Point the telescope to -35 dec on the meridian. Standing on the mezzanine, facing the back of the secondary, there are 3 bolts at the back of the secondary to adjust the collimation. The bolts are located at the 2 o'clock, 6 o'clock, and 10 o'clock positions. Adjust the bolts with the 5/16" ball driver, they are stiff but have never needed excessive force or allen key. At the moment they are pretty far in, so they can be backed out 2-3 full turns if necessary.
With the telescope outside of focus, looking at the guide tv: 6-in will move the star up 2-in will move the star to the upper-right 10-in will move the star to the upper left Move the star away from the bright limb of the defocussed comatic image to center the donut hole.
When collimated and focussed, stars in the center of the guide tv should be tight and round. Stars in the upper-left corner of the guide tv should be highly stretched out and pointing towards the center of the guide tv. Stars in the lower-left corner of the guide tv should be moderately stretched and pointing toward the center of the field. When going out of focus, both inside and outside, the donuts should be round, as opposed to astigmatic squashed donuts.
The goal should be a view of the pupil as in the image that follows (created by MC 11/05/07 off the guider screen using a defocused bright star).
Perry and Gil tweaked the collimation on 11/08/15, which resulted in a major improvement in image quality, as shown in the image that follows.
There are 2 basic steps to alignment: (1) align the laser with respect to the rotator plate, and (2) align the various optics to the laser.
Normally, one would define the optical axis to be the axis of rotation. It is thus best to have the axis of rotation be the axis of the cell. To line up the rotator then, do step (4), which would otherwise be skipped.
(1) Put laser in flange at cass rotator plate approximately centered and perpendicular to focal plane. The laser needs to have tilt as well as translation motions. Put flat mirror in rod-mount in hole of secondary mirror tilt the secondary using the three large outboard bolts till laser beam returns on itself. Other screws on secondary do not have to be loosened for tilting (only translation).
(2) Bring laser parallel to rotator axis via: Rotate plate, generally, the return beam describes a circle whose center is not at laser. Tilt secondary mirror till returned beam is at center of circle, tilt laser to return beam on itself. Repeat till tolerance achieved.
(3) Now look at beam on the secondary mirror (or anything) as rotator moves, translate laser till no movement is seen. Step 2 may need to be repeated. The laser is now on the rotator axis.
(4) Now the rotator itself must be aligned, in tilt and centration. This is done by adjusting the large black plate above the rotator. The large bolts must be loosened to do this. Centration and tilt defining bolts are used for adjustment. The idea is to bring the laser axis to the axis of the mirror cell, as defined by the central hub. Use crosshairs to define the center of the hub, running a wire around bolts at the top of the hub. A piece of paper should be placed at the top of the hub, too, to act as a projection screen for the laser. Rotate the rotator around and note the behavior of the laser beam. If the beam does not move, then translation of the plate is all that is needed. If the beam describes a circle, then do as above in (2), changing tilt and translation of the plate. The new rotator (9/97) appears to have uneven motion in the bearing, causing jumps when rotated. This affects all of the collimation described below. The effect is about 90arcsecs, or 2mm at the secondary.
Now the laser is on the defined axis and perpendicular to the rotator.
(5) Now recenter primary using the inner hub of the cell as a reference point. Micrometer measurements are used here, referring the mirror hole itself to the inner hub.
The next step is not normally done. Actually primary is set in tilt be referencing to bottom of mirror cell, using micrometer measurements of the glass-cell distance.
(6) Now bring primary tilt to laser axis. Put illuminated pinhole at radius of curvature of primary (outside of dome, we used an aluminum plate with holes in it, a flashlight or lamp, a large tripod, and a ladder). Move telescope so that laser falls on pinhole. Tilt primary till return pinhole matches laser and outgoing pinhole. As stated elsewhere, it is best to tilt the mirror at the zenith, then check the effect of the tilt. The position of the telescope was about -1h16m12s, -35d17m52s, the pinhole was located on the SW roof, 96 inches high, about 2.5 feet from the corner.
The hardpoints have locking screws that must be unlocked.
(7) Recenter secondary mirror, using a crosshair fixture mounted in the mirror hole. To translate, both the set of 6 screws as well as the three large screws holding the mirror to the focus ram must be loosened. Loosen the set screw and drive the positioning screws inward.
(8) Correct secondary tilt using a flat mirror attached to the secondary, look for return beam. Final adjustment should be made using a star.
(9) Place coma corrector cell in telescope. Usually, the internal reflections can be used to line up the cell in centration and tilt, though there are crosshairs and a mirror if needed.
(10) With an instrument mounted, observe a bright star in the TV. Tilt the secondary until the best images are on-axis.