MODS2 Commissioning Night 5 - 20140504 UT

Observers: R. Pogge (OSU), O. Kuhn (LBTO), J. Hill, D. Miller (LBTO, Remote from Tucson)
Telescope Operator: S. Allanson (LBTO)
AO Support: D. Miller (LBTO)
Telescope Support: John Little et al. (LBTO)
Software Support: T. Golota (Remote from Tucson)
Instrument Support: J. Power (LBTO)

Summary

Improvements to Active Optics

We've made the following improvements to MODS2 active optics tonight:

1. We are using new gains for all the Zernike terms based on last night's measurements.
2. Doug installed a new 9x9 reconstructor with the corners clipped off into GCS
3. We adjusted WFS_fluxSigmaFactor to get better rejection of stray faint spots.

The above items have recovered the robustness of active optics convergence that we were missing at the beginning of tonight.

There are three remaining issues that make the MODS2 active optics less than perfect:

1. There is pupil wobble as a function of rotator angle that moves the pupil around in ways that GCS is not yet able to correct. We believe the cause is a misalignment of the axis of the wfs (but don't have hard proof). This adds changing aberrations to the converged wavefront at different rotator angles. There is not any available optomechanical adjustment to correct the problem. We can imagine a GCS patch which allows us to calibrate the wobble.
2. There is a problem with sextractor measuring incorrect spot positions which may vary from frame to frame. The result is that the star image may appear triangular when the seeing is 0.7 arcsec or better.
3. The wfs pupil position wrt/ the reconstructor is quantized in integer lenslets. This makes some changing aberrations at different guide star positions. This effect is worse by ~20% for the MODS2 wfs with the 20% smaller pupil. A long term improvement is to make a set of fractional lenslet reconstructions that are optimized for different pupil-lenslet shifts. Doug and Lee have already been working on this.

Tasks for Tomorrow Night:

1. Measure the pupil wobble
2. Check the transformation
3. Field aberration data

MODS2 Dice, Photometric and Astrometic Calibration Data obtained.

• Photometric night - no clouds seen from the balcony at sunset or sunrise and no sign of clouds on the all-sky.

Details

Trying to Stabilize Active Optics

02:40 Active Optics is consistently running away in coma and spherical.

02:45 Set Active Optics Gains in GCS MODS_R_RM_sub11.cfg to some nominal lower values consistent with last night's data - pending a more detailed analysis by Doug.

sextScale_Z4    -0.8    double
sextScale_Z5    -0.8    double
sextScale_Z6    -0.8    double
sextScale_Z7     0.8    double
sextScale_Z8     0.8    double
#sextScale_Z9   -0.8    double
#sextScale_Z10  -0.8    double
#sextScale_Z11  -0.8    double
sextScale_Z12   -0.7    double
sextScale_Z13   -0.7    double
sextScale_Z14   -0.7    double
sextScale_Z15   -0.7    double
sextScale_Z16   -0.7    double
sextScale_Z17   -0.7    double
sextScale_Z18   -0.7    double
sextScale_Z19   -0.7    double
sextScale_Z20   -0.7    double
sextScale_Z21   -0.7    double
#sextScale_Z22  -0.7    double
sextScale_Z22   -0.35   double
sextScale_Z9    -0.4    double
sextScale_Z10   -0.4    double
sextScale_Z11   -0.4    double

02:54 Clear Active Optics on M1 and M2 - now Z9, Z10, Z11, Z22 have gains set to zero in PSF. Now Z7 coma is running away by itself.

03:00 Clear Active Optics on M2 hexapod only

03:05 Clear Active Optics and move to a new target.

03:07 reflatten shell, and recover from wierd failure

The initial image after clearing everything has several arcsec of coma, in start contrast to last night where the initial image was 1 arcsec and round without any active correction. Telescope structure and mirrors seem in decent equilibrium albeit at +8.5 degC.

03:27 ACQUIRE preset for manual adjustment of coma and pointing.

03:36 Now active optics seems stable after Steve manually dialed out the initial coma.

