[ATM] Re:A new kind of folded Newtonian

[Vladimir Sacek <vladis.2@juno.com>]

Surprisingly, there is no reference in the literature (that I know of)
giving specific relationship between the curvature on Newtonian's
diagonal and induced astigmatism.
Doesn't seem as much of a problem to come up with the longitudinal
astigmatism geometrically, and it is easy then to convert it into
wavefront error. What I get is not quite in accordance with the common
notion. For any given smooth error on the diagonal surface, induced
astigmatism wavefront error is only 70% as large (the result I posted
previously was in error probably due to doubled angles, or who knows
what).  And that compares to 2/3 as large spherical aberration error
contrast-wise. In other words, a 
1 wave curvature on the diagonal surface will result in 1/1.4 wave
astigmatism, which compares to 1/2.1 wave of spherical aberration. And it
is independent of focal ratio.

So, given smoothly distributed error, there is no difference between
amounts of astigmatism induced by a full aperture diagonal vs. regular
(sub-aperture) diagonal. The former will have much longer radius, and
smaller angles of deviation, but it is offsetted by the latter being
closer to the focal plane. Both will produce same amount of longitudinal
astigmatism. That should be a geometric consequence of  already mentioned
equality of wave retardation. It would be interesting to hear what others
have to say on the subject of Newtonian diagonal astigmatic properties.
Most of us have it as a part of optical system, yet it is so little known
about its effects.

Vlad

_______________________________________________
ATM mailing list http://www.atmlist.net/