(Borja)

Several issues has not allowed me to drive the ETMY ESD until late afternoon today. At this point I was able for the first time to test, with real data, the automation code for the ESD charge measurements develped at Livingston. I did have previously adapted it for Hanford's slightly different configuration but this was the first time I was able to test its results. Unfortunately the automation on the injection, data request and analysis is not robust, not allowing for the whole process to finish several times. Also the code does not take into consideration conversion factors on the V BIAS from Voltage to counts and viceversa. I assume this is taken care in Livingston outside of the code but certainly that solution does not make it universal.

Looking at the procedure with Rai I realized that we have to be careful on the level of the driving signal amplitude to be below the minimum V BIAS used in the analysis otherwise linear approximation assumptions in the methodology are no longer valid.

Rai is leaving on Saturday and we need to apply his discharging technique before then (optimally tomorrow). Before this takes place we need to have some ESD charge measurement data so that we can compare with data taken after the discharge and see the effects observed. This time constrains has made me decide to do the measurements manually tonight. I may be able to run the automation code afterwards and compare it tomorrow with the manual measurements but this may not be possible. See manual measurements in the attached document.

Today I checked the RF level at various points in the ISC R3 rack (which is the one by HAM6, containing the AS-related demodulators) as a part of the DRMI preparation.

The next thing we need to in this rack is the I-Q balance adjustment for all the demodulators. This is something we can do before the table is moved in place.

RF level re-adjusted:

I have adjusted the RF level by inserting or changing RF attenuators such that the power at the input of every RF device is at 10 dBm except for the WFS quad demodulators. Sine those quad demodulators have a 4-way splitter in the box, they did not have to be as low as 10 dBm. So I simply did not put an attenuator for their paths. This gave me about 13 dBm at the inputs of all the quad demodulators.

Two RF cables replaced:

When I was checking the RF level, I found two wacky cables which showed anomalously high loss. One of them was pretty obvious -- the SMA connector part was the cause and stress on the connector changed the amount of the loss. So I simply cut the SMA connector part off of the cable and put a new SMA connector. This fixed the issue. In the other cable, I could not identify what parts were giving the high loss. However Filberto replaced both N-connector and SMA connector for me and this fixed the issue. I put them back in the rack and confirmed that the RF levels were at the expected level. Good.

As suspected and reported earlier while doing coil balancing (cf logs 1 and 2), SR2 M2 UL actuation appears to be non functioning. An in-chamber cable would be disconnected.

First, in order to confirm the problem, an old range-of-motion matlab script was resurected and used : it ramps positive then negative offsets (+/- 132000 cts) to each coils of M2 stage, and looks at the motion response. The plots of the results are attached and are clearly showing a difference with the UL response, moving less than 0.1um whereas the other osems show more than 1um range of motion. 

Then, to see if the problem was in-chamber/in-air, the resistance of the four coils of middle mass was measured at the satellite box and the in-air side of the feedthru - in both cases a breakout board was used, plugging it to the sat-box side of cable H1:SUS_HAM4-31 in the first case, and the HAM4-D6-3 feedthru connector in the second case (wiring ref : D1000599). Measurements with the multimeter at the two ends roughly gave the same results described below :

LR (pins 1-14)~= 17 Ω
UR (pins 4-17) ~= 17 Ω
LL  (pins 7-20) ~= 17 Ω
UL (pins 10-23)  = ∞ Ω

The resistance of UL shows the circuit is open between the flange connector and the coil, somewhere in the chamber.

SR2 was tested by Stuart after HAM4 doors closeout and wasn't showing any actuation issues.

Since Borja balanced the coils with a non functional actuator yesterday, I will run undamped M2-M2 TFs overnight using the gain values he found, and see how the dynamic response is affected.

Stefan, Sheila, Alexa, Kiwamu

We locked the IMC on a 00-mode this evening. The next step is to wait for the pump-down to finish and  perform a ring-down measurement to check if the absorption is not too high.

Preparation:

After talking to Mike and John, we agreed that we can do an IMC-related interferometry with a low power laser even though we are not yet on the turbo pump.The pressure at that time was below 10 torr according to John. The PSL power was decreased to 56 mW by using the rotational stage because we needed to be lower than 100 mW which is the limit for the in-air IMC locking. We then locked the rotational stage such that we won't accidentally exceed the limit. Also I had a chance to talk to Dennis about this laser power issue and he agreed that we can do some interferometry with a low power laser. Thank you for the discussion, Mike, John and Dennis !

