Earlier today I took data for redoing the MICH and SRCL feedforward filters, following my instructions from alog 47185.

* Measured MICH FF with both mich and srcl ff off

* Turned on mich ff (feb 28 filter), then measured Srcl ff.

I apparently don't actually understand the nuances of the fitting scripts as well as I thought, so I haven't actually created new filters.  I checked the current coherences of MICH and SRCL with DARM, and they are both quite low (1e-2 level), so they're already fine for now anyway. 

Note that the PRCL coherence is still around 5e-2, so we will want to address this. Hopefully it'll be fixed by the TCS tuning to minimize frequency noise coupling, but if that doesn't work, we may need to consider using the PRCL FF filters.

I am accepting things in SDF, so that we can go to Observe.

Why does IMC-MCL_FM_TRIG_THRESH_[ON/OFF] change?  This filter bank isn't in use after we acquire lock, so it shouldn't matter.  I've reverted to the value that was in both the safe and observe snap files.

We are also leaving on several lines, for monitoring overnight.  Dan's script will kill these if we lose lock.  This does, however, mean that we are *ignoring* excitations in the LSC and ISS models.  We will revert these to being monitored when we are done with this test, end of next week at the absolute latest (i.e. before O3).

To make this work, Dan has made some changes to the TCS guardian, that we likely want to keep anyway. Also, in order to prevent dropping out of Observe, we've not-monitored the CO2 rotation stage go-to-power buttons.

The OPS Overview FOM that displays the Observatory Mode is displaying only a white box when Observing is selected as the mode.  See screenshot.

Then back to Commissioning at 0:44 due to SQZ Guardian.

Sheila exported our currently-in-use inverse sensing filter, and checked it into the svn.  It is at: /ligo/svncommon/CalSVN/aligocalibration/trunk/Runs/O3/H1/Measurements/Foton/2019-03-16_H1CALCS_InverseSensingFunction_Foton_SRCD-2N_Gain_tf.txt

We exported the D2N filter, but not the actual gain filter, which has a gain of the order 1e-7.

In the attached figure, I plot the February filter versus the new one.  I have imported the .txt files to make these plots, so we're pretty sure that we've exported the correct filters. 

FRS12545.

If any of the component HnEPICS_*.ini files used by h1edc to create its full channel list are modified, the CRC bit in h1edc's STATE_WORD will turn red, and its MSGDAQ will report "Modified file detected". In the same way as for real-time front end models, this indicates that a h1edc and DAQ restart is necessary to add/remove the new slow channels to the DAQ.

I wrote a bash script (h1edc_generate_daq_ini_file.bsh) which generates the H1EDC.ini file in exactly the same way h1edc does when it is started. The script runs every minute, as a controls cronjob on h1fescript0. If a modified H1EDC.ini file is generated, the original is archived and the file is updated. The h1edc model then detects the CRC change of this file and raises the CRC condition.

Attached snippet of the CDS overview shows the CRC bit (green square). Note the DAQ bit and DAQ-CRC are still blank until h1edc is added to the DAQ tomorrow.

TITLE: 03/18 Day Shift: 15:00-23:00 UTC (08:00-16:00 PST), all times posted in UTC
STATE of H1: Lock Acquisition
INCOMING OPERATOR: None
SHIFT SUMMARY:
LOG:

15:54 FIl into CER

16:00  Meeting

16:33 LVEA Sweep - Richard, Betsy, Jenne, and DS

17:15 Robert, Kara, and Sharan out to LVEA - coil

17:30 LVEA Sweep done

17:56 Robert, Kara, and Sharan back

18:22 Dave Barker restarted h1edcu (CDS Overview)

18:35 Patrick and Sheila out to LVEA - ISCT1

19:03 Sheila and Patrick back

19:16 Kara, Sharan, and Robert out to LVEA

20:28 Lockloss - maybe due to PEM group activity in the LVEA

21:56 NLN 96.3Mpc

Ed, Sheila, Jenne, Patrick

We locked at 35W this morning following work done last night reported in 47606 and comments.  

• POWER_30W is now going to 35W.
• We measured the PRCL gain at 30W, the ugf was already a bit low, once we powered up to 35W it decreased a bit more so we set the gain to 16 to get a ugf of 45Hz, we could make it higher. 
• We measured the MICH P gain which was not changed from 30W yesterday, so we left the gains of both MICH ASC gains as we had found them yesterday.  
• We found that the POPX centering PZT was railing, which was also happening during last night's 35W lock, Patrick and I relieved it by turning knobs on the steering mirror that points to POPX (near the edge of the table).  It seems likely that there has been some confusion about how to offload this,  the output of the integrator is circled in the attached screenshot, that is what should be driven to zero.  Since the voltage applied to these PZTs goes from 0 to positive voltages, they are at the center of their range when the "MONITOR" is around 16000 counts.  If we adjust the steering so that the "MONITOR" is around 0, we will be near the limits at the output of the integrator, which might be why we were so close with this that a small change in power caused us to rail.  
• Jenne made measurement to re-tune MICH and SRCL FF at 35W.  We turned on the old FF and the coherence is low, so for now we are just continuing to use it. 
• Patrick and I measured the CARM OLG, which was at 15kHz with 6dB IN1 GAIN on the CM servo.  We powered off the Aglient which was connected to the IMC board, but left the SR785 connected to CARM (excitation disabled).  We couldn't find the ethernet cable we used to use for GPIB, so it is not connected to GPIB right now.  
• We thought that the guardian would not increase the gain beyond 6dB in "LASER_NOISE_SUPPRESSION", based on looking at the code, but it set the gain to 9dB, which seemed OK for that lock.  
• We put Georgia's changes for the ISS back into the IMC_LOCK guardian 47607.  We may want to put more thought into how to make the ISS DC coupling work for a variety of laser powers.  

