The body of the collimator is made of two pieces of T-6061 aluminum tubing, the larger of 1.25" outer and 0.75" inner diameter and the inner of 0.75" outer and 0.5" inner. This is very hefty and solid aluminum tubing. I need a 1.25" outer diameter to fit in the telescope focuser. But the laser module has a 0.4" diameter (approximately, going from memory). So one needs to bridge the gap, and as I couldn't find laser tubing of 1.25" outer and 0.5" inner diameter (the last 0.1" is bridged with electrical tape, and one needs some wiggle room for adjustment), I had to use two. Plus, I only want the 0.5" diameter in the last 1.5" of length, so there is more room for the cross-hairs on the return beam. So nesting the tubing is the trick. Unfortunately, 0.75" outer diameter doesn't fit in 0.75" inner diameter, as the tubing is not made to tolerances that make telescoping possible. Moreover, aluminum on aluminum friction is very high--as high as rubber on dry cement.
Hence, one has to sand the larger tubing from the inside and the smaller tubing from the outside. I ended up sanding the larger tubing from the inside by attaching sandpaper (with duck tape) to the handle of a screwdriver, with some foam in there to add pressure, and spinning the contraption with a hand-drill to sand. I went through a bunch of sandpaper that way, but finally the small tube started sliding in. And then it stopped. Aluminum on aluminum friction is very high. I hammered it in a bit more, but an inch was left sticking out. Which is perfectly fine. There was not going to be any problem with sliding away, because of that amazing coefficient of friction. Anyway, so I had about an inch in and an inch out. Also, the outside of the 1.25" tubing needed sanding to fit inside the focuser tube (which was also aluminum, so I was afraid it would get stuck there if there wasn't some wiggle room).
I then cut a 45 degree slice out of the tubing. A little too close to the end--I should have done it past where the smaller tubing was. (As it was, to allow more room for the cross-hairs, I ended up expanding the hole with a Forstner bit.) I used two pieces of steel bolted around the tubing at 45 degrees as a guide for the hacksaw while cutting. Took a long time. Aluminum may be relatively soft, but cutting through such thick tubing at an angle still takes a while.
Next, I drilled three holes for adjustment screws--the collimator needs to be itself collimated. I then planned to tap it by driving #6 screws into it. One went in fine. Another broke as I was screwing it in--it just sheared its shaft. I managed to remove it. A third broke off flush with the tubing. That one was nasty. With a friend's help, we tried to remove it in various ways. Finally, a punch made a small hole, we drilled a ways in, and then I just drilled a new hole right through the screw shaft. My friend also told me that when one uses a screw to tap a hole, one needs to go two partial turns in, and one out, and so on, while I was foolishly screwing straight in. I tried that, and broke a second screw in that same shaft. Oops. Around then I realized I had drilled the holes one size too small, and that's why the screws were breaking. I sanded it flat again, drilled a slightly larger hole than before, and tapped it successfully. Whew.
The laser assembly is much easier than the housing. The laser is a <5mW unit, and I assume is safer than the typical unit as the crosshair grating diffuses the beam quite a bit. The laser comes with two nice wires. I attached a switch and a battery pack. The battery pack was made of three 1.5V LR44 batteries taped together, and some scrap pieces of thin metal for contacts for the wires touching them. The laser also has a focusing ring. (If one removes it all the way, one can remove the crosshair grating, and get a laser dot.) I focused it for roughly the length the beam would have in the scope, namely focal length times two, though it'll probably get defocused before I'm finished and I'll have to do it again. I used some heat shrink tubing (nice stuff) to make the connections more solid. And I transferred the warning label from the laser module to the housing. The batteries and switch will stick out of the back of the collimator. Not so pretty, but usable and easy.
Then, the bottom of the laser got wrapped in electrical tape to fit snugly in housing, adjustment screws were added, and the laser got aligned so its beam is parallel to the housing. The latter step is done by making two V-shaped supports out of nails, putting the laser in, and rotating it and adjusting its screws until the center of the cross projected on a wall a couple of meters away doesn't move beyond the desired tolerances.
There are two ways of collimating with a cross-laser. One way is to center the returning cross on a target in the 45-degree slit area, and the other way is by aligning the out-going and returning cross-marks on the secondary mirror. First I collimated a scope using the out-going and returning cross-marks, and I glued in a thick cardboard target with a hole in the middle (if it's too thin, the outgoing laser light peeks through). I then traced black lines where the returning beams hit the target. From then on, one can collimate the telescope by getting the laser cross-lines to align with the black lines on the target. There are photos here.