E-time - 20161011 UT

Observers: JHill, DMiller, OKuhn (Tucson)
Telescope Operator: SAllanson (LBTO)
AO Support: DMiller
Telescope Support: MWagner (LBTO)
Software Support: TSargent (Tucson)
Instrument Support: ACavallaro (LBTO)
Engineering Support:

Summary (First Half)

Andrew got some LBC twilight flats although some cirrus moved in on him at the end.

We went down to horizon for MikeW to reseat fiber modules in the DX AdSec box. DX AdSec started working after that, although it is unclear if there was a causal connection. For the time being, we are giving Mike the credit.

We've updated all the guiding and active optics parameters for the modified MODS1 wfs. Thanks to Rick and Tom for nice looking focus and alignment on both cameras. MODS1 guide stars can now be anywhere in the patrol field without fear of vignetting (like MODS2).

We took data to measure LEFTZEROPOINT, but we ran out of time before getting a full set of MODS transform data.

We handed over for reconfiguration to LUCI at 06:37 UT. That would have be loads of time except that the AdSecs started having an IDL license problem (Steward license was timing out even though it worked OK earlier in the evening?).

With MODS2, the Blue channel IMCS would not lock. Tried multiple times (and, yes, unlike at the end of the night, I did remember to wake MODS! sigh... But, when trying again in the morning, I did wake MODS, but then after the lock on both sides failed, I checked the laser and it reported "disabled"... ongoing investigation)

With the change made to the agw.ini XCenter, YCenter, I think the SW limits on the probe may need to be adjusted.

If the left rotator zeropoint is really off by 1.4 deg, then we should try to obtain a set of transform data. But I am suspicious of this. John entered a new value into PCSInstrument.conf, we cycled PCS, and yet the next set of measurements showed a 1.7 deg offset (same sign as before --- as if we made it worse, but not by nearly as much as the expected amount). A mask alignment would confirm the rotational offset (it really should have been done to determine this offset - faster and more accurate), so if these masks could remain loaded, any beginning-of-night time might be useful for checking this. Stone_O is up then.

We also did not obtain a spectrophotometric standard or photometric imaging. Given the newly coated red collimators, it would be nice to obtain these data soon.

Preparations

Rick and Tom installed the modified MODS1 wfs at the end of last week.

Kellee checked out MODS1 and MODS2 cameras during the day.

Alessandro installed some MODS masks.

John and Olga modified the MODS_L.cfg file:

guidecam_ROI		"78 12 447 379"
guidecam_hotspot_x	235.0 (preliminary guess)
guidecam_hotspot_y	200.0 (preliminary guess)
guidecam_center_x	235.6 (not sure if this has any remaining use)
guidecam_center_y	195.6 (not sure if this has any remaining use)
WFScam_offset_x		0.0	
WFScam_offset_y		0.0	

and with Doug:

WFScam_spacing_x	13.12
WFScam_spacing_y	13.12
WFScam_hotspot_sext_x   228.5
WFScam_hotspot_sext_y   137.5

and later on sky:

WFScam_hotspot_x	        392.0
WFScam_hotspot_y	        248.0
guidecam_hotspot_x	254.8
guidecam_hotspot_y	189.5
WFS_r_to_pupil_shift	0.124
WFScam_pixelscale	        0.1866

Details

LBC Sky Flats

01:16 TRACK preset for LBC flats. Sky was clear to start, but cirrus moving through by the end of flats.

01:56 Flats complete

Reseat Fiber Converters for DX AdSec

01:57 Move to horizon to reseat fibers

02:23 reseating fibers at AO box completed (without successful resolution)

02:25 Move to zenith to reconfigure swing arms for MODS

02:39 Authorize as MODS/MODS (even thoough DX AdSec still having issues)

02:44 Restarting (or trying) both AdSecs.

MODS1 AGw wakeup and alignment

Initial Pointing and Hotspot location

02:56 ACQUIRE Preset to M5 star (M5_0835 = pi Peg) - succeeded after a couple tries.

SX IE=-21 CA=-51 DX IE=+10 CA=-208

03:05 ACTIVE Preset to AO1042 = BPM 96787 - Send zenikes disabled on left

03:08 Changing hotspot at GUI hung with WFS "paused"

wfs images not updating - why?

03:13 255,200

03:17 225,200 from gscclient

03:22 245,200 from GUI

03:25 stop/start GCSL and resend preset (hotspot is correctly reset)

03:31 set hotspot 235,240 from GUI, pauses to move probe and loses star (it paused to move probe correctly, but why did it lose the star?)

change maxHotspotfromProbe from 0.5 to 5.5 so it doesn't try to move the probe.

