Telescope and AGw4 Commissioning at LDG - 05-06 May 2009 MST

Observers:: J. Hill, J. Storm (mtn), A. Rakich (Tucson) , D, Miller (Firenze)
Telescope Operator: W. Wack
Software Support: N. Cushing (mtn) Telescope Support: J. Rousseau, J. Morris

Summary

We measured the probe / camera angles on-sky, but they came out 180 off from the expected value.

We got Active Optics working on the y-axis by adjusting the the offsets to the rotator angle. However, there are still two more x-flips needed in GCS to get the corrections right for non-zero probe positions.

We confirmed the hotspot position. And we adjusted the probe zeropoints to move the hotspot onto the rotator center.

Lost half an hour to LDG rotator complaining about proximity sensor.

Lost 45 minutes to GCS having problems with selectAGW. These may not be random as some instances seem to be stimulated by us moving the rotator manually. Does a cancelled preset (because of rotator off-source) do something to GCS? Note the similar failure at 02:00 AM which was not stimulated by the rotator.

Tentative Plan for Next Few Days

  • Wednesday - test patched GCS (for pupil position), and new version IDL, measure field aberrations at DG with AGW4 (This is Jesper's last night on the mountain)

  • Thursday - test TCS Build 26 at Bent Gregorian with IRTC (needs M3 swing arm moved in)

  • Friday - Move IRTC to Direct Gregorian and take data for AGw Transformations

Preparations

From last night: We don't yet understand completely why LEFTZEROPOINT wants to be +5.5 and not something near -180 degrees.

I (DJT) expect roughly this value if you have oriented the on-axis acquisition image to have N-up. Recall that LEFTZEROPOINT is to orient the on-axis science instrument on sky, not the AGW. The +5.5 offset instead of -7.3 is likely from a half-24-bolt hole pattern offset and the fact that the sky is flipped with respect to LFBG. You should set LEFTZEROPOINT to +185.5 (or -174.5) and then verify in ACQUIRE mode that the AGw camera is oriented S-up (an offset to the north should move the star on the acquisition image up when at PA=0).

Details

Open to sky covered with thin cirrus.

19:34 On lbtmu107 moved 4GB /var/log/debug to debug.1, and then gzipped it. This gives us a managable size file to work with, but I may havscrewed it up since GCS didn't pick up the new descriptor.

19:44 Stop and restart GCS to see if it picks up new debug, but it didn't.

Field and camera orientation

19:52 pointto BS9145 Tel="ACQUIRE" gs=8 PAM+ PA=60 - Adjust IE=-50 CA=+14 (but this was off-axis)

20:01 offset 0 5 (RADEC) moves star lower left, but this experiment is too hard at PA=60. Go back to PA=0

20:02 pointto BS9145 Tel="ACQUIRE" gs=0 PAM+ PA=0 - Adjust IE=-60 CA=12

20:05 offset 0 5 (RADEC) - star moves down.

20:08 LEFTZEROPOINT changed from +5.5 to +185.5, stop and restart PCS.

20:08 pointto BS9145 Tel="ACQUIRE" gs=0 PAM+ PA=0 - Adjust IE=-65 CA=+12 (probably the M2 offload got us)

20:11 offset 0 10 (RADEC) - star moves up on acquire image as DJT requested, so this is apparently PA=0.

20:13 pointto BS9145 Tel="ACQUIRE" gs=0 PAM+ PA=INDEF

20:14 pointto BS9145 Tel="ACQUIRE" gs=8 PAM+ PA=INDEF=344.29 - no star in acquire image (not counting two very faint ones)

20:15 wayne turns rotator by 180 deg - and we see the star. so the world (or perhaps the camera) is upside down compared to Dave's prediction.

20:18 LEFTZEROPOINT changed from +5.5 to +185.5, stop and restart PCS. Clear M2 offload.

20:21 pointto BS9145 Tel="ACQUIRE" gs=8 PAM+ PA=INDEF=344.29 - Adjust IE=-64 CA=+14

20:22 offset 0 10 (RADEC) - moves star down (a little right of straight down as we are at PA=344).

offset 0 0. getxy -u 1 give x=0.111 y=515.691

20:24 setxy -x 0 -y 520.69 -u 1 - moves star up. This is moving probe closer to the center of the field, or South. This means north is up on the acquire image.

