1.0 Abstract: The question of collimation came up during the setting of the specifications for the instrument.
2.0 The drawing set which will be used to build DEIMOS will contain approximately 2500 drawings. The specific drawings that deal with collimation or alignment are:
3.0 Jumping right in:
3.1 There are 2 slit masks which are positioned in the curved focal plane.
These do not extend to the telescope's Elevation Axis. At the center of this curved focal plane where it intersects the instrument's rotation axis will be a 0.75" diameter tooling ball. This will be the first thing built. All distances will be measured relative to this front reference ball. This will be removable so that there will be a clear line of sight through the instrument. A cross hair module can be installed in place of the tooling ball for collimation.
The slit-masks will be held by a rigid frame. This frame has 3 tooling balls which locate in 3 kinematic mounts: a cone, a groove which points at the cone, and a flat. Three locking clamps will keep the balls in the 3 mounts. Each of the kinematic mounts has a motor so that the slit mask, once locked in, may be moved to align it with the sky. When the slit mask frame has been fitted with mask material (0.003" thick stainless steel) it will be held in three similar kinematic mounts on the laser engraving machine. This machine will now have a common reference with the DEIMOS instrument while making the slitlets.
An alignment slit-mask frame will be built. This will serve to initially align the three kinematic mounts. There will be two sets of these mounts for the dual sided spectrograph and so there will be a second alignment frame which will be used to insure the two sides of the spectrograph are 180 apart. There will be tooling balls for doing this. We will also be able to take advantage of the fact that the rotating axis is horizontal: we can use a machinist's level to locate the second side of the spectrograph once the first side is lined up. (This would not be possible if DEIMOS were assembled with its axis vertical.) The accuracy of a machinist's level is 0.001" per foot (17 arc-sec) and this is good enough.
4.0 The first problem is the hole in the center of the Collimator Mirror:
A fixture will be built to locate a tooling ball at this mirror's vertex. The distance from the slit ball and this ball will be 86.5" center-to-center. This will be easy to measure. The translation in the other 2 directions will be done by rotating the instrument in its bearings. (This will have been done as step number 1.) The rod used for measuring this distance will be stainless steel and will be kept in a long wooden box.
The 2 angular adjustments of this mirror will be harder to adjust. A dial indicator can be positioned to measure the runout in the z direction of the mirror right next to the hole. We plan to have a 4" diameter central hole. There will be an area approximately 1" wide around the hole which will not be used by the specified FOV. Typical "Last Word" indicators will easily see 0.001" runnout. This is 1 minute of arc and meets the spec.
Re-coating: After removing and replacing this optic, the alignment procedure can be repeated. The fixtures and tooling must be stored in a safe place.
5.0 The next problem is the Tent Mirrors:
These 2 mirrors are flat. They divide the spectrograph into 2 sides. There will be a nominal 1/4" gap between them.
It will be easy to locate these mirrors along the instrument centerline by measuring from the slit tooling ball. This distance will be defined in the optical layout. The angle will also be easy to measure in the plane of the optical layout. What is difficult to measure is the radial distance around the spectrograph's rotation axis. That is, they must be 180 apart and aligned with the slit-mask frames, which were previously aligned. We plan to use the machinist's level here, too.
We will reach down from the 3 slit-mask kinematic mounts on each side to the upper corners of the mirrors.
These mirrors must have hard stops in the tip and tilt stages. These stages are the ones used as flexure correctors. The initial collimation should put these mirrors in the center of travel of the tilt and tip mechanisms.
6.0 The grating slide:
This slide is tolerant. Both sides must be the same so that grating cells can be interchanged. The gratings in their cells must be aligned more carefully. There will be an alignment fixture which we will call a "dummy grating". It will be a cell with a tooling ball. This can be used for spacing to the camera lens cell and back from the tent mirror.
The linear slide must be parallel with the slide on the other side of the spectrograph. They must both be parallel with the 2 tent mirrors. The slides can be surveyed to be the same distance from the slit ball in the direction along the instrument rotation axis (z axis).
6.1 Grating alignment fixture: This device establishes the front surface of the ruled grating at the proper location so that the pivot axis of the sub-cell is parallel to the rulings. All sub-cells must be the same so that future gratings can be aligned and used. This work will be done by technicians at UCSC.
6.2 The flat mirror:
This mirror travels with the grating slide. It is used for direct imaging and thus the slit-mask will be removed when this is in place. This mirror maps the FOV onto the CCD mosaic. It must be repeatable, of course. It has an outline drawn on it to help focus things. The CCD will move to one of the focus limits or else a thinner filter will be inserted so that the CCD looks at an out-of-focus image of the primary mirror. This mirror is at the exit pupil formed by the collimator mirror.
7.0 The camera lens:
The lens cell assembly will be aligned internally. A target will be installed in both ends of the cell for aligning with the flat mirror in the grating slide. The slit mask used for alignment can have a small hole at the FOV center. We will define the 5 arc-minute radius as the center of the FOV. We know where the center of the beam going through this hole will hit the collimator mirror. We can place a small dot in the center of this 6" diameter beam on the collimator mirror. Looking back through the camera cell assembly it will be possible to align the cell with these marks (the slit mask hole and black dot). We will be looking at the flat mirror in the grating slide and the tent mirror. The camera lens cell will then be positioned along its axis by measuring the spacing between the cell and the flat mirror. The lens cell will now be pinned.
The pins are now removed.
The lens elements can be installed into the lens cell in a clean area. Covers will be installed on both ends of the cell and it will be returned to the spectrograph and the locating pins re-installed.
8.0 The shutter and filter slide:
These will be mechanically aligned by mating surfaces. The rear of the cell will have a machined pocket which fits a raised boss on the top of the shutter. The filter slide has a similar locating fit into the shutter. There are no collimating adjustments for these 2 items.
As new filters are built, they must be made the correct thickness (0.50") to the specified tolerance (0.005").
The slide axis of the filter slide has no critical alignment spec.
9.0 The dewar:
The dewar is mounted to a focus stage and course alignment will get it to 0.020". This should put the CCD mosaic at the mid-point of the focus travel. The tip and tilt of the dewar will be done only after the spectrograph is finished and we can take and evaluate images of the slit mask with all the holes in it. This mask will have a grid of holes spaced on 0.25 inch centers. The holes in the grid will have 182 micron diameters (0.007"), every 10th hole (in both directions) will be twice this size (0.014" Diameter). A 182 micron hole represents 0.25 arcsec on the sky.
9.1 The CCD Mosaics:
Each dewar will have a 2 x 4 mosaic of CCD's. The warm adjustment will be made with adjusters inside the vacuum chamber. These adjustments will be made on a milling machine base using a microscope with a depth of focus of 0.001". Once the 8 CCD's are within 0.002" of being in the same flat plane, the vacuum is sealed and the mosaic is cooled with liquid nitrogen. Any changes that happen during cooling and evacuating can be corrected by external adjustments. There will be 3 for each CCD in the mosaic. This is 48 total adjustments (for the dual-sided spectrograph). The range of each one of these is 0.002" and the smallest motion is 0.0001" (about 2 microns or 0.1 pixel).
10.0 The TV guiding system:
There are 2 TV cameras. Each has an alignment problem.
11.0 Calibration light sources:
These will need no special alignment fixtures.
This will not be too much trouble.
This location is to 0.5" and is easy.