You are here: Foswiki>Commissioning Web>Com20080412 (16 Apr 2008, NormCushing)Edit Attach

Commissioning Notes 20080412 UT

Observer: J. Hill and a cast of dozens others both in the control room and in Tucson
Telescope Operator: D. Huerta

Summary

We fixed a number of problems with PSF, OSS, MCSPU, PCS, irs, LBTtools, and we seem to have everything working. Unfortunately, the wind was too high and we were only able to open the dome for half an hour. We spent the half hour raster scanning by hand trying to find the star image. We did see a star flash across the screen once.

Misc. Tests in the late afternoon

a) Italo set IRTC server time to UT. (fixed)

b) We need Jose to change the IRTC filenames to UT (rather than MST)

c) Dave T. is updating LBCtools.IRTC to control IRTC exposure time (fixed)

The daytime engineers have centered the Rigid Secondary over the SX Gregorian Rotator using the Hexapod and alignment telescope measurements. The nominal centered position of the SX M2 Hexapod is X=+5.50mm Y=+3.50mm Z=0 RX=-75arcsec RY=-280 arcsec. (There were conflicting notes about the sign of RY, so we operated during this night at the incorrect value of RY=+280 arcsec.)

01:30 - Ross is trying to measure appropriate primary tilt with the multiple reflection of the secondary in the primary in the secondary. This experiment didn't have enough illumination, so we will try again tomorrow afternoon.

Nighttime Ops

02:40 - Daytime engineering is done so we are getting ready to swing in the tertiary and balance the telescope. Joar stopped his upstairs OSS, and we've restarted the new one downstairs (on the new TCS that includes the TO station). Tom S. has switched us to the rotator capable version of MCSPU. Dave A. is preparing for a balance run with SX M2, SX M3 and DX PF (he claims that isn't a standard configuration!)

02:45 - The PSF LEFT Secondary sub-GUI still has the primary mirror position limits operational. Chris is working on disabling this and putting in the hexpod limits which are loosely Z +-12.5 mm X,Y +-15 mm. We don't know the angular limits so we'll apply a large number like 10000 arcsec (Chris fixed it at 03:20)

While we are still making progress, the bad news is that the wind is blowing at 38+ mph outside.

03:00 netconfig stop and re-start OSS, to try to get it talking to the new MCSPU, but it didn't help. Paul is working on it......

03:05 Dave A. reports HBS not giving enough pressure to keep the elevation friction down.

03:20 pointto command isn't able to preset the telescope, and that is still true after we authorize as IRTC. First problem was that GCS was not running. We started GCS LEFT.

03:30 Next problem is: [I] PCS setNewTarget failed; target is below horizon. RA: 0.000000000000000e+00 Dec: 0.000000000000000e+00 Az: 5.356042590117504e+00 El: -7.516779179983517e-01

Normal presets from the Preset GUI work, but the preset coming through irc/irs fails.

03:45 - MCSPU is not getting rotator polynomials, and the MCSPU rotator page is not updating even though it worked a few hours ago.

03:50 - MCSPU crashed approximately coincident in time with the building stopping when Jon R. switched the oil pressure. Jon and Dave are running a friction versus oil pressure experiment, so Dave is doing another balance run.

03:58 - test takepic, and it correctly controls the camera exposure time.

But: [E] GetMultiParam: The list of DD entries is empty. [E] GetMultiParameter execution failed: GetMultiParam: The list of DD entries is empty.

04:03 Stopped (irc -s) and restarted irs, but it still is the GetMultiParam error. along with [E] Exception was caught:[Error] IRSocket::GetImage::GetPar -> Could not read from socket.

04:11 Dan is running diag21k to chase down a problem that seems related to MCSPU telemetry.

04:27 MCSPU display hangs on the "rot lfbg" page.

04:30 Tom recompiled MCSPU software.

04:35 Dan killed zombie mcstemp in the basement.

Did various stops and restartings of PCS, MCS, PCSGUI, PRESET GUI.

04:52 MCSPU test with MCS subsystem off, but the MCSPU hanging persists.

05:04 Restart mcstemp and rpcserver on jet. This videocon of a roomful of programmers at work is a powerful new form of reality tv. And more seriously, it is like they are right here in the room adjacent to the control room. The only drawback is that we can't eat their snacks.

05:15 Control room getting warm.......David goes off to restart the chillers in the utility building. Wind is still 39 mph.

05:20 Rotator polynomials are now flowing to jet and the rotator is tracking!

05:43 Restart mcstemp without restarting RPCserver, and it works.

05:50 Restart mcstemp again, after the rotator page hung.

06:04 Restart mcstemp and PCS again, after the rotator page hung.

Recipe for restart: start MCSPU, start PCS, Preset

MCSPU fails when you execute a slewtotrack if polynomials are already running, or is it because there are bad polynomials.

