TITLE: 12/11 Eve Shift: 00:00-08:00 UTC (16:00-00:00 PST), all times posted in UTC
STATE of H1: Observing at 72.7965Mpc
INCOMING OPERATOR: Jim
SHIFT SUMMARY:

Lock is going on 31+hrs with a range drifting slightly down from 75 to 70Mpc over the last 8-10hrs.  useism has sorty of flattened out just under the 90 percentile.  

• Investigated A2L during this lock.
• Continued work to figure out what channel knocked us out of OBSERVING during Cheryl's shift.

LOG:

• Continue to have Dust Monitor alarms throughout the shift.
• A few EY glitches:  0:28, 0:44, 6:49, 7:20, 7:21
• Heard noisy clicking noise in the OSB area above Hugh's Office.  I looked above the ceiling panels and noise was coming from electrical boxes attached to Ventilation shafts.  Left messages with Bubba, Joh, & Richard.  Richard called and said it was probably the heater.
• 6:41 Noticed RED on CDS Overview!  Timing Error for H1IOPASC0
  • 7:50 While in OBSEVING, hit DIAG RESET to clear the Timing Error & it went away (still have an overflow on it).

Cheryl observed an instance of H1 being dropped out of OBSERVING due to SDF changes tracked down to the Computer/SDF Node:  sysecaty1plc2

(This was for the Yarm.  We noticed this same issue a week ago for the analogous Computer/Node for the Xarm:  sysecatx1plc2.)

I continued the work of figuring out who is our pesky channel dropping us out of OBSERVING.  The first thing I did was look at the (3) channels Keita found for X last week and see if the Y-arm counterparts changed today---found nothing in dataviewer.  I then ran the scripts Cheryl ran and came with the same result of seeing a change with the channel H1:FEC-1031_SDF_DIFF_CNT.  But this is just a name for a channel SDF uses.  

I then just went to where sysecaty1plc2 is on medm.  This is related to Beckhoff, so maybe the channel can be tracked down by snooping around medm land.  To get to a baseline/starting point, I went to:

SITE MAP / SYS / EtherCAT overview / H1 Y1 PLC2 /

From here you have several different subsystems (Als, Asc, Isc, Lsc, Sys).  So, I went through all of these subsystems and the screens nested within them.  The first thing I did was to find the "*_ERROR_FLAG" status box for each subsystem (it's green for all, and I reckon if there was a change to the system, it would go red).  So I grabbed this channel for all the subsystems mentioned above, and the only one which changed when we dropped from OBSERVING was the Als one.  I then played the same game--go into the nested windows within and trend "*_ERROR_FLAG" channels for each component within Als.  Ultimately, I ended up finding a single channel which had activity around the time in question.  It was found here:

SITE MAP / SYS / EtherCAT overview / H1 Y1 PLC2 / Als / Y / Fibr / Lock / Temeraturecontrols  (i.e. H1ALS_Y1PLC2_Y_FIBR_LOCK_TEMPERATURECONTROLS.adl)

And on this medm, the channel in question is:  H1:ALS-Y_FIBR_LOCK_TEMPERATURECONTROLS_ON

I'm not saying this is the ONLY channel which could be culprit for the OBSERVING drop, but this is one I saw drop out at that time (see attachment#1), BUT there is a caveat.  If I look at 20min before the Drop, the ALS channel in question had some similar drop outs (see attachement#2).  For the earlier one, the drops only lasted about 10sec (attachment#3).  For the drops which took us out of OBSERVING (attachment#1), after 15sec of drops, then we dropped out of OBSERVING (& overall the ALS ON switch went off/on for about 40sec).  So maybe the SDF changes have to happen for a certain amount of time before latching us out of OBSERVING?

As another check, I looked at the last 12hrs of this lock and the only time H1:ALS-Y_FIBR_LOCK_TEMPERATURECONTROLS_ON had these fits of turning OFF for a handful of seconds were in that 20min time period when we dropped out. 

Question:  Is this enough to warrant NOT MONITORING H1:ALS-Y_FIBR_LOCK_TEMPERATURECONTROLS_ON?  Or should we keep searching?

WP 6385

I installed Conlog on the conlog-master and conlog-replica machines. I am going to leave it stopped and not acquiring data until I finalize the channel list with Dave. The work permit should be left open.

Keita made a template to passively measure the Quality of A2L for H1 & he said it would be good to look at this over a long period (alog#32106).  We are currently on a 25+ Hour Lock.  During this lock:

• 1st A2L was run 1-2hrs after lock.
• 2nd A2L was run 24hrs later

Before running the 2nd A2L I ran Keita's template for once an hour during the lock.  I saw that many hours (~16hrs) after the 1st A2L that the DARM & ASC Coherence for PITCH began to increase again, and then I ran the second A2L.  (I mainly focused on PITCH because it was the DOF which showed the most change.)

