Commissioning Notes from 11-12 May 2008 MST

Observers: J. Hill and J. Storm on mountain, plus Biddick, Delapena, Grenz, Hooper, Cushing in Tucson
Telescope Operator: A. Ceranski

Summary

Spent the first 60% of the night chasing down various problems and or tests related to the new TCS features. There were plenty of minor issues to deal with, so the software folks in Tucson were not bored. The simple summary is that all the new features work: north is up at PA=0 for ds9 display of IRTC images, pointing free coma correction seems to work, as does coma free pointing correction. The pointing model coefficients can be adjusted on the fly. GCS can offset the star to the hotspot (we think). We were guiding on-axis at all position angles, but guiding was not very good. That may be a combination of high wind, really bad seeing (1-3 arcsec), and some problems with GCS. See the guiding discussion at the bottom of this note.

The next part of the night was spent adjusting the configuration file of the AGW so that a better x,y value of the center position was used. After we changed the center to 0,612.5 we saw a large out-of-focus pupil. We changed the configuration back, and we still saw the out of focus pupil. Finally we worked out that this pupil was a ghost, and the real bright star was 25 arcsec away. It was away with the old coefficients because John made a typo and reversed the sign of X on the probe positom. It was away with the new coefficient because Jesper made a mistake in the conversion of encoder coordinates to X-Y that was exactly the same amount as John's typo. It took us a couple hours to sort that out.

We got back to guiding for the last half hour or so. See the notes at the bottom.

Nighttime Activities

Times reported in UT for 20080512.

01:20 D. Ashby completes a balance run for the Gregorian configuration.

02:10 Open Dome, sky mostly clear with a few scattered cirrus. Wind around 10 m/sec. T=8 degC.

02:25 Penguins over Flyers 4-2 with an open-net goal in the closing seconds! (We were in danger of postponing observing if they had gone to overtime.)

Jesper set the UMAC temp thresholds to -20, 65.

Testing the new TCS Build

02:30 Having a reflective memory problem that seems related to the new TCS build this afternoon.

03:15 Reflective memory seems straightened out now, but software folks are still debating why the build/install went wrong.

03:20 IRTC camera connected now, but we had to cycle the power in the treehouse a second time.

03:27 pointto WT10_292 - Needed a spiral search, found at offset 60,20 before collimation, at offset 60,0 after collimation.

04:01 by-eye collimation finished. Seeing is fairly bad. Still showing vibrations from the swing arm. The new pointing-free coma corrections in PSFL seem to work.

04:04 pointto WT10_292 again, now centered at 60 -10.

040640 takepic 0.1 at offset 60 -10 050730 takepic 0.1 at offset 60 -20 (image moved down on live IRTC display) North is up on the DS9 image.

04:17 now at offset 55 -25. ptmodify IE claims success, but doesn't modify the trajectory.

04:25 RY=+15, reduces offset to 20 -20.

Short break here for John and Jesper to finish their dinner, for Michele to rebuild PCS, and for Aaron to turn down the speeds of the fans in the treehouses.

04:45 Tested the new PSF command line command adjustPointing.

05:00 pointto WT10_292 again.

05:10 We are centered now with no offsets, and no global tilts. This is with IA=-450 and IE=-32 (previously -263, -30). Now the problem is that we have X=+2.600 mm

Images are certainly showing less evidence of vibration, but there may be vibration induced by the wind. [Note added the next day: It seems that one of the lower treehouse fans was still running at high speed.]

05:14 adjustPointing left M1 -5 0 (X=-0.233 RX=+5), then readjust IA=-415.

pointto command failed fro M5_1332 gstar=2 , with some problem about /par

05:27 pointto M5_0531 gstar 5 landed star within 2 arcsec of center of IRTC, but GCS failed with RPC error talking to AGw.

05:31 stopped and restarted oacserver

05:35 pointto M5_0531 gstar 0 on axis, but GCS reports transfer of correction to PCS failed. Besides that, the star was too bright.

