This is a very brief description of the Kast Double Spectrograph. For more complete information, please refer to Lick Observatory Technical Report No. 66, which may be available from your local reading room or from a fellow observer. 120" observers may obtain single copies for their own use from the Lick Observatory Publications Office (publof@ucolick.org).
All new users must be checked out by a resident astronomer on
their first night.
Please ask for this on your time request.
Please direct questions to Rem Stone (rem@ucolick.org) or Tony Misch (tony@ucolick.org).
The converging light from the telescope is focused at the slit of the slit-decker assembly, and then passes through three independent filter wheels known as the upper, lower, and user wheels. These components are in common for both sides of the spectrograph.
The beam then passes through a region where one of three possibilities may occur. The light can (1) strike a dichroic mirror which passes the long wavelength light (the "red side") and reflects the shorter wavelengths (the "blue side"), (2) the light may strike an aluminized mirror which reflects all the light to the blue side, or (3) nothing can be in the beam, which allows all of the light to go to the red side. These options may be selected remotely.
On the red side, after striking a silvered collimator, the beam is dispersed by a reflectance grating and then passes through a filter wheel that can accommodate four filters up to 5.5-in. in diameter, and then goes on to the camera-shutter-dewar assembly. Any of three positions in the grating carrier can be selected and tilted by remote control, and the user may have either three gratings, or two gratings and a flat mirror (for direct imaging), mounted in the carrier. The dewar is mounted on an x-y stage to allow positioning of the spectrum on the CCD.
The blue-side light path consists of an identical but aluminized rather than silvered collimator, followed by a filter slide capable of holding three filters up to 4.5-in. in size. The light then passes through one of three remotely selectable grisms (or no grism for direct work) and then to the camera- shutter-dewar assembly. This dewar is also mounted on an x-y stage, but since the Reticon 1200x400 CCD can record very nearly the entire useful spectral range produced by the camera, translating the dewar in the dispersion direction will not gain much. However, it is not uncommon to move the blue side stage in the x-direction in order to gain a hundred or two angstroms at one end or the other.
There are no manual dark slides. The shutters are quite close to the CCDs and are pretty light-tight with respect to the dome lights, at least.
The scale is very nearly the same on the two sides.
A very convenient aspect of the instrument is the ability to switch quickly between direct imaging and spectroscopic modes of operation. This capability, combined with a good telescope offset program, is sometimes crucial for centering the spectrograph on very faint objects.
Exposure lengths can be different for the two sides, and they must be started independently, so it could turn out (forexample), that several red-side exposures are taken and read out while the blue side is taking a single exposure.
Observations from both sides are numbered in a single sequence, but each image number is prefixed by a color code; e.g., b105.ccd and r106.ccd.
Spectra from both sides are displayed with blue on the left.
It is very simple to make on-screen plots from the CCD images. Quick-look data analysis is usually done with Vista.
| Scale at Cassegrain focus | 3.86 arcsec/mm | |
|---|---|---|
| Maximum slit opening | 145 arcsec | |
| Maximum decker opening | 145 arcsec | |
| Monochromatic beam size (both sides) | 82.6 mm | |
| Collimator focal length (both sides) | 1416 mm | |
| Camera focal length (both sides) | 188.7 mm | |
| Arcsecs/pixel (neglecting anamorphic mag) | 0.798 | |
| Approximate range of good performance | ||
| Blue lens | 3000-7000 Angstroms | |
| Red lens | 4000-11000 Angstroms | |
The following paragraphs follow the format
of the spectrograph
controller screen, which
is related to how the light flows through the
spectrograph
You may refer to a drawing of the Kast lightpath.
