This morning I tweaked the beam alignment into the PMC and the operating currents and temps of the 70W amplifier pump diodes (PSL power is down by ~1W, FE output power remains unchanged).  All work was done with the ISS OFF.

When I started, the PMC was transmitting 51.6W and reflecting 11.7W.  When done tweaking the PMC was transmitting 51.8W and reflecting 11.45W.  Not much of a gain, which meshes with my findings from October that the increase in PMC reflected power is not due to alignment drift of the PMC but alignment drift of the 70W amplifier itself.  The alignment into the PMC continues to hold fairly well.

With that complete, I began tweaking the operating current and temps of the 70W amp pump diodes.  The 70W amp was outputting ~67.1W when I started.  The changes are summarized in the below table; I attached a screenshot of the current 70W amp pump diode settings for future reference:

At these settings the 70W amp is outputting ~67.4W (as read by H1:PSL-PWR_HPL_DC_LP_OUTPUT, the power monitor PD that is picked off from 70W amp DBB path), the PMC is transmitting 52.1W and reflecting 11.4W.  While I only increased  the overall output from the 70W by 0.3W, the these setting provide a better transmission/reflection through the PMC than simply brute forcing the 70W amp output power up by 1W; doing that increased the PMC reflected power but did not add to the transmitted power, indicating most of the power increase was in higher order modes and therefore no use to the IFO.

When I turned the ISS back ON, the diffracted power percentage was ~2.8%.  This is slightly high, so I adjusted the ISS RefSignal to -1.99V, bringing the diffracted power % to ~2.3%.  At this setting the PMC is outputting 52.5W.  This changes was accepted in both safe and OBSERVE SDF files.

I reset both PSL power watchdogs at 18:37 UTC (10:37 PST), thereby completing FAMIS 10744.

I centered the IMC WFS.  First I offloaded the MCWFS using the IMC guardian, then moved the IMC to offline (which misliangs MC2) and used the prompt relfection off MC1 to center the WFS.  I had difficulty using the pico mototrs, in part because of large cross coupling between pitch and yaw, and decided it was easier to just go to the floor and center by hand in the end.  I also added a ndscope for centering the MC WFS to the IMC overview screen. 

The spot has returned to a very similar location on IM4 trans after relocking the IMC (0.3P, 0.3Y). 

Looking at a trend of the wind speeds shows that the end stations are consistently below the other buildings. This is obviously dependent on the direction of the wind, but this looks very promising.

The attachment shows the last ~15 hours with the end station traces (blue and purple) consistently below the other buildings. Zooming out got messy, but you get the idea with this shot.

Most sdf diffs were reverted to the values that they would be after a fresh NomLowNoise lockloss.  They only showed up as diffs this morning because the last lockloss wasn't from NomLowNoise (there were lock attempts after the earthquake, as well as an initial alignment). 

---- LSC -----

*  Output Matrix columns 10,11,13 and rows 1,2

Revert.  They are set to +/-1 in ISC_LOCK's lownoise_length_control, so should normally be these values when we have a fresh lockloss.  But ALS_DIFF sets them to zero in PREPARE, so if we lose lock after having tried to do ALS locking, then they'll show up as diffs.  But normally will be found as +/-1.

*  SRCL1 offset

Revert.  Value of 50 is set in lownoise_length_control, and what we'll see after a fresh lockloss.  Value of -800 is modehop value during DRMI acquisition.

* TR_CARM_TRAMP

Revert. Set to 3 in prep_tr_carm.  Set to 10 in carm_offset_reduction main.  Set to 5 in carm_offset_reduction run.

* ALS-C_REFL_DC_BIAS_TRAMP

Revert. Set to 0.1 in prep_tr_carm. Set to 2 later in prep_tr_carm. Set to 10 in carm_150_picometers.

---- OMC ----

* ALS-C_DIFF_PLL_CTRL

Revert.  Offset set to OFF in ALS_DIFF prepare state.

