User's Guide to the Kast Double Spectrograph

Table of Contents

Quick Reference
Hardware Overview
Common Path
Blue Side
Red Side
Detector Characteristics
Kast Controller
Data Taking System
Position Angle
Arc and Flat-field Lamps
Diagonal Mirror
Kast Focus
Telescope Offset
Setup and Observing Hints
Setup Procedures
Observing Hints
Calibration Lamp Spectra
Exposure Time Calculator
Data Quicklook

Data Archive
Mt. Hamilton Homepage

Common Path

Decker | Slit | User Filter Assembly | Polarimeter Module | Dichroics and Mirror


The decker is a long reflective plate which overlies the slit. The observer may select a variety of options including slits of varying length, rectangular blockers of varying widths, circular blockers of various diameters, a set of non-symmetric slots for bizarre sky subtraction schemes, or the decker may be withdrawn completely for unobstructed direct imaging. One positon called "the finger" places the end of the rectangular blocker just off of the slit (for slits 1" or less wide) for use as a reference for positioning objects at the center of the slit length (it's usually a convenience in the reduction if everything falls along the same rows).

Notice that there are two full length positions (1 and 10). Position 1 may be used for either spectroscopic or direct. Position 10 will allow a full length slit for "normal" slit sizes, but will be paritally obstructed by adjacent portions of the decker if the slit is fully open, as for a direct window. Position 10 is closer physically to the normally used finger (position 11), so if you are switching between large and small spectroscopic slits, you may wish to use these adjacent positions. In fact the decker moves very rapidly, so 1 will work just as well. But don't use 10 for directs. A full-length decker is 145 arcseconds, which is 333 pixels on either the red or blue side detector (both have 15 micron pixels). This corresponds, of course, to the full length of the slit itself.

The decker slide is presently configured as follows:

PositionNameEncoder numberNotes
1Open-4375For direct imaging
2Asymmetric slots 1-2942
3Asymmetric slots 2-2601
4Asymmetric slots 3-2259
5Asymmetric slots 4-1918
61 arcsec long-1398For slits up to 14" wide
72 arcsec long-1228For slits up to 14" wide
815 arcsec long-1058For slits up to 14" wide
930 arcsec long-885For slits up to 14" wide
1060 arcsec long-712For slits up to 14" wide
11Spect-170Full length, not for directs
12Finger-75End of blocker in slit for slits larger than 1" wide
132 arcsec rectangular blocker-29For slits up to 9.5" wide
144 arcsec rectangular blocker108For slits up to 9.5" wide
156 arcsec rectangular blocker217For slits up to 9.5" wide
168 arcsec rectangular blocker335For slits up to 9.5" wide
179 arcsec circular blocker1154*
186 arcsec circular blocker2348*
193 arcsec circular blocker3522*
*The occulting disks are alumininized spots on a quartz plate which is not AR coated. Occulting disks are designed for use in direct imaging mode and do not line up with the slit. To occult objects in spectroscopy mode, use the rectangular blocker.

Figure 1: Decker Occulting Spots (for Direct Imaging, not Spectroscopy), 3, 6, and 9 arcsec diameters.

Figure 2: Slit Length Settings (left) of Spect, 60, 30, 15, 2, and 1 arcsec long, and Assymetric Slots (right).

Figure 3: Finger (Rectangular Blocker) on left, Slit Length Settings (right).


The aluminized slit opens bilaterally, and the smallest available step is a tiny fraction of a pixel. For convenience, the most commonly used slit sizes may be selected from the spectrograph controller simply by selecting the size in arcsecs according to the table below. Alternatively, you may enter the desired slit width in arcseconds from the 'Other...' option in the Slit selection menu.

Minimum slit size is set to 0.4 arcsec. Remember that you need at least a two pixel slit to minimize aliasing problems with narrow emission lines.

Scales and Conversions
scale at slit = 3.86 arcsec/mm
1mm = 3.86 arcsec = 8.885 pixels (for 15 micron pixel size)
1 pixel = 0.43 arcsec = 0.1126 mm (for 15 micron pixel size)

PositionName Encoder Number
0Home 0
1Open -34344
20.5 arcsec 3989
31.0 arcsec 3857
41.5 arcsec 3724
52.0 arcsec 3591
62.5 arcsec 3459
73.0 arcsec 3326
84.0 arcsec 3061
95.0 arcsec 2796
109.0 arcsec 1734

Other... User enters desired slit size in arcsec

User Filter Assembly

There are three stacked filter wheels called, from top to bottom, the upper, lower, and user wheels. One position of each wheel (position 0) is always empty.

Filters in the top three wheels will affect both beams of the spectrograph. If you want to use different filters on the two sides, they must be installed in the holders in front of the cameras. Since these later filters are in the collimated beam they must be bigger than 3.5 inches in diameter in order to avoid vignetting for direct imaging (a 2" square filter will reduce the effective aperture to about 1-meter; if this is your plan, it would surely be more politic to ask for one-meter time in the first place).

Order separating filters should not be necessary in typical double-beam use, except in the far red.

