First night checks
The following are the essential checks/calibrations to be carried out on the
first afternoon/night after LIRIS has been mounted on the telescope. For a
more complete list, see this slightly-outdated version hosted
at the IAC, and the quality control tests here.
Retarder plates
Check that the retarder plates (#996 and #997) have been mounted to the main
colour filter slide in the A&G box. For example, search the filter database by
the filter number. They should be mounted in the AGCA Main colour filter
Position 2 and 3, respectively. If not contact Neil and/or Emilie.
Preliminaries
To load the LIRIS utilities package, do:
SYS> source /home/whtobs/liris/setup_lirisutils
This will load the corresponding alias for every utility (limageallslit, slitprof, findmask, etc.) in that directory.
During afternoon
Make one 10s dark, with storemode diff_pre:
SYS> storemode liris diff_pre
SYS> lblanks
SYS> dark liris 10
SYS> storemode liris diff
Alignment of slits
(If you haven't, do the preliminaries!)
The purpose of these measurements is to update the information about mask positions which will facilitate the target acquisition during the night.
Data acquisition
- Check that the telescope is pointing to zenith.
- Check that the primary mirror petals are open.
- Check that the tertiary mirror is out (on the right side of the mirror petal pointer there should be light on CASS focus, not in GRHIL nor in GRACE), specially after a GLAS run.
- Switch on the top ring flat lamps 9W and 25W.
SYS> fflamp 1 on
SYS> fflamp 2 on
- Check that the agmirror is out. If not:
SYS> agmirror out
- Set the hc filter in imaging mode:
SYS> limage hc
- Select the 0.65 arcsec slit and do a test exposure (3 s):
SYS> lslit l0p65
SYS> glance liris 3
Check that you can see an image of the slit, with a reasonable number of counts, in the range 1000 to 10000 ADUs. If not, modify the exposure time accordingly in the following steps.
- Obtain the images, with a proper exposure time of all slit masks without grism in the light path:
SYS> limageallslit (exposure time)
This will take images of every mask and slit in the slit wheel and store the slit name in the image title. With 3s exposure time this will take less than 9 minutes. Repeat it:
SYS> limageallslit (exposure time)
and check the slits stay in the same positions (so the wheel works ok). Now, we will treat long slit and multiobject masks separately.
Long slit case
The orientation of all long slits has to be checked, use for example imexam +j, if they are tilted more than 1 pixel
(top-bottom) then one have to correct the wheel position. See
correction of the mechanical positioning of the slit wheel.
The positions of all the overlays for DS9 (stored in the configuration files) have to be checked against the real slit positions.
- Display the image corresponding to each individual
long slit mask in a DS9 frame and overlay the reference position using the
"LIRIS slits" button in the DS9 display*.
- If the overlaid box does
not coincide with the observed position, then:
- Move the region in DS9 until it is correctly aligned with the image of the slit, specially in the central region of the detector. Double clic on the region to get the region parameter values.
- Update the file /home/whtobs/liris/reg/liris_<slitname>.reg using the right X-CENTRE value (leave the Y-CENTRE value unchanged!!). An entry in such file should look:
image;box( x-centre , y-centre , slit_width , slit_length ,0.) # text={SlitId}
AND
- Update the file taurus:/wht/var/liris_slitdb_pos.dat, using the X-CENTRE value (leave the Y-CENTRE value unchanged!!), which are used by the acquisition scripts. An entry of such file (long slit case) is as follows:
slitname 1 x-centre y-centre
Polarimetry slits
Use the same method as above for fixing the x-direction for polarimetry slits i.e. clear0p25 and l0p75f0p25, used for imaging polarimetry and
spectropolarimetry, respectively.
To adjust in y-direction see correction of the mechanical positioning of the slit wheel.
*Note: If you see any odd behaviour of the slits, such as a strong misalignment, do
SYS> lsethome all
and go over to the Alignment of long slit section again.
Multiobject masks
(If you haven't, do the preliminaries!)
In case of multiobject masks the procedure becomes slightly more
complex:
- Check the new MOS mask names with a
SYS> lwheelconfig
and update the /home/whtobs/.ds9.ini file with them: update the button definitions (the text shown in the DS9 buttons) and the names of the .reg files those buttons refer to. If necessary, add new buttons or delete superfluous ones.
- For every MOS mask, measure the position of the slitlets and
reference holes: open a new terminal shell in whticsdisplay1, load the LIRIS utilities package and copy the images of the MOS slit masks previously generated to a temporary directory (e.g. named as the date, /home/whtobs/liris/slits20080609):
> source /home/whtobs/liris/setup_lirisutils
> cd /home/whtobs/liris/slits20080609/
Then do:
> findmask r<NNNNNN>.fit <maskname>
where r<NNNNNN>.fit is the image of the mask without grism in the optical path, and maskname is the name of the mask as written in taurus:/wht/etc/lslit_config.dat.
