Repaired dust monitors were put in the LVEA in the "Beer Garden" (H0:PEM-LVEA_DST5) and near the previous location of the H2 electronics racks (H0:PEM-LVEA_DST7). One was also put in the OSB bake oven room (H0:PEM-LAB_DST3).
The latest SUS QUAD build in the Staging Building is designated "QUAD 03 BUILD 03" in the SVN directory. All data should be saved in the "/ligo/svncommon/SusSVN/sus/trunk/QUAD/X1/QUAD03/BUILD03/" directory.
J^3 (Bartlett, Garcia, Kissel), A. Ramirez Jeff and I grabbed the first set of transfer functions from X1 SUS QUAD03 after Jeff and Andres have brought it up-to-date with retrofits. Note, that this is the "first brush" suite -- These are taken before reaction chain cables are installed, to ensure the dynamics are expected. The results are attached. I've compared these measurements against the three other ETM-flavor QUADs we've build across the project, X2SUSQUAD13 @ LLO X2SUSQUAD14 @ LLO X1SUSQUAD04 @ LHO (soon to be H2SUSETMY) Not that in this latest round of measurements, the OSEM Open Light Current has not been normalized, so the ~25% discrepancy in overall DC amplitude of the transfer functions (for example R0 V and R) should be attributed mostly to uncompensated variation in sensitivity of the OSEM sensors. The results look great; I approve the suspension to go onto installing reaction chain cables. I've left the suspension with damping loops on and running. If I were to be picky, the only thing that I've found that raises a yellow flag, which we should keep an eye on in future measurements: Reaction Chain Lowest Pitch mode is high, but no other pitch dynamics appear to be effected. As a side note, now that we've got more statistics on ETM-style QUADs -- it looks like there's a good deal of excess cross-coupling in the measurements of X1SUSQUAD13, and the other SUS's show that we can build these with much less. We should keep an eye on this when we move to retro-fit/upgrade/add cables. Data Taken from DTT files here: /ligo/svncommon/SusSVN/sus/trunk/QUAD/X1/QUAD03/BUILD03/ SAGM0/Data/2012-01-06_X1SUSQUAD03_BUILD03_M0-*-WhiteNoise_TF.xml SAGR0/Data/2012-01-06_X1SUSQUAD03_BUILD03_R0-*-WhiteNoise_TF.xml and exported to .txt in F1F2F3LFRTSD LTVRPY order to .txt files of similar name as above (with "_tf" tag). Data initially processed by: /ligo/svncommon/SusSVN/sus/trunk/QUAD/Common/MatlabTools/plotquad_dtttfs.m Data compared with others with: /ligo/svncommon/SusSVN/sus/trunk/QUAD/Common/MatlabTools/plotallquad_dtttfs.m
Before closing the doors on the HAM-ISI in the staging building (assembly 1), we measured ASDs of the instruments. The horizontal GS13 in corner 3 (POD SN68 – GS13 SN805) had a bad response at low frequency (see spectra) which is a typical characteristic of a stuck proof mass.
After opening the vacuum pod, we verified that:
- The kinetic feet could not move and the cross bar was not loose.
- We could feel the mass moving when the geophone was tilted
- The mass is centered when the geophone is horizontal (Level on the second circle)
We measured a spectrum while making sure the geophone was horizontal and the proof mass was centered. The spectrum still looks bad.
After taking apart the geophone, we could feel that the mass is not moving freely (sticky). We will replace this geophone with a new one.
(Corey, Eric, Jim, Mitch, Vincent)
The first part of the day was devoted to getting some "before" measurements on the BSC8ISI; that is, before the Vibration Aborbers were installed. Unfortunately, it sounded like clean measurements couldn't be made for both Stages (Sensors were saturating on the system [most likely due to nearby loud hammering/drilling]). Because of this, we were only able to do work on Stage1.
The Stage1 Vibration Absorbers are to be installed on the Outer Walls/Doors of the Trillium & Horiz L4C area. It was quickly discovered that we could not install the Vibration Absorbers. This is because there are a pair of dowel pins (for masses) which are exactly within the bolt pattern for the Vibration Absorber (!).
