Detected a large shift this morning on the IPS readings for the ITMX HEPI.  Turns out the Pump Station Controller Servo had died this morning just before 0800pst.  This would have affected all corner station HEPIs.  The pressure data stayed locked at 80psi even though it was very obvious we lost pressure.  The servo must have stopped pushing motor speed and it just went to zero.  I'll see if Dave has any suggestions.  We'd have gotten an invalid data alarm (if we had alarms on this, like we should) and maybe that is good enough.

There was construction activity up on the Mezzinine with the TCS chillers this morning; possibly related but it certainly isn't the first time work has been going on up there.  Anyway, it has been sometime since this or any of the HEPI PS controllers have hung up.  We found the computer running but unresponsive to local or remote I/O.  So we power cycled the DC power supply and once it responded :

ssh controls@h1hpipumpctrll0  (password is not standard)

cd target-new

sudo ./run

This restarted things in a safe non-running mode.  The motor speed was manually ramped up to get near 80 psi on the supply pressure and the servo was then engaged with no problem.

Looking at the ITMX Inductive Position Sensors we see that the system did not return exactly to where it started but that is expected.  But the changes are pretty small with the largest local shift less than 1 mil and the rest all being less than 1/2mil (.0005")

Sheila, Keita, Alexa

Temperature changes in end X (from turning on or off cleanrooms) can mislaign the refl diode.  The behavoir in the X end seems similar to what was seen at the Y end in July (alog 7146).

After Alexa realinged the TMSX yesterday (alog 8987) and turned the cleanroom on again, there was a temperature change by about 0.5 deg C in 202B.  as the temperature settled you can see the drfit in the ALS input PZT sensors,  the retro-reflected beam started to fall off the diode, and there were drifts  in HEPI sensors, and bosems.  (attached plots)  I fixed the calibration for the PZT sensors (they are urad now) just before Alexa did her realignment, that is why there are no earlier trends that are usefull. 

I recovered the beam on the diode by moving TMSX just now (11:50 UTC), I moved it by +4urad in YAW.  The current alingment offsets are 13urad PIT, -319 urad Yaw. 

HEPI moved less than 1urad in RZ, not enough to explain the shift we saw.  Fabrice said that this amount of motion is not surprising, but that it would be reduced once position loops are commisioned. The TMS bosems do indicate a change of 4urad in yaw (I'm not sure if there is a relative minus sign or not)  Jeff suggested I look at the ISI CPS, it seems like the changes due to the temperature chage are less than 100 nrad in RZ and RY, for both stage 1 and stage2. 

Conclusion:

Temperature changes of less than 1 degree C can shift the TMS alingment by 4urad, this is enough to misaling the ALS refl diode.  The last plot attached is a one year trend of the temperature at end X, temperature shifts by about a degree C are not unusual, so we may need to find a long term solution for alignment onto this diode.  The short term solution is just to leave the cleanroom off. 

I will leave things alone for a few more hours to make sure the alignment stays stable after my realingment, then turn the cleanroom off again tonight. 

The 2 Thorlabs photodetectors on the PSL table monitoring the DC power after the EOM (IO_AB_PD3) and after the rotation stage (IO_AB_PD2) had their readback cables connected to the EtherCAT auxiliary signals concentartor 3 in rack ISC-R4. IO_AB_PD2 is connected to Extra1 (rear) and IO_AB_PD3 is connected to Extra2 (rear). Slow controls channels have been added to H1ECATC1 PLC1. They are available as H1:PSL-EOM_A_DC and H1:PSL-PERISCOPE_A_DC.

This is just a note of which camera is directed to what:

We've turn the two test mass Capacitive Position Sensors off to investigate the comb feature seen at LLO and possibly starting to reveal itself at LHO.  The thought is that all the CPSs have their own master timing for each chamber and the individual chambers may be beating against one another.

We'll restore these in a couple hours.

[Sheila, Kiwamu]

Looking at the ITMX camera, Sheila found a good alignment of PR3 which centers the beam on ITMX. The below is the slider settings:

PR3 (to get the beam centered on ITMX)

• YAW bias = 758 counts
• PIT bias = -1058.5 counts

It seems that this alignment also kind of centers the beam on ITMY without touching BS. Indeed the biases look similar to what Keita had when he got it centered on ITMY (see alog 8978). However, with this alignment, we were unable to get the PRX fringes even with ITMX steered. To study what is happening, I looked at the PR3 camera which is an analog camera (see alog 8937) attached on the MC spool. Then I realized that the ITMX return beam was hitting the swiss cheese baffle. This was the beam propagating from PR2 to PRM. I tried to steer it with ITMX, but the actuation range was not big enough and also the spot tended to disappear as I steered ITMX. So instead, I steered PR3 to let the beam hit the PRM.

As a result, currently multiple bounces is happening in the PRX cavity. A clear beam spot is visible on the BS baffle of its HR side and the spot is sensitive to both PRM and ITMX, indicating that this is a multiple-bounces beam. Note that I didn't change the ITMX biases so that both pitch and yaw sliders are zero. Apparently the horizontal alignment is not good because I can see the beam is slightly hitting the BS baffle on its right side. I am guessing that we need to re-tweak PR2 horizontally to clear BS.

PR3 (to get multiple bounces in the PRX cavity)

• Y = -107 counts
• P = -312 counts

Blue bolts bad - see picture.  Swap these out if there is an opertunity.

