Travis and I spent the morning in BSC10 re-lacing the ETMY PenRe, UIM, and ESD actuator cables through the reaction chain and re-securing them in the usual manner so they float with the masses and don't interfere with anything. 

We also reassembled and reattached the PUM and UIM flags in their nominal configuration.  Since we decided to do this after reinstalling the unit in the chamber, we had to remove the PenRe cans again and insert the flags for that stage through the holes in the PenRe mass.  Usually we attach the flags before installing it, however the unit then knocks around a bit while being maneuvered on the duct jack and the arm and we worried they could come free (although never has happened in the past).  Not sure which assembly order is better since today's activity took a lot longer than installing them outside of the chamber.

In the afternoon, we reattached the UIM-Top Mass wire segments and worked our way through the suspension to suspend all 8 masses.  The new ETM16 monolithic hangs!  With the reaction and main chains fully suspended, we observed a pretty good pitch misalignment of the reaction chain and a bit on the main.  We used the (now) easy PenRe Pitch mass slider bar to correct the gross reaction chain mis-pitch and the UIM coarse pitch adjusters on the main chain to do the correction there.  We dropped the ACB back into place (removed the wedge) and observed that the oplev beam was headed back to the oplev receiver port to within ~a cm or so of the diode.  Will continue from here tomorrow with reseating OSEMs and fine tuning alignment.

*Note to self, there will NOT be an oplev reflection from the reaction chain anymore - someone put a hole in the reaction mass!

Hard closed GV 5,7 around 8pm after checking in with commissioners.

Valved back in all three turbos. Left RGA and all six IPs valved in.

Pressure gauges PT 114,124,134 tripped after closing valves and disengaging switch at the LOTO pin. Ongoing Beckhoff issue.

GV 5 hard closed at 20 psi. GV 7 was stubborn and hard closed at 50 psi. I valved back in the gate annulus volumes at both GVs.

[Sheila Jenne TVo Georgia]

While trying to lock ALS we found that x-end aux laser was glitchy. An example of the sort of behavior we saw is shown in the first screenshot and indicated by the pink arrow, where noise was seen in the green and IR intensity. We thought this was due to the aux laser mode-hopping, so Sheila and I went to the end station to adjust the diode current and crystal temperature. Initially the diode current was 1.949 A and the crystal temperature was 23.73 C, we adjusted this to 1.902 A and 23.78 C.

After the initial adjustment to the laser current and temperature the green and IR power out of the laser was drifting with the crystal temperature (shown in second screenshot), also possibly due to mode-hopping. We returned to the end station to adjust the laser parameters again. The aux laser diode current is now 1.997 A and the crystal temperature ~23.67 C.

Later we determined the glitching was possibly due to the aux laser noise eater ringing, and is fixed by resetting the noise eater. In the future it would be good to have a way to monitor if the noise eater is ringing and toggle it on and off. This used to be achieved using the demod RF mon, but this was not working today.

[For the commissioning team]

We did a quick measurement of the loss in the x-arm after improving the pointing into the cavity using IM4 and PR2.

It is important to note that there were no angular loops closed when doing this measurement, so the power build up wasn't optimized ( LSC-TR_X_QPD_B_SUM ~ 0.95).   We were running short on time and verified that the DC angular alignment was good enough so that the locked vs unlocked state gave us a big enough dip to estimate a visibility.  

Comparing the above numbers to alog-38493. They are close, but I can't say anything definitive about the change in loss between the two ITMXs (before and after swap), we can redo this measurement when we have more time to close angular loops (and maybe try a more sophisticated loss measurement).  Also important to note, mode-matching contributes to the loss as well and this is not taken into account with this measurement.

[Georgia, TVo, Sheila, Jenne]

Summary: We were able to lock and align the Xarm and input beam for both green and IR.

This morning, we locked Xarm with the green laser, and aligned TMSX, ETMX and ITMX to the green beam.  As usual, PR3 was adjusted to get the transmitted green beam out onto ISCT1.  (The ALS light pipe was also opened.)  The ALS WFS didn't end up working for us today - the yaw signals seemed to make sense, but the pitch signals didn't really.  We decided to move on rather than investigate this in detail.  The TMS and ITM pointing were both set using the bafflePD alignment script, then ETMX was adjusted to maximize the transmission through the cavity.

After lunch, we succeeded in moving IM4 and PR2 to get the IR beam aligned to the Xarm. 

We ended up using LSC-POP_A_RF45_I to lock the Xarm, rather than the planned ASC-AS_A_RF45_I_SUM.  POP has worse SNR, but seemed to work a little bit better.  We did not go back and re-try to use AS_A after the cavity was well-aligned, but we need to check and confirm that the signal path is doing what we think it is, sometime using MICH (it all looked fine, and was going to Xarm_IN1, but we just weren't catching any good locks).  We used an input matrix element of -10 for POP, so that the XARM_IN1 was the same roughly +-200 counts that it was during O2 initial alignments of Xarm-only flashing with the old ASAIR_RF45.  Otherwise all the Xarm locking settings were as in guardian. 

