TITLE: 09/17 Eve Shift: 2330-0500 UTC (1630-2200 PST), all times posted in UTC
STATE of H1: Observing at 148Mpc
INCOMING OPERATOR: TJ
SHIFT SUMMARY:  Nice quiet shift with H1 now locked for about 26hrs.

LOG:   2340 Betsy & Jason out of the optics lab

TITLE: 09/17 Eve Shift: 2330-0500 UTC (1630-2200 PST), all times posted in UTC
STATE of H1: Observing at 148Mpc
OUTGOING OPERATOR: Tony
CURRENT ENVIRONMENT:
    SEI_ENV state: CALM
    Wind: 5mph Gusts, 2mph 3min avg
    Primary useism: 0.02 μm/s
    Secondary useism: 0.18 μm/s
QUICK SUMMARY:

H1's been locked since last night!  (20.5hrs now)  H1 was in need of an ISC_LOCK node LOAD, but the H1 SQZ briefly dropped us out of Observing, so Tony was able to take this LOAD off the To Do List.

Seismically, there were a couple of EQs  which required EQ Mode, but H1 rode right through them (last night and during Tony's shift....no "ASC Hi Gn" transition was needed.

From last night, it looks like TJ & Jim took care of BRSy and also the violins were fixed by Ryan C (and thanks about the note on its finicky gain!)  :)

TITLE: 09/17 Day Shift: 1430-2330 UTC (0730-1630 PST), all times posted in UTC
STATE of H1: Observing at 145Mpc
INCOMING OPERATOR: Corey
SHIFT SUMMARY:
H1 Has mostly been locked and Observing all day.
But we dropped from Observing at 15:18 UTC to apply violin damping.
Back to Observing at 15:22 UTC

And again dropping to commissioning when the SQZr lost lock at 23:07 UTC.
Back to Observing at 23:11:12 UTC.

LOG:

Since we have been struggling with the carm offset reduction sequence, here is a quick look at the data from our most recent lock compared with a model that I borrowed from Sheila, see alog 62110.

I plotted the TRX norm data versus the REFL LF data, normalizing REFL LF to the value it has when DRMI is locked only (3.7 mW). Here are the TR CARM offsets (uncalibrated) as well for reference

I found Sheila's code in the alog above, and added these data points to it to plot the transmission versus the refl dc power for different PRGs. This suggests, as we expect, that our PRG is quite high, the data points showing a PRG between 48 and 54. However, the very early points indicate a much lower PRG.

I also added Sheila's plot with the arm transmission versus carm offset in picometers. Based on our transmission, it seems that a "CARM 150 picometers" state, our CARM offset is actually pretty close to 150 pm. Similarly so for CARM 5 pm, assuming that our PRG is close to 54. These points are marked with stars.

Anyway, this doesn't really help us understand what's going wrong with the offset reduction.

Jeff and I noticed that today's lock may have some elevated Bounce and roll modes.
9.7Hz bounce modes.
13.7- 13.9 hz roll modes.
Some fundemantal 500 hz Violin modes for fun.
Another plot from 6-600 hz from ealier in this lock.


I also took a look at a different lock ~ 10 days ago pre power outage.
9 hz bounce mode
13 hz roll mode
Some exquisite looking Violins.

Tagging SUS.
 

We have now been locked for over 16 hours.

IMC REFL DC power is steady at 18.5 mW

IMC WFS A is at 0.95 mW and IMC WFS B is at 0.75 mW

The IMC power in is 62 W and the power at IM4 trans is 56.7 W

MC2 trans is about 9670 [mystery units]

This is reasonable power for IMC refl, but the WFS power is very low. These are the jitter witnesses, and jitter subtraction is not performing as well as it was before the power outage. I can think of several possible reasons for this, but I'm sure that having less than a mW of power isn't helping.