20:32 pointto BS9145 Tel="ACQUIRE" gs=8 PAM+ PA=0 x=26.33 y=519.34

20:33 offset 0 10 (RADEC) moves the star straight down. It moved only the telescope and not the probe.

20:36 pointto BS9145 Tel="TRACK" gs=0 PAM+ PA=0

Images in /Repository/AGW_Data/090506/

20:39 setxy -x 0 -y 612.5 -u 1 guider 0001

20:40 offset 0 10 irc OffsetPointing 0. 0. 10. RADEC MOUNT true ABS left guider 0002 - star moves down on ds9

20:42 offset 5 10 irc OffsetPointing 0. 5. 10. RADEC MOUNT true ABS left guider 0003 - star moves right on ds9

20:43 offset 0 0 guider 0004

20:39 setxy -x 0 -y 610 -u 1 guider 0005

20:39 setxy -x 2 -y 610 -u 1 guider 0006

20:46 pointto BS9145 Tel="TRACK" gs=8 PAM+ PA=0 then setxy -x -26.324 -y 518.914 -u 1 (coords from pointto) guider 0007 - no star

20:48 setxy -x -26.324 -y 518.914 -u 1 (pointto didn't have the x-flip) - star seen in guider0008

Active Optics calibration

20:50 pointto BS9145 Tel="GUIDE" gs=8 PAM+ PA=0 Zglobal=0.1 x=26.32 y=519.34

PSF Gains for Z5-Z10 are still zero.

Confirmed that the patched GCS does the proper correction of the guider image for probe rotation of agwNum 1.

20:57 pointto BS9145 Tel="ACTIVE" gs=0 PAM+ PA=0

But it isn't taking WFS image. Waiting for AzCam - no hint of errors on GCSGUI.

21:04 Stop and restart GCSL.

21:05 pointto BS9145 Tel="ACTIVE" gs=0 PAM+ PA=0 and clear M2 offloads.

                                                                 WFS: z(04) = -225.0823
May  5 21:07:05 lbtmu107 LBT_left_GCS: WFS: z(05) = 502.0019
May  5 21:07:05 lbtmu107 LBT_left_GCS: WFS: z(06) = -828.0373
May  5 21:07:05 lbtmu107 LBT_left_GCS: WFS: z(07) = -986.2821
May  5 21:07:05 lbtmu107 LBT_left_GCS: WFS: z(08) = -58.5264
May  5 21:07:05 lbtmu107 LBT_left_GCS: WFS: z(09) = -127.3665
May  5 21:07:05 lbtmu107 LBT_left_GCS: WFS: z(10) = 86.4234
May  5 21:07:05 lbtmu107 LBT_left_GCS: WFS: z(11) = 238.7671

21:08 Send manually +1500 Z7 - was accidentally correct and improved the pupil.21:

21:10 Send another +1500 Z7

21:12 Send another +1500 Z7, but we hit the +Y limit on M1

21:15 Remove active corrections from PSF

21:16 pointto BS9145 Tel="ACTIVE" gs=0 PAM+ PA=0

21:19 Sent +1500 Z7, but then GCS compensates, before we get the gain back to zero

21:20 Sent -1500 Z7

21:22 Sent +1500 Z7

21:24 Sent +1500 Z7

21:26 Set PSF gain for Z7 to -1 that seems tp work

21:29 Set PSF gain for Z7 to zero.

21:30 Send +1500 Z8 manually - sign Z8 is also reversed

21:33 pointto BS9145 Tel="ACTIVE" gs=0 PAM+ PA=90

Flip WFS_ZrotatorDirection from -1.0 to 1.0.