05:57 pointto WT10_279 gstar 0, but this star is over at the left edge of the acquisition image. Adjust IA=-445.

06:05 Stephen fixes the missing runtime library needed for the guider_image command.

06:11 Michele gives us new version of GCS with the error commented out.

06:12 pointto WT10_279 gstar 0, irs reports RPCERROR (from where?), GCSL died unexpectedly.

Guiding Tests

0616 Restart GCSL and pointto again. Rotangle=155 deg. GCS is guiding (not running away) but star is oscillating wildly inside the box. It looks like the correction could be off by 90 deg?

06:26 Discover PCS "guide" button was off. Turn in on, and the image stabilizes somewhat.

May 11 23:31:05 lbtmu107 LBT_Networkserver: GCSL died unexpectedly

06:34 Restarted GCSL and pointto WT10_343 gstar 0 (trying to move out of the wind)

06:38 spiral search

06:40 setfocus -p 24; setfilter -f 3

06:58 IA=-420

06:59 pointto WT10_343 again. This time we acquire guide star, but still bouncing around.

07:02 pointto WT10_343 PA=90 but hit rotator limit.

07:06 pointto WT10_343 PA=-90 Found it there, but then couldn't find star at PA=180. Then couldn't find it at -90. AI=-470

07:24 pointto WT10_343 PA=-90 (Guide rate = 0.1 Hz in LBT.conf, apparently doesn't do anything)

07:26 pointto WT10_343 PA=180

07:30 pointto M5_0555 at several PAs ending with +90 where GCSL is happy

07:49 IA=-480 IE=-35 May 12 00:49:41 lbtmu107 LBT_left_GCS: fatal error: GuidingThread: guide image didn't arrive within timeout of 1388 ms. But, then it restarted on its own after ~20 sec ????

07:51 PCS died when I did a ptmodify command.

Restart PCS

IA=-480 IE=-40

07:59 PCS dies again on ptmodify IE -25 (a special deadly command?) 0 and -30 are OK -10, -25, -26 kill PCS

IA=-500 IE=-30 Star in corner of IRTC

May 12 01:08:18 lbtmu107 LBT_left_GCS: TCPSocket::recv error: Bad file descriptor May 12 01:08:23 lbtmu107 LBT_Networkserver: GCSL died unexpectedly

May 12 01:24:08 lbtmu107 LBT_Networkserver: GCSL died unexpectedly

08:30 We send the software people home.

08:35 Stop GCS on purpose while we are adjusting constants in the AGw configuration file.

Now the center coordinates are X=0.0 Y=612.5.

09:02 edit same values into AIP_L.cfg so GCSL can read them.

09:08 Clear active optics and recollimate

Z8=+3400 nm (in the new pointing free definition) and it really is pointing free within the arcsec or so of image motion that we have.
adjustPointing left M1 -10 0, which gives X=-0.465 and RY=+10, more negative moves the live image left at this position
All global offsets are back to zero.

09:21 I checked to make sure the adjustPointing moves with M2 are same amplitude and same direction in both axes.

09:33 IA=-340 IE=-32 (nominal) This azimuth term is somehow connected to the large coma term, but I don't understand how it is related to 2 weeks ago.

Adjusting configuration for probe, and problems with pupils and ghosts

09:35 pointto M5_0555 for guiding on axis. But why does the acquisition image show a huge pupil?

May 12 02:42:32 lbtmu107 LBT_left_GCS: fatal error: GuidingThread: guide image didn't arrive within timeout of 1388 ms. May 12 02:42:32 lbtmu107 LBT_left_GCS: TCPSocket::recv error: Bad file descriptor May 12 02:42:32 lbtmu107 LBT_left_GCS: guide image file name = /tmp/guiderimage002684.fits May 12 02:42:32 lbtmu107 LBT_left_GCS: guide image is invalid! May 12 02:42:53 lbtmu107 LBT_Networkserver: GCSL died unexpectedly

./setFilter 2
filter should be no. 2
05/12 2:57 LBTO@lbtmu107 /home/telescope/GCS> ./getFilter
filter is 4

The above caused us to stopAGW and stop oacserver, and restart everything.

