While the laser was down this morning, I took a quick measurement of the ETMX pit oplev damping loop. It has about 5 dB of damping in a narrow range around 0.46Hz, and has zero phase and -10dB of gain at 1.5 Hz with no light in the arm.
Because the noise from these loops is limiting DARM at 10 Hz (alog 34735), I made a slightly more aggressive cut off (LP8 is intended as a replacement for LP10) which can be tried next time we have a chance. This is the reason there are unloaded changes in the filter files for all 4 quads right now.
I used this measurement to update the quad model with gain of the electronics and software between the PUM pitch actuators and the optical lever. See CSWG log 11209.
The gain was found simply by comparing the measured loop gain to the modeled loop gain with the installed filter, and scaling the model until matched the measurement. The gain applied to the model to make them match is 1/1500.
More measurements like this for the other suspensions will be useful for telling us if this gain factor is the same for all suspensions, or at least the same across sites.
Thanks for including the measured data, that made the analysis a lot easier.
Very good, quiet 4 days. Overall duty cycle was 91.4%, with range reaching just over 70 Mpc at times
- 2 24+ hour lock stretches
- Some wind issues Thursday morning (UTC) and Saturday evening (UTC)
- Strange trend in low-frequency noise on Sunday: started low, peaked at 0800-1000 UTC, then came back down slowly; possibly periodic with 24-hour period, uncorrelated with microseism
Summary of the data quality shift for Monday March 13 to Wednesday March 15, 2017 (inclusive):
In the MAIN SITEMAP area under the second PEM pulldown select Weather Wind for Ski's direction/speed plot. The arrow directions needed updating to reflect Robert's remapping (LHO aLog 28456) of the anemometer directions. These can be a bit confusing as wind is typically described as 'being from', e.g., a west wind is from the west. But these arrows point toward the direction the wind moving toward so that should be intuitive. The numbers can be confusing though with 0 being from the CS to EndX and CW from there. So a number like 180 might be thought of as a south wind but it is actually wind moving from EndX toward the Corner Station. Enough words.
Starting CP3 fill. LLCV enabled. LLCV set to manual control. LLCV set to 50% open. Fill completed in 53 seconds. LLCV set back to 16.0% open. Starting CP4 fill. LLCV enabled. LLCV set to manual control. LLCV set to 70% open. Fill completed in 129 seconds. TC A did not register fill. LLCV set back to 43.0% open.
We took a few minutes to do excitations on HPI HAM1+HAM6 before going back to observing. This was to follow up the measurements made in 34623. Robert observed that some of the upconversion that we have been having in DARM in times of higher 3-10 Hz ground motion has harmoincs of 11.6 Hz, so one suspicion could have been that this is a multiple of the tip tilt resonance around 6Hz, and this would have been missed by the excitations that concentrated on 10-40Hz. However, we found nothing resembing this up conversion. As we knew HAM1 motion does get into DARM at lower frequencies. (probably through CHARD)
PSL tripped. We were observing for a total of 2min and 27sec
PSL tripped due to a trip of the 35W FE power watchdog; this trip was NOT due to a flow problem. The first attachment shows the output power of the lasers that make up the PSL (NPRO, 35W FE, HPO). It's somewhat hard to see in the picture, but the NPRO begins to lose power ~1 second before the FE and HPO lose power. The second attachment shows the NPRO output power compared to the relevant interlock signals for this trip. The NPRO loses power, causing a trip of the 35W FE power watchdog; this in turn causes the HPO to lose injection locking and ~6 seconds later a trip of the HPO power watchdog. The third attachment shows the NPRO power output compared to the output power from the NPRO pump diodes. Timeline of the trip at this point appears to be:
At this point it is unclear what caused the NPRO to lose power. This, on the surface at least, appears to be another occurrence of those rare PSL trips due to some glitch in the UPS the NPRO is plugged into (it has been some time since we've seen a trip like this); while I have no direct evidence of this, the behavior of this trip is identical to that of the previously linked trip from September 2016.
The system came back up without issue.
FRS 7665 was filed for this trip.
This morning I saw some noise in the 60-200Hz area on DARM that lastest for a few minutes and then went away. I came back about an hour later, and an hour after that as well. The SEI BLRMS are clearly coincident with this breathing noise, and with the slight drop in range. Robert is thinking this may be tilling or truck traffic on the Hanford site.
I attached a plot of some of the BLRMS. It is seem the most in the corner station, but the ends see it a bit as well.
