Temp dogs were removed.
Isctex green periscope readjusted as it was close to the edge.
Tele beam position readjusted so the beam is centered on the primary.
Red extraction path was done.
High power beam dump path was done.
Working on IR QPD sled path.
*** HPI ***
Unlocked
*** ISI ***
Undamped
*** SUS ***
Damped
I've made some changes in the core switch (sw-msr-core) to support the new workstations subnet that we need to build to support the burgeoning number of hosts we have. In addition, I've updated the network interfaces on cdsegw0 to account for the new VLAN, and created associated EPICS gateway processes to match. cdsfs0 is now configured to allow mounts from 10.22.0.0/20 as well.
Cheryl, Sheila, RObert
The PSL shutdown because the power watch dog tripped. (second attached trend shows power slowly dropping) Cheryl and I restarted it about an hour ago.
All week, the reference cavity alignment has been fluctuating. right now I have turned the TPD resonant threshold down to 0.3 V so that the FSS is is locked. This is supposed to be 0.9V, so this is pretty bad. Although we have had drifts over the last few months, this week has been especially bad. The PMC transmitted power shows the same trend, so the problem seems to be upstream of the FSS. (first screen shot shows trends)
Robert and I had a look at some temperature trends, we saw that the new higher termperature commisioning mode the temperature on the table has been fluctuating a more than Robert expected, and that this temperature fluctuation is correlated with alignment fluctuations. This has been true since https://alog.ligo-wa.caltech.edu/aLOG/index.php?callRep=7564
So, in summary there seem to be 2 unrelated problems, one that has cause the PMC trans power to fluctuate more than usual this week, and the temperature fluctuations that have been present since late august.
- Filiberto and Frank to EY for electronics verification
- Jeff and Thomas to BSC3 to work on ACB
- Bubba/Apollo to install viewports on HAM3
- Jodi and ? to EY to vacuum out bellows/beam tube where spool was removed
- Joe and Greg - laser hazard in H2 PSL enclosure for alignment work
- Sheila - transitioned EX to laser safe
- Jeff and Andres to EX to remove pushers (in chamber work)
- around noon, starting to put HAM2 doors on
- Jeff and Andres back out to EX after lunch and talks
- HAM2 doors are on
- Reverse Osmosis alarm around 3PM
- Sheila restarted the H1 PSL laser
- I redamped HAM2 and HAM3 optics
After the removal of 12 bolts from the Table top balance masses (to add viton dampers) we expected needing to replace that lost weight. We added 3 D071201 (0.3lbs total) to the NW corner--see image. Betsy removed some more teflon stuff later and we see almost no effect from that change, it was small and fairly uniform in distribution. ISI CPS vertical readings are a few hundred counts, zero is perfect on this HAM and well under the 2000 cts absolute requirement. Lock/unlock differences are essentially unreadable; this ISI locks very easily.
Concern though about this added mass: The masses are secured in place with a couple 1/2-13 nuts on the long studs. There is a recess into which the mass is pulled with the tightening of the nut. It actually bends the mass in the middle where it is unsecured. It is light but if there is an unwanted resonance from this, we need to scheme some mitigation: something behind it in the middle to prebend it, a bushing in the well to keep ends from being pulled in...
J. Kissel, B. Weaver, M. Landry, S. Biscans, C. Remet H. Radkins, T. Vo, B. Gateley
*phew*
The last known in-chamber payload suspect in the misdemeanor seen earlier with the ISI-ITMX vertical transfer functions was the teflon rails that were temporarily installed on the SUS-ITMX for first-contact cleaning. So, Betsy and I went out and removed them. Preliminary measurements (attached below) indiciate these were the problem, the ISI looks fine now. We are a go for final cleaning and close up of BSC3 (as agreed by all those listed above). In the next half hour that it takes me to finish writing this aLOG, Sebastien is going to get a few more transfer functions just to be sure we're OK, but Celine (from afar) has been watching the story, and has informed us that she had seen similar badness when she had taken ISI transfer functions with the QUAD locked (with EQ stops), so we 100% confident that things have been restored to goodness already.
To repeat for clarity,
- It wasn't the dome installation that caused the badness
- It wasn't the arm cavity baffle's eddy current dampers or mechanical rubbing
- It wasn't the in-chamber cabling
It was 300 kg of mass (the suspended elements of the QUAD) that had been temporarily locked up with teflon rails for the (new, yet) normal first contact routine.
--------
Plot Caption:
[[ISI-ITMX, Stage 0-1 Actuator drive, Stage 1 CPS sensor response, driven transfer function.
HEPI is floating and balanced, ISI is floating and balanced, ISI control is entirely OFF,
ACB is freely suspended and Eddy-current Damped, SUS-ITMX is freely suspended with main
and reaction chain locally damped with top-stage OSEMs]]
BLACK Known good reference measurement taken over Tuesday-to-Wednesday (2013-Nov-19 to 20) night.
