MODS1 AGw Commissioning at LDG: 23 September 2010 UT

Observer: J. Hill (LBTO), R. Pogge (OSU), R. Stoll (OSU), M. Pedani (LBTO), B. Atwood (OSU)
Telescope Operator: DHuerta (LBTO)
Telescope Support: JLittle (LBTO)
Instrument Support: JMorris (LBTO)

Summary

Lost to weather..........clouds and rain in all directions. We've celebrated Chilean Independence Day (a few days late) with some Pastel de Choclo, and we celebrated the Autumnal Equinox. DCox ran a rotator balance test on MODS. We did some closed dome testing, and found a serious problem with the Leach Controller of the MODS1 guider AzCam that we were not able to solve. See Issue #2866, and the related Issues #2867 and #2868.

Preparations

Even with bad weather, we'll plan to start at the normal (2 UT) time with some rotator balance tests with Dan Cox. Some work on the MODS CCD controllers delayed that a bit. Heavy rain continues outside.

Marco edited the existing file (now renamed: PCSInstrument.20100922.conf) to put the working wavelengths for MODS1. We will do the transformation measurements with the Red camera and the I-filter so the LEFTWAVELENGTH is 0.750 microns.

We've consciously decided to keep the MODS pointing model derived from LDG testing with AGw4 in Oct 2009. We've got no reason to to think that this is worse than the present LUCI-derived model.

We woke up webcam lbtmc01 by cycling the power. Nobody's had the time/energy to hike up to lbtmc04.

Details

LDG Rotator Balance

03:50 LDG Rotator Balance measurements begin with DCox remotely. Telescope is at horizon. JMorris fills LBCs at the same time.

Fake Observing Tests and debugging Issue 2866

04:20 Authorize MODS fails.

Thu Sep 23 04:41:40.093 2010 55462.19560293372 1 IIF  iif.IIFCommand.authorize.alert left IIF TCS authorize alert for left MODS at directGregorian focus: AGW selection failed! side: ''L'', instrument: ''MODS'', FS: ''directGregorian''
Thu Sep 23 04:41:40.099 2010 55462.19560300052 1 IIF  iif.IIFCommand.authorize.alert left IIF TCS authorize alert for left MODS at directGregorian focus: searchKey: ''L_MODS_DIRECTGREGORIAN'', found: ''OSU_L.cfg'', error: No Error., in file ''/lbt/tcs/current/etc/../etc/GCS//Instrument2AGW.cfg''
Thu Sep 23 04:41:40.105 2010 55462.19560307725 1 IIF  iif.IIFCommand.authorize.alert left IIF TCS authorize alert for left MODS at directGregorian focus: selection of AGW failed.

We find the "error: No Error" syntax very peculiar!

The selectAGW left OSU_L.cfg command also failed. We restarted GCSL, but that didn't help. So we called Torsten. He thought the error message was peculiar too. By process of elimination we found that the AzCam server process in mods1-azcamg was failing at the PowerOn (which turns on the amplifiers in the controller) step of the startup script. (Even though Torsten has run this many times in the last few weeks.) Note that the AzCam server processes have to be restarted manually at the moment - pending a power switch on the AzCam computers in LLTH.

The message in the AzCam server window is:

RunScript:PowerOn
*** Reply: ERROR "Power_On reply:unexpected reply ERR"
RunScript error for file \azcam\systems\lbtguider\initlbtguider.txt

In the meantime, we have retracted the SX M3 Swing Arm, and turned the cooling in URTH back on.

See Issues 2866, 2867 and 2868.

05:15 Rick and Rebecca check the fibers from the Leach Controller to the AzCam computer and finds that the fibers are fine at the instrument and in the LLTH.

Torsten and Paul have the next idea that it could be the power supply of the Leach Controller, but we don't have an immediate way to test that. So debugging continues to see if we can isolate the problem to the Leach controller.

Torsten points out that we can set GCSL.no AzCam flag in lbt.conf to ignore this problem. (see below)

05:21 Turn off controller power for several minutes (on MODS interface). That didn't help any. Same error at PowerOn in the script.

