LFBG Re-Commissioning with LUCI: 16 September 2010 UT

Observer: DThompson, JHill (LBTO)
Telescope Operator: SAllanson and GBechetti (LBTO)
Telescope Support: JUrban (LBTO)
Instrument Support: DOfficer (LBTO)

Summary

Seeing wasn't so great tonight (0.7 - 1.5 arcsec), perhaps because we had more solar heating during the day. We updated the collimation models and the pointing model for LFBG. We checked the transformation measured last night and analyzed by Jesper. After getting the flips and rotations straightened out, the residuals were 33/31 microns over a small range of position angles. When we rotate over all position angles in a given place on the sky, the residuals (measured on the LUCI science detector) are dominated by LUCI internal flexure (even with flexure compensation on). We tested automatic field aberration data collection scripts in IRAF and IDL. We checked the DETXY offsets with the new transformation. Seeing improved to 0.5 - 0.8 arcsec at the end of the night, so we got some decent field data with the -20 arcsec Tilt correction on M2 from last night.

3-Night Overview

We've nearly finished the important AGw3 and LUCI recommissioning activities. Just a few tasks remain to be done:

• Refine pupil position parameters for WFSing in IDL and GCS (Doug off-sky), and install those.
• Complete the Field Aberration Measurements and M2 alignment (Andrew, et al.). This needs about 0.5 night of sky time in decent seeing. Andrew says more time could be needed to sort out the M3 angle. (21-SEP)
• Adjust the zeropoint of the new pointing model after the Field Aberration adjustments are completed. (22-SEP)
• Adjust the global offsets in the M2 collimation model to include the Field Aberration adjustments. (22-SEP)
• We have not explicitly measured the AGw-to-LUCI focus offset, but it seems to be very close to optimal as we had images at 0.3 arcsec FWHM on LUCI.
• We have not re-measured the photometric or astrometric performance of the guider CCD, but that is not very important. Nor do we have reason to think it will have changed.

See Com20100921 for the next round of field aberration measurements

Preparations

Tony has patched Telemetry r10 to fix Issue #2832.

Jesper has created a new transformation based on last night's measurements. We have installed it in oacontrol.conf.

Added these global offsets from last night to the SXPMLUCIFERCollimation.dat file.
Xglobal=-0.684 Yglobal =2.196 Zglobal=+0.186 RXglobal=+23.73 RYglobal=+6.83
These globals were only used for collecting the collimation model data until 04:00 UT.

Details

Open

01:40 Open T=12.2 degC, D=2.4 degC, Wind light from west. A few scattered low clouds.

01:46 Dave adjusts the pointing on a bright star. IE=-58.4 CA=-23.4 New collimation adjustment a big improvement.

01:53 Preset ACTIVE BS9178 PA=180. Dense field, take astrom data. Luci#0006, 0007=sky Ks+Br_gam 6x10.0s.

Update AGw transform with Jesper's new solution from last night's data.

02:03 Preset again to BS9178 PA=180. Luci#0008, 0009; 0010 at PA=150

Check current collimation model.

Collimate on BS9178 EL=52
Check on ACT0081 EL=20 not good
Check on ACT0302 EL=83 OK

Take collimation model data.

Star	EL	NNNN	Notes
BS9170	52.5	0013	parallel offset Z -1.0 mm after this image
BS9160	34.1	0014
BS9155	21.1	0015	Poor collim, elongated, shading across pupil?
BS9160	32.5	0016
BS9164	44.1	0017
BS9170	49.0	0018
BS9183	55.6	0019	correct the pointing, then...
FS149	81.7	0022
FS151	69.8	0023
FS152	49.0	0024
BS9169	56.5	0025
BS9177	69.8	0026
BS9183	60.3	0027
BS9182	50.7	0028	parallel offset Y -2.0 mm after this image
FS149	87.2	0029
FS151	78.3	0030

Header data for these images collected in /home/lbcobs/Collimation/new/LUCI_20100916.txt. That file has already been corrected for the parallel offsets of M1/M2 implemented above.