Initial Alignment:

To recover a good alignment, we started from the PZT input pointing. We checked the spot position of the scattered beam from the HAM1 viewport, projected on the wall of the PSL enclosure. The beam was lower than the previous marking, by 5 mm or so. We then touched up the input PZT to bring the beam back to the previous making position. The PZT count in PIT had been 679.6 and after this alignment it became 683.6 counts. The we checked the alignment on the table, but surprisingly, we did not have to touch any optics at this point. This means that Bubba and company placed the IOT2L table very precisely, maybe with a precision of a few mm. Thank you, Bubba !

We then placed an analog camera in the transmission path because the GigE was not functioning for some reason. At the beginning the alignment seemed off largely in pitch. So we started tweaking the things. Starting with the PZT again, we could not get a decent 00-mode. We decided to touch the MC mirrors. We touched MC2 in pitch and was able to see a decent 00-mode flash. Good.

Locking and further tweak:

As soon as we turned on the IMC board with a high gain, it grabed a 00-mode. We then did a futher alignment  on MC2 to coarsely maximize the power-buildup. The visibility was 75 %. In lock, the reflection was about 40 and unlocking We then re-centered the WFSs beam by hand on the table and tried the ASC servo. But, after an hour of struggling, we realized that the MC2 TRANS QPD did not have a good enough signal-to-noise ratio. So we need to crank up the laser power once the pumping is over. This is now preventing us from engaging the ASC loops.

An enigma:

Signals from the IMC_REFL_DC signal occasionally behaved quite funny -- the REFL_DC signals went up by a facto of 5 or so, while the WFS DC signals remained the same. At this point, we have no idea what was going on. We will investigate a bit more tomorrow.

(Alastair and Greg)

Over the last three days we have worked on three separate areas:
1) Bonding the Wavelength thin film polarizers into new mounts
2) Working on the X and Y tables
3) Testing laser #20510 that arrived today from Caltech

1) Thin film Polarizer Bonding 
The thin film polarizers from Wavelength, of which we use 4 per site, had pretty flimsy mounts where a plastic screw was used to hold the optic in place. This resulted in a tendency for the optic to move around causing quite large angular changes that could not be accommodated with our optical layout. A new design of mount was made and here we bond them into place.

Initially we measured the pitch of the old mounts for each of the 4 polarizers.  We measured the initial position of the optic, then tested how far it could pitch up and down. 
#1) Initial position 26mRad, Highest pitch 31mRad, Lowest pitch 26mRad 
#2) Initial position 18mRad, Highest pitch 31mRad, Lowest pitch 13mRad 
#3) Initial position 9mRad, Highest pitch 12.5mRad, Lowest pitch 9mRad 
#4) Initial position 5mRad, Highest pitch 5mRad, Lowest pitch 5mRad 

The optics were then removed from their old mounts and were drag wiped. First contact was applied to the front surface to protect it during mounting. We then bonded the polarizers to the new mounts using RTV Silicone. The silicone was applied with a syringe to three points on the mount, and using 5mil spacers the optic was held with a set spacing to the mount. After 24hours (listed as max strength setting time for the RTV) the spacers were removed and the pitch of the newly bonded polarizers was measured. The mounts have 1/4-20 threads top and bottom, and we measured each polarizer both ways up to be sure that we don't have any issues with how parallel the faces are, and since the optic is on a post on a table we could also have some error from this also which we would see by checking the polarizer both ways up. Serial numbers below are those stamped on the new mounts 

#001 Writing right way up 1.5mRad pitch Writing wrong way up 1.5mRad pitch 
#001 Writing right way up 2.5mRad pitch Writing wrong way up 6.3mRad pitch 
#014 Writing right way up 15mRad pitch Writing wrong way up 11mRad pitch 
#008 Writing right way up 5.6mRad pitch Writing wrong way up 5.6mRad pitch 

Polarizer #014 showed a pitch that was outside what we think acceptable (requirement 7mRad), and was then removed from the new mount and rebonded. Since we have spares we used a new mount to rule out problems it. The new mount is #015. This polarizer was tested for pitch after 4 hours of curing, but the RTV was found to not be fully cured. It was reset to the correct height and will be left the full 24hours before retesting. 

2) Work on X and Y tables 
The X table initially had a full version T1200007-V11 of the layout (not using masks in flipper mirrors). The Y table had most of the layout roughly in place for T1200007-V12, though mostly not aligned. The two polarizers from the power control stage had been removed from both tables for bonding. 

We began by turning on the laser on the Y-arm table and aligning through the first polarizer and AOM, up to the power control stage. The extinction ratio at the first polarizer was measured at ~100:1. The AOM was checked to have a 1st order efficiency of >90%. No alignment was done after the power stage since polarizers were not available. However we did go through and mount the few missing optics in roughly the right place, so the table is complete but not aligned. 

