Started TMSX transfer functions from DTT. Measurement will run overnight

[Dave Kiwamu Arnaud]

Today, we discovered that top mass OSEM2EUL matrices were set to 0 for TMSX and IM1. Also, they were frozen, and would go back to 0 right after trying to set a value in any of the slots. The recorded burt files show that values were 0 since august 30th, when rcg was updated to 2.7.

The problem was fixed by restarting the models

Stuart has been reporting a similar issue at Livingston

Picomotor controller 5 is not sending the correct readback to beckhoff, (the power on, pin 7 on the readback cable for example).  I pulled the unit and it is in the shop, I will find a tester for it in the morning.  Controller 4 is having similar problems, but I did not pull it. 

- PCAL working on LLO-Y periscope

- PSL crew work ongoing

- Sprague spraying down beam tube enclosure

- TMS crew at EX

- EY ion pump signal goes down for rerouting cables

- LN2 delivery at EX

- Kyle at GV5

- Cyrus reconfiguring and rebooting the vacuum LX VME crate

- Upgrade H1 PSL front end models

CP2 overfilled following Cyrus' reboot and BURT restore of the LX VE rack this morning -> The instrument air for CP2 indicated red or "STOP" on the MEDM screen as expected but was actually enabled as if in the "START" state -> Normally, we would be required to manually enable or "START" this via the MEDM screen following a reboot/restore and it would then indicate green -> Investigation showed a "fail" light associated with a 2210 card was on following the initial reboot -> Cyrus replaced this card, rebooted and BURT restored again -> I confirmed that the LLCV valve was now closed when the MEDM indicated "STOP" -> Later I noted that the LLCV remained closed even with 100% open PID output and with the instrument air "START" condition, i.e., the LLCV was not opening regardless of the PID output value or the enable state of the instrument air -> I confirmed that there weren't any rack fuses "lit" or open 

For now, I manually filled CP2 to 105% (boil-off rate ~2.5%/hour) and will come in early in the morning and manually fill again etc. until we find the problem.

[Stefan, Kiwamu]

 As a part of the commissioning wrap up porjects we characterized the IMC ASC control.

Adjustment of UGF:

• UGF of the all 6 loops was adjusted to be about 0.08 Hz, measured with an impulse response method
  • we had a difficulty in exciting the suspension with swept sine signals and ended up with the impulse response technique where we inject a step function at the error point and see the behavior of the error signal.
  • Later Stefan managed to measre the open loop with swep sine. See tha attachement.

Change in servo shapes:

• Looking at the spectrum of the WFS error signals we noticed there was a sharp peak at 1 Hz which is a pitch resonance of the triple suspension.
  • The height was about 10 times higher that the floor around this frequency
• We impleted a notch filter at 1 Hz with Q of 2 and heiht of 30 dB.
• Instead of enabling FM6(0.06, 0.1) we decided to go with this simple notch filter.

Adjustment of Output matrix for DOF3:

• We put non-zero numbers in the cross-coupling elements in the outmatrix for DOF3.
  • We have been using only PZT for feedabck, but this made the response of the control slow because DOF1 and 2 needed to wait and follow the movement of DOF3.
  • Instead of doing this cascaded loop we tried a simpler scheme by feeding the signal back also on the suspensions.
  • This worked well.

Some other changes :

• we increased the whitening gain of IM4 TRANS from 15 to 45 dB.
• we enabled the top two whitening stages for each WFS

Leaking joint that had been externally "VacSealed" a few weeks ago may still be leaking, albeit less than it had been -> 9 x 10-8 torr*L/sec helium background rose to 1.1 x 10-7 torr*L/sec after 150 seconds of helium applied to the bagged joint -> This needs to be repeated to confirm -> ideally the background should be lower in relation to the leak being measured for better confidence

Christina, Patrick

Yesterday, we aligned the DBB for both input beams and started checking on the measurement scripts. A tiny misalignment of the PMC input was also fixed. 

In the morning the position of the photo diode inside the Front End was changed due to ECR E1300621-v1. During that the laser was shut down because of the lid-overwrite. We restarted the laser, the power of the oscillator droped a little on the MEDM screen. However we couldn't find the reason. Now that photo diode is hit properly, the status of the noise eater can be watched on the laser MEDM screen.  

We found that the reference cavity was misaligned a little, thus, we optimized that. 

The DBB scripts were fixed and we run all the measurements for both laser systems. We also run the ISS measurement script. The performance of the lasers seems to be good. The out-of-loop noise of the ISS is a little to high, but the inloop performance indicates that the loop is working correctly. (At the moment the show references in the pdf are the measurements from Livingston. We will create our own references soon.)

BOSEMS:

- cabling issue found this AM

- BOSEMS installed and tested - two were replaced - TMSX now has 6 good BOSEMS

- BOSEM channel LEFT is reading a known good BOSEM (20K+ counts on any other channel) as 13K counts
--- Filiberto is at EX tracking down the issue.  Last info from him was that all cables have checked out OK, and he's looking at the electronics.


