[ATM] ATM: Collimating by concentric perimeters (first installment)

[scarab2 at cox.net (David Harbour)]

Mike,

You are 100% correct, here. You hit the nail on the head, so to speak. These
things that you mention are the first requirements for sound collimation. As
a rule, it is not hard to get the primary centered inside of a geometrically
"true" tube. And the secondary mirror assembly does have to be centered with
regard to the centerline of the primary mirror (its optical axis) and the
focuser's centerline must intersect the OTA's centerline. This is tricky,
when installing the secondary holder; that's why commercial ones have the
holder mounted on a threaded rod, so that the location of the secondary can
be tweaked forwards or backwards until it is actually centered on the
intersection of the optical tube assembly centerline and the focuser
centerline. Once these conditions are met, one may start collimation by
first squaring on the primary. I have no remedy for those who practice the
offset secondary thing, for making the field of view fully symmetrically
illuminated; this will prevent one from using the method of "concentricity"
to align all components. But you are right; the geometric fidelity of the
tube with its primary mirror mounted so its optical centerline (axis) is
congruent with the centerline of the tube will make the rest easy. If it isn
't, then one has to find a remedy for the lack of these geometric
considerations being true. The best way to "compensate" for them not being
true is to make sure that your tube and the mirrors mounted in it fit the
requirements initially; then one does not have to worry about finding a way
to compensate for one of these things not being true. In other words, one
should construct the tube or tube truss as truly as one can. Then the simple
method of "concentric reflections" will be easy to accomplish.

In order to keep this post not too overly long, I will put in your first
paragraph, here, Mike, and then with my next post on the subject, include
your next paragraph, and then with my last installment, I will include also
your last paragraph. You have understood this so well that your notes to me
should also be posted here. We will do it in installments so that each post
is not too long. You have noted that my large Warren truss square cross
section tube was built with extreme care, approaching the geometric fidelity
of perfection, assembled from parts made from templates that were laid out
to machinist's tolerances, and the tube was made to nearly machinists'
tolerances; it had to be, to be able to make collimating of its Cassegrain
components even possible. Some literature for Cassegrain reflectors comment
on this- that the tube, as a platform for the optics, has to have a high
fidelity of geometric accuracy. Same should apply to Newtonians, so that
they can be easily collimated.

The method of collimating by appearance of concentric perimeters for
components will not work for one who offsets his or her secondary mirror, as
is advocated in most literature. Offsetting the secondary mirror is a
practice of dubious value, and makes collimation more difficult. In my next
post I will describe the first steps in collimating the mirrors of a
Newtonian, assuming that one has built the tube "truly" and located both the
primary and the secondary "truly"; also I will put your second paragraph in.
By doing this in installments we can keep the post not overly large.

>Hey Davey,

>I revisited the basics of collimation, and noted that the fundamental
>issue is one of ensuring that the focuser is aligned relative to the
>OPTICAL PATH, of course, and not necessarily relative to the tube/truss,
>since, in many cases, the tube/truss itself cannot be used as a reliable
>frame of reference (except in a case like your yoke-mounted scope, where
>the truss is very carefully constructed!).  So the problem definition
>now becomes something like 'given the <optical axis> of the primary
>mirror, how does one then align the other components in the optical path
>with respect thereto?'.  The solution to this issue is of course
>applicable to OTAs with supposedly round/straight tubes/trusses as well...

(snip)

>Mike Byorick