March BG Com Night 12, UT 2009 March 18
Observers: Dave Thompson, NAgeorges, MJuette, WSeifert (@LBT)
Telescope Operator: DGonzalez-Huerta
SW Support: TSargent until 04:00
Telescope Support: DOfficer, JLittle
- Jesper's latest transformation appears to work!
reality check: now it needs more detailed testing, verification
- New pointing model OK, but a large offset was needed tonight.
- Imaging data taken, use for testing overheads.
- Ringing in AZ(?) servo noted, takes a majority of the offset time
- Took collimation dataset
Twilight flats with LUCIFER.
Startup problems. oss.side.tertiary_mirror: [adjustCurrFocalStation] Not at a focal station. Apparently we need to de-authorize and re-authorize LUCIFER because OSS must know what instrument to expect commands from? Time lost: 16 minutes
Verify new pointing model
New pointing model generated from previous night's data. Changes are evident from prior models. Current rms = 5.61 arcsec, but there are additional terms that could be added to reduce this. I emailed the input data and my model to Pat Wallace asking for comments and suggestions. Initial correction of the pointing tonight is quite large, ~72 arcsec. DGH reports that the temperature seems to be fluctuating, and is a couple degrees warmer than yesterday. Can this make this big of a difference in the pointing? Can it be from being way out of collimation? If so, then the startup procedure should be:
- correct pointing to be able to acquire a reference star
- collimate on that star
- correct the pointing now that telescope is collimated
The new pointing model is saved as LUCIFERSX.20090317.ptmod and saved in the appropriate place. I also made it the default pointing model for LUCIFER on restarts of PCS.
Take data on LUCIFER field now because it is setting (S255 06:12 +17:59)
Initial collimation bad, very comatic. We had to go to a brighter star and allow the WFS to get close (after a couple corrections, the bright star was saturating the WFS in 15s), then to a fainter star to complete the collimation.
MST20:33 script started. Collimation not converging well (went below 400, then up to 580, then back to the 350 range. Seeing is ~0.6 arcsec on the guider so it should be able to do better.
20:40:37.345 2009 Offset alert for left LUCIFER: Primary mirror collimation reached limit (Y=-2.6 mm). We had to add 2mm to M1 and M2 to get the primary off its limit. 20:50 back to being collimated (10 minutes lost).
Offset while integrating (done on purpose)
luci_20090317_0273, 0275. The basic offset seemed to be OK, but there is a clear oscillation or ringing at the final position. See image. The offset between the left and right stars (black trail) was one arcmin.
- AZ servo ringing?:
Note that the ringing takes up a majority of the time to execute the offset (the telescope motion part), as evidenced by the brightness of the trail there vs. the part covering the initial offset between positions.
MST22:19 Done with the LUCIFER imaging script.
*Test of new AGw transformation*
Jesper generated a new transformation for oacontrol from the data taken two nights ago. While the rms is higher than earlier fits, it is still in the 200um range. We see large offsets from the rotator center on LUCIFER, so some step in the transformation must not be correct.
22:20:47.135 Preset to BS9143 GS=24 PA=90. Star not found...correcting the pointing (15 arcsec off in CA).
|| successful, using new transform
|| successful, also using new transform
copied oacontrol.conf.sfp20090316 over oacontrol.conf. Old version saved to oacontrol.conf.20090318. *No scale change in AIP_L.cfg put in place because it is not clear what to change! Preset again to star 25 successful.
The on-axis position for the two images using GS24 lie on top of each other, as do the two for GS25, however the pairs of 24-25 are still separated.
MST00:25 Restarting the oacontrol server in case that was necessary to have it use the new transform. Also did a soptAGW/startAGW and homed the motors. That worked, but now the internal/external transformations disagree, so we do need toi change the scale factor in AIP_L.cfg:
Mar 17 23:05:36 lbtmu107 LBT_left_GCS: AGWUnitAIP: internal transformation to AGW native coords: 96.808531, 481.271252 mm
Mar 17 23:05:36 lbtmu107 LBT_left_GCS: AGWUnitAIP: external transformation to AGW native coords: -0.976179, 611.840753 mm (oacontrol)
Mar 17 23:05:36 lbtmu107 LBT_left_GCS: AGWUnitAIP: moving probe to native coords: -0.976179, 611.840753 mm using external transformations
MST23:12 Changed rad_to_mm to 1.0 (previous value 124410.0. Now the collimation is going haywire. John suggests we clear everything back to the default collimation.