This will produce 5 files. The most important one uses the liris_maskname as root with the suffix .reg (used for display purposes in the
DS9 tool).
- Check r<NNNNNN>.fit in ds9 against the new liris_maskname.reg (load manually as the buttons in ds9 do not work yet). Sometimes the script findmask cannot properly identify the slits or the reference holes, in particular
if they fall onto special places on the detector*. In those cases one
must edit liris_maskname.reg and provide better values.
- Additionally, if the .reg files needed editing, you need to edit maskname_slitdb.dat and
liris_maskname.mask
with the new X-centre and Y-centre values of the modified slitlets, and the file maskname_specreg.dat with the following syntax:
slitlet N x-centre 1 1024 y-minimum y-maximum
- Open a shell in taurus, and update taurus:/wht/var/liris_slitdb_pos.dat with an entry per mask (add one if it didn't exist before). The line should contain, in the same order as in the
liris_maskname.mask
file, the coordinates of the REF positions of the mask:
maskname N x-centre1 y-centre1 x-centre2 y-centre2 x-centre3 y-centre3
Where N is the number of REF holes the mask has (2 to 4).
- Copy all the lines of maskname_slitdb.dat (another file generated by findmask) into the same file, taurus:/wht/var/liris_slitdb_pos.dat, appending them after the last line.
- Copy (replacing any old, obsolete version) the liris_maskname.reg files to /home/whtobs/liris/reg/. Copy liris_maskname.mask files, copying them to /home/whtobs/liris/mask/ AND to taurus:/wht/var/
- Restart any IRAF DS9 session and the RTD:
SYS> startudasrtd
*Note: If the mask is not properly positioned onto the
detector, the wheel configuration files have to be
updated. Follow these instructions.
Additional checks
Positioning of the pupil and grism wheel
From time to time, or if a specific mechanism (e.g. pupil, grism) has been recently
repaired/modified, some additional checks must be done. Please refer to the LIRIS Positioning of
the Wheels page for instructions on correcting the mechanical position of the
grisms and pupil wheels.
Also, from time to time, the detector should be monitored. Please refer to the LIRIS quality control page for instructions on dark current and hot pixels monitoring, and gain and linearity calibration.
Wavelength calibration
Take arcs of the most common configurations, for instance with the l0p75 slit and the ZJ and HK low resolution grisms.
IMPORTANT: Keep in mind that the grisms descriptors for lspec command are not the same as the names given in the configuration page. For example, the psgrj, psgrh and psgrk correspond to r3000j, r3000h and r3000k when dealing with lspec.
During twilight
Rotation centre
Select a bright star and take two exposures at two different rotation angles 0 and +180, and then
calculate the centre of the line connecting the two points. Update the file:
/wht/var/liris_rotcent_pos.dat
with the new rotation centre. If the two positions of the stars are separated by more than ~80 pixels, ask the OSA to do a 7-star pointing model centering the star in the rotator center.
Aperture Offset for quick acquisition
This test will take less than 5 minutes.
GOCAT to any star with magnitude about 10-12. You can use the one from the last part of the 7-star calibrate, or any other star you like, e.g. one from the pointing grid. Take a 1 second image with LIRIS in any filter as long as you can see the star. It doesn't matter if it's saturated. It should be obvious and quite close to the centre of the image. Find the approximate coordinates of the star: x y (to the nearest pixel is fine) and use the command:
SYS> lobject_inslit l1 x y
to calculate the offset to the default slit acquisition position at y=400. When asked if you want to move the telescope, answer no, we just want the offset required. Pass this offset to the telescope operator who can note it in the logbook along with the calibrate results and set it up in the TCS when long-slit spectroscopy is used.
After this test has been done a few times, we will know how stable the position is over time and it may be that it is not necessary to be done on every LIRIS first night. Please email the results to Fiona if possible.
Array alignment with celestial axes
Select a bright star and take two exposures at PA=0 or PA=90 and offset the telescope along
the RA or Dec axis by 1 arcmin. The vector connecting the two positions determine
the orientation of the array. In principle the angle should be
zero.
If it is not zero, ask the OSA to update the TCS LIRIS setup and retry. Otherwise it will happen that during spectroscopic nodding the target will be decentred or even drift off the slit when larger nodding offsets are applied.
Additional checks
From time to time, the photometric zero points should be monitored (follow procedure here, only in photometric nights), and the local background measured (here).
Also, the grism alignment and spectral distortion should also be checked from time to time (here).
|