A variety of pre-programmed positions are available,
of which
the 2 arcsec (about 3 pixels) is most often used.
| Position | Upper | Lower | User's |
|---|---|---|---|
| 0 | open | open | open |
| 1 | -- | BG14++ | -- |
| 2 | Spinrad NS | OG570 | -- |
| 3 | ND 5.0 | ND 6.25 | -- |
| 4 | ND 1.25 | GG455 | -- |
| 5 | ND 7.5 | CuSO4 | n/a |
| 6 | ND 2.5 | GG385 | n/a |
| 7 | calcite | GG495 | n/a |
All three grisms are always present, and are remotely selectable.
| grism no. | A/pix | range | approx. coverage |
|---|---|---|---|
| 1 | 2.80 | 3200 | 3200-6200 |
| 2 | 1.85 | 2220 | 3225-5445 (match D55) |
| 3 | 1.13 | 1350 | 3195-4545 (match D46) |
Additional grism info:
| grism | grooves/mm | straight through (A) | blaze (A)* |
|---|---|---|---|
| 1 | 452 | 4780 | 3306 |
| 2 | 600 | 4340 | 4310 |
| 3 | 830 | 3880 | 3460 |
*To maximize efficiency in the UV, blaze wavelengths
are shorter than central wavelengths.
| Grating | grooves/blaze | E/pix | range | tilt* | useful range |
|---|---|---|---|---|---|
| 1 | 600/5000 | 2.35 | 2820 | 2.21(c)+5631 | 3800-10000 |
| 3** | 830/8460 | 1.70 | 2040 | 3.15(c)+5097 | 3800-10000 |
| 4 | 1200/5000 | 1.17 | 1404 | 4.72(c)+4178 | 3800-7310 |
| 5*** | 300/4230 | 4.59 | 5508 | 1.10(c)+5716 | 3800-11000 |
| 6*** | 300/7500 | 4.60 | 5520 | 1.09(c)+5742 | 3800-11000 |
*where c is the desired central wavelength in angstroms
**830/8460 grating in second order yields 0.78 A/pix, range is
1020A, and the second order central wavelength is given by
6.51(c)+4260. The max tilt of 35,400 restricts lambda < 5293A.
*** Note that the 300/7500 is more efficient than the 300/4230
for all lambda greater than ~5400 A.
There is provision for four round filters up to
5.5 inches in diameter, plus an open position.
| Position | Contents |
|---|---|
| 0 | open |
| 1 | GG 455 (3mm) |
| 2 | GG 495 (3mm) |
| 3 | OG 550 (3mm) |
| 4 | Spinrad NS |
Both sides use uv-flooded Reticon 1200x400 devices with 27 micron pixels, which corresponds to about 0.8 arcsec per pixel. These have excellent UV response (Q.E. at 3200A ~40%) and are relatively free of blemishes.
The CCD's must be kept cold to preserve the flood, and they should be protected from unnecessary exposure to bright lights.
The full well depth is in excess of 200,000, but the ADC saturates at about 32k minus the baseline, or usually about 27,000 DN. An easy to remember not-to-exceed number might be 25,000 DN. Gain on both sides is about 3.8 e-/DN with the attenuators in (default), or half that without the attenuators. The readout noise should be about 6 e-.
Suggested operating temperature for both sides is in the range -110 to -125. A warmer temperature will increase the noise but diminish charge transfer inefficiencies, and vice versa.
The chips read out fairly fast: about 15 secs for a spectrum.
The flexure situation is considerably better, but is not completely resolved. There is still about five pixels of total shift on each side, moving between extreme positions on the sky. The blue side may shift as much as five pixels parallel to dispersion, but fringing will not generally be a concern, and the shift may ordinarily be accounted for by reference to skylines. The red side may also shift as much as five pixels in the dispersion direction, and because of fringing this may be a more serious problem than on the blue side.
The usual red fringes start to appear at about 7000A. This may become a particular problem due to the flexure described just above, because if observations at large zenith distances are flattened with straight up flats, the object and flat fringes may not match. If this is a concern, you may wish to take "local" flats.
The x-y stages should in general not require adjustment. Both sides are set so as to put the image near the top of the chip in order to minimize readout time. The blue side useful range very nearly matches the chip size, and the red side wavelength region is chosen by tilting the grating, not by translating the chip.
Direct images may be taken on both sides.
The unvignetted field of view is about 145 arcsec square (180 pixels at ~0.8 arcsec/pixel).
An obvious difficulty is that most of our narrow band filters are 2" square and must go in the user's filter wheel. Their use in the red or blue side filter holders where the beam is 3.5" in diameter will reduce the effective aperture to about one meter; you may prefer to use the 40" and avoid the wrath of Practically Everybody...