COSMIC is unique in many respects for Palomar. It is the only narrow line
imager. It is also the only multiobject spectrograph. Some general notes:
It is very labor
intensive to use (like most multislit spectrographs), so do not underestimate
the amount of work needed!
- The readout time is shockingly long for a detetcor this size, by modern
standards (several minutes). Factor this into your time allocation for
calibration and on-source exposure.
- COSMIC has some significant stray light and light-leak issues. You
will want the dome to be completely dark for your calibration frames. You
will also need to be careful about turning the cal lamps on and off.
- COSMIC has a pretty good
- You will find particularly useful the
lamp exposure time table.
1) Making the Slit Masks
- There are many places for the slit making process to go wrong, so you
need to check this at each step. If you arrive at the telescope and the
slit masks are bad, well, too bad, there's no way to fix them. If they are
bad, there are some pre-made single-slits for COSMIC, so at least you can
do one object at a time.
- When you print the masks on the printer, make one set with numbers,
and one set without. The one with numbers you will need as a reference
during the alignment stage (at the telescope), but the one without numbers
is the one you must give to the photo lab.
- When you print out the slits, get a ruler and make sure the box is
exactly 9.5 inches across. Some printers and software (like Acrobat) make
adjustments to the print size at the 2-3% level. This is not acceptable,
since the masks must be precision photoreductions from the printed
original. If you print from unix "lpr" you should be ok.
- Since the slit masks will be photoreductions of your printouts, make
sure the printouts are clean, with bold lines. I.e. good toner and no crap
on the rollers. Any flaw on the printout will be a flaw in mask.
- It takes about a week for the lab to make the masks. Sometimes they
screw it up. It would be best to get this job to them 2 weeks ahead of the
run. This means you should start the process of selecting targets and
making the mask files at least a month ahead of time.
- Here is an example postscript version of a mask without numbers, and one that is labeled with
numbers. Also, here is an example of a finder chart that was very useful for placing
the slits. You could make such a thing with skyview.
- When you lay out the slit masks, make sure there are at least a few
objects in the slits that you will be able to readily see in an (imaging)
exposure of a few minutes. You will need these to align the slit mask on
2) Mounting and Installing the Masks
You will probably need to start
mounting and installing the slits around noon, so that you will have time
to take all the needed calibration data before dark. The reason is that
you will need to take calibration data for every slit mask you intend to
use that night, and this can take a really long time, given the COSMIC
The stuff you will need to mount the slits. There should be two bags full
of machined slit mask holders waiting for you (the day crew will get them).
There will also be tape in the bags. Bring some liquid paper with you. The
lightbox is probably in the office off the control room. You'll also need
to borrow some screwdrivers from the office as well.
Trim the slits so that that they have a border like this. They must lie
flat in the holder, if you have to pinch them to fit them in your slit
alignment will be off. You will probably see some tiny pinholes in the mask. These
will show up as out-of-focus doughnuts in your spectrum. Using a toothpick
and some liquid paper, cover these up. There is a type of marker you can
get at a professional photography place to do the same job.
This is a slit mask holder. There are two types in the bag. You need the
ones that look like this (they notches and such are machined a little
different). Note that it is laid out as if you were about to put a slit
mask on it, with the notch closest to the mask at left. Place the mask such
that the writing is backwards and in the upper right corner, when the slit
holder is positioned as it is here. Try and line the box marks up as
closely as you can with the slit holder. After you've attached the cover,
use thin slices of tape to cover up the box marks along each side, as well
as the writing on the mask. When you are done, the only light shining
through should be from the slits.
Looking down from the prime focus elevator into the prime focus cage. You
will unzip the black shroud over COSMIC in order to place the slit holders
into the slit wheel. You do not need to remove the shroud. The blue box is
the COSMIC controller. You will use the inc and dec buttons to spin the
filter wheel. Note - tell the day crew you are heading up to prime focus.
Also, the interlocks on the elevator are pretty finicky - be sure to
solidly close all the gates, or they will lock out the telescope.
You place the slits into the slit wheel, using the blue box to advance or
reverse the wheel. Write down which holder is in which wheel position. The
holders are inserted and twisted slightly, they will snap into place. Some
of the holders are a bit sloppy in the wheel. Note the tape used to cover
the four edges of the slit mask to blot out the writing and alignment box. Also note the little bit of tape
used to secure the holder. There's very little clearance for the wheel,
though. You will need to leave at least one position open, and you
will probably want to leave another with a long-slit in, for standard
stars and such.
3) Basic Theory of Slit Alignment
This part was incredibly unclear to me from the instructions, so I will try
to explain it here.
The basic problem with using a spectrograph like COSMIC is that it is
impossible to assemble all the components (like the slit masks) with such
precision that they simply work when pointed at the sky. Instead, the masks
must be actively aligned after they are inserted. This is critical, since
if the targets aren't right down the slit you will lose some if not all of
your sensitivity. The basic steps are
- After you have inserted all of the slit masks into COSMIC, you
will turn on the cal lights, withdraw the dispersion grating, and take
images of each slit mask. Each FITS image will just be an image of the
mask. Note that this has to be done for each slit mask each time it is
inserted into the wheel. If you change masks around during the night,
or remove and put one back in, you must repeat this step. This
happens during the day.
- Then, you will run a piece of software that will step through the
slits. For each slit you will mark the center and each end. This will
generate a text file to be used later. You will need that printout of
the slits with numbers on them for this step.
- After you've opened and done the usual stuff (focus, pointing
check), you will slew to your target. Withdraw the dispersion grating,
and rotate the slit wheel to the open position (remember?). Now you can
take an image of the sky.
Now you'll need that finder
chart, labeled with the numbers of the slits. Using the provided IRAF
scripts, you will identify a few of those bright objects you included
on the slits. It will generate centroids, compare them to the known
slit locations (from the second step), and generate offsets and
- The offsets get sent to the telescope. The rotation is sent to
COSMIC - it has a rotator that will rotate the detector relative to the
sky. You will probably need to repeat this whole process again in order
to fine-tune the positioning. While this all sounds complicated, once
you have done it once or twice it is really easy. You'll waste much
more time waiting for the ccd to read out.