Mark,
You don't need to use an eyepiece at the front end. The fibre doesn't need a collimated beam entry and could handle say an f6 input cone; this would give about f3-f4 output.
There's some headaches getting the star image onto the fibre... Tom Kaye in his PAS article, p183 shows a solution using a hypo needle to hold the fibre and allow guiding on the "cut nose" of the needle; still pretty small but obviously doable.
|