As per WP 11669 I upgraded the link between sw-msr-core and sw-gc-sysadmin to a 10G link. This is done to simplify the configuration of the new core switch when it gets put in later this month, by moving everything to 10G everywhere.
TITLE: 02/08 Day Shift: 16:00-00:00 UTC (08:00-16:00 PST), all times posted in UTC
STATE of H1: Planned Engineering
INCOMING OPERATOR: None
SHIFT SUMMARY:
Today's main activity was closing out HAM6 (more specifically--being able to steer the OMC trans beams out of the chamber (after the installation of the modified OMC shroud plate)---the beam they worked with was very faint and hard to see (most of the LVEA lights needed to be off for this work!). LVEA was then transitioned to SAFE, and then HAM6 ISI was unlocked and balanced for SUS transfer functions.
There was wind fence work and lots of VAC team work as well.
LOG:
While Jason and I were doing some FARO surveying we noticed a cable being crushed under a computer rack leg in the bier-garten. Following the cable led to an accelerometer by GV6 on the Y-arm beamtube.
I think this accelerometer is H1:PEM-CS_ACC_BEAMTUBE_YMAN_X, in which case its signal went very noisy around 9am Tue 16th Jan (first day of the break).
Oli P, Ibrahim A, Betsy W
The first transfer functions were taken for the top stage of the BBSS a couple of weeks ago (75211), and we have been working on plotting these transfer functions against their models. We have finally been able to do this (pdf), and what we see is a pretty good match to what we expected to see. Plots 1 - 6 of the pdf (also attached as pngs) are the transfer functions for each individual degree of freedom compared against the model's result for that degree of freedom. Plots 7 - 12 compare the measured and model TFs between dof's that tend to couple into each other. Pages 13 - 18 compare each measured and modeled dof to the OSEMs that make up that dof.
Here are the plots for M1 L to L, M1 T to T, M1 V to V, M1 R to R, M1 P to P, and M1 Y to Y. The results are similar to what we noticed when manually comparing the location of the peaks.
Scripts Used
The original BBSS model is a Mathematica notebook found at $(sussvn)/Common/MathematicaModels/TripleLite2/mark.barton/20211212bbss/production/stdcalc/ASUS3L2ModelCalcPlots.nb.
The default model parameters in Matlab can be found at $(sussvn)/Common/MatlabTools/TripleModel_Production/bbssopt.m. Then, $(sussvn)/BBSS/Common/MatlabTools/plotBBSS_dtttfs_M1.m was run to grab the data from the transfer functions we ran and plot them against the theoretical models.
The results are found in $(sussvn)/BBSS/X1/BS/SAGM1/Results/2024-01-05_1000_X1SUSBS/.
Questions we have:
- Why are the measured magnitudes higher than the model magnitudes?
After looking over it for around 4 hours and getting some help from Jim and Rahul, we could not figure out whether the discrepency was due to:
- an incorrect calibration factor (the 1.5404 is a calibration factor that was a remenant from the BSFM script who's values seem to make sense)
- an issue with the filter units on the test stand or dtt
- an issue with the model
- turbulance from cleanrooms
- something else??
- What's up with Pitch?
Seems coupled to "V", which should have expected coupling but not in this peak (according to model). Ideas are:
- issue with model? Were "new wire lengths" added to model params?
- issue with how we built it? Pitch being off can sometimes indicate an issue with the location of the center of mass
- something else?
- What's with noisy P to "any other DoF" coupling traces?
There is terrible coherence.
- Some mechanical issue? No rubbing from inspection ~month ago
- Some DTT issue? There are no GDS overflows/railing (which is what caused the problem last time) when running TFs
- Something else?
Next steps:
- Take many measurements of different parts of the BBSS and make sure they are within their build tolerance, as well as double checking that all updates to the build have been included in the model parameters.
- Make any adjustments to the build or model that are needed.
- Rerun TFs and see if the new results make sense.
Thu Feb 08 10:09:10 2024 INFO: Fill completed in 9min 6secs
Gerardo confirmed a good fill curbside.
