Science Prep Night 20130208 UT


Checked out MODS, LBC and at the end of the night switched to LUCI to do MOS and long-slit test acquisitions. The instruments are working, as well as before, and ready for science.

MODS: Before sunset, we obtained through-slit twilight flats in dual, red-only and blue-only modes - just for practice, in the rare case that observers request these, and to add to a database of flats. Next we observed a spectrophotometric standard star (G191b2b) in all configurations. The sky looked clear, but towards the end of these observations, cirrus came in and the guide signal dropped by a few tenths of a magnitude. We ran a long-slit acquisition and quick observation of a Nova. Then, concerned about whether collimation was holding, we ran a 5-dice imaging script of 30-second observations.

We noticed that collimation was not holding --- the pupil image jumped by 1 column of subaps and coma started to be introduced. We called Doug. He worked on this and ultimately made changes to the hotspot and mask in the configuration file. For certain guide probe locations, the pupil is vignetted by a structure in the MODS AGw. We thought this was handled by the active optics, but it was not adequate. After making these changes, Doug forced the problem again, and the collimation seemed to hold --- no coma was being introduced. But we do not have complete confidence that collimation will hold for all guide probe positions. Doug said he would use archival data to map of the pupil vignetting as a function of guide probe position; we will give this information to Rick, to incorporate into modsView.

LBC: Prior to LBC observations, the DX collimation lookup table was edited to add a Z global offset of -0.6mm. This is so that the starting point, after the hub move last month, would be closer to collimation. On the first dofpia run, z4 was within the range (|z4| < 10,000 nm) where the other corrections (coma, astigmatism, sperhical) would be sent. Encountered problems with the field of a pointing star when only 1-3 pupils were selected and had to move to another (observer-supplied) field. Pupils there were faint but measurable using dofpia, /x2.

Neither the blue elongation nor pupil "skirts" were seen. Seeing was ~1" and the pupil structure not so detailed as when the "skirts" have been seen. The telescope was in good thermal equilibrium. SX face-back delta T ~ 0.15C and DX face-back was better, delta T ~ 0.1C.

The horizontal striping that had been reported on LBC -Red images was seen in all images where I looked for it. It seems to be on only chips 2 and 4.

LUCI: Test acquisitions for MOS and long-slit using observer scripts; for MOS (J0952 program) and long-slit (SN2011fe program). MOS acquisition at N1.8 worked well and gave small errors. For future reference: there were 4 alignment stars not seen and, through the boxes, there were several stars not seen, but fortunately there were also stars visible, and these apperaed to be centered after the MOS ACQU offsets were made. Test done at N3.75, too, and it gave offsets in the correct sense. Long-slit acquisition also appeared to work, though I did not confirm the centering. Thicker cirrus came in.

23:45: started MODS services. Took sieveSnap and imSnap.

00:15: Geno sent test preset with MODS. I had not put it in observing mode yet it did complete without issue. As a precaution, I homed AGW probe after this. But, when a second test preset done with MODS in observing mode did not complete/display am image, Geno asked me to check whether AGw camera was on. Both AGw Guide and WFS were OFF on the MODS User Interface, Instrument Utilities. I unlocked power and turned these on. Next test preset - GCS reported that acquisition image did not arrive in time. Geno powered off/on GCS. Finally image appeared in time. Test preset successful.

00:25 LBC test biases: Look OK with count levels ~450-460 Blue and 480-490 Red. Vent doors open. A gradient on chip 3 with lightest part at top left. May very well have been there before and visible when chamber was light, but I had not noticed it.

00:43 LBC test for background striping with light exposure: /home/lbto/SciPrep/LBC/test_background.ob, test_win_background.ob and test_win2_background.ob. Windowing because the time the vertical striping appeared during the night, it was upon switching readout (20121010).


