From: Mark Holm 
To: atm{at}shore.net
Reply-To: Mark Holm 


Richard Schwartz wrote:


> The trick here is that for a given star only a tiny spot of the diagonal is
> used.   Surface errors beyond the boundaries of that spot are meaningless.
> So the diagonal can be badly warped, but will still function well if it has
> a smooth curve, as was shown by Anthony Stillman.   I am looking at having a
> minimum obstruction scope with the diagonal just far enough from focus to
> avoid seeing the dust on it.   A relay lens or gregorian secondary will
> re-form the image where it can be accessed by eyepieces.
>
>


Richard,

I can't agree.  Let's try a concrete example:  A 200mm diameter mirror of
1200 mm focal length (f/6) with the Newtonian focus 150 mm to the side (100
mm for the radius of the mirror, 25 mm tube clearance, 25 mm focuser
height.)

Assuming for the moment that a single star image is a perfect point (close
enough for the argument).

The diameter of the illuminated "circle" on the diagonal
(actually the minor axis of the illuminated ellipse) will be given by

150/1200 * 200 = 25 mm

The algebra comes from simple proportions because the illuminated
"diameter" is simply the result of applying similar triangles.

Clearly the full aperture of a typical diagonal isn't used to form the
image of a single star, or even two closely placed stars, so, for fine
detail resolution,
one need only consider the error over a portion of the diagonal, but it
isn't a tiny portion, it is a fairly sizable amount like 3/4 or 2/3.

Mark Holm
mdholm{at}telerama.com

--- BBBS/NT v4.00 MP