Cassegrain Interface Unit
The light enters the fibres in the Cassegrain Interface Unit (Fig. 1). This is mounted on the standard SAAO 74" acquisition/guider box, which is used for target acquisition and off-set or direct guiding, using a fast readout CCD camera.
Star light is focussed onto one of two aperture plates, consisting of 0.45 mm diameter (2.7 arcsec) entrance apertures. The `star' plate is in fact a silvered metal mirror, which is tilted at to the optical axis. A field of 20 arcsec diameter, centred on the entrance hole, is imaged by a lens onto the input face of a 6 6 mm coherent optical bundle. This fibre bundle relays the image up to the acquisition/guider box, which can be viewed with the CCD camera, when it is driven to the off-set position X = 0, Y = -5700. This enables the target star to be correctly positioned onto the entrance hole. Once this is achieved, the telescope may be either guided manually (e.g. for short exposures on bright stars) on this image, or an off-set guide star acquired, all using the same camera.
After passing through the entrance hole, the f/18 star beam is converted to an f/2.7 beam by a focal reduction lens below the hole. This images the entrance hole (and stellar image) onto the input face of the 50 m optical fibre. The speed of the input beam is well matched to the numerical aperture of the fibre (N.A. = 0.22), which ideally accepts a beam slower than f/2.2.
Adjustments are made to both the (X, Y, Z) position of the fibre, and the tilts , to maximise the light throughput. This is usually done with an arc during the first afternoon of the set-up, and maybe checked on subsequent days.
A second, identical, entrance aperture, which is not viewed by the CCD camera, is displaced South from the `star' aperture by 90 arcsec. This is referred to as the `sky' hole, and can be used to obtain either simultaneous sky or arc spectra, using the second fibre. These spectra will appear interleaved with the `star' spectra on the CCD chip. This mode is probably only of use for faint stars and where precision radial velocities are required. Note that the extraction of the data is somewhat more complicated for this mode of operation.
The Cassegrain unit incorporates both arc and halogen flat field lamps. Light from the lamps is collimated and fed by a beam-splitting prism to a pair of lenses which focusses the light, via an insertable 45 mirror, onto one or other (or both) of the entrance holes. A second focal reduction lens, mounted on the movable mirror, further concentrates the light into the entrance hole(s). This is done so that arc and flat field exposure times can be kept to a minimum (typically a few seconds). Both a Thorium-Neon and Thorium-Argon arc are available, although usually the latter will be employed. Control of the lamps is through the Giraffe menu item in QUARTZ.
The three-position mirror is controlled using the QUARTZ program. Position 1 is for all normal program observing, while position 3 is for arc and fibre flat exposures. Position 2 is not useful at present. Although a well exposed spectrum using the Th-Ar arc can be achieved in 5 sec, the flat-field lamp is too bright and needs suitable attenuation with an ND filter. A circular variable filter is used, and at present a setting of ND 4.0 with a 10-second exposure gives well-exposed fibre flats with the blue filter on the telescope Cassegrain unit set to `IN'.