I used a router and a hand saw to make a simple secondary cell out of 1" square poplar rod. It will eventually attach it to a simple two-vane hacksaw blade spider (I used a cheap Harbor Freight rotary tool with a cut-off disk to grind the teeth off the blade), and it will adjust about two axes by direct rotation.
The mirror came with a plastic primary cell. Unfortunately, that would have required quite a bit shimming to fit in the 8" inner diameter tube, as the cell was rather small in diameter. So I made one. As I often do, I laid it out in
Inkscape. (The annotations are for the benefit of the blog--they weren't there in my original file.) I am very bad at measuring (but getting some cheap digital calipers for $8 from a Hong Kong seller on ebay helped a lot--and then I got $4 back as one of the jaws was chipped and a set screw was crooked, so it was a great deal) so my best way for precision is to print out a cutting template, attach it to the piece and drill and cut right through it, or at least make marks with a pencil or drill bit through it.
The plan was to use scrap material. The cell itself was to be a triangle with corners snipped off made of two layers of 1/2" Baltic Birch plywood, cut from the circular inner waste generated by routing the tube reinforcement rings. I cut the cell with a jig saw, not very precisely, and cut a ventilation hole with a hole saw. I then drilled and countersunk holes for collimation screws
in only one of the two triangles. I JB Weld'ed 2" long machine screws into their countersunk holes, and then I glued the other piece of plywood on top of this one (with Titebond II), covering up the screw heads, and clamping hard. As a result, the heads of the screws are trapped within what is basically a solid piece of plywood (two pieces of plywood glued together are basically one thicker piece of plywood!)
There would be a rectangular baseplate made from a piece of scrap red oak that I had. The baseplate would be mounted in the tube, near its bottom (not quite at bottom, as the plywood reinforcement ring on the tube would I rounded the corners of this scrap piece with the router using approximately the right diameter to fit in the tube, bearing in mind I could always sand it smaller (as I indeed had to) but couldn't make it larger. (By the way, here is a hint for how to measure things off from the router. I cut things with a 1/4" up-cut spiral bit. For measuring, I replace the spiral bit with a small length of 1/4" stainless steel rod, and measure from that with the calipers.) Notice from the diagram that the rectangle would be somewhat off-center in the tube--this was on purpose, to fit better with the collimation screws. I drilled holes for the collimation screws (actually, I did that while drilling them in the triangular cell, because they needed to line up exactly).
I also made a bit of a countersink around the holes, so that the collimation springs that would go on the screws (I had some springs that someone once sent me) would be countersunk, which would make the assembly have a lower profile. I didn't have a paddle bit of the right size for these countersinks, and I didn't think I could control my router for such small work, so I did something wacky. I chucked my 1/4" spiral cut bit into my cheap Harbor Freight drill guide, locked its height, and used a drill to route out the countersinks. It wasn't very neat, but it was good enough, and the lower speed was less scary. May not have been good for the bit, I already somewhat damaged the bit in an earlier episode when it was accidentally cutting against cement.
I also cut three ventilation holes in the baseplate with a hole saw. The result coincidentally (honest!) looks like a certain rodent from a company that protects its intellectual property with a zeal that one may think violates the Greek maxim
meden agan. In defense of this design, I will say that the placement was entirely functional. I made the inner hole as large as I could while keeping strength around the collimation screw holes, and the small holes are also located in such a way as to be fairly symmetric, and not too close to any of the alignment holes or the inner hole. Four wood screws hold the base plate to the tube.
The mirror is glued to the cell with three 3/4" blobs of silicone sealant, with the positions optimized with
Plop. I used some removable 1/4" particle board spacers to make sure the blobs wouldn't flatten out. One wants the mirror to float on the blobs, and not have stresses put into it by the differences between thermal expansion of wood and glass. By the way, when I removed the mirror from the original cell, I was horrified to see it was attached with two blobs of 1/16" thick rubbery stuff.
