The H2 FMY front end computer (h2susb78) sent IPC Dolphin error glitches which IOP tripped the SEI for BSC8 in several separate occasions between 9am and 09:30 this morning local time. At 10:10 the front end locked out, no ping, no ssh, console stuck at last log message. All lights looked good on the front end, but the only recovery was to power cycle the computer. This in turn crashed h2susb478 (ITMY front end) because of the Dolphin drop out. We took the opportunity to recompile the SUS ITMY model since it had not been built since Apr 11th and was missing the DACKILL part (RCG branch 2.5 build).

I took the opportunity to rebuild the temporary optlev model to make it generic. A H2 DAQ restart was also performed, the master file was changed and the OVERVIEW medm regenerated.

The cause of h2susb78 problems is not known, but there is some work progressing around BSC8 at the times in question.

HEPI:
Last week, the L4C gain of the pier modules were increased from 22 to 333. The transfer functions measured over the week end showed that this new gain is appropriate. No saturations on the L4Cs during the measurements.
I will start to work on controllers for the HEPI pretty soon.

ISI:
This morning, I fixed the isi2stagemaster.mdl and recompiled ISI-BSC6 and ISI-BSC8 due to an error mentioned in a previous aLOG.
I installed the new aLIGO BSC-ISI matrices freshly computed. I posted them in the SVN at /BSC-ISI/Common/Basis_Change_BSC_ISI/aLIGO/Matrices/Y/aLIGO_BSC_ISI_Matrices_Direction_Y.mat
I installed using one measurement:
- Symmetrization filters (actuators, geophones)
- Damping filters (stage 1 and 2)
- Blend filters (basic filters with a blend at 750mHz).
- Isolation filters. I haven't spent time on tuning. The goal is just to make sure the scripts and the model are ready. (UUG=15Hz, comfortable gain and phase margin).

I have attached spectra of the ISI undamped and controlled:

• LHO_ISI_BSC8_ASD_m_CART_ST1_L4C_Undamped_vs_Isolated_2012_05_29_172914.fig
• LHO_ISI_BSC8_ASD_m_CART_ST1_L4C_Undamped_vs_Isolated_2012_05_29_172914.pdf
• LHO_ISI_BSC8_ASD_m_CART_ST2_GS13_Undamped_vs_Isolated_2012_05_29_172914.fig
• LHO_ISI_BSC8_ASD_m_CART_ST2_GS13_Undamped_vs_Isolated_2012_05_29_172914.pdf

Every hour on the hour a copy of the INI PAR and FILTER files are copied to the LHO CDS web server. The URLs are:

https://lhocds.ligo-wa.caltech.edu/exports/running_config/h1/daqfiles/

https://lhocds.ligo-wa.caltech.edu/exports/running_config/h1/filterfiles/

https://lhocds.ligo-wa.caltech.edu/exports/running_config/h2/daqfiles/

https://lhocds.ligo-wa.caltech.edu/exports/running_config/h2/filterfiles/

Attached are plots of dust counts > .5 microns.

Jason, Travis and Doug,
Jason measured the ETMy pitch just prior to the First Contact removal from the H2 ETMy optic.
the pitch was 629urads down (req is 639 urads down)

After the first Contact was removed it change to 379 urads down (a change of 250urads)

This is near what was predicted for a 1.74g weight First Contact coating.

Note: This was not accounted for on the H2 ITMy or measured.

We should possibly measure this for each optic or compensate for it in the calculations as we will also be adding additional weight with the new plan for using First Contact which will remain in place until just before we close up.

Vern, Hugo,

During the chamber side testing HAM-ISIs are locked. Part of the testing consists in measuring the DC voltages outputted by the CPSs to make sure that the probes and the electronics are still functional. This measurement is taken at the BNC output of the ADE boxes. 

We recently observed that the voltages we measured depended on the number of probes that were connected to the ADE Boxes. We also noticed that the voltages measured with only one probe connected to the ADE boxes were high in comparison to what we expect on a locked ISI.

Vern and I investigated it this afternoon: 

• The CPSs have a driven shield. This driven shield is exposed on the BNC connector of the in-air CPS cable (feedthrough side). Accidentally grounding this shield to the test stand would significantly affect measurements. To avoid such situation, we adjusted the position of the feedthrough to make sure that this driven-shield could not get in contact with the test stand. 
• We bolted the ADE boxes together to make sure that they share the same ground.
• We grounded the ADE boxes to the test stand.

The voltage measured at the BNC output of the ADE boxes would then match our expectations (0V +/- 4V), without varying with the number of probes that were connected.

We tried disconnecting the grounding cable between the test stand and the ADE boxes. The DC voltage read on the ADE boxes increased and became dependent on the number of probes that were connected, again. This latest observation confirms that the issues we experienced on the CPSs were related to the grounding of the ADE boxes.

conclusion:

• ADE boxes need to share the same ground. 
• ADE boxes must be grounded to the test stand of the ISI. 
• Not having the ADE boxes grounded to the test stand cause the CPS readouts to vary with the number of probes that are connected to the ADE boxes.

Voltages and noise spectra were re-measured for each CPS and HAM-ISI Unit #4 (HAM2) chamber side testing report is being updated.

