The MOS Environment

The MOS environment is composed of the following main elements:

Mask Slides

Changing mask slides is the only operation that the observers (or the OA) may have to do at night in the dome. The mask slides are introduced through a port in the bottom cover of the central octagon (see Chapter 5 for practical procedures). Presently, MOS slides can accomodate three 78 x 78mm mask-holders in addition to the open position (Figure 5). The mask slide is locked into position at each location to ensure that there is no lost motion or flexure.

FIGURE 5; Front View of the MOS Mask Slide

Filters

The tolerances for filters designed to fit in the MOS/SIS filter wheels are quite tight. If you intend to bring your own filters or have them fabricated for your observations, they should have the following specifications (CFHT standard):

Size: 76.2mm (3") diameter. 75mm round filters can also be accomodated if necessary.
Thickness: MUST not exceed 7mm.
Imaging quality: 1/4 wave or better, peak to valley over the surface.
Coating: Hard, anti-reflection coating on both surfaces preferred.
Blocking: 10-4 or better between the low cut-off of the filter and 3000Å and above the high cut-off of the filter and 11000Å.
Band width and central wavelength: specified for 0°C and a parallel beam.

Filters currently available at CFHT for use with MOS may be found on our filter web pages.

Grisms

The grisms mounted in the MOS/SIS cassettes have circular cross sections and 65mm diameters. The maximum practical thickness is 60mm (?). If you think that a new grism should be purchased for MOS, first check that it can be designed within the above specifications. The CFHT grisms available for MOS are listed on our Grism web page, along with the relevant parameters. Figure 6 shows the grism efficiencies vs. wavelength.

FIGURE 6; MOS Grism Efficiencies

CCD detectors

Currently, two CCD detectors are available for use with MOS: Loral 3 and EEV 1. Loral 3 has several drawbacks, including larger pixels, reduced sensitivity, and slower readout. The only significant drawback of EEV 1 is the relatively large amount of fringing beyond 7000Å. In practice, EEV 1 is recommended for most applications, unless weak line fluxes beyond 7000Å are desired. A more complete description of these CCDs is given in our CCD web pages.

Calibration Lamps

The system for introducing uniform illumination of wavelength and flat-field sources was designed by Y.P. Georgelin and G. Monnet. The calibration unit fits into one of the ports of the cassegrain bonnette and uses the rear surface of the existing central 45° mirror to introduce light onto the optical axis of the telescope. Either one of two spectral lamps or a flat field halogen lamp can be selected to illuminate a transmissive diffusing screen. The optical scheme uses two commercial 9.87 grooves/mm fresnel lenses as field lenses and one biconvex lens for a relay. Two symmetrically-mounted lamps are used in tandem for each set to ensure better than 5% uniformity in the blue and better than 3% in the red. The calibration system is controlled by the data acquisition and instrument control computer, independently from the MOS/SIS control system.

LAser MAchine (LAMA)

Mask Preparation

Once an image of a field has been acquired, the mask preparation is carried out with another HP terminal by starting a LAMA session (login as lama; ask your support astronomer for the current password).

To design a mask, you will display the field image and interactively superimpose the aperture contours on the objects of your choice. The details of the procedure are given in Chapter 5; we just note here that you need to select the size of your slitlets (for objects) and round apertures (for centering stars), then use the MOS and PAN icons to move the apertures from one object to the next, and to center them precisely.

The useful area for slitlets is less than the whole area of the CCD. A first limitation comes from the mask holder geometry: on a 2048 x 2048, 15µm CCD, the slit coordinates should lie in the range 105 < X < 1898 and 185 < Y < 1793. Then, unlike for long slit observations, multi-slit spectroscopy implies that each slit has a unique spectral domain. This domain is directly related to the slit position on the aperture mask. For a slit located at Yccd, and a CCD with Ymax pixels, (e.g. Ymax = 2048 for LORAL3), the spectral domain is defined as

where Lambda 0 is the zero deviation wavelength in Å and D is the dispersion in Å/pixel. One important consequence is that if your program requires a given wavelength range, you will be restricted to a fixed area in which to position slits. The LAMA mask-design tool assists you in this respect by representing the full spectral range for the grism of interest, making it clear if the spectrum of a particular object will fall within or off of the detector.

A few weeks before your observing run, CFHT will ask you how many blanks you will need for the entire run, in order to have them ready in advance. Try to estimate how many fields you could observe if everything goes well, then multiply by 1.5 for safety. For instance, if you plan to use 10 masks ask for 15 blanks.

[Contents] [Prev] [Next]

Send comments to: eugene@cfht.hawaii.edu