So, we decided we'd hammer out the pairs of 1/2" dowel pins on the three doors (that's a total of 6 which needed to be pressed out). And we obviously didn't have the equipment to do this, so we MacGyvered tools with what we had (a big mass for a hammer and an allen wrench as a punch). To hammer these pins out, we had to completely remove all three doors which entails removal of lots of bolts, removal of mass from doors, powering down the equipment at the SEI rack, disconnecting cables from the Trillium and L4C, and then carefully taking out these doors to a work table for pin removal.
The doors were eventually re-installed (not torqued), the ISI was covered, and left in that state overnight.
This morning the crew is already set for starting the work we intended to start last night.
We're running transfer functions over night of both ITMY and FMY top stages, hopefully it doesn't crash. cdsimac0 cannot put in excitations through Matlab I did not have the patience to investigate, so I logged into cdsws5 and am running them there "remotely". The 3 line script calling the three items to run (ITMY M0, R0 and FMY M1) is in SusSVN/sus/trunk/Common/MaltabTools/SchroederPhaseTools/. That will give you the info on which scripts are used, all committed to svn. For future reference, the BSC-ISI is locked and possibly in a funky position.
Attached are plots of dust counts > .5 microns. High levels of dust in the LVEA.
After Friday's FMY damping loop sign changes, further investigation indicated the *M1_COILOUTF_*" actuation gains were flipped for each OSEM Coil. This was most likely caused by an outdated burt restore file used after the front-end reboots performed the previous day. This resulted in the damping loop gains needing a positive sign for signal damping. The signs on the *M1_COILOUTF_*" gains are now flipped from Friday's values. The FMY M1 damping loops now have all negative gains and have been closed successfully. FMY M1 COILOUTF_* 01/06/2012 before changes 01/09/2012 after changes F1 = 1 F1 = -1 F2 = 1 F2 = -1 F3 = -1 F3 = 1 LF = 1 LF = -1 RT = -1 RT = 1 SD = -1 SD = 1 A new burt snapshot file was taken for this state to prevent this issue, located here: "/opt/rtcds/lho/h2/cds_user_apps/release/sus/h2/burtfiles/fmy/h2susfmy_20120109_new_gains.snap"
With the concern expressed by the AOS/SLC/ACB crew, I thought it reasonable to look at the level of the Solid Stack Panel(SSP) to which the ACB is bolted. The components of the ACB cover most of the panel but there is a few inches available at the East end and just a couple at the West end. The SSP is ~10" wide (N-S) and 30" long (E-W). Our 'as-built' measurements using the Sokkia Optical Auto Level put the SW corner of the SSP 0.9mm lower than the opposite NE corner. The other corners, NW & SE, were within .2mm of each other: Start level wrt average: NW 0.0mm SW -.5 SE +.2 NE +.4 0.9mm/30"(762mm)==1.2mrad. For some reason out on the floor I calculated about 3x this...not sure what I did. Anywho...Lisa said the requirement was 0.14degrees or 2.44mrad so we were already there. Still given my miscalculation earlier we made it even better. We loosened the bolts holding the SW corner of the SSP to the I-Beam on which it sits. This allowed that corner to lift up. I then slipped a pseudo ClassB 20mil shim between the SSP and the I-Beam and re-tightened the bolt. Despite the 20mils, this only improved things a little with other things playing into the game. The West side to East side tilt only improved to 0.7mm but that is of course well within the 0.14degree spec. Final level wrt average: NW -.3mm SW -.3 SE +.4 NE +.3
Attached are plots of dust counts > .5 microns. Very high levels of dust in the LVEA (H0:PEM-LVEA_DST11_5). High levels of dust in the optics lab (H0:PEM-LAB_DST1_5). Some dust in the clean room over BSC8 (H0:PEM-LVEA_DST15_5).
EricA CoreyG HughR On Wednesday we started this process of connecting the Actuators to the Crossbeam (Foot). We completed this yesterday long with 130lbs on extra weight on the West Crossbeam to mimic the ACB. We then released the HEPI suspension and then made adjustments to the big HEPI Springs to bring the dial indicators back to pre-disturbed position. We got pretty close to this but there remains some differences. I'm confident we remain in tolerance for the translational position but the Yaw may need tweaking. Today we made gap checks of the HEPI Actuator actuator-plate/bellows shield. This is our motion limit and we need to insure it remains centered. We've made some adjustments of the Actuator positions to improve this centering. We still need to do a couple more. We also ended the day yesterday with a stuck L4-C (Horizontal NW(Pier2)) but Eric pulled that out and gingerly freed it. It has been reinstalled.
