Kyle reported on Aug 12 that he thought there was a correlation between his RGA signal and locking attempts.

https://alog.ligo-wa.caltech.edu/aLOG/index.php?callRep=29079

On July 19 there was a programming change to the FPGAs;

https://alog.ligo-wa.caltech.edu/aLOG/index.php?callRep=28535

During this process there was a pressure burst in PT410 at the YEND. Dave  believes that as the system was brought back, dacs railed while connected to the ISI. This caused all actuators in the chamber to go to their maximum output. We believe this caused some heating which resulted in the pressure bump.

Perhaps Kyle saw pressure changes related to ISI activity?

Dont forget to get work permits in today for work tomorrow. Let's try to make tomorrow a short maintenance day so comissioners can get to work early.

SEI - HAM6 will trip with often the shutter now. HAM6 now runs in Damped for the beginning of locking then then switched later but is a WIP. SDF will be red.

    CS HEPI pump maintenance tomorrow.

SUS - HAM6 sus look good.

CDS - ISS outer loop work, FRS work.

    New version of GDS code (DTT updates).

PSL - PSL seems good. TCS has CO2 laser swap planned (Alastair on his way). BS OpLev investigation.

Vac - Fire up pump to bake out RGA at vertex, Yend worked well. A couple of days of pumping for HAM6

Facilites - LED lights along walkway need to be coiled up (broken?)

Ran through work permits.

Julia Kruck and I buried the high-sensitivity LEMI magnetometers this week near the vault.  This is at the approximate location for which trial spectra were taken (https://alog.ligo-wa.caltech.edu/aLOG/index.php?callRep=12525), and determined to be good enough to study subtraction of the effects of globally correlated fields due to, for example, Schumann resonances.

We buried them directly in the soil at a depth of about 18 inches. We used a bubble level to ensure that they were not tilted in Z, and aligned them with beam tube X and Y using lines constructed and measured on Google satellite images. We think that this alignment technique was more accurate than could be made with the simple compass that we had, though we made sure that the compass agreed with our alignment.  The connector to the X-axis magnetometer is 17m in the -X-direction from the  +Y corner of the electronics vault, and the magnetometer extends in the -X direction beyond the connector (see photos). The Y-axis magnetometer has its connector 1.22 meters +X and 1.22 m -Y of the connector for the X-axis magnetometer, so the closest parts of the magnetometers, the connectors, are 1.72 m apart. The connector  is on the +X or +Y end of the magnetometers so that the signals will have the proper phase relationship. 1.22 meters of cable is buried directly in the ground before the point where the cables come together to enter the conduit (yet to be installed) in order to damp any cable vibrations. The LEMIs lie between the lines of orange stakes that parallel them to either side (see photos). The electronics had not arrived so they have not yet been tested.

Absolute location: X-axis magnetometer  is 1020 m +X,  188 m +Y of the vertex

Robert Schofield, Julia Kruck

Evan, Sheila

This morning we had truoble with DRMI ASC.  THere are a few things that seem to have helped (in addition to intal alingment and some by hand alignment), and how the guardain is reliably getting to analog CARM. 

• The lock point of the SR3 cage servo was changed friday, changing it back to 931 (M3_WIT_P) and redoing SRC alignment seemed to help the SRC and MICH loops.
• Since the optimal yaw alingment of PRM seems to have changed quite a bit, I added an offset of 0.24 to POPA Yaw.  I also took out the lines in DOWN that were resetting this to 0 for each lock.
• Adjusting the current to the BS oplev laser reduced the glitching.
• The MICH ASC loops are now ramped on more slowly
• Evan increased the matrix element for REFL 135 Q to MICH by a factor of 2 (to 2.5) so that the MICH length ugf is around 10 Hz even when we are locked on 3F.
• This morning the mode cleaner was not relocking, and I was clearing the WFS history almost every lockloss.  This problem seems to have gone away for now.

Currently we are in a similiar situation as the last two nights, we can lock the interferometer but loose lock within a few minutes.  We have lost lock even while sitting on analog carm, we have been able to switch PRCL and MICH to POP sensors but suddenly lost lock about a minute latter. 

There are a few things that still needed worktonight with the fast shutter, and will need more work:

• It trips the HAM6 ISI often. According to the detchar sumary pages I have tripped the ISI 18 times in the 3 hours I've been locking tonight (it trips once when the shutter triggers and often a second time while reisolating)
• I lowered H1:SYS-MOTION_C_SHUTTER_G_TRIGGER_LOW from 0.014 to 0.1 because the power was frequenctly dropping below this in DRMI
• I commented out two of the checks on teh AS WFS DC power that Stefan added, they are both checks that there is light on the diodes after the shutter test has finished and the shutter is reopened.  These test were failing because the DC centering loops are turned off for the shutter test.  Since the beckhoff code already check that there is light on AS_C after the shutter is open again, I don't see any safety reason we need to check that power is restored on the WFS. 
• I also moved the check that there is no light on the WFS once the shutter is triggered so that it is only done once in the main state.  As originally written it made sure that there was no light on the diodes for 8.5 seconds after the test started, but this failed because the shutter was reopened faster than the guardian expected.  It seems like checking once should be good enough.
• I changed the threshold on the LOCKLOSS_SHUTTER_CHECK guardians so that it should request a check of the shutter each time we lose lock from an on resonance interferometer with 2 or more Watts in the arms.  We may need to adjust this further, but for now I have added a check in down so that it will not attempt to relock until someone manually changes the state of the LOCKLOSS_SHUTTER_CHECK guardian.  The hope is that this will serve as a reminder for everyone to do the check until we have an automated check in place. 
• There were a few bugs in the way the shutter testing was handled in ISC_LOCK, which I think I have fixed all of now. 

