Re: ATM More Spheroidalage

[Brett Schaerer <schaerb@home.com>]

Anothony-
    I think that you, Texereau, and Couder are all right.  Raytracing a few of your example
spherical mirrors in OSLO LT, I come up with your numbers when I directly enter the mirror data
(i.e. at 587.6 nm, with an 80mm f/14 spherical mirror, I get a P-V wavefront error of .09677 , for a
750mm f/21.7 sphere, I get a P-V error of .24, which are essentially your wavefront error numbers).
When I adjust the focus (aoutofocus for best spot size, then tweak it to minimize P-V error), the
P-V error is greatly reduced (for the 80mm f/14, the P-V wavefront error becomes .02395  for the
750mm f/21.7, the P-V wavefront error is .06156).  If I input Texereau's example of a 6" f/8.2
mirror, I get a P-V wavefront error of .9285 before refocusing, and .247 after refocusing.
    As I understand it, before refocusing, the light from the center of the mirror is in perfect
focus, while that from the edge is blurred out to a large area (much more than that from the 70%
zone)  When the light from the 70% zone of the mirror is perfectly focused, the light from the
center and the edge are both blurred out by about the same amount.  The trick here is to find a
focal ratio, such that when the 70% zone is in focus, the blur from the center and edge fit inside
the airy disc.  So if you allow for refocusing (which is a reasonable assumption for a visual
telescope), then you can essentially find the best fit paraboloid of a spherical (or any other
shaped) mirror by eye as you're looking through it..
    Of course, this is assuming that you can make a perfect spheroid, which is probably why it's
been suggested to go with at least f/10 for a 6", to allow for a little bit of breathing room.
--
                                                                     -Brett