Startup Night 3: 2017 Sep 04 UT

Observers: JHill, OKuhn, BRothberg (Tucson), SWalsh (LBTO)
Telescope Operator: SAllanson (LBTO)
AO Support: GTaylor
Telescope Support: KNewton (LBTO)
Software Support: CBiddick, KSummers (Tucson)
Instrument Support: ESolheid (LBTO)
Engineering Support: TMcMahon


Dave T installed new flexure compensation tables for LUCI1 and LUCI2 (can fill in modes here, but I suppose N375 imaging, N180 G200 and N180 G210 as well as N375 G210, the calibrations which I know were done).

In LBCR, Elliott cleaned the filters and vacuumed the moths out. In LBCB, he cleaned the lenses.

Barry created and sent around scripts for LUCI test to be done.

John updated Pointing model and Collimation models.


At the zoom, we converged on the following plan:

Start with LUCIs in the beam
  1. Pointing model checks
  2. Collimation model checks
  3. collection of field aberration data (if seeing <1")
  4. LUCI rotator center check, using M92 field and track presets at 2 diametrically opposed PAs.
  5. LUCI MOS acquisitions to measure rotator zeropoints:
    1. Acquire mask1 and note dtheta needed to align
    2. Acquire mask2 and note dtheta
    3. Edit rotator zeropoint in PCSInstrument.conf file by adding or subtracting <dtheta> (50% chance of getting the sign right...) and restart PCS
    4. Acquire either mask1 or mask2 again and see if the sign was right. If not, update PCSInstrument.conf. Strictly speaking, that should be it, although I would be tempted to repeat the acquisition just to insure that dtheta is close to zero now.
  6. Collect transform data (we have a question for Doug M regarding this: does he prefer we use his scripts or are LUCI scripts OK?) and this may need to be deferred.
Reconfigure for LBCs:
  1. Repeat sidereal mode science-like OBs. (Questions remain about NS guiding and dithering, but these are lower priority tests than simply exercising LBC in the mode which will get the most use as we start partner observing.
  2. Skyflats if clear. (Today will focus on Us LBCB, and on LBCR V and i).


Checked new pointing and collimation models. Pointing model performs similarly to the old one - it was left in place throughout night. Collimation model had some errors, but even after these were corrected, it put M1 near the limits and so was not used during the night.

Collected data for measurement of LUCI1/2 rotator centers. (Aladin does some weird things subtracting images taken at 2 different PAs!)

Through the offsets computed on two MOS alignments, computed adjustments to left/right rotator zeropoints for LUCI and implemented these in PCS (The dtheta output by RTD needed to be subtracted from the current rotator zeropoint. Of course we got the sign wrong the first time...) Checked the new left/right rotator zeropoints and these result in a reduction of dtheta (dthetas now ~0.03 deg) output by RTD.

Tried to start on transform data collection using IDL scripts on StoneA, but we ran into problems: left side wanted to talk to IRTC, while right side wanted to take a LUCI1 image. John could hack the code to fix the right side problem, but we didn't know how to tell left side it should talk to LUCI1 (error messages are in log below).

Switched to LBCs: As it looks clear on the allsky, observed photometric standards: Galadi, SA95. Next a face-on spiral IC342 (beautiful! and seeing 0.6"), the latter using guiding and dithering and at g,i. On the next, IC342 Usz, script, we had problems with the LBCs: first a B TRACKERS error, then a R TRACKERS ERROR and then the UI slowed to a halt. Also trying to bring the up from the CMU took ages.


01:45 Open - cirrus but mostly clear.

First 1.5 hours: Pointing and collimation checks. Steve is sending bino presets to WT/ACT stars at a range of elevations/azimuth.

New pointing model: The performance of it is similar to that of the old pointing model; we sometimes miss stars outside the 25" acquisition guide camera FOV. To see if this was due to collimation, the new collimation model was implemented.

New collimation model: But the new collimation model did not work - on SX, it drove M1 beyond its limit. Had to revert to the old collimation model.

Collimation tests: On DX, there is a bad actuator which is preventing collimation. Called Greg and he fixed it.

03:37 Winding down on pointing and collimation model tests. Steve and John are concluding that the new pointing model is not any better than the old one, but we are leaving it in to evaluate it more as we do further tests.

03:39 Preparing for field aberration data collection. Steve going to BS star... but we are changing our minds. It would be better to collect field aberration data after putting in a new collimation model.

