h1fw0 crashed at 23:05 Tue 5th march. I restarted it at 10:30 today, Wed 6th. The cause of the crash is under investigation.

For the 35W laser.

I replaced the dust monitor in the optics lab (lab location 1) which is indicating a sensor failure. It is labeled 'U'.

After running a week at 125psi, the LVEA HEPI hardlines were vented in preparation for filling with fluid.  The plugs at BSC2 South were fitting with pressure transducers for aLIGO operation.

This report is for yesterday.
North door is removed so chamber is ready for entry.

This report is for yesterday.
Feedthrough/blank install is complete except for one 12" conflat. Hugh is locating an appropriate conflat. We will stand-down on this work to concentrate on BSC2 Cartridge install.

This report is for yesterday.
Dome was removed
Walking plates were wiped down and put into place
Both staging cleanrooms were populated
Chamber cleanroom function was checked (2 fan-filter units were on low rather than high, 1 fan-filter unit was not functioning and was replaced)

Attached are plots of dust counts > .3 microns and > .5 microns in particles per cubic foot requested from 5 PM March 4 to 5 PM March 5. Also attached are plots of the modes to show when they were running/acquiring data.

I have not included plots of the counts in the LVEA. The IOC for the dust monitors in the LVEA was restarted today putting 'nan' (not a number) in the data.

The plot of the status for dust monitor 1 in the optics lab shows when it started indicating a sensor failure. (transition from 0 to 1).

Data was taken from h1nds1.
1440.0 minutes of trend displayed

J. Kissel (after being asked by M. Landry and K. Izumi and contacting S. Biscans, B. Lantz, H. Radkins, G. Grabeel, because V. Lhuillier is on vacation)

Kiwamu was looking to re-align beams from the ALS Table into BSC6, and asked me to restore the chamber to an H2OAT-kind of alignment. I took this as "untrip all the watchdogs, and get damping loops and alignment offsets running." The suspensions were easy, the ISI was not-so-bad, but HPI would not restore -- the process of turning on the isolation loops (which are unfortunately necessary in order to apply the large RZ offset needed to align the ETMY yaw) would trip the HPI WD, with actuators as the trigger, indicating that the error signal was too large.

After some digging, and some question asking, I found the problem arose from the inductive position sensors (IPSs). A trend of the last 30 days (see attached) shows that on Feb 25th, at ~21:30 (1:30p Local) something jolted HPI, most predominantly in the Corner 3, 
V3 = 4500 [ct] * 0.001/655 [in/ct] = 0.0069 [inches] ~ 6.9 [thou] DOWN 
H3 = 3600 [ct] * 0.001/655 [in/ct] = 0.0054 [inches] = 5.4 [thou] -RZ 

with V2 and V4 moving roughly 2 [thou] DOWN, and it has stayed offset as such since. (See G1000125 for indication of location and pointing of BSC6 HPI sensors)

LHO aLOG 5576 indicates that both Apollo (chamber cleaning) and tours (for our future Indian colleagues) were in the end station that day.

A similar trend of the ISI ST1 capacitive position sensors (CPS) shows a very small (~100 [ct] * 0.001/800 [in/ct] 0.00013 [in] = ~ few micron) jolt at the same time, but otherwise continued on unaffected; certainly no large, several [thou] DC shift.

This implies that somehow the IPS were bumped / moved / mis-centered somehow, and that there wasn't really a permanent gross misalignment.

Greg and I went down to the end station and confirmed there was no obvious sources of rubbing or fixed forces (say from the work platform that surrounds it). We found nothing obvious. Unfortunately, though there were dial indicators on Corners 3 and 4, comparison with previous positions yield nonsense, so they were probably moved or bumped into uselessness (with is easy to do). Noteably, we forgot to look at the stops inside the boot, but that should be a quick check when we head back again.

Other large-scale activity has gone on since the signal drop on the 25th, e.g. the clean room move and staging for door removal on 2013-03-01, and is visible but does not result in any such large shift in DC signal offset. 

Since Kiwamu's alignment into the chamber has been delayed slightly waiting for PZT mirror assembly parts, we've (Greg, Hugh, Kiwamu and I) resolved to check the alignment of HEPI with an optical level over the next day or so (with priority given to tomorrow's cartridge install, naturally). If that reveals no such change in alignment, we'll recenter the IPSs and move on.

Today I installed a new router for the DMT system, with the latest version of Vyatta Core - 6.5R1.  It is functionally identical to the old system, but we are trying to standardize on a hw/sw config for these systems in the MSR, and this is the first step.  There should be no visible difference in accessing the DMT machines from off site (unless you tried today while I was in the middle of swapping things around).

