See computer startup instructions on wall above each system console. Usually computers are left up and running, since installation of UPS. Check for response from keyboards.
TCS3210 (TCS) - will give UI> prompt. ICS3220 (ICS) - will give * prompt.
* SI XYZ,n,n XYZ=observer n= PE account (see whiteboard). Write initials on board, dates resident and write NO in the 'ok to delete?" column. * ADAM and answer prompts: f/15, A&G+RBS, EEV7/TEK4 etc, run#(n) and D to dump to disk only. To abort from an Adam command press break, then SEND CON:, unless the Adam terminal hangs during readout, in which case type RESET at the ICS system terminal, which forces an automatic restart. To abort from Adam altogether press break then * CA (cancel).
Note that the new SCSI ICS disk drives are 2Gb capacity, and can store up to approximately 450 full-frame big chip images, a factor of 3 improvement over the old drives, with the Sparcstation having 18 Gb capacity.
The RBS has exchangeable optics, optimised for blue or red (>5000A).
Changing the optics is straightforward, and can be done in about 20
minutes. Follow the instructions in Roger Edwins The St Andrews
RBS User Manual, and you can't really go wrong. There are up to five
components to change: slit assembly (either 1 or 1.5 arcs), collimator,
grating, camera mirror and corrector unit. The change procedure for
each optical component is accurately described in the
Manual: note only that the screw on cap protecting the collimator is
very stiff, due to some unfortunate impact with the telescope pier,
so care must be taken when replacing the cap, to prevent cross-threading.
All the RBS optics are stored in an aluminium case in the left-hand
cupboard below the worktop in the "darkroom", adjacent to the dome
area. You will also need Allen keys to remove the grating and camera
mirror assemblies.
Before you start to set up the RBS, ensure that you have the correct
optics mounted, and with the grating angle set for the appropriate
spectral region, for the grating in use.
Good, so you've changed the optics. The RBS is a REAL spectrograph,
all RBS functions being controlled from the observing floor. Externally
there are four controls of interest during setting up, and to a lesser
extent during observing: grating angle clamp, slit and dekker slide,
Hartmann mask, and finally the micrometer on the camera mirror unit,
used for focussing.
The grating angle is changed by first unclamping the grating using the
bronze knurled clamp on the grating scale, rotating the grating to
the desired angle using the knob on the side of the spectrograph,
labelled Unclamp grating before rotating, then re-clamping.
The slit and dekker slide has two detents, although finding the inner one can be a little tricky, as it is not well defined:
Fully In - whole slit covered Inner detent - star slot uncovered - central 2mm exposed Outer detent - arc slots uncovered - 2 x 2mm exposed Fully out - whole 20mm slit exposed - Longslit
A 2mm slit length projects to about 0.5mm wide spectra at the camera
focus. During observing stellar and comparison spectra are usually
exposed on separate frames, using Longslit mode (slide fully out).
The Hartmann mask has three positions:
In - first Hartmann position Detent - second Hartmann position Out - for normal operation, secured in stowed position using fork clamp.
The micrometer on the camera mirror is used to adjust the spectrograph
focus. Initial focus for the red optics is about 7.5mm, and 7.4mm for
the blue. Unless the focus is apalling, micrometer increments of 0.1mm
should be used initially. The Hartmann focus test is described later,
but beforehand the mounting of the cryostat itself should be discussed.
The CCD cryostat is mounted on the RBS using the familiar RGO kinematic
mounting plate. This has three capstans with micrometers - the capstan
settings are contained in the black RBS logbook in the control room,
and should be set-up initially by the Duty Engineers. Hope and pray that
these are correct, and that the CCD is orthogonal to the RBS focal plane,
as correcting any tilt on the CCD is not for those of a nervous
disposition - suitable tranquilisers may be found in the RBS optics
case: these will almost certainly be required when squaring-on TEK4.
The capstans are mounted at about 8, 4 and 1 o'clock, and are referred
to by the letters A (Cone), B (Slot) and C (Flat), engraved on the
kinematic mount. The
cryostat is mounted so that it's dispersion direction is horizontal
when viewed from behind the cryostat so A and B should correct tilt in
the dispersion direction, and C, spatial tilt. If you do determine new
capstan settings for a particular detector, please ensure that these
are recorded in the RBS logbook for future reference.
The other crucial adjustment is the micrometer controlling cryostat
rotation, thus allowing the dispersion direction to be accurately
aligned with the rows of the CCD. Before attempting to adjust the
micrometer the six 10mm faceplate bolts must be loosened.
The Ellis A&G, on which the RBS is mounted has 2 hollow cathode calibration lamps and a Tungsten lamp for flat fields. All three can be fitted to the integrating sphere on the A&G box, or alternatively either of the hollow cathode lamps can be mounted directly on the same A&G port, secured by a locking collar. Note that the Tungsten gives flat-fields about as flat as La Palma when used in the integrating sphere; use dome or sky FF's instead. The three lamps are connected to a power supply on the A&G as follows:
CuNe Position 1 CuAr Position 2 W Position 3
Since the control room reorganisation during 1995, with the electronics racks being moved into the ICS room, the RBS handset can no longer be used (as the cable is not long enough), instead all handset functions are now controlled by Adam commands, as follows:
CUNE ON/OFF turns on/off the cune lamp CUAR ON/OFF turns on/off the cuar lamp TUNG ON/OFF turns on/off the tungsten lamp COMP IN/OUT moves the Comp/Acq mirror into/out of the beam SLIT IN/OUT moves the slit view mirror into/out of the beam
The point to watch here is that if one mirror assembly is in the beam
and you want the other mirror, you MUST move the unwanted mirror
out of the beam before attempting to move the other into the beam.