BLUE The bad state the ISI has been since measuring quickly Wednesday afternoon, after
Teflon rails are installed,
QUAD is first-contact painted,
ACB is B&K hammered
Dome is installed
ACB is re-inspected
ISI and other in-vac cabling is inspected (from underneath the ISI)
RED Now good, current measurement, after Betsy removed the temporary teflon rails.
Just to confirm what Jeff said: I took a few more quick transfer functions (ST1, ST2, vertical, horizontal), and everything looks back to normal (the teflon stops were the issue). We're ready to move on.
Today, Calum and I cleaned HAM2, focusing on the table and SUS surfaces. We then pulled the First Contact on MC1, MC3, and PRM and made an inspection. Hugh also made a final balance and unlocked the ISI. I took short average TFs of L, P, V DOFs on MC1, MC3, and PRM to verify no major mechanical rubbing. I also checked SUS health via the ability to close damping loops of these suspensions and also the PR3. We pulled the first FC sheet at 11:10am and the last door went on at 1:15pm.
Note, during the last vented period, we noted that the PRM cleaning was not very sucessful and streaks were observed on the HR surface. We chose to leave the mess at the time. The FC cleaning we did yesterday/today seems to have worked as I could not see any such streaks on the optic.
Today, in the final steps before putting the doors on HAM2, we laid a fresh witness plate (wafer) near to the center of the table. It is in the area slightly behind PRM. This is the witness plate that should be picked up and analyzed immediately after the next vent and the doors are off.
We also added a 1" witness optic in front of the MC3. It is standing up, clamped between 2 PEEK cable clamps.
The Vertex ISC Cable Pull table (E1200408) has some inconsistencies in bechoff chassis port labeling for the picomotors. The pull table indicates:
The OL cables were connected to port 3, while the HAM 6 cables were connected to port 5. Based on the EtherCat Systems drawing (D1100683),
I have correctly connected these cables according to the EtherCar system drawings. The pull list remains to be updated. **
Mark Layne and I went down to Y end to vacuum the manifold between the spool opening and BSC10 (~10 ft). The manifold section showed very little dust accumulation between the time of chamber cleaning and now. There were a few (5-6) whitish flakes close to the spool opening but other than that there was very little for Mark to vacuum up. There is BSC flooring in BSC10 and the flooring hardware is in place. We left the cleanroom vacuum in place so that the PCal and CPB teams can use it for "Clean as You Go" work during assembly and installation.
Find attached the plots of last night transfer functions.
As you can see, something bad happened at the resonances around 1Hz since we put the dome on. This badness appears on the vertical sensors, both stages. It mostly seems like a rubbing cable.
Hugh and I are going in the chamber right now to see if we can find something.
PS: L4C-H2 presents a gain of 0.5. We'll fix that later.
Recall - as I pointed out below, when we go the "handoff" from SEI to close the dome and move on with other chamber closeout items, I clamped the ITMx optics.
Thanks for the recall Betsy. I think we found our issue here! I'm running quick measurements right now to double check, but we should be good to move on soon.
J. Kissel, T. Vo After a quick set of transfer functions on the ISI-ITMX showed more irregularities (see LHO aLOG 8653) after Thomas and I finished B&K hammering the Arm Cavity Baffle (see LHO aLOG 8656), Thomas an I went in again, first thing this morning, to investigate the baffle for signs of rubbing. We found nothing out of the ordinary: - The baffle suspension is free, - The eddy current damper's copper plate (which had caused problems before) were completely free of the suspended magnets, and the gap size was roughly 2-3 [mm] - The the cable situation looked no different that when we had a clean set of ISI TFs. We spend about an hour, moving things around, wiggling cables, ensuring things free by manually exciting things, but could not find anything that would cause such a drastic effect on the ISI. I attach the same quick transfer function that Seb took yesterday, BLACK shows the nominal result; what we expect to see if everything's peachy-keen. BLUE shows the result just before we went into the chamber this morning. It indicates the same bad result [as expected, since nothing should have changed over night]. RED shows the result just after we came out of the chamber, having inspected the ACB. It's the same as blue, indicating that (a) we didn't change anything after messing around with the baffle (we were worried that just looking at the ACB's ECD system the wrong way would cause badness), and (b) we can't change the ISI result by messing around with the baffle (moving things around, shaking stuff, etc, etc, i.e. the ISI is rather insensitive to the baffle). The important point we realized just as we were about to go into chamber: the other thing that happened between the noon (good) measurement of the ISI and the evening (bad) measurement of the ISI, was that the BSC dome was installed (see LHO aLOG ) After discussion with Sebastien, it makes much more sense that such a big operation is far more likely to have caused the badness that just tweaking / B&K hammering the ACB. The seismic team has now launched been launched on a thorough investigation. All apologies ACB, you're not such a bumbling oaf after all; maybe just a big, loveable, misunderstood, simpleton. HODOR.