00:50 Twilight sky spectrum through 1" slit. (skyspec.obs)
  • r b 3: 8000 counts in 1 sec (guide probe not homed)
  • r b 4 @ 00:55. 5 seconds b 30k and r 47k
  • b 5 blue-only 5 seconds, 16k
  • b 6 blue-only 12 seconds,
  • r 5: dual 20 sec 36k
  • r 6: red 30 sec, 25 k but only about 1500 at red end.
  • r 7: red 40 sec, 20k
  • r 8: red 90 sec,
  • b 7: blue 120 sec... pretty low about 2k counts. 01:28

01:29: Try MODS imaging flats.
  • r 9 1-sec red-only some stars and low counts 1k

01:32: Pointing/collimation. For the second time now, the acquisition image did not arrive in time. Geno stopping/restarting the GCS. This time it worked. Another time it failed - keeps flashing "timed out waiting for image" and guide image does not update. Geno calling Joe & MIchelle.

02:10: Geno stopped GCS, rebooted AZCAM server and AZCAM-G (I also again powered off the AGw Guide and WFS on MODS Instrument Utilities, but they could have been left on as these buttons only power off/on the Leach controllers). Started all these again. Seems this helped as we do not see the "timed out waiting for image" messages.

02:17: acqMODS g191b2b.acq It looked clear overhead at sunset. There were low clouds on the S horizon.
  • Collimation looks pretty good - guide stars look round. Seeing about 0.8-0.9".

02:27: execMODS g191b2b.obs Seeing 0.8-0.9" but blowing up to 1.5-1.6" at times. Guide star images still look round, without tails.

02:30 - modsDisp crashed:
 Image mods1b.20130208.0011.fits STD - G191-B2B blue grating - Clear - LS60x5 IN
Traceback (most recent call last):
  File "/home/MODSeng/bin/modsDisp", line 317, in <module>
  File "/home/MODSeng/bin/modsDisp", line 226, in watcherLoop
    display.set('scale mode zscale')
  File "/lbt/mods_runtime/python-mods/lib/python2.6/site-packages/", line 361, in set
  File "/lbt/mods_runtime/python-mods/lib/python2.6/site-packages/", line 308, in _selftest
    if self.verify and not xpa.xpaaccess(, None, 1):
  File "/lbt/mods_runtime/python-mods/lib/python2.6/site-packages/", line 183, in xpaaccess
    if errmsg: raise ValueError, errmsg
ValueError: XPA$ERROR: no response from server during handshake (?:?)

02:56: seeing up to 1.6" and clouds - guide plot shows dimming of about 0.2 mag. (see plot attached).

03:16 Finish with acqMODS -i and acqMODS -a g191b2b.acq to get a final image and see how collimation has held up. Guide star has looked good throughout.
  • See mods1r.20130208.0025, 0026, 0027 and compare with initial image mods1r.20130208.0010.fits
    • 10 fwhm = 6.6 (0.811)
    • 25 fwhm = 4.5 but not round
    • 26,27 also show stars near the axis which are not round.
    • => concerned that collimation did not hold up well over the 1 hour of spectroscopic observation since near on-axis stars did not appear
round on images 25,26,27.

03:43: Long-slit acquisition and observation. IQ on 5-sec acq image looks OK. But compare initial and final acq images: r 30 with 37,38,39.

04:24 Imaging script to check image quality and collimation - guide star does not look perfectly round. Images are 30 sec each. Seeing ~ 0.8". Taking a 5-dice of 30-sec exposures.

04:36 about: MODS blue channel hung on "Exposure Done, cleaning up...". Gui is hung after two images each of blue/red.
> go
STATUS: RED GO Finished 1 of 1 exposures. ACQTAG=MODS1R.20130208042759 EXPSTATUSSTATUS: BLUE GO Finished 1 of 1 exposures. ACQTAG=MODS1B.20130208042801 EXPSTATUSTATUS: RED GO Red Channel Exposures Done - Blue Channel still acquiring...

