I thought this post might be helpful to anyone experiencing the same issue as me.
I recently got frustrated with the fact that my C8 produces nowhere near collimated images when the original cheap "Celestron model #94115-A" diagonal in it (ca 1999 vintage), unless it's been collimated with the diagonal mounted at exactly that same angle.
Collimating with the eyepiece in the visual back means that the image in the diagonal is always out of collimation. Collimating with the diagonal in means the image is fine if you don't touch anything, but then it's even more out of collimation if the diagonal is rotated.
I always figured this was due to some eccentricity of the clamp screws, of the visual back, or an angular miscollimation of the diagonal. But today I read on CN that angular miscollimation of the diagonal causes mostly focal plane tilt, not collimation issues (since it's so close to the focal plane it more tilts the wavefront than moves the coma-free center of the image). Someone else (https://www.cloudyni...l/#entry5465272) also explained it in a way that clicked: the reflecting surface can be displaced, not just tilted. That's what really moves the center of the image away from the center of the eyepiece. And I knew there was eccentricity between the eyepiece and the reflected image of the OTA's axis because when I rotate the diagonal in the visual back, the image rotates around a point not even in the FOV with a 12mm/60deg EP.
Ok, so what's up with this diagonal then? Well, I actually finally looked at the thing critically!
Oh. My. God. What is going on here?!? No wonder the darn thing is rotation sensitive! The reflected image must be translated by at least 5-6mm!
Ok, can this be fixed? People talked about padding/shimming prisms and mirrors to fix these sorts of things.Carefully opening the thing up (and making sure to only touch the edges of the prism, and that with gloved hands) I found that the prism is spring-loaded inwards.
Well, ok: which direction to push it? Working it out, the prism needs to move inwards. Dang how's that supposed to work? This is the direction the spring's already pushing it! Well turns out the thing was mis-assembled or assembled with faulty parts. Both of the threaded ends of the nosepiece and the (other thing?) protrude well into the plastic body, as seen below. That's apparently not where the prism is supposed to go: after screwing around with things, I determined it's supposed to rest on the plastic housing.
With the two threaded interfaces unscrewed far enough that they weren't pushing on the prism anymore, the image looks like this:
Better! But still not perfect, and I saw that there was a burr raised in the plastic from the threading:
Cut that out with a utility knife blade (in progress above). Now the prism can seat flat. I also noticed that the set screws which hold the prism from rattling around from the sides were super tight. In fact you can see that the pressure actually spalled the glass near one of them:
Having allowed the prism to move inwards and getting it as good as I could, I now couldn't leave well enough alone! While I was definitely planning on tightening those set screws, I wanted the spring pressure higher. Now that the diagonal had moved the spring was more relaxed and applying less pressure. I tried to bend it to increase the pressure...
Hot melt glue! My favorite low quality adhesive! Appropriate here, I think, because I think I may need to get back in there at some point. I may choose to add set screws so that the threaded interfaces can be adjusted but then locked in place, to push the prism around in there in an adjustable way.
Edited by InductorMan, 07 March 2021 - 10:19 PM.