A few other things:

The ISS diffracted power has been at around 4% since August 2nd (the first day of the vent). Before the vent it was around 16%, but DIAG_MAIN thinks the diffracted power should be between 1 and 6%.  I moved it up to about 14%, so DIAG main is complaining. 

There is a problem with switching DRMI to POP signals in full lock.  We can stay locked stably for ~10 minutes at ANALOG CARM, but lose it shortly after transitioning DRMI to POP signals each time. Before transitioning off the 3F signals, the MICH UGF is about 5 Hz.  I manually changed the digital gain from 1.4 to 3.6 to restore the loop to a 10 Hz ugf, although I don't know why the gain would be low.  The PRCL and SRCL UGFs were normal.  After adjusting the MICH UGF I was able to transition to POP, but did this in the middle of an earthquake.  I don't know if the lockloss about a minute latter was due to the DRMI problem or due to the earthquake. 

Our DRMI lock times have been a little long in the last 2 days, it seems like the beam splitter optical lever is glitching.  The attached screenshot shows a PRMI lock where the BS oplev damping is on for the first minute, you can glitches in both POP90 and POP18, when the oplev damping gain is ramped to 0 the glitches go away although there are some oscillations in the build up.  Turning the oplev back on brings back the glitches. 

Sheila, Stefan, and Koji worked on relocking.

23:37 UTC Kyle to LVEA to move two carts from high bay to near HAM5, connect pump cart to vertex RGA
23:44 UTC Kyle done moving carts, starting connecting pump cart to vertex RGA
00:49 Kyle out of LVEA for the night

The astronomy club has set up telescopes for public viewing.

I am leaving the ISC_LOCK guardian in down.

Today we have had some difficulties with alignment in the process of recovering the IFO. 

This morning we moved ETMX by 20 urad in yaw, Patrick checked trends and the top mass osems roughly agree with the alingment sliders that we moved by about 20 urad, while the optical lever indicates that this was moving the test mass back to the position it was in last time we were locking, on August 1st. 

We also saw that even after doing initial alingment twice, the arms atill had a large relative misliagnment which caused a low recycling gain and therefore a lower cam offset than expected while CARM is locked  which cause random locklosses at different points.  Earlier tonight we we seeing a 5% drop in the refl power when the arm transmission was at 5 times the single arm power, while the same arm transmitted power level on August 1st coresponded to only a 1% drop in reflected power.  Koji and Stefan worked on solving this problem, so I will leave it to them to log their solution. 

After decoupling the pumping components used during the recent bake out of the Y-end RGA, I exposed the RGA to the Y-end vacuum volume, energized the filament and let it come into equilibrium for an hour or more.  I then let the RGA scan continuously with the multiplier (SEM) on for an additional hour or so while I gathered up my mess(es).  I periodically checked the scanning as I walked past the screen.  At one point, I noticed that the spectrum was changing rapidly towards the "dirty".  I monitored the scanning and noted that after reaching a temporary maximum, the amus which had increased then returned to near their original values.  

After consulting with the Jeff B. (the operator on shift), I feel that the observed changes in partial pressures were likely the result of IFO locking attempts as they coincide closely in time.  Perhaps something gets hot when the IFO is locked or when mirrors are steered?  

See attached scans 

Using the calibration of alog 28133 (3.0554e+04 cts / Watt from SRM, or 3.273e-05Watt/ct) I added FM10 filters to the segments of ASC-AS_C to cast the output into Watt entering HAM6. I also updated H1:ASC-AS_C_SUM_OFFSET (which avoid astronomic PIT and YAW values).

SDF is updated.

The attached is just one instance and is fairly representative of the trips.  Most of the time, the ACTuator is called to blame, but as I reported in 29056, I'm not sure the Last Trip is absolutely robust.

The 2 seconds of full data show the ISI in Full Isolation (GRD STATE) when Stefan throws the shutter (FASTSHUTTER A MODE.)  You can see a fraction of a second later, the Actuator MASTER DRIVEs going very large.  All the actuators are contributing to the saturation counting as the threshold is only 32760 counts.  The watch dog trips near the end of the two seconds (WD MON STATE.)  It was going to happen anyway but I can't be sure why the WD tripped because none of the counters (iii SAT COUNT) actually reached the trip threshold.  This is a mystery to me, still.  The SAFE COUNT is 8192, 4096, and 10 for the ACT, GS13, and CPS.  None of the counters reach the SAFE COUNT level....  Regardless, the Actuators where going to reach it so that is another problem to figure out.

Unless the shutter softens its blow on the ISI, the trips are going to happen almost everytime.  For now, we are running with HAM6 in the DAMPED state.  There the HAM6 sees a few saturations on the GS13s when the Shutter triggers but literally just a few.  The HAM6 will be transitioned to ISOLATED by the ISC after the FAST SHUTTER is tested.

In an effort to eliminate the need to connect and disconnect shutter cables during vent we have moved the shutter logic controler to the rack from the table.  Also replaced the BNC cables with Yellow BNC cables so they can be easily identified.
See attachment.