MODS1 AGw Commissioning at LDG: 15 November 2010 UTC
Observer(s): R. Pogge (OSU) and M. Pedani (LBTO)
Telescope Operator: S. Allanson (LBTO)
AO Support: T. Leibold and D. Miller (LBTO remote from Tucson)
Telescope Support: M. Midkiff (LBTO)
Instrument Support: T. Shih (LBTO)
Clouds rolling through at sunset, so a delayed opening. Started an IMCS telescope scan before sunset,
will get about 4 sequences before starting.
Got two Collimation table files by using the mkcoll_irtc.pro script. Conditions were terrible but we wanted to exercise the procedure
and find any bottlenecks for MODS. See files: /home/lbcobs/Collimation/new/Collimation_MODS20101115.dat and Collimation_MODS20101115_optimized.dat for
an analysis of what we got. These were not
implemented since we have to discuss them with JH (Global offsets etc.) on next monday.
- 0.5h to two failures of the LDG rotator. Fixed by swapping fibers on a communications module. See IssueTrak #2481.
- 2h to failure of the azimuth drive. Pressure was increased to try to get us through the rest of the night. See IssueTraks 2988, 3002, and 3003.
Got a set of guide probe transformation data and astrometric calibration images using a section of the Stone E astrometric reference field.
The 20 star transformation run was a full shakedown of the scripts, but seeing was less than ideal. Hopefully we can use the many
stars in this rich field to get better precision.
Winds increased after about 2am, we closed at 5am due to high winds.
Dewars filled at 2240 UTC (Nov 14). So far the hold time is longer than 24h, at each roughly
24h interval thus far there is still liquid in the reservoirs. We will be taking data this week with
the environmental monitors to quantify this during this week, and between now and the
December 14-20 commissioning run.
IMCS Flexure Testing
- Starting an IMCS telescope scan in both MODS channels
- Select a rotator position angle
- At fixed rotator angle, scan the LBT in elevation 90 to 30 and back in 15-degree increments
- At each elevation, allow the IMCS to lock and take a sieve mask image in both channels
We repeat the sequence for a set of rotator angles. Tonight we have time for 4 sequences,
each of which requires about 20-22 minutes per sequence. We will run 210, 300, 180, and 30
degrees rotator angle, which represents the cardinal points of PA on the meridian.
- Opened up just after sequence 3 (rot=120) was complete.
- started 4th and final sequence for today (rot=30)
- Weather Today: T_air =-2.6 RH=38% T_Glass=-0.1 Wind ~9 m/s. Many clouds now.
- Preset to pointing star WT10-323 (parallactic acquire)
IE = -12
CA = 37
Torsten has loaded the newest SExtractor file, everything else looks set. We are using BP6
- BS9188 for collimation
Fat and comatic at the start, but slowly starting to converge and correct, seeing looks poor: 1.8 arcsec.
Mirrors are 2.5C warmer than outside still and coming down - ventilation is on. We have some
thermal catch-up to play.
- guide star lost, seeing was degrading through 2.5 to 3 arcsec just before that.
Oops, I was running a bias calibration test script while the seeing was still poor and forgot that the script contained a
"hatch close" command. Mea culpa.
- Seeing 2.6-3 arcsec. Wind is up, now above 10m/s with gusts up to 14m/s so the vent doors are closed. Mirrors are within
0.2C of the local ambient air.
- Mirrors still about 0.6C above the air temp. Seeing improved a little to 1.8-arcsec.
- Mirrors now flatter in temperature, seeing more like 1.4-1.6 arcsec, so continuing to improve.
The all-sky camera autogain system has gone awol again, so despite having moon we can't tell what clouds are around
except visually - something should be done to just grab the FITS images and short-circuit the stock software.
- Seeing still not good, dress rehearsal of a collimator map run to get the procedure down.
- preset failed, SExtractor labeled star as flag=28, in the new code it rejects Flag larger than 16. Torsten restored the
old flag threshold and present succeeded. The flag threshold was too stringent for out-of-collimation images.
- Doug tweaked a parameter in the WFS analysis and we reran the preset. The correction is better.
- finished the sequence for what it's worth. No stars at El=70 in our catalog. Marco is analyzing, and we're waiting for Doug who is in
transit to the remote control room in Tucson...
Guide Probe Coordinate Transformation
- Doug is in Tucson and we're ready to start working the coordinate transformation
Test frames in the Stone-A1 field, guiding on a 16.5 mag star. Seeing is still poor, about 1.5-1.8 arcsec. Exposure times of 10-20sec bin 2x2 will be fine in this seeing for a run at the coordinate transformations. The initial work will be to shakedown the scripts.
- Rotator fault, Steven had to restart MCSPU to recover.