03:42 LUCI Rotator Center checks: Run M92 scripts which use track preset, Ks, N30_FS mask, N3.75 camera, 2.51s x 22 NDIT
  • first pair at PA=0 was done before correcting pointing and collimating on nearby star
    • (L1,L2) (4,4) are skies @ 300", 300", (5,5) are field centers
  • 04:00 Pointing check and collimation on nearby star
  • 04:02 PA=0. 13893011 04:03:21 IIF PresetTelescope command for left side(s) failed! 0- PresetTelescope result status: Error 1 - Stop{FLAO}[96]2-parkActiveProbe: parkProbe() command failed: OK
  • 04:04 Tried again at PA=0, OK. 6 (sky) and 7 (obj)
  • 04:0? PA=180 8 (object only, no sky)
  • 04:0? PA90 9
  • 04:0? PA=270 10
  • Aladin subtractions of sky (6) from the images at different PAs (8,9,10) shows that the positions of the field stops do not overlap much. Called Dave who confirmed that the N30 FS is not really shifting very much on the detector (by examining the images in ds9) and aladin is not simply subtracting one image from another but must be doing some shifts based on the wcs info.
Doug called and found a typo in the new collimation model. We'll try the edited version later.

I cannot see /Repository/ from the rm525 machines.

04:40 Rotator zero point checks with MOS masks.


  • N375 acq images show: L1/SX W-unit not parked and L1 and L2 both need at least field stop alignment. Pupil alignment OK? That was done earlier, but I'll adjust the L2 FM4 position using RTD/AFC tab.
    • Current = 96,1958 target=5,2043. Adjust, mask to turnout, take new field image: 14. One more adjustment this time to upper right corner: current=2015,2021 target=2047,2047. New image 15.
    • New mask image 16.
  • L1 12,13 and L2 15,16
    • On L1, I see only stars 30, 32, 34, 35, 36, 37. 32 and 35 appear like double stars. 34 is near the W-unit glow. 37 is way off. Use only stars 30,32,35 and 36 and obtain:
      • first, for mask: x-shift=-52.6673px (-6.2042") and Y-shift=-12.3385pix(-1.4535") Angle=0.5008 deg.
      • x-shift = 0.99076", y-shift=0.94301", angle=-0.48958 deg. errors are 0.13174", 0.15486"
    • On L2, I see only stars 30, 32, 34, 35, 36, 37. Same as for L1.
      • for mask: x-shift=-6.0589, y-shift=16.5424 angle=-0.5764
      • and with same stars used 30, 32, 35 and 36, we get: x-shift = -2.93804", y-shift=1.65724" and angle=-0.90879 deg but errors are in the orange, like L1, at 0.12810 and 0.14995.
  • L1: 14 and L2 17 are the thru-mask images. I see stars 30,35,36,32,37,34 in boxes but they are not well centered.
  • Repeat acquisition on this field & mask: PY315.
  • 05:40 restart script.
  • L1: 15, 16 and 17,better x-shift=1.266, y-shift=-0.399, dtheta=0.06622 deg. (tweaked)
  • L2: 19, 20 and 21, better result. x-shift=-3.11, y-shift=1.16 and dtheta=-0.37329 deg
  • L1: 20,21,22(mask): x-shift=0.88", y-shift=-1.4" and dtheta=0.076 deg.
  • L2: 22,23,24(mask): x-shift=-2.27", y-shift=-0.44" and dtheta= -0.44 deg
Average dthetas are 0.071 deg for L1 and -0.405 deg for L2.

Editing PCSInstrument.conf.newpointing. For LUCI1, LEFTZEROPOINT was -187.734. We find dtheta=0.071 deg, so we are changing LEFTZEROPOINT to -187.663, and RIGHTZEROPOINT is being changed to 176.24, adding the dtheta calculated to the old value. Changes made both for LUCI and LUCI/ARGOS.

cp -p PCSInstrument.conf.newpointing PCSInstrument.conf

06:52 Steve stopped and restarted PCS.

06:55 Ran PCY315 script. MOS error. Meanwhile, Steve loaded new collimation model. Still it has issues, and after some typos were fixed, still SX M1 is not in but close to the limit.

07:20 MOS error recovered. Init MOS on LUCI2. Going back to field using PCY315 mask.
  • L1: 24,25 and 26(mask) dtheta=0.16 (dx=0.83, dy=-0.68) looks like we had the wrong sign
  • L2: 26,27 and 28(mask) same for this one, dtheta =-0.795 deg.
07:43 Edited PCSInstrument.conf.newpointing: LEFTZEROPOINT now -187.805 and RIGHTZEROPOINT now 177.05. cp -p PCSInstrument.conf.newpointing PCSInstrument.conf. Steve bounced PCS.