---- ASC -----

* AS_A_RF_36_Q pit and yaw offset

Revert.  Set to zero and offset turned off in ISC_LOCK engage_asc_for_full_ifo. They are turned on and set to finite numbers in ISC_LOCK's drmi_locked_check_asc, in prep for transitioning to them for mich asc.
Reverted everything else.  Input/output matrices, and main ASC filter banks are set in the asc prepare states.  They are here as diffs due to the lockloss not being from full NLN.

---- ALSEX ----

Revert. All diffs are related to WFS.  These will be set by down / prepare states.

---- ALSEY ----

Revert. All diffs are related to WFS.  These will be set by down / prepare states.

--- EX_ECAT_PLC1 ----

* Shutter_A_STATE

Revert.  (Can't revert, is being requested open right now).  But should normally be closed after a NLN lockloss, until LOCKING state on ALS_ARMs.

--- EY_ECAT_PLC1 ----

* Shutter_B_STATE

Revert.  (Can't revert, is being requested open right now).  But should normally be closed after a NLN lockloss, until LOCKING state on ALS_ARMs.

--- CS_ECAT_PLC2 ----

* LSC-REFL_SUM_B in1 and in2 enables

Revert.  They are set with in1 off and in2 on in carm_to_analog, so this is how they should normally be found after a NLN lockloss.

* LSC-REFL_SUM_A in2 enable

Revert.  Set with in2 on in carm_to_analog, so this is how they should normally be found after a NLN lockloss.

* IMC-REFL_SERVO_IN2_GAIN

Revert.  Should be left at -22 at the end of carm_to_analog.

* ALS-C_COMM_VCO_CONTROLS_SETFREQUENCYOFFSET

Revert.  Should be left at -200 at the end of prep_tr_carm.

--- CS_ECAT_PLC4 ----

Leave alone.  This is 3 sqz channels, but the sdf is shared safe & observe.

While performing the SDF cleanup of the SAFE.SNAP files this morning, we also cleaned up some of the inconsistancies related to the alignment offsets and switches.  We referred back to the rule that the OFFSET ON/OFF toggle switches are adjusted via Guardian on numerous suspensions during alignments and were therefore set to be not-monitored, while the OFFSET values themselves were set to be monitored.  If there is a suspension that does not get misaligned via a guardian script, then the OFFSET value and button are still set to be monitored.

   Performed quarterly maintenance on all accessible deployed dust monitors. 

CS HAM6 and the Biergarten dust monitors passed the Zero Count and Flow tests. No problems noted. 

Optics Lab - Both dust monitors passed the Zero Count and Flow tests. No problems noted. 

Diode Room - The dust monitor passed the Zero Count test. The flow is a bit on the low side but still within spec. Will continue to watch for change. 

PSL - These check were made before the holidays. All was normal. 

Did not check the End Station dust monitors due to tumbleweeds blocking the access roads.    

J. Kissel, H. Radkins 

Hugh and I reconciled the seismic systems SDF systems this morning, with 
- The sdf file set to the safe.snap
- the platforms (both ISI and HPI) offline (i.e. the SEI_{CHAMBER} gaurdians were set to OFFLINE), 
- the site-wide sensor correction OFF (i.e. SEI_CONF manager was in SC_OFF_NOBRSXY), and 
- the SEI_DIFF (aka CPS DIFF)  also off (i.e. SEI_DIFF is in the DOWN state).
This is the "no seismic systems or control loops are functional, but they are ready to be made functional" state that we anticipate we want the platforms to be in upon return from a site-wide (or at least seismic wide) computer failure.

In doing so, we made sure that each front-end model that controls these operations (h1isi{chamber}, h1hpi{chamber}, and h1seiproc) had these channels of their safe.snaps appropriately monitored and settings accepted in the above mentioned configuration.

Note, a lot of the sensor correction channels and CPS diff channels are *unmonitored* in the OBSERVE.snaps (the file against which these models are almost always compared, since we expect the seismic system to always be *fully* ON and functional, i.e. rarely-if-ever in its "safe" state), because (a) we're constantly adding states and configurations to these systems and guardians, and (b) we still regularly explore using these different configurations without wanting to take the IFO out of OBSERVATION READY (because we believe that the changes, while they may cause different IFO behavior when *in* a given configuration, the *switching* of configuration is slow enough compared to g-wave signals that it does not warrant not observing.)

