The
preliminary design,
June 1998, by John Hart of MSSSO,
includes dewar and window and a TNG interface. Only one arm is drawn, although
the baseline is currently two. Click to enlarge. More recent
images are available in the Image Gallery
or on the project page.
In the two arm case, the question arises of whether to opt for a
convex or concave "V" arrangement of gratings. In both cases the
grating assembly is to be translated along the direction of
one of the gratings so that the path differences remain the same.
The diagram shows
(a) convex arrangement,
(b) concave arrangement,
(c) one arm arrangement created by sliding the grating assembly
(here a concave one) across the collimated beam.
In all cases the cameras are at the same position
although differently illuminated. Note that, contrary
to the impression given by the sketch, the diameter of
the camera lens cannot be made significantly smaller,
even if the instrument were permanently set up in two arm mode,
because the full collimated beam diameter still passes through
the optics.
The thumbnail at left
shows a proposed FOCUSING MECHANISM (click to enlarge).
John Hart wrote (13 May 1999):
The dewar is carried an a plate which is incorporates three equally spaced focus screws around its periphery. These screws are mounted in preloaded angular contact bearings, which provide excellent control (the type of bearing used in machine tool spindles). The thread of the screw engages with a thread in the fixed mounting plate. The clearance in this thread pair, combined with the short engagement length, allows it to accommodate considerable tilt in the focus plate.
Backlash between the screw and nut is eliminated by means of preloading. A preload nut is mounted in a recess under the fixed plate, and prevented from turning by means of a guide pin near its edge. A belleville washer (a slightly conical spring washer) is mounted between the fixed plate and the nut to provide a plenty of preload force.
The focus travel for a one microrad (0.2 arc sec) image blur is 0.013 mm for TNG and 0.016 mm for WHT. The pitch of the focus screws is 1 mm, and we engrave a scale housing bearing housing so that the position can be read to 0.01 mm. Because the focus plate is movable, there is a light leakage path into the spectrograph. To suppress this, and also to limit focus travel in the outward direction, a labyrinth ring (the one with the "Z" shaped section) is screwed to the focus ring and protrudes through the centre hole in the fixed plate. The gap between the fixed plate and the labyrinth ring is kept small, and the surfaces are painted matt black.
The travel we have allowed at the moment is 3 mm. We can make it more, but at the cost of some stiffness. The idea is that it would only be used for fine adjustment. Each dewar would have an adapter arranged to put the CCD in the nominally correct position.