We've eliminated all ground loops that can be easily eliminated in chamber at this point. There are two remaining grounding-related issues.
-
There seems to be a serious ground loop problem for OMC DCPD and OMC QPDs, which is described in a separate alog (https://alog.ligo-wa.caltech.edu/aLOG/index.php?callRep=12349).
-
Shield for beam diverter cables (there are two) appear to be grounded to the ISI via mighty mouse connector. This is Beckhof and there's no signal that needs a great S/N or anything on these cables, and I don't know how the grounding of Beckhof chassis is done, but we might want to modify the in-air DB25 cable to cut the pin13 wire at the feed through.
The first problem will NOT be addressed until I hear back from Rich.
The second issue is dealt with in air (if the solution is to cut pin13 wire).
We "fixed" another tiptilt with more annoying problem than yesterday (see below).
We "fixed" a bunch of other things by changing four in-vac cables (also see below).
Apart from the issues noted above, we checked picos, OMC REFL QPD (sled), AS_C QPD, WFS(DC) and OMC SUS, and these are all good.
Everything is connected back except the DCPD and OMC QPD cables at the field rack.
We didn't like the way a big coil of cables was dangling from ISI table top to the side, no idea why this was done (just an extra weight without real merit), so we moved it to downstairs. Due to the reduced weight on the ISI, Hugh might want to rebalance ISI.
Tomorrow I'll be out, but Corey and Dan will have to do the following:
-
Re-center BOSEMs of TT2 and TT3. These are the units that we pulled and reinstalled. Most of the coils are OK, but one of the upper coils for each TT is not well centered.
-
Check if QPDs, WFS(DC) and DCPDs work using a flash light, because we messed with the cables. There are 5 QPDs, 2 WFSs and 2 DCPDs. Right now all corner station ISC whitening are crazy, ask Daniel to fix before you go in chamber. You also need to connect the DCPD cable and OMC QPD cable back to the field rack.
-
Check if beam diverters and picos work, because we messed with the cables.
-
Unlock ISI and give it to Hugh.
After this I don't have a problem closing HAM6 for now.
Annoying TipTilt, part 2
When "fixing" OM3, one of the BOSEMs was really persistent. Both the side and the bottom were touching (first picture), and the bottom gap was non-existent no matter what, plus the maximum side gap I was able to achieve was less than the thickness of aluminum foil.
I used a folded piece of aluminum foil as a shim to raise the micro DB connector (second attachment) to have enough bottom gap, and cocked the connector as much as I can to maximize the side gap (third attachment).
FYI, the shell of the female connector in the second picture is not isolated by design. But the male connector shell that is on the cable on the first and the third picture should be isolated.
Annoying cable and connector problem.
We replaced total of four in-vac cables.
Three of them is due to an issue with yet another design feature to rely on razor-thin gap.
In the attached picture, the front shell is PEEK but the screws that fix the front shell to the metal back shell are metal. The screw heads should be somewhat recessed, but if they are sticking out for any reason, they will contact with the metal part of the feed through, thus the metal back shell is connected to the chamber.
Some of the screws are recessed less than the others, and some of them are sticking out because of the screw head slot was somewhat deformed by the screw drivers or whatever. There can be three modes of failure:
We have identified three cables that were either 1. or 3., and replaced them with new ones which worked OK. There's no reason these screws should be really tight (I think we need to discourage people to do that), but if it fails because the screws are a bit tight, that's not a good cable either.
We identified another cable that was 3., but this was a picomotor cable with thicker gauge wires, and we don't have replacements. Corey made it such that the connector doesn't fully seat (or maybe he just didn't fully tighten the connector, I don't remember). This cable was tested good yesterday and today, but it turned bad when we were doing some other unrelated work at the feed through and its vicinity.
We also found one cable that makes or breaks ground loop depending on where/how we route and how we breathe, and replaced that with a new one:
The serial number markes as ?????? is not unknown, it was recorded before by Corey, it's just that I don't remember.
Closing the documentation loop on these guys (also updating ICS Assembly Load #ASSY-D1300122-LHO). I'll mark the damaged cables as such in ICS as well.
According to ICS, this cable (the elusive D1000223) is most likely one of these: S1202641 (most likely) or S1202643. And as Keita mentions, we have no spares of these at LHO (and NONE for 3IFO!). :(