UT 20130203 Technical Observing

Observer: JHill (Tucson)
Telescope Operator: DHuerta (LBTO)
Additional AO Support: DMiller (Tucson)
Telescope Support: JLittle (LBTO)
Instrument Support: JMorris (LBTO)

Summary

A GCS problem that turned out to be something wrong with a new reconstructor configuration cost us 1.5 hours at the beginning of the night.

We did a long (30 minute) open-loop tracking and collimation and astigmatism test in magnificent seeing (0.35 arcsec on guider), plus another (20 minute) test in a different part of the sky. The results were good, but the bumps and wiggles were also interesting. See plots and summary below.

We installed a refined pointing model based on the data from last night. We verified that it was working OK at LFBG on 12 stars.

We had to close at 05:25 UT because of heavy clouds.

John's note to telescopework

Dear Telescope,
        Over the course of the last 2 weeks, we've installed new
pointing and collimation models on the SX side of the telescope
to cure the various issues with SX field aberration.

ALL SX Gregorian and Bent Gregorian focal stations (everything but
LBC) now have the pointing model SX_Gregorian.20130202.ptmod.  This
was a pointing model measured from 55 stars in stable conditions with
AGW1 @ LFBG (using the collimation model described below).  The rms
fit was 4.5 arcsec (including limitations from the AGW1 transform).

Those same Gregorian focal stations use an improved collimation model,
which is the "Ashby3" model (known to Ashby as
Update_Collimation_NO_Pointing_Sep2012_plus_LT) plus a small
correction for the coma zeropoints measured on-sky.  This is the model
that was implemented to correct the field aberrations and included
pointing that we've been fighting since November.  The binodal
astigmatism is matched to the Laser Tracker measurements, and the
binodal has been confirmed to be low on-sky at multiple elevations.

Ciao,
        John (plus Doug, Dave, Dave)

P.S. The DX side of the telescope has the same old (Dec 2011) pointing
and collimation models.  The pointing model is a bit stale, but we've
confirmed that the DX side does not suffer from the field aberrations
that were plaguing the SX side for the last 6 months.

Preparations

Dave T. tested LUCI1 all day long.

Doug M. is analyzing the pointing model data from last night, and the results look better than those from 20130124.

Implemented as SX_Gregorian.20130202.ptmod
(Doug's name was - LUCIFERSX.20130202.ptmod)

       coeff	   change     value    sigma

  1     IA         -0.233   -394.00   12.395
  2     IE         +0.092    -90.05    2.829
  3     NPAE	   -0.468    +92.82   16.034
  4     CA         +0.485    -49.34   19.272
  5     AW         +0.238     -3.66    0.716
  6     AN         +0.079    +23.84    0.743
  7     TX         +1.172    +50.03    4.539
  8     TF         -1.986   -124.15   10.599
  9     A1A        -0.129     +8.10    6.288

Sky RMS =   4.45
Popn SD =   4.92

LBT left Pointing Log, 2013 Feb 2
T   50   4.4546    0.000   0.0000
  IA       -393.9951    12.39474
  IE        -90.0463     2.82941
  NPAE      +92.8218    16.03368
  CA        -49.3423    19.27163
  AW         -3.6550     0.71631
  AN        +23.8383     0.74347
  TX        +50.0340     4.53852
  TF       -124.1523    10.59893
  A1A        +8.1012     6.28838

Details

Open

00:55 Open shutter doors

01:08 Start mirror ventilation

Sky is full of cirrus, while heavier clouds are approaching from the west. Temperature is about 4 degC.

Initial Collimation LUCI1 @ LFBG

ACTIVE preset

01:35 Restart GCSL as guide image was doing funky things - alternateing bright/faint images.

ACTIVE preset - now the full frame WFS image is displayed, and we are not doing active optics.

01:40 Stop and restart AGW1.

