(The Multi-Aperture Red Spectometer)

Kitt Peak National Observatory

National Optical Astronomy Observatory

Jim De Veny,  Instrument Support Group
Arjun Dey,  Instrument Scientist
MAY 20, 2002

Version 2a

Staff Contacts:  Daryl Willmarth ( and Arjun Dey (dey@

Table of Contents

Observing Programs That May Benefit From MARS

System Overview

  The MARS instrument  is a reincarnation of the old Cryogenic Camera, but now upgraded with the following: The instrument  is designed to obtain low-to medium-resolution (8-30 Angstroms) spectra of faint objects using a high throughput spectrograph and  CCD detector. Spectra (S/N~8-10 per resolution element) of objects of magnitude R~21 can be obtained in ~45 minutes of integration under good observing conditions.
    The system consists of the 4-Meter R.C. Spectrograph (see Figure 1) with the main MARS assembly mounted in place of the normal collimator mirror housing. The straight-through design employs a lens collimator and a "grism" (transmission grating replicated on a prism) preceding the camera. The camera consists of an evacuated f/1 classical Schmidt with a thick LBNL (Lawrence-Berkeley National Labs) 1980x800 pixel CCD (15x15 microns) at the focus. This chip is sensitive over the 3400-10,500A spectral region and has a readout noise of  ~8 electrons rms. Since this is a thick device, there is no fringing. The quantum efficiency of the chip peaks at ~92% at  8000-9000A. The overall system efficiency, including the atmosphere, telescope, and instrument peaks in the ~40% region with the new VG8050-450 grism (see  Figure 2).
    The normal slit of the spectrograph is replaced by an aperture plate assembly giving a coverage of 5-arcminutes. The system can be operated in two different modes: long-slit or multi-object. For long-slit observations, a collection of metal aperture plate slits of various widths is available. For multi-object work a user-designed slitlet mask is used.

Click here for the MARS optical diagram (Figure 1).

    Multi-slit masks are individual masks fabricated for each target field. They require substantial effort to prepare, since good astrometric prositions are required for each object. To maximize the number of objects that can be observed without overlapping their spectra, an ancient UNIX-Fortran77 program, MSLIT, designs the mask pattern*. With this observing mode coupled with the nod-and-shuffle technique, one can observe up to 30 objects simultaneously. Refer to the manual "Multi-Slits at Kitt Peak" which is available from the KPNO documents website  for details on design and construction. Please note that the multi-slit design software is being significantly updated with an anticipated release date of summer 2002

* Soon to be upgraded.

Exposure Time Calculator

Total System Throughput - The Optimum Case



DQE Curve for the MARS/LBNL CCD

   Click here to go back to the Table of Contents

Instrument Observing Configurations

    Three observing modes are currently supported:

Any of these modes can optionally use the "nod-and-shuffle" technique for accurate sky-subtraction.  See the section below for more details on using this technique.

Standard Long Slit Masks

Mask # Width- Arcsecs Width-mm Length-arcmins Length-mm
1 1.0 .150 4.5 41
2 1.7 .255 4.7 43
3 2.5 .375 4.4 40
4 2.5 spare .375
5 3.2 .480 4.4 40
6 3.17 spare .479
7 3.7 .555 4.2 38

Offset Slit Aperture Plates

    This set of slits are laterally offset from the normal on-axis position  and shift the central wavelength up to + - 500-600 Angstroms from the normal grism centers.  Slit lengths vary, so consult with your Instrument Assistant if the length is critical to your program. Nearly all of these slits are at least 4.4 arc-minutes in length.