Dice5 tests for repeatability of offsets

03:45 Going to start dice5 test.

• Script locations:
  • Persson scripts are in /home/lbto/MODS2Comm/Dice/PerssonScripts/
  • dice scripts are in ~modseng/modsScripts/Support/Dice/
• Preset to BS9145. elevation 57 deg
  • At PA=0, red IMCS did not lock. We tried PA=270, but it did not lock here either. At PA=180, it locked.
  • 03:45 Executing dice10.img

03:56 Executing dice10.img script.

• The first one was executed by hand (execMODS2 dice10.img)
• After that, the series was run using the runDice5 script.

04:30 Executing dice1.img script

05:10 AO475 star chosen for high elevation field. max elevation 78 deg

• IMCS lock could not be achieved for PA=0.
• Created script for PA=180. IMCS converged.
• Start runDice5
  • note that field is 3Kx3K for these.

05:13 Set Active Optics Gains for Z9,Z10,Z11,Z22 to 0.1 in PSFR

05:43 Turn on -0.35 gain for Z16 on in MODS_R_RM_sub11.cfg, and high factor 0.3 in MODS_R.cfg, but not other terms in range Z12-Z21.

06:16 Change WFS_fluxSigmaFactor from -2.5 to -1.5 in MODS_R.cfg (also other MODS and LUCIs). This significantly improves the smoothness of the wavefront by eliminating faint stray spots from the fits.

06:40 Why is IDL picking the wrong lenslet center compared to GCSR? e.g. wfsc 0272

wfsc 0345 is a good image to test out sextractor parameters

06:55 El40 field.

• 07:08 data taking stopped - comm glitch?
• aborted and changed roi to 2Kx2K
  • 1 5-dice on this field with 3Kx3K
  • 4 5-dice 10" and 5 5-dice 1"
• 08:00 Done

MODS2 Photometric Calibration

08:10 Slew to M92 (=NGC6341) which has many stars with ugriz from Clem et al. (2007).

• NGC6341f1_PA180_ugrizPhot.img script in /home/lbto/MODS2Comm/SciPhot (repeated script 2x)
  • 2x g and r
  • 1x u and 2x i
  • 1x u and 1x z
  • blue images mods2b.20140504 NN=7-14
  • red images mods2r.20140504 NN=158-169

09 Slew to Stone_M field and obtain data field for science camera astrometry.

• StoneM2_PA180.img in /home/lbto/MODS2Comm/SciAstrom/
  • This corresponds to TCS Stone catalog M5 (though center is slightly offset).
• blue images NN = 16-20
• red images NN=170-175

08:55 Set Active Optics Gains for Z9,Z10,Z11,Z22 to 0.2 in PSFR.

09:20 Slew to Stone_M7.catalog and take a few more guide camera images, for guider astrometry.

• Stone_M7 field is much less dense than the M92 one done last night, but here already.
• /home/lbto/MODS2Comm/GuiAstrom/Data/20140504/right_guider0000NN.fits
  • NN = 3-12 500 ms 10x Clear
  • NN = 13-22 500 ms 10x LP

09:29 Doug and John taking over. ACTIVE preset to StoneM7 on-axis

New Reconstructor Testing

Doug has installed in GCS the new 9x9 reconstructor with the corners clipped off == 20140504_mods2.

Robustness tests

09:34 Send +1000 Z8 to DX M2 hexapod

09:37 Send +3000 Z7 to DX M2 hexapod - The angle between the Z7/Z8 correction seems slightly wrong.

09:40 All of the Active Optics Gains in PSFR are back to 1.0.

09:43 Send +300 Z9 to DX M1

09:50 ACTIVE preset gs=1 - x=0 y=-8.7

09:52 ACTIVE preset gs=25 - x=0 y=?

Doug fixed the sextractor configuration files in the new reconstructor directory.

09:57 Set Z9,Z10,Z11 back to full gain in GCS (0.8), add Z17 at half gain (0.35). Also set the COW weights back to the same values we set for sext at the beginning of the night -- this was part of the initial stability problem seen at the start of tonight.