[Michael, Jim, Tony, Shoshana, Neil] We've spent the week working on installing a remote mass adjusting/balancing system for the End-X BRS. Upon opening the End-X BRS we discovered that it is slightly different than the other BRSs (most likely because it was the first one built, photo 1 below) so we had to get a couple extra parts machined for the parts we brought from UW to fit properly. The hardware has finished being installed (photo 2), and we've run a quick test of the picomotor (Newport 8341-UHV) to make sure it has full range of motion. We've finished closing the BRS up and are now waiting for it to pump down before doing some additional in-vacuum tests. We are writing up some installation and balancing instructions as well as leaving another set of parts, so a remote balancing system can be installed on the other BRS. We assume the End-Y BRS should fit the drawings on the dcc (D1800045), so no additional parts should need to be machined. We are also leaving 16 flexures here with Jim. The pico motor controller chassis was not working, so we are leaving our UW picomotor driver (Newport 8742) here so that the End-X BRS can be adjusted until the chassis is fixed. Once it is fixed the in-air wiring will need to be re-connected to get the motor to the chassis. For now adjusting will have to be done with a laptop at the End-X station, but once hooked up to the chassis, adjustments can be made from the control room.
Yeah! Jim's back will be grateful.
Notes about the design and additional details on the installation are in SEI log 1886 and attachments.
Tagging EPO for these new BRS innards photos.
For FAMIS #26281: All looks mostly well for the last week for all site HVAC fans (see attached), EXCEPT a TINY increase for the (lvea) MR FAN3 almost 5-days ago.
For FAMIS 26230:
Laser Status:
NPRO output power is 1.81W (nominal ~2W)
AMP1 output power is 67.75W (nominal ~70W)
AMP2 output power is 136.8W (nominal 135-140W)
NPRO watchdog is GREEN
AMP1 watchdog is GREEN
AMP2 watchdog is GREEN
PMC:
It has been locked 5 days, 20 hr 2 minutes
Reflected power = 18.46W
Transmitted power = 108.3W
PowerSum = 126.8W
FSS:
It has been locked for 23 days 0 hr and 30 min
TPD[V] = 0.6882V
ISS:
The diffracted power is around 1.8%
Last saturation event was 23 days 1 hours and 2 minutes ago
Possible Issues:
PMC reflected power is high
FSS TPD is low
ISS diffracted power is low
NOTE: Perhaps an old issue, but noticed TCSy Chiller's innards were noticeably more green than x.
Addressed TCS Chillers (Tues [Jan23] 828-838am local) & CLOSED FAMIS #26166:
While running the dust monitor check this morning, LVEA6's dust monitor has a warning requesting it to be checked.
TITLE: 02/08 Day Shift: 16:00-00:00 UTC (08:00-16:00 PST), all times posted in UTC
STATE of H1: Planned Engineering
OUTGOING OPERATOR: None
CURRENT ENVIRONMENT:
SEI_ENV state: CALM
Wind: 4mph Gusts, 3mph 5min avg
Primary useism: 0.02 μm/s
Secondary useism: 0.28 μm/s
QUICK SUMMARY:
HAM6 & HAM7 work continues as we are still laser HAZARD in the LVEA.
The ITMX cps had started glitching this week, I tried reseating the cards to fix the glitching, but I guess I forgot to reconnect one of the sensors. The St1 h1 sensor is railed, so the watchdogs nominally can't be reset. I've temporarily bypass the st1 cps in the wd (via an epics variable that raises saturation threshold above the adc range), but the bypass will reset, eventually. I'll reconnect the st1 h1 sensor tomorrow morning.
I plugged the sensor back in this afternoon. Bypass not needed now.
Shroud panel update:
For the OMC Trans video beam, we further abused the shroud panel successfully move it enough such that the beam comes out. However, that made it somehow impossible to mount the panel on the cage using at least three screws.