Other Setup Aperture Plate Masks

    --- SETUP MASKS ---
    #19 Lynds Test line of holes, 32 holes, 380 micron diam. 
    #20 Knife-edge Hole 1.95mm diameter
    #21 Knife-edge Hole 2.05mm diameter
    --- TEST MASKS ---
    #22 Test Centroid 6 holes
    #23 Garth No.I alternating double hole pattern
    #24 Garth No.II another double hole pattern 
    #25 Fowler Test X-pattern centered at 45 deg., 200 micron holes
    #26 1 hole, 2.5" diam. centered
    #27 2 holes, 2.5" diam. (star/sky)
    #28 2.5" slit with 2" center occultation
    #29 2.5" slit with 4" center occultati

Imaging Mode

This observing mode is still in a somewhat experimental status.  The setup involves removing the grism and slitmask from the instrument. Note that the physical slitmask must be removed but the slitmask assembly (decker), which carries the field lens, must remain in the beam (i.e. the decker must be "IN") .   The FOV in this mode is a circle of diameter 5.2 arcminutes.  The instrument has two filter bolt slides (an upper and a lower) which can hold up to four filters each plus an open position.  There are actually four interchangeable physical bolts for each slide. One of the lower bolts has been modified so that two slots will hold 4x4-inch filters.  There is also a set of Bessel BVRI  filters (3.875x3.875 inches, lower bolt) dedicated to MARS use. Adapters for using 2x2-inch filters exist but the FOV will be reduced.

   Click here to go back to the Table of Contents

The Optical System

Atmospheric Dispersion Compensation -  The ADC  or Risley Prisms

A set of Risley Prisms is available for atmospheric dispersion correction at the RC focus of the 4-meter telescope. This ADC covers a field slightly larger than MARS field (~5 arcminutes) and is automatically controlled by the telescope control system. One inconvenience  caused by the Risleys is difficulty of locating guide stars.  The Telescope Operator must be aware that suitable guide stars must fall inside the annular area just outside the instrument FOV but inside the field of the prisms. Choosing a guide star in or too close to the MARS FOV will result in vignetting the instrument field.  The ADC is used at the option of the observer and may not be suitable for all types of observations.

Click here for a plot of the ADC throughput in the 5500-10,000A spectral region.

The Collimator Lens

The collimator optics consist of a re-imaging field lens, located just beneath the  aperture plate near the focal plane of the telescope, and a collimating doublet. The field lens is a simple plano-convex lens that images the pupil of the telescope onto the collimating doublet and also reduces vignetting  when the full 50mm spectrograph aperture is used.The plano-side of the lens goes upward (skyward).

The doublet collimating lens is a classical crown-flint cemented lens with the achromatic range centered at a wavelength of 0.63 microns. The clear aperture of the lens is 4.2 inches, and the effective focal length is 31.5inches.  The flint element faces the incoming beam. Both the field lens and collimator are anti-reflection coated. The flint element is not a good  UV transmitter and contributes to the poor throughput below 4000A.

The Grism Dispersive Element

The dispersive element consists of a transmission grating replicated on the hypotenuse exit face of a right angle prism. The grism follows the collimator and was chosen for this application because of its high efficiency. The purpose of the grism is to obtain dispersion without deviation at the center of the spectral range. The instrument performs as a direct-vision spectrograph with the ``zero" order of the grism falling outside the field of view of the detector. The grism is fixed and there is no provision for adjusting the central wavelength by tilting the assembly. Adjustments of a few hundred Angstroms are possible by using offset slits as described in an earlier section of this manual.
This instrument uses two types of grisms, a "conventional" grism, shown below, and a new volume-phase holographic grism (see NOAO Newsletter, 67, Sept. 2001, p 32). The new "VPH" grism employs a sinusodial varying index of refraction medium produced by holographic interference as the disperser.

The undeviated central wavelength (UDCW) depends upon the wedge-angle and refractive index of the prism, and the particular ruling used. When the vertex of the prism  equals the groove angle of the rulings,  maximum energy throughput is obtained.

Available Grisms

Grism Number 4950-400 5970-300 8010-300 8050-450** 9700-300 9700-300 7300-300
Old Grism Number 650 770 730 none 780-1 780-1 780-2
g/mm 400 300 300 450 300 300 300
Order 1 1 1 1 1 2 1
UDCW* - Angstroms 4950 5970 8010 8050 9700 4850 7300
Resolution - Angstroms@ 12 15 15 8 15 8 15
Dispersion - A/Pixel 3.2 4.3 4.3 2.0 4.3 2.2 4.3
Spectral Range - Angstroms ~4000-

* Undeviated central wavelength
@ With a 2.5" aperture
**   New Volume-Phase Holographic Grism, see NOAO Newsletter  #67, September 2001, p. 32. Has a built-in OG-550 filter.
Note(5/04): Grism 6560-300 (old#810) out-of-service

   Click here to go back to the Table of Contents

The Upper Filter Bolt

The upper filter bolt is normally reserved for up to four order separation filters. These filters are 3-1/2 inches square and are NOT interchangeable with the lower filter bolt.