We didn't go to observing, instead PEM injections went ahead, but we tried to do some checks that will make it easier to go into observing soon. 

• We changed the nominal state of the laser power guardian from 30W to 35W.
• Most SDF changes were understood from work over the weekend including the change in input power. 
• MC2 M3 LOCK has a low pass filter engaged, which looking back at the SWSTAT over the last week has been different for an unknown reason between different locks (sometimes FM9 is engaged, sometimes FM10 instead). To be continued...
• The squeezer has a power stabilization servo for the green power, which does not engage if there is too large of an offset.  Daniel engaged this manually by adjusting the offset, but we will need to put some thought into how to run this servo without putting more than 20mW into the squeezer fiber.  To be continued...

This morning, Daniel, Richard, Jenne, and I performed a sweep of the LVEA in prep for logging some ER14 Observation Mode overnight this week.  We did not prep everything for O3 since this and next week are still heavy in PEM injection and Calibration work which need equipment plugged into lots of places.  We also DID NOT do the end stations since we have quite a few Tuesday Maint items at the end stations tomorrow and will perform the list then.  The checklist item status after the morning sweep is as follows:

LVEA Sweep

• OPERATOR TO DO before button pushed - Make sure no one is in the LVEA
• Confirm cranes in their "parking spots" & their lights are OFF - WILL FIX DURING TUES MAINT and when PEM IS FINISHED NEXT WEEK
• ALL OFF EXCEPT PEM SET - Monitors/work stations are turned OFF (except VAC computers)
• DONE - Phones unplugged & batteries pulled from handsets (Phone locations here)
• Confirm no mechanical shorts onto HEPI or ISC Tables - WILL FIX DURING TUES MAINT
• DONE - Cleanrooms OFF
• DONE - PSL in Science Mode
• DONE - ISC Table fans OFF
• Confirm wifi access points are unplugged (instructions) - WILL DO IN 2 WEEKS 
• Electronics racks (i.e. make sure no test equipment connected to a rack, unless work permit for it.) - PEM and SQZ Still have activity next 2 weeks - See other notes below
• Forklifts NOT connected to charger - WILL NEED TO CHECK AGAIN IN 2 WEEKS
• DONE - Unplug unused power supplies/extension cords
• Lights OFF (check lights via operator cameras) - DO NEXT WEEK
• DONE, SDF SET BY DANIEL - From the controlroom double check that PICO motors are powered off
• Turn off the PSL phone via the 48v PS in the GC Computer servers room - DO NEXT WEEK
• Power down the intercom - DO NEXT WEEK
• ALOG the LVEA has been swept

VEA Sweep (for End Stations) - NOT DONE TODAY, STILL OK TO GO TO OBS MODE FOR TESTING

• Make sure no one is in the LVEA
• Crane in "parking spot" & lights are OFF
• Monitors/work stations are turned OFF
• Phones unplugged & batteries pulled from handsets (Phone locations here)
• Confirm no mechanical shorts onto HEPI or ISC Tables
• Cleanrooms OFF
• ISC Table fans OFF
• Confirm wifi access points are unplugged (instructions)
• Electronics racks (i.e. make sure no test equipment connected to a rack, unless work permit for it.)
• Unplug unused power supplies/extension cords
• Power down the intercom via blue box in Airhandler room
• Lights OFF (for endstations check lights via webcams)
• ALOG the VEA has been swept

Items which required attention this morning, or are noted here as being the configuration for future sweeps:

IN COMPUTER ROOM

• IFR Signal Generator left on in Rack ISC-C4 - Used by D Gustafson,
• COIL INJECTION INPUTS in ISC-C4 Rack - Used by D Gustafson, WILL DISCONNECT BEFORE RUN
• WIFI STILL ON

NEAR H1 PSL and RACKS

• Agilent 4395 Analyzer still plugged into IMC for CARM Loop commissioning - WILL PULL NEXT WEEK
• Confirmed PSL DBB powered off at Rack

NEAR HAM2/3 RACKS

• 24V PS in bottom of facilities rack powering Beckhoff - RMCCARTHY TO REMEDY, OR APPROVE FOR O3

NEAR Y-arm Manifold

• Put back panel on VE Rack "LVEA-Y"

NEAR H2 PSL enclosure

• Power supply in bottom of empty rack - looks to go into the enclosure - likely a dust mon OK FOR O3

NEAR HAM6 SQZ area

• IFR signal generator on in HAM6 rack due to RF AM Modulator to replace the Aeroflex unit currently used to drive the AOM. - WILL FIX WITH NEW PARTS in ~3 WEEKs.  https://dcc.ligo.org/LIGO-E1900035
• Lots of messy stuff needs cleanup by crew next week, but all temp equipment unplugged.