ATM More Tri-space info.

[Roger Ceragioli <rogerc@dakotacom.net>]

Today I messed around some more with raytracing 8 and 10" tri-space apos.
What I seem to have found is this: 1) corrector triplets are relatively
insensitive to tilt errors (they produce coma at about 1/2 the rate of an
equal tilt in the objective), but are much more sensitive to decentration,
producing image flare.  I made this analysis on a 10" f/15 design; 2)
though you can produce an all-spherical design with excellent axial and
abaxial sharpness at f/15, it quickly becomes harder and harder at faster
f/ratios--after much labor I succeeded in designing a good f/10 that had
very decent performance on and off-axis, but it was hard for ZEMAX to find
the solution--3) I found two other glass combinations that could work for a
corrector triplet, but neither worked as well as the B58-53/KzFSN2/BaK1
combination; and 4) not surprisingly, it's much easier to make these
correctors work well if they fall at about 1/2 rather than 2/3 of the way
to focus.  My 10" f/10 was designed in that way.  Unfortunately, a
1/2-sized corrector will be much more expensive than a 1/3-sized.  What
Roland Christen fails to discuss in his 1985 S&T article is how off-axis
coma quickly becomes troublesome in f/ratios faster than f/15.  That's the
one drawback in these correctors.  Ironically, a standard achromat can
easily be made completely coma-free.  But there's that uncorrected
color....  Once again, there's no easy out.  It's all a choice of evils.

Roger Ceragioli