Draft 2007 Aug 30 - *** indicates query or unresolved issue

The recipes below supplement the NAOMI setup and observing recipes.

• Safety
• Startup
• Mirror-flattening
• LGS WFS
• Measuring the WFS offsets
• GLAS control
• GLAS light-path mimic
• On-sky laser checks
• Tip-tilt correction
• Observing
• End of night

• Command glossary
• Troubleshooting
• Technical information

The laser is mounted on the top-end ring of the WHT, feeding the beam-launch telescope behind the secondary mirror. The intense 5149-A light from the laser is an eye-safety hazard anywhere in the dome area, particularly if any part of the beam reflects (accidentally, or intentionally) off the inside surface of the dome.

Access to the dome is strictly controlled (by the laser safety duty officer) during observing.

• On the aodisplay screens, login to sextans as whtobs. This brings up the exceed windows manager. The left and right screens are 8-bit and 24-bit displays respectively, as before.
• Open xterms by clicking on the whtobs@se.. icons. One of the icons opens an xterm on the 8-bit display, the other opens one on the 24-bit display.
• As of 24/8/07, guis run on sextans are very sluggish. It's better to run the Codeso software (oasis gui) from an lpss94 xterm running on aodisplay:
  • ssh -X whtobs@lpss94
  • obssys (then select 1, and select 1 again)
  • startoasisserver [only one of these can be running]
  • startoasisgui
  If you need a copy of this gui on the observing system, repeat the above steps there, but omitting startoasisserver.
• In each xterm window, run obssys, 1, 1 (this will also load the usual .aliases)
• TopGuiGlas & (not TopGui, to force colour lookup table to be correct).

To switch to LGS WFS mode:

• Continuous WFS display off, stop framing on the WFS page, select LGS from setup menu on top bar of topgui.
• In an xterm window, type: CameraSwitch.py, select slave, check that the status updates to slaveSynch on the GUI, then close the GUI. [*** should this step be before or after the re-init?]
• Re-init SDSU, start WFS framing, select continuous display.

Note that on the LGS WFS, there are only 8*8 spots (and the bottom four are very faint), due to vignetting. Each spot is accompanied by a ~ 1%? ghost due to reflection from the back surface of the beam-splitter.

The laser is started up by the daytime engineering staff (ideally at least 6 hours before observing). The DEs' laser start-up and laser/BLT checks instructions can be found on the ops-team home page.

The laser traffic-control web page is usually brought up with the observing system.

To kill a window on aodisplay, you can no longer use right mouse button, 'kill window', like you could on solaris: 1.Type xkill in a terminal. 2.A message appears saying Select the window whose client you wish to kill with button 1...., and a skull and bones pointer appears. Click on the appropriate window with the left mouse button. (If you change you mind, then click on one of the other mouse buttons.)

At this stage, if required, the DM / NGS loop (i.e. just about everything apart from GLAS) can be tested by measuring offsets on the NGS WFS, closing the NGS WFS loop and checking what happens the pinhole is shifted 0.1 mm (the usual test).

For high-order correction with the LGS, it's necessary to measure offsets on the LGS (see below), *not* on the NGS.

• Power up the high voltage unit located in the electronics room (ask an engineer).
• On top of the shutter unit (the square double-decker arrangement on the optical bench with lots of optics sandwiched in between) sits a switch unit. This must be 'on' and the top switch in the correct polarity. Check with white paper at the shutter output to see if green light comes out.
• For calibration work, ND filters need to be placed in the WFS optical train (don't forget to remove them for normal on-sky operation).