03:35 stop/start GCSL and resend preset starting is 235,200

03:38 235,220 (seems to have moved left on wfs)

03:40 235,180 (seems to have moved right on wfs)

03:41 215,200 (seems to have moved down on wfs)

03:43 255,200 (close) so start sending zernikes - now Doug is analyzing the well collimated images with IDL. Seeing is 0.5 arcsec.

pupil on-axis position 392,248 applied to WFScam_hotspot

03:48 255,190

03:50 254,189

03:53 254.5, 189.5

03:56 255.0,189.5

03:57 ACTIVE preset to BS for longer exposures on guider and wfs.

04:10 254.8,189.5 ADOPTED VALUES FOR HOTSPOT (and returned maxhotspot value to 0.5)

Stage Centering to Rotation Axis

/lbt/data/share/tcs/lbcobs.20161011/

stop guiding

move rotator

ptincrement 0 CA 5

gcsclient readGuideCam left -e 1000 (fails as expected)

gcsclient readGuideCam left -e 1000 ---> guide 0001

gcsclient readGuideCam left -e 90000

move rotator

04:17 guide 0002 centered at 243,186.8 (image space, no binning)

Olga's calc -2.42 mm X and -0.593 mm Y (center of circle needs to move left and down)

Reduced xcenter from 92 to 89.58 mm in agw.ini

Restarting GCSL and agwservice (not in a well coordinated way)

04:36 guide 0003 - correct direction even though image is saturated

change ycenter value more negative by 0.593

stop/start agwservice

04:41 guide 0004 centered at 203.75,193.5 (so Y went in the wrong direction)

stop/start agwservice but we can't go beyond the previous value of ycenter = 204.

adopted 89.58,204.0 for xcenter, ycenter

stop/start agwservice

Olga's Summary

We measured the coordinates of the hotspot on the MODS1 guide camera to be:

Xhs, Yhs = 254.8,189.5 (these are the 'physical coordinates' which do not
depend on the ROI).

 With the original agw.ini values (XCenter, YCenter, Focus0) = (92, 204,
38.58) and when the guide probe was on-axis, the center of rotation on the
guide camera was measured to be 292.8, 198.8.  To make the guide probe
on-axis position match the hotspot required offsets to (XCenter, YCenter)
of (-2.42mm, +0.593mm), assuming 0.0637 mm/pixel.

 For (XCenter, YCenter, Focus0) = (89.58,204,38.5), and with the guide
probe on-axis, the center of rotation is close to the hotspot. We could not
set YCenter > 204. When we tried,  the agwserver complained. From my notes
from March 2014, I think 204 is not a physical limit, but is very close,
and the HW would not allow the 204.593 mm needed to reach the hotspot Y
position.

My notes from March 2014 say that the range of the AGw stage is
 X: 0...185.684 mm
 Y: 0... 204.342 mm

I remember that, at that time, we talked about the limits - both in the
software and hardware. After this change, I think the SW limits, especially
in X, may need adjustment.

What are the current limits in the agw service? Should I move the guide
probe (gprobe -90 0, gprobe 90 0, ...) while recording stage coordinates,
AGWXS, AGWYS and AGWFS to explore the range?

Rick's Explanation the Following Day

For gprobe coordinates -2.42mm, +0.593mm the corresponding XCenter,
YCenter, and FOCUS0 coordinates are as follows:

  XCENTER 87.160
  YCENTER 204.590
  FOCUS0 38.502

The soft limit defined in the agwutil.h header file is 204, which was
somewhat conservative but consistent with MODS1 stage operational range
before the upgrade.

My action will be to bump up the raw stage coordinate limit to 205 and
recompile the AGw server.  I can have this done within the hour.

Note, that to get the XCENTER, YCENTER, and FOCUS0 values, you do this:

1) Set the guide probe the desired XY field position using gprobe

    isisCmd 'm1.ie gprobe -2.42 0.593'

   mind the single-quotes around the m1.ie ... 0.593

2) Query the AGW stage to get the raw actuator positions corresponding to
those values:

    isisCmd m1.ie agw

    then pick off  AGWXS, AGWYS, and AGWFS, which report the raw stage
actuator position in millimeters.

Is that what was done to get the XCENTER et al. values?

I set the agw.ini file parameters as follows based on the original email:

   XCENTER 89.58
   YCENTER 204.593
   FOCUS0 38.5

and restarted the AGW server.  No complaints, and sending the command
"GPROBE 0 0" puts the guide probe at these actuator coordinates.

You may wish to revisit them, but this gets around the 204 limit.

