Sheila, Daniel, Terry, Nutsinee

35.5MHz I/Q optimization, 70.78 degree offset

We injected 105Hz excitation to SHG PZT then adjusted the phase offset while watching I/Q on the SR785. 70.78 deg offset gave a factor of 40dB in ratio and that is now the current setting.

SHG Open-loop gain TF, unity gain frequency ~ 3kHz

We adjusted the common mode board gain until the UGF is ~3kHz. When the SHG crystal was not heated (25C) this gain was 4dB.

Later we changed the crystal temperature to 35C to get green light out of the SHG. The gain since then has been 20dB.

Then we reduced the power going in to the SHG cavity. Was 400mW, now 100 mW. This didn't change the UGF by much so I didn't touch the gain setting (still 20dB).

Here's a snapshot of the final open-loop gain TF as of Friday evening. Once lowered the input power I was only able to drive the PZT at 50mV stably.

Two problems prevented acquiring the SHG fast channels:

1. The AA was not connected to the ADC
2. The channel assignment in the model was wrong

Now fixed.

We found a loose cable inside TTFSS which prevented the readback of the LO monitor. We then tested the remote controls for the TTFSS and fixed a problem with the excitation enable logic and the gain slider calibration. The laser crystal temperature was also a factor of 5 too high, since the slow controls signal through the TTFSS implements a voltage divider. We also connected the fast channels. The only thing missing now is the HV.

Most of the DC PDs had the wrong configuration in TwinCAT. Now fixed.

We now have a guardian for the squeezer SHG locking loop which seems to be working fine.  It is located in sqz/h1/guardian In order to run it Dave had to add the sqz guardian directories to the list of places where the guardian machine checks for code.  

It is a very simple guardian, with 3 states, down, lock, and locked.  

It currently uses the transmitted IR power through the SHG to determine if the cavity is locked. Once we have a calibrated green diode (which Terry is working on as I'm writing), we can use that instead. 

There is some basic error checking but much more could be added. 

It locks the cavity by toggling the state of the first boost.  If this doesn't work, it starts a scan of the cavity using the slow output offset, and toggles the boost at each step of the way.  The step size and speed of the scan can be adjusted, but it seems to lock the cavity within about a second with the current settings.  

- Jenne, Keita, Cheryl

• Keita used the PZT scan data to find the real MC2 Trans beam, and center it on the MC2 Trans QPD
• after aligning to flash the IMC, changes in MC1 alignment walked the beam on IOT2L REFL far enough that the table needed to be realigned
• I realigned the REFL path, which had changed enough that after realigning the REFL PD, WFS, and camera paths, I also had to realign the path to the high power beam dumps
• I continued IMC alignment, and used mostly MC3 and MC2, and did need to use MC1 (small changes that maintained REFL beam path)
• Jenne locked the IMC - screenshot attached
• WFS centering has moved away from (0,0), and the beams are now at around (-0.3, 0.6)
• Jenne has the WFS up and running
• centering on MC2 Trans is at (0.2, 0.3) and the SUM is currently over 15, and has been as high as 19, at 5.4W input power to the IMC
• turning the IMC locking off for the weekend, and will resume on Tuesday

TITLE: 01/12 Day Shift: 16:00-00:00 UTC (08:00-16:00 PST), all times posted in UTC
STATE of H1: Planned Engineering
LOG:

