The collimation of this telescope is difficult mostly because the secondary mirror cells don't have position gauges (allowing control of what you are doing) and the f/13.5 cell is not kinematic at all but rather a complex puzzle where center and tilt are sometimes coupled on a single screw... So follow this guideline to avoid a nightmare. For your reference, the best images (FWHM) obtained for typical R-band 30sec exposures are: 0.60" at F/13.5 on-axis (classical Cassegrain) and 0.85" at F/7.5 over the FOV. (RC Cassegrain)

• By design and fabrication, and as our starting hypothesis, the primary mirror cell is such that it is centered with the instrument rotator. The primary mirror (M1) is centered in its cell through 4 radial adjustable pads inside the chimney (check exact procedure).

• Set the instrument rotator in the nominal position (30deg), install the 3 dial gauges in the back of the M1 cell (make sure you see the indicators move as they touch the mirror and you tighten them). The dial gauges fit in open holes of the cell at the same angle as the 3 collimation screws: East, Northwest and Southwest. The dial gauges, which have 3 readings each (tenths, hundredths and thousandths of an inch) are about 680mm off-axis thus a 1/1000" (smallest division) corresponds to a 7.7 arcsec tilt of M1. The collimation screws have 20 threads/inch so 1 turn moves the mirror up and down by 1.27mm. Located at an off-axis distance of 500mm, one turn of the screw thus corresponds to a 0.14deg tilt of M1.

• Install the K&E base plate on the rotator with the 3 short conic-head screws at 120deg (they provide an autocentering of the baseplate). Install the K&E in its support. Secure it with the 3 tightening screws on the left side of the support (the left one is used to open up the tube if the K&E is hard to remove). Through the window in the GAM, install the K&E target on the extremity of the optical tube. Make sure the target itself is not loose and slide at maximum on the tube. Orientate the target marks with the cardinal directions. Make sure the dials positionning the reticle of the K&E are set at 0. Never touch them later and check regularly that they are at 0.

• Install a light at the GAM window so that it shines on the target AND upward to the corrector (in the chimney), and another lamp at the top of the chimney shining upward to the secondary mirror.

F/7.5 : start with this focus, it is easier!

• The F/7.5 cell has 2 basic adjusments: radial (3 points at West, Northeast and Southeast) for centering, and longitudinal (3 points at East, Northwest and Southwest) for tilt. Two of the centering screws push directly on pads in contact with the secondary mirror (M2) and the third one (West) pushes on the mirror via a strong spring and has a locking nut. In thee back of the mirror, there are 3 hard points (screws with square head) that one needs to rise by a few mm when starting to tilt in order to open us space for M2 to go up, and tighten up when the collimation is done (use a 5 to 7/1000" shim to leave a tiny gap between these hard points and the glass for thermal expansion). The central hole in the mirror is fitted with a sleeve and a bolt which pulls up on the mirror to a system of counterweight supposed to hold the mirror figure against gravity. The 3 counterweights bars should go through the center of their support (if they don't, screw/unscrew the bottom nut holding the sleeve until they are centered). This has to be done at the end once the mirror tilt is set.

• Looking through the K&E, focus on the sleeve and the bolt/nut on the surface of M2. There is a tiny dot in the center of the bolt: this is the reference that needs to be centered with the reticle. The bolt diameter is 1/4" to give you the scale (you can also measure any decenter with the reticle dials but DON'T forget to reset them to 0 after measuring).

• The next step is very important: MAKE SURE that the K&E is perfectly perpendicular to its base plate (the centering has good locking screws but check it with a caliper, measuring the gap between the centration bolt blocks and the disk holding the K&E). Sight the bolt and make a zoomed drawing (direction and distance) of where the reticle is compared to the dot. Now, taking extreme care at not bumping the K&E in its baseplate (which has some play unfortunately -to be checked-), unscrew the 3 conic head bolts and rotate by 180 the baseplate onto the rotator (beware: you can not use the rotator to rotate because the bearing is not perfect and the rotation doesn't happen in a plane and thus your alignment will never converge!). Tighten up again and sight. Draw the new relative position. If the position of the reticle with respect to the dot has not changed before and after the 180deg rotation, it means the K&E is perpendicular to its support, so you can proceed. If it wasn't the case, determine the center of rotation (half way between the 2 reticle positions) and use the 3 tilt screws (with nut) of the K&E mount to set the reticle on that virtual rotation center. This is difficult because you need to maintain the wished position while you tighten the tilt screws. Once you have passed that tricky exercise, sight the bolt again, draw the respective positions of the dot and the reticle and rotate the baseplate of the K&E as described before. The new sights should show no difference. If this is not true yet, iterate the previous process.

• The K&E is now perpendicular to the rotator and thus defines the mechanical axis of the telescope onto which the optical axis of M1 and M2 must be brought. Use the radial screws in the M2 cell to center the mirror. Next, defocus slightly the bolt and you shall see the white K&E target imaged by M2: by tilting the mirror, center the image of the target with the reticle. The optical axis of the F/7.5 M2 coincides now with the mechanical axis.

• Drive the telescope to 45deg zenith distance in the 4 cardinal directions and sight each time at M2 with the K&E to detect any loss of alignment (sign that the mirror might not be held properly in its cell, i.e. not tight enough). Next flip to F/13.5.

F/13.5: the tough one!!

• The F/13.5 cell also has 2 'basic' adjusments:
  • Centering : 4 radial push-pull screws at the cardinal directions, each of them has a locking nut (but the East one), which need to be loosened first with a 25/32 tube key, and tightened at the end. These screws (use a 1/4" allen key) are accessible through holes in the wall of the Secondary Cage but you need to run the focus to the correct value to have these holes open up and be able to put the tools through them. These 4 screws are push-pulling on a belt attached to the mirror. Normally, you will be able to push one screw when the other 3 are loose (principle of the push-pull system). The problem of these centering screws is that they also introduce 'random' tilt and astigmatism if too tight!!