[Chris W, Koji A]
The OMC DCPDs and QPDs were tested with a flash light. All of these are responding to the illumiantion.
Here is the confirmed connections:
DCPD (Transimission side of the BS) -> The DCPD B channel in the OMC model
DCPD (Reflection side of the BS) -> DCPD A
QPD1 (Short arm) -> QPD A
QPD2 (Long arm) -> QPD B
The modules were turned off after the test.
[Rich A, Koji A]
The OMC cables at Rack 5 were checked.
The OMC uses the split variant of the whitening chassis (D1002559) at U18, instead of the installed one.
Rich found the correct front panel module at the EE lab. The new front panel was installed. This allowed us to connect
the DCPD cable (#307 cable) which had been hanging at the side of the rack. (Note to myself: The module should be labeled)
We also found the DB25 cable between U8 and U10 was missing. We brought a DB25 cable from the squeezer remnant.
Cables for Binary IOs are not yet connected. (Note to myself: The DB25 cable should be labeled).
The HV cables are not yet prepared.
The TMSX crew are done for the day and have left the suspension in a hopefully good state, so I'm starting Matlab TFs at 16:45.
TFs had completed when checked at 08:10 Thu:
Undamped: /trunk/TMTS/H1/TMSX/SAGM1/Data/2013-11-20-1069030477_H1SUSTMSX_M1_0p01to50Hz_tf.mat
Damped: ^/trunk/TMTS/H1/TMSX/SAGM1/Data/2013-11-20-1069046943_H1SUSTMSX_M1_0p01to50Hz_tf.mat
Plots are attached. Yesterday's undamped and damped data are magenta and cyan. Included for comparison are undamped H1:TMSY of 2013-04-30 (orange) and undamped L1:TMSX of 2013-11-19 (black). Everything looks good.
Picomotor cable mystery.
It seems as if picomotor ordering is flipped somewhere between Beckhoff and in-chamber.
Picomotor in-vac cable has one DB25 (called J1) on one end and four smaller cables with mighty mouse connector (J2, J3, J4 and J5) on the other. In the cable drawing (https://dcc.ligo.org/D1000238) , these four cables are numbered such that the one closest to pin 1 of DB25 is J5, the farthest J2.
Both sites agreed that J2 goes to the green 50-50 splitter (M3), J3 to the steering mirror just upstream of the periscope (M6), J4 to the dichroic that steers the IR to the sled (M4), and J5 to the steering mirror just upstream of the IR sled (M14).
Now, the aluminum collar for the picomotor connectors are labeled as "A", 'B", 'C" and "D", and it's these labels that I and Corey used, assuming that A is J2 and D is J5.
However, in the cable harness routing diagram (https://dcc.ligo.org/LIGO-D1300007), the same cable is specified such that the one closest to pin1 of DB25 is "to M1", the farthest "to M4", it is ordered starting from pin 1 of DB25, so it's the opposite of the cable drawing. So A might be J5 and D J2, and this might be it. If so, we will re-rout the in-vac cables.
I have a vague memory that Corey confirmed that A is the closest to pin1, but I'm not 100% sure.
Since there's no way we can find which cable is closer to pin 1 without disconnecting DB25, and since there's a certain risk to disconnect and reconnect in-vac DB25, I'm asking Rich Abbott if he remembers which convention was used when labeling mighty mouse connectors.
Checking Cable (D1000238) In Lab
I pulled out two random (s/n S1104585 & S1105221) Picomotor cables to check how the four Mighty Mouse connectors are marked & where they are located. On both cables, they are marked in a way which matches the drawing/our guess for EX TMS (but not for Keita's memory for what I confirmed in-chamber in Sept):
I'm re-attaching photo/note of what we saw when we tested the Picomotor cable at the Test Stand next to the BSC (we observed the switched around order back in Sept):
(Alexa, Stefan, Kiwamu, Keita)
The convention at both LHO and LLO for the picomotor in-vac cables is as follow:
Unfortunately, the in-vac cables in HAM6 had cable "A" = J2 closest to pin 1 and cable "D" = J5 closest to pin 13; however the opposite is desired. This explains the switch. We have swapped the ABCD cables in vacuum, and now the picomotor controls are all correct. Success!
Moral of the story: be careful with the ABCD labellings on the picomotor in-vac cabling because they don't seem to be consistent. It would also be nice to have pin 1 labeled on the outside of the cable connector so that we do not have to disconnect it to know where pin 1 is located.