15:27 UTC Hugh to HAM2, looking for Mark and Tyler
15:35 UTC Hugh back
15:57 UTC Hugh starting an excitation on end X HEPI
16:11 UTC Christina and Vanessa to end Y
16:43 UTC TJ to optics lab
16:43 UTC Hugh, Mark and Tyler to HAM2 to install scaffolding (WP 7281)
16:47 UTC Christina and Vanessa leaving end Y
16:57 UTC Christina and Vanessa to LVEA to clean
17:19 UTC Filiberto to mezzanine to connect TCS to Beckhoff safety interlock
17:20 UTC Robert to LVEA to look at damaged panel in IOT2L
17:25 UTC Robert back
17:30 UTC Gerardo to LVEA, isolate pump on top of HAM5 and 6
17:43 UTC TJ back
17:56 UTC Hugh to HAM2 to check on scaffolding work
18:01 UTC Chandra to mid Y and end Y
18:04 UTC Fire department through gate, touring site
18:05 UTC Gerardo back
18:10 UTC Mark and Tyler back
18:15 UTC Fire department off site
18:23 UTC Travis and Betsy to end Y
18:25 UTC Cheryl to HAM3, WP 7284
18:46 UTC Mark and Tyler back to HAM2
18:49 UTC Greg starting WP 7285
19:09 UTC Nutsinee and Daniel to squeezer bay
19:34 UTC Gerardo to mid Y to replace VEA camera
19:39 UTC Daniel and Nutsinee back
19:56 UTC Greg and TJ to LVEA to install laser enclosure panels around TCSY table
19:58 UTC Mark and Tyler back
20:07 UTC Gerardo back
20:15 UTC Betsy and Travis back
20:15 UTC TJ back
20:36 UTC Mark to HAM5 to tighten door bolts
20:36 UTC Greg done
20:39 UTC Peter to PSL racks to remove a power supply
20:44 UTC Peter done
20:56 UTC Cheryl to IOT2L to adjust cameras on table
20:56 UTC Filiberto done adding TCS laser to safety system. Going to install power supply that Peter removed in the CER.
21:04 UTC Sheila to LVEA
21:12 UTC Hugh to HAM2 to look at scaffolding
21:18 UTC Hugh back
21:35 UTC Mark done
21:40 UTC Filiberto done in CER. Going back to cabling in LVEA.
21:42 UTC Gerardo to mid Y with camera
21:45 UTC Keita to IOT2L
21:46 UTC Darkhan and Yannick to end Y, PCAL work, transitioning to laser hazard
21:57 UTC Darkhan transitioning end Y to laser hazard
22:02 UTC Gerardo back
22:03 UTC Terry to squeezer bay
22:13 UTC Marc to CER
22:57 UTC Nutsinee to squeezer bay
23:04 UTC Dave restarting h1sqz model
23:05 UTC Darkhan and Yannick back. end Y back to laser safe.
23:08 UTC Nutsinee back
23:29 UTC Sheila and Nutsinee to squeezer bay
23:47 UTC Chandra to mid and end Y

WP7287

Chandra, Dave:

The Y2 beam tube cold-cathode gauge alarms were sped up from 15 minute to 5 minute reporting. The alarm level was kept at 5.0e-09 torr. The CP4 cold-cathode gauge (PT245) was added to the alarms system with an alarm level of 1.0e-07 (it is operating around 5.8e-08 torr). This work permit was closed.

Daniel found some more ADC connection issues, which were fixed this afternoon. h1sqz model was restarted, no daq restart was needed.

This morning Travis and I went into BSC10 (EY), swung back the ACB to gain space around the ETMY QUAD structure, and started the disassembly of the QUAD - removing QTY 4 VA's, cross braces, Sleeve wedges, and started on some of the sleeve fasteners.  We then did a fit-check to see if we could get the install arm elevator into place without having to also swing back the TMS.  We didn't get the sleeve removed but it looks like we'll be able to get the elevator in the spaces around the QUAD structure without having to pull the TMS back.  We'll try next ~Tues-Wed and see how far we get.

(Note, swinging back the TMS involves a lengthy process of adding support brackets and equipment to the TMS lower table section, loosening some of the suspended cable sections, stopping up the upper mass, then pulling it back for a few extra inches of clearance.  Resetting it all after the maneuver is complete will result in a many day period of realignment of the TMS.)

The TCS Y chiller that has a leaky seal (SN# 0110193301120813) has been replaced with the spare (SN# 0127640801150617). New quick connect fittings were used as the old ones had been cross-threaded. I used the old settings from the original chiller on the spare chiller (attached here, couldn't find a document containing them). No leaks or water level dropping could be detected after several hours of running, will recheck tomorrow.

Measured dew point of EY purge air for about 12 minutes this morning, with no demand (closed off metal valve to get flow through pressure relief valve). DP crept down to -25.3degC. Then found a "switching failure alarm" on at the control panel (visual only - no sound). I reset and re-measured DP. Quickly dropped to -22degC in several seconds. The two drying towers appear to be cycling between regenerating and drying.

Compressors are running between three and four units with current flow. Output pressure is 0.5-0.7 psi. Soft cover was on BSC door, but arm is in chamber.

Not really too important for the IFO maybe but as a test/proof of concept, this DOF was addressed as it was easier.