Pictures are attached

Kyle, Gerardo 

Summary ->  
Looking for large leaks -> We found a total of (7) leaking conflat flanges that will need to be addressed (>10-6 torr*L/sec typical).  In all cases, these were the large 10" and 12" joints and not the smaller feed-through joints.  We wrapped the leaking joints in RED UltraTape for indentification.  Also, both O-rings of the septum between HAM3 and BSC2 are leaking (inner high 10-4 torr*L/sec, outer mid 10-5 torr*L/sec), these are the O-rings on the BSC2 side which are integral to the septum plate.  


Kyle 

Vented H1 input MC

We took down h1boot for about 3 hours from 10am to 12:20 during this morning's maintenance in order to clone its boot drive. Michael unlocked the H1 PSL prior to the shutdown, and relocked it after h1boot was restored.

The cloned drive will be used to create x1boot as an identical copy of the production H1 and H2 systems.

A new backup VLAN network cable was installed to h1boot, and the switch reconfigurations were submitted to SVN.

The code which reads the data from the FMCS web server and writes it to EPICS has been moved over to new code. I had tried this earlier, but it appeared to have some bugs, which these changes address. The same things noted here apply.

First vacuum and wipe down of BSC2 were completed on Friday, 25 May. The dome was also documented and cleaned. Second vacuum of BSC2 was started on Friday and completed this morning. Final inspection was finished before lunch. FTIR samples were taken right after lunch and then dust barriers were removed. One dust barrier was wiped down and moved to the transition area between BSC2 and HAM4. The dome was replaced on BSC2 and torqued down. 

J. Garcia, J. Kissel

Jeff launched a set of TFs on ETMY before heading out on Friday, just after Team TMS finished up in chamber, and Travis  had set ETMY's EQ Stops to pump-down ready ~0.75 mm distance to confirm that the SUS was still free. Attached are the results. Everything looks just as it did before, matches up with other SUS's up its type, and agrees well with the model. Things are looking good, couldn't ask fo mo betta!

Attached are plots of dust counts > .5 microns.

For posterity, I wanted to show how the picomotor signal from the driver box is broken up into the individual DOF signals.  In the pictures attached, the signal from the driver box goes to the Picomotor Break-out Feedthrough Panel [image 1].  This box takes the DB25 cable and splits it into two DB9 cables [2,3].  These DB9 cables go to a Picomotor Breakout Box, which rests inside the table on the cable tray.  This box has a PCB which splits the signals on the DB9 cable into four 4-pin RJ45 connectors, which are the pitch and yaw signals for two different picomotor assemblies [4,5].

- Travis visited the Y-End station, EQ stops and FC removal of ERM, see his entry for more.
- Jason at Y-End station, alignment check up of ETMY, see his entry for more.
- TMS team was at Y-End station, see Keita's entry for more.
- Filiberto visited Y-End station to check on reports of odd smell, no smell found in or out of the building.
- Kyle checking for leaks in the H1 input tube.

The various activities at Y-End generated some dust, counts up to 1500.

This was the first attempt after the viewport was rotated back, and our alignment was still at the same bad position and the beam was not hitting the ETM. So apparently something moved, but we couldn't figure out what.

Anyway, we steered the beam back by adjusting the ALS table periscope mirrors, using the green QPDs on the TMS ISC table. then steering the TMS itself so the beam retro reflects back to the ALS table through an aperture we set up on the ALS table. We didn't use the picomotors as we knew that the alignment was pretty good two weeks ago and we were sure that that the picomotors were not moved since then.

After some tweaking, the beam looked really good at the secondary, F2, and the primary (Cheryl took pictures). It was low on F1 (a bigger of the two folding mirrors on the TMS telescope), but in the end I think it's more important to have the secondary and the primary reasonably well centered than the folding mirrors, and the beam retro reflects, so I think we're in a good shape.

The new offsets: 16000 counts for PIT, -5100 counts for YAW.

Before we started, Cheryl noticed that the EQ restraint bar is not centered any more (it used to be reasonably centered). We readjusted the bar position.

We also relieved that crazy PZT offset. It was on YAW of the upstream PZT mirror. Since the PZT mirrors themselves don't have any fine manual adjustment, we used the closest steering mirror. The input to the MCL PZT controllers were terminated. This made a tiny change in the alignment (which we were able to measure by QPDs) but these were taken care of manually by the periscope mirrors later. 

The diff-single converter on the field rack as well as the anti-imaging chassis on the ISC rack was turned off.

The report for HAM-ISI Unit #5 - Assembly Validation is now posted under the DCC for validation.

This unit was tested at the begining of the month with testing horizontal GS13s (alog #2802). These testing pods were replaced with production pods. All tests involving GS13s were performed again this week. This report is now waiting for approval. This Unit will then be our second HAM-ISI coming out of the staging building with a full set of production GS13s (the first one was HAM-ISI Unit #4).

Reports regading the Assembly validation of previous units at LHO are also available on the DCC:

HAM-ISI Unit #1 - Assembly Validation
HAM-ISI Unit #2 - Assembly Validation
HAM-ISI Unit #3 - Assembly Validation
HAM-ISI Unit #4 - Assembly Validation