After both A2Ls, didn't really notice an improvement in range.  Did notice a better looking DARM spectra (on nuc3).  

Question:  If the DARM & ASC Coherence increases like this, do we want to run the A2L?  Is this the right thing to do?  Or do we just let the Coherence increase?

Attached is a look at the PITCH coherence for every hour during this lock (with A2L moments marked).

Timeline for A2L on current 24+hr lock:

• 12/10 1:02utc:  OBSERVING
• 12/10 2:32:  A2L run
• 12/10 16utc - present:  Pitch Coherence between DARM & ASC gets higher
  • Ran Keita's DTT template at many different times during this current lock and started to see the Coherence increase around 16utc.  It was even higher around now.  So decided to...
• 12/11 1:00utc:  Ran a 2nd A2L of this lock.

SDF Channel File Changed

Also wanted to take this time to look at the channels involved with SDF node which dropped Cheryl out of Observing during her shift.  (Do we really want to be dropped out of OBSERVING for looking at some SDF channel files??)

OBSERVATORY_MODE

Took to "CALIBRATION" for this 4mins of downtime from OBSERVING

TITLE: 12/11 Eve Shift: 00:00-08:00 UTC (16:00-00:00 PST), all times posted in UTC
STATE of H1: Observing at 74.5957Mpc
OUTGOING OPERATOR: Cheryl
CURRENT ENVIRONMENT:
    Wind: 4mph Gusts, 3mph 5min avg
    Primary useism: 0.03 μm/s
    Secondary useism: 0.47 μm/s
QUICK SUMMARY:

Can see a slight increase in the "secondary useism" over the last 12hrs. 

Looking at Violin Mode peaks on DARM, we have:

• primary at 2x10^-19  (~1000x smaller than yesterday at the beginning of this lock!)
• secondary at 4x10^-16

LiyuanZ, PeterK, BetsyW, AlenaA, RickS (with support from EddieS, CalumT, StephenA, DennisC, GarilynnB, MalikR, et al.)

Some time ago, LLO removed PMC SN08 from operation due to a glitchy PZT.  

We recently removed and replaced the PZT and the curved mirror it actuates, using an original-style PZT ordered by Pking and a spare mirror from the original PMC mirrors (supposedly) provided by BennoW.

We characterized the losses in the cavity using a setup in the LHO "Triples Lab" (upper floor of Staging Building) that utilizes an NPRO and three Pcal-style integrating spheres and associated photodetectors (see LIGO-T1600204-v3).

We made some improvements to our measurement setup that preclude direct comparisons of the estimated losses before and after replacing the M4 mirror, but we estimate that the total round trip losses were reduced by about a factor of two by replacing the one mirror (investigations of other highly contaminated PMC indicate that the PZT is the source of the contaminants and that the mirror bonded to the PZT is the most contaminated).

Our current best estimate of the average losses per mirror for this cavity is about 60 ppm (see attached table).

We were surprised to find that the transmitted light level through M4 is about 40 times smaller than M3.  The spec for the M3 and M4 transmission is 60 ppm and we calculate the M3 transmission to be about 65 ppm, but the M4 transmission of only 1.6 ppm is a mystery.  It doesn't seem it was from the same coating run, as expected.  However, discussions with DanielS indicated that this might be acceptable.  The M4 transmitted light is used for the ISS path and is typically attenuated by about a factor of 100 on the PSL table.

We also tested a new concept for fabricating the PMCs that relies on machining tolerances for setting the orientation of the four cavity mirrors and eliminates all gluing from the assembly.  Two original PMC bodies were re-machined at a local machine shop in a single setup with the hope of achieving relative accuracy between the points on which the cavity mirrors register on the level of 5 micrometers.

We assembled both "all-bolted" prototypes under a clean bench by mounting the mirrors against three balls that register at the bottoms of counterbores in the aluminum bodies and holding the mirrors (and the sandwiched PZT) in place using off-the-shelf SS flexures (see attached photos).  We used mirrors recently procured by PeterK from ATF.

We discovered that there was an error in the coating of the new PMC flat mirrors; the reflectivity is only 2,400 ppm when it was supposed to be 24,000 ppm.  Thus the cavity finesse is 10x higher than desired.  While this won't work for the PSL, it aids in measuring the mirror losses.  The results of two measuremens are tabulated in the attached table for the S/N10 body.  The average losses per mirror are estimated to be about 11 ppm.  We have not measured the losses for the other "all-bolted" cavity yet.

These measurements confirm the abiltiy to machine the bodies to the required tolerances.  We will test the PZT performance as best we can in our lab setup when time allows.