Robert and I watched with the roof camera and saw that there are about 5 large trucks that will pass the site on Route 10 within a few minutes of each other. The trucks are the kind that haul dirt/gravel/rocks and have two trailers (excuse my lack of truck terminology there). As these trucks pass, noise is DARM is clearly seen, and in the 3-10 and 10-30Hz SEI BLRMS as well. This much ground motion caused by trucks is abnormal, and we are hoping that it was a only for today. Our range has definitely seen this ground motion and it is show as going from 48-68Mpc for the last 3 hours, and is also seem in the previous lock ending around 16UTC. I attached a 24hr trend of some of the SEI BRLMS and range attached.
While I was watching for trucks to go by, I also saw a FedEx delivery truck show up on site. As this truck made the turn from Route 10 onto our driveway, it raised DARM similarly like the above trucks. On the straight the noise dropped, then the turn toward the LSB it rose again, and finally it had to backup and the noise popped back up.
Laser Status: SysStat is good
Front End Power is 34.12W (should be around 30 W)
HPO Output Power is 163.8W
Front End Watch is GREEN
HPO Watch is GREEN
PMC: It has been locked 5 days, 21 hr 27 minutes (should be days/weeks)
Reflected power = 16.03Watts
Transmitted power = 64.35Watts
PowerSum = 80.38Watts.
FSS: It has been locked for 0 days 0 hr and 45 min (should be days/weeks)
TPD[V] = 3.319V (min 0.9V)
ISS: The diffracted power is around 2.0% (should be 3-5%)
Last saturation event was 0 days 0 hours and 45 minutes ago (should be days/weeks)
Possible Issues:
Our good friend Arnaud made an medm showing the BLRMS motion of all the ISI platforms:
"I made a medm screen to show the performance from all platforms at once.
This should help when trying to catch a problem like we had last month with elevated HAM6 motion (https://alog.ligo-la.caltech.edu/aLOG/index.php?callRep=31831)
The file lives under the svn at /opt/rtcds/userapps/release/isi/common/medm/graphics/grid_all.adl (or grid_mini.adl, a smaller version).
and only needs the site/ifo variable as an input (which the sitemap should provide).
At LLO it is currently under sitemap->ISI->PERF_OVERVIEW"
These BLRMS performance views are all visible on the ISI platform overview screen but having them on one screen should be useful. I've updated the H1 area with this file and edited the SITEMAP.adl; LHO Operators can find it in the MAIN area under the SEI pulldown at PERF_OVERVIEW.
Remember green is good, red is bad, yellow and orange in between. See attached.
model restarts logged for Sun 19/Mar/2017 - Tue 14/Mar/2017 No restarts reported.
Ran well over the weekend.
Went over maintenance items for tomorrow (see whiteboard for plans).
There are also a few commissioning, calibration, and PEM items that will require time with a locked IFO either tomorrow or over the week.
not sure of the cause yet. The 60-200Hz noise came back a few times before the lock loss, maybe related?
Back to observing after getting to low noise and losing lock to an unheard PI, and then a brief period of comissioning when we were back up again.
Adjusted the PSL Beckhoff computer time to daylight savings. The clock was about 10 minutes off in addition to the hour.
EX had a dip in pressure around March 9th from 97.7 to 97.5, but other than that all looks good.
I forgot to link to FAMIS4530
Closed
Looking over the past week, the daily EY gliches are at 20:44 PDT (03:44 UTC). On Saturdays a second glitch occurs at 21:04 PDT (04:04 UTC). Attached plots show 03:25 to 04:10 UTC for 3/19 and 3/12 (two successive saturday evenings in local time).
Here are data from End Y with most of the pressure channels and the controller output to the pump motor.
The attached plot gives support to this argument:
Something is happening with either the Pump motor or the VFD causing an increase in pressure out from the pump. This is evident by the increasing delay in pressure response as we proceed down the Pump Station Manifold. PRESS1 is the first after the Pump, the rest are after Laminar flow resistors and the filter. BSCSUP_PRESS is way down the line at the Chamber. The PRESS1 clearly leads the down stream pressures and it steps more sharply than the others. So this is a real pressure change increasingly softened by the accumulators and resistors etc as the wave moves down the line.
The VOUT signal is the PID loop responding to the differential pressure (diff between BSCSUP and BSCRET[not shown].) There is a big step down on this channel just before the pressure step up but that wouldn't cause an increase in pressure. So, I don't believe the motor is being directed to drive harder.
I'll look over the VFD manual to see if it has any scheduled items that might cause this? That just hardly makes sense...
Edit--Extra--If the output voltage of the PID controller could actually be manipulated without the VOUT data signal registering it, that too would be a possibility...Dave is looking closer at the computer.