The LHO GC border router has had its media converter and transceivers replaced with a single transceiver to improve reliability and diagnosability, along with the matching equipment at our primary ISP.

Dave, Jonathan, Greg, Daniel

Last Tuesday (June 23rd 2015 between 12:30 and 15:50 PDT), Dave and I took a GPS time output signal from one of the rear BNC ports on a Timing IRIG-B module in EY. We temporarily fed the signal into the 16kHz rack microphone channel (see https://alog.ligo-wa.caltech.edu/aLOG/index.php?callRep=19299):

H1:PEM-EY_MIC_VEA_MINUSY_DQ

in order to record the signal in frame. Dave then found the frame file, and Jonathan extracted the first second of the H1:PEM-EY_MIC_VEA_MINUSY_DQ channel (each raw frame file has 64 seconds of data and is named according to the first second) into both text and binary formats. I then read the binary file (a bunch of 32 bit floats) using Julia, and compared the resulting array to the NDS-generated text file to confirm that the values were the same. I plotted the signal and read it twice, recording the long/short/control pulse sequence each time to make sure I'd read it correctly. I then decoded it using the IRIG-B signal spec. Finally, I compared it to the ostensible recording time, as indicated by the filename. c00010000c010000100c100000100c001001110c100000000c101001000c0

The signal as I recorded it (control = c, short = 0, long = 1) along with the decoded time; it matches the GPS time of the frame (which is 16 seconds later than the UTC time of the frame). A plot of the data is attached.

c00010000c010000100c100000100c001001110c100000000c101001000c0
 0001 000 0100 010  1000 01   0010 1110 10        1010 1000 8
 8s       22m       21h       174d                15y

Which corresponds to June 23rd, 2015 at 21:22:08 GPS time.

The GPS time (from the name of the frame, as recorded in the attached text file dump) is 1119129728.

This corresponds to Jun 23, 2015, 21:21:52, which is 16s early, as expected, due to the 16 current leap seconds.

It seems, then, that the IRIG-B Timing Module's signal isn't misaligned with the DAQ (at least not catastrophically).

The GDS tools (foton, diaggui, awggui, diag, etc) have been updated to gds-2.17.1.1 to handle conversion of UTC to GPS times for the upcoming leap second.  The leapsecs.dat file which dataviewer reads has also been updated, so dataviewer sessions started after the leap second should do the UTC to GPS conversions properly.

This change affects Ubuntu 12.04, Ubuntu 14.04, and SL6 workstations.

Reset the accumulated WD counters for HAM2, HAM5, ETM-X, and ETM-Y

Reset the accumulated WD counters for HAM2, HAM5, ETMX, and ETMY

Scott L. Ed P.

6/25/15
Cleaned 68 meters ending 7.6 meters north of HNW-4-091.

6/26/15
Cleaned 51 meters ending at station HNW-4-094. Test results posted here.

Removed lights, begin relocating cords and lights to next (AND LAST SECTION ON X-ARM).

6/29/15
Complete hanging of lights and begin vacuuming support tubes and spraying diluted bleach/water solution on (many) soiled floor areas.
Expansion bellows are excessively dirty, therefore we will do a pre-clean on these and go back over with regular cleaning. Filled 200 gallon D. I. water tank.
Only able to clean 5 meters of tube after all of that.  

The contractor is on site to sweep the gravel and rocks off of the access roads at each arm. 

The sweeper machine moves at a very slow pace so please be aware of this when traveling up and down the arms.

I've collected transfer functions for all optics and all DOFs in HAM6.  Everything looks good and we should pump down.  A subset of the results, for OMCS L and OM1,2,3 P are attached; the xml files have been saved and committed to the SUS SVN.