21:37 pointto BS9145 Tel="ACTIVE" gs=0 PAM+ PA=90

Change WFS_rotAngleOffs from 9.0 to 189.0

21:41 pointto BS9145 Tel="ACTIVE" gs=0 PAM+ PA=90

May  5 21:47:52 lbtmu107 LBT_left_GCS: [D] error: Mem.setOwnedValue('gcs.side[0].WFSCam.imageSaturated', var) returned: Data size is greater than space allocated. (-13)

21:48 Send +1500 Z7 manually

Guiding stopped. Tracking glitch. Didn't Andrew say only a few minutes ago "Things are going really well."?

21:50 Set PSF gains to 1 on Z7 and Z8.

21:51 pointto BS9145 Tel="ACTIVE" gs=0 PAM+ PA=0 Seems to be waiting an excessive time .

21:54 Stop and restart OSS. Preset timed out.

Adjust pointing to get away from +Y limit

21:57 pointto ACQUIRE - hangs because the DG rotator won't track. It complains about hitting the prox limit in 30 sec, but the reported position is 185 deg. Called Tom Sargent.

22:17 Now rotator is happy, but we don't understand what we did to fix it if anything. During this event we also saw the bug where doing a rotator operation delays PCS by 6 sec. Lost 30 minutes.

Tom's analysis from studying the logs the next day was that the rotator really was near the end of travel at -90, even though the display was showing the rotator angle as +185. If this happens again, it would be useful to take a photo of the rotator.

22:18 pointto BS9145 Tel="ACQUIRE" gs=0 PAM- PA=0 - Adjust IE=-57 CA=17

22:25 pointto BS9145 Tel="ACTIVE" gs=0 PAM+ PA=0 - now testing with Z7 and Z8 gain of 1.

Sky is now beautifully clear. Seeing is varying 1-1.5 arcsec, but we aren't fully collimated.

22:30 Turn on Z5 gain to 1.0 - it converges.

22:32 Turn on Z6 gain to 1.0 - seeing degrades to 1.7 arcsec.

22:34 Send +500 Z9 and leave gain=1

22:39 Adjust WFS_rotAngleOffs from 189 to 180 deg

22:40 pointto BS9145 Tel="ACTIVE" gs=0 PAM+ PA=0

22:42 Send +500 Z9

22:44 Send +500 Z10 with Z9 and Z10 gains both 1. Last wfs reported 99 nm, but don't believe it too much.

22:49 pointto BS9152 Tel="ACTIVE" gs=1 PAM+ PA=INDEF=229.72 x=0.00 y=500.9

22:55 Send +500 Z9 - converges OK

22:58 pointto BS9152 Tel="ACTIVE" gs=1 PAM+ PA=259.72 x=55.72 y=516.06

22:59 Send +500 Z9 - diverged

23:02 Changed WFS_probeRotationFactor from -1.0 to +1.0

23:02 pointto BS9152 Tel="ACTIVE" gs=1 PAM+ PA=259.72

It converged here, but the images are remarkably triangular. Z7 seems to be converging quite slowly.

WFS_ZrotatorDirection   1.0             float   # multiplication factor for rotator angle (to change direction use -1.0)
WFS_probeRotationFactor 1.0             float   # used to change how the probe rotation is added to the rotator angle (use -1.0 to subtract it from rotator)

23:12 pointto BS9152 Tel="ACTIVE" gs=0 PAM+ PA=0 - back on axis -seeing improves to 0.8 to 1.2 arcsec

23:18 pointto BS9152 Tel="ACTIVE" gs=0 PAM+ PA=45

Data from GCS in /Repository/GCSfiles/dailyimages (later to be moved into 20090506)

23:19 Send +300 Z9 and set gain to 0. and Z10 gain to 0 This starts wfsc 0128

23:22 pointto BS9152 Tel="ACTIVE" gs=0 PAM+ PA=120

May 5 23:25:26 lbtmu107 LBT_left_GCS: WFS: error: triggerWFS: WFS image didn't arrive within timeout of 37716 ms.

23:28 Stop and restart GCSL

23:29 pointto BS9152 Tel="ACTIVE" gs=0 PAM+ PA=120 This starts wfsc 0133

23:31 pointto BS9152 Tel="ACTIVE" gs=0 PAM+ PA=65 This starts wfsc 0135

23:36 pointto BS9152 Tel="ACTIVE" gs=19 PAM+ PA=INDEF=68.98 - couldn;t move probe that far.