23:25:14.329 2009 Preset on-axis to collimate on BS9143, OK.
Changed AIP_L.cfg back and trying a preset to GS25 (same as above), external transformation definitely bad....so AIP_L.cfg rad_to_mm changed back to 1.0. Now the magnitude of the coordinates are OK, but there is an Xflip, so we change pcsmx=+1 and restart oacontrol, stopAGW/startAGW.
No on-axis star...correct pointing again? Why is this so bad?
|| after changing both rad_to_mm and pcsmx, plus restart oacontrol
|| it looks like it worked
|| Yes, it did!
00:13:41.487 Preset to gs=25, succeeded
00:18:08.106 Preset to gs=24, succeeded no offset on LUCIFER!
00:21:40.545 Preset to gs=26, succeeded
Now for multiple off-axis guide stars the on-axis source on LUCIFER stays basically stationary. This needs to be further verified and calibrated, but the large offset now appears to be gone.
Testing another field with LUCIFER (FS33). See luci_20090317_agw2_*.fits. Testing multiple fields/PAs with a similar dataset as from yesterday.
Utility building chiller is off...David is going down to restart it. John suggests turning off mirror ventilation until things recover.
01:21:33.326 couldn''t move probe to position (SFP) -71.2183, 120.615, focus: 21.3744 :AGw command status: AGW1: Error on UMAC
But then the next offset after this succeeded with no intervention or warnings. We did nothing to recover from this, and the system seems to be going OK.
Offsets seem to be placing the star much closer to the hotspot than before, an added bonus of the better transformation. However, the guiding offsets do still need to be looked into (check/fix the CCD2SFP
rotation used in GCS and/or have a native transformation in the oacontrol library).
Spoke to Jesper. He will try to make code to correct for the hysteresis in the AGW positioning for tomorrow night, so no need to take more data on this tonight.
Correct pointing for next task. IE changed from 25.00 to , CA from 30.00 to.
MST02:30 Take COLLIMATION dataset
Secondary AO cleared (now at its nominal position), M2 offload cleared from M1, all collimation now sent to M1 (not "split" mode). AO on M1 and c00 forces left as they were.
Even found star at EL~14, FWHM~1.5
MST03:36 Jesper called and said he found a bug in his code, resulting in a small improvement He is still looking into things and will email and of his day with more info.
MST03:55 GCS-AGW communication problem. Restarting GCS (and AGW).
MST04:09 IE corrected 25.0 to 13.0, CA from 30.0 to 23.0
MST04:56 Done with collimation data collection.
LUCIFER to run a script next (0.4-0.5 arcsec on guider).
05:48:50.585 Offset hung with GCS waiting for a guide image. Restart GCS at 05:56:07.363.
06:00:46.820 Offset again hangs with "error requesting an image in GuidingThread
. Lost connection?" GCS and AGW restarted, AGW homed.
06:07:37.073 Preset again
Did not do anything below except maybe practice obs for (1) below...
Practice run on Walter's official demonstration OB's
We propose here four observing blocks, each with 1 hour on source integration:
(1) Imaging: 2sec DIT (+ 2 sec read out) (+ imaging acquisition (inclusive preset) ~ 5min)
(2) Imaging: 300sec (+ 2 sec read out) (+ imaging acquisition (inclusive preset) ~ 5min)
(3) Longslit spectroscopy: 300 sec (+10 sec read out) (+ spectroscopic acquisition ~ 15 min)
(4) Longslit spectroscopy: 900 sec (+10 sec read out) (+ spectroscopic acquisition ~ 15 min)
Optimization of new GCS acquisition algorithm
Collect data for new collimation model
Missing data for performance report
BG Berformance Verification
- 16 Mar 2009