• Put in the LGS calibration source (when not in use, it's usually positioned behind the fold flat):
  • Centre it over the magnetic contact as best as possible, being careful not to touch the DM or fold flat.
  • Flick the switch to the left. Ensure that it is correctly mounted.
• Switch on/off lamp by plugging in/out the lead at the left side of the optical bench.
• Switch to LGS mode (as indicated above)
• Set LGS WFS mode with WFS SetGlasMode 1
• Check the intensity of the LGS calibration source. This should be as bright as possible without causing saturation on the WFS. The intensity can be controlled from GRACE.
• On the WFS page of topgui, click on alignment and on zernike to bring up the alignment and zernike tools (2 windows).
• The first step is to bring the average X/Y centroids to zero by offsetting the LSM postion. The values are given at the top of the WFS alignment gui.
• WFS ZeroWFSoffsets slave - explicitly zeros the offsets file before starting.
• WFS SetLSMtipTilt 4096 4096 - adjust values until average x and y are within ~ 0.01 of zero Values late Aug 2007 wewre 4050 3580, but we don't know how reproducible this is.
• On the zernike display, focus should be ~ 0 (within 0.1?)
• Bring up FSM traces for segment 1 (corresponds to LSM on the plot pages). *** The A and B values should be selected as A and B correspond to x and y(?). FIXME which plot?
• Zero average x and y by moving LSM, to maximise dynamic range
• Get LGS WFS offsets with the following commands *** on lpss42 *** (it doesn't work on ***sextans, and don't use auto-offsets on the autolauncher - there will in future be an AutoOffsets script for this).
  • WFS SetWFSoffsetsToCurrentCentroids slave n
    takes n frames (100 if not specified), determines the average centroid for each subaperture and automatically subtracts average values from centroids when computing AO correction. Note that the offsets are already 'loaded' once this command has been issued, however they should be saved (see below) so that they can be recalled later.
  • WFS SaveWFSoffsets slave
    saves the offsets, returns the name of the file
  • A warning message will be given if the offsets on a particular segment are too large (>0.3? of a box). There is currently a bug, meaning that you have to add 1 to the segment number quoted, in order to get the real segment number (as displayed on TopGui).
  • If loaded successfully, average x and y and rms x and y should all be ~ 0
• LGS calibration source out, LGS ND filters out To remove the LGS calibration source, reverse the instructions above: hold it and flick the switch to the right, carefully remove it from the light path (be careful of the DM/fold flat) and place it behind the FSM.

To obtain a good AO correction, the best strategy was to take offsets
on-sky. To do this, (i) centre the LGS on the WFS using the BLT, and
(ii) use the LSM closed loop to take out any motion, then (iii)
measure the offsets. (The spots should be as well centred as possible
to maximise the dynamic range.)

At the moment, the LSM loop is only controllable by close the LGS
loop. Therefore Tim had to change the gains to 'freeze' the DM, so
that the offsets could be taken while the LGS loop was closed:

WFS SetGain 0.0002 (NB. Unilluminated segments stay fixed at 0.)
WFS SetLSMgain 5000 5000

He then set up plots of segment 1 (LSM) and a different segment (one
of DM segments) to check that only the LSM was moving, not the DM,
once the LGS loop was closed.

2.5 ms exposure time on WFS.

Check 'None' loaded for Offsets.
WFS ZeroWFSoffsets slave [he wasn't 100% sure if next command did this anyway]
WFS SetWFSoffsetsToCurrentCentroids slave
WFS SaveWFSoffsets slave 

If a large offset on one segment, then open the loop, maybe increase
to 5ms integration time, then close the loop and repeat taking the
offsets.

N.B. Tim originally tried using larger LSM gain and smaller DM gain,
but numerical errors caused problems.  Also, half way through the 100
frames taken for the offsets, he lowered the DM gain to 0.0001.

What should the gains be reset to?

[Aside: Currently the LSM is not offloading to the BLT.]

To load offsets (e.g. if you are testing the AO correction for offsets
taken on-sky vs off-sky), use TopGui. (There is also a command line
option, but Tim is not sure whether it works.)

Below is a snapshot of the GLAS control gui:

Click here for the web page on which the GLAS internal-camera images are displayed. Alternatively, type 'gracecamserver1' in a web browser (this currently does not work on aodisplay - try using whticsdisplay1).

1.The laser ON/OFF status is not known by the observing system
yet. For the purposes of commissioning, the mimic now has the status
fixed to be ON all the time.  

2.The beam dump should show the correct
status. (However do not rely on this at this point.)  

3.The boresight
camera mirror should show the correct status, but it hasn't been
possible to thoroughly test this.  

4.The BLT x, y, focus values are
correctly displayed. (At a later date we may ask Craige to change the
precision, but it depends if we continue to use the x, y values or if
a different coordinate system is used.)  

5.The power values should be
ignored for now, as this is awaiting further engineering work.  

6.The
Laser calibration unit is deployed manually, therefore is fixed to be
'out' on the mimic.  

7.The Pockels cell and Image/Pupil View mirror
status are not known by the observing system.  

8.No LGS WFS info is
displayed yet.  

9.The NGS WFS status is correctly displayed.  

10.There
is no LSM information (I need to check the original specifications).

11.The tiptilt info and NGS loop status is correctly displayed.

12.The LGS loop status isn't being updated.  

13.In the NGS WFS, the
(i) camera focus, (ii) lenslet and (iii) filter mechs update okay. I
don't think the pickoff probe does, and I haven't tested the
foreoptics focus (although I suspect this will be okay).  
14.Most of
the realtime status (certainly the mirror flat and the DM gain) seemed
to be okay apart from the SG gain.

