Support tubes were picked clean of fibers and bellows were vacuumed. Dust barriers were installed and then support tube protection was installed. Brushing began and all but three sections of the floor were completed. Four drills were used.

Here are some ResourceSpace photos of yesterday's placement of one of the H1 output tube sections.

Mark L. Chris, and Mick worked at HAM3 all day.  Checklist was updated and reviewed with crew. Viton strips were attached to dust barriers so that the barriers were ready for install. Feedthroughs and viewport were removed from the chamber and preserved clean. Initial inspection was completed and documentation photos taken. The support tubes had so many fibers on them that the fibers must be "picked" prior to bellows protection being installed so no "brushing" went on today.

•  One of the H1OMC Tubes was installed.  Looks like they staged the second one for install as well.  
•  HAM3 In-Chamber Cleaning continued
•  Dave and crew continued SUS/SEI Testing work (including a reboot of the DAQ (David Stone went out to EY to power down Actuators before testing).
•  During OMC work, noticed fairly high dust counts (Dust Monitor #1's 0.3um counts hovered between 8,000-15,000 counts over a 75min timeframe.
•  Dale had a school tour in the afternoon.

Alex, Dave, Richard, Jim, Vern, Rolf, Joe

We had a successful day testing the new SUS OSEM software watchdogs on the BSC8 systems. Rolf released advLigoRTS/branch-2.5 which we installed on H2. We built test models for h2tcsl0 (to generate the OSEM signals) h2susb478, h2susb78  and h2seib8 (to drive local DAC channels and verify the watchdog kills the DAC outputs). We modified the IOP models on h2susb478 (ITMY) h2susb78 (FMY) and h2seib8 (BSC8 ISI and HEPI). Representative channels for DACs which are under test were routed back to the ITMY ADC card, such that we could verify the analog voltages coming out of the AI chassis were being zeroed by the watchdogs.

Alex installed the Dolphin system such that h2susb478, h2susb78 and h2seib8 could communicate using PCIe IPC.

By driving the ITMY OSEM signals with the TCS DAC, we were able to test that the h2susb478 IOP zeroed its DAC (DAC0 was tested) and disabled the DAC's anti-imaging chassis (DAC1 was tested. Richard injected a function generated electronic signal into the AI in parallel with the binary control signals to perform this last test.

We tested that both ITMY and FMY can independently kill the DAC channels on the SEI frontend (as well as killing their local DAC channels). We constructed MEDM screens to conduct these tests.

We also tested:

• the bypass mechanism for the watchdogs
• loss of IPC connection causes SEI to fail safe and kill its DACs (if not bypassed)
• the panic button allows the operator to manually trip the watchdogs
• the watchdogs are latched in their tripped condition, requiring manual intervention to untrip them

All systems passed their tests. Tomorrow we will test the EY BSC6 systems.

Transfer function measurements are running tonight on H2 SUS ETMY.  The estimated completion time is ~4:00AM PST.

Today, Gerardo, Travis and I removed the PUM from the ITMy lower suspension and stowed it in it's cake tin and transport case.  We only cleaned the front face since we were using the ergo arm on that surface.  Most of the large glass rubble has shaken off the glass mass, but there is still particulate on it.

Travis then cleaned up the horns on the ITMy in anticipation of gluing metal prisms near them.  We fit checked the new prisms and prism gluing jig this morning.  We had to do some filing on the jig as the clearance was too tight on the prism.  The prisms and jig are now in C&B for processing.  ETA of the prisms is likely next Monday.  Tomorrow we will work on making the ITMy wire loop.

The third cleaning was completed first thing this morning and the "dirty room" was moved away from the chamber. 

The Apollo crew worked in and around HAM3 through-out the day. This morning, both doors and the support table were removed.The support table was preserved clean for later use. Later, the garbing and staging cleanrooms were moved into place on the west side of the chamber and loaded with all the appropriate accoutrements. Hoses were pulled to the chamber. Tools/tooling were retrieved from CNB. The checklist was updated, reviewed, and changed  to reflect lessons from the H2 experience. Carolyn's documentation station was staged. It looks like things are ready to start in-chamber tomorrow.

Jeff Garcia is out at EY performing some tests on ETMy (B&K Hammer).

Richard, Vern, Jim, Dave

We have started building the test system for the new SUS software watchdog system. We are initially testing using the ITMY/BSC8 system. Richard verified all coil drivers for BSC8 suspensions (ITMY and FMY) are disabled.