The dome popped up smoothly, no issues with the o-rings.  Access is tight coming in through the small overlap.  The handrails are also out making the cleanroom ready for lowering--first task Wednesday AM.

The PSL environment variables went invalid around 13:00. It appears that the h1ecatc1 Beckhoff computer rebooted itself. I do not know why. It restarted itself again when I was navigating through the folders. I uninstalled OpenOffice.org 3.3, Java 7 update and Adobe Flash Player 11 ActiveX. I managed to restart the IOCs after it restarted again.

The dust monitors in the LVEA all lost communication. Restarting the IOC seemed to bring them back. I burtrestored to 2013/03/04/00:00. Dust monitor 1 in the OSB optics lab is indicating a sensor failure. Dust monitor 4 in the LVEA had a low battery status again, but it has now gone away.

Mark B. added IPC inputs for ISI->SUS to the H1 SUS MC2 model. H1SUSH34 was rebooted.

Lots of fun spending most of the day contorted in tight places.  The work platform stabilizers prevent them from being in final place.  Will be revisited.

Following power spectra taken on friday february 27th (see LHO aLOG entry 5596), transfer functions have been ran over the week end for PRM phase 3a testing.

The two attached pdf show comparison between hsts model, LLO Phase 2b (March 2013) LHO Phase 2b (March 2013) and LHO Phase 3a (March 2013) respectively with suspension damped and undamped.

Transfer functions are nice and smooth, showing peaks at the expected frequencies.

Phase 3a testing is therefore complete

Kyle valved out the Y1 Module for a few hours today. Tomorrow  we will valve out the Y2 module.

Attached is the rate of rise for the two end vacuum gauges.

If I have done the math correctly this translates to a very low outgassing rate with no significant leaks :  1.3 e-6 tl/sec for the 2km module. If it is all H2 this yields ~2e-14 tl/s/cm2

h1susauxh56 was rebooted to investigate the problem of multiple awgtpman processes running. Only h1iopsusauxh56 was configured to start, two versions of its awgtpman process was found to be running. This problem will be resolved later, for now we need to be vigilant on FE reboots.

The problem of the SUS QUAD models running long was tracked down to the awg process by Alex. We tested this on h1susb6 by running three modified awgtpman processes. We reverted the system back to the original code at the end of this test, so h1susetmy cpu is running long again.

The three new IOP models h1iopsush56, h1iopsusauxh56 and h1iopseih45 were started now that the GDS 64 awg limit was removed by Jim earlier.

I've restored (both sensor and alignment) calibration on all mode cleaner HSTSs, in hopes to get the IO team back to where they started. With the alignment offset calibrations in place (i.e. 1.8751 [ct/urad] in P, and 2.6808 [ct/urad] in Y -- see LHO aLOG 4763), the DRIVEALIGN offsets are

        [ct]           [urad] 
MC1 P  -360.9 / 1.8751 = -192
    Y   420.0 / 2.6806 =  156.68 

MC2 P    -9.8 / 1.8751 =   -5.22
    Y   367.9 / 2.6806 =  137.2

MC3 P     710 / 1.8751 =  378.6
    Y     220 / 2.6806 =   82.07 

(where original "good" offset numbers are from LHO aLOG 5631)
I've also ensured that damping was on for all three suspensions.

NOTE: I HAVE NOT YET TAKEN A NEW SAFE.SNAP. This requires turning off the offsets and damping loops, such that there's no drive going to the suspensions. Since Cheryl wants to get going, I'll just capture later in the night when everyone's gone.

Stuart had flagged that the h1susmc2 model did not have the proper ISI->SUS IPC inputs H1:ISI-HAM3_2_SUS_GS13_* (*=X,Y,RZ,Z,RX,RY), so I added them. 

The h1suspr2 model had already been receiving these signals from h1isiham3, so there was no need to add them on the ISI side. 

EPICS error channels H1:SUS-MC2_ISI2SUS_GS13_*_IPCERR had previously been defined, so there was no need to add them, just connect them up, which meant no need to restart the DAQ.

Following my previous log on PR3 power spectra, showing suspicious resonnances at 2.7 Hz for several degrees of freedom, Travis went in chamber and found a blade tip EQ stop touching.
A new set has been ran, and look very good, as what we expect.
To sum up phase 3a power spectra, the attached pdf shows comparison between LLO phase 3b (Sept 2012), LHO Phase 2b (Feb 2013), LHO phase 3a (before moving the EQS March 2013) and LHO Phase 3a (after moving the EQS March 2013)

data and plotallhlts_spectra have been commited on the svn