Here are photos from Actuator installation (most of them are for the Vertical Actuator install).
Table legs were set into place and leveled. On Monday the table legs will be grouted into place.
H2 SUS FMY's damping loops are no longer stable with the settings I'd confirmed working yesterday. HOWEVER, they are stable if the signs of all the loops are changed (i.e. change H2:SUS-FMY_M!_DAMP gains from negative to positive). I had logged in this morning (~10a ET, 7a PT) to take some transfer functions on ITMY, noticed in passing that H2 SUS FMY's watchdog had tripped and thought "no big deal," something must have glitch and tripped it over night, and left it be. When I had finished ITMY's measurements, I went back to clean up my work and re-damp everything so that seismic could get to work, and noticed (1) FMY's Master switch was off (was on when I glanced at it) (2) FMY's COILOUTF filter's outputs were off I, by-hand, turned the switches back on, and blindly turned on the the damping loops. The slowly but surely became unstable within a few seconds. Totally confused, I flailed about for a bit looking for sign errors etc. Regrettably I didn't BURT capture FMY after yesterday's confirmation that the damping loops were running yesterday. So, I've BURT restored to the snapshot where the loops were confirmed stable last, in ${RTCDSROOT}/userapps/release/sus/h2/burtfiles/fmy/fmy_dampfilter_111004.snap as per Jeff G.'s aLOG 1506. All loop signs are the same as what I expected (and as I had them set to yesterday) -- the only tweak I needed to do was to select the correct filters in the COILOUTF banks, because Jeff G's snapshot was taken before the filter bank upgrade, mentioned here. After this they're still unstable! In trying to figure out what the heck is wrong, I flipped the sign on the DAMP Filter gains from - to +, and they're stable (?!?!). Still investigating ... Richard has to make some changes to the DC powr supplies at 11a PT, so for now I've turned off all digital control to both ITMY and FMY. I'm looking for something electronic (chassis, ADCs, computers, networks, etc.) or digital (MEDM setting, model changes, CDS software changes) that might cause the sign of the actuation to flip -- and only on FMY. It's VERY strange. Perhaps the timing glitch, and Dave's subsequent reboot of everything and BURT restore, mentioned in passing here, changed something digital upstream of the FMY signal chain (like the IOP)? Sounds unlikely, but possible ...
Upgrade of control room and MSR update. Floor tiles were removed, under floor cables are being removed. Timing system in the MSR was glitched, requiring all front ends plus their IO Chassis to be power cycled. All systems were burt restored to 15:00 local time.
Attached are plots of dust counts > .5 microns. The dust monitor in the clean room over BSC8 had a low battery alarm when I came in today, and I restarted it. The system was down intermittently while the servers were moved from the MSR to the electronics building.
Final work on linoleum was completed today. All seams were welded, and grout dams were caulked into place. Table legs were also moved over to the South bay.
(Corey, Hugo, Jim, Mitch)
Assy #3
For all intents and purposes, this Assembly is ready to be pulled off the Test Stand.
Assy #1
After powering up the FMy satellite boxes and checking that the EE on the outside of the chamber were functioning properly (thanks, Richard), I went in the chamber to look at the in-vacu stuff. Indeed, the cable that needed an end-to-end swap before the holiday break, had not finished making it to it's connecting octopus cable. When I tried to seat it, I found that the 225 cable needed helicoils, preventing good connection. We're hunting tools now to finish helicoiling these (all 3 225 cables need the helicoils) connectors.
Today I finished helicoiling and connecting the FMy cabling at the optical table. While this eliminated a possible source of problems, it turned out the real problem was a bent pin on the air side of the feedthrough. After straightening the bent pin, the cable was reconnected and successfully tested for functionality.
Though there was a bent pin on the Vac. flange I believe this was a result of troubleshooting and not the initial cause. Per my directions this cable was removed and changed and replugged in a number of times in trying to narrow down the problem.