Richard cabled the DAC output from TCS to the ITMY ADC system. In addition, the following DACs were connected back to the ITMY ADC0: ITMY DAC0, BSC8-DAC0 (HEPI) and BSC-DAC1 (ISI). Please see drawing D1200460 for details:

https://dcc.ligo.org/cgi-bin/private/DocDB/ShowDocument?.submit=Number&docid=D1200460&version=

Jim and Dave created a test model to run on h2tcsl0 to drive its DAC outputs. The model is called h2cdsswdawg and runs on an unused core on h2tcsl0 as dcu122. This model was tested to output analog voltages from h2tcsl0 DAC0 which are then digitized by h2susb478 ADC. One channels from TCS DAC0 is also fed back into a TCS ADC channel for verification, this was also tested.

We built a new model to replace h2susitmy on h2susb478. This model reads the ADC channels which are being used for the OSEM simulation in these tests, and it can be used to excite ITMY DAC0 channels. This model is called h2cdsswdb478 and has dcuid30 (replaces itmy model). We took out the h2itmyoptlev model for these tests.

All models are built using advLigoRTS/branch/branch-2.5. The DAQ was reconfigured to support these new models.

We made a first attempt at installing the software WD onto h2iopsusb478 but that is still a work in progress.

- LVC meeting is underway in Cambridge, MA. Assorted staff is off-site this week.
- Apollo installing cable trays. Work was not as dusty as expected (very brief dust alarm).
- TerryS and crew cleaning around HAM3.
- Kyle leak-checking by BSC8.
- Bob Rhodes on site working on HVAC.
- CDS guests visiting this week to review code changes.
- HEPI pipe welding at EX.
- CDS people rebooting DAQ throughout the day.
- Rear exterior LSB door is out of order (by Dale's office), Rodney is having a locksmith fix it.
- OSB Doors locked at 16:10 PST.

Flooded O-ring at leak detector inlet with He after check and observed predictable permeation response

Rich Abbott reminded me of the fans in the I/O chasis, so I looked for coherence between a magnetometer (which detects the fan rotation frequency) and channels passing through the A to D in one of the chassis. I found coherence for all channels examined. Figure 1 shows coherence between an OSEM channel and  my magnetometer channel and also shows that I was able to repeatedly move the coherence peak back and forth in frequency by covering/uncovering the fan ports (which changes the rotation frequency). We have had problems with fans in electronics for a long time – I found that the fans in coil drivers showed up in AS_Q, using the fan port covering technique, before S1 (here). My recollection was that, in this previous case, the coupling was through power supply ripple so Richard, Filberto and I tried powering the fans of the aLIGO chassis with a separate supply. Coherence was no longer evident. While a separate supply would mitigate this coupling, I do like Rich’s idea to eliminate all fans with vacuum suction cooling.

Robert S., Rich A., Richard M., Filberto C.

Figure: When there is no port covering, the right hand fan produces the blue peaks at about 570 Hz, the left hand fan produces the blue peaks at about 562 Hz. I adjusted the port covering to put both peaks at about 547 Hz (red). I went back and forth between the two states several times (the multiple red and blue traces).

Damping loops are now implemented on HAM-ISI Unit #2.

Watchdogs are set to their safe value:

CPS limit:            20,000 counts

GS13 Limit:         20,000 counts

Actuator Limit:    20,000 counts

Transfer functions are running over the weekend.

Baseline is too high -> should be low enough on Monday

The majority of staging for re-starting In-Chamber Cleaning at HAM3 is complete. The chamber cleanroom is in place. Both Phase 1 and Phase 2 cleaning are done: this means full garb is required to enter the cleanroom from this point forward. (I posted signs on the east and west side of the cleanroom and sent out an email.)Dust barriers, wrapping table, tooling, etc have been craned over to HAM3. A garbing cleanroom is in place. We will move the staging cleanroom ASAP. Drills have been checked with ~30 running for at least 10 seconds despite the fact that they have been sitting around for months. Paperwork has been generated. 

The ISI was strongly excited on march 1 at its first rigid body mode (1.24Hz). Since L4Cs and GS13s have their resonance at 1Hz and T240 are fragile instruments, they may have been damaged during the intense shaking. The attached calibrated spectra shows that all geophones seem to be working properly. Spectra were measured with the ISI locked and unloaded (lower stages of the QUAD removed). Then motions measured by stage 1 (L4C, T240) and stage 2 (GS13) geophones are very close. CPSs seem to work too.

