4 now 41%
model restarts logged for Thu 09/Mar/2017 - Mon 06/Mar/2017 No restarts reported
J. Kissel Recently, Shivaraj, Arnaud, and Joe have noticed that during large tidal excursions, the actuation strength of the test mass ESD system changes (see LLO aLOG 31460). The mechanism is as such: - LLO feeds its low-frequency tidal signal to the ETMs at the M0/TOP stage (LHO feeds its tidal signal to L1/UIM) before offloading very-low-frequency control signal to HEPI. - This actutation force displaces the main chain from the reaction chain (the reaction chain does not move), relative to the QUAD's cage because that's M0 stage OSEMs actuation reference (as opposed to the L1/UIM stage, in which the actuation reference is the reaction chain). - The ESD actuation strength depends on the distance between the test mass and reaction mass. Thus, as the tidal control forces increase the distance between the chains, the actuation force of the test mass actuator decreases. After hearing of this, and knowing the difference between H1 and L1 tidal control, I'd asked "have we seen similar effects at H1?" Yesterday's tidal excursion (see LHO aLOG 34711) is an excellent test bed... and no similar change in actuation strength is seen when the Tidal control has this 80 minute excursion . However, there are are changes in strength that appear at the beginning of each lock stretch. These *might* be tidal, but they also might be correlation / covariance with the as yet unmeasured SRC detuning parameters... I attach several pre-made plots to kick-start the investigation: - Jim's plot of the tidal control signal sent to the test masses - Relative actuation strength of the ETMY test mass, as continuously measured by calibration lines.
Arnaud and Shivaraj The tidal control in this case was send to ETMX L1 but we only track the actuation strength of ETMY (whose HEPI was in control all this time).
The attached 5+ day trend plot shows some PSL Table Temp, the LSC drive to HEPI and a lock state indicator.
Clearly the slow warming after the AC ends its rapid cooling or maybe the slow propagation of the cooling effects in the PSL room are seen on the common arm length. Shame, we used to have the Ref Cav 20 C above the room...
The kinks in the Tide signal are not tidal and now that the AC has been disabled, this resulted in the tidal runaway the IFO experienced last night.
Of course, if we reach a new and tolerable steady state, I see no issues running at the higher temp.
The AC is only meant to be enabled whenever personnel are working inside the laser room.
The reason being that whenever personnel are working in there, the HEPA FFUs are supposed
to be on. They generate heat and make it a tad uncomfortable to work in.
Otherwise the Laser Room just inherits whatever the temperature of the LVEA happens to
be (with a small difference due to the make up air fan and the walls).
The reason for the AC turning on otherwise is still a mystery to me at least.
Attached is a plot of the operator selected observatory mode (H1:ODC-OBSERVATORY_MODE) for March 8 2017 (March 8 2017 00:00:00 UTC - March 9 2017 00:00:00 UTC). A value of 10 corresponds to observing.
As of now the summary page (https://ldas-jobs.ligo-wa.caltech.edu/~detchar/summary/day/20170308/lock/segments/) looks like the attached, and the top right panel doesn't agree with Patrick's plot.
Summary says 69.1% observing, 28.9% not locked. It might be related to the fact that some DMT segments were not generated on that day due to DMT problem (look at the white gap in "observing" and "calibration" on the top left). These numbers are skewed here but I don't know how exactly. What is the numerator and what is the denominator?
TITLE: 03/10 Owl Shift: 08:00-16:00 UTC (00:00-08:00 PST), all times posted in UTC
STATE of H1: Observing at 59Mpc
INCOMING OPERATOR: Patrick
SHIFT SUMMARY: No issue relocking at the beginning of shift. The rest of the night were quiet. Broad bump at low frequency (~10-20Hz) was present all night but went away after I ran a2l.
LOG:
13:30 LLO down. Out of Observe to run a2l.
13:44 a2l done. Back to Observe.