We may want to consider either a) increasing the power on the IMC refl path b) changing the splitter between IMC refl and IMC WFS to be a 50/50 instead of a 90/10, or c) some combination of the first two options that gets us reasonable power on both IMC refl and IMC WFS.

Wed Sep 17 10:08:25 2025 INFO: Fill completed in 8min 22secs

Gerardo confirmed a good fill curbside.

Yesterday it initally looked like the BRSY was run up by activity at EY, but by the end of the day there was still no damping going on and DIAG_MAIN as well as SEI_CONF had notifications about BRSY. I spoke with Jim and he walked me through the same procedure that he and I did back in July (alog86074) when this happened last. The difference between last time and this time is that the checks I put in worked and we caught this much sooner. I also now have the correct way to remote into the brs machine.

After logging in and recapturing frames two times, the BRS is damping nicely.

TITLE: 09/17 Day Shift: 1430-2330 UTC (0730-1630 PST), all times posted in UTC
STATE of H1: Observing at 154Mpc
OUTGOING OPERATOR: TJ
CURRENT ENVIRONMENT:
    SEI_ENV state: CALM
    Wind: 2mph Gusts, 1mph 3min avg
    Primary useism: 0.01 μm/s
    Secondary useism: 0.17 μm/s
QUICK SUMMARY:

H1 is locked and Observering for 12 hours.
The Violins are very upset, seems like EMTY mode 1 is angry.  DCPDS are very diverged.
Main DARM screen is being restarted.

A new Diag_Main:
SEI_STATE: SEI_CONF might be stuck

TITLE: 09/16 Eve Shift: 2330-0500 UTC (1630-2200 PST), all times posted in UTC
STATE of H1: Corrective Maintenance
INCOMING OPERATOR: TJ
SHIFT SUMMARY:

H1 Back To OBSERVING Again & Leaving so overnight (Elenna messaged TJ [owl] about status).

Day 6 post-Power-Outtage from noon PDT last Wednesday.

Hand-off from TJ at beginning of shift:

• Last night H1 was able to have some Observing time. 
• Today, there was work/changes on the PSL table, so now we have a different H1. 
• TJ ran a MANUAL alignment, but he claims our usual automatic Initial Alignment should work.
• Currently waiting for H1 to lock DRMI and then will see what other changes will be needed for H1 to get to Low Noise.
• Will see how the night goes to see if we will have an OWL shift (which is also TJ!)

End Of Shift Notes:

1. Violins were mostly damped down at the beginning of the lock, but ETMy Mode1 started ringing up and I tried several gains in that 1st hr of lock, but toward the last hr of the shift, the gain I left in rung up the mode to almost 6.0 and it was taken to 0.0 by Guardian.  I conctacted Rahul and for the night we will leave this gain at 0.0.  The mode is slowly coming down, but it is now off-screen on DARM.
2. BRSy continues to be "Not In Use"
3. NUC30's DARM DTT timed out and clicking the Start button didn't help.  I restarted nuc30 a couple of times, but the startup script would not restore this computer to its nominal state.  Leaving for the morning. 

LOG:

Locking Notes

• DRMI #1 took about 50min (with CheckMichFringes)
  • Lockloss at DHARD WFS
• DRMI #2 locked within 1min!
  • Lockloss at MOVE SPOTS
• DRMI #3 locked in 2min
  • lockloss Carm 5 Picometers
• DRMI #4 locked in 3min
  • lockloss Carm offset Reduction
• DRMI #5 locked in 4min
  • lockloss Carm 5 Picometers with a quick glitch from something in this state.
• DRMI #6 locked in a few seconds!
  • made it to Observing!

H1 made it to NLN, and there were a couple SDFs:

1. LSC  TR_CARM_GAIN had a set point of 1.0, but was at 2.1---Elenna said this should be ACCEPTed.
2. SQZ OPO_SERVO_COMEXCEN had a setpoint of "Off", but was at "On"---Elenna said to REVERT this.

Violin fundamental was just above 1e-15, but now violins are being damped (a handful have some extra gain to help the expedite).