The alignment on the X-arm table should have been correct up to the power control stage, but was found to be off after the first mirror. The mount was checked and it was found to be held securely in place with the actuators locked and the mirror firmly secured. The laser mount was also checked and found to be secure. It's not clear how this had moved, but by the time the beam reached the rotation stage it was back in the correct place, so we hypothesize that the first mirror was knocked during the previous alignment and prior to its actuators being locked. We realigned the beam through the first polarizer and AOM, checking the extinction ratio of the polarizer and the efficiency of the AOM (similar values found to the Y table). The alignment was again performed up to the power control section but again the lack of polarizers meant we could go no further. This table has all its optics, but they still need placed and aligned for V12 layout (some small changes were started, but most optics have not yet been moved). 

3) Testing of laser #20510  
The laser that arrived from Caltech was found to have damage to the RF driver from shipping. There were three substantial pieces of damage : a dent on one side where the end of the laser had hit, and both sides of the driver had bent water connections. There were no signs of damage to the laser (you can see the linetracker on the end is straight in the photo attached, though it had clearly been hit into the RF driver) and the outside of the Pelican flight case shows no damage either. 

We decided to test the laser immediately so that any repairs can be carried out. The laser was tested in the squeezer bay at the end of the LVEA, and required power, RF, water cooling etc to be put in place. 

Using the driver for the laser we immediately found problems, with a maximum power output of 40W. It was found that one of the fuses on the RF driver had blown (clearly burned out and not just broken). We replaced it and immediately the fuse blew again. We then replaced the driver with a unit from a different laser. The power output was ~50W at 95% duty cycle (using synrad controller this is maximum output). The CW output should be in excess of this and matches the specs for the laser. 

The RF driver will need to be returned for repair and we will be looking into improving the packaging for the lasers. There are two more lasers currently at Access having their water cooling problems repaired, and Greg may drive out to pick these up just to avoid more damage.

J. Kissel, G. Valdez, R. McCarthy

I grabbed the recently found GS13 578 with the busted locking mechanism (see LHO aLOG 12932) from EX, opened the can, and inspected the locking rod and gear mechanism inside the can. I didn't find any obvious obstructing stuff that would prevent the locking mechanism from functioning -- it just seems as though the main gear controlling the locking pin has ceased. Without further dis-assembly, with which I'm uncomfortable doing, there's no way to assess the problem further.

While I had the can open, I converted 578 from horizontal to vertical configuration -- a much more robust configuration if we don't have a locking mechanism. Unfortunately, during the process, one of the three fragile metal tabs that engage the corresponding vertical spring (already pretty mangled when I got it) broke in the upper-left corner such that the screw-hole is no longer whole. Luckily there was still enough of the tab left, that I could put the tab in place, and tighten down the square nut enough that it held the tab engaging the vertical spring. 

Attachment Captions:
Pg1 - A picture of the corner of the broken tab, showing the vertical spring engaged regardless of the missing chip.
Pg2 - The bottom of the locking mechanism, showing the locking knob gear, and the main locking pin gear. The gears are engaged, and show no signs of damage.
Pg3 - Looking under the proof mass, at the locking pin. Nothing seems to be interfering.
Pg4 - A shot of the top of the instrument, for ID.

After restoring the can, I increased the tension on the vertical springs to bring the proof mass up into the locking containment ring, as the locking mechanism would do. It's *lock* locked, but it's held against the retainer ring. In this configuration it can now withstand a few g's without bouncing around, and there's much less danger to the fragile horizontal flexures (still the stock flexures, as these were installed before the Stanford team innovated better ones). Note, this GS13 also has a LIGO custom pre-amp (D050358) in it, so the Q is a sludgy 5-ish (ref. T0900457), which is also good for robustness.

From here, I suggest one of two courses of action:
(1) Replace the stock horizontal flexures with the rugged Stanford flexures, such that this GS13 doesn't need a locking mechanism any more. We should have at least a set of these at LLO left in the spare stock.
(2) Send the instrument back to GeoTech, and have them repair the instrument. If so, then then can also replace the broken vertical spring tab.

I'd say the likelihood of instrument failure if shipped it it's current condition is pretty high. 
I'd say swap out the flexures, but will discuss with the SEI/PEM/CDS teams. 

For now, the instrument resides in my office, leaving the summary of these eLIGO PEPI Feed-Forward GS13s as
S/N     Config      Current Location
568        V        EX, in change-room, by bench
574        V        CS, In my office
578        ??       EY, on change-room racks
584        ??       EY, on change-room racks

This morning I spent sometime updating the ops overview screen, it is projected in the control room again.