10:04 Send +3000 Z8 on DX M2 hexapod - The Z8 correction was good, but made some crosstalk to Z9.

10:10 ACTIVE preset gs=25 with PA offset by 20 deg. x=-28 y=-76

Trefoil is still bouncing around from one frame to the next. We see excess noise in Z9,Z10,Z11. Whatever happened on wfsc 0501 is likely the source of our remaining issues. The temporary fix is to reduce the Z9-Z11 gain by x3 or so to make things less sensitive to frame-frame noise.

Active Optics Calibration

Conditions and active optics are better than last night, so we'll take more calibration data for Zernike gains.

10:33 stopWFS

10:39 zernike_collect with sext_7_16.list

Sext
Zernike	Amplitude	Measured	Average	Factor	wfsc #
0	0	12			1
7	500.000	-690 -819 -661	-723	-0.69	2-4
0	0	118			5
0	0	167			5
8	500.000	-847 -789 -777	-804	-0.62	6-8
0	0	-203			9
0	0	-283			9
9	500.000	978 1024 913	972	0.51	10-12
0	0	373			13
0	0	-233			13
10	500.000	589 596 554	579	0.86	14-16
0	0	116			17
0	0	90			17
16	300.000	439 430 477	449	0.67	18-20
0	0	390			21
0	0	173			21-25
17	300.000	477 440 387	435	0.69	26-28
0	0	173			29

10:xx zernike_collect with sext_4_6.list

Sext
Zernike	Amplitude	Measured	Average	Factor	wfsc #
0	0	-178			30
4	500.000	601 610 724	645	0.77	31-33
0	0	263			34
0	0	113			34
5	500.000	785 688 715	729	0.69	35-37
0	0	27			38
0	0	187			38
6	500.000	364 398 279	347	1.44	39-41
0	0	-168			42
0	0	-106			42
11	300.000	458 124 211	264	1.13	43-45
0	0	-109			46
0	0	-245			46
22	300.000	3 154 -17	46	6.45	47-49
0	0	-337			50
Sext
Zernike	Amplitude	Measured	Average	Factor	wfsc #
0	0	25			51
0	0	-311			51
22	300.000	8 48 59	38	7.73	52-54
0	0	-283			55

11:xx zernike_collect with sext_12_22.list

Sext
Zernike	Amplitude	Measured	Average	Factor	wfsc #
0	0	-461			56
12	300.000	88 75 85	83	3.61	57-59
0	0	-385			60
0	0	62			60-61
13	300.000	482 489 474	481	0.62	62-64
0	0	78			65
0	0	-105			65
14	300.000	574 429 455	486	0.62	66-68
0	0	-100			69
0	0	-12			69-70
15	300.000	547 492 432	490	0.61	71-73
0	0	4			74
0	0	118			74
18	300.000	649 402 422	491	0.61	75-77
0	0	-6			78
0	0	-313			78
19	300.000	214 190 257	220	1.36	79-81
0	0	-231			82
0	0	-6			82
20	300.000	659 598 639	632	0.47	83-85
0	0	128			86
0	0	-231			86
21	300.000	612 566 536	572	0.52	87-89
0	0	-203			90
0	0	-190			90
22	150.000	-127 -117 87	-52	-2.86	91-93
0	0	-250			94

Test of High-Order Zernikes (Z12-Z21)

11:37 ACTIVE preset gs=14 which is brighter

11:40 Turn on the high order zernike gains. ACTIVE preset gs=14 - The high order corrections appear to be stable.

11:48 Offset by 10 arcsec RA (abs) to get a different pupil position.

11:48 Offset by 40 arcsec RA (abs) to get a different pupil position.

Changed rms low order (only) threshold from 700 to 500. Set Z9,Z10,Z11 gains to 0.4 (%50) in GCS.

Close

11:58 Close for impending dawn.

-- %USERSIG{JohnHill - 2014-05-04}%