Elenna messaged TJ (owl shift), to give him a heads up he may need to Guardian Code work.

Tagging OpsInfo and guardian since this log includes some big changes.

Tonight after several tries and failures I have found a workaround for the carm offset reduction sequence that should get us locked. I have adjusted the code and proofread it multiple times, but it is currently untested in the sense that I have not run it, I have only replicated in code the steps that I took to get us locked.

The major problem I faced tonight is that I could not engage DHARD at DHARD WFS; the CARM offset that is set in CARM_150_PM is just too far off the fringe to make the DHARD signal any good. I found that as soon as the CARM OFFSET REDUCTION state ran, the DHARD signal was actually useable for control. The first thing CARM OFFSET REDUCTION does is set H1:LSC-TR_CARM_OFFSET to -7, whereas CARM 150 PM ends with this offset at -3. I found that getting this offset to -7 or -8 sometimes is close enough to make the DHARD signal "real". I tried just setting the value in CARM 150 PM, but I had one lockloss with that strategy, not sure if it was the cause. I finally decided that I think the correct order here (for now) should be CARM 150 > DARM TO RF > PARK ALS VCO > SHUTTER ALS> CARM OFFSET REDUCTION > DHARD WFS. DHARD WFS usually comes right after DARM TO RF so this involves moving the DHARD engagement up a bit. This is a little risky, as anyone who watches the AS camera during carm offset reduction knows, because the arm alignment starts to get really shaky as the arms get closer to resonance. However, I've been doing a version of this for a bit now as a part of debugging this sequence and I think it generally works.

Besides that problem, we had several locklosses tonight around CARM 5 PM, which is where Sheila made several changes to follow our "new recipe" for getting CARM to REFL. A major change we are making here is that we are further reducing the CARM offset and further bumping up the PRCL gain while it is on REFLAIR 27I. PRCL seems to be losing too much gain as we reduce the offset and it causes locklosses when we reach resonance. However, once PRCL is on POP, the gain is fine, so Sheila and I chose to edit the LSC input matrix value. I had to correct some errors to ensure the correct intrix value is changed and that it is changed to the correct value. Then, I realized that in our "recipe" steps, we actually ran the entire usual CARM 5 PM state first, then we an our new steps which hard codes a TR carm offset and such. So, I edited the code to run the usual CARM 5 PM steps, then if we set this "after_power_outage" value to True, it will also run the additional recipe steps.

That is a lot of words, so below I am going to write down the steps of what should happen, if all goes correctly:

• guardian is run as normal through to DARM_TO_RF state
• instead of going to DHARD WFS, the guardian should go next to PARK_ALS_VCO (and then shutter als)
• after the CARM_OFFSET_REDUCTION state, the guardian should go to DHARD WFS (this is out of order numerically and ladder-wise, but I confirmed the new graph shows this)
• after DHARD WFS, the guardian should run CARM 5 PM
• during CARM 5 PM, the following should happen:
  • the "normal" CARM 5 PM code should run which goes from about line 2648 to 2672. at the end of these steps the TR CARM offset should be a number between -48 and -49, depending on what the code calculates
  • then, if self.after_power_outage is TRUE the code should:
    • set the TR CARM offset to -52 and wait 5 seconds for the ramp
    • increase the TR CARM gain to 2.1 and wait 5 seconds for the ramp
    • set the TR CARM offset to -56, and set the LSC AIR3F matrix value for PRCL (REFLAIRB27I) to be 1.6 times the value it already is, so 1.6*-1.22 and then wait 5 seconds for ramp
    • load the AIR3F matrix which should ramp the intrix value set above, and increase the DHARD P and Y gains to their "resonance" values, -30 and -40 respectively
    • then it should return True
  • if self.after_power_outage is FALSE, the state should return true instead of running the extra steps
• Then, CARM_TO_REFL should run and the rest of the steps can proceed as normal

Here are some other details:

Since DHARD WFS was behaving bizarrely, I did leave the green shutters open and check the green alignment once the ASC converged, using the normal technique of running the green QPD offsets while the IFO ASC converges so the green alignment follows the IFO alignment. Once that completed, I checked and all the offsets appeard to be the same, the largest difference being less than 0.1. Therefore, I concluded that it is unlikely that we need to reset the green alignment, and that whatever is causing the DHARD issue is not because we are in a bad alignment.

