There has been progress with the second round of focuser prints, but I’m going to take a moment to sidetrack to the primary mirror linear bearings. This is for adjustment of primary mirrors to make sure the images merge into one stereo picture rather than two side-by-side. Up to this point, I’ve been planning to use a linear bearing made by 8020.net, something like this. By orienting the bearings on each primary perpendicular to one another, this would allow adjustment strictly along X and Y planes.
One of the things I liked about this solution though was that it allowed me to glue the primary to the mirror cell (which is perfectly acceptable for 8″ primaries), and that it allowed for a completely standard push/pull collimation configuration to control primary tilt.
The bad news was that they would cost over $100/each and add extra length to the mirror box. Also, they might impede airflow to the back of the mirrors, making them cool more slowly. Last, it’s hard to say how easy it would be to adjust the brake underneath the mirror without some kind of weird contraption. I also wondered whether the teflon smoothness would be a help or a hindrance in locking down the perfect spot.
More fundamentally, it was brought to my attention this is probably overkill for an adjustment that should only need to be done at first light, possibly very, very occasionally afterwards.
After a few days of pondering, I returned to the now-common theme of, what would it cost to 3d print? Could a custom design help? Here’s the result:
A big air hole aids cooling circulation. Little cups hold RTV while it’s being applied. (Additional ventilation holes may be added to assist in RTV drying time.) Big 1/4-20 bolts go through to the back to lock down the position from behind, like the rest of the collimation screws. Also, since the bolt is fixed in place in the bearing, you can grab the bolt from the back and use it to move the mirror. It’s half the height of the 8020 part. And less than half the price.
It still might be overkill, but it does now have some real advantages over both the 8020.net part and the standard binoscope method of using screws on the bottom of the mirror set at 90 degrees from one another.