I attach a screenshot of a few examples of the channels I captured for the HAMs (Hugh did the BSCs, but didn't take screenshots).

I'll note that the last attachment shows the h1seiproc model showing diffs for the ON/OFF switch of the CPS diff system, but we've left this as a DIFF because we want the system to be OFF (0.0) when the computer reboots, but we currently have it turned *on* because we want it on, and understand why its on (see LHO aLOG 54322).

Corner station dust monitor vacuum pump is running well. The vacuum pressure was good. The temperatures were well within operating range. There was no carbon dust noted on the muffler. 

Could not check the End Stations due to the tumbleweeds. 

Closing FAMIS #13004

Check the 4 pump skids on the CS Mezzanine. Fluid level is 6 6/16, + 1/16 over last check. No leaks observed. 

Could not check the end stations due to tumbleweeds. Will check at next available opportunity. 

Closing FAMIS task #13494

Jenne confirms that they will re-evaluate the R0 alignment efforts to mitigate potential scattering between the chains possibly at the ESD, so I have reverted the P and Y offsets to zero.

H1:SUS-ETMX_R0_OPTICALIGN_P_OFFSET (and then Y and again for ETMY)

TITLE: 01/07 Owl Shift: 08:00-16:00 UTC (00:00-08:00 PST), all times posted in UTC
STATE of H1: Preventive Maintenance
INCOMING OPERATOR: TJ
SHIFT SUMMARY: Tricky night with big earthquakes and the winds just toped 70mph
LOG:

• 08:34 SEI_CONF to LARGE_EQ for incoming 6.5Mag from Puerto Rico
  • This EQ was really big, 0.03-0.1Hz up to 10μm/s, but no WDs tripped!
• 09:44 SEI_CONF to EQ
• 10:28 LSC tidal errors on verbal alarm. From alogs it seems that going from EQ to Windy will help.
• 10:31 SEI_CONF to WINDY
• 11:10 NLN and Observing
• 11:39 SEI_CONF to EQ incoming 5.6Mag from Puerto Rico
• 11:47 Lockloss see alog 54329  I had trouble getting the IMC to lock.
  • I have had the observatory mode set to wind. Not sure if that's correct.
• 14:19 Started Initial Alignment
• 15:19 Forced ISC_LOCK to DOWN during initial alignment as initial was waiting too long at ACQUIRE_SRC for WFS and I couldn't see an easy way to get it to skip past.

TITLE: 01/07 Day Shift: 16:00-00:00 UTC (08:00-16:00 PST), all times posted in UTC
STATE of H1: Preventive Maintenance
OUTGOING OPERATOR: Camilla
CURRENT ENVIRONMENT:
    SEI_CONF state: SC_OFF_NOBRSXY
    Wind: 48mph Gusts, 35mph 5min avg
    Primary useism: 0.21 μm/s
    Secondary useism: 0.40 μm/s
QUICK SUMMARY: Maintenance starting, very high winds on sites.

TITLE: 01/07 Owl Shift: 08:00-16:00 UTC (00:00-08:00 PST), all times posted in UTC
STATE of H1: Lock Acquisition
OUTGOING OPERATOR: Jim
CURRENT ENVIRONMENT:
    SEI_CONF state: WINDY
    Wind: 27mph Gusts, 22mph 5min avg
    Primary useism: 0.21 μm/s
    Secondary useism: 0.54 μm/s
QUICK SUMMARY: Down for a few hours now, there's been high winds and a big EQ since lock loss preventing relocking. Hopefully there earth is settling down a bit now.