Analyzed this data with /home/lbcobs/Collimation/new/mkcoll_irtc.pro. See attached plot of the results below.

Take pointing model data.

pointing corrected IE=-73 CA=-26 after implementing new collimation model SXPMLUCIFERCollimation.20100916b.dat and SXSMLUCIFERCollimation.20100916b.dat

mount log file started: /home/telescope/20100916.ptlog
04:23 Preset to BS9183 to collimate

P5	UT	SourceID	LUCI	PM-Corrected Coordinates	Notes
01	04:29	WT10_373	0033	22:25:52.5870 53:56:40.714
02	04:33	ACT0407	0034	5:00:53.9876 77:46:31.820
03	04:35	WT10_401	0035	15:37:45.7965 71:09:13.378
04	04:36	WT10_358	0036	18:56:32.1475 46:45:58.154
05	04:38	ACT0281	0037	15:44:55.8963 21:26:41.171
06	04:40	WT10_179	0038	18:24:30.0223 -8:10:30.615	unwrap AZ
07	04:45	ACT0253	0039	20:09:00.7119 13:00:57.024
08	04:48	WT10_068	0040	19:57:37.1584 -37:44:41.772
09	05:02	WT10_168	0041	22:54:07.1594 -12:11:26.003	AZ glitch @4:51, IT#2836
10	05:06	WT10_274	0042	22:32:27.4868 19:37:25.416
11	05:08	WT10_225	0043	1:24:54.5826 3:20:57.709
12	05:10	ACT0325	0044	1:09:17.3673 34:48:21.132
13	05:12	WT10_338	0045	0:19:46.6816 36:57:44.574
14	05:13	WT10_278	0046	1:22:24.7100 22:05:11.843
15	05:15	WT10_242	0047	0:52:28.8808 13:38:55.429
16	05:16	ACT0306	0048	22:11:23.1754 29:15:16.449
17	05:18	ACT0185	0049	22:32:34.2678 -6:28:01.540
18	05:20	ACT0085	0050	22:22:04.5978 -34:29:19.770
19	05:22	WT10_115	0051	21:31:51.3537 -25:24:45.856
20	05:23	ACT0289	0052	21:35:34.2115 23:58:39.601
21	05:26	WT10_270	0053	19:34:45.3084 16:38:29.937
22	05:28	ACT0213	0054	17:21:06.9382 1:26:37.913
23	05:30	WT10_300	0055	17:52:32.1803 27:11:36.962
24	05:31	WT10_334	0056	20:51:30.9327 37:59:22.148
25	05:34	ACT0398	0057	21:09:31.6912 68:29:26.728
26	05:36	ACT0406	0058	14:18:36.9088 75:40:07.535
27	05:37	WT10_410	0059	1:55:42.7541 81:57:32.080
28	05:39	ACT0321	0060	21:56:08.7664 34:33:01.307
01	05:42	WT10_362	0061	22:51:50.8461 46:44:07.784	repeat first 12 pos'n
02	05:44	WT10_407	0062	6:35:26.9137 80:34:23.631
03	05:46	ACT0396	0063	17:33:18.0427 67:12:13.905
04	05:48	ACT0347	0064	20:13:23.7982 43:48:40.236
05	05:49	ACT0283	0065	17:09:27.2491 22:05:29.951
06	05:52	WT10_181	0066	19:40:45.1713 -7:30:45.061
07	05:54	WT10_255	0067	21:32:59.1655 12:32:22.292
08	05:55	WT10_070	0068	21:29:49.6151 -37:41:34.184
09	05:58	ACT0170	0069	0:27:06.9215 -9:52:23.220
11	06:00	WT10_227	0070	2:41:21.6144 4:25:39.460	(cell 10 empty)
12	06:01	ACT0327	0071	2:46:06.7721 36:16:00.305
10	06:03	WT10_276	0072	23:58:19.1232 19:41:36.791

Pointing model plot from TPOINT:

Sky RMS = 4.87 arcsec. File LUCIFERSX.20100916.ptmod copied to the usual place and made the default in PCSInstrument.conf. PCS restarted ~UT7:40 to read in the new model.