• Select LGS WFS mode with WFS SetGlasMode 1
(*** 1 needs to be specified? Is this the same as selecting from TopGui?) The fast readout should be selected. (*** Are there two options?)
• Tweak BLT x,y position in 0.05 steps, e.g. to 3.0 4.57, to steer the LGS behind the 8-arcsec stop (as viewed on the video camera). Later, tweak in smaller steps to centre on WFS.
• Tweak BLT focus if required (minimise size of spots on LGS WFS).
• Set WHT focus to 98.45 (measured 24 Aug 2007).
• The importance of the telescope focus requires some explanation. The LGS-WFS has a fixed focus position where the reference is the laser calibration source. The laser calibration source is conjugated to some fixed distance from the telescope (typically, 15km). That distance is defined by the timing delay set for the Pockels cell shutter. If the WHT focus is not at the correct value then this will affect also the laser beacon focus and hence introduce an unwanted focus term on the LGS-WFS. The system will want to take out this term by demanding the DM to correct for it. Whether there is a focus problem can be tested by checking the focus demand from the LGS-WFS in open-loop and closed-loop mode. There should be no difference between the two situations. To get readout of the focus term, type PrintFocus 10. The output is in (approximately) the same scale of the WHT focus. If there is a significant difference, the WHT focus may be wrong, or the range-gate distance must be adjusted.
• Adjust range (A) and duration (B - A) using control in GLAS rack (keys 2, 8 to increase, decrease, 'delay' key to switch between A and B) Matches range to position of calibration source.
• Laser alignment test (not necessary as regular test):
  • Determine the rotator center on any camera
  • Put a bright star on the rotator center
  • Defocus the WHT to bring the star in-focus on LGS-WFS; this will require ~ 7mm positive shift to the WHT focus
  • Put the LSM at the mid-range values of 4096 in x and y
  • Find the star in the WFS by moving the telescope, center it up, and record the positional offset.

Open Jure's TT control page. To start the TT server, using sextans (with an obssys applied):

• ttServer -n taurus -p taurus

Note that a log created by the TT control system can be found at tiptilt/2007xxyy. A new log is started every day at 12 UT. Severe problems are colour coded in red.

At the start of the night you should take bias frames (using the 'Take bias' button) for the TT configurations you plan to use.

Note that the tiptilt real time process is automatically started again if it stopped (e.g. with the 'kill' button on the GUI). If the controller is switched off, the RT process goes to sleep. It will recover when the controller is switched on again.

Below are screen dumps of the top part of the page in camera and plotter modes respectively:

Click on topgui NGS display to bring up tit-tilt display at centre of main WFS display (currently not very useful, intensity levels are wrong).

• Set default parameters for GLAS operation with WFS SetGlasDefaults
• Set loop gain, e.g. 0.2?, start reconstructor, from topgui.
• WFS SetLSMlimit 200 200?
• Move telescope to target, acquire on NGS WFS (i.e. the TT display), close the NGS loop (from the TT gui or topgui).
• Select 'Autoguiding' to be 'On'. (Note that when the loop is open, zero packets are sent to the TCS.)
• Check that the TO has set the black autoguider switch to ¡È#2 Tiptilt PC¡É and ask the TO to start guiding.
• Steer the star onto the LGS WFS, using the BLT x,y steering control on the GLAS gui, and watching the image on the 8-arcsec inspection camera (NB the coordinate system is not orthogonal to the WFS). Steps ~ 0.02 mm are needed.
• Set the display maximum value to 500 for integration times 2.5 - 5 ms.
• Close the LGS loops.
• Monitor the trace for 1 DM segment, centre it using the BLT steering control.
• Useful GLAS-mode commands are given in the command glossary later in this document.

The acqtool has been calibrated (with dichroic 4). When the pickoff is at 0,0, the acqtool pixel is 286.0 283.5. The acqtool has successfully been used, but some of the OASIS acquisition features may not work if they require control of the AO loop.

To perform open loop tests, but with autoguiding, keep the TT loop closed, but with the gain set to zero. (Jure can give more details.)

Follow the usual end-of night procedures for NAOMI and the science instruments, and cover the GLAS optics.