  • I refreshed the GUI. Still hung
  • Then I ran "blue reset"
  • I had to abort (cntl-c) the script
  • running a script to do the last 3 dithers and run with execMODS -e
  • collimation is falling apart...
  • when script ended, I changed to 8Kx3K (to use modsBias on it) and took an image ---
  • now repeating the entire script with 8Kx3K mode. Guide stars show tails and images are not collimated. See mods1b.20130208.00(30-34).fits and mods1r.20130208.00(47-51).fits.
  • Geno calling Doug.
  • entire sequence ran from b 23 and r 40 b (23-34) and r (40-47) corresponding to UT 04:26 to 05:07.

The shack spots had shifted one subap during the observation. See wfsc images 000205 and 000232. Doug changed the hot spot in the configuration file, and we were able to collimate for this guide star position, but unsure it will work for all positions. This is the 3rd field observed so far and the first time we saw this problem - other image may not have been well collimated but we did not notice this large shift.

I attached finders which show the positions of the guide stars for the 3 fields observed with MODS tonight.

05:55 "Warning - stream of elevation polynomials has stopped. The GCS reports "ERROR: probe failed to move to offset position. Guiding paused.

This happened on "offset 10 -10 abs". We had obtained 3 of the 5 images already: red 52,53,54 and blue

PCS had stopped - window was closed (spontaneously, we don't know why).

06:06: Resending the preset. Script ended at 06:24. Doug updated the configuration file now. He changed the mask, used by source extractor, from 11x11 to 9x9 shack spots.

06:25 Resending preset after the GCS configuration file was updated. Grabbed wrong guide star. Tried to preset again but again it chose the faint one. Running with this one.

06:51 Doug forced the problem to see whether changes to the configuration file handle the shift. He shifted the pupil one subap to the right and watched to see whether coma would be added. Looked OK.

07:05 To summarize work done. Two changes were made to the MODS configuration file:
  • the hot spot was changed
  • the mask used by the source extractor active optics algorithm was changed from one using an 11x11 grid of spots to one using a 9x9 grid.

07:05 One more MODS field. Shell rip during offset to center up on slit. (07:17). Recovered but ripped again (~07:20). Geno calling Juan Carlos, as the shell is not resetting. Weather is closing in so we have to give up on this preset and go to LBC.


03:00 taking some LBC images while observing with MODS to see noise patterns. On lbcr.20130208.030134.fits there is the horizontal striping - seems it may be on chips 2 and 4 only, but would need to look more closely at all.

08:00 Switching from MODS to LBC

08:13 Slew to ACT0392. Run dofpia, /first. Had updated z global offset on DX to -0.6 mm this evening. first dofpia images. 081354/081353. The first z4 to Red was +500nm, to Blue -500nm.

dofpia is not converging --- there are only 1-3 pupils found.

Slew to 11:01:30 64:02:40 instead. Collimated in 2 iterations.

08:38 Slew to a new field to correct pointing/copointing (did not go back to ACT0392 field since so few pupils were found). The pupils are very faint. Ran dofpia, /x2. There may be some cloud as later there were more pupils and slightly brighter. 084514/084511 b/r copoint images.


repeat acq of images 084832/084829. B/R have 5.2/

08:48 run OB offset.ob to guide and test application of offset given in Interactive Pointing part of Target page when it is not absorbed into the Target Coordinates.
  • guiding appears to work - see guide offfsets in LBC log and some lines in 20130208.. But also in
ri Feb  8 09:31:37.926 2013 56331.39696674038 2 PCS  pcs.command.applyRBTipTilt.warning left PCS ApplyRBTipTilt Warning: application of PSF range balance tip/tilt request is disabled on side left.
Fri Feb  8 09:31:40.089 2013 56331.39699177899 2 PCS  pcs.command.applyRBTipTilt.warning right PCS ApplyRBTipTilt Warning: application of PSF range balance tip/tilt request is disabled on side right.