15:22 Bubba+Apollo to MidY
TITLE: 03/10 Day Shift: 16:00-00:00 UTC (08:00-16:00 PST), all times posted in UTC
STATE of H1: Observing at 61Mpc
OUTGOING OPERATOR: Nutsinee
CURRENT ENVIRONMENT:
Wind: 28mph Gusts, 23mph 5min avg
Primary useism: 0.09 μm/s
Secondary useism: 0.31 μm/s
QUICK SUMMARY:
No issues to report.
Relocking is going okay. Stopping at Violin_Mode_Damping_2 to squish one of the fundamental modes that rung up high.
A bunch of ETMX CDS overview went red briefly, middle ish column (either DAC or DAQ). Came with OMC DC PD saturation warning. And the red went away before I could look closer.
Accepted ETM and ITM SDF differences after making sure ETMX bounce mode was not ringing up. The rest are gain settings. Back to Observe.
An accepted PI differences. Had to changes mode 28 filter and PLL set freq. Back to Observe again.
Nutsinee is correct, at the same time we got a SUS-ETMX glitch the OMC PD DC ADC channels saturated. The time was 08:42:49 UTC. The saturations stopped 2 seconds later, the glitch was cleared at 08:44:04 UTC (cronjob running on the minute). Could this be a wierd coincidence?
This morning I turned on 6 new wireless access points (2.4GHz / 5GHz) in the OSB in preparation for replacing our existing system.
The access points are installed in the following locations:
They were powered up at:
$ date Tue Mar 7 11:11:49 PST 2017 $ date -u Tue Mar 7 19:11:50 UTC 2017
This afternoon I turned on 5 additional wireless access points (2.4GHz / 5GHz) in the LSB in preparation for replacing our existing system.
The access points are installed in the following locations:
They were enabled at approximately 15:30 PST (23:30 UTC)
A major question for the H1 interferometer is whether one of our arm cavities has anomalously high absorption. To look into this, Aidan asked me to run mode scans of the arm cavities immediately after lockloss, so that we can watch the frequency separation of the higher order modes change as the interferometer cools down. Prep work for this was done during last week's commissioning window (alog 34512), following the technique that Kiwamu used in 2014 (LLO alog 13768).
Prior to breaking the lock, Sheila and I went to the LVEA and bypassed the /10 frequency difference divider in the ALS COMM electronics chain, such that I can scan several FSR.
Travis got us back to ALS locked about 3.5 minutes after I broke the 9.5 hour lock, and the scan was started immediately after that. After the first scan or two, I remembered to slightly misalign the IR beam into the cavities, which I did by moving PR3 0.8urad in pitch (which was enough that I started seeing the TEM10 mode in the arm cavity transmission powers).
The scan ran for 30 minutes.
I put the FDD back in place, and Travis is now re-locking the IFO. The attached screenshot shows my striptool of the sweeps, mostly just that indeed it went on for about 30min.
The HOM spacing change over 30 minutes (t=[1200s, 3000s] in the attached plot) estimated by the online SIM model is approximately 90Hz for the XARM and 125Hz for the Y-ARM. This assumes:
| Optic | Absorption (ppb) |
| ITMX | 210 ppb |
| ITMY | 280 ppb |
| ETMX | 160 ppb |
| ETMY | 250 ppb |
I analyzed the data from the cavity scan measurements. There's a lot of noise in the data (partly from ALS COMM frequency noise) and I need to do a proper estimate of the best measurement we could expect to get.
In spite of the noisy data, we see a distinct downward trend on the XARM HOM spacing that is larger than the YARM downward trend. I did a linear fit to the data to estimate the change in frequency over time (from t = 310s to t = 2000s) and compared that to the change in frequency expected from the SIM model. The results are:
Based on the standard error of this measurement, we expect that the total absorption in the XARM could be anywhere from 1x to 5.2x the total absorption expected in the SIM model. The YARM could be anywhere from 0x to 2.2x the total absorption in the SIM model (it can't be negative).
Obviously I need to repeat this measurement taking care to get the best precision possible.