There are two things that would be helpfull from operators on shift or anyone on site:

•   Checking that the intent bit is accurate  This bit should be set to undisturbed when no one is actively working on the interferometer, and set to commisioning if there are measurements running or people working in the LVEA. This will let detchar know when to look at the data.  Commisioners will try to remember to set this as well when we leave for the night and when we start work, but it would be helpfull if everyone keeps an eye out to make sure it is accurate, especially when work first starts in the morning.
• Untripping watchdogs if there is no known reason why something should be tripped, we should try to keep things isolated/damping and keep the PSL running. 

While I was doing this I noticed that the PSL ODC state bit was bad.  I toggled the noise eater switch to get rid of the NPRO relaxation oscillation, after which stefan readjusted the ISS setpoint.  The theory is that the setpoint was adjusted while the NPRO was oscillating, once the oscillation was gone the ISS could no longer lock at that set point. 

Lastly, the FSS mixer voltage has been around 0.8 since june 16, when the PSL came back from the maintence.  Peter King tells me the PD was swapped out durring that time, so I have raised the threshold for the ODC to 0.9V from 0.6V.  The PSL ODC bit is now green. 

Jeff K., Krishna V., Erik S.

We managed to reduce the frequency to about 6 mHz. We then attached the left arm of the vacuum can. The heat-shield cross-links were then connected, the reference beam-splitter was installed and we did a rough alignment of the optics. The top flange was then put in place and we bolted the autocollimator on and turned on the LED. With some foil-shims under the autocollimator, the alignment was good enough to get clean images of both the reference beam-splitter and the balance-mirror. After a few attempts of moving the adjustment rod very carefully, we managed to get the balance-mirror in range of the autocollimator.

We then enclosed the tiltmeter in the foam-box and started the data-taking program on the laptop. The next step will be to measure the tilt transfer function to measure the distance between the center of mass (COM) and the pivot. We will then adjust masses on the balance to minimize this distance.

We're already measuring sub-nanoradian precision.... at 10 [Hz] ;-). See attached read-out display. On the upper right display is the difference between the reference mirror and balance mirror position as measured by the autocollimator (left and right image pattern on the upper left panel, respectively), where 1 [pixel bit] = 3.5 [urad]. You can see the spot images that the CCD is sampling on the first images of the "Day3b.pdf". The ASD in the bottom left corner shows the measured tilt as a function frequency. 

Source mdl and C files used to build the H1 front end models:

Files which are locally modified (need to be checked into the repository):

Jeff B, Arnaud, Kiwamu and Dave

I'm instigating a weekly resolution of partially loaded or modified filter files. The philosophy is that pending filter changes shoud be applied regularly, preferably during Tuesday maintenance. Unloaded FM changes are time bombs which can impact on recovery time following a model restart, computer reboot or power glitch.

Running the command "check_filtermodule_load_status" this afternoon gave the output:

Similar to HAM2 & HAM3 but the shift in position wasn't enough to trip the ISI but the outputs were decidedly large.  So, I reset the target position to the free hang position.  Unable to safe.snap as SUS folk are using platform. 

Day Shift Summary
LVEA laser hazard

09:02 Justin – Transition LVEA to laser safe
09:15 Doug & Jason – OpLev alignment at X-Arm
09:22 Nathan – Working in the optics lab
09:25 Paul, Jordan, & Sudairsham – At Y-1 in beam tube taking PEM measurements
09:30 John & Gerardo – Vacuum pumping in corner station
09:37 Hugh – Unlocking HEPI at HAM2 & HAM3
09:45 Richard – High Voltage on at End-Y
10:21 Betsy – Cleanup work in west bay
11:15 Richard – Going to End-Y
12:24 Hugh – HAM3 HEPI work
12:44 Gerardo – Going to End-Y
13:00 Karen – Cleaning at End-Y
13:38 Kiwamu – Escorting a group from KAGRA on a tour of the PSL
14:20 Apollo – Installing feed through at HAM6
15:02 Jeff – Going to End-X to work on tilt meter
15:21 Doug, Jason, & Apollo – OpLev preparation work at Y-Arm
15:26 Vern – OpLev preparation work at Y-Arm  
16:00 Greg & Allister – Energizing CO2 laser in east squeezer bay
16:05 Shelia – Transition LVEA to laser hazard    

Appolo Team installed the feedthrough on HAM6 for the ISC fast shutter and the feedthrough emulator.
The fast shutter was tested with a DC power supply and it was confirmed the shutter mirror popped up.

Since the prepared feedthrough has the form factor of 2.75" CF, a zero-length adapter 2.75"-4.5" was inserted (great thanks to Gerado and John).