We had some struggles throughout the day, see in particular Sheila and Georgia's alogs.  TVo is summarizing our Xarm visibility measurement in yet another alog.

Attached is our alignment after the arm peek.  IMC WFS were on, but no other ASC is engaged.  I also attach the IMC overview, for the IM4 trans and POP_B values, so we don't have to trend them later.

Since the HPO has been off, the FSS has been oscillating at 238Hz according to the PZT fastmon.  THe attached screenshot shows some examples.  

This caused us some problems during the arm peak, and it would be difficult/futile to do DRMI with this happening.  It would be great if someone could look at this loop in the morning. 

I unplugged a read back signal cable (that we never have utilized) which indicates the OPEN/CLOSED status of CP4's 10" gate valve and, shortly thereafter, noticed that PT246B had tripped off -> I re-enabled it in CDS but it isn't likely to "fire" at these Y2 BT pressures -> I assume that this is another example of the inter-dependencies with the Beckhoff "daisy chain" setup?  I haven't yet adjusted to the new "normal". 

Revisited leak checking the new hardware bolted to CP4's 10" gate valve -> No leaks detected with 6 x 10-10 torr*L/sec background -> Moved clean room etc. and began wrapping components for baking -> Expect to be baking by e.o.b. tomorrow.

TITLE: 02/06 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 light was flashed down the X-arm.  Locked with green and currently working on locking in red.  Not sure if they'll stay here the full time to 8pm, but Jenne said, "will be definitely be here past 4pm."  :)
LOG:

• Seismic DMT viewers were down the first few hours in the morning, and came back around 11am.
• Had issue with video4 not being able to run StripTool and medm (Dave tracked to an OS issue....these Macs will soon be swapped out to have the same set up for all FOM computers, so this should be a moot issue soon).
• 15:50 EX electronics work
• 15:56-17:20 Cleaning LVEA (christina and later joined by vanessa, karen)
• 15:57-16:14 Opening GV7 (Chandra)
• 16:11 MY work (Kyle)
• 16:11 Turning OFF ITMx oplev, turned back on later (Sheila, Thomas)
• 16:29-18:03 LN Truck at CP2
• 16:50-17:35 Clean cable work around HAM6 (Hugh)
• 16:55-17:15 Parts at MY (Marc)
• 17:12-18:06 Diode Room checks (Peter)
• 17:22-17:54 EX cleaning
• 17:23 Opening PSL light pipe (Sheila)
• 17:40 ETMy tuning/alignment (Travis, Betsy)
  • 22:47 cont. work 
• 17:44-17:53 Moving Contamination Control parts (Bartlett)
• 17:54 LN Truck at CP2 again
• 17:58 Squeezer work (terry)
  • 19:51 nutsinee joins
  • 22:33 Continuing work (terry) & nutsinee joining at 23:01.
• 18:17 Cintas(mats) vendor
• 18:39 Banging noise at EX begins, tripping ETMx table & disrupting BRS.  This was manlift dirving & drilling by Ken.  He was later notified to stop noisy work.
• 18:45 Commissioners to EX
• 18:48 Contamination Control/Cleaning Room work throughout day (Bartlett)
• 18:59 portapotty contractor
• 19:00-19:21 PEM photos
• 20:10 MY walk (Peter)
• 22:30 Stickers on elec racks (Fil)
• 22:40-0:00 Craning large parts out of the way at MY (Chandra, Mark)
• 22:48 Cable work in optics lab (Hugh)
• 23:45 EY Temperature alarm

• LVEA's GV7  for the X-arm opened just after 8am.  
• Green beam from EX seen at LVEA's ITMx shortly afterward.  Need to align all optics upstream to allow IR light to shine back to EX.
• PSL Shutter opened for IR light.
• ETMx Seismic Table tripped.  This was later determined to be due to manlift moving outside of the building and drilling (Ken electronics work); this work was stopped.
• The Seismic DMT viewer was down for first few hours (Dave & Jonathan worked on getting it back).
• Alignment work continues during the lunch hour....

Some recent modest ASC full data trend plots have shown gaps in the data with an accompanying popup error:

[Error] Purging failed. Increase 'Max drawing path length' in prefs.

It is not immediately clear what "purging" and "max drawing path length" mean, but if you increase it from the default of 20k to 100k the gaps disappear.

Attached is the 'gappy' plot with the popup, and the preferences with the increased max drawing path length

Opened GV 5,7 for commissioners x arm work. Opened up GV 5 to utilize cryopump #1. Pressure at beam splitter is currently 1.14e-7 Torr. Pressures just past CP 1,2 are 6.64e-8 Torr and 7.75e-8 Torr, respectively.

Recorded another RGA scan just before opening valves, after the filament was turned on overnight (thanks Jeff K.!). Will scan again later today.

(This RGA computer in control room is not functioning properly. An upgrade is in its future.)