TITLE: 12/10 Day Shift: 16:00-00:00 UTC (08:00-16:00 PST), all times posted in UTC
STATE of H1: Observing at 73.9225Mpc
INCOMING OPERATOR: Corey
Assistance:

•     Corey and Keita

SHIFT SUMMARY:

•     locked all shift
•     in Observe most of the time
•     SDF kicked H1 out of Observe twice
•     first time was EY ECAT PLC2, which had a channel toggle three times - very possibly no change to H1 config
•     second time I made a bad click and SDF went red and knocked H1 out of Observing - absolutely no change to H1 config
•     alog 32420 has all the details of the drop in Observe
•     one earthquake - H1 stayed locked and in Observing
•     no PI modes needed to be damped
•     some wind up to 20mph, but mostly under 10mph during the shift

• 19:19:45UTC - H1 kicked out of Observe
• 19:24:52UTC - H1 back in Observe
• investigating

• 17 hours of lock
• TCSY chiller flow (bottom), power out (top)
• flow glitches:  16:44UTC, 17:04UTC, 17:59UTC
• power out is zoomed in quite a bit, so some changes but small

• Reported on Terramon
• Magnitude: 6.1
• Location: 131km WNW of Arawa, Papua New Guinea; LAT: -5.7, LON: 154.5
• Noticeable on BLRMS around 16:53UTC
• State of H1: remained in OBSERVING
   

H1 has been locked 15+ hours, and I ran some dataviewer trends looking at 2 hours, and found some optic alignment changes that I think are interesting, and some that might show up in DARM.

• plot 1: IM3
  • around 15:53UTC, glitches in DAMP_P_INMON and DAMP_Y_INMON, small but a clear change
• plot 2: PR3
  • around 15:40UTC, amplitude of OSEM pitch and yaw signals decreases, around 15:49UTC, Oplev shows a clear ramping of pitch that goes from 0.22 to 0.45
• plot 3: SR3
  • around 15:36UTC, oplev pitch starts to change, around 15:44UTC, oplev yaw starts to change
• plot 4 and plot 5: RM2
  • four glitches in yaw starting at 15:12UTC (-9urad), 15:35UTC (-8urad), 15:36UTC (-22urad), and 15:46UTC (+11urad)

TITLE: 12/10 Day Shift: 16:00-00:00 UTC (08:00-16:00 PST), all times posted in UTC
STATE of H1: Observing at 70.6121Mpc
OUTGOING OPERATOR: Jim
CURRENT ENVIRONMENT:
    Wind: 19mph Gusts, 14mph 5min avg
    Primary useism: 0.02 μm/s
    Secondary useism: 0.38 μm/s
QUICK SUMMARY:

• H1 locked 15+ hours
• wind increased about 2 hours ago
• GRB arrived 16:06UTC

TITLE: 12/10 Eve Shift: 00:00-08:00 UTC (16:00-00:00 PST), all times posted in UTC
STATE of H1: Earthquake
INCOMING OPERATOR: Jim
SHIFT SUMMARY:

Once I sat down to H1, able to get to NLN with no issues.  Have had a handful of EYSaturations during this lock.  Seismic trends are steady & winds are getting quiet.

LOG:

• Locked up H1 (only really needed TMSx tweak, and the usual BS & PRM adjustments).
• 1:02  Undisturbed
  • Mode28 was acting up and I hammered it down as soon as I could (fairly easy)
  • Then Mode27 slowly started acting up.  It took me about 35min of work to get it to finally drop & stay down.  I tried addressing it before the alarm came on (as I did for #28).  This was the 18038Hz mode.  I tried phase changes with only a little bit of luck.  Then I adjusted the gain.  And then I tweaked on both.
• 2:32:  Out Of Observing.  Ran A2L
  • For Observatory Mode, I wasn't sure what to select....I figured Calibration or Corrective Maintenance.  I opted for the latter, but now I'm thinking I should have selected "Calibration" (we should really have a differnt name for this one...something like "OBSERVATION MEASUREMENTS" (so Calibration, injection, A2L, and the like could all be inclusive).  CORRECTIVE MAINTENANCE implies a failure and thus an FRS (I did not file one).  
• 2:36:  Back to Observing
  • 3:04 Rode through an EQ
• Adjusted the PSL Dust Monitor #101 Alarm Levels

With the ASC IMC model now running at 16384 Hz we look at the coherence of jitter as measured by the IMC WFS and other channels up to 7.4 kHz. Not sure we can conclude anything except that pointing errros contaminate everything.

We can compare this with an older 900-Hz bandwidth measurement from alog 31631 which was taken before the piezo peak fix (alog 31974).