IO work 13 Dec 2017

• PSL beam aligned to the MC1 and MC2 irises
• IMC flashed, good 0-0 modes
• 0-0 modes were used to align MC2 Trans
• this means that we have a known good IMC alignment that centers the beam on MC2 Trans
• strong 0-0 modes flashed in each MC2 Trans QPD quadrant at around 700 (raw signal)
• variations in the height of the flashes in all quadrants was 20-30 counts, which is around 5%
• beam from the PSL was monitored, and checked on MC1 iris more than once today, and it appears that later in the day something happened, because my last check of the beam showed that it was off in yaw
• IM3 required a pitch adjustment, and this was accomplished without removing the baffle, because the locking collar on the pitch adjustment screw wasn't tight
• tightening the collar required removing the IM3 SiC baffle
• the Viton remained stuck to the SiC when the baffle was removed, which made reinstalling the baffle easier, however
• the weight of the SiC baffle further adjusted the IM3 pitch, so there is a pitch slider actuation
• the locking collar could be put on the back side of the adjustment screw, which would potentially make pitch adjustable without removing the SiC baffle
• each time it's required that we remove and handle a SiC baffle the risk of having one shatter increases
• IM4 is in the same situation as IM3, the pitch adjustment screw and collar are not currently adjustable with the SiC baffle installed on the IM4 cage
• the flashing beam from the IMC is aligned to the PRM Trans and PR2 irises

Replaced ITMs ESD Feedthroughs

(Kyle R, Gerardo M)
Right before lunch both feedthroughs were removed and replaced while all door covers were on.  We removed the old style feedthroughs and installed the new type (Accuglass PN100520) on BSC1 and BSC3 (flange F2, subflange 3 for both BSCs).  Both need to be leak checked.

completed install of new python scripts to set/store ISI cartesian bias settings

WP7245

To complete the transition of the ISI Cartesian bias setting scripts from Perl to Python, today I edited all the MEDM files which call these scripts. In all cases the scripts are called indirectly via an xterm running a wrapper bash script. I changed the dimensions, font and colors of the xterm to make it more readable. Again I used ISI ITMY ST1 as my test system, returning the settings after I had finished.

H1 MC2, PR2, SR2 Scrapper Baffle B&K Results, Comparison against L1 Results

J. Kissel

I've process several sets of data to show results of B&K measurements which characterize the new scraper baffles in HAM3 / HAM4, the MC2 scraper baffle (), the PR2 scraper baffle (), and the SR2 scrapper baffle (), all of which are assembled with viton built into the structure, so we expect very little high-Q features.

Attached are results from all three LHO baffles, in addition to comparing the same DOFs measured at LLO (though, LLO used a laser vibrometer, and only measured the response to 2D excitation in one direction).

It's difficult to make any broad sweeping statements about any of these baffles.
I would have naively expected that the PR2 and SR2 baffles from both sites would have similar resonant features in the comparable excitation + response directions, but they're all different. Similarly, the MC2 scraper baffle shows a different response. I could easily imagine the response being different because of the differing hit locations -- from the looks of pictures, Arnaud / Stuart hit the centers of big surfaces on each baffle (think excitation of drumhead modes of those panels), where Siddhesh and I were hitting the top corners of structures (think excitation of flag-pole modes of the entire structure). Also, the response sites were different between observatories: the laser vibrometer can directly measure (at least one) surface of each baffle, where as the accelerometer bolted to the table picks up ISI table resonances (i.e. the high Q features around 10-15 Hz which are the locked / HEPI cross-beam resonances).

If we're really looking to confirm that these are built identically, and have confirm that *all* resonances are damped to a sufficiently low Q, then we're going to need to standardize our measurements.

For now, all we can do is acknowledge what frequencies we see where, and hope the don't show up in the IFO's sensitivity!

For those who need more detailed information, I attach the data (zipped up in an archive that you'll need to decompress), and the script to process the data to make these plots.
 
Details of the measurements can be found in the following aLOGs:
H1 MC2 / PR2 LHO aLOG 39590
H1 SR2 LHO aLOG 39288

L1 MC2 / PR2 LLO aLOG 36908
L1 SR2 LLO aLOG 37020

BSC1, 2, 3 SUS ESD and cable ground work finished

After another marathon day in chambers, Fil, Richard, Travis, and I finally finished connecting the newly fabricated ITM ESD feedthru connections in-chamber and their subsequent testing (FRS Ticket 5033).  The testing, of course, was not straight forward and we walked around troubleshoot a few things here and there.  We struggled with the ITMX BIAS intermittently failing the HIPOT testing and continuity checks (see issues with this reported last year 27562).  We tracked down the issue to the pin at the other end of the cable we reterminated (and have not touched during this vent) which led to me disassembling the connector a bit and looking for shorting mechanisms.  I didn't find any, but upon reassembling the connector, things seemed to test fine.  Sigh.