23:38 pointto BS9152 Tel="ACTIVE" gs=18 PAM+ PA=INDEF=65.44 This starts wfsc 0141

23:40 pointto BS9152 Tel="ACTIVE" gs=15 PAM+ PA=100 x=64.28 y=514.45 This start wfsc 0143

23:43 pointto BS9152 Tel="ACTIVE" gs=15 PAM+ PA=25 x=-78.09 y=525.12 This starts wfsc 0146 (I think).

Editing Doug's pupil parameters into AIP_IRTC_DG_L.cfg
WFScam_hotspot_x	390.2		double	# ccd x coordinate of pixel that represents leftmost x-center of WFS spots
WFScam_hotspot_y	251.1		double	# ccd y coordinate of pixel that represents leftmost y-center of WFS spots
WFS_lenslet_edge_x	381.65		double	# defines the edge of the virtual lenslet grid in x
WFS_lenslet_edge_y	240.45		double	# defines the edge of the virtual lenslet grid in y

23:49 pointto BS9152 Tel="ACTIVE" gs=0 PAM+ PA=0 This includes wfsc 0156 but I lost track of where it started.

23:55 pointto BS9152 Tel="ACTIVE" gs=15 PAM+ PA=120 This includes wfsc 0162 - 168

FILE /Repository/GCSfiles/dailyimages/wfscimage000172.fits
GUIDE_X -0.03
GUIDE_Y -0.10
GUIDE_Z 30.99
DELTA_Z 0.01
AGWFILT 3
PUPIL_X 390.20000
PUPIL_Y 251.10000
ZERN2PSF 1
OFFAXIS 1
ROTANGLE 340.15   <----- The actual rotator angle at LDG
ZERNROT -160.16    <----- rotAngleOffs - ROTANGLE (a.k.a. Doug's angle)
ROTDELTA 0.00       <----- probe rotation correction
FLDANGLE -90.00

May  6 00:04:17 lbtmu107 LBT_left_GCS: WFS: probe x/y sfp: -0.030716, -0.096332
May  6 00:04:17 lbtmu107 LBT_left_GCS: WFS: radius = 0.168518   Doug's Angle: -4.354904 rad
May  6 00:04:17 lbtmu107 LBT_left_GCS: WFS: X = 0.157864   Y = -0.058968   phi = 339.517609 deg
Why is Doug's angle in the syslog much different than ZERNROT? The ZERNROT value in the .info file seems to be correct.

00:22 pointto BS9152 Tel="ACTIVE" gs=15 PAM+ PA=120 x=94.00 y=542.28 - hung waiting for WFS image

00:25 pointto BS9152 Tel="ACTIVE" gs=15 PAM+ PA=120 x=94.00 y=542.28 - still hung

May 6 00:28:30 lbtmu107 LBT_left_GCS: WFS: error: triggerWFS: WFS image didn't arrive within timeout of 37716 ms.

00:29 Stop and restart GCS

00:29 pointto BS9152 Tel="ACTIVE" gs=15 PAM+ PA=120 x=94.00 y=542.28 wfsc 0207 - 0211

00:34 pointto BS9152 Tel="ACTIVE" gs=15 PAM+ PA=1 x=112.36 y=578.67 wfsc 0212 - 214

Recall that Z9/Z10 gains are still turned off, and note that we see Z9/Z10 changing as we move farther off the y-axis.

00:37 pointto BS9152 Tel="ACTIVE" gs=15 PAM+ PA=160 - couldn't move probe that far

00:38 pointto BS9152 Tel="ACTIVE" gs=15 PAM+ PA=0 x=-107.73 y=566.31 wfsc 0215 - 216

00:41 pointto BS9152 Tel="ACTIVE" gs=15 PAM+ PA=0 wfsc 0217

Andrew notices that the pupil shift correction versus probe X is going the wrong way in IDL. Probably this is also happens in GCS because of the X-flip.

00:41 pointto BS9152 Tel="ACTIVE" gs=0 PAM+ PA=0 wfsc 0220

00:52 Stop WFS so Andrew can try it in IDL. Put Z9, Z10 gains back to 1.