From tcslog - these messages started at 01:34:xx when we restarted GCS.
Feb  3 01:54:54 tcs4 LBT_left_GCS: WFS: WFS full image file name = /Repository/GCSfiles/dailyimages/left_wfscimage000058.fits
Feb  3 01:54:54 tcs4 LBT_left_GCS: WFS: Subimage could not be created, viewing full WFS frame in GUI instead 
Feb  3 01:54:54 tcs4 LBT_left_GCS: WFS: createSubImage(386, 208, 254, 254) failed.
Feb  3 01:54:54 tcs4 LBT_left_GCS: WFS: supposed to attempt new WFS approach but Sextractor object not instantiated!
Feb  3 01:54:54 tcs4 LBT_left_GCS: WFS: using old style WFS instead ...
Feb  3 01:54:54 tcs4 LBT_left_GCS: PROFILING: WFS: triggerOldWFS() entered.
Feb  3 01:54:56 tcs4 LBT_left_GCS: reading WFSingThread configuration from '/home/telescope/TCS/Configuration/GCS//LUCI_L.cfg'
Feb  3 01:54:56 tcs4 LBT_left_GCS: trying to use Source Extractor for WFS.
Feb  3 01:54:56 tcs4 LBT_left_GCS: WFS: sext hotspot: 273.00, 125.00
Feb  3 01:54:56 tcs4 LBT_left_GCS: Assuming WFS_xy_transpose is false.
Feb  3 01:54:56 tcs4 LBT_left_GCS: WFS_lenslet_rotation set to -1.30 degrees
Feb  3 01:54:56 tcs4 LBT_left_GCS: WFS_redefine_hex_z configuration for AGW is: '456'
Feb  3 01:54:56 tcs4 LBT_left_GCS: WFS_redefine_med_z configuration for AGW is: '78'
Feb  3 01:54:56 tcs4 LBT_left_GCS: Assuming WFS_pupil_dir_x = +1
Feb  3 01:54:56 tcs4 LBT_left_GCS: Assuming WFS_pupil_dir_y = +1
Feb  3 01:54:56 tcs4 LBT_left_GCS: Assuming WFS_pupil_dir = +1
Feb  3 01:54:56 tcs4 LBT_left_GCS: WFS_subtractFlat = 0
Feb  3 01:54:56 tcs4 LBT_left_GCS: WFS_subtractBkgrd = 1
Feb  3 01:54:56 tcs4 LBT_left_GCS: WFS: reading WFSingThread Zernike scale factor configuration from '/home/telescope/TCS/Configuration/GCS//LUCI_L_RM_sub11.cfg'

From event log:
Sun Feb  3 01:54:10.801 2013 56326.07929168583 5 GCSL gcs.wfsthread.alert left pauseLoop request received, WFS cycle interrupted
Sun Feb  3 01:54:12.817 2013 56326.07931501307 5 GCSL gcs.interface.azcam.wfsc.readout.started left WFScam full ccd readout started
Sun Feb  3 01:54:25.196 2013 56326.07945828894 5 GCSL gcs.wfsthread.alert left pauseLoop request received, WFS cycle interrupted
Sun Feb  3 01:54:27.210 2013 56326.07948160599 5 GCSL gcs.interface.azcam.wfsc.readout.started left WFScam full ccd readout started
Sun Feb  3 01:54:39.617 2013 56326.07962520729 5 GCSL gcs.wfsthread.alert left pauseLoop request received, WFS cycle interrupted
Sun Feb  3 01:54:42.132 2013 56326.07965431301 5 GCSL gcs.interface.azcam.wfsc.readout.started left WFScam full ccd readout started
Sun Feb  3 01:54:54.509 2013 56326.07979756566 5 GCSL gcs.wfsthread.alert left pauseLoop request received, WFS cycle interrupted
Sun Feb  3 01:54:56.523 2013 56326.07982087568 5 GCSL gcs.interface.azcam.wfsc.readout.started left WFScam full ccd readout started
02:18 stop GCSL, GCSR, PSFL and restart network servers on tcs4.