Jim, Eric, Corey, Hugo

We recently received 5 production-GS13s from LLO: 3 verticals (Pods #: 30, 14, 88) and 2 Horizontals (Pods #: 93, 24). They are intended for HAM-ISI Unit #4.

We huddl-tested them before installation. One of them (Horizontal, pod #24) features the behaviour we have observed on GS13s with whether a broken flexure or a stuck mass. We tried tilting the instrument back and forth to unlock the mass. We could feel the mass moving. The ASD measured after tilting the instrument is the same as before.

J. Garcia, J. Kissel, B. Shapiro

After unlocking both the QUAD and the ISI, here are the latest results from H2 SUS ETMY. In summary, the chains look good and match other equivalent chains reasonably well, save the details below. My only concern is
- There seems to be cross-coupling, most notably in the M0 Transverse direction (but present in other degrees of freedom *and in both chains*), between T/R at 1.25 Hz, and T/R/P 1.6 Hz.

Other things that are not worthy:
- The upper two Vertical mode frequencies are consistently higher than ITMY by ~2% (for main chain, same SUS but with metal erm for reaction chain). BUT, this would mean an equivalent increase in lower blade stiffness of sqrt(2%) = 0.04% -- by which I'm not alarmed, merely interested, since almost all other instances of SUS project wide have had *remarkably* consistent vertical modes.
- Again focusing on the vertical TF -- Brett identified that we had accidentally left the UIM horizontal center of mass parameter, h1, at 0.005 m instead of the nominal production value of 0.000 m for the main chain, monolithic, 'fiber', parameter set. This offset, when present, (among other things, most notably) creates a cross-coupling of Pitch to the Vertical TF @ 2.76Hz. I've corrected this in the production model parameter set. Further, as you'll see in the attached, ITMY has some of this offset, where ETMY does not (good!).

------------

For persepective, one should probably start with allquads_120315_H2SUSETMY_ALL_TFs.pdf
Here, I compare 2 chains, 4 measurements of monolithic main-chains, and 3 measurements of equivalent reaction chains.
Main Chain
H2 SUS ITMY 2011-12-02 -- Pre-cartridge install, Floating ISI, After IAS Adjustments, Stiffening Elements + Vibration Absorbers in Place, L1&L2 Aligned
H2 SUS ITMY 2012-02-16 -- Post-cartridge install, Pre-ISI-Shakedown-Fiber-Failure, Floating ISI
H2 SUS ETMY 2012-02-14 -- Pre-cartridge install, after major intra-chain rubbing was relieved
H2 SUS ETMY 2012-03-14 -- Pre-cartridge install, after further UIM flag rubbing relieved, stiffening elements installed, etc (only thing missing is VA, which shouldn't effect anything).

A quick scan through these TFs show that, for the most part, the most recent ETMY measurement matches the best ITMY measurement. In fact, comparing CYAN with MAGENTA, shows that the most recent round of UIM flag adjustment has greatly cleaned up all DOFs measurements.

The only flaws visible are those mentioned above: 
- Cross-coupling, most notably in the M0 Transverse direction between T/R at 1.25 Hz, and T/R/P 1.6 Hz.
- Slightly higher Vertical stiffness.

Reaction Chain
QUAD04 2011-11-22 -- Phase 1 approved measurement of H2 SUS ETMY, with metal ERM (on a solid stack)
QUAD03 2012-01-24 -- Phase 1 approved measurment, with metal ERM (on solid stack)
H2 SUS ETMY 2012-02-14 -- Pre-cartridge install, after major intra-chain rubbing was relieved
H2 SUS ETMY 2012-03-15 -- Pre-cartridge install, after further UIM flag rubbing relieved, stiffening elements installed, etc (only thing missing is VA, which shouldn't effect anything).

Here, in the most recent ETMY measurement, the reaction chain looks OK at first glance, but upon closer investigation shows the same 1.25Hz and 1.6Hz modes as the main chain, and the mid-frequency band just looks kinda noisier (though not as bad as 2012-02-14).

--------------
The ALL_ZOOMED_TFs.pdf plot just shows a tighter X-axis for a better look at the details, and the first two attachments show a lot more information (cross-coupling, plus OSEM basis) on the CYAN measurements.

Though it's difficult to blame a particular thing (*sigh* isn't that always the case?), my gut tells me there's still something rubbing between the chains. I know we've been having trouble with UIM flags... Also it's difficult to imagine a scenario where Transverse couples to Pitch...

We'll keep thinking.