Order Separation Filter List


Filter Thickness Filter Thickness
WG-345 1mm OG-515 3mm*
WG-360 2mm OG-530 3mm*
GG-375 3mm OG-550 3mm
GG-385 3mm OG-570 3mm
GG-400 3mm* OG-590 3mm*
GG-420 3mm RG-610 3mm*
GG-455 3mm RG-645 3mm
GG-475 3mm RG-695 3mm*
GG-495 3mm RG-830 3mm*
CuSO4 ** 8mm* KG-2 2mm
BG-38 2mm KG-3 2mm
BG-39 2mm Kopp 4-96 5mm
* Uncoated filter, all others are hard AR coated
** See the section below for special  information on this filter
These 3.5-inch square filters can be used in the upper filter slide ONLY

Filter Transmission Curves

   Click here to go back to the Table of Contents

The CCD Detector

The CCD is a Lawrence Berkeley National Labs 1980x800 chip with a readout noise of ~6-8 electrons rms..  The device is a thick chip (~300 microns) and, as a consequence,  has a significant cosmic-ray hit rate.  Exposures should be limited to 30-40 minutes to avoid excessive number of hits. The pixel size is 15x15 microns and the scale on the chip is ~0.86 arcsec/pixel. The chip operates at a gain of 2.0 electrons/adu. and the dark current is ~70 electrons/hour/pixel.  The current device shows no appreciable fringing.

Click here for a system DQE curve.

Click here for measured DQE values

Click here for a link to the LBNL CCD website.

Nod-and-Shuffle Observing Mode

The CCD can be operated in a standard single exposue mode or in a multiple-exposure "nod-and-shuffle" mode.  In the "nod-and-shuffle" mode,  the central one-third of the CCD is used as the active collection area while the top and bottom thirds are used as sky and target collection areas.  The telescope then nods between the target and sky fields while the charge is shuffled into the appropriate accumulation area. Note that this mode restricts the available field of view for either long slit or multislit observing modes, but results in excellent sky subtraction.  The following links provide additional information.


ICE Observing Commands

observe Take an object, zero, dark, flat, or comp image 
more  <n> Repeat the last 'observe' n times 
flpr Flush the process after a control-c abort
zero Take a series of zero  or bias images 
object Take a series of object images 
flat Take a series of flat images 
comp Take a series of comparison spectra images
dark Take a series of darks 
test Take a test image and overwrite previous test image 
detpars Set the detector parameters for CCD format and gain 
instrpars Set the instrument parameters & header keywords,
  such as grism, filters and DISPAXIS=1
obspars Set observing paramenter;  image root name and sequence number
ccdinfo Display current CCD format, gain, binning,  temperatures etc.
wfits Write FITS format data frames to tape 
tele Test TCS link for transmission of telescope header data

   Click here to go back to the Table of Contents


Recent Documentation:
"Report on the Nod-and-Shuffle T&E Run on the Mayall Telescope - 2000 March 15 and 16, UT", internal NOAO report
"Upgrading the Cryogenic Camera", NOAO Newsletter  65, March 2001, p35.
"The Multi-Aperture Red Spectrometer: CryoCam Resurrected!", NOAO Newsletter 67, Sept. 2001, p31.
"Update on the Performance of MARS", NOAO Newsletter 68, December 2001, p 25.

Older Dated Documentation:
"Multi-Slits at Kitt Peak: A Manual for Designing and Using Entrance Masks for Low/Moderate-Resolution Spectroscopy",  March 1996, KPNO documentation website

"A Setup and Reference Manual for the Kitt Peak Cryogenic Camera",  August 1998,  KPNO documentation website
Click here for the previous version (ver1) of this manual (no figures).