See this on tcs4 console where telescope did netconfig start:
IST: accepted connect
image server receiving image.
CamImage: createSubImage: coordinates do not match up
xmin: 259  xmax: 513
ymin: 81  ymax: 335

02:55 After bunches of restarting we reverted one round of GCS configuration file and that made active optics happy.
We moved LUCI_L.cfg to LUCI_L.cfg.short (the one JMH editted at 11:41 UT to shorten wfs exp time).
We moved LUCI_L.cfg~ to LUCI_L.cfg (the one from 09:13 UT before Doug changed reconstructors).

diff LUCI_L.cfg LUCI_L.cfg.shortwfs
35c35
< WFScam_exposureTime	30000		int	# ms of exposure time that's set by default
---
> WFScam_exposureTime	10000		int	# ms of exposure time that's set by default
39c39
< WFS_subDir		"20120816"		string	# subdirectory relative to the path of this config file where to find WFS specific files
---
> WFS_subDir		"20121213"		string	# subdirectory relative to the path of this config file where to find WFS specific files
03:00 At EL=53.8 the X part of coma on M2 is 0.1 mm, and the Y part of coma is +0.3 mm. That means our collimation model offsets are good (about the average of 3 nights).

So the problem is something in Doug's new reconstructor directory that only gets read when GCS restarts. We don't know what was initially making the guider image goofy.

Open-loop tracking and open-loop collimation and open-loop astigmatism test @ LFBG

These open-loop tests were taken to characterize current telescope performance for Pathfinder, and for general diagnostic purposes.

First Test

03:05:04 ACQUIRE preset to BS9107 gs=0 - Center pointing on this star IE=-54 CA=-5

03:09:54 ACTIVE preset BS9107 gs=0 (on-axis) at AZ=292 EL=73 Seeing =0.35 arcsec on guider. multi_step_focus is running to give collimation updates while active optics is blocked. Initial Z5,Z6 on M1 was -852,+2091 nm. Initial Zactive on M2 was -0.170 mm.

03:14:00 block guiding and zernikes for open loop tracking test.

03:44:00 resume guiding at AZ=289.6 EL=67.6
03:44:34 resume wfs

03:47 On M1 Z5=-1108, Z6=+924 nm; Zactive on M2 =-0.114 mm

Second Test

03:56:50 ACQUIRE preset to BS9118 gs=0 - Center pointing on this star IE=-57 CA=-13

04:01:39 ACTIVE preset BS9107 gs=26 (far off-axis = 148.7 mm) (missed star by ~ 3 arcsec)

04:08:00 block guiding and zernikes for open loop tracking test. AZ=162.1 EL=55.3 Seeing =0.50 arcsec on guider. multi_step_focus is running to give collimation updates while active optics is blocked. Initial Z5,Z6 on M1 was -1578.3 +523.7 nm. Initial Zactive on M2 was -0.079 mm.

04:28:00 resume guiding at AZ=170.4 EL=56.3
04:28:44 resume wfs

04:33 On M1 Z5=-1187.7 Z6=+608.6 Zactive=-0.090 mm

Summary of Open-Loop Test Results (JMH:20130221)

We did open-loop tests at two positions in the sky for 30-minutes and 20-minutes respectively. The atmospheric conditions were quite stable, with thin cirrus, so we got good centroid data on the uncorrected guide star. Guiding corrections were blocked, while the guide loop still recorded images. The star positions as measured by GCS with sextractor are used to measure the drifts in the open-loop tracking of the telescope. Active optics corrections were also blocked, so we were also measuring the open-loop drift in collimation and focus. The multi_step_focus script was running, to provide collimation LUT updates every 10 sec or so. The AdSec was also updating the astigmatism LUT open-loop. This night with cirrus clouds was thermally stable with temperatures fluctuating around 4 degC during the night. The first 30-minute test was with an on-axis star, and the second 20-minute test was with an off-axis star so it also includes instrument rotator errors. Immediately prior to the open-loop tests, the telescope pointing was updated (ACQUIRE preset) on the same star. The test results are shown in the attached plots.

The simple summary is that the telescope pointing drifted by 2 arcsec in radius over a timescale of 20-minutes for both tests. The longer test had an inflection point (flattening the drift rate) so also had 2.2 arcsec of drift in 30 minutes. The image FWHM degraded from 0.35 arcsec FWHM to 1.2 arcsec FWHM in 20 minutes due to a combination of focus and astigmatism (astigmatism appears to dominate because of the hysteresis of the AdSec - 1000 nm Z6).

Note: These drifts would be worse on nights with strong thermal changes, and would be worse at high elevation if we had not recentered the pointing. "worse" means perhaps double the drift rate (see such a test on 20130123 UT).

