For Rai. Both the pirani and cold cathode data.
no restarts reported
We had some trips durring the earth quake, the attached plot of ETMX is from the first trip, a later trip wthe plot of ITMX is from a trip that happened after Chris untripped the watch dog and the guardian tried to bring it back (the ground motion was still high and we were still using Tcrappy).
Since everything is tripped and those we have untripped are tripping again, I am not going to try to post all of these plots Anyone who is interested is free to find the data in the frames.
some things about guardian:
It would be nice if the manager displays an more usefull error message if the ISI trips, the message we got was: node ISIS_BS_ST2: NOTIFICATION
The ITMY manager says WATCHDOG TRIPDAQKILL, a more usefull message.
We tried moving ITMX to T750, it tripped again tryng to isolate. Chris was able to get ITMX to damped, he had to figure out first that it is necessary to go to init, which is not obvious. ETMX did suceed in isolating with the blends on Start. Since we have had huge earth quakes the last two nights it is not surprising that guardian couldn't bring the ISIs back on Tcrappy. We will need to wait for more normal trips to test that more.
Also, Rich said that I should write about problems in the alog in red letters, so here are a few, but I'm making them yellow since they are more like annoyances than huge problems:
ISI overview screens bring the control room work stations to a halt. 73% of my cpu time is curently devoted to medm (Xorg), and all I have open are 2 ISI screens and 2 ISI watchdog screens. I'm not sure if this is a problem with the workstations or the screens, but it is pretty annoying.
It does seems like the control room workstations are crashing much less than they were, thank you for that CDS!
The plotting tool for WD trips doesn't alway work, here is a sample error message:
File "/opt/rtcds/userapps/release//isi/common/scripts/wd_plots/main.py", line 170, in _BufferDict
abs_threshold_mask=abs_threshold_mask, host=host, port=port)
File "/opt/rtcds/userapps/trunk/isi/common/scripts/wd_plots/pydv/bufferdict.py", line 243, in __init__
self.add_channels(channel_names, wd_state_mask=wd_state_mask, abs_threshold_mask=abs_threshold_mask)
File "/opt/rtcds/userapps/trunk/isi/common/scripts/wd_plots/pydv/bufferdict.py", line 382, in add_channels
abs_threshold_mask=abs_threshold_mask)
File "/opt/rtcds/userapps/trunk/isi/common/scripts/wd_plots/pydv/bufferdict.py", line 344, in __fetch_data
self.conn.clear_cache() # clear any saved information from a previous connection
RuntimeError: Input/output error
It would be nice to have the brief guardian screen in the ISI screens, so you can tell what is going on. I would greatly prefer if this is the brief version that still allows you to select the requested state (at least for the manger), because these screens allow us to use gaurdians despite the bugs that plauge us with other guardian medm screens.
It would also be nice if the guardian medm screens themselves worked no matter which workstation you use, how you opened medm, or if they worked at all at an end station. There seems to be no way to open a guardian screen from an end station, not even using guardmedm. My current approach durring the day is to call the operator to ask them to misaling/ realign optics, I'm sure they are going to get sick of this soon.
Chris, Sheila, Alexa, Daniel,
This morning Chris rephased the ALS WFSs, and balanced the gains of the different sections. Results were verry sensitive to the cavity alingment, so a few iterations were needed. Then he measured the sensing matrix again several times, this also depends on the cavity alingment. These three measurements were with as good of an alingment as we could find by hand (phase in degrees in parens for those elements where the phase was inconsistent):
| ETMX PIT | ITMX PIT | ETMX YAW | ITMX YAW | |
| WFS A | -944, -994, -1211 | 858, 807, 613 | -61.8, -61.9, -59.3 | 1043(-27), 1043(-107), 1031(40) |
| WFS B | -580.2, -669, -877 | 766(15), 935(114), 801(-179) | 52.0(-132), 52.1(-123), 51.4(-10) | 568, 541, 570 |
With this information we zeroed out WFS B to ITMX PIT and ETMX YAW and WFS A to ITMX YAW, the elements where the phase wandered, and then inverted this. We also offloaded the WFS alingment signals to the top stage, to avoid saturating the DAC.
Chris has also written a script that centers the WFS using the picomotors. It is userapps/als/h1/scrpits/als_x_wfs_center.py
We have also written a wfs relieve script at userapps/als/h1/scrpits/WFS/alsWFSreleive.sh. This just relives the top stage, onto the offset of the M)_LOCK filter bank.
