|
1
|
- This procedure includes enough checks to serve as “pre-AO run” check
list. Not everything would be needed on a normal day after the run has
started
- In GRACE
- Outside of GRACE
- Starting up the system
- Laser flattening the Deformable Mirror
- White light flat
- Simplex flat and checking image quality
|
|
2
|
- In GRACE
- Environment
- Check if humidity is acceptable (Hygrometer on the table); switch on
or off the de-humidifier accordingly, empty its tank
- Inspection of the light path
- Take out all mirror covers (DM, OAP1 flat, FSM, Fisba small return
flat, OAP2 flat). Place covers in the dedicated plastic rack
- Check if nothing obviously obscures the light path (forgotten tools,
mask). More subtle obscuration will need imaging with WFS to reaveal
itself.
- Check which dichroic is set: 50/50 for observations with OASIS, IR
dichroic (Barr) for INGRID/OSCA. From Nov 2005, dichroic changer are
controlled from oasisgui or a dedicated PC in the Control Room. There
is no physical contact anymore.
- Check which pinhole is set at the Simplex NCU (need 2mm)
- Electronics
- Reboot the FISBA PC every other day (all automatic; simply “Restart”)
- Check if all VME racks are alive (all green…). This can be tested
later on from TopGui or an engineering Gui at Navis.
|
|
3
|
- In the Control Room
- Check on the air-conditioning control panel if temperature is stable
- Run up obssys on taurus, check cameras:
- AG3, AG7, INGRID, MITLL3 (for OASIS)
- From NAVIS rightmost console and keyboard
- Start a whtobs sesion with a telnet to Taurus (obssys, 1, 1). Before
telnet don’t forget to xhost + to authorise graphics from Taurus onto
Navis.
- startoasisgui ; if not working try first shutdownoasisserver; startoasisserver;
startoasisgui
- From oasisgui float the bench; move NCU to simplex
- Check if all mechanisms are in green on USP, Ingrid, OASIS pages
(otherwise a corresponding VMS rack will need reboot)
|
|
4
|
- From NAVIS leftmost console and keyboard
- See if TopGui is running, otherwise issue a restart or NaomiRestart and
topgui & at the NAVIS prompt. There is a “special case” recipe in
the green Naomi Software folder, at the tab Trouble (hand written in
the margin)
- From the Light Path page load the latest DM flat. After a long period
of inactivity (> week) try out several previous flats if needed.
For Fisba to work fine, interference fringes should be seen on every
segment.
- From the C40 page Start WFS framing, close the SG loop
- Check if the SG gain is 0.8, otherwise type SG SetGain 0.8
- Check if the WFS is ok and
Click on Continuous
- Check noise level on the WFS, visually or with a Grab
- Check on the NAOMI logbook where was the pickoff to see spots. If the
pick-off doesn’t move, INIT+START and INDEX+START from the engineering
gui launched from EngGuis – Mechanisms DM
- Sanity check for the SDSU: try changing to Mode 10 (while the WFS is
framing) then switch back to Mode 1
|
|
5
|
- From oasisgui switch on the simplex lamp
- Spots should be visible on the WFS window. Otherwise (and if the pickoff is at the right
position, NCU light is on, shutter open) something has been left on
the light path!
- As a sanity check, from TopGui try moving the FSM in X and Y, from
Light Path Page: Mid range is for 4096 adc; go to 5192, 3072, you
should see the spots move to the corner of a WFS cell if they were
centered (1 arcsec = ˝ cell). Then Midrange (not Zero!)
