h1susauxex froze around midnight last night. I rebooted it using front-panel-reset-button. Since this is not on dolphin, no other restarts were needed.

All spectra appears to be within acceptable, nominal range.

TITLE: 04/18 Day Shift: 15:00-23:00 UTC (08:00-16:00 PST), all times posted in UTC
STATE of H1: Planned Engineering
OUTGOING OPERATOR: None
CURRENT ENVIRONMENT:
    Wind: 5mph Gusts, 3mph 5min avg
    Primary useism: 0.07 μm/s
    Secondary useism: 0.18 μm/s
QUICK SUMMARY:

• Rick and Peter are in the PSL.
• Cleaning crews are at End Stations
• Mark and Tyler at EX doing GV and Chevron baffle work

I am starting the process of moving all guardian nodes from h1guardian0 to the new h1guardian1 machine for testing.  We believe that we may have fixed the problem in guardian that was leading to anomalous segfaults (full explanation on the problem and fix in a follow-up post).

Craig, Niko, Georgia

Today Georgia, Niko, and I went down to Xend to take more careful measurements of the Electric Field Meter (EFM) Prototype built by Rich, Calum, and Luis.  

We drove the ISI to check if we could see the electric fields above our EFM noise floor.  We drove the ISI both vertically and horizontally until the watchdog tripped, but never saw a signal in the EFM.  This is surprising, since Rai's prototype was sensitive to ISI driving in the Y chamber.

We measured a calibration TF for both the positive X EFM sensor plate and negative X EFM sensor plate (Plot 1),
the common mode rejection from the X sensor plates (Plot 2),
and an ambient electric field ASD in the X-end chamber (Plot 3).  

Finally, I roughly estimated the ETMX displacement noise we might expect from ambient electric fields given our usual ESD setup. (Plot 4)  
The details of the calculation made to get plots 3 and 4 are in a jupyter notebook on git.ligo.org.

At 100 Hz I estimate around 10^{-22} m/rtHz displacement noise, far below our noise floor.  However, I'm getting a stronger frequency dependence than we perhaps anticipated: our measured electric field falls like f^{-1.5}, until it runs into what is probably the EFM circuit noise floor at 200 Hz.  This gives an overall displacement frequency dependence of f^{-3.5}.

Pictures are courtesy of Niko and Georgia who did all the in-chamber work today.  

Things to do:
1) Get capacitance measurement between calibration plates for a better volts to electric field TF.
2) Understand and create a noise model of the EFM circuitry.  This will probably help explain the strong frequency dependence we're seeing with some low frequency pole.

Sheila, Nutsinee, Terry, Dan Brown, TVo

To get an idea of the mode-matching from the OPO to the OMC, we wanted to do a beam profile measurement of the squeezer and use Alamode to fit the data points.  The hope is that the fit will match a Finesse model that includes the squeezer path from the output of the OPO, through the two lenses, ZM1/ZM2, and into the Faraday.  Then the beam reflects off of SRM HR surface and propagates to the OMC. 

The squeezer crew was able to take profiling measurements between OM2 and OM3 with the nanoscan using a pick-off mirror. They repeated this for three positions of the lens on the VOPO translation stage which allows for about +/- 2cm in travel.  The translation stage position that moved the lens closest to the VOPO has the best overlap estimate with the OMC at about 83% (using alamode) which means it's possible that we have to move the translation stage on the breadboard to get better results.  It's clear that we should move the lens+translation stage closer to the VOPO, which will move the waist location into the right spot, but this will alter the waist size.  If our model is correct, we can estimate the amount of mode-mismatch if we have to live with this lens positioning.

To double check our distances to the , which is not clear because the OMC is covered up with the shroud, we measured the OMC_REFL beam as well as the beam coming right after ZM1.  One thing that is quite odd, is that the beam was pretty astigmatic ~18% on some of the measurements.  This was not seen in the previous measurement but was seen when we've measured the IFO beam profile.  However, rotating the profiler didn't change the effect so it might be real.

To do:

- OMC PZT scan with Sqz beam this will allow use to double check our mode-matching estimate.

- Hammer on the model a bit more to get real-world measurements, Finesse, and Alamode to match up.

Peter King went back into the PSL so that we could try to lock the IMC for some work in HAM6 this evening.  While Peter was in the enclosure, the IMC was locked and the alignment was decent, however as Peter was leaving it looks like something misaligned in yaw which is not in the same gouy phase as the PZT, so we can't re-aling it without going back into the PSL.  Since I'm not sure which irises Peter and the PMC crew were using, we are not going to try to align it tonight.  If we can lock the IMC tomorrow that will help us to finish of HAM6 work. 

