ATM few comments

[Bratislav Curcic <epabcc@epa.ericsson.se>]

I just have to comment on two statements I've seen in some of the 
postings I read. 

Marinus van der Lugt <lugt@space-elec.dofn.de> writes :

> (Note: only a star on the optical axis of a
> collimated scope using a non-spherical mirror will have no coma.)

This can be read also as :
"only spherical mirrors are free of coma"

This is a VERY common misconception. Optical aberrations in the system
are defined by both 'active' elements (a primary mirror in case of
Newtonian) and not so obvious 'passive' like aperture stop. In 99% of
newtonians the aperture stop is equal to primary's edge, i.e. entrance
pupil is defined by primary size. A Newtonian with spherical primary
built in such way is NOT free of coma. In fact, it will have almost
exactly the same amount of coma as equivalent scope with a paraboloidal
primary. Only and ONLY if the aperture stop is placed at radius of
curvature we achieve the necessary symmetry (so called 'offence against
sine condition' criteria is met) to get system free of coma. But fact
is, if we place the same stop ar ROC of paraboloidal mirror, we will
get rid of most of the coma from that one as well. Of course, because
paraboloidal shape isn't sompletely symmetrical, we'll have some
residual abberations, but they will be greatly reduced.

Mark Cowan <cowanm@open.org> wrote :

> It seems some people figure the errors in the secondary as equivalent to
> errors in the primary.
> This is wrong.  Texereau states (briefly) in the first few pages of his
> chapter on diagonals that defects in the secondary have an effect
> dependent on the ratio of the distance from the focus to the secondary
> over the primary.  In my 8" f/5.75 that ratio is 7 to 1, so a 1/20 wave
> secondary *ought* to be as good as a 1/140 wave primary.  Show me the
> reference flat that could test *that* sucker! 
> 
> Anyway, Mel stated he's never had a *bad* secondary, and cited the same
> kind of ratio.  Since we all hold Mel in almost as high regard as Tex,
> although perhaps for different reasons :), it's probably true.

The truth is, a quarter wave aberration is a quarter wave aberration,
NO MATTER WHERE WE INTRODUCE IT. The flatness of Newtonian's diagonal
is very critical. Mel may have been lucky, others (me included) haven't.
First, lets get some things clear : it is FAR easier to get 2" piece of
optics not to depart 1/10 wave or whatever than a 20" one.
Secondaries are mass produced on planetary polishers, and it is quite
common to let the machine run for some time, simply break the mold,
test/categorize the secondaries into 1/20, 1/10, 1/8, 1/5 wave or whatever,
aluminize and sell. That's why secondaries are so cheap.
Now back to secondary flatness requirement. The '7 to 1' is another
misconception, and very dangerous one indeed. Get a 1 wave convex
secondary (should be good to 1/7 wave, shouldn't it ?) and have a look.
I bet you haven't seen such astigmatism in your life !
It all really comes down to a proportion of the active (used) surface. The
secondary must not depart from ideal plane more than 1/8 wave (or whatever
is your requirement) over the area used to form one diffraction image. 

[BTW this dimension can be found by dividing the BFL (= secondary to focal
plane) with primary's F-ratio; 
in your case this is about 6.6/5.75 = 1.14" - so if your secondary is
_indeed_ a 1/20 wave one, and it is say 1.8" m.a., assuming uniform
sphericity (worst case) it will be good to 1/20 / (1.8/1.14) ~1/30 wave 
over the used area]

If your secondary is a 'minimal' one, it WILL have to be at least as
good as your primary. Period. If you don't use 100% of secondary all
the time (i.e. it is larger, to allow for some unvignetted field - as
in great majority of scopes out there) it can be a BIT worse (so 1/10 
secondaries suffice for most scopes). But again, not all '1/10 wave'
secondaries are the same ... so beware.

Bratislav