05:32 Send Rick and Rebecca to the LLTH. Swap fibers to run WFS Leach Controller with guider computer lbt1. That worked OK.

05:35 Swap other fibers to run Guider Leach Controller with WFS computer lbt7. That failed same as before.

05:38 Restore the fibers between AzCam PCs and Leach Controller to the normal configuration.

05:40 Power cycle the guider computer lbt1 in the LLTH. The script failed at power on after reboot. Cycled the controller power again, and it still failed.

We are suitably impressed that nobody has modified the initialization script since January 2007. The configuration seems stable enough.

05:49 set GCSL..noAzCam flag in lbt.conf to false. Now the authorize succeds because GCSL is ignoring AzCam. (Danger - this flag affects all focal stations.)

Test Preset

05:55 Send Closed Dome Preset from MODS to StoneA1 - Telescope moves, but preset fails.

Sep 23 05:55:02.658 2010  Preset alert for left MODS mode TRACK: no AGW unit selected.

The selectAGW left OSU_L.cfg command still fails.

05:59 preset track on meridian to equator - still fails for no AGW unit selected.

This puts LDG rotator at 175 deg. Parallactic angle is roughly -4 deg. But MODS was clocked 30 degrees at mounting, so we need to change LEFTZEROPOINT so this rotator angle is 210 deg instead of 180.

06:04 Change MODS LEFTZEROPOINT from 0 to +30 degrees in PCSInstrument.conf

06:05 Resend preset and rotator goes to approximately the right place. (sometimes you get the sign right!)

06:07 Another test preset at Dec=+10

Change GCSL.noAGW to true in lbt.conf. Stop and restart GCSL.

06:10 Another test preset --- succeeds, and Rick gets the proper return.

06:13 Send offset in angle, and it works. But we noticed that the global angle wasn't what we expected. The reason is that the telescope GUI in the MODS control panel is picking up status info and putting in the current rotator angle in place of the requested celestial position angle, screwing up the next preset. Rick knows the bug (he thought he'd stomped that one a while ago) and will fix it tomorrow. In the meantime we can use the script/command-line interface commands in the MODS control panel that work fine. A bad old version of the TCS gui panel object must have gotten recycled back in (we don't get to test the telescope stuff much until we get to the telescope).

Another bug noted is that the MODS IIF agent is returning the HA with a bad branch-cut calculation as -23h instead of +1h every now and then. Oops. Rick won't even try to debug that until he's more awake tomorrow. It will mean some FITS headers will have odd-looking (but formally correct) HA values in them.

IMCS Test with Telescope Tracking

06:31 Send preset tracking for an hour crossing the meridian at EL=80 with IMCS Red running.

RA=23:50 Dec=+25:00, within ~30min HA east of the meridian. This gives us about 1h of tracking with a fairly aggressive rotation on zenith approach (1 deg/min or more).

MODS1 setup in red-only imaging mode, sieve mask + calibration illuminator (VFLAT level 2.5). 15s integration on 3Kx3K roi readout. Script imcstrack.pro executed. Exposes 15s, reads out with IMCS locked, then sleeps 2 minutes "with one eye open" using the IMCS to lock during its sleep interval. The result should be equivalent to a series of long exposures covering 1h of observations across the meridian. Sampling of elevation and rotator change will be roughly 10 images before and after transit, so images every 3 minutes.

07:34 preset to another field on the equator. This is a similar setup to the previous in terms of the instrument, but at the equator we have a less aggressive change in the rotator angle to maintain PA=0 on the "sky". This new sequence starts at image 0020. One image (0026) has an apparent jump due to a readout glitch interpolating extra rows in at the top/bottom quadrant boundary split in the 4-amp readout.

07:43 Turn on automatic transmission of MODS images to archive. Earlier images were sent manually to the archive.

08:30 data collection ends. Last image is 0036.

08:31 Up to zenith to top off MODS dewars.

08:45 End of MODS testing. Hand telescope over to Nancy for LUCI tests.

-- JohnHill - 22 Sep 2010