I reworked the IO_AB_* path on the PSL. This incuded a small change to the layout to accomodate a 2" mirror mount. The beam is routed to IO AB_PD2, and the PD is powered.  IO_AB_PD3 is not yet on the table.  IO_AB_PD1 is on the table but not hooked up.

Moved one inch away from the periscope foot:

IO_AB_M8

IO_AB_BS2

IO_AB_M9 - also moved one inch toward the center of the table to keep the path length from the lens 15"

IO_AB_PD - also moved one inch toward the center of the table

IO_AB_M10

IO_AB_CCD1 - not yet on the table

Need for DC PDs:

- power cable for IO_AB_PD3

- BNC cable for IO_AB_PD3

- connection to DAC for PD2 and PD3

Need for camera:

- camera - didn't find a WATEC

- power cable 

- BNC

Other to do:

- swap optic from manual rotation stage to powered rotation stage for IO_MB_HWP2

- install and test the rotation stage in the main beam

What has been done so far :

- Filters were installed in the lockin demodulator and were tuned for a 10Hz signal excitation.

        *9:11Hz bandpass for the optical lever signal

        *1Hz ellip low pass and 20Hz notch for the demodulated I and Q signals

- Lock2Osem matrix filled to excite the pringle mode of the second mass from the pitch oscillator (UL, LR = 1 LL, UR=-1)

- Sent a 10Hz, 50000cts signal from the oscillator, to excite the pringle mode of M2, but I couldn't see the demod Q and I changing when modifying the coil gains, so I quit after few iterations

- M2 length to M3 Pitch transfer functions, without coil balancing. It matches the model below 2Hz, cf attachement. The signal/noise ratio of the tf is too low above that. I had to divide the model supposed to be in rad/N.m by 100 to match with the op lev TF, and I'm not sure where this factor comes from. The red curve is a fit of the modeled tf, and the cyan is the inverted tf.

What I intented to do this evening is to install this inverted filter in the L2P M2 drivealign matrix, take a new L2P TF and compare it with the previous one, to see the improvement., but didn't have time to finish.

If anyone wants to take over : 

Script live in /SusSVN/trunk/BSFM/H1/BS/Common/Scripts/L2Pdecouplingfiltersdesign.m

TF data is in /ligo/svncommon/SusSVN/sus/trunk/BSFM/H1/BS/SAGM3/Data

Hugh, Mitch, Jim

After TMS removed their lower half this morning, SEI went to EY to remove our mass stacks from the optical table. Relatively painless, we were done by 3pm. Our parts are currently stored in foil the AOS cleanroom over the gate valve. NEXT!

Today, Travis and rerouted the SUS testing cables from the SRM satellite boxes to the SR2 satellite boxes and cleaned up all the riff raff cabling left over from OMC work in the area.  We labeled the cables and the SR2 sat boxes.  Note, currently SRM is still plugged into the table end of the cables so we need to just move them the the SR2 suspension on the table.  I also moved the SR2 optic (in a special holder) to the table so be aware.

Next up:  Level and suspend the SR2 all metal sus and pull back all of the OSEMs in order to take open light gain/offsets.

900 Corey doing TMS work at both EX and EY

1000 RCG upgrades begin

1020-1600 Cheryl working in PSL enclosure

1040 Fil and Aaron doing cable survey at ISCH1TL

1120 Townsend onsite

1140 Jim/Cyrus working MX

1240 Praxair delivery

Today, TMS removed the lower TMS structure.

SEI shuffled weights and cleared the table of the H2 weights.

SUS located the ETMy SUS cookie cutters and assisted in the hunt for the BSC10 TMSy cookie cutter.  We'll plan to reposition the SUS QUAD upper structure on the table tomorrow unless the TMS cookie cutter pops up and TMS wants to press on.

After the rcg update and front end restart, hepi, isi and suspension systems have been restored back to their automatically generated snap from 8:00 am this morning.

I used the following commands :

for optic in mc1 mc2 mc3 prm pr2 pr3 itmx itmy bs im htts etmx etmy srm tmsx
do
burtrestore h1sus${optic}epics 08:00
done

for isi in ham2 ham3 bs itmy
do
burtrestore h1isi${isi}epics 08:00
done

for hpi in ham2 ham3 bs itmy
do
burtrestore h1hpi${hpi}epics 08:00
done

We finished welding ETMY on Friday evening. It was hanging with about 5 or 6 mrad of differential pitch wrt the PUM. It was left hanging over the weekend. When we looked at it this morning, two fibres had broken and the other two were lying against one of the earthquake stop rails. All the horns and stock were still intact. We scribed off the stock from all 8 horns and filed them back to 6mm above the ears. Work is ongoing to clean up the remains of the two shattered fibres and to pull enough fibres to allow welding again first thing in the New Year.
Angus, Travis & Betsy

Yesterday we adjusted WFS1 path.

From some calculations by myself and Lisa Barsotti plus an assumption that MC mode cannot be THAT bad, I concluded that there isn't that much of a difference between moving everything to optimize VS moving only RM2.

Detailed analysis will follow shortly, but it seems like we're at not too bad, not too good position and if that turns out to be the case we'll stop here.

For example X waist location is already perfect, the waist size is smaller than ideal and the Gouy separation seems to be 114 deg or some such. Y should be worse.

Before calling it a day, we looked at RMs and one of the eddy current dampers for RM2 was significantly closer to the mirror holder than everything else, so it was backed off.