12 May 2015 UT MODS2 Commissioning

Observers: MWagner, BRothberg, OKuhn
OSA: SAllanson
InstSp: JPower

Conditions: Cirrus - it was thicker earlier today, but seems to be thinning out, nearly clear around middle of night, but again clouding up towards twilight.

Rotator ZeroPoint and Transform:
When we took a series of images between which a detxy Y offset was made in active mode (what observers would do), we found the line of stars tilted still further CCW from the slit (line of stars at -0.9 deg from +Y axis, while slit at ~-0.14 deg from +Y axis). Offset in Y direction in track mode to confirm/check yesterday's result. Reverted to initial rotator zeropoint. Again found the same slope for line of stars (-0.9 deg, vs. ~-0.14 for slit). A new transform (or a rotation to the transform) is needed.

MODS2 acquisitions:
  • Responses to "Use all alignment star boxes as shown (Y|N)?:" should be case-insensitive (need to type "n" for no, since "N" does same as "Y")
  • rotation offset has:
    • sign error
    • seems to be consistent at 0.7-0.8 deg, which reflects difference in detxy Y offset vector and long-slit.
  • are distortions corrected in mms or in mods2Align?
    • for mask2, which has boxes closer to the edges, even after offset is made, stars are not all perfectly aligned in boxes - some in corner, some below -see image 57.
    • for mask1, alignment was good. But a dY = +1.0" needed to center; mods2Align alone did not perfect the alignment.

Timeout for large dY offset when low point close to bottom of guide patrol field. This led to a series of trailed images during which offset and go commands appeared to be out of sync.

Saturation on red as last night. Trails persist into next images (see mask field images).

Frequency of Blue sequencer problem in data from yesterday (mods2b.20150511.*.fits). Problems for the following images: 34, 42, 44, 45-48, 51 and 54. 1-33 fine. In the 21 images taken between 34-54, 9 are bad.


03:00 Offset timeout leading to race condition: Running slitscan with active mode, but we had a problem with offsetxy timing out. Seems like offset of -120" down, when low point is close to patrol field edge, times out. dY = -120" for a star a little closer to the MODS FOV was OK. Retried offset, but exposure may have already started, and offset/go were out of sync. Doesn't go wait for offset to complete?

>> offsetxy 0 -60 abs
** ERROR: Command 'offsetxy 0 -60 abs' timed out after 30 seconds
** Abort, Retry, or Ignore? > r
** Retrying command 'offsetxy 0 -60 abs'

I retried and it did the offset, but we got into a condition where we got the errors and saw images where the star is trailed as if it was offsetting while exposing.

>> offsetxy 0 -60 abs
DONE: offsetxy ABS Slit Offset Completed lbtSide=right
>> go
DONE: offsetxy ABS Slit Offset Completed lbtSide=right
>> offsetxy 0 20 abs
DONE: offsetxy ABS Slit Offset Completed lbtSide=right
>> go
** ERROR: GO cannot start a red exposure - red channel already exposing
** Abort, Retry, or Ignore? > i
** Ignoring error and continuing...
>> offsetxy 0 -20 abs
STATUS: RED GO Finished 1 of 1 exposures. ACQTAG=MODS2R.20150512030737 EXPSTATUSDONE: RED GO Exposures finished.
>> go
DONE: offsetxy ABS Slit Offset Completed lbtSide=right
>> offsetxy 0 40 abs
DONE: offsetxy ABS Slit Offset Completed lbtSide=right
>> go
** ERROR: GO cannot start a red exposure - red channel already exposing
** Abort, Retry, or Ignore? > 

Aborting the script and repeating. Ahh, but the script really did not end, although it said it did. But the dashboard shows the last slitmask image of 120 sec still ongoing.

>> exptime 120
DONE: RED GO Exposures finished.
>> go
DONE: RED EXPTIME ExpTime=120 seconds. -CLAMP

** Script GD153_act.img done
lbto@obs3:11 % execMODS2 GD153_act.img

** Executing GD153_act.img

>> OBSMode
DONE: OBSMODE Instrument in Observing Mode INSTMODE=OBS CALLAMPS='None' HATCH=OPEN AGWXGP=-31.863 AGWYGP=-164.646 AGWFOCUS=0.000 AGWXWFS=-31.863 AGWYWFS=-164.646 AGWXS=59.447 AGWYS=23.902 AGWFS=52.916 AGWFS0=39.504 XGPLAST=12.617 YGPLAST=-111.804
>> nimgs 1
** ERROR: nimgs Cannot change this parameter during image acquisition.