- Got a warning about loss of sequencer access from the MODS1 blue channel IC. Repeated SEQINITs did not clear, had to quit and warm restart. Problem cleared. First encounter with this after Bruce swapped out the old Blue IC before the run, but the character is different than with that unit. Will watch.
- Repeat of the rotator fault, D.Cox consulted on the phone. Steven swapping fibers on the communication module in the Aux Control Room and then informiing MCSPU of the problem.
- Rotator is back online now after the swap. IssueTrak
#2481 appended with this incident, and priority elevated to high.
- Doug back to running the script test (preset and expose with MODS red science camera)...
- Seeing still pretty poor, temperature has dropped to -5 and mirrors are not quite keeping up. Script testing continuing, seeing is pretty bad and making
the WFS convergence slow.
- Preset failed after moving the star onto the WFS pickoff. Best we can tell is that the seeing is so poor and the star sufficiently faint that when the star went into the 50/50 beam splitter, it dropped below the detection threshold. Marco pulled up the acq image off the repository and this is our best interpretation - the star is there, but the seeing puffed and it is below threshold. The images in question for the acquisition are: left_acquire000062.fits, 63, and 64 (the sequence is acquire, center, and then offset probe into WFS pickoff). The image that failed was 64.
- Moving to another field around BS9107. A much richer field, and higher up.
- Ran into primary mirror limits on the preset, Steven backed out the mirrors by hand, but now out of collimation. Convergence is very slow as a consequence.
- Doug changing WFScam_hotspot_x from 368 to 370 in the GCS config file for MODS. We were having a problem with the off-axis pupil. Failed on the move into the pickoff. This is acquisition image acquire000070.fits.
- Got an acq fail on sequence 71, 72, and 73, the latter on the probe move. Rotator was PA=90. The probe offset sent was correct but, the star fell short of the box.
We repeated with PA=0 and a different guide star and it succeeded (images 74, 75, 76), then PA=-90 and a third guide star, and this moved the star into the WFS pickoff
there (images 77, 78, 79) and succeeded. Image 79 shows the WFP pickoff in about the right spot for the guider to be able to finish putting it onto the hotspot. Consulting
the logs on MODS1 shows that the stage offsets to go from guide center to probe position were correct.
- The Az torques maxed out, so the drive stopped. Steven restarting the MCSPU, reported in IssueTrak
- Tried an PA=180 preset. Got the same Az failure. Steven restarting again...
- Got Mike up to investigate the Az failure. Still no joy... Lots of phone calls down to Tucson...
- Focus was on regaining control of the HBS. Control of the telescope seems to be coming back, but not clear at what level.
- Going to a pointing star (parallactic acquire mode).
New pointing constants:
CA = +40
- Preset (active) to Stone E field at 07:45/+00
Slewed OK, started guiding OK, WFS collimation started... we are hopeful
- Doug taking over WFS correction, PA=-90. Seeing about 1.5 arcsec still. Humidity is rising to about 68% and temp is -5C
- Doug's transformation data acquisition script running OK for 4 stars for the first test run. It ran successfully. Seeing about 1.2 arcsec.
Pupil centering on the WFS is not quite correct, Doug is tweaking it...
A complicating factor looking at the first 4 test images is that the "brightness veto" feature of the GCS, whereby it assumes the brightest star
to appear in the guide field is of necessity your guide star (even if it isn't) tends to pull to the wrong star for this measurement. It complicates
selection of the test stars for the transformation, as we have to make sure there isn't a brighter star nearby that might confuse it.
- Starting the "monster script" with 20 stars. Preset to PA=-90, first collimation image will be wfsc000022. wfsc000024 first image while taking
a MODS red image (mods1r.20101115.0073).
- After image mods 0077 and wfsc 0027, script was stopped by a star over the edge of the patrol field (1mm over the line).
- MODS image 78 corresponds to wfsc 031 after the restart and recollimation.
- MODS image 86 corresponds to wfsc 039. We are at number 16 out of 20 in the sequence
- Sequence complete, last image MODS 91, wfsc 044.
- Astrometric calibration image, PA=0. Started with an on-axis acquisition to tweak collimation.
The seeing is worse - we noticed the wind is now up to 15 m/sec.
Second active preset to an off-axis guide star. Seeing variable 1.3-1.5 arcsec. Will keep binning 2x2 on these images.
Taking images in ugriz. Failed to notice the first two blue were 8x3K bin 2x2. Oops.
- Going to the azimuth "sweet spot" to try to assess the true wind speed, since we're seeing gusts at about 17m/sec, and hearing the wind against the building.
Looks like the mean speed is 13-15m/sec. We'll continue to watch it closely.
- Bin 1x1 at Stone E PA=0, g and r filters 60s astrometric calibration frame.
- Same, but PA=180 - wind still gusting, but staying below the limits.
- Gust over 20m/s, we're closing due to wind.
- 14 Nov 2010