10:25 The out of focus pupil was a ghost, and we find the star at offset 25 0 at PA=0 even though it was at offset 0 0 on IRTC. ????

10:40 The star is at offset 0 -25 on guider at PA=90. Guider probe is back at the old coordinates of -7 600.5.

10:50 Stop and start AGW and oacserver again. Home motors.

10:57 offset command hangs with rotator at 351 deg. Had to kill PCS and restart. IA=-320.

11:11 oops, John was using -7.5 600.5 when the old position was +7.5 600.5.

11:21 back to the new zero values of 0, 612.5. But now we are missing the star again!! (apparently not all John's fault). Found it again at offset 25 0 at PA=0. So somehow we have the same error again in X ???

11:30 It turns out that Jesper had made an error in the encoder conversion to X-Y that was exactly the same as what John had made

11:34 pointto WT10_396 gstar=0. It guided for a few iterations and then GCSL died unexpectedly.

11:39 changed guide filters to 0.2 Stop and restart PCS. No evidence that these guide filters do anything at all.

More Guiding Tests

11:53 Guiding on this star, but still jumping all over.

11:56 setGuideRate to 0.1 Hz, and it slowed to 7 sec intervals instead of 1 sec.

11:57 setGuideRate to 0.3 Hz about 3 sec intervals. Az=-23 EL=48 Rot=36

12:01 setGuideRate to 0.05 The jumping seems more calm,

12:05 took a manual image and crashed GCS guiding thread.

05:10 Close dome.

The main problem with the guiding is that it seems to have the star systematically at the wrong X position even though it clearly know the location of the star relative to the correct hotspot. Can anyone explain the calculation below? Why is the final guide correction in Y, when the centroid error is in X?

May 12 05:04:02 lbtmu107 LBT_left_GCS: guide image file name = /tmp/guiderimage003460.fits
May 12 05:04:02 lbtmu107 LBT_left_GCS: ESO succeeded
May 12 05:04:02 lbtmu107 LBT_left_GCS: star offset from hotspot: 10.206779, 0.063881   [This is correct value in pixels.]
May 12 05:04:02 lbtmu107 LBT_left_GCS: star offset in SFP: 0.552748, 0.003459              [This is the value in arcsec.]
May 12 05:04:02 lbtmu107 LBT_left_GCS: star SFP should be: 0.302532, 0.275795
May 12 05:04:02 lbtmu107 LBT_left_GCS: centroid img: 281.793213, 254.936127 (abs) 39.793221, 49.936119 (rel)    centroid sfp: 0.081433, 0.274412
May 12 05:04:02 lbtmu107 LBT_left_GCS: star delta: -0.495728, -0.075049
May 12 05:04:02 lbtmu107 LBT_left_GCS: pcsClient.guider.applyGuideCorrection(0.081433,0.274412)

Meanwhile, the PCS reported:
May 12 05:04:02 lbtmu102 LBT_PCS: Centroid Xarcs: 0.135720993081729 Yarcs: 0.457352697849274 Xmm: 0.081432595849037 Ymm:0.274411618709564

Plan for tomorrow night

1) Try to close the Active Optics loop with Doug. Hopefully the seeing will be better.

2) Try to understand the remaining goofiness with on-axis guiding.

3) Position the probe manually to locate off-axis globular cluster stars. This should let us map the probe coordinate system.

-- JohnHill - 12 May 2008
Topic revision: r6 - 13 May 2008, JohnHill
This site is powered by FoswikiCopyright © by the contributing authors. All material on this collaboration platform is the property of the contributing authors.
Ideas, requests, problems regarding Foswiki? Send feedback