A report on the performance of the seismic isolation subsystem during ER7 at both sites has been compiled on the DCC: G1500811

Summary: 

• High wind adversely affected uptime at LHO during ER7
• SEI watchdogs still trip without any apparent ground motion cause at both sites, but less frequently than one year ago 
• Noted SEI watchdog trip trends (sensors that trip semi-regularly with no apparent cause):
  • BS ST2 GS13 at LLO 
  • HAM 6 ISI CPS at LLO
  • EMTX HEPI actuator and ST1 T240 (within ~10 minutes) at LHO
• No sign of upconversion from trains at LLO, but not many instances of elevated ground motion in 1-3Hz range

ER7-driven requests from DetChar for the SEI team: 

• Work to improve uptime via mitigation of earthquakes and wind-driven ground motion and addressing WD trips uncorrelated with ground motion

To test if NDS clients after tomorrow will correctly apply the upcoming leap second, I injected a sub-second signal into a DAQ DQ channel today. The H1:PEM-EX_ADC_0_08_OUT_DQ channel is a PEM test channel with ADC bit noise on its input. I ran awgstream to inject a signal which is one second long, and all zero except for 1/16th of a second starting at the 0.5S mark. This 1/16th second block is a step function, with an amplitude of 1000 counts.

The time of injection is chosen to be modulo 64 in GPS time, so it will be the first second of a 64second DAQ frame. The time of injection is

The model runs at 16kHz, this channel is acquired to the DAQ at 2048Hz. The decimation filter is producing the ringing seen at the step transitions.

Gerardo, Bubba, Kyle 

Installed HAM6 East and South Doors 

***Noticed many scratches on HAM6 East Door chamber flange - yes, where the door O-rings contact -> This is due to the fact that spring clamps have been used to secure the soft-covers to prevent the purge-air from blowing them off as is the norm but, unlike the normal case, for this flange, the O-rings are located in grooves on the door instead of the chamber.  As such, the O-ring protectors are on the door and not the chamber.  So, the spring clamps clamp against the machined sealing surface of the chamber flange and only the C3 fabric acts as a barrier between the steel clamps and the steel flange and scratching is inevitable.  Let us not do this again!  In the future, O-ring protectors should be installed on the chamber flange in addition to the door flange for this special case.  


Kyle, Gerardo 

Connected pump carts to HAM5 and HAM6 annulus pump ports -> Began pumping annuli(?) -> Will pump HAM6 tomorrow.

I repeated the charge measurements on both ETMX and ETMY. Plots are attached.

I was working on FF on HAM2 and when I looked at HAM3 for comparison, I saw that the GS-13's have a 23 hz comb. My first thought was some vacuum component, but Kyle says he doesn't have anything running in the area. Near as I can tell it starts at about 9 am local on Friday. Unsure of the cause, but I need to run. I'll look some more tomorrow.

LVEA: Laser Safe
Observation Bit: Commissioning 

07:15 S&K Electric – Running conduit on Y-Arm near end station
08:35 Christina & Karen – Cleaning in LVEA
09:07 Jodi – Going to End-Y and End-X to check property tags
09:10 Bubba – Going to End-X – Will be in VEA and mechanical room
09:43 Richard – Going to End-Y to check on electrical work
10:17 Hugh & Jeff B. – Going to HAM6 for CC closeout tasks
10:23 Jason & Ed – Going to End-X to adjust OpLev laser
10:23 Jim – Going to HAM5 to check for rubbing
10:32 Richard – Back from End-Y
10:52 Jason & Ed – Back from End-X
11:00 Jodi – Back from end stations
11:05 Corey – Going to HAM6 to take photographs
11:12 Corey – Out of LVEA
11:19 Christina & Karen – Cleaning at End-Y
11:24 Hugh & Jeff – Out of LVEA
11:54 Christina & Karen – Cleaning at End-X
12:34 Jim – FF testing on HAM2
12:35 Christina & Karen – finished at End-X
12:36 Hanford alarm testing
13:25 Gerardo – Going to HAM6 to prep for door install
13:43 S&K Electric – Out of Y-Arm beam tube – Electrical work at End-Y
13:45 Door crew at HAM6
14:27 Filiberto – In LVEA near X-Arm to stage for vacuum Beckhoff work
15:06 Brutzman’s on site to finish cubical installation
15:20 S&K Electric – Out of End-Y
15:27 Jeff – Going into CER
15:45 Door crew finished at HAM6
15:57 Kyle – Going to End-X to recover parts and equipment