After a few iterations of the HEPI  Y to RZ IPSALIGN factor (H1:HPI-ETMX_IPSALIGN_3_2) the value that reduced the YAW seen on the Oplev YAW sufficiently was determined.

As I showed in 40106, driving the HEPI platform in X or Y produced Yaw seen on the optical lever.  Including the Y to RZ factor above in the control scheme reduced the yaw seen on the optical lever about a factor of 10.  The attached plot shows the same channels as in log 40106.  As X is not being driven, there is no OpLev Length to Angle correction.  This time period shows 2 measurement with different IPS_ALIGN factors: -0.007 & -0.0055.  This was just to convince me that the process was effective.  This really needs to be done for the in-line drive, ETMX X and ETMY Y drive.  These effects are pretty small and the ASC has managed to deal with the YAW induced when a long lock stretch drive HEPI 100s of um.  Still if the ASC doesn't have to deal with this, it should be a good thing.

Once I figure out how to a priori compute this with just a couple measurements, I'll do this for the in-line drive which will be a bit more involved given the OpLev Len2Ang correction.  Notice too, the RX_OUT16 (middle left plot); I'll look at other sensors to see if the horizontal drive is actually producing real tilt.  It may not be real unless our local sensors on HEPI are being fooled just like the YAW loop is as I showed in 40106.

Pressure is stabilizing at CP4 after isolating it and valving in a small ion pump. Working with CDS team to set/adjust pressure alarms.

Wondering was these perturbations are in corner station pump down. Peaks from Wed. morning and step down Thursday morning.

- Jenne, Cheryl, Keita

Jenne and I confirmed that the beam from the PSL did not require any adjustments. The REFL beam was present on IOT2L and the beam was on the REFL camera.   On IOT21L we adjusted the alignment of the REFL path to clear the iris/HWP/CalciteWedge combination of optics and cleanly dump beams before the 2" steering mirror to the REFL diode, and after those adjustments we restored the beam to the REFL camera.  Because we're not using WFS at this time, the beams onto WFSA and WFSB were not adjusted.

Searching for the good beam found yesterday, it was found again today, however, as we walked the PZT and MC1 the power dropped steadily.  I went to HAM3 West door and took a few pictures with the IR camera, and saw that there was a large beam on the beam dump post that is in from of MC2 Trans.  We confirmed this beam as the reflection from MC3, and drove MC3 and MC2 alignments to ensure that there are no IMC mirror reflections coming back to HAM3.

Talked to Keita, plan is to scan the PZT to look for a beam(s) on MC2 Trans QPD, because we calculated the angle needed to move the beams on MC2 Trans (600urad) and the PZT should have the range to put the real beam on the QPD.

Jenne is running two AWGGUIs on ZOTWS1 that has a sticky note on the left monitor saying "In Use" to control the PZT overnight.  The PSL power is at 2.7W (requested 4W).  Alignment of the IMC mirrors should not be changed until tomorrow morning (end of test).

The PZT scanning will be complete well before 6:30AM, so HAM2 HEPI locking and scaffolding work should proceed as planned.

Keita noticed just now that there is a readback error on AS B 36 whitening gain.  We should fix this before we start needing AS WFS.

On Tuesday, Dave and I drove to EX to restart the seismic computer and most of the rack components to see if we could get the .085 hz feature that RichM found in December. Sadly, turning on then off again hasn't fixed the problem. I have been able to narrow down the source, I think. In the asds we took before, the feature seemed biggest in the RX/RY loops, so I started by turning those off. This made the feature go away in the X/Y dofs ( or maybe it just got smaller than the background?), it also made the peak disappear in all sensors other that the T240s. I think this shows that the peak is coming from something in the Z1/2/3 (local) T240 signal chain. To dig in a little further I collected an hours worth data with the ISI in this configuration ( ST2 damping only, ST1 RX/RY/Z isolation loops off, X/Y ISI sensor correction off, HEPI Z sensor correction off) and looked at the coherent subtraction between the 3 local Z T240 signals, as shown in the attached figure. The first plot shows the asds of the 3 local channels, Z1 & Z2 are similar at .085hz, but Z3 has a small bump. The second plot shows the result coherently subtracting each of the channels from one another. The fact that the .085hz goes away for subtractions that don't include Z3 makes me thing that there is something going on with the corner 3 sensor. I don't have time tonight to see if this extends to the horizontal dofs on that corner, but I think I want to try power cycling the T240 chassis again at EX.