We debated how we should secure the panel robustly, but in the end concluded that enlarging existing hole (by milling, good job Tyler!) is the way to go. That way we don't have to do crazy things for hardwares, the panels can stay where they are supposed to be. See Betsy's pictures for the milling vs drilling vs whatever-ing. The modified panels will be installed tomorrow.
We already installed a viewport simulator on HAM6.
Grounding problem was fixed:
ASC-AS_C grounding was fixed. We first touched the AS-AS_C cable on ISI table and nothing changed, we jiggled the cable coming from the top of the ISI to stage 0 and nothing changed, then we jiggled the section between stage 0 and the feedthrough and nothing changed.
Just to make sure, Rahul disconnected the DB25 connection on ISI table and the grounding was gone, which was a surprise. In the end, it was the unused QPD cable assy that was laying on top of the ISI table. ASC-AS_C and the unused QPD cable assy share the same DB25 connector. The shield collar on the unused PCB was touching the ISI table, causing problem for ASC-AS_C. We relocated the unused PCB so that it won't touch. See Rahul's pictures.
HAM6 grounding issues:-
Picture 1: shield collar on the unused PCB touching ISI table.
picture2: shield collar on the unused PCB readjusted and then grounding issues were gone.
picture3: showing in use QPD cable assy cable clamping/routing.
After some failed attempts to cut a hole out of a glass panel and retain the panel (not a thing stain glass'ers do apparently with standard cutting tools, according to google), we switched methods and got diamond glass hole saws. These chewed up the glass pretty badly (quite a few practice rounds attempting to improve such as faster, slower, more water, taped, sandwiched, etc). And we were leary of where the paint from the hole saw went. #thumbsdown.
Then Tyler decided to try milling a square out of the glass. Some more practice rounds and a new tooling fixture plate later, he managed a clean-ish looking square cut around the original 0.6" beam hole. There are 2 chips at the square but the square is 1.5" on a side so these chips should be well out of the beam way and we'll tyr to improve the setup tomorrow for panel 2. It may have some chips, but this solution seems to be the best mediocre solution we have come up with with the HAM6 team. Attached is the photo assay, ala Tyler and I, of the evolution in glass hole cutting which took most of the day. Note, the last picture showing the new square hole has egregious reflections of the room light in yellow rectangles and dirty smudges all over which will be cleaned, both of which you should ignore.
https://dcc.ligo.org/DocDB/0117/D1500044/002/D1500044-v2.PDF Is the back panel modified today (not shown on drawing, but for part number ref).
Tomorrow I will clean this piece and help install it at the -Y OMC side of the shroud assy, while Tyler performs the same type of square-hole cut on the +X panel spare (D1500045).
Today's activities: - EX pumpdown status: it is at ~8.2E-8 - EX RGA bakeout: now it is ramped up to 150 deg C: this is the set temperature. RGA scans next Monday-Tuesday - EX-BRS: it is ready for pumpdown. Everything is set, tomorrow the pumpdown starts - The BSC8 AIP railed this morning: it pumped the BSC8 Annulus volume, while it was also backed with an Aux cart - the gas load is pretty high through an adjacent O-ring from the vented volume, so in the future all GV Annulus volumes will be pumped with aux carts with the AIP powered off, to avoid AIP overload - The cable tray of west-side BSC8 has been placed back to its original position - A leak was located on the HAM8 RGA calibrated leak. A pre-vent RGA scan will be done though, just to make sure if there is nothing severely out of the ordinary, and the issue will be handled during the vent
WP 11672
Entry of work done earlier this week
ECR E2200448 1/5/2024 – EX
In support of the installation for centering the BRS by picomotor, a DB9 cable was pulled from ISCTEX (picomotor driver) to the BRS enclosure. Breakout cable D2400020 breakouts the fifth channel from the DB25 picomotor cable.
ECR E2300367 1/5/2024 -1/6/2024 EX/EY
Network and power cables pulled from the SUS-R1/FAC-R1 racks to the Oplever pylon. Part of the installation for remote centering of the Oplevers. Network and power cables terminated. Network cable to be connected to the Beckhoff CU1128 hub.
