Left Direct Gregorian Technical Observing with IRTC: UT 2009 November 26

Observer: JHill (LBTO)
Telescope Operator: ACeranski
Instrument Support:
Telescope Support: JLittle
SW support: (Tucson)


It was another beautiful clear night. Unfortunately, the seeing was much worse despite the telescope being in better temperature equilibrium.

We sorted out the IA coefficient versus azimuth encoders that was bothering the pointing last night. Then we spent the first half of the night collecting a large set (100 stars) of pointing data. Essentially all the stars landed on the acquisition image, and the eyeball rms of the old pointing model with corrected IA was about 5 arcsec. So this big slug of data is mostly to investigate the fine details. But, SEE ISSUE #2437: It seems likely that the azimuth axis was running on motor encoders.

We spent the last couple hours doing various tests related to guiding on faint stars in bad seeing. Probably getting more mileage on the system is useful, but the productivity per hour seems low. Seeing was varying between 0.8 and 2.5 arcsec for most of the night. We did work out that a larger guide box is better for guiding in very bad seeing.

We observed large correction forces building up on the primary mirror again, but we haven't figured out what is causing them.


The biggest technical failure of the day was that John H. broke the pivot axle of his reversible belt buckle (before he ate Thanksgiving dinner!). That resulted in some poorly collimated pants. Super glue seems to have cured the problem until a replacement part can be obtained.

Doug logged in and cured our IDL problems from last night. It was a known but rarely encountered problem relating to a configuration file in /tmp/.

We verified that the WFS pupil hotspot coordinates were correct in this build.
current/etc/GCS/AIP_IRTC_DG_L.cfg:WFScam_hotspot_x      388.0           double  # ccd x coordinate of pixel that represents leftmost x-center of WFS spots
current/etc/GCS/AIP_IRTC_DG_L.cfg:WFScam_hotspot_y      246.0           double  # ccd y coordinate of pixel that represents leftmost y-center of WFS spots
current/etc/GCS/AIP_IRTC_DG_L.cfg:WFS_lenslet_edge_x    383.0           double  # defines the edge of the virtual lenslet grid in x
current/etc/GCS/AIP_IRTC_DG_L.cfg:WFS_lenslet_edge_y    243.0           double  # defines the edge of the virtual lenslet grid in y
Those might have explained the spurious wavefront corrections had they been wrong.



00:25 Open to another beautiful night. Photometric except for a little patch of clouds very low on the SW horizon. T=1.7 degC. D=-22.3 degC. Wind 6 m/sec from the West.

00:28 Aaron goes upstairs to reset the Rotator MCR, but it wasn't really tripped. It was only a software artifact in MCSPU.

Initialize Pointing and adjust the IA term of the pointing model

00:37 pointto M5_0036 gs=0 TelM="ACQUIRE" PAM- - after ptreset, ptspiral Adjust IE=2 CA=144 and IA=-404 at EL=38.5
Then rebalance terms to IE=2 CA=32 and IA=-260 . This correction was based on last night's observed trend of azimuth error CA as a function of elevation IE.

00:52 pointto M5_0050 gs=0 TelM="ACQUIRE" PAM- - Adjust IE=0 CA=27 at EL=38 (so not really a fair test).

00:55 pointto M5_0881 gs=0 TelM="ACQUIRE" PAM- Adjust IE=-10 with no correction needed for CA at EL=68.8 AZ=110.

00:59 pointto M5_1568 gs=0 TelM="ACQUIRE" PAM- Adjust IE=-13 CA=18 at EL=75.8 AZ=-22

01:05 pointto M5_0893 gs=0 TelM="ACQUIRE" PAM- Adjust IE=-5 CA=29 at EL=43.9 AZ=6.5

Test GCS display of bright stars

01:09 pointto M5_0893 gs=0 TelM="ACTIVE" - The preset failed because the image was horribly saturated so it could not guide stably, but all the images on the GUI were displayed optimally. That Issuetrak #2388 can be closed.

Initial collimation

Note that the images have already been round and decently focussed without operator intervention or active optics.

01:12 pointto BS9186 gs=0 TelM="ACTIVE" - didn't find the star although see a bright rectangular artifact that may be moonlight.

