SV: ATM A cage as secondary holder

["Nils Olof Carlin" <nilsolof.carlin@swipnet.se>]

Andy S. said 
> ................All the
> optical elements are made adjustable in several degrees of
> freedom for optical collimation, this means that the final
> optical axis may end up not being orthogonal to the alt/dec
> axis after collimation tweaking. Getting the OA perpendicular
> to dec/alt is not a trivial exercise to begin with so I've
> been thinking of ways to do collimation