  • tilt : 3 outside screws driving pinions and teethed sectors at West, Northeast and Southeast. They require a special key that is a tube with 2 pins at one extremity and a central long allen key. First, locate the allen key in the lock screw of the pinion, loosen it, and rotate the tube until the pins enters the pinion. You can then rotate the tube and see the pinion rotates. The full range of each pinion corresponds to a tilt of about 5 arcmin (the star goes Northeast if you rotate the Northeast pinion to see the teethed sector go to the left). Tighten the lock screw at the end. Finally, there is also, inside the Secondary Cage (enters the tube standing up on the chimney), one push-pull screw rising up and down the entire cell mechanism from the south side thus allowing tilt in a quasi North-South direction (in theory this is in addition to the West pinion, also tilting the mirror in North-South). Use a special 1/8 long allen key for that adjustment.

• Once again, sight M2 through the K&E. You will see the large central hole in the glass (about 60mm in diameter). Center the hole with the K&E field of view (the 2 circles are quite close so this is easy and accurate typically within 0.5mm) or use the special aluminium target in the center of M2 (and center the tiny dot with the reticle). Next, defocus very slightly the glass and you will see the image of the corrector (down in the chimney just above the K&E) and especially its perimeter as a thin bright line just outside (very close) the hole of M2. Make that circle concentric with the hole in M2. The optical axis of the F/13.5 M2 coincides now with the mechanical axis. In principle, you ought to center first and then attempt tilting with the 4 dedicated tilt screws (but if you don't have enough range, use the tilt introduced by some centering screws and iterate...)

• Now comes the funny part (hey,hey the game is not over!): tighten the mirror without loosing the adjustments... Don't forget all the locking nuts (6) and make sure that the 4 radial screws for centering are all pushing hard against the belt (otherwise the mirror will move when you flip). With a bit of patience and practice, all that goes actually quite well. Always check any screw adjustment you are doing with someone looking in real time through the K&E (or do it alone but be ready for about 100 trips back and forth between the K&E and the mirror). Finally, check that there is no loss of alignment versus telescope position. Eventually, do a flip to check F/7.5, and then back to F/13.5.

• LAST STEP: bring the optical axis of M1 onto the mechanical axis (therefore also the optical axis of the M2s). In theory, you can use anyone of both foci to do that final adjusment (but whatever focus you use, it is highly rercommended to check the other focus later). It has to be done at night-time with direct imaging (look at through-focus sequences and 'zero the coma' on axis) or with the Hartmann screen. Use the tilt dial gauges of the M1 cell to monitor the collimation iterations.

IN PRACTICE, start with F/13.5. Tweak the tilt of M1 to get proper collimation at F/13.5. Then flip to F/7.5 and if the collimation is not adequate, use ONLY the adjustment screws (center and tilt) of M2 to get it right. DO NOT modify the tilt of M1 anymore obviously. The reason of that strategy is that you can't re-tweak the F/13.5 M2 at night-time while you can with the F/7.5 M2. I repeat: DO NOT in any case use the adjustment screws of F/13.5 M2 at night time because you will mess up everything.

Notes about gravity effects and mirror flips:

• After doing the K&E alignment, move the telescope 45deg off the zenith in one cardinal direction. Even when the mirror is tightly held in the cell, you will probably notice about a 1mm 'fall' of the mirror, probably due to a sag of the secondary cage. I find about the same fall amplitude at both focii and in all cardinal directions. Curiously the tilt seems to remain unchanged. This 1mm decenter is obviously a collimation limitation (at ZD45deg). This sag must somehow be related to the M2 focus encoder drift by about 30um when moving from zenith to ZD45deg (the drift was more like 100um before some maintenance was done to the focus drive mechanism in July 00). Actually, when the mirror is held tight in the cell so that it doesn't move on its own, adding to the cage sag, it is frequent to see trefoil and/or astigmatism, so that you need to loosen a bit the pressure, taking the risk of allowing decenter.
• Reading the axial dial gauges of M1 when going off-zenith shows that M1 is stable in its cell, within a 1-2/1000" at ZD45deg (7-14 arcsec tilt of M1).
• Doing repeated flips from f/7.5 to F/13.5 can alter the collimation (try it by sighting with the K&E), decentering the mirrors by up to 1mm, which is another collimation limitation. In order to minimize that effect, the flip has to be done by making sure both conical pins of the top ring are tightened slowly and equally.

Some useful notes for the Hartmann screen method:

• Use the model newha1.coo (116 holes) and Baldwin's shapzz.help
• Reduce the data on ctio60 (Solaris) with the following version of ximtool (type: /usr/local/combin/ximtool -gui /usr/local/combin/ximtool-alt.gui)
• Use binning 2x2 (faster), one amplifier so you access the axis properly
• Use star mag 9 (clear sky), defocus the telescope by about 8000 units (don't put spacers, the focus mechanism axis is sufficiently colinear with the mechanical axis)
• Angle convention of Shap: 0deg is South, 90deg is West....(this is with the normal Tek2K orientation, i.e. arcon box facing North). With that orientation in mind, lower M1 on the side indicated by the coma angle in order to decrease the coma (for example, coma angle is 270deg=East, lower=unscrew the East collimation of M1).
• Equivalence between wavefront error (1 micron) measured with Shap and the resulting image blur (EE80 in arcsec) for each aberrations: spherical 0.29", coma 0.37", astigmatism 0.88", trefoil 1.04", quadrafoil 1.12".
• Check out the focus thermal sensitivity on this graph

Maxime Boccas, 9sept00, last updated 19june01