Attached are plots of dust counts > .3 microns and > .5 microns in particles per cubic foot from approximately 6 PM Dec. 11 to 6 PM Dec. 12. Also attached are plots of the modes to show when they were running/acquiring data.
HAM3-ISI was unlocked uppon IO's request.
Ollie, Robert, Michael R
In an effort to mitigate table noise from the chiller we swapped the quick connect fittings on the oscillator manifold with barb fittings. Robert will post an aLOG later describing the status of this test in more detail. In the process we found/created a small leak in the return path valve in the crystal chiller line. This looked like it was coming from the threaded insert on the valve, so we replaced those. No leaks were found after.
Following the mysterious crash of the SUS IM model yesterday at 12:40pm (looks like a Dolphin or timing error, SUS IM model froze up, IOP model has hugh timing error, DAQ status 0x4000), we just restarted both models on the h1sush2b front end (h1iopsush2b and h1susim). The user model was burt restored to the safe.snap. The user watchdog continues to be tripped, so zero DAC drive throughout this restart.
Toward the end of shift I heard rumblings of rebooting HAM-AUX (requested by IO group & performed by Dave)
I got into the chamber about 11 am and exited at about 12:15. I took 8 samples (7 samples of large particles, 1 small of fine particulate) using the new collection kit, which worked well for the large particles. My overall impression is that this section of beamtube is not as "dirty" for large particulate as the HAM3-->HAM2 run: Large particles found in 3 sections in HAM9-->HAM8 vs multiple large particles found in all but one of the sections in the HAM3-->HAM2 run and less fine particulate on overhead nozzles in HAM9-->HAM8 vs lots of fine particulate (enough to show up on glove with light swipe)on overhead nozzles in HAM3-->HAM2. See pix below. The samples will be sent to CIT for characterization. Robert was getting ready to go in just as I was coming out so we'll see what he finds.
- Jodi, Cheryl Last Thursday (Dec 6) Jodi and I inspected the H1 input beamtube. Pictures attached show that we found white powdery stuff (HAM3 table, beam tube, MC2 cage), white fuzz (beamtube), part of a latex glove (baffle), fibers, and uniform particulate on the overhead nozzles (picture of gloved finger shows what came off with a swipe).
Regarding the last picture in the above collection: I took a look at the MC2 suspension today and see that this pile of particulate is metallic and is from the install arm brackets which mount at this location. We have not vacuumed this suspension since in-chamber, if ever.
Mark B. Starting TFs on MC1 following the fix.
Mark B. Data taken, damping restored. TFs look clean.
Note - the fix was performed this morning. Namely, I removed all 3 top BOSEMs and 2 of the flag mounts in order to run the 4 blade tip stops out 2-3 mm.
JimW, HugoP,
HAM3-ISI is locked.
Attached are plots of dust counts > .3 microns and > .5 microns in particles per cubic foot from approximately 6 PM Dec. 10 to 6 PM Dec. 11. Also attached are plots of the modes to show when they were running/acquiring data.
HughR, HugoP,
Installed the TMDs on HAM2-ISI.
Corner 1: TMD S/N 16
Corner 2: TMD S/N 17
Corner 3: TMD S/N 18
We also started payloading the ISI. We bolted a mass stack of 200lbs (4x50lbs) to account for the absence of PR3. 34lbs are still needed on that stack to achieve PR3's mass. We might do it with wall masses as it is easier to secure them to the ISI.
Attached are plots of dust counts > .3 microns and > .5 microns in particles per cubic foot from approximately 6 PM Dec. 7 to 6 PM Dec. 9. Also attached are plots of the modes to show when they were running/acquiring data.
During the day it was noticed that the two wall clocks in the control room would reset themselves, either by showing all dashes, or by showing all eights. By late afternoon we noticed that the UTC time was one second in th future while the local time agreed with the front panel of the NTP server. We power cycled the NTP server and then soft reset the UTC clock and all systems now agree with each other.
The cause of the glitching is still not clear, please email me if you notice any clock problems.
MX rack rebuild. WP#3600
We powered down the foundry rack in the MX mid station to relocate it. We also installed new fiber optics cables and replaced some of the ethernet cables. The MX vacuum and FMCS channels were unavailable during this downtime, which was from 11:00am to 12:18pm local time.
H1DAQ h1ldasgw0 upgrade to Sol11. WP#3601
The primary LDAS QFS gateway machine for the H1 DAQ was upgraded to solaris 11.0 by Dan this afternoon. There were no frames written by h1fw0 between the times
local time | GPS time |
14:04 | 1039298656 |
16:10 | 1039306208 |
For data between these times, please use h1nds1 instead of h1nds0.
EPICS Gateway reboot.
After the reconnection of MX, the EPICS gateways would not reconnect and would periodically freeze. The cdsegw0 machine's console was showing multiple CPU lock-out errors, and I could not restart the machine from the console or ssh login. So I had to reset the computer using the front panel switch. It had been running for more than 365 days and performed a disk system check.
I then restarted all the gateways except any related to the H2 system. This means that the H2DAQ does not have access to any IOC outside of the H2 FE LAN, i.e. no vacuum or fmcs data. After 3 hours of running we are not seeing any processor creep on the gateways, normally they consume increasing cpu time after running for a while. Perhaps the reduced processor loading has fixed this problem.
Attached are plots of dust counts > .3 microns and > .5 microns in particles per cubic foot from approximately 6 PM Dec. 6 to 6 PM Dec. 7. Also attached are plots of the modes to show when they were running/acquiring data.
Closed the O-ring valves and the gate valves @ the turbos overnight, pumping will continue tomorrow morning.
Pumping restarted at 8:12 am this morning.
Mark B. (This work was mostly done Friday 12/7 but I wanted to check a few things before posting.) After the M3UL OSEM was sorted out (see 4840), I went on to get all the matrices and settings checked out and the damping working. I ran prepare.m to be sure all the filters were loaded. I entered all the M1/M2/M3 OSEM2EUL and EUL2OSEM values from the output of make_sushlts_projections.m. I set SENSALIGN and DRIVEALIGN to the identity matrix. I set the signs in M1 COILOUTF to give +ve sensor response for +ve offset. The final values were -1, -1, +1, +1, -1, +1. This agrees with the settings previously used on the triples test stand but differs in SD from E1100109-v2. I set the signs in M2 and M3 COILOUTF per E1100109-v2. I found by trial and error a set of damping gains that gave stable damping, but these can probably be improved: -0.5, -2, -1, -0.02, -0.01, -0.04.
Mark B. According to Jeff B, the apparent sign flip for SD is expected - the magnets are put on per E1100109-v2 but the OSEM is on the opposite side for HLTS (the right as viewed from the back), so it sees the opposite sign of magnet.