Check the New Transformation

06:17 pointto StoneOagw gs=20 PA=180 Tel="ACTIVE" 0074 1036.4 1046.0 # EL=55
06:22 pointto StoneOagw gs=12 PA=180 Tel="ACTIVE" 0075 1036.4 1033.5
06:25 pointto StoneOagw gs=14 PA=180 Tel="ACTIVE" 0076 1028.3 1012.1

Turn on LUCIFER Flexure compensation and re-check
06:28 gs=14 PA=180 0077 1028.2 1013.1
06:29 pointto StoneOagw gs=12 PA=180 Tel="ACTIVE" 0078 1036.6 1033.6
offset to PA=225 0079 1028.0 1039.9
offset to PA=135 0080 1037.6 1020.5
06:38 pointto StoneOagw gs=20 PA=180 Tel="ACTIVE" 0081 1036.5 1046.0 # EL=53

Replace previous transformation:
06:44 pointto StoneOagw gs=20 PA=180 Tel="ACTIVE" 0082 1027.7 1046.4 # EL=52
06:46 pointto StoneOagw gs=14 PA=180 Tel="ACTIVE" 0083 1023.1 1029.0

06:52 pointto StoneOagw gs=3 PA=180 Tel="ACTIVE" 0084 1028.6 1035.4
06:54 pointto StoneOagw gs=3 PA=225 Tel="ACTIVE" 0085 1029.2 1029.7

06:57 pointto StoneOagw gs=7 PA=270 Tel="ACTIVE" 0086 1028.6 1035.4
07:00 pointto StoneOagw gs=9 PA= 90 Tel="ACTIVE" 0087 1031.2 1028.7

Neither transformation is very good!
For now, AGw rotation center at ~~ 1029.90 1032.05 on LUCIFER (use for pointing model)

Script Development

07:02 Stop GCS WFSing on the present star

after some fooling around with output redirection and fixing typos, I've got agw_collimate.py working correctly so it can call Doug's script collimate_multiple.pro.

Practice Field Aberration Data on BS9107

This is mainly practice data as the seeing is worse than 1 arcsec.

07:52 Apply -20 arcsec of Tilt to M2 (Andrew and Doug have a complicated debate about this.) This gives adjustments of X=+1.021 RY=196.40 to M2.

working in /OldRepository/AGW_Data/100916/
This first run has the interactive threshold set to a large number (400) for testing.

08:01 cl < home$field_BS9107c.cl

GS	wfscimage
0	5-8
82	9-11
0	12-13
35	14-16
0	17-18
19	19-21
0	22-23
64	24-26
0	27
96	28-30
0	31-34

manually interrupted here........

Testing Doug's Field Data script

08:35 Doug starts testing his script on BS9107 - seeing is 0.72 - 0.95 arcsec.

He's working in the parallel directory /OldRepository/AGW_Data/20100916/
Note that once again Doug and John have parallel directories 20100916 and 100916. Doug's numbering of the wfsc images starts again at 0001.

08:40 Probe couldn't reach gs=70 (that it could previously reach).
08:49 Probe couldn't reach gs=68 y=262 - This is because he was accidentally using a different central star as the target.

09:09 retesting with the normal central star. Now we have the same star, but Doug(IDL) and Dave(IRAF) don't agree on the proper motion corrected coordinates of this star (gs=0 = gs61).

Note that the IDL pupil parameters need fine tuning when the guide star is far off-axis. Doug will need to fix that when he analyzes this data. Then we will propagate it to AIP_L.cfg.

09:28 now John running field_collect as LBTO. Seeing 1 arcsec. Doug goes to bed.

GS	wfscimage
0	19-24
82	25-27
0	28-30
35	31-33
0	34-39

09:47 didn't converge in 6 iterations, so we stop and do something else.