01:06 Start WFS again while Doug works on an IDL bug. wfsc 0232 - We are getting guider images as good as 0.5 arcsec, but the seeing is fluctuating quite a bit.

01:13 Stop WFS so Andrew can try it. Andrew sending Z9= +500 and gain = 0.

01:16 Start WFS ---through wfsc 0259

01:22 Flip the pupil shift value in X
WFS_pupil_shift_x	-0.20		float	# pupil shift factor in x
WFS_pupil_shift_y	0.19		float	# pupil shift factor in y

01:23 pointto BS9152 Tel="ACTIVE" gs=15 PAM+ PA=0 wfsc 0260 - 265
Andrew turned the Z9 gain back on ~01:25
That clearly diverged so the change in pupil shift is more complicated than we expected from the X-flip in oacserver, but see below. We think pupil_shift_x is not used anywhere.

01:29 Flip the pupil shift value in X +0.20

Hotspot Verification and Rotator Axis Check

01:31 pointto BS9152 Tel="ACTIVE" gs=0 PAM+ PA=0 wfsc up to 280

01:38 Block guiding in GCS

x=1.0 y=612.5 wfsc 0281 (star moved much farther on guider than expected)

x=-1.0 y=612.5 wfsc 0283

x=-2.0 y=612.5 wfsc 0285 - 0286

01:45 x=2.2 y=612.5 wfsc 0288

x=2.5 y=612.5 wfsc 0289

01:48 x=0 y=613.5 wfsc 0292 - 0293

x=0 y=611.5 wfsc 0295 -0296

Conclusion: Hotspot is in the advertised location within 0.5 arcsec. Seeing got worse, so the measurement wasn't as precise as we hoped. See also wfscimage000412 taken in twilight.

01:54 pointto BS9152 Tel="ACQUIRE" gs=0 PAM+ PA=0

01:56 pointto BS9152 Tel="TRACK" gs=0 PAM+ PA=90 then move probe in

02:00 rotating 0-360 with readGuideCam -e 70000 AGW selection failed

These messages from the second attempt. We don't understand what made GCS hiccup and decide that it was switching to LUCIFER. Should add note to Issue 2164.
May  6 02:01:37 lbtmu107 LBT_left_GCS: AGW::getFilter returned an error: 202
May  6 02:01:37 lbtmu107 LBT_left_GCS: [E] AIP AGW unit error: AGW1: Network Error
May  6 02:01:50 lbtmu107 LBT_left_GCS: [E] AIP AGW unit error: AGW1: RPC Communication Error
May  6 02:02:02 lbtmu107 LBT_left_GCS: [E] AIP AGW unit error: AGW1: RPC Communication Error
May  6 02:02:02 lbtmu107 LBT_left_GCS: [E] AIP AGW unit error: AGW1: Network Error
May  6 02:02:02 lbtmu107 LBT_left_GCS: AGWunit init(): _filterPosition or _focus read back from AGW server is negative indicating an error.
May  6 02:02:02 lbtmu107 LBT_left_GCS: init() returned an error, indicating the AGW unit is not accessible.
May  6 02:02:02 lbtmu107 LBT_left_GCS: GCS Subsystem COULD NOT change active AGW configuration to AIP_L.cfg
May  6 02:02:02 lbtmu107 LBT_left_GCS: GCSL:setAGW execution time: 26670 ms
May  6 02:02:02 lbtmu107 LBT_left_GCS: wait done for handle block:2:1 0 94
May  6 02:02:16 lbtmu107 LBT_left_GCS: [E] AIP AGW unit error: AGW1: RPC Communication Error

02:03 selectAGW junk

selectAGW AIP_IRTC_DG_L.cfg - worked OK

02:04 rotating 360-0 with readGuideCam -e 70000 guider 0012 but focus was off. It got parked by TRACK mode.

02:08 rotating 360-0 with readGuideCam -e 70000 guider 0013

02:09 pointto BS9152 Tel="ACTIVE" gs=0 PAM+ PA=0

02:22 stopAGW -u 1, stop and restart oacserver to get new probe zeropoints. startAGW -u 1, home We forgot to stop guiding before doing this, so GCSL is complaining. Stop and restart GCS.