PSL Status: 
SysStat: All Green, except VB program offline & LRA out of range **
Output power: 32.9w  
Frontend Watch: Green
HPO Watch: Red

PMC:
Locked: 6 days, 2 hours, 22 minutes
Reflected power:    2.6w
Transmitted power: 23.1w 
Total Power:       25.7w 

ISS:
Diffracted power: 5.6%
Last saturation event: 0 days, 4 hours, 44 minutes 

FSS:
Locked: 6 days, 2 hours, 21 minutes
Trans PD: 1.190v

**   The Crystal chiller had a low water alarm. I added 400ml of water.  

SEI:	
HAM6 close out and door install
Looking into some issues with HAM5
Running FF testing on HAM2

SUS:	
ETM charge levels are still good. They will be monitored to see if they return
Investigating 2 BOSEMs on TMS-X which are not responsive 

CDS:	
Working on mods to Beckhoff chassis in corner station

NOTES:	
There will be an Internet access outage on Tuesday from 09:00 to 10:00
Stripe Rite will be on site Tuesday to sweep along the X and Y beam tubes  

While following up some hVeto results, I came across some very strange signal in ISI-HAM4_BLND_GS13Y. It's also in X and Z - I haven't checked other degrees of freedom or the raw GS13 signals. For several days, there's a sweep from ~700 Hz to 1024 (Nyquist) once per second. It doesn't seem to be quite aligned with the second. The first plot is an Omega scan that shows what it looks like.

The second plot shows the onset at June 9 16:49:54 UTC. The only thing I can find in the log is a restart of h1susb123 at 11:50 local (alog). I think this channel is SEI not SUS but maybe they share a front-end machine? The problem looks to be continuous until about June 13 21:30 UTC, third plot. I don't know what fixed it. There's more digging in we can do if needed. This is not a huge problem for detchar, the sudden change in the channel just caused some weird results in hVeto. We probably should be ignoring the high-frequency part of the signal anyway.

I noticed this morning (through the Detchar pages, very sad that last nights SEI data isn't available) that the HAM5 ISI has not been doing well since about the time HAM6 vent activities started. Sadly this casts some doubt on the feedforward data I posted the other day, so I re-evaluated the performance at HAM4 while I was digging into HAM5. My first image shows the difference between HAM's 4 & 5 in the Y direction, all the other DOFs look similar. Second image shows the rotational dof's of HEPI L4C's on the two chambers, and HAM5 sees al lot more motion than HAM4 (dashed lines are HAM5, solid are HAM4). The translations are messier, (third image) but still seem to indicate more motion at HAM5. Hugh reports finding a few mechanical contacts at the chamber while he was working on HAM6, so I'll wait to panic about HAM5 until after that mess is cleaned up. Jeff is filing a report with FRS on this.

On to doing a better evaluation of RY feedforward on HAM'S : the fourth image shows the healthy HAM4 RY feed forward, dashed lines are off, solid are on. Fifth image shows that turning RY FF on has not spoiled the X performance.

Last two images show HAM4's Y and Z feedforward performance, with the same dashed/solid off/on convention.

Filiberto, Patrick

I didn't realized it had been removed, and except for communication errors with the missing chassis, the PLCs ran on without apparent incident.

Filiberto made the same modifications as to the one at end X and reinstalled it. I updated the system manager accordingly, checked it into svn and activated it.

This time I didn't restart the PLCs, but I still had to restart the EPICS IOC. I burtrestored to 6:10 this morning.