MODS1 Active Optics Pupil Position, Rotation, Reconstruction

WFS_r_to_pupil_shift 0.124 (copied from MODS2)

04:56 restart GCSL

MODS1 Off-Axis Guiding and Pupil Position

04:58 ACTIVE preset both sides to off-axis star (x=0 y=-109)

05:05 offset +30 deg abs (-54.5,-94.5)

05:09 offset -30 deg abs (+54.5, -94.5)

05:14 preset fails for Y=-189 mm

05:14 preset succeeds for y=-165.9

05:16 Auto exposure correction is broken for both guide and wfs in this version of GCS (5.4c). I added a note to Issue 6184 as the two problems are tightly coupled in software.

05:20 offset +30 deg abs (x=-82.5 y=143 - MODS reports the opposite x sign)

05:24 offset -30 deg abs (x=+82.5 y=143)

05:25 left primary mirror panics because of the moment problem

restart GCSL

MODS Rotator Zeropoint Check

05:32 ACTIVE preset to Stone field - failed because Olga's MODS presets were asynchronous

05:37 pseudomono on left just to move to the new field.

05:42 ACQUIRE mode to adjust pointing SX IE=-45 CA=-52 DX IE=+10 CA=-202 at EL=47 AZ=225

05:47 pseudomono on left just to move to the new field.

Olga is running TRACK presets from MODS and offsetting along the slit, so these notes are fairly sparse.

Ran a script to take a slit image and then dither up/down along the slit dY={0,+30,-30,+15,-15}. MODS1 images 10,11,12,13,14. Determined offset -1.4 deg between line of star centroids (1.89 deg CW from Y axis) and slit (0.446 CW from Y axis).

06:20 stop and restart PCS to absorb the new LEFTZEROPOINT for MODS1 of +26.230 deg (previously was +27.630)

06:22 pseudomono on left just to move to the new field. Took another set of data here. We think the previous move was with the wrong sign, so LEFTZEROPOINT should be +29.630 (we have not installed that value).

Now star centroids -1.7 deg from slit. Line of star centroids (2.23 deg CW) and slit (0.48 CW from Y axis).

I think we went the wrong way, but the magnitude of the change was not 2.8 deg, only 0.3 deg (It is like the PCS did not get this newly updated value). Still, if the line of star centroids should be rotated CCW to align with the slit, the angle should be added to the initial zeropoint, as written in the MODS2 AGw Commissioning Report, which also, with prescience, states that this will be forgotten. So it was.

In retrospect, should have gone straight for the more definitive MOS alignment to determine rotator zeropoint offset.

MODS Transform Data Check

We ran out of time, so didn't take any transformation data. Olga still needs to analyze the two sets of LEFTZEROPOINT data in the daytime.

Reconfigure for LUCI/LUCI

06:37 slew to zenith for reconfiguration to LUCI/LUCI

06:50 start fighting IDL license problems..... too long delay for Steward license from cosmo. And now the crate3 communication problem has returned on DX AdSec (see DougM for details).

07:02 Preset ACQUIRE for LUCI/None

Reconfigure for MODS/MODS

12:17: LBTB's target set early. Once they were done, reconfigured MODS1+MODS2 to try to take a spectrophot standard.

12:30: acqBinoMODS g191b2b.acq ... well into twilight, but at least we'll test acquisition (acqBinoMODS and long-slit).

• Works! So it would seem that the *img scripts don't work with the binocular lbttcs.
• Both red imcslocks just exceeded their 120-sec timeout. Retry worked. rotator angles 337 deg. elevation 63 deg.
• MODS1 "offsetxy -11.034 8.226 rel" need dx = 1.7"
• MODS2 "offsetxy -8.045 3.076 rel" need dx = 2.84"
• need to check/adjust reference positions for 5" slit
• execBinoMODS g191b2b.obs --- 12:44 (ridiculously close to sunrise).
• I DID NOT WAKE UP EITHER MODS ... HOW THEY LOCKED BEFORE IS UNCLEAR, BUT THE OFFSETS FROM REF POSITION NOT VALID
• wake up both MODS
• acqBinoMODS
• MODS1 -9.583 8.130 rel
• MODS2 -6.021 6.213 rel
• MODS1 looks centered (slit center 503.5) MODS2 is a bit left of center (slit center 513.5).
• too bright! Putting mods to sleep and closing up.

• Decide to try MODS2 IMCSLOCK on blue.
  • woke MODS: irlaser on, irlaser enable
  • ran grlamps.cal. Check irlaser and it reads disabled. Sounds like it could be an instance of the laser logic problem. Further investigation.

-- %USERSIG{JohnHill - 2016-10-11}%