02:28 TRACK, setfocus, setxy AGW selection failed

02:32 readGuideCam -e 70000 guider 0015 - probably we went the wrong way

02:34 pointto BS9152 Tel="ACTIVE" gs=0 PAM+ PA=0

stopAGW -u 1, stop and restart oacserver to get new probe zeropoints. startAGW -u 1, home

02:40 TRACK, setfocus, setxy , selectAGW manually before taking the image.

02:43 readGuideCam -e 70000 - selectAGW still failed.

02:45 readGuideCam -e 70000 guider 0017

02:54 pointto BS9152 Tel="TRACK" gs=0 PAM+ PA=0, setfocus, setxy

readGuideCam -e 1000 works OK, but then readGuideCam -e 70000 fails selectAGW

02:59 readGuideCam -e 70000 guider 0019 - poor seeing or maybe the focus moved from an M2 offload

03:00 readGuideCam -e 70000 guider 0020

Probe is now adjusted to have hotspot on the rotator center with AGW4

Back to Active Optics

03:07 pointto StoneMagw Tel="ACTIVE" gs=25 PAM+ PA=0 - but it says couldn't move probe.

03:11pointto StoneMagw Tel="ACTIVE" gs=0 PAM+ PA=0 - seeing varying between 0.8 and 2 arcsec FWHM.

Gains in PSF are 1.

03:18 Z5=+1000 wfsc 0327 - 329

03:22 Z6=+1000 wfsc 0331 - 0334

03:24 Z9=+300 wfsc 0335 - 0337

03:27 Z10=+300 wfsc 0338 - 0340

See elongation in wfsc 0338. Image is jumping around alot and seeing is varying alot.

03:37 pointto StoneMagw Tel="ACTIVE" gs=44 PAM+ PA=INDEF=352.45 x=0 y=457.26

Up until now, IDL was still reading the LFBG rotator angle in cases where IDL was taking images.q

just collimation through wfsc 0363

03:45 Z5=+1000 wfsc 0364 - 0366

03:48 Z6=+1000 wfsc 0367 - 0372

03:52 Z9=+300 wfsc 0373 - 0375

03:54 Z10=+300 wfsc 0376 - 0398

Trying to get a better image for the rotator center measurement

04:10 pointto StoneMagw Tel="TRACK" gs=0 PAM+ PA=0, setfocus, setxy, selectAGW AIP_IRTC_DG_L.cfg

readGuideCam -e 1000 guider 0021

04:13 readGuideCam -e 70000 guider 0022 - selectAGW failed How does it know? Is it because moving the rotator cancelled the previous preset? Yes, the first time the rotator moved off-source, IIF cancels the preset and sends a stopGuiding command to GCS. This interupts the process of taking the guider image. After the preset is already cancelled, it works fine.

selectAGW AIP_IRTC_DG_L.cfg

04:15 readGuideCam -e 70000 guider 0022 - better seeing/focus than above, but not an especially round ring

04:17 readGuideCam -e 90000 guider 0023 -

04:19 readGuideCam -e 90000 guider 0024

04:25 readGuideCam -e 85000 guider 0025 - this is the best of the batch. Isn't it remarkable what the atmosphere can do.

We can't see anywhere WFS_pupil_shift_x (or y) is used in the code, so try to repeat the experiment that caused divergence.

04:38 WFS_pupil_shift_x -0.20

04:40 pointto StoneMagw Tel="ACTIVE" gs=0 PAM+ PA=0 - did not diverge on-axis thru wfsc 0411

04:49 pointto StoneMagw Tel="ACTIVE" gs=68 PAM+ PA=0 x=-102 y=504.57 - It doesn't seem be be diverging here. Maybe the earlier divergence was a fluke.

04:54 stop WFS in GCS.

Doug and Andrew burned into twilight making IDL test.

05:07 Close dome T=5degC D=-3.8 degC

What about the geometrical field aberrations?

-- JohnHill - 05 May 2009
Topic revision: r7 - 07 May 2009, JohnHill
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