- Light should fall on AG7 (OASIS+Simplex mirrors); run 0.1 AG7 from
Taurus, or check from the simplex:
IngridAlign.py --camera AG7
- or mirroralign –camera AG7
- At this stage if the image looks good enough (fwhm <3 pixels) the
checks are done
|
|
6
|
- Perform a laser flat: equalise elements tip and tilt
- Go to the air conditionning console (upper left corner above the TO
console, next to the met station) and set Fan speed to 20%
- From TopGui Engineering Menu launch FISBA
- Click on One Shot , if nothing happens the FISBA PC needs rebooting
- If all DM elements are not visible (i.e. Fringes not present on some
segments) tweak by hand (if loading the latest flat was not enough)
- Click on Iterate
- Save this flat from the Light Path page DM controls
- All DM elements should show one fringe only. Now they are all parallel
but maybe not in the same plane
|
|
7
|
- Perform a white light flat: to equalise in piston
- Switch on Simplex light, Iris = -150
- Set WFS Integration Time to 25ms or 50ms
- Spots have to be bright enough for the algorithm to converge
- From the General Scripts window do an Autocentre
- Check focus; the spots separation has to be close to 8 pixels. Adjust
the WFS fore-optics focus from the EPICS guis
- Type AutoWhite from Navis prompt
- Answer 0 (leave alone) to the question about Integration Time
- The process takes more than 5 min; the X-shifted and Y-shifted lenslet
- Note the result: the algorithm gives satisfactory result if:
- There are more than 120 measurements
- X and Y gradients are less than 200 in absolute value
- For specialists: If Gradients are high, take out a global tip (X) or
tilt (Y) with AddZernike tip –(gradient/100) then the return flat will
need adjusting in Grace, and a new Laser/White light flat
- After Laser Flat and White light Flat, don’t forget to restore air
flow to >80% in Grace.
|
|
8
|
- Perform a simplex: to optimise image quality
- Run the simplex (with AG7 visible light or Ingrid for infrared)
- Set WFS Integration Time to 25ms or 50ms
- To simplex with Ingrid, take out OASIS and Simplex mirrors (from USP
tab on oasisgui),otherwise (simplexing with AG7) you need both mirrors
in.
- Adjust the Simplex Iris so that the maximum intensity measured is
between 10000 and 20000
- Click on Simplex
- Monitor the Merit Function (MF): image is considered good when MF <
-0.015 (with AG7) or MF < -0.03 (with Ingrid)
- Usually image starts improving after 800-1000 cycles
- If after 2000 cycles MF doesn’t improve, stop simplex then start
again: the algorithm may be stuck in a local minimum
- When the MF is satisfactory, stop simplex
|
|
9
|
- Check image quality from the observing system
- run AG7 0.05
- Display the .fits file, use imexam j and k commands
- Image is good when its fwhm is between 2 and 3 pixels
- Complete description of the NAOMI checks:
- http://www.ing.iac.es/~crb/wht/aorecipe1.html
|
|
10
|
- Derotator, NCU, Flat, DM. back
|
|
11
|
|
|
12
|
- NCU and pinholes, DM. back
|
|
13
|
- Fisba PC (in electronics room). back
|
|
14
|
- oasisgui (showing here usp mechanisms).
back
|
|
15
|
- oasisgui (showing here ingrid mechanisms). back
|
|
16
|
- oasisgui (showing here oasis mechanisms). back
|
|
17
|
- Tools menu in TopGui. back
|
|
18
|
- Light Path page in TopGui. back
|
|
19
|
|
|
20
|
|
|
21
|
- Close Strain Gauge loop. back
|
|
22
|
|
|
23
|
- Shack Hartmann Spots on WFS. back
|
|
24
|
- How to check the FSM is alive. If the spots don’t move, suspect a fault
or simply go into GRACE to turn it on! (along with the FSM filters, in
the first rack from the left in the electronics room) back
|
|
25
|
|
|
26
|
|
|
27
|
- Engineering gui’s menu. back
|
|
28
|
|
|
29
|
|
|
30
|
- Mirror control page. back
|
|
31
|
|
|
32
|
- Autocentre. Press Enter to make window disappear. back
|
|
33
|
|
|
34
|
- Adjust Fore-optics focus from mechanisms engineering gui. back
|
|
35
|
- Result of AutoWhite. back
|
|
36
|
|
|
37
|
- Merit Function of the simplex algorithm. back
|
|
38
|
- Result of the simplex algorithm.
back
|
|
39
|
- Checking image quality with imexam.
back
|
Notes