Waited until idle and now, finally, repeating the script.

The offsetxy 0 -60 abs timed out again. It could be that the guide probe could not reach this position? No, we verified that it could reach this position by manually entering this offset (0 -50, 0 -55 and 0 -60 all OK). But in the script the offset 0 -60 is preceded by offset 0 60, and the 120" offset may take longer than the defined timeout.

  • execMODS2 GD153act.img (ACTIVE MODE)

Co-added the series of images and measured the angle of the line of stars. This is now -0.989 deg. It is tilted more CCW than in track mode. This seems to confirm that we need to adjust the transform to account for the change of rotator zeropoint.

We are repeating this script in track mode, to confirm the results of yesterday.
  • execMODS2 GD153.img (TRACKMODE)
  • slope is -0.00135 (-0.077 deg) (vs -0.00248 or -0.142 deg for slit). Difference is 0.065 deg. This confirms that the rotator zero point needs to be changed.

But, since we don't have a matching transform, either by collection of data using this rotator zeropoint or by rotating the existing transformation matrix, we will return the rotator zeropoint from 26.95 to the initial value of 27.55.

Edited PCSInstrument.conf after which Steve restarted PCS.

Now running the slitscan script in active mode to check how well the detxy y offsets in active mode go up/down along the slit. Script got the wrong guide star, but we are guiding OK. r 37...47.
  • offsetxy 0 -60 did not time out this time. Earlier (with correct guide star chosen), the probe was driven closer to the edge of the patrol field.
  • slope of line of stars is -0.0162 or -0.928 deg, similar to what I got before with the new rotatorzeropoint. (I think this makes sense - the transform determines how the dY moves on the detector regardless of zeropoint.)

Switch to MODS1 for science observations while waiting for Stone L mask fields to rise.

08:17: Back to MODS2

acqMODS2 mods.7.StoneLf2.acq Preset was taking a long time and Steve said there was a problem with the rotator. He's resetting the rotator.

mods2Align:
  • drawn boxes are not aligned with actual mask (see Mask2/*boxes.jpg)
  • Use all alignment star boxes as shown (Y|N)? ... I chose N but all boxes are automatically marked on field image. Use "n" (lower case).
  • Used 3 boxes, for which it computed.
MODS2 Offset Command:
   offsetpointing -0.7550 -1.878 -0.291 detxy rel

mods2r.20150512.0052. Stars are not in boxes except in middle.

Did rotation go the wrong way? offset +1.51 0 0 rel ==> 53. Looks better.

offset 0 0 1.1 to move stars up. => 54.

acqMODS2 mods.7.StoneLf2.acq
lbto@obs3:114 % mods2Align mods.7.StoneLf2.mms mods2r.20150512.0055.fits mods2r.20150512.0056.fits
yes
z1=302.3052 z2=367.7603

** Use all alignment star boxes as shown (Y|N)?: n

** Mark Alignment Star Boxes manually with an 'x'.
   and then press 'q' when finished.
   *** You must mark at least TWO (2) boxes ***

Log file _modsalign.log open
2432.48 2601.86   1367.
2370.89  769.44    617.
 615.46 1334.05    600.
z1=503.378 z2=730.8497

** Mark the center of each alignment star with an 'a'
   and then press 'q' when finished.
   Stars must be selected in the *SAME ORDER* as the boxes.