01:17 pointto M5_0850 gs=0 TelM="ACQUIRE" PAM- Adjust IE=-1 CA=35 at EL=57 AZ=173

01:19 pointto BS9186 gs=0 TelM="ACTIVE" - still no hint of a star.

01:22 pointto StoneOagw gs=0 TelM="ACTIVE" - Image was 1.4 arcsec without any focussing or collimation. Active optics only improved things to 1 arcsec.

Back to Pointing and Encoder Adjustments

Dave A. calls to point out that there is a significant Azimuth encoder offset.

01:28 Slew to Az=120 to adjust the encoder. We adjusted azimuth relative encoders by approximately 23 arcsec.

01:32 pointto M5_0881 gs=0 TelM="ACQUIRE" PAM- - Star is just in upper left corner of acq image. Adjust IE=-15 CA=+26 IA=-290 at EL=76 AZ=128.

01:38 pointto M5_0864 gs=0 TelM="ACQUIRE" PAM- - Star is just in lower right corner of caq image. Adjust IE=-2 CA=+26 IA=-270 at EL=49.7 AZ=176.

01:43 pointto M5_0896 gs=0 TelM="ACQUIRE" PAM- - Star is well centered (3 arcsec off) at EL=27 AZ=164.

01:45 Slew back to az=120. Offset repeats within < 1 arcsec.

01:49 pointto M5_0893 gs=0 TelM="ACQUIRE" PAM- Adjust IE=10 CA=20.5 IA=-270 at EL=44.8 AZ=3
Note that IA=-274 was the value from the 12-Oct-2009 LDG model, so there may have been an azimuth encoder offset on 26-October.

Now unwrap azimuth, and verify that pointing is still the same. It was, so there are no azimuth scale issues. At the same time John L. and John H. investigate why bogie B is making popping sounds. Maybe it has a steering issue and is slipping on the track during long slews?

02:14 pointto M5_0036 gs=0 TelM="ACQUIRE" PAM- - Adjust CA=25.


02:26 pointto StoneAagw gs=0 TelM="ACTIVE" - Seeing is varying from 0.9 to 1.8 arcsec, so this decides what we will do next - we'll collect more pointing model data. Wind has rotated to 10 m/sec from the East.

Pointing Data

02:33 Start pointing logging in PCS using file /home/telescope/20091126.ptlog

The working directory is /OldRepository/IRTC_Data/20091126 but only the gopoint.log file is written there. I've also attached the gopoint.log file to this wiki page for the complete list of times, grids and targets. These observations are pointed in GUIDE/ACTIVE mode, so the star is well-centered on the IRTC image in nearly every case. Images are accumulating in /newdata as expected.

But, SEE ISSUE #2437: It seems likely that the azimuth axis was running on motor encoders.

02:37 gopoint 1 grid=66 target=ACT0364

02:42 gopoint 2 grid=1 target=WT10_404 irtc.20091126.024526.fits /newdata/irtc.20091125_00094_cube.fit

Adjusted minimum exposure times for guider to 200 ms and wfs to 1000 ms while working on these bright stars. This is a temporary change.

Changed default irtc cube name to 26 here.

02:47 gopoint 3 ploop+ grid=39 target=ACT0403

After the first 10 stars, the eyeball rms of our open-loop pointing looks to be about 5 arcsec rms. So we hardly need a new pointing model. The goal will be to get a large dataset so we can study some of the finer details of pointing.

The C-ring lower steel is about 1 degC warm, but otherwise the telescope seems to be in decent temperature equilibrium. (see the attached plot below).

03:47 working on gopoint 20 grid-85 target=WT10_293

03:49 stop and restart GCSL

03:50 pointto BS9107 gs=0 TelM="ACTIVE" - Seeing is 0.8-1.2 arcsec on guider at EL=60 after some collimation.

03:57 gopoint 21 ploop+ grid=57 target=ACT0257

04:25 didn't find a suitable star for position 29.

05:07 working on position 43

06:05 IRAF error during position 62
GOPOINT: Taking IRTC sequence 0.1 100 
INTERNAL ERROR: dictionary full
     line 131: IRTC$takeseq.cl
     called as: `takeseq (exptime=0.1, Nimages=100, frmrate=100000, take1=yes, verbose=no)'
  "takeseq ( Tint, Nimgseq, frmrate=100000, take1=yes, verbose=no)"
     line 151: home$gopoint.cl
     called as: `gopoint (gridID=21, ploop=yes)'

Nothing more than flpr needed to restart.