Fil will likely post the full numbers above or below this alog for both ITMX and ITMY testing (which included HIPOT).

Meanwhile, the ground checks of the SUS cables that were checked on ITMX, ITMY, and the BS which failed a few days ago were fixed via a variety of reseating the connection fasteners at the feedthru end of the cable, remedying pinch points of cables in-vacuum, or resolving to have EE fix it somewhere on the outside.  Most of the cables had not been touched during this vent, with the exception of the ITMX PUM one which had migrated out of it's viton pad seat during the ITMX rebuild last month.

So, really, really we could do the BSC 1, 2, 3 closeout tomorrow.

Ops Day Shift Summary

0.85Hz Peak on the ETMX-ISI

Over the last week we have noticed an anomalous peak in the amplitude spectrum of various sensors on ETMX-ISI, reminiscent of the 0.6Hz peak on HAM 2(?) back when

Looking more closely it seems to be mostly showing up on  a few of the CPS channels, St1-H1, ST2-H2 and H3, see the attached screen shot. This data is with the loops on

H1 SUS IM Cages: Viton Added, Further Torquing of Cage -- Now Much More Consistent Resonances; Transverse Mode Damped

J. Kissel, J. Warner, C. Vorvick

After Cheryl / Sheila / Hugh made another pass at torquing down IM cage bolts and applying viton between their baffles and cage (LHO aLOG 39654), Jim and I measured the new cage response with the B&K hammer / accelerometer system (LHO aLOG 39682). 

Attached are the results.

The cages now look *much* more consistent with each other. Sadly, it appears that the viton only effectively damps the transverse mode of the cage. However, the longitudinal direction's Qs are still not *really* high, and the frequency (now just one! yay!) is still above ~150 Hz, i.e. in the "safe" region for ISI control. 

Note, that this was taken before Cheryl's final battle with the IM3 cage (see LHO aLOG 39702), which she now believes has been resolved. Before that change (i.e. this data set), IM3 looked good and consistent with the other cages (see attached), so we can imagine it only has gotten better / more consistent with others. Thus, we won't bother to re-hammer unless someone strongly wishes it.

 
New data has been committed to the svn here:
/ligo/svncommon/SusSVN/sus/trunk/HAUX/H1/
    IM1/BandK/2017-12-07_H1SUSIM1_X_LongExc_Xresp.txt
    IM1/BandK/2017-12-07_H1SUSIM1_Y_TransExc_Yresp.txt
    IM2/BandK/2017-12-07_H1SUSIM2_X_LongExc_Xresp.txt
    IM2/BandK/2017-12-07_H1SUSIM2_Y_TransExc_Yresp.txt
    IM3/BandK/2017-12-07_H1SUSIM3_X_LongExc_Xresp.txt
    IM3/BandK/2017-12-07_H1SUSIM3_Y_TransExc_Yresp.txt
    IM4/BandK/2017-12-07_H1SUSIM4_X_LongExc_Xresp.txt
    IM4/BandK/2017-12-07_H1SUSIM4_Y_TransExc_Yresp.txt

Data may be processed with the following scripts:
/ligo/svncommon/SusSVN/sus/trunk/HAUX/Common/MatlabTools/
    haux_bandk_datalist.m
    plotHAUX_bandktfs.m

H1 ITM Elliptical Baffle (Surrounding BS in BSC2) Structural Resonances

J. Kissel

I've processed the B&K measurements taken of the ITM elliptical baffles (see LHO aLOG 39678). 

These baffles are under suspicion after Robert found they couple into DARM (see LHO aLOG 39290).

They have since been rebuilt peppered with viton washers (see red lines in D1101806), and this data is *after* that reassembly, while the baffle was fully suspended.

Results, raw data, and script to plot it attached.

IO work 12 Dec 2017

• the IMC is flashing and the beam is through the IO Faraday
• I had to align from the PSL twice today, due to finding the beam on the MC1 iris was off in yaw about half way through the day (plot attached)
• HAM2 tool stock has dwindled, so it took a while find a short allen key that I could use on the bottom left IM3 OSEM (clearance issues with MC3)
• there was significant fringing at times during the day, that made alignment a challenge, and this appears to be due to ground/beam tube motion
• IM3 needs a major realignment, and while this is going well, I am still working on getting the beam aligned to where it needs to be, and at the same time, getting all of the OSEMs centered enough that I can damp the optic
• I adjusted, for the first time, the pitch of IM3, and found that I could easily turn the bolt under the optic that controls the pitch
• I should not have been able to easily adjust that bolt, since that bolt has a collar that is supposed to be tightened to secure the bolt in place
• my finding today suggests that that bolt has not been secured with the collar for a while, so likely it's been loose since before O1
• chalk this up to another way the alignment of IM3 could have changed while the chamber is under vacuum, for example, during an ISI trip (shaking)
• I'll attempt to tighten the collar tomorrow without taking the baffle/Viton off of the front of the IM3 cage, but if I can't, I'll have to take the baffle and Viton off to access the locking collar

ITM ESD feedthru connection work today

More details coming, but Fil, Gerardo and I spent most of today rebuilding the ITM ESD feesthru connectors to the latest rev.  This involved pulling it from the feedthru on the inside of the chamber port, cut the connector off the 5-way cable, strip and recrimp pins, and reassemble inti a new connector. We had a few failed attempts on the ITMY set but finally figured it out. 