  • see focus offsets coming through in lbc.log

09:11 Stopping OB. Running dofpia, /x2 and Loading and running offset_absorb_test.ob This has the offset dRA = -10 in it. *One needs to absorb the offset entered into the Offset and Rotation panel. Otherwise it has no effect.*

Red image 093627 (120s r-SLOAN) has fwhm 3.6 pix Blue image 092558 (360s SDT_Uspec) has fwhm 5.7 pix and avg ellipticity 0.04

Run dofpia, /x2 again and this time run OB offset_absorb_test (using V-BESSEL instead of SDT_Uspec and i-SLOAN instead of r-SLOAN).

First images at V-BESSEL look OK, not elongated. Seeing 1.2-1.3" in Blue.

LBC seems to be working as well as before.
  • Guiding looks OK (see warning messages above which may be harmless - I've never looked for these or seen them before). None of the blue elongated images seen.
  • Dithering is OK. In the instance I looked at, 0 0 was not the same but there may have been an fpia between these.
  • No skirts around blue pupils - the pupils looked more blurry/worse seeing. Seems we see the skirts for the best seeing.


Switching to LUCI at 10:00 UT.

10:43 sending MOS acq J0952M1. Seems to have worked. Images are:
  • sky #4
  • obj #5
  • mask #6
  • MOS alignment: I had to exclude stars 2,6,7,8 from the fit as I could not see these in the sky-sub image. Fit results:
    • Fit results:
      • rot = 0.0617 deg (error 0.0391 deg)
      • Tx, Ty (arcsec) = 0.4586, 0.2421 (errors 0.25 and 0.32 pix)
  • confirmatory image #7. I can see stars in some boxes. In other boxes I see no stars, but I see nothing very bright for these on the sky-sub field image.

Condensation on window. Chamber DP = -10.5 deg; Chamber Temp = -2.8 deg; outside RH = 61%

11:22 cirrus. We'll try N3.75 MOS acq.
  • sky #9
  • obs #10
  • mask #11
  • MOS alignment - again had to exclude 2,6,7,8
  • Fit results:
    • rot = 0.0745 (error 0.0526 deg)
    • Tx, Ty (arcsec) = 0.3949, 0.1333
    • T-Error (pix) = 0.7393, 0.92
  • confirmatory image is #12.

11:40: trying a long-slit acquisition using a ND script.
  • sky #13
  • obj #14
  • slit #15
  • guider seeing 1"
  • barely see target (sn2011fe). Offset required is detxy = -2.40",13.00".
  • thru-slit confirm #16 and field confirm #17 (there is now a lot more cirrus...)

12:00 thickening cirrus --- guide star flux plummetted by about 1.5-2 magnitudes from UT 11.2 to UT 11.6 - now off the plot. We see no stars from the balcony. Going to close.

12:20 Putting LUCI in end-of-night configuration: dark mask, blind/blind,mirror and N3.75 (00_Scripts/Misc/EndOfNight).

12:41 Dome dark, starting MODS calibrations:
  • grlamps (dual, blue and red grating comparison lamps) 0.6" slit
  • mosflats for A0773m1 blue 69 (9500 adu o 5 sec) - blue 81 and red 83 (25000 adu) - red 95
  • repeating the blue flats with ND1.5 for which counts were only 9500 adu with 12-sec instead of 5-sec. blue 82-86.
  • --- did not finish these as day crew arrived and was ready, and we can do these during the next days.

-- DavidThompson - 15 Jan 2013
I Attachment Action Size Date Who Comment
NovaMon.pngpng NovaMon.png manage 225 K 08 Feb 2013 - 05:52 OlgaKuhn  
g191b2b.pngpng g191b2b.png manage 386 K 08 Feb 2013 - 05:51 OlgaKuhn  
guideplot20130208.jpgjpg guideplot20130208.jpg manage 184 K 08 Feb 2013 - 03:19 OlgaKuhn guide plot during g191b2b
qso8k.pngpng qso8k.png manage 193 K 08 Feb 2013 - 05:52 OlgaKuhn  
Topic revision: r8 - 08 Feb 2013, OlgaKuhn
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