• PrintFocus n - average n frames to obtain equivalent mm of secondary-mirror motion.
• WFS - list WFS commands

• WFS Get* commands return the values of parameters:
  • WFS GetGlasMode
  • WFS GetLSMlimit
  • WFS GetLSMtipTilt - get the LSM x,y demand values
  • WFS GetLSMgain
  • WFS GetTiltRemoveWeight
  • WFS Get8x8mode
• WFS SetGlasDefaults - set default values of WFS parameters for GLAS operation
• WFS SetGlasMode 1
• WFS SetLSMgain x y
• WFS SetLSMlimit n - limit LSM travel in DAC units about mid-range.
  Total range is 0 - 8195. 8/07 - set to 2000.
• WFS SetTiltRemoveWeight xw yw - set weights for calculating tip-tilt to be removed.
• WFS Set 8x8Mode 1 - set 8x8 pixels per subaperture, disabling the normal 1-pixel guard band.
• WFS SetWFSoffsets filename
• WFS ZeroWFSoffset masterinslave

TopGui very slow to respond.

Kill duplicate processes, restart topgui (type: ps -ef | grep TopGui; kill XXX; TopGuiGlas &). Try running oasisgui on lpss94 instead of sextans.

Beam dump can't be taken out while parked at zenith.

There could be a discrepancy with the TCS (if telescope in ENG mode?), meaning that the laser safety system won't permit the beam dump to be moved out.???

Acqtool can't move the pickoff probe.

Ensure that the NAOMIHOST environmental variable on taurus has been set to 'sextans.ing.iac.es'. Check that the EPMDataServer is running on sextans.

Loss of c40 communications

Type ¡Èc40CommunicationsRestart¡É as naomi@aocontrol1

Sequencer dies

Type: Sequencer2.py

TopGui won't come up in LGS mode

GlasRestart

Serious GLAS RTC failure

Rene's notes 26/8/07:

- kill TopGuiGlas process
- GlasRestart
- Look out for error messages of scripts that fail to load. Currently three
fail always (this will be fixed; is easy).
- To load the three failing scripts, enter the Python script names in
  sextans xterms:
     Sequencer2.py (this should be run from a separate terminal)
     SubscribeBridge.py (this and PublishBridge.py can be run from the same
       xterm)
     PublishBridge.py
- WFS SetGlasDefaults     to reload defaults
- TopGuiGlas   to run up TopGui
- on CPU page, close the strain gauge loop
- on CPU page, start LGS WFS
- If there is a problem with the framing it may 
  be necessary to re-init the 
  LGS WFS on the SDSU page. However do not do this by default.
- on mirror control page, LOAD last flat
- on real time page, LOAD NAOMI reconstructor
- on WFS page, load offsets
- on SDSU page, set integration time
- on an Xterm type: CameraSwitch.py and in the box that appears click
'slave'.
Check that slaveSynch is set as the camera. 
Then exit that box.
- on TopGuiGlas activate continuous readout and adjust display control if
necessary.

The main components of the system are shown below (figure from the newsletter article):

The GLAS cradle (top-end ring) and the layout of the optics on the GRACE optical bench are shown below (also taken from the newsletter article):

Individual components in the GLAS launch system:

• IR pump laser, housed in the GRACE electronics room.
• GLAS laser (5149 A), mounted in a cradle on the WHT top-end ring.
• Beam diagnostics
• Beam-launch telescope (BLT) mounted behind the secondary mirror.

The beam is 35 cm wide when it leaves the BLT, and focuses to a spot 10 cm wide at altitude 15 km (i.e. subtended angle ~ 1 arcsec). The laser pulses at 5000 HZ?, and each pulse lasts 400 nsec?, total power ~ 17 - 19 W, yielding an artificial star with equivalent mag ~ 9.5?

Individual components in the GLAS LGS WFS light path:

• The beam-splitter (notch filter, 5149 A +- 75 A) reflects returning laser light to the LGS WFS arm, and passes the remaining wavelengths (throughput 95%?) to the NGS / science camera. The loss of wavelengths around that of the laser affects the OASIS LR610, MR516, HR502 and HR530 configurations. The filter is in principle removable, but in practice it is a lot of work to remove and replace it.
• Laser steering mirror (LSM).
• Field stop and 8-arcsec camera view (useful when steering laser spot onto the LGS WFS).
• Pockels cell shutter.
• Shack-Hartmann array.
• CCD

• Rotator centre on AG3 determined 24/8/07 to be at 594.5, 452.3.
• To focus a natural guide star (rather than the LGS) on the LGS WFS, offset the WHT focus by 7 mm (positive?).
• Turbulence above La Palma - > 70% is from < 3 km (JKT scidar?).

---

Last Updated: 2007 Aug 26 Chris Benn crb@ing.iac.es, Sam Rix srix@ing.iac.es Rene Rutten rgmr@ing.iac.es