07:06 Jose figured out that the preset problem is because the wavelength in target.dat needed to be microns and not nm. I copied Dave's LBTtools to /home/LBTO/iraf/LBTtools and made this change. Then the preset worked OK.

07:20 The get multiparameter problem was a typo in the filename. So the error was a GetParameter call with a non-existent file. The action for Jose should be to check for the existence of the file so it makes a cleaner error. The action for Dave is to build in the new file /home/LBTO/iraf/LBTtools/IRTC/std_irtc_dd.dat Note that I have editted Jose's getdd.dat file to correct some typographical errors (renamed to Dave's name above) and emailed it back to Jose.

Next Issue: If PCS supplies a Nan, you get the following error from irs:

[I] GetMultiParameter command status: SUCCESS FITSIO status = 402: bad float to string conversion Error in ffd2e: double value is a NaN or INDEF

07:45 Install new build of MCSPU (rebuilt to the very latest commonsw), although it takes a couple tries to start correctly. This fixes a crashing problem caused by the event software, but it still hung. Perhaps we need to start PCS after rotator is up and running?

08:20 The theory is that there is a timing problem between the polynomials and the start tracking commands.

OSS crashed for an unknown reason. After Paul restarted it, we were able to test the moveOPE command in both relative and absolute modes. It adjusts the PSF M2 Instrument Offsets as advertised. I've also verified that moveOPE to M1 serves the same purpose as StepFocus to update the primary mirror collimation at the present elevation.

08:45 The programmers in their sleep-deprived delirium think that have fixed something related to the timing of slewtotrack relative to the first good or bad polynomial. Or maybe the problem is still there, but the freezing up that it creates is less.

Made an script (inex.cl) to take intra/extra focal pairs:
# update the M1 collimation
moveOPE M1 X=0.0 Y=0.0 Z=0.0 RX=0.0 RY=0.0 RZ=0.0 ABS+ Magic-
# move to the extra/intra position (invented offset)
moveOPE M2 X=0.0 Y=0.0 Z=0.01 RX=0.0 RY=0.0 RZ=0.0 ABS+ Magic-
# take a picture
takepic exptime=0.1
# move to the other position
moveOPE M2 X=0.0 Y=0.0 Z=-0.01 RX=0.0 RY=0.0 RZ=0.0 ABS+ Magic-
# take another picture
takepic exptime=0.1
# put the secondary back for good housekeeping (must use Magic+)
moveOPE M2 X=0.0 Y=0.0 Z=0.0 RX=0.0 RY=0.0 RZ=0.0 ABS+ Magic+

Do we need to label the images as intra or extra focal? It won't be easy to tell with the M2 position, although it would be easier (and slower) with moving M1.

09:23 Tom S. has the aha moment, and realizes that mcstemp is neglecting to unlock the mutex on early exit.

Opening the Dome

09:38 Wind has dropped to ~25 mph, so we decide to open the dome and go for it. T=-2degC

10:12 offset 0 -40 command has irritated the PCS gui in a bizarre way. The apparent pointing seems to be jumping by minutes at a rate higher than 1 Hz, but the telescope seems to be responding correctly to my offsets.

10:19 Preset to Beta Her

Saw the star go by while telescope was settling...!!!! The communal impression was that it was something resembling an image, and not just a whiff of diffuse light.

Managed to scan in 8 x 20 sec steps in each of the cardinal (RA/DEC) directions. But, did not find the star. SX primary mirror was at X=-0.5 Y=+2.5 Z=0.2 RX=38 RY=-45 for EL=78.7 Probably we should make it more centered to start next time. SX secondary mirror was at X=5.5 Y=3.5 Z=0 RX=-75 RY=280 (RY was incorrect - see above)

10:33 Close Dome due to increasing wind (18 m/sec on the ground)

10:48 Software guys go home. The hardcores will hang around and hope for another gap in the wind.

11:46 PSFR has a problem that it is not getting a collimation lookup table. Stopping and restarting PSFR doesn't help. Then the semi-asleep realize that IRTC is authorized, and nobody has ever made a collimation table for DX IRTC. It works fine if we change the authorization back to LBC.

Things We Need for Tomorrow (none are an emergency)

  • RFG is bringing up a new computer to run EF on
  • Stephen is installing a new build that includes the bug fixes made to TCS tonight.
  • Instructions for starting and recovering the Hexapod from OSS. (I've learned $$$, but I suppose there are more tricks than that.)
  • Instructions for starting the rotator in MCSPU
  • If anybody really knows, we could use tooltips or postits that say which physical direction +X, +Y and +Z move the hexapod.

-- JohnHill - 12 Apr 2008
Topic revision: r5 - 16 Apr 2008, NormCushing
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