06:10 pointto M5_0151 gs=2 TelM="ACTIVE" - seeing is still 0.9 to 1.4 arcsec on guider.

06:15 gopoint 29 grid=43 target=WT10_379 irtc.20091126.061900.fits /newdata/irtc.20091126_00153_cube.fit

06:20 gopoint 62 ploop+

07:41 We've finished all the gridpoints, as Marco's program only does 88 positions.
GOPOINT: Presetting telescope to grid position 89 
Warning: Attempt to delete a nonexistent file (junk_gopoint.tmp)
ERROR: scan limited to 1022 char lines
     called as: `scan (gopoint.gridnum, gopoint.srcID)'
  "print ('!./Pointingstars5 < ', junkfile )| cl | scan (gridnum, srcID ..."
     line 90: home$gopoint.cl
     called as: `gopoint (gridID=62, ploop=yes)'

07:43 gopoint 1 ploop+ - Let's do a few more stars to see how the telescope has changed since we started. The steel has cooled about 1 degC in the last 6 hours, so the gradient is somewhat less than an average night. Note that the mirror cell ambient saw a 1 degC increase in temperature near 7:30 UT.

08:25 Finish 100 pointing stars! But, the telescope has made its first error of the night. The secondary position being reported back to PSF from OSS was not updating. We speculate that the secondary may actually have been moving, and that it was the reporting back of the position that went wrong. We killed OSS, verified its death and restarted. The secondary was in the correct position as soon as OSS finished its initialization.

Collimation and WFS Tests

08:29 pointto StoneEagw gs=0 TelM="ACTIVE" but the command hung in IRAF. Killed and restarted my IRAF window.

08:32 pointto StoneEagw gs=0 TelM="ACTIVE" - Seeing is even worse at 1.2 to 1.9 arcsec on guider. We're just going to sit here for a while and collect some WFS performance data in horrible seeing.

Testing IDL WFSing

08:57 Press stopWFSing button on GCSGUI so we can test IDL WFSing.

08:58 agw_collimate, exp=5 => /OldRepository/AGW_Data/091126/wfsc000001_irc.zern

08:59 agw_collimate, exp=30

09:00 agw_collimate_loop, exp=30

This ran through wfsc000037.

Note that we saw no evidence of spurious wfs corrections with GCS or IDL. Although the typical wavefront rms was 400-500 nm, so Z22 isn't sent in these conditions.

Guide Stars of Different Brightnesses in Poor Seeing

09:30 pointto StoneEagw gs=13 TelM="ACTIVE" (mag 14.98)

09:39 pointto StoneEagw gs=2 TelM="ACTIVE" - This star (mag 16.6 in unknown band) looks to be near the faint limit when the seeing is more than 1.5 arcsec FWHM and using 2.1 second exposures. GCS starts to legitimately not be able to detect it when the seeing is more than 2.0 arcsec FWHM. At that point, the photons are spread over the whole 100 pixel guide window. Peak counts reported by GCS are about 30 above background when seeing is 2.0 FWHM, and about 160 above background when seeing is 1.0 arcsec FHWM.

09:54 Changed guideWindowSize from 100 to 200 pixels in AIP_IRTC_DG_L.cfg.

09:55 pointto StoneEagw gs=2 TelM="ACTIVE" - No change in the guide window.

09:57 Stop and restart GCSL.

09:58 pointto StoneEagw gs=2 TelM="ACTIVE" - Bigger guide box seems to work fine. And it makes the bad seeing look much better! getStatus reports the effective guide rate at 0.305 Hz with 2.1 sec exposures. The serious advantage seems to be that the detectability of this faint guide star is improved when the seeing is worse than 2 arcsec. This seems to be a good reason for using this larger guide box as the default.

10:20 pointto StoneEagw gs=4 TelM="ACTIVE" PA=INDEF=300.21 (mag 17.40 in unknown band), but requested probe position is out of range.