07:45 Run script again.
  • L1: 27,28 and 29(mask) Using same stars as before: 30,34,35, calculate dx,dy,dtheta=1.3,-.37 and 0.035 (better).
  • L2: 29,30 and 31(mask) " " dx,dy,dtheta= -5.11, 0.42 and dtheta=0.00992 deg (better).
  • Send corrections for thru-mask confirmatory images. Look OK (not all stars are in boxes, as we saw earlier... mask design), but those stars that we saw in boxes before are there.
08:11 Going to StoneA for collection of transform data.
  • mkdir /home/lbto/20170904_rfbg and _lfbg.
  • start idl in two term windows, in each, at idl prompt, run iif_register, /LEFT (or /RIGHT)
  • Configure LUCI:
    • N30_FieldStop
    • N3.75 camera
    • mirror
    • flex comp off (?)
    • filter = K
  • transform_collect in each window. Chose Stone_A3. On the left side, we have an error because transform_collect is acting as if it wants to take an image with IRTC, although we tried iif_register,/tcs and iif_register,/left and on the right side, the script that tries to take a LUCI image runs, but returns an error and the arguments to this command refer to LUCI1, not LUCI2, although we did iif_register,/right. LUCI is authorized on both sides. agw_set_station command is not found.
  • left side error message: IDL> transform_collect

    IRCL: preset, Stone_A3, MODE=ACTIVE, GS=0, TS=0, PA=-90.0000 (iif_preset)

    180 nm rms error in image left_wfscimage000342.fits
    => Collimated in 1 iteration

    IRCL: preset, Stone_A3, MODE=GUIDE, GS=0, TS=0, PA=-90.0000 (iif_preset)

    => guiding stopped
    => guide star set to SFP (x,y): 0, 150 mm
    return # = -1
    ===> Error in irc_getirparams
    err = Resmsg[0] = IRTC2 GetIRParams result status: ERROR
    err = Resmsg[1] = [Error] - IRSocket::IRSocket -> [::connect] Could not connect to port.
    ===> Error in irc_getfov
    err = Resmsg[0] = IRTC2 GetFOV result status: ERROR
    err = Resmsg[1] = [Error] - IRSocket::IRSocket -> [::connect] Could not connect to port.
    ===> Error in irc_getfilter
    err = Resmsg[0] = IRTC2 GetFilter result status: ERROR
    err = Resmsg[1] = [Error] - IRSocket::IRSocket -> [::connect] Could not connect to port.

  • Right side error message:
  • 417 nm rms error in image right_wfscimage000389.fits
    => Collimated in 3 iterations

    IRCR: preset, Stone_A3, MODE=GUIDE, GS=0, TS=0, PA=-90.0000 (iif_preset)

    % Compiled module: GCS_GUIDING_STOP.
    => guiding stopped
    % Compiled module: GCS_SET_GS_SFP.
    => guide star set to SFP (x,y): 0, 150 mm
    % Compiled module: LUCI_TAKE_IMAGE.
    => Requesting LUCI image
    % Compiled module: LUCI_SEND_REQUEST.
    ===> Error in luci_send_request: Error taking a Luci image
    request = ssh echo /home/lucifer/lcsp/bin/ wait pause 0.00000 1 1 | /bin/bash
    Exit 127
    /bin/bash: line 1: /home/lucifer/lcsp/bin/ No such file or directory

09:00 Switching to LBCs. Though there were clouds earlier, the allsky now looks clear. We'll run some photstds scripts and a science script that will further test guiding and dithering. Shane is driving.
  • Tried SA113, but it was setting and very close to the moon. WRS crashed - detection threshold too low (sky background 20,000).
  • moved on to Galadi.
    • collimate. Converged in 3 iterations of FPIA after WRS was done.
    • copoint. ds9/iraf conflict requires restarting both ds9 and iraf (like last night).
  • next observed SA95
  • 11:00 observe IC345. Confirm z4 computed for both red and blue. Last g image had 0.6" FWHM (2.7 pix) and about the same for i.
  • about the time we started the next Usz ob, we had a blue trackers RPC error. After trying (just in case) to play OB again. Powered off and back on other systems. Red trackers now showing an error. Meanwhile, remember btec is hardwired not to turn off upon turning off other systems, so I cycled power on it via (Perhaps only the blue controller needed to be turned off/on, not the PC, to resolve the RPC error?) Now they are turning back on the LBCs.
    • was very slow to come up. My user interface is very slow... Shane's also.... Steve is going to check the KVM in CRB.
  • Red HK did not come up: problems on power line. We tried to turn off, turn on but still the same problem. I rebooted the CMU, and ran lbckill/lbcstart, but when we try to turn on HK, the browser is hung on "Waiting for lbccontrol". Perhaps we need to reboot the dataprobes in the electronics cabinets?
  • End of night with LBC issues.
05:30 Closing

-- %USERSIG{OlgaKuhn - 2017-09-04}%


Topic revision: r6 - 04 Sep 2017, 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