Continued from previous ALOG75713
Replaced Meggit (Endevco) Accelerometer Power Conditioner with LIGO Accelerometer Power Conditioner. WP11653
S2300076 installed, all signals attached and powered up at 4:03pm PST.
We did have inrush current blow a fuse on the line, we replaced the 10A fast blow with a 3A slow blow fuse and this solved the issue.
The IO chassis did trip when we switched the accelerometer chassis on, I worked with EJ and Jonathan to restore the connection.
To Do:
MX power adapter cable install and signals transferred from Endevco boxes to the LIGO units.
This log goes into more detail from yesterday's work (alog75754) and what Jason and I did today.
After we checked the chiller lines had no leaks without the table in loop, we added the tables in next. The Y table was free of leaks, but the X table had a slow drip from a few od the push to connect fittings on the RF driver (attachment 1). We swapped all of these fittings for new ones and cut the 1/4" diameter tubes ~0.5", with a cutter that we borrowed from PSL, for fresh connections. This worked great and there were no more leaks.
To fire up the laser to wanted to avoid the fried opamps that we've had in the past, so we turned on the CO2X electronics in this order:
No fried opamps this way. When we turned the newly refurbished laser on, it output 53W according to our Thorlabs S322C power head placed just after M1. The alignment was far off in yaw, not even hitting POL1. Jason and I first yawed the laser itself within its mounting bolt holes, which didn't move much, then started to align downstream. We eventually got the beam to the irises before the periscope, but had moved a large majority of the optics along the way - M1, ISSPO, PM1, PM2, M4, M5, M6 (table drawing). Using the FLIR One camera we got the beam to hit the irises in a similar way to before the laser swap (before, iris1 now, iris2 now). We then checked on the table FLIR for mask alignment. Here we noticed that the annular mask was a little off horizontally, so we translated the mask to better match the previous position (before, annular now). Once we inject into vacuum we might need to make some more minor changes on table, but this looks good for now.
The last steps today were to recalibrate the thermopile at the laser head and verify that the power meter is still centered on the beam. The power meter need a small translation to center the beam. The gain in the laser head WATTS filter was previously 8.58755 and we eventually ended up at 13.904. Jason used new_gain=old_gain*Pm /PE , where Pm is power measured with the Thorlabs S322C just after M1 and PE is the power reported in Epics with the old calibration. We iterated this a few times to make sure it was as close as we could feasibly get.
Next steps are to inject the beam into vacuum and check centering with the HWS. The plan is for this to be done just before commissioning starts in a few weeks.
We left the CO2X laser keyed off and ready for the upcoming laser safe work.
Pico alignment onto ITMX done in 75987. Rotation Stage recalibrated in 76008.
The new laser is reporting an incorrect temperature via H1:TCS-ITMX_CO2_LASERTEMPERATURE, plot attached. There is a RTD temperature sensor on top of the laser but there's no reason it should have change.
It could be related to an EDC change /DAC restart 23rd Jan 75522 as IY is also reporting a lower (but still believable) temperature.
H1:TCS-ITMX_CO2_LASERTEMPERATURE is a SLOW channel, is a DAQ(SLOW) model = H1EDC rate = 16 type = FLOAT
Naoki, Camille, Julian, Rahul
Summary: We measured beam profiles in HAM7 and HAM6 using the SQZ beam, after the PSAMs offloading work from 75677 75709 and the alignment work from 75732. We noticed that there is a lot of astigmatism on the beam in HAM6, but not in the HAM7 location that we were able to reach with the profiler.
Details:
In HAM6 we measured at three locations marked by the letters a,b,c in the sketch that Julian will attach. The red lines in the sketch show the irises.
Location 6A is 1.75 inches from the iris, Location 6B is 2 inches from the iris (closer to OM1). At location 6C we measured the beam in reflection off OM1, 23.25 inches after OM1. (The fast shutter is further in the -Y direction in chamber than what is shown in the CAD drawing). We also would like to measure at position 6D, once the second iris is removed.
