March BG Com Night 2, March 7-8 2009 MST
Observers: J. Hill, A. Rakich and T. Schmidt (Tucson)
Lucifer support: M. Juette (on-call)
Telescope Operator: W. Wack
SW Support: M. Delapena, T. Leibold, N. Cushing (Tucson) plus P. Grenz (remote)
Telescope Support: J. Rousseau, T. Shih
Worked until 1AM getting M3 into position. Still not sure what exactly the problem is.
Then spent another 3 hours trying to acquire a star. Finally worked out that the M3 was at the wrong selector angle. We turned the UMAC back on and remarkably we were able to initialize and move the selector to the proper position.
The next crisis is that no starlight is making it to the WFS. See critical issue #2056. We need the
Work Platform mounted on the left side, and we need the calibration light source mounted on the AGw.
This should take priority over all other telescope activity for Sunday.
The only good news is that Michele fixed the PCS problem with back-calculating the RA/DEC.
It was caused by a change in the rotator position update rate.
Details (times reported in MST below)
In anticipation of seeing real stars, Torsten has put GCSL back in real stars mode and not starOverlay simulation. (controlled in lbt.conf)
18:30 Open Shutter Doors
18:44 Switch M3 (both) out of simulation mode for OSS.
18:54 Stop and restart OSS. Selector amp for SX is M3 is still faulted. Startup and init, but that seems to hang.
So we send Wayne up to recycle the power.
19:04 Stop and restart OSS. Kill and restart our frozen GUI. Amp was not initially faulted, but it faulted as soon as we tried to do "init".
19:10 Sent Tim to measure the UMAC power supply voltage.
19:28 Paul runs initialize script without OSS. It fails again with the script.
19:30 Again .....
20:00 Tim measured UMAC power supply voltages, and the +5VDC for the Selector Hall Sensors without joy
21:00 Turn off the UMAC to let it cool for half an hour.
In the meantime, Michele has installed and tested a new version of PCS to fix the back-calculation problem.
21:40 Turned UMAC back on, and now it seems to have a fault immediately rather than when we do the init. After several tries we did get the selector to home once, but then it failed on the other axes at +limits, so M3 is far out of the needed position.
22:00 Switch Left M3 onto Right Side of UMAC, but we get the same selector fault.
22:45 Put OSS into full M3 simulation mode
22:54 pointto BS9122 ACQUIRE - failed, OSS not at focal station
22:57 Authorize LBC, then LUCIFER.
22:58 pointto BS9122 ACQUIRE - succeeded, but GCSL not getting AzCam
23:00 stopAGW, startAGW, home, restart GCSL
23:05 pointto BS9122 ACQUIRE - still no images.
23:xx pointto BS9122 GUIDE -GCSGUI displayed a simulation guide image from last night first.
23:xx pointto BS9122 GUIDE - GCSL failed to identify that there was no star in the acquisition image.
23:20 Close dome. Then move to zenith to try to install spare M3 cables.
23:28 OSS back to real (not simulation) mode.
00:15 Ran tests with temporary cables bypassing to SX M3, but we still get the same selector amp fault
after we start init.
00:30 Had problems getting OSS to talk to the UMAC.
00:40 Tried some more combinations of power cycling and finally got all 4 axes to home.
Moved them to -615 arcsec -20 arcsec 0 piston and -8.35 deg selector. We later learned it should be +8.35 degrees.
00:45 Turned off power and removed temporary cables. Reconnected permanent cables, but
left power to UMAC off.
00:52 Editted lbt.conf to put M3 (both sides) back in full simulation mode.
Authorize LBC, then LUCIFER again.
01:00 Deploy M3 swing arm from OSSGUI. Doesn't work until we turn on the drives.
01:10 Retract stow pins, but doesn't work until we drive the telescope up to exactly zenith.
We shouldn't have retracted the swing arm with the oil on, as this makes the telescope lean over
on the stow pins.
01:20 Reopen the dome.
01:25 Preset failed for not at a focal station. Authorize LBC, then LUCIFER again.
01:29 pointo M5_0444 gs=0 ACQUIRE PAM- - Preset succeeded, but GCS is not getting
any image, or is not displaying any image.
01:33 pointto M5_0444 GUIDE - doesn't find any star
01:34 pointo M5_0444 gs=0 ACQUIRE PAM- - Preset succeeded, but GCS is not getting
any image, or is not displaying any image. Must be a new bug. Andrew made a new Issuetrak.
01:45 start spiral search taking images from command line and displaying them with Guider.py.
Failed to find the M5 star after the full 2 arcminute radius search.
There are some clouds, but not enough to stop use from seeing an M5 star.
Tim claims he took off the retroreflector from the AGw. Andrew claims he did not install the AGW dust cover. Shutter door is open. Guide camera can see some lights in dome. M3 does not have a cover.
M2 cover was off yesterday. M2 is correctly collimated. M1 is correctly collimated within 0.1 mm.
M3 was homed and repositioned the best we know how (This turned out to be the error.).
Prime focus swing arm is retracted, M2 and M3 swing arms are deployed. Telescope is tracking OK.
Shutter doors are open. Guide probe is at the center of the field.
02:10 Telescope collides with building bumpers. Why?
02:25 Wayne is repeating the encoder reset.
02:42 pointto M5_444 TRACK, then reposition probe with setGSsfp -x 0 -y 0
02:44 100 sec exposure gives 11K counts, so we are seeing the sky.
02:47 start the spiral again with 1 sec exposures. Didn't find anything out to 90 arcsec radius.
02:56 Slew to zenith for visual inspection of M3 position. Wayne's description sounds like it is
out of position.
03:20 Slewing to moon for the final sanity check. Moon also seems to be missing AGW, but the result isn't very definite. In the meantime we read OSS.conf and discover that Dave/Paul gave us the wrong selector angle earlier and it should be +8.35 degrees (emphasis on PLUS
03:48 Change lbt.conf and restart OSS
OSS didn't connect to UMAC on the first try.
But, we were able to move the Left M3 without getting an amp fault.
04:04 pointo M5_0547 TRACK - failed for no focal station; authorized LBC/LUCIFER
04:07 pointto M5_0547 TRACK - setGSsfp -x 0 -y 0 FOUND A STAR IN THE CORNER
Adjust IE=-13 CA=-20
04:17 pointto BS9177 gs=0 ACTIVE - acquired the star, but preset failed because star too dim.
04:23 pointo BS9152 gs=0 ACTIVE - acquired the star which is brighter, but GCS claims too faint.
In DS9 we see 1000 counts above background. GCS reports that sextractor didn't find any star
in the guide image at 04:25:06 MST. See Issue #2058.
04:32 edit lbt.conf to use GCS internal centroiding rather than sextractor.
04:36 pointto BS9177 gs=0 ACTIVE - succeeded, but a few seconds later we lost the guide star because of clouds. The guiding comes and goes, but WFSing isn't working.
GCSGUI doesn't display the WFSimage. See Issue #2057.
04:49 Manual collimation still doesn't give us counts on the WFS, so we switch to gs=12
05:50 Even after we went back to M5_0547, we are getting any focussed starlight on the WFS chip.
So the conclusion is that the Laser Tracker activity screwed up fold mirror 5, and the starlight
is missing the pinhole, or something like that. See Issue #2056.
05:55 Close the dome, although Wayne had problems with the left shutter door.
06:00 Talking to Jesper on the phone to work out a strategy.
- 02 Mar 2009