Another Transformation Check

Jesper thinks he may have gotten his X-flips and Y-flips confused. So he's now sent us a new transformation in oacontrol.conf.20100915new2

09:48 Installed the new transformation file

09:50 Preset to ...... and try different guide stars

The good news is that oacontrol.conf.20100915new2 is a much better transformation. So whatever Jesper flipped was the right thing.

Jesper reports:

... I got tangled in the usual flipping and
rotation circus but now things are clear and I believe I understand exactly
what is going on. I have attached my little awk-script for preparing the data
for the fitting program so you can include the LUCIFER variant in your IDL
scripts. I suppose that using LUCIFER will be the standard calibration
procedure in the future, so I thought I might spare you some flipping and
rotating...

  I should note that the last fit had stddev values of (x,y)=(40,30)microns.
If I reject one of the 22 stars (possible proper motion error?) I get a stddev
of (33,31), so this should really be as good as it gets. At this level it
seems that the LUCIFER flexure contributes significantly to the errors so to
reach this low stddev the PA should be kept within a small range.

DETXY Offsets (check offset orientation along slit)

10:12

Script offset	NNNN	Xcen___Ycen
OFFSET = 0.0 60.0	0103	1012.9 1541.8
OFFSET = 0.0 -120.0	0104	1013.2 544.1
OFFSET = 0.0 120.0	0105	1012.9 1541.8
OFFSET = 0.0 -120.0	0106	1013.3 543.9
OFFSET = 0.0 120.0	0107	1012.9 1541.8
OFFSET = 0.0 -120.0	0108	1013.5 543.6
OFFSET = 0.0 120.0	0109	1013.0 1541.4
OFFSET = 0.0 -120.0	0110	1013.4 543.7
OFFSET = 0.0 120.0	0111	1012.9 1541.3

Had problems with NaNs in the Rotator Polys. See Issue 2839. Stopping and restarting PCS got rid of them.

Another shot at BS9107 Field Aberration Data

10:56 field_collect with bs9107_field.list - seeing is 0.5 - 0.9 arcsec on guider

GS	wfscimage
0	40-44
82	45-47
0	48-50
35	51-53
0	54-55
19	56-58
0	59
64	60-62
0	63
96	64-66
0	67-68
40	69-71
0	72-73
35	74-76
0	77
48	78-80
0	81
8	82-84
0	85
68	86-88
0	89-90
96	91-93
0	94-95
39	96-98
0	99
97	100-102
0	103

12:03 field_collect with bs9107_field.list - seeing is 0.5 - 0.7 arcsec on guider

GS	wfscimage
0	104-105
82	106-108
0	109
35	110-112
0	113-114
19	115-117
0	118
64	119-121
0	122

Andrew reports the next afternoon that we got the magnitude and the axis of the M2 pointing adjustment right, but the sign was wrong. We should have applied +20 Tilt to M2 (see below).

Note that IDL was sending Zernikes Z4 through Z20 for this dataset. See Com20100921.

Close

12:26 Close for impending dawn, T=11 degC, D=1 degC, Wind light from West. Sky is clear.

After observing, I changed WFS_rmsLimit was 600.0 now 800.0 in AIP_L.cfg for AO5. We've verified in 0.45 arcsec seeing that after you've sent a correction of 800 nm, the image is good (below 0.5 arcsec). I even observed a 0.5 arcsec image after a 2700 nm correction, but let's not get carried away.

On the afternoon of 17-SEP-2010 UT, I also added the +20 Tilt to M2 in a permanent way in the global offsets. The previous global offsets (also M2 total collimation) were: X=-7.010 Y=-2.600 Z=-1.000 RX=148.650 RY=181.800, and the new values are X=-8.031 Y=-2.600 Z=-1.000 RX=148.650 RY=-14.599. This is pushing us far out on the X-range of M2, so a swing arm adjustment may be needed if the next round of focal plane tilt measurements confirm this position. Note also that the Y-value of -2.6 was influenced by a parallel offset to M1 and M2 to get collimation range in warm weather.
See the revised values in Com20100921.

-- JohnHill - 15 Sep 2010