TITLE: 06/23 Day Shift: 15:00-23:00 UTC (08:00-16:00 PST), all times posted in UTC
STATE of H1: Commissioning
SHIFT SUMMARY:

• Locked the whole shift, 09:23 locked so far
• We were able to ride out several EQs today, two were big enough to put seismic into EQ mode. Both being from the North MidAtlantic (5.2 & 5.6)
• Some more comissioning work; PEM, injections

Locked at NLN at 13:37, in Observing at 13:58.

All of the Picket-fence seismometers except TOOM (South OR) and PNT (South BC) lost connected around 18:30.

We went out of Observing at 18:53:34 by mistake but we went back into observing 30 seconds later 18:54:04.

We went out of Observing at 19:08 for some planned PEM work at EndX, back into Observing at 20:03. We went out of Observing at 21:37 for some injections.

While trying to stop a transient injections exitation, the CW injections were accidentally stopped at 22:26. Still ongoing trying to get these back up and running.

In and out of Observing as we try to figure it out

LOG:                                            

Ran stochastic short and long hardware injections, in coincidence with LLO.  These are of the style described in alog 69723.

hwinj stochastic short --gps 1371592362 --run

hwinj stochastic long --gps 1371593299 --run    (Interrupted by earthquake.  LLO lost lock, so I ctrl-c'd)

I used a little under an hour of commissioning time, from about 12:05 to about 1 PM PST, in part when LLO was down, to improve on the shaker coupling measurements at EX that I made yesterday.

Attached is a strain noise spectrum comparing the sub-100 Hz performance of H1 and L1 (inspired by a similar plot that Elenna made). The data was taken in the recent 5-hr long coincident lock, so H1 has its improved low-frequency noise that came from reducing the input power to 60 W. The L1 noise is lower by a factor of 2-3 in the 15-30 Hz band, and the H1 noise is lower below about 13 Hz.

Also noteworthy are broad noise peaks (as might be produced by scattered light) in H1 at 26 Hz, 39 Hz, and 52 Hz -- i.e., these peaks are separated by 13 Hz, though there's not a noticeable peak at 13 Hz.

Enabled H1:SQZ-LASER_RFAM_LF, adding ct2V filter and setting gain to 1. SDF updated.

We've been locked for 06:04, we went out of Observing at 19:08 for some planned PEM work at EndX which is still ongoing. We are currently feeling an EQ from the North MidAtlantic (5.2), seismic went into EQ mode at 19:42.

I wrote some guardian code to decrease the lockloss time. 

For improving DRMI wait time:
The code starts a 60 second timer. When the timer ends (if 5 minutes hasn't elapsed) then it checks the last 60 seconds of POPAIR_B_RF90 data to see if the max value of the flashes is below the threshold of 200. If so, then it will send guardian to ACQUIRE_PRMI. If the max value of the POPAIR data is equal or above the threshold then it will reset the 60 second timer and check again on the next passing of run(self). This should decrease the wait time at ACQUIRE_DRMI_1F from 5 minutes to 1 minute if DRMI_LOCK won't be achieved.

For improving PRMI wait time:
Does the same as the DRMI code above except the POPAIR max threshold is set at 20. And if the threshold isn't reached, then it sends guardian to CHECK_MICH_FRINGES. This should decrease the wait time at ACQUIRE_PRMI from 10 minutes to 1 minute if PRMI_LOCK won't be achieved.

These thresholds were chosen after looking at various times during lockloss.  Finding when the guardian state would change from ACQUIRE_DRMI_1F (101) -> DRMI_LOCKED_PREP_ASC (102) or when ACQUIRE_DRMI (101) -> ACQUIRE_PRMI (50) and comparing the max values for POPAIR_B_RF90_I_ERR_DQ. Similarly for ACQUIRE_PRMI (50) -> PRMI_LOCKED (51) or ACQUIRE_PRMI (50) -> CHECK_MICH_FRINGES (48). These threshold values can be easily updated as needed.

My code's location is /ligo/home/bricemichael.williams/DRMI__PRMI_POPAIR_check_speed_up.py and just needs to be copied into the ISC_LOCK code file. I made notes in my file where the parts of the code go and discussed with Camilla and TJ about testing it on Tuesday (6/27) during start-up after commissioning.

Enabled H1:SQZ-LASER_RFAM_LF, adding ct2V filter and setting gain to 1. SDF updated.

I compared the sensitivity with 75W vs 60W. The sensitivity below 50Hz significantly improved with 60W. On the other hand, the sensitivity at high frequency does not really change. As shown in the SQZ measurement in alog70686, the shot noise without squeezing improves by ~1dB with 75W as expected (sqrt(75/60)~1dB), but the squeezing level is also lost by ~1dB with 75W and therefore the shot noise almost does not change in the end. 

Fri Jun 23 10:09:21 2023 INFO: Fill completed in 9min 20secs

Jordan confirmed a good fill curbside.