We were then sucsessfull in turning on all 4 DOFs, and we turned up the gain, we estimate that we got a ugf of 1 Hz for both of the Pitch loops. We made the attached measurement. In the left plot, the COMM noise measurement (no cavity pole removed) from monday night is the blue reference, and our measurement from tonight is in red. The rms down to 0.1 Hz is 10Hz, still almost 30Hz down to 0.02 Hz. We also have higher noise at 1-3 Hz, we don't know if this is caused by the changes to the WFS, or if it is just a difference in the ground motion. Because of the earth quake we aren't able to repeat the measurement without WFS right now, but we looked at some coherences to try to understand. The top panel in the middle plot shows that the oplevs have coherence with the WFS control signals up to about 1 Hz. We looked at some end X seismometer coherences and don't see anything (bottom panel of middle plot). The right plot shows that the oplevs do have coherence with the COMM Noise from 0.4-0.8 Hz.
This data is saved as COMM_NOISE_April_2.xml in my COMM folder (sheila.dwyer/ALS/HIFOX/COMM)
While going over these sensing matrix measurements we found some problems with the templates, we can probably disregard the measurement. A btter one is coming soon, we hope.
Similarly to what has been done for ETMX, the plant inversion filters (pitch and yaw) of ITMX UIM were installed today in the drivealign matrix.
Measurements were taken using the dtt templates living in :
sussvn/trunk/QUAD/H1/ITMX/SAGL1/2014-03-29_H1SUSITMX_WhiteNoise_P2PY[...]
sussvn/trunk/QUAD/H1/ITMX/SAGL1/2014-03-29_H1SUSITMX_WhiteNoise_Y2PY[...]
will run for ~ 16 hours (undamped / damped for main and reaction chain)
FInished when checked 4/7 (Arnaud is away).
M0 undamped: 2014-04-02-1080519986_H1SUSETMY_M0_0p01to50Hz_tf.mat
M0 damped: 2014-04-02-1080537980_H1SUSETMY_M0_0p01to50Hz_tf.mat
R0 undamped: 2014-04-03-1080555846_H1SUSETMY_R0_0p01to50Hz_tf.mat
R0 damped: 2014-04-03-1080573824_H1SUSETMY_R0_0p01to50Hz_tf.mat
Everything looks good except that as usual the R0 cabling has pushed the pitch modes up in frequency.
Comparison plots pending.
Almost all of the bolts for the mirror holder bases etc. were not tight enough, so we retightened them. In one case a bolt that attaches the post to the base plate was not tight, and I actually ended up rotating the entire post while tightening the bolt for the mirror holder. That was a steering mirror upstream of the green EOM. I was able to recover the alignment easily.
We tried to see the red beat note but couldn't. The fiber power is decent (50-60uW), but we're throwing away a lot for polarization analyzer PD, then we throw away 50% on the BS, and we're only getting 5-8uW at BBPD. But we're sending almost 19mW from Prometheus.
Blocking the beam changes the noise floor of BBPD, we changed the NPRO temperature up and down and saw nothing. BBPD itself seems to be working OK, we connected AM laser to the network analyzer and BBPD was responding.
The mode matching looks suspicious though it could be just a trick on the eye as one beam has much more power than the other. Tomorrow we'll go back and try to improve mode matching.
Ideally the fiber polarization should be adjusted so you're not pitching ~75% of the beam onto the polarization monitor PD. I'd noticed the polarization drift last week and just didn't get around to adjusting the controller.
We started welding in fused silica fibres this morning and now have 3/4 fibres installed. We plan to weld the final fibre tomorrow morning, followed by de-stress and hanging. Travis, Gary, Giles
Aidan, Alastair, Dave H, Dave R, Albert L, Eric G, Thomas V.
Tidied up the plumbing on TCSX. Installed plumbing on TCSY. All three picomotor mirrors installed and confirmed to be working on TCSX.
Hooking up proximity sensors for flipper mirrors. Installed shutter on TCSX laser.
Morning shift covered by Corey
130 - TCS crew energizing laser on optical table at BSC3. Beginning to populate table at BSC1. Did I see Dave Reitze working on this?
various ISI/SUS trips throughout the afternoon as a result of commissioning.