We also found the 5th (LL quadrant) pin at the CP on ITMX not well seated in it's connector which caused us some grief. We eventually just squeezed the inner pin to make it work since we cannot fix this in any other way. The continuity checks then passed.  

TBC...

Purge air update

I had a little time to investigate the seemingly inadequate purge-air issue after the in-chamber folks were done for the evening and all of the soft covers were on.  Nothing definitive was determined but perhaps one clue was added.  I'll need to have a second person to confirm (one in LVEA and one at the Kobelco) but tonight it seems that the frequency of "low flow" is greater than the drying tower switching frequency.  If so, my original observation was wrong.  

For tonight, I increased the pressure setpoint at which the compressor loads from 80 psi to 90 psi..  This will only help if the issue isn't drying tower dependent.  Chandra and I should be able to investigate tomorrow but we may need to shut off the system.  

As an aside, I noticed that none of the lights are working on the HAM1/HAM2 or HAM3 clean rooms?  Both appear to be plugged in and, for the case of the HAM3 clean room, the breaker isn't tripped.   

porting ISI cartesian bias set/store scripts from perl to python

ECR1700056, WP7245

Hugh, Jim, Rich M, Dave:

I have ported the perl scripts storeCartBiasTargets.pl, setCartBiasSetpoints.pl to new python scripts (storeCartBiasTargets.py, setCartBiasSetpoints.py). The new scripts take exactly the same arguments as the original scripts, so no changes to the calling systems are needed.

Hugh and Jim allowed me to test my code on H1 ISI ITMY ST1 this afternoon.

In the old system, the 'set' command was called from MEDM (ISI_CUST_CHAMBER_ST1_BIAS.adl) within an xterm calling a wrapper shell script, while the 'store' command was called directly. The xterm allowed any errors from the script to be displayed.

In the new system, both commands are called via an xterm window calling a wrapper shell script (storeCartBiasTargetsWrapper.sh was created for this purpose). The new scripts perform more error checking and user feedback which is shown in the xterm.

After testing on H1 ISI ITMY ST1, I restored all the settings back to their original values (using a ramp time of 5 seconds). The new code is now ready for rollout on all ISI systems.

 

TCS Steering Mirrors in BSC2 (SM2's) Support Structure Resonances

J. Kissel, S. Pai

I've processed the results of using the B&K hammer accelerometer system to characterize the resonances of the support structure for the TCS Steering Mirrors in BSC2 (SM 2, D1101851). These have been of concern after Robert identified them as potential scattered light sources in LHO aLOG 39290. Since, they've had standard SUS vibration absorbers (D1002424) installed to help reduce the Q of their resonant features (see preliminary testing at T1700535).

We have no before-vibration-absorber data other than Robert's spectrograms, but I attach the new results with vibration absorbers installed.

I also attach a crude screenshot comparing these results against Robert's spectrograms, and the resonant features roughly line up. #sanitycheck

The results show that, while the majority of the resonant features *above* 30 Hz have a nice low Q, there remains a feature at 25.62 Hz (for TCSY's mirror) and 20.88 Hz (for TCSX's mirror) that is still particularly high Q. 

Edit -- I attach a zoom of these ~25 Hz resonances, as well as the raw data and a script written to analyze it.

Ops Day Shift Summary

Cheryl & Ed Still at HAM2

Cheryl & Ed are continuing IO work at HAM2 past 16:00. 

Corner station closeout progress

• ITMy had its "pizza crust" FirstContact painted on the perimeter of the FC spray sheet that was applied last week.
• CPy had its FC painted on.
• CPx had its FC painted on.
• ITMy PUM cable ground loop fixed.  Cable was routed through a pinch point in the Top Mass.  The cable was moved out of the pinch point and the short was alleviated.
• BS Top Mass ground loop checks were started, but aborted until tomorrow until better access to the outside of the feedthrough can be established.

For tomorrow, we will need to swing the ITMy ACB back into its nominal location such that we can recover our only ACB swing back wedge (the second wedge we borrowed from LLO was returned earlier this month).  Swing back ITMx ACB and apply spray FC and "pizza crust".  Finish out BS ground loop checks and start ITMx ground loop checks.