Explanation of the Plots

The attached plots are from guideplot.py. The upper panel shows FWHM (arcsec) and ellipticity of the guider image as a function of time (UT hours). The times for starting and stopping guiding are indicated. The middle panel show the brightness (r' magnitude) of the guidestar as a function of time. The lower panel show the X and Y positions of the guidestar relative to the hotspot (arcsec, each offset from zero for clarity), and the radial distance R (arcsec, not offset from zero, orange points). The two plots have different timescales on the x-axis.

guide.sx.20130203test1.jpg

guide.sx.20130203test2.jpg

Checking new pointing model

Installing new pointing model from last night SX_Gregorian.20130202.ptmod

04:34 Stop and restart PCS. Restart Pointing Log.

04:36 ACQUIRE preset to WT10_197 Adjust IE=-61 CA=-45

Data/logs live in /home/aoeng/data/20130203/pointing/

04:40 ptmodel_collect, LIMIT=1000 (still with 30 sec wfs exposures)

Missed second star because of clouds.

Star Name Target Star GStar WFSC Image
1 ACT0359 1 1 left_wfscimage000175.fits
3 WT10_400 3 3 left_wfscimage000177.fits
4 WT10_329 1 1 left_wfscimage000179.fits
5 ACT0295 5 5 left_wfscimage000182.fits
6 WT10_229 1 1 left_wfscimage000184.fits
7 ACT0267 6 6 left_wfscimage000186.fits
8 ACT0110 5 5 left_wfscimage000188.fits
9 WT10_182 15 15 left_wfscimage000190.fits
10 WT10_271 3 3 left_wfscimage000191.fits
11 ACT0239 1 1 left_wfscimage000195.fits
12 WT10_322 4 4 left_wfscimage000197.fits
13 WT10_290 1 1 left_wfscimage000198.fits

Close

05:24 Closing for heavy clouds

05:25 Stopped pointing log.

Closed Dome Test

06:20 Doug and John running closed-dome tests with fake guide stars to solve the GCS problem. We do this by send a GUIDE preset with fake guide stars and then doing the triggerWFS command to take and process a wfs image. The erroneous behaviour was reproduced.

It turned out to be missing sextractor configuration files in the reconstructor directory. Doug will make an Issuetrak about the long term solution. We are now running the new reconstructor with its own copy of sextractor files.

07:20 Doug fixed the astigmatism LUT for DX M2 (replacing the one that got trashed in software upgrade).

Astigmatism Hysteresis Experiment Plan (not yet implemented)

Active Preset (wait 4 min)
Slew to hold EL-10 (wait 0.5 min)
Active Preset (wait 4 min)
Slew to hold EL+8 (wait 0.5 min)
Active Preset (wait 4 min)
Slew to hold EL-6 (wait 0.5 min)
Active Preset (wait 4 min)
Slew to hold EL+4 (wait 0.5 min)
Active Preset (wait 4 min)
Slew to hold EL-2 (wait 0.5 min)
Active Preset (wait 4 min)
Slew to hold EL+10 (wait 0.5 min)
Active Preset (wait 4 min)
Slew to hold EL-8 (wait 0.5 min)
Active Preset (wait 4 min)
Slew to hold EL+6 (wait 0.5 min)
Active Preset (wait 4 min)
Slew to hold EL-4 (wait 0.5 min)
Active Preset (wait 4 min)
Slew to hold EL+2 (wait 0.5 min)
Active Preset (wait 4 min)

-- JohnHill - 03 Feb 2013
Topic attachments
I Attachment Action Size Date Who Comment
daq.20130203telescope.jpgjpg daq.20130203telescope.jpg manage 570 K 05 Feb 2013 - 21:37 JohnHill Plot of Telescope Temperatures
guide.sx.20130203test1.jpgjpg guide.sx.20130203test1.jpg manage 524 K 21 Feb 2013 - 21:48 JohnHill Open-Loop Test1
guide.sx.20130203test2.jpgjpg guide.sx.20130203test2.jpg manage 454 K 21 Feb 2013 - 21:49 JohnHill Open-Loop Test2
Topic revision: r8 - 22 Feb 2013, 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