Failure to observe this may lead to Adam becoming confused and acting
strangely (nothing new there!). So for example, if you've just done an
arc, and want to return to slit view, you must type COMP OUT and then
SLIT IN. Each mirror movement will take about 25 seconds, so be patient.
First set-up priority is to ensure that the CCD rotation is correct, and that the dispersion direction co-incides exactly with the rows of the CCD. The RBS is very critical here, since a 2mm slit length projects to about 0.5mm at the detector. Ensure that the micrometer on the camera mirror has been set to 7.5mm for the red optics, and 7.4mm for the Blue optics, or check the RBS logbook for the most recent values. Rotation is best done using arc lines, so take an arc with the dekker in longslit mode and the Hartmann shutter out, and measure the centroids at top, middle and bottom. Adjusting the cryostat rotation requires the six bolts on the cryostat faceplate to be loosened, and the cryostat rotated away from the stop. Do not attempt to rotate the cryostat with the micrometer - back off, move the micrometer, rotate the cryostat back until it touches the micrometer, then retighten the bolts, and try again. Ideally the rotation along the dispersion direction should not exceed 0.1 pixels.
Now take an exposure with the right Hartmann shutter in, and then one with the left Hartmann in, still in longslit mode (dekker out). Measure the centroids of the arclines in the right and left Hartmann images. The difference between the left and right Hartmanns gives the focus shift as a function of CCD position (i.e. the tilt). Once the tilts are known, adjust the capstans, and repeat, until the tilt in both spatial and dispersion directions is <0.3 pixels. Capstans A and B should correct for tilt along the dispersion direction whilst C, or C and A/B should correct for tilt along the less critical spatial direction.
In longslit mode (with dekker out), make two exposures as above, with the right and then the left Hartmann shutters in. Measure the centroids of the arclines in both images, the difference being the focus shift as a function of CCD position. Adjust the camera focus until the shift is zero.
After all your painstaking efforts please ensure that you record the capstan settings, rotation micrometer reading and camera mirror focus micrometer readings in the RBS logbook in the control room, date it, sign it, and indicate clearly which detector was mounted and which optics set was installed.
If the observers central wavelength is known, then set the grating angle accordingly for the grating in use (from the RBS Manual), and take an arc exposure. Identify the lines in the resulting spectrum using the arc maps in the IDS Calibration Lamp Arc Maps manual ,La Palma Technical Note No. 70. SQUISHing the image may help here since the Lexidata only shows the central region of the chip. Should the orientation of the spectrum be incorrect there are three commands available which will flip the image:
Adam:> XYSWAP=1 LRSWAP=1 TBSWAP=1
Finally, in preparation for observing, ensure that the Hartmann mask
is fully out, and secured in place with the plastic fork clamp, and
that the dekker slide is also fully out, in longslit mode.
Put the kettle on - You've earned it!
With the introduction of the Perkin-Elmer to Sparcstation networking,
image and header files are transferred to the Sparc as soon as they
become available at the Perkin-Elmer. Consult Gary Mitchell's notes
on how to set the networking in motion, which should be near the
Sparc.
For all data, two tapes are generated. The D tape is made by the
observer, normally at the end of each observing session, and goes
into the ING data archive. Before then, it is copied, either by
the observer or by the tape copier, to produce the copy or C-tape,
for the observer.
General rules for D tape usage are to append data to them, not to
use a new one each night, and to not append more that say 300-500
files to a single D-tape, since this makes data retrieval from the
archive laboriously slow. The D tape must always be a DAT tape,
although the C tape can be either DAT or EXAB.
At the end of the session, insert the D-tape into the lpss6 Sparc DAT drive. Make sure that all your files have been transferred to the Sparc, then pull up a new Xterm window and type:
AUTOFITS and answer the prompts
Within AUTOFITS is an option to initialise the tape. DO NOT do
this unless it is a NEW tape. If you initialise a tape which
contains data, then any data on that tape will be
irrecoverably lost.
Whenever you are dumping data, please leave a note on the terminal to say which files you are dumping, and give your name and time you started dumping.
You have two choices here. Either copy the data yourself and verify
it using the FITSLIST option within the TPCOPY facility in the WHT,
or fill in a Mag Tape Copy Request form and let the tape copier
do it.
To do it yourself, load a C-tape in the DAT drive again of the Sparc and repeat AUTOFITS as described above. This tape MUST then be verified using the TPCOPY facility in the WHT terminal area.
Alternatively, fill out a MTCR form for the tape copier and leave it in the filing tray in the control room.
After all files have been dumped to D tape and copied to C tape you will have to delete them from the disk, to maintain your working space. The command to use for this is PURGE, which will only delete those files which have previously been archived to tape. Before you PURGE any files contact your SA who will check the JKT tapes status (maintained by the tape copier), which will indicate which D tapes have been copied and verified. The tape copier (or observer, if producing and verifying their own C tapes) should write YES in the 'ok to delete?' column on the whiteboard, but the tape status file must always be checked before purging any data.
lpss6% more /home/lpss1/vbf/tapes/jkt.tapes Adam:> PURGE which will delete a specified range of files from the disk. The prompts are: which disk volume> SYS Detector> CCD Initials> XYZ (as for Login) Files to be purged> R1-R150 The above will purge files: SYS:CCDxyz.001 to SYS:CCDxyz.150 incl.
Adam:> RUN 0 "BIAS" does a zero second exposure (reading out CCD with shutter closed), and creates a bias frame, with title "bias". Adam:> MULTRUN takes multiple bias frames. Simply reply to the prompts.
Check bias value using STATS mode of PHOT. If different from default value when you initialised Adam, enter:
Adam:> NEW_BIAS n