J. Kissel

Following up on all the work from yesterday replacing the V2 OSEM on H1SUS OPO (ID of problem: LHO aLOG 41468, Replacement LHO aLOG 41470), I've had a bit of time between HAM6 SQZ Team Closeout Activity to grab a set of undamped Euler basis TFs.

Attached are the results.

One can see -- especially in pitch -- that swapping the V2 OSEM has moved out the PD-to-Coil cross-coupling zero to the typical ~10 Hz, and fixed the problem.

The TFs are pretty noisy, but not unexpectedly so: while the ISI was unlocked, the chamber was still half-open letting in SQZ light, so I expect purge was still flowing heavily over the OPOS.

I'd like to get
    - A Repeat of the Euler basis TFs,
    - OSEM basis TFs,
    - Damped TFs,
one morning when we have the chamber door parachutes completely closed, and pulled away from the ISI (as was true for the 2018-04-02 data).

J. Kissel, T. Shaffer 

While in HAM6 fixing the OPOS (LHO aLOG 41468), TJ noticed that the ECD magnets for ZM1 were still far away from being engaged and functional. He moved them into place, so I've retaken TFs to ensure there's no *new* rubbin. There is no new rubbing. We're otherwise still as it was left at the end of March LHO aLOG 41181.

Attached are the usual metrics.

The remaining flaws: 
   - L2L doesn't show the primary pitch mode
   - Y2Y shows the primary L mode
   - The Y2Y primary Y mode is lower in frequency than other HTTSs.
   - When damped, the primary Y mode's Q is reduce, but the unwanted L mode still shows up with the original Q in the Y2Y TF.

I'll talk with TJ again, in hopes to recreate he and Betsy's two-OSEM test -- hopefully if I witness it, I'll understand the results better and offer further suggestions.

Gerardo M., Tyler G., Kyle R. 

Today we replaced the "legacy" 10" Cetek gate valve with a new 10" VAT valve.  I am pumping the turbo volume with a leak detector overnight and will leak check the turbo-side joint tomorrow.  

Fun Fact -> I, necessarily, started and will be running overnight the instrument air compressors (located in the chiller yard)

PeterK, JasonO, EdM, RickS

Today, we swapped out the original PMC for one of the new "All-bolted" PMCs (see LIGO-T1700543-v2), S/N007 (photos attached)

We aligned the PMC, FSS and ISS paths, locked the servos and adjusted the light levels and gains.

When we increased the power into the PMC to ~ 20 W, we noticed that the modematching to the PMC was degrading.

We found that the first turning mirror downstream of the new 70-W amplifier is damaged (extremely bright glow where the beam is reflecting from the surface (if you zoom on the photo below you will see the bright spot on the surface - 9:00, slightly left of center).  This is likely the cause of the change in the beam characteristics with power.

We plan to swap this optic out first thing tomorrow.

Aidan, Daniel V-H., TVo, Dan B., Terra

This is the initial analysis of the HAM6 output beam scan we did last week. Attached is the overall picture, in the style of LLO's beam study.

Measurements are done in HAM6 during single bounce configuration (no TCS) either directly in beam or using pickoff. Data is then fit by varying input beam parameters and, to some degree, OM optical/path length parameters. These are compared against the original output beam design. (There is a caveats to this, see below.) Since there was large astigmatism, what I show is beam fit prioritized to fit vertical measurements. [R,w] directly after SRM for design and vertical fits shown in plot. X-axis distances are in meters from SRM.

Main points:

•  we find up to 20% astigmatism coming into HAM6 
  • Finesse modeling for the single bounce beam parameters using the up-to-date design configuration shows about 3% astigmatism coming out of the SRM
  • table-top check of beam profiler gives 2-3% variation between vertical and horizontal, even when tilting scanner various ways with respect to incoming beam (NanoScan NS-SI/9/5 scanning slit profiler)
• looking at just the vertical beam, fit mismatch from OMC is ~5%
• we find similar shifts in OM placements (towards SRM) as LLO compared to original design; the distance between the edge of the ISI and OM1 & OM1 to OM2 was measured in situ. We find the latter agrees with up-to-date Zemax distance.

Caveat/to do: the design beam includes a 50km ITM lens, so I need to subtract that for more meaningful comparison; in the works.

For reference, the parameters used are:

• SRM to OM1: 3.4 m (Zemax gives 3.454 and I tweaked this slightly for fit), OM1 to OM2: 1.39 m (Zemax and in situ measurement)
• OM1 S: 1/4.6, OM2 S: 1/(1.7)