Log file _modsalign.log open
#   COL    LINE     COORDINATES
#     R    MAG    FLUX     SKY    PEAK    E   PA BETA ENCLOSED   MOFFAT DIRECT
2432.46 2609.16 2432.46 2609.16
  20.64  10.00 1.002E6    620.  17090. 0.07  -30 3.94     6.57     6.86   6.88
2402.15  777.68  2402.15 777.68
  17.76   9.42 1.712E6    636.  34785. 0.09  -19 5.30     5.96     6.00   5.92
 636.72 1308.53  636.72 1308.53
  17.83  11.21 327303.    616.   6668. 0.02  -27 6.75     5.99     6.04   5.94

Computed Mask Alignment Offset:
   dX:  -1.913 arcsec
   dY:  -0.356 arcsec
  dPA: -0.8387 degrees

MODS2 Offset Command:
   offsetpointing -0.8387 -1.913 -0.356 detxy rel

Send the offset to the Telescope (Y|N)?: n

and in dashboard command window entered offsetpointing 0.8387 -1.913 -0.356 detxy rel

Confirmatory image is 57. Stars are in boxes, however they are displaced within the boxes in a regular, field dependent way. How is distortion handled in mms? for mods1? for mods2?

acqMODS2 mods.6..StoneLf1.acq => 58 59 and confirmatory thru-mask 60. mods2Align yielded

lbto@obs3:122 % mods2Align mods.6.StoneLf1.mms mods2r.20150512.0058.fits mods2r.20150512.0059.fits
yes
z1=302.8784 z2=367.1859

** Use all alignment star boxes as shown (Y|N)?: n

** Mark Alignment Star Boxes manually with an 'x'.
   and then press 'q' when finished.
   *** You must mark at least TWO (2) boxes ***

Log file _modsalign.log open
2704.52 1446.98    635.
1308.39 1975.66    644.
 866.97 1133.87    639.
z1=516.3771 z2=745.8526

** Mark the center of each alignment star with an 'a'
   and then press 'q' when finished.
   Stars must be selected in the *SAME ORDER* as the boxes.

Log file _modsalign.log open
#   COL    LINE     COORDINATES
#     R    MAG    FLUX     SKY    PEAK    E   PA BETA ENCLOSED   MOFFAT DIRECT
2722.14 1455.48 2722.14 1455.48
  22.32  10.61 572145.    635.   7473. 0.11    1 9.26     7.32     7.65   7.44
1312.47 1959.96 1312.47 1959.96
  23.05  10.54 608102.    639.   7507. 0.07   39 10.6     7.57     7.96   7.69
 887.49 1111.30  887.49 1111.30
  21.22  10.93 424204.    642.   5969. 0.03  -78 4.80     7.16     7.20   7.07

Computed Mask Alignment Offset:
   dX:  -1.306 arcsec
   dY:   0.901 arcsec
  dPA: -0.6966 degrees

MODS2 Offset Command:
   offsetpointing -0.6966 -1.306  0.901 detxy rel

Send the offset to the Telescope (Y|N)?: n
   OK, not sending the offset...

/home/MODSeng/bin/mods2Align done.

Followed by: mods2Cmd "offsetpointing 0.6966 -1.306 0.901 detxy rel" ---> yields image 60.

Needs still a +Y = 1.0" offset. Image 61 looks pretty good.

Summary of MODS alignment:
  • new question is buggy: Use all alignment star boxes as shown (Y|N)?: n
    • If upper case N is used, then it is like Y.
  • sign error in rotation offset.
  • are distortions corrected in mms or in mods2Align?
    • even after offset is made, stars are not all perfectly aligned in boxes - some in corner, some below - field dependent, especially on mask2 (first one, see image 57) where boxes are closer to edges.

Putting MODS2 to sleep (oops --- after I wake up MODS1 again, I'll put MODS2 to sleep).

Using end of night to do science + test IQ on LBCR.

-- %USERSIG{OlgaKuhn - 2015-05-12}%
  • mods2Align_boxes_allY.jpg:
    mods2Align_boxes_allY.jpg

  • stars used for first/second alignment (green/magenta):
    selstars.jpg
Topic attachments
I Attachment Action Size Date Who Comment
mods2Align_boxes_allY.jpgjpg mods2Align_boxes_allY.jpg manage 112 K 12 May 2015 - 09:11 OlgaKuhn  
selstars.jpgjpg selstars.jpg manage 16 K 12 May 2015 - 09:12 OlgaKuhn stars used for first/second alignment (green/magenta)
Topic revision: r5 - 12 May 2015, OlgaKuhn
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