Upper Filter Wheel
PositionName*Encoder number
0 Home 0
11 Open 1275
12 None 2299
13 Spinrad NS** 3323
14 ND5.0 4347
15 ND1.25 5371
16 ND7.5 6395
17 ND2.5 7419
18 calcite prism251
*ND values are in magnitudes.
**Spinrad night sky filter pass band is ~6100 - 7600 A

Lower Filter Wheel
PositionNameEncoder number
0 Home0
11 Open 1435
12 BG14++ 2459
13 OG570 3483
14 ND6.25 4507
15 GG455 5531
16 CuSO4* 6555
17 GG385 7569
18 GG495 411
*This is a good quality, full slit length crystal.

The user filter wheel accepts up to four filters of your choice mounted in our 2" square holders. Filter may be up to 8mm deep. Someone will definitely need to show you how to mount the filters on the first occasion. Please have a staff member (telescope technician or support astronomer) update the Kast motor control software with the current filters. Please remove your filters at the conclusion of your run (and have a staff member update the motor software accordingly). It's definitely safest not to assume that if the motor control software says the wheel is empty that there are no filters in place. We suggest always setting this wheel to the open position if you're not using it, just in case.

We have a fairly large library of narrow and intermediate band interference filters available, mostly on loan from astronomers within the UC system. It's clearly best not to assume that any specific filter will be here when you need it, as the owners are certainly free to remove them at will, and they may on occasion be loaned out. Since they're not really ours, we don't really control them. Any arrangement to take them elsewhere should be made with the owners, and duly recorded on the mountain as well.

User Filter Wheel
PositionNameContentsEncoder number
0 Home N/A 0
0 Open Empty 1178
(user determined) 2816
(user determined) 4454
(user determined) 6093
(user determined) 7731

Polarimeter Module

Instead of the User Filter Assembly (described above) a Polarimeter Module can be installed (the current default Kast setup has the polarimeter installed instead of the user filter assembly). The polarimeter module contains two filter wheels and a waveplate. The polarimeter module will affect both beams of the spectrograph.

The upper and lower filter wheels accept up to four filters mounted in our 2" square holders. Filters may be up to 8mm deep. The lower filter wheel usually has B, V, R, and I filters installed, though this may not always be the case. If you install other filters, please remove them at the conclusion of your run. We suggest always setting these wheels to the open position if you're not using them. The tables below list the usual contents of the two wheels, though if wish to use them you should double check the contents.

Polarimeter Upper Filter Wheel
PositionNameEncoder number
1 Open 2319
2 Filter681
3 Polaroid7234
4 None 5596
5 Empty 3957

Polarimeter Lower Filter Wheel
PositionContentsStep number
1 Open 5914
2 B 4275
3 V 2637
4 R 998
5 I 7552

The waveplate can be moved in and out of the light path and rotated via the Kast motor control software.

Waveplate Insertion and Rotation
PositionNameEncoder numberNotes
Out Out N/A Waveplate out position
0 Home0
1 0.0 deg 485
2 22.5 deg 673
345.0 deg 860
467.5 deg 1084
590.0 deg 1235
6112.5 deg 1423
7135.0 deg 1610
8157.5 deg 1798
9180.0 deg 1985
10202.5 deg 2173
11225.0 deg 2360
12247.5.0 deg 2548
13270.0 deg 2735
14292.5 deg 2923
15315.0 deg 110
16337.5 deg 298

Dichroics and Mirror

There are two separate carriers, either one of which (or neither, but not both) may be in the beam. Either carrier may contain one of the overcoated dichroics, or an aluminized flat mirror. The carrier called 1 (or X) is the one farthest away from the access door, and moves in an E-W sense with the TUB at its standard (90 degree) position angle (or left-right as seen from the access door); number 2 (or Y) is closer to the door and moves N-S (or toward and away from the door).

Crossover for the blue dichroic (D46) is about 4650 A, and for the red dichroic (D55) it is about 5500 A. The crossover for the new D46 is about 4600 A and for the D57 at 5700 A. About 200 angstroms of the spectrum are affected by the crossover of the dichroic. The new D46 will replace the original D46 in the 2017A semester. The old D46 will remain available by special request for observers with long term programs that desire no changes in hardware. The new dichroics have superior transmittance/reflectance and more distinct crossovers than the original D46 and D55. However, the new dichroics have brighter ghost images 45 pixels (and additional fainter ghosts about 90 and 135 pixels away) from the main trace than the old dichroics.

Insertion or removal of the D46 dichroic shifts the red spectrum by less than one pixel, but the D55 dichroic shifts the spectrum by about 10 pixels, compared to when no dichroic is in place.

If you're only using one side at a time, in order to switch sides move the mirror in for blue or out for red.

To get Select Side illuminated
clear Clear red
dichroic D46 or D55 or D57 both
mirror Mirror blue

Dichroic Transmittance and Reflectance Plots

Figure 1: New D4600 Reflectance-Transmittance Plot. Note that the reflectance greater than 1 is because it is a relative measurement compared to the flat mirror's aluminum coating.

Figure 2: D5700 Reflectance-Transmittance Plot. Note that the reflectance greater than 1 is because it is a relative measurement compared to the flat mirror's aluminum coating.

Figure 3: D5500 Reflectance-Transmittance Plot.

Figure 4: Old D4600 Reflectance-Transmittance Plot. Scheduled to be retired 2017A semester.

Support Astronomers (
Last modified: Sat Mar 6 16:21:05 PST 2021