Once the feedthrough was installed, the in-vac shutter cable was connected to the feedthrough.
Then the shutter connection was tested by giving some current. With a current of 50mA, I could confirm that the shutter mirror popped up with this current.
At CIT we tested the shutter at 150mA in a vacuum pressure and it survived. So as long as we don't exceed this current, the shutter coil is completely safe.

The viewport emulator was also installed on the North door flange.
The heights of the emulation rings, particularly the east-most one which was moving, should be confirmed and adjusted.

With the new cart biases, this is certainly a time to update the safe file.  I've also taken on the campaign of turning off outputs of unused filter banks feeding into the active loop paths.  Since these can get an offset etc and one might be none the wiser.  If the nominal state for the unused banks is output off, seeing that the output wires not green is sufficient to know that those other banks are not causing a problem.

I instructed conlog to stop recording 98 *_THRESH_MAX channels. These can count down every second over a long period of time.

The names were added to the exclude list in '/ligo/lho/data/conlog/h1/input_pv_list/exclude.txt' and the script to generate a new channel list was run. The script removed these and added 13 TCS channels and 1 SYS channel.

In total 14 channels added, 98 channels removed.

Very similar story to previous post on WHAM3.  As soon as the target position was reset to the current free position the controller locked right in.  I don't see as large a shift at balance time but may be clouded by whether the ISI is locked or not.   Attached are the local CPS, Cartesian CPS, and changes in the Target positions.

The ISI was tripping and I saw it was because the deltas between free position and the targets were too large.  I didn't realize we were restoring all DOFs, I thought we only restored the RZ (Yaw) except for a few other special case locations.  This tripping was happening during the RX & RY target restores and maybe this makes sense.  The payload/balance changed on HAM3 (as on many others) and the balance was done well within spec.  See the first attachment for the CPS Local shifts during my balance on July24. This looks pretty bad even though our spec is +-1600 counts.  This is a 2 day plot though and there may be some detail in the time I actually balanced, like, since the payload changed here so might some distortion on the table and a new CPS zero/lock comes about.  Regardless, it is still within spec even though I feel this could have been balanced better.  See the second plot though for the real problem.  Since the free hanging position moved, of course so did the cartesian position.  Since the current ISI control scheme is restoring the target bias of all DoFs, the controller has a lot further to drive to the target and hence the trip.  Tried several times with Guardian and ISI scripts and same thing.  I reset the Targets to the current free hanging position and all is well. The third plot is the shift in the target position for the record if someone needs to steer back.

WHAM3 is currently under managed Guardian control at ISOLATED.

I slightly modified the previously made LSC ODC block and added it to Kiwamu's latest LSC model. The model compiles, but I am holding off with installing on running. They are not in svn yet because I want to verify the model is running well.

Affected files:

common/models/lsc_asKiwamuLeftIt_20140806.mdl               : What the name sais. My starting point.
common/models/lsc_addingODC_inprogress_20140805.mdl  : my changes
common/models/lsc.mdl                                                           : currently same as common/models/lsc_addingODC_inprogress_20140805.mdl
h1/models/h1lsc_asKiwamuLeftIt_20140806.mdl                    : What the name sais. My starting point.
h1/models/h1lsc_ODCAdded_20140806.mdl                          : my changes
h1/models/h1lsc.mdl                                                                : currently same as h1/models/h1lsc_ODCAdded_20140806.mdl

While doing cabling for HAM6 in the CER, noticed that the following cables for ISCT1 in ISC-C2 were plugged in backwards. Swapped the following at the EtherCAT Corner 5 Chassis. Cables are connected as listed below and match the pull list E12000408.

Cable_ISC_328 connected to Port 9 (DC PD's ISCT1)
Cable_ISC_326 connected to Port 10 (Auxiliary ISCT1)

Filiberto Clara

J. Kissel, R. Schofield

I discovered a lonely, forgotten, disconnected GS13 hiding under BSC9 (H1 ETMX) chamber today. I attach pictures. 

Things that Robert and I can surmize:
- It's serial number is 574, and has a "property of the USGS, greater than $5k" sticker on the side with the number G12077 and a bar code.
- It may have been used for S5/S6 feed-forward studies (Will check with Mike, Keita, Richard)
- Ski may have purchased it, given the bar-code on the side (Will check with John)
- It's a 2005-ish era model, based on the color of its can and unfinished look of the feet.
- It's currently configured as a horizontal
- It claims of a broken locking mechanism, but after a few turns the mass seemed locked enough for transport
- It's functionality is unknown.

Finder's keepers ...