We also measured beam profiles on the -X side of the HAM7 table, location 7A is 6.25" in the +X direction (closer to ZM5) from the iris, 7B is 8.25" from the iris in the + X direction, and location 7C is 12" from the iris in the +X direction. We weren't able to move the profiler further into the chamber without blocking the beam heading towards FC1.
We measured 13.5% widths, with A1 as the vertical axis and A2 as horizontal.
| location | ZM4 PSAMs strain gauge [V] (PZT request [V]) | ZM5 PSAMs [V] | A1 Width [um] (standard deviation) | A2 Width [um] (standard deviation) |
| 6B | 0.82 (0) | -4.14 (0) | 1392 (8.7) | 1573 (9.9) |
| 6B | 4.38 (100) | -4.14 (0) | 1406 (7.3) | 1584 (6.3) |
| 7B | 4.39 (100) | -4.14 (0) | 2001 (6.4) | 2017 (6.1) |
| 7B | 1.0 (0) | -4.14 (0) | 1928 (5.6) | 1924 (5.9) |
| 7B | 0.98 (0) | -0.52 (100) | 2200 (8.2) | 2212 (9.4) |
| 7B | 4.31 (100) | -0.52 (100) | 2276 (9.5) | 2303 (9.3) |
| 7B | 8.49 (200) * | -0.52 (100) | 2363 (11.0) | 2412 (13.8) |
| 7B | 8.53 (200) | 3.57 (200)* | 2611 (9.0) | 2691 (7.2) |
| 7B | 5.74 (100) * | 3.59 (200) | 2539 (13.7) | 2603 (18.1) |
| 7A | 5.72 (100) * | 0.64 (100) | 2387 (13.5) | 2446 (13.1) |
| 7C | 5.715 (100)* | 0.63 (100) | 2374 (11.0) | 2415 (9.5) |
| 6A | 5.7 (100) * | 0.625 (100) | 2074 (9.9) | 1611 (7.4) |
| 6C | 5.7 (100) | 0.612 (100) | 1670 (10.2) | 1618 (7.0) |
* strain guage still drifting
Stll left to do:
Useful distances from Don:
Don Griffith sent us some useful distances in HAM7 which are useful for interpreting these measurements, based on the CAD model that he updated in in March 2022 based on the as built photos of HAM7 (D1900365-v1).
Attached are new and final sketches (these should be used for future analyses) of the beam profiler positions with distances.
Note: we measured a distance in HAM7 from the edge of the table to ZM5 of 1841.5mm, whereas Don Griffith CAD model shows a distance of 1822.7mm.
Attached are screenshots of the beam profile measurements. The screenshots are labeled to indicate position, ZM4 voltage, and ZM5 voltage.
Attached is the sketch of HAM6 and pictures of the beam profiler position.
Sheila, Naoki
We measured several distances as follows.
From first iris to OM1 in HAM6: 53.75''
From ZM5 to second iris in HAM7: 68.5''
From edge of table to second iris in HAM7: 4''
Then we removed the irises in HAM6 and measured the beam profile at the position of second iris, which is 13.5'' from OM1.
A1: 1285 (6.7)
A2: 1426 (9.2)
We also measured the seed power at several locations. The seed coming out of HAM7 is 1.05 mW and the seed in HAM6 is 0.71 mW, which makes sense since the SRM reflectivity is 68%. The seed reflected by OPO without dither lock in SQZT7 is 30.5 mW.
We tried to take a photo of the beam and the ruler in ZM5, but it was difficult to read the ruler by camera or our eyes so we roughly measured the beam position in the baffle of ZM5. The hole diameter of the baffle in ZM5 is ~3 cm and the beam is ~1 cm below the top of the hole so the beam in ZM5 is higher than the center of the hole of the baffle.
The beam profile measurement in 75770 was done at 6D in this figure. The ZM4 and ZM5 PSAMs strain gauge [V] (PZT request [V]) are as follows.
ZM4: 5.68 (100)
ZM5: 0.52 (100)