TITLE: 01/07 Eve Shift: 00:00-08:00 UTC (16:00-00:00 PST), all times posted in UTC
STATE of H1: Lock Acquisition
INCOMING OPERATOR: Camilla
SHIFT SUMMARY: Bad environment
LOG:
Niko had a hard time with the wind, I made progress and got locked turning of the arm CPS differential,  but then a 50 mph gust broke lock. Haven't locked since then because as the wind came down and Malayasian earthquake came in and made locking difficult.

Just got locked again after Niko had several hours of rough weather. Wind is only somewhat less bad than before. The only thing I changed to successfully get locked was turn on the arm cps differential controls (I took SEI_DIFF to FULL_DIFF_CPS). Microseism is high, so scatter is bad, and our range is taking a beating, which is not unexpected, but the ADS lines are not swamped, we seem to be stable.

First attached plot is the state numbers for the ISC LOCK and SEI DIFF. The dip in the blue line is where I switched SEI_DIFF to FULL_DIFF_CPS at about 4000 seconds in, you can see after that after that I was able to get to NLN, and we've been in OBSERVE since.

Second attached plot shows 2.5 hours of winds at the ends and corner for the same time as the first plot. The blue corner station wind is consistently higher, maybe because that sensor is more affected by the flow accelerating over the building? Or Maybe the fences are doing there jobs? The low frequency blrms on the wall seem to show less separation between the CS vertical eq band blrms and the end horizontal blrms than we used to see. Worth a different study. But, you can also see that CS winds actually increased while we were locking after the SEI_DIFF switch at 4000 seconds. The ends rolled off a bit for a while, but seem to be back up again, now.

The third plot are some channels from the endstation BRS, being the best direct ground tilt sensors we have. The top plot is the .03-.1hz blrms from each brs, you can see that EY is consistently higher, and it kind of rolled off around the time I switched SEI_DIFF, but it comes back higher at the end of this window.

These BRS blrms channels (just like we have for the ground STS) seem like they could be really useful for environmental studies, but I don't think they've gotten a lot of attention. For, instance, I would guess that without FULL_DIFF_CPS 5 nrad/s in the .03-.1hz blrms is a useful threshold, we couldn't seem to lock when BRSY was hitting this level, but when we had the arm cps diff on we got up to 7 nrad/s and it was okay. Worth looking at more.

Bottom subplot for the third plot is the velocity readout from each brs, it's a bit contaminated by the brs resonance, not sure what else this shows other than it didn't get quieter.

The takeaway, I think, is the even if the microseism is high, if the wind is bad and preventing locking, it maybe worth it to go to the FULL_DIFF_CPS. This is something we will have to think about

It's pretty breezy out, and the microseism is creeping up above our 90th percentile.  As a result, we have pretty terrible range.  Sheila pointed out that a simpler way to test whether the scattering that Robert postulates (in alog 54298) does indeed come from scatter between the end test masses and reaction masses is to change the angle of the end reaction masses.  If we can affect the scatter by changing the angle between the etm and reaction mass, then it points to Robert being correct that this relative displacement is a source of scatter. 

Keita okay-ed this work to happen during Observation while the environment is bad, as long as the ramp times for moving the reaction chains are long.  In order to do this in Observe, Niko unmonitored R0 offset and tramp channels for the optic align sliders for both ETMs. 

I started moving both the ETM reaction chains at the same time, only in pitch, up to about 8 urad from their nominal positions.  I was getting some pretty bad glitches (which I think were due to increasing wind), so just in case, I moved them both back to their nominal positions.  About 80 seconds after those ramps finished, we lost lock.  I attach a screenshot of my ndscope, but mostly as a reminder to myself what channels I was looking at.  I don't think I had moved enough yet to see anything difinitive.  I had been hoping to move a few tens of urad, both in pitch and in yaw, and then do a series of ~10 min new positions, 10 min old positions, 10 min new, etc. to see if the angle really makes a change.

I wonder however if I was moving the ETMs by moving the reaction chain, and since the glitches are loud enough to drown out the ADS lines, the ETMs weren't really following.  Something to try when we get relocked will be to turn up the ADS line amplitudes to get better signals, and make sure that the SOFT angular degrees of freedom are doing what they need to do.