Here's a new BruCo scan for a good range time after the LSC FF improvements from yesterday:

https://ldas-jobs.ligo-wa.caltech.edu/~gabriele.vajente/bruco_GDS_1371513618

In summary:

• no MICH coherence anymore
• a little SRCL and PRCL coherence: it might be worth trying an iterative improvement to the SRCL FF
• next in line: ASC-DHARD_Y and ASC-CHARD_P

TITLE: 06/23 Owl Shift: 07:00-15:00 UTC (00:00-08:00 PST), all times posted in UTC
STATE of H1: Observing at 140Mpc
SHIFT SUMMARY:

LOCK#1:

• Green arms looked bad (not surprising after the long lock), so I decided to tweak manually.  However, instead of going to the GREEN_ARMS_MANUAL state, I went to the MANUAL _INITIAL_ALIGNMENT.  At any rate, I tweaked their alignment up, ran a down and then proceeded with locking, but directly to Acquire PRMI.  PRMI flashes looked dead and the video showed an obvious pitch misalignment.  Started by taking BS negative for Pit, but this didn't help, so I went positive, and this quickly locked PRMI.  Continued going taking BS positive and then optimized POP18/90 by tweaking PRM in pitch.  Then, I let ISC LOCK take over by quickly locking DRMI and continuing on.  Lockloss at ENGAGE DRMI ASC.  And on ASC signals, PRC2_P_INMON & PRC2_Y_INMON became noisy, as well as MICH_P_OUT16 & PRC2_Y_OUT16.  Verbal also mentioned BS saturations.

LOCK#2:

• Once again had a lockloss at ENGAGE DRMI ASC.  And on ASC signals, PRC2_P_INMON & PRC2_Y_INMON became noisy, as well as MICH_P_OUT16 & PRC2_Y_OUT16.  Verbal also mentioned BS saturations.  Attempted a few more locks all to end the same way (no surveys submitted for these repeat locklosses)
• Going for an Initial Alignment

LOCKs Post Alignment & with Laser Noise Suppression issue

• DRMI post-alignment looked noticeably better (POP18 was double what I had for the lockloss locks)---so H1 was just misaligned after the long lock.  Locking was fine until I encountered the LASER NOISE SUPPRESSION issue.
• Earlier there was DIAG MAIN message stating PSL Low Diffracted Power.
• IMC LOCK message:  "Second loop AC Coupling out of range, waiting 1 second to try again"
• LASER POWER message:  "Power not the same as requested"
• This was fixed by addressing diffracted power & iss 2nd loop ac coupling offset....but range does look a little lower.

LOG:

• 0751 LOCKLOSS (Lock Clock:  26hrs51min)
• 0852-0913 Initial Alignment
  • NOTE:  PRC Align had repeated IFO_OUT saturations.  Ryan S taught me this is due to saturations for the OMs.  I saw that OM1 & OM2 had Yaw saturations, so I cleared histories for this path and this cleared the saturations.
• 1358 Back to OBSERVING!  (and handed off to Ryan C)

TITLE: 06/23 Day Shift: 15:00-23:00 UTC (08:00-16:00 PST), all times posted in UTC
STATE of H1: Observing at 134Mpc
CURRENT ENVIRONMENT:
    SEI_ENV state: CALM
    Wind: 6mph Gusts, 4mph 5min avg
    Primary useism: 0.01 μm/s
    Secondary useism: 0.06 μm/s
QUICK SUMMARY:

• Locked at NLN for 1:23, in Observing since 13:58 UTC
• Range is a little lower than yesterday
• Dust/CDS/VAC/Violins looks good

As posted earlier, there were issues getting through the LASER NOISE SUPPRESSION step.  Luckily, Evan saw my posts on ligo.chat and gave me suggestions for stuff to check from status descriptions I posted.  In addition to the note earlier, we had various messages:

• DIAG MAIN:  "PSL ISS Low Diffracted Power"
• IMC LOCK:  "Second loop AC Coupling out of range, waiting 1 sec to try again"
• LASER POWER:  "Power not the same as requested"

From this Evan pointed me to ISS 2nd Loop Offset alogs:

1. Alog 66309
2. Alog 58264 Used procedure Jenne wrote up in here to increase diffracted power to 2.5% (via REFSIGNAL) & then adjusted 2nd Loop AC Coupling Offset such that toggling the 2nd Loop would not cause a transient (see scopes for the changes & then zoomed-in shot of transient test by increasing PSL power to 10W---similar to scope Jenne posted).

• H1:PSL-ISS_REFSIGNAL:  -2.011 [was -2.010]
• H1:PSL-ISS_SECONDLOOP_AC_COUPLING_OFFSET:   0.000 [was 0.050].  SDF was ACCEPTED for this change (see attached)

Luckily, with the above changes, for the next lock attempt, LASER NOISE SUPPRESSION was not an issue.

Once at NOMINAL LOW NOISE, it took the Camera Servo about 22min before we could go to OBSERVING.

And after looking at the range forthe last hour, it appears to be lower than what we had for the last lock, so perhaps my changes were not optimal and could be improved.

H1's Longest Lock Duration:  26hrs 51min

There was no obvious environmental reason for a lockloss at 0751utc.  

Green arms were a bit off (as well as the BS for PRMI).  First locking attempted failed while ASC was engaging for DRMI.

R. Short, J. Driggers

Daniel had noticed that the ISS secondloop gain had been set back to -5 dB when the IFO relocked to 60W last night. Since we now want this at -2 dB in our 60W configuration (alog 70684), I've updated the ISS_acquisition_gain in lscparams.py to be -2 and committed to svn. While we were out of Observing this afternoon starting at 22:45 UTC, I changed the gain to -2 dB, accepted the change in SDF, and reloaded the IMC_LOCK guardian.

Plot 1 shows the dark noise in LSC-REFL_A_LF and REFL_B_LF (yellow/black), with 10W laser input with ISS second loop (red/blue) and w/o ISS second loop (magenta/cyan). The photocurrent in REFL_A and REFL_B was about 16mW each for the 10W measurements. It will turn to about 10mW in full lock. So, we are now about a factor 4 above dark noise.

Plot 2 shows the RIN of the 2 REFL PDs and their average, together with the ISS second loop sensors. The ISS inner and outer sensors have about 7-8mW of light each, so the measurements are limited by the shot noise of the inner PD. In full lock these sensors see about 60mW.