New screens showing oplev path on the QUADS after updating what Stuart did at LLO. I made a small correction and commited SUS_CUST_QUAD_OVERVIEW to the svn. cf attached screenshot
Betsy and I stowed ITMx (see attached picture). Plan for this guy is to get the sealed door on the box later today or tomorrow, then leave it in the cleanroom until its ready to move over to the beer garden. Margot
[Rich M. Arnaud]
ETMY GS13 V2 input signal was sitting at around 6000cts whereas the others at ~100 cts. Turning off/on the digital Gain + DWH switches brought it back to ~1000cts. Better. Though, it seems like there is still an offset that adds up to the signal.
Arnaud, Eric G, Apollo, Thomas We've installed the optical lever with two levels of 1:10 whitening in hardware and complimentary de-whitening in software. They are calibrated to micro-radians and I've attached a plot showing the linear response. The difference in their gains is most likely due to their difference in beam width. Pit_Gaussian_width = 2.02 mm Yaw_Gaussian_width = 1.41 mm H1:SUS-ETMY_L3_OPLEV_PIT_GAIN = 95.50 H1:SUS-ETMY_L3_OPLEV_YAW_GAIN = 69.33 H1:SUS-ETMY_L3_OPLEV_SUM_OUT = 20,000 counts.
(Daniel, Alexa)
We measured the dark noise and shot noise of the LSC RF PD in the refl path of the IMC.
From this measurement we calculated the AC transimpedance of the PD to be 2.1kOhm (with a 100 Ohm DC transimpedance).
Started off morning with many RED systems. Jeff B walked me through getting everyone (SUS, HEPI, ISI) in BSC2 back & then went from there.
Days Activities:
-Had to leave shift at 1:30 and was spelled by Justin B.
Log-Log plot attached with T0 set to start of rough pumping (~700 torr).
There was a magnitude 8 earthquake in Chile, all BSC ISIs tripped.
How did the new BSC SEI guardians do in recovering everything? Where there any problems or hiccups? Did it go smoothly?
Thanks for posting the data, Sheila. We (by which I mean Hugo) will look at this a bit more. A quick look makes it seems that all three vertical drives on 3 of 4 platforms were pushed to the max within 1 second of each other. Many people (W. Hua, Rana, Dan Clark, me, Peter F, et. al.) have suggested that using that using the sensor correction to only isolate against differential motion, rather than trying to get rid of the absolute motion as the whole site heaves up-and-down would be a smart thing to do. I think this will be a good example to look at. -Brian
TIme series were collected on both BS (ground STS, high gain) and ETMX (ground T240, low gain) at LHO.
They can be found on the svn at:
ligo/svncommon/SeiSVN/seismic/Common/Data/2014_04_01__Chile_Earthquake_Data/
For reference, LHO ground sensor time series were also colected for previous earthquake data, recorded on March 3rd of 2014. ligo/svncommon/SeiSVN/seismic/Common/Data/2014_03_10__Earthquake_Data/
Chris, Sheila
Our measurements of the COMM noise with the WFS locked have an rms of 20Hz, measured down to 0.1 Hz. To make sure this was reasonable, we turned off the refl bias path so that the laser was locked only to the green beat note. The transmitted power fluctuations are noticably less than the HWHM of 40Hz, so this seems believable. (StripTool is the second screenshot attached).
We wanted to try locking the arm half way down the fringe so we could make a measurement of the noise using the transmitted power. Since the IR trans PD is normalized to 1 when the cavity transmission is maximized, the calibration of this signal is just a high frequency gain of 84 (FWHM of the arm cavity), and to correct for the refl bias loop gain we have a pole at zero and a zero at 3Hz. The correct calibration has a more complicated frequency dependence, so we would have to do the real calibration to make a comparison above about 42 Hz. Neither of these traces have the cavity pole at 42 Hz corrected for. You can see that the side of fringe measurement agrees with the normalized refl PD signal well up to 100 Hz . (1st screen shot attached)
We repeated the measurement in the normal configuration, where we lock on resonance using the refl bias path, and the noise is the same (30Hz rms down to 0.02Hz). The RMS is dominated by the pitch mode, so we will work on increasing the WFS bandwidth tomorrow. Tonight we tried OpLev damping (this added noise around 1 Hz) and turning up the pitch OSEM damping gains (inconclusive so far).
We redid the measurement with the WFS outputs held, and saw that the noise was high again (100Hz rms down to 0.1Hz), with the WFS on again we were back to the low noise state. So we can safely say the WFS are helping.
We also checked that the noise is the same from 3-100Hz with and without the refl bias path engaged.
It's very nice to at last see a benefit from the WFS aignment control! good job.
Here is the plot of the Noise with WFS on and with WFS off, for those who want to see it.