PREP FOR MOVE:

- telescope safety support beams attached to the top assembly


STILL TO DO - before transfer functions:
- LEFT BOSEM channel fix
- center BOSEMS
- remove protective wipes from the table, note some have visible particulate from installation activities
- check cables are not touching wires
- lower genie to allow TMS to hang freely

I have to leave for a dentist appt. but expect to resume leak checking GV6 when I return (1530?) -> The Corner vacuum equipment is being pumped only via the YBM Turbo pump which is being backed by a leak detector and CP2.  IP1, IP2, IP5 and IP6 are valved-out.  This is a temporary pumping arrangement.  Also, the instrument air used to control CP2 is not enabled currently which is also a temporary state.

W/ expected pending approval of the WHAM6 ISI Optical Table position, went in to fine tune the elevation & level of the HAM6 Optical Table.  Also, we had done a rotation last week as we assessed the horizontals so vertical did need to be revisited.
Using LVEA vertical monument 802, the Optical Table was set to -220.1+-0.1mm Gz as per D1100398 & E1000403.
Three pages of survey, dial indicator, & load cell readings are attached

I moved the vacuum VME crate for the X beam manifold (h0velx) over to the new boot server today, as specified in WP 4090.  This move did not go as smoothly as the one for LY.  Two issues:

1) The paths in the startup file pointed to /src - all the other files have referenced /cvs instead, I'm not sure why this one is different; it may be that the startup file for this crate had been updated with the 'new' fs scheme (/src) at some point, but the others never were (/cvs).  In any event, I need to re-scrub all the startup files to verify they point to the correct place on vxboot (/cvs in this case).

2) The 2210 VME card failed on reboot - this is the relay/BIO board.  The symptom was that the fail light would not go out once the crate was booted.  Tried re-seating the card (which did not work), so the card was swapped out with a spare.  One symptom of this failure was that the CP fill valve went open to 100%; it's not clear if this is the design intent, or a side effect of how the card failed.

In any event, h0velx is now running as expected from the new server.

This work done last week Wednesday.  Opportunity arose so we (Jim & Mitchell) uncovered the upper ISI and located the 600 lbs on the Keel.  Location of the mass is per D1002191-v2.  This will have to be removed for Cartridge installation and then replaced in chamber but meanwhile, the ISI is ready for all auxiliary payloads, balancing and integrated testing.  The upper ISI was recovered with C-3.

LLO X-arm Pcal Periscope, is wrapped outside the H2 PSL, waiting for the crate from LLO, we expect the crate arrives this week.

The structure is on a four wheel dolly, and attached to a hoist.

Today we will start the assembly of LHO Y-arm inside the H2 PSL.

pablo

I added, compiled and installed the green arm initial alignment code as part of the h1asc.mdl We are at revision 5464.

The installed version has a single ON/OFF switch to turn on the loops and integrators, as well as ramp in and out the dither drives.

Still TBD:
- IPCs to ETMs TMSs and end station PZT actuators.
- MEDM screens for the setup.

For usage in Robert Schofield's magnetic coupling measurements and general optical lever optimization, we were playing around with the whitening chassis to optimize the QPD signals going to the ADC.

The optical lever QPD electronics have the ability to engage up to 3 levels of 10:1 whitening; all with a pole@10Hz and a zero@1Hz.  These filters are turned on via a daughter board connected to the whitening chassis (D1001530), which are exactly the same as ISC's whitening chassis.

Figure 1 shows the spectra of the three levels of whitening to make sure the electronics work the way we think they do, notice that we only go up to two levels of whitening since engaging all three levels completely saturates the ADC.  At two levels of whitening, we see that the higher frequency (~50Hz)amplitude is comparable to the low frequency (~.6HZ) amplitude but not yet dominating the RMS.  Although after monitoring the optical lever for a few hours with two levels of whitening engaged, we saw that the ADCs railed briefly about once an hour, not good.  So in the end, one stage of whitening was optimal and of course, a de-whitening filter was also implemented in digital land.

Figure 2 shows the whitened+de-whitened signal and compares it against the unfiltered signal to make sure we didn't screw up the digital de-whitening portion.

Figure 3 and 4 compare the dark noise currents of one stage of whitening engaged vs. no stage of whitening engaged.  There is about a magnitude of difference in the signal to noise ratio between the two set-ups at >10Hz, as we expected.

Computer in the EX TMS was swapped out for one that can run MEDMs.  BOSEM signals were displayed on the screen, but did not change with BOSEM flag alignment.  We swapped cables at TMS to verify that that wasn't the problem.  Jeff spent a fair amount of time on the phone with Dave looking into the software situation.  The outcome of that was that it appears to be a hardware issue.  Dave did some further diagnostics, however it's late enough in the day that a fix will come tomorrow.

Christina, Patrick, Sheila

we transitioned the PSL to commissioning mode at 2:45, it sill stay in commissioning mode for the rest of the week