TITLE: 02/08 Eve Shift: 00:00-08:00 UTC (16:00-00:00 PST), all times posted in UTC
STATE of H1: Observing at 67Mpc
INCOMING OPERATOR: TJ
SHIFT SUMMARY:  Observing for ~3 hours after relocking post-Maintenance Day and Sheila's OMC work.  PI modes have been better behaved than my previous shift.
LOG:  See previous aLogs.

 

In alog 33915 I used my images of the steering mirror under the thin film polarizers to identify possible clipping.

In alog 33986 I posted images of the clipping I found today.

This shift in the beam on the steering mirror would come about from a Nothward shift in the beam on the thin film polarizers, which would indicate an Eastward shift in the beampath of the EOM.

Here are images supporting the beam shift:

Attachment 1: images of the wedge in the EOM path shows the beam moved Eastward

Attachment 2: images of the steering mirror shows the beam moved Northward

Attachment 3: image of the iris down stream of the bottom periscope mirror shows the beam moved Southward

Attachment 4: Drawing that shows that all of these beam changes are consistent.

Both of our locks since today's maintence period have had large glitches at low frequency.  In the first lock we also had a reappearance of the 1083Hz OMC length glitches (and the corresponding peak in our spectrum).

The OMC length feature coming back may be due to more low frqeuency motion of the interferometer because BRSY is not in use, https://alog.ligo-wa.caltech.edu/aLOG/index.php?callRep=33978

I added a boost to the OMC length loop to keep it on resonance better, measurements attached.  This seemed to help the 1083 Hz peak.  I turned the boost off from 4:14 UTC to 4:21 UTC Feb 8th UTC

If we want to make this boost permanent we should rearrange filters in OMC-LSC_SERVO and update the OMC guardian.  For now I have put it in OMC-LSC_I_SWSTAT FM3.  Travis accepted this in SDF, and has gone back to observing.  This data may be ususable because of the huge glitches. 

Operators: If you have to relock later tonight, you can engage the boost for the OMC by turning on FM3 in OMC-LSC_I (you can find this under OMC> CONTROL> green button saying I in the center near the top).  

We have had several (~6-8) instances tonight of the PSL FSS PZT oscillating and knocking the IMC out of lock.  I have been toggling the Autolock on/off to fix it.  I had the same issues last night while relocking.

J. Kissel

I've gathered this week's effective bias voltage measurement which provides an alternative confirmation that the actuation strength of the electrostatic drive (ESD) systems is not varying significantly over time from the accumulation of charged particles. This is no longer an issue because of the regular sign flipping of the ~400 [V] of requested bias that we began in November of last year (see LHO aLOG 31929 for ETMY and LHO aLOG 31172).

Attached are plots showing the effective bias voltage trend over the past year for each quadrant of control, as measured by the optical lever, in both pitch and yaw. The first is for ETMX, and the second for ETMY. One can see the marked change in effective bias voltage accumulation starting in November, as the trend turns over and becomes flat over the past few months. The last two plots re-cast this trend in terms of relative actuation strength of the ESD systems. One can see that the relative actuation strength change caused by the accumulation of charge remains small and static in time, at the level of a few percent. 

We can use this confirmed static angular actuation strength as a proxy to confirm that the longitudinal actuation strength remains static, as shown in LHO aLOG 24241. That means the actuation strength we use in our model of the interferometer's differential arm length control system when it was updated last in January (see LHO aLOG 33004) is still valid, and time-independent filters we used to calibrate the real-time output of the detector -- which are based on that model -- are accurate to within these limits. (Any remaining small time-dependent change has been corrected for in the almost-real-time, low-latency correction algorithm since the start of O2; see LHO aLOG 31732.

Sheila is working on OMC length dither while LLO is down.

Python standard environment work

Jim

HVAC networking between OSB and LSB

Carlos, Jim, Apollo

Installed a new HVAC unit on the network link between OSB and LSB

Remove timing status of DTS fanout from Beckhoff SDF

Dave

Modifed the ecatc1plc1 observe.snap to make IFO observation mode agnostic to DTS status.

Patch and reboot h1dmtlogin

Dave:

The debian 2FA authentication machine for DMT was patched. It had been running for 245 days, so it was rebooted. It came back quickly, indicating an FSCK completed quickly or was not ran.

Note to control room, the EPICS readback of H1 range goes to -1.0 when h1dmtlogin is rebooted, it acts as the proxy server for this GDS data.

Conlog channel lists

Patrick

Generation of the channel list for conlog was improved.

TITLE: 02/08 Eve Shift: 00:00-08:00 UTC (16:00-00:00 PST), all times posted in UTC
STATE of H1: Aligning
OUTGOING OPERATOR: Jeff
CURRENT ENVIRONMENT:
    Wind: 14mph Gusts, 11mph 5min avg
    Primary useism: 0.21 μm/s
    Secondary useism: 0.41 μm/s
QUICK SUMMARY:  Just finishing up IA after Maintenance Day.
 