J. Kissel, D. Barker

A while back, I'd put together Simulink infrastructure for the OPOS (optical parametric oscillator suspension; then called VOPO) while we were beginning to rearrange the SUS HAM56 ADC / DAC allocation to accommodate the new squeezer suspensions (see LHO aLOG 38827). Since then, we've decided several things we'd like to change:
    - We don't like a suspension type and an instance of that suspension type to be the same. So, VOPO and VOPO gotta go.
    - We don't call the output mode cleaner the "vacuum output mode cleaner," so, we've dropped the V in VOPO for the instantiation, and now the suspension type is OPOS.
    - We don't like the channel ordering H1 V1, H2 V1, H3 V3, in the digital system, because that's not how team seismic orders there sensors named the same way, so we convert to H1 H2 H3, V2 V3 V3 (see E1700390 for sensor arrangement)
    - The important axis of the OPOS is *not* aligned with the HAM6 ISI's Cartesian coordinate system, so we convert to every other suspension's coordinate system, i.e. Euler, i.e. L T V R P Y (see G1701821).
Also, I never actually got the MEDM screen infrastucture up to speed to match *any* model, let alone the above new requests.

Today, I updated the front-end code to meet all of the above requests, compiled it, Dave installed it and restarted the h1susopo process. After that, I've moved things around in the sus/common/medm/ userapps repo folder to changing things from VOPO to OPOS, updated a bunch of screens and the associated marco file, and committed it all to the svn. 

I've also *begun* to fill in this infrastructure where I can, but it's not fully complete. OSEM2EUL and EUL2OSEM basis transformation matrices we copied by hand from TST aLOG 11396, 'cause the scripts that calculate the values need to be updated for the now-preferred channel order. I've still gotta 
    - copy over damping filters,  
    - re-check-in with TJ regarding magnet polarity for the COILOUTF gains, 
    - fill in the watchdog signal processing filters, and of course, 
    - we need OSEMs to include any open light current stuff in the OSEMINFs). 
Certainly good enough until we actually hook up our OPOS to the readout and control system.
We also need to change the naming convention in the h1susauxh56 model (VOPO to OPO) so we can pick up the coil driver monitor channels that are always useful when debugging whether suspension is being driven.
Finally, I've made sure the SDF system's safe.snap is up to date with all new settings, and committed that to the repo as well.

Attached are a few screenshots and a text file with more detailed notes about what needed changing -- especially the details of the top-level front-end model changes. 

Valved in vertex RGA and all six main ion pumps and valved out all three main turbo pumps (all pumps reading ~1 mA). Tomorrow will open GV 5,7 for x-arm commissioning (8am-8pm local).

Sheila, Nutsinee, Terry,

This is an update on the progress so far on the VIP. We have aligned and mode matched both the CLF/seed 1064 path and the 532 pump path to the OPO, and are ready to think about swapping M1 for the higher green reflectivity version.  

• We have collimators which were labeled SN18 for 1064 and SN4 for 532.  Maggie measured the modes out of these collimators and posted her data here.  Terry and I also measured both wavelengths in both collimators, here, but at that time the 1064 and 532 collimators were somehow swapped (so the collimator called "supposedly for 1064" in that log is now being used for 532 nm and vice versa.).  Nutsinee also made a better measurement of the mode out of the collimator now being used for 532 in the far field here.
• The first attached photo shows the path from the 1064 collimator to M2 where it is injected into the OPO.  The distance from the collimator to the steering mirror is 173mm, the total distance from the collimator to the AR side of M2 is 270 (+/-~1) mm, from the collimator to the edge of the OPO mounting block is 226mm, and the distance from the collimator to the ROC+150mm lens (f=334mm according to this) is 41mm.  After the lens is a thin film polarizer type 2 which transmits P polarized light, followed by a half wave plate to rotate the polarization back to vertical as required for the OPO.  
• The second attachment shows the 532 pump path.  The distance from the collimator to the first steering mirror is 180mm, the total distance from the collimator to the edge of the metal block the cavity sits on is 190mm, the total distance to the AR surface of M1 is ~234mm.  The first lens is ROC+75mm (f=162.8mm) 152mm from the collimator, while the second lens is 199 mm from the collimator (19 mm from the steering mirror).  This path has a thin film polarizer and half wave plate similar to the 1064 path.  
• We scanned the cavity using the M2 PZT and a thorlabs PZT driver.  Using the free spectral range of the OPO 853.452MHz from T1700104, I did a quadratic fit of the measured 00 peaks of 1064 transmission through M1 to rescale the voltage applied to the PZT driver into resonant frequency of the OPO.  The higher order mode peak is the mode mismatch. From these four measurements the mode matching is 81.1+/- 0.2% standard error.  We did not try to optimized the mode matching of this path, but we can come back and attempt to adjust it if there is time before installation.  
• The dips in reflected power are 95.7+/-0.1% of the power off resonance, without correcting for the higher order modes.  While this is consistent with our expectation for a 99.8% reflectivity M2 and 87.2% reflectivity M1, this give us a measurement of the cavity losses which is dominated by the transmission through M1, and doesn't give us a good measurement of the OPO escape efficiency. 
• Although the documentation calls for two dichorics in the squeezing path, we had only one superpolished dichroic so we used it to separate the green and red and use 1064 HR mirrors for the rest of the squeeze path. 
• Nutsinee has and will post data from the 532nm cavity scans.