When I moved DHARD WFS around in the ladder, I realized that this messes up the IDLE_ALS option. I'm hoping that this DHARD change is temporary, so I just commented out the ladder options that include IDLE_ALS.

We had a mystery lockloss during MOVE SPOTS that I don't understand.

Based on our calculations of the PSL>IMC throughput, I determined that the PSL power should be set to 62 W requested, which should give us 56 W on IM4 trans (we get about 91% now verses 93% before the power outage). I confirmed that is the case, IM4 trans is 56.7 W now, before we went to laser noise suppression (where ISS second loop is engaged). I edited the NLN power value in LSCparams and I reloaded both ISC_LOCK and LASER_PWR guardians.

TITLE: 09/16 Day Shift: 1430-2330 UTC (0730-1630 PST), all times posted in UTC
STATE of H1: Corrective Maintenance
INCOMING OPERATOR: Corey
SHIFT SUMMARY: Lots of work today to understand our troubled IFO as well as some maintenance items. We perhaps have narrowed down the change from the power outage and have changed the input power into the PMC to bring us back to a similar place. We have run an initial alignment, which probably could have run automated, and we are now going through PRMI/DRMI for a second time. I accepted some SDFs in safe before we started locking, see screenshots attached.

BRSY was rung up during some work at EY today, and doesn't seem to be damping. I've contacted Jim, but we will keep an eye on it.

LOG:

• 2154 Started initial alignment, no issues
• 2223 Started main locking

I created an updated blend for the SR3 pitch estimator. These are in the SUS SVN next to the first version. (see LHO log 84452)

The design script is blend_SR3_pitchv2.m

the Foton update script is make_SR3_Pitch_blend_v2.m

these are both in {SUS_SVN}/HLTS/Common/FilterDesign/Estimator/  - revision 12608

The update script will install the new blends into FM2 of SR3_M1_EST_P_FUSION_{MEAS/MODL}_BP with the name pit_v2. pit_v1 should still be in FM1. Turn on FM2, turn FM1 off. 

Jeff and Oli tried the first one, and see that the first 2 modes (about 0.65 and 0.75 Hz) are seeing more motion with the estimator damping than the normal damping. To correct this, _v2 add OSEM signal to the estimator for those modes. See plots below - First 2 are the _v2 blend and a zoom of the _v2 blend. figure 3 shows the measured pitch plant vs. OSEM path - the modes line up pretty well. There is a bit of shift because the peaks are close together. I expect hope this will not matter. Figure 4 shows the plant vs. the model path. Now all 4 modes are driven by the measured OSEM signal instead of the model. 

It is interesting to see that the model was not doing a good job of predicting the motion at the first 2 peaks. This is (I guess) because either (a) the model and the plant are different - or - (b) there are unmodeled drives pushing the plant (the suspension) that the model doesn't know about. 

I'm guessing the answer is (b - unmodeled drives) and is likely from DAC noise. I think this because
1 - The plant fit is smooth and really good.
2 - In the yaw analysis that Edgard and Ivey are doing (not yet posted) the first mode of the yaw plant can be seen with the OSEM, but the ISI motion is much too small to excite that level of motion. But the OSEM can see motion, so something is exciting that motion.
3 - The DAC noise is the only thing I can think of.

Quick chat with Jeff indicates that the DAC noise models at those frequencies are not well trusted. We'll try something anyway and see if it is close. I don't see how to use the estimator to deal with that noise - we'd need to have an accurate realtime measurement.