I modified the CDS overview to:

1. show the top level status of the Timing system, plus the associated information string

2. conlog has returned

Ops Shift Log: 02/07/2017, Day Shift 16:00 – 00:00 (08:00 - 16:00) Time - UTC (PT)

Jim sat cross legged looking through a 1.5" viewport for close to 5 hours attempting to move a few gram bar a few microns with no tactile knowledge that he was actually touching it.  Back and forth, back and forth.  Jim has successfully done this before; I never have.  Krishna and Michael have done so many times.  Finally we decided that at least in the current position it will drift into a useable place in a few weeks.  It is a little hard to predict with all the thermal activity but historically, this is where we believe we are.

We, commissioners, may decide we need to speed this along with another intervention but with the fickleness of this we can make no promises.

Cheryl, Kiwamu, Sheila

This morning Cheryl went into the PSL and saw that the rejected beam from the thin film polarizer after the rotation stage was hitting the edge of the steering mirror which sends it to a beam dump.  (She has nice photos of this).  This aftrernoon she, Kiwamu and I went into the PSL to fix this.  

We first adjusted the position of the iris after the bottom periscope mirror, and added two steering mirrors, and iris and a beam dump to the path to the OSA so that we would have good alingment references for the beam leaving the table.  We then turned the power down before the PMC, and using base plates as markers moved the entire TFP assembly 5 mm closer to HAM1.  Kiwamu took some photos of the resulting beam position on the steering mirror which are on Cheryl's camera.  The beam was then high on the beam dump apperature, so we changed the spacer in the beam dump mount for a thicker one.  We turned the power back up, and checked the beam on the irises and the persicope PD.  

Cheryl's photos from this morning suggested that there could be clipping on the output apperature of the EOM.  We tried to determine if this was clipping or a back reflection, but can't say for sure.  Now we are out of the PSL, the IMC relocked without problems, and Jeff B has started inital alingment. 

Jim B., Patrick T.

The command line clients for Conlog have been installed. They are 'conlog-at', 'conlog-diff', 'conlog-change' and 'conlog-list'. There is a -h help option for each.

Examples:

conlog-at '1 min ago' UTC 'H1:SUS-ITMX*'
conlog-diff '1 hour ago' '1 min ago' UTC 'H1:SUS-ITMX*'
conlog-change '1 hour ago' '1 min ago' UTC 'H1:SUS-ITMX*'
conlog-list '1 hour ago' '1 min ago' UTC 'H1:SUS-ITMX_R0_WD_OSEMAC_RMS_MAX'

The searches are described in the user guide presentation: G1401113 v2.

The web interface is at h1conlog.

Conlog is currently only accessible from inside the CDS network.

See 33648 for previous work.  Basics--There is an STS2 seismometer piggy backed with the BRS.  It however, exibited some noise and I found it mis-centered last week.  Despite many centering attempts last week I could not get it to center the masses in an expected timely nature.  After this work, there appeared a glitching behaviour.  On Thursday, I checked and found the masses with the same voltage reading and all the other bad things persisted.  I hit the mass center button and walked away.  After this, the X Y Z signals suggested that the masses were now better positioned but the glitching behaviour remained.

This morning when we arrived, the mass positions where much closer to okay than they ever had been:  0.1 -2.2 and 5.5V for U V W. Jim checked the feet for tightness and found them secure.  First thing was to power cycle the system to see if I could clear the glitches.  After the power cycle, the mass positions weren't any better but not too much worse.  Three mass centerings left the positions in an okay place: -.3 -.5 & -1.6V. Steckeisen recommends <2.0volts is acceptable.

Sadly though, it still appears to have positive going glitches and early looks at the spectrum suggest the glitches are still there.  Also, since we spent a few hours in the BRS box, we may now be in a thermal instability that we must wait out. We could end up out of position after the temperature stabilizes but this can be corrected with another mass centering when possible; but, the glitches are another thing.  Might have to attempt another power cycle if they continue.

I'll post data later after  things settle.

After moving the EE timing fanout from its direct connection to  the timing master over to the DTS fanout, we realised that if the DTS fanout unit was powered down this could cause H1 to go out of observation mode. This is because the Beckhoff SDF is monitoring the ACTIVE channels for all ports. If a client is powered down, the timing status would go RED. The only way to green it up, would be to reactivate the client or press the ACTIVE MEDM button, which would then take the IFO out of observation.

this morning I un-monitored the ACTIVE and SLAVE_NAME channels for port_8 on the master.

david.barker@zotws2: svn di h1sysecatc1plc1sdf_OBSERVE.snap

-H1:SYS-TIMING_C_MA_A_PORT_8_ACTIVE 1 Active 1
-H1:SYS-TIMING_C_MA_A_PORT_8_SLAVE_NAME 1 "Fanout Test Stand" 1
+H1:SYS-TIMING_C_MA_A_PORT_8_ACTIVE 1 Active 0
+H1:SYS-TIMING_C_MA_A_PORT_8_SLAVE_NAME 1 "Fanout Test Stand" 0

   Sweep the LVEA up to the PSL. There is still alignment work ongoing in the PSL. When the crew if finished in the PSL they will complete the sweep.