10:22 pointto StoneEagw gs=4 TelM="ACTIVE" PA=0 - AGW4: Error moving the motor(s) setProbePosition: couldn't move probe to position (SFP) (x,y,z): 170.052, 99.1903, 14.4792 mm.

10:23 pointto StoneEagw gs=2 TelM="ACTIVE" - same error. Home the motors on AGW4.

10:28 pointto StoneEagw gs=2 TelM="ACTIVE" - Failed to find the star in the acquisition image. I can see the very blurry star, but I'd classify this as a legitimate failure.

10:30 pointto StoneEagw gs=4 TelM="ACTIVE" PA=0
   irc PresetTelescope 0. POSITION ACTIVE left
Resmsg[0]  = PresetTelescope result status: Error
Resmsg[1]  = OK
Resmsg[2]  = OK
Resmsg[3]  = AGw command status: AGW4: Error moving the motor(s)
Resmsg[4]  = error: setProbePosition: couldn't move probe to position (SFP) (x,y,z): 170.042, 99.1915, 14.4807 mm.
Resmsg[5]  = guidestar list successfully changed but probe couldn't be moved to position.

10:33 pointto StoneEagw gs=2 TelM="ACTIVE" - same error. Home the motors on AGW4.

I made this Issuetrak #2432, although it might be the same as #2411.

10:45 set guideWindowSize to 150 pixels. Stop and restart GCSL. This makes the effective guide rate 0.33 Hz with 2.1 sec exposures. I'm proposing to make this the default value. See the Issuetrak.

10:46 pointto StoneEagw gs=2 TelM="ACTIVE"

10:49 set WFS default exposure time back up to 25000 ms.

Dice5 offset data in poor seeing

I don't know if this data will be good for anything or not. Maybe it can be a comparison to good seeing data? Seeing was verying between 1 and 2.5 arcsec.

11:06 dice5 7 7 DETXY 1 sec 50 images 3 cycles starting with irtc.20091126.110743.fits

11:29 dice5 7 7 DETXY 1 sec 50 images 3 cycles starting with irtc.20091126.113015.fits

Note that we've now built up large correction forces on the mirror (-117 pounds max on actuator 101). Now the question is what is different here than we had 2 hours ago when the forces were half this big. We are presently at EL=48 with seeing varying between 1 and 2 arcsec.

More IDL WFSing

11:51 Stop GCS WFSing. Start IDL to see if it has a different opinion about the high order corrections. The first image wfsc 39 had a negligble amount of Z22, but 66 nm of Z10. Could there be an angle issue? After 10 iterations, the force on actuator 101 decrease only slightly from -113 to -106 pounds.

12:01 Clear C00 on SX primary

12:02 continue IDL WFSing using agw_collimate_loop. starting with wfsc 49. First correction had +1034 nm Z4. wfsc 54 had -63 nm Z22 (326 nm rms) and wfsc 55 had -49 nm Z22 (267 nm rms). After 15 iterations, the force on actuator 101 has built to -44 pounds, but half of that came from wfsc 54 and 55. After 31 iterations, the force on actuator 101 was only -40 pounds. Last image was wfsc 79.


12:30 We declare a holiday and close the dome. Happy Thanksgiving!

-- JohnHill - 19 Nov 2009
I Attachment Action Size Date Who Comment
daq.20091126.gifgif daq.20091126.gif manage 49 K 27 Nov 2009 - 00:14 JohnHill Plot of Telescope and Mirror Temperatures
gopoint.clcl gopoint.cl manage 5 K 16 Jan 2010 - 00:27 JohnHill IRAF script to automatically collect pointing data
gopoint.loglog gopoint.log manage 11 K 26 Nov 2009 - 12:35 JohnHill Logfile for the pointing model targets. (not PCS log)
pmcc.dx.thermal.20091126.gifgif pmcc.dx.thermal.20091126.gif manage 32 K 27 Nov 2009 - 00:19 JohnHill Plot of DX Primary Mirror Temperatures
pmcc.sx.thermal.20091126.gifgif pmcc.sx.thermal.20091126.gif manage 34 K 27 Nov 2009 - 00:19 JohnHill Plot of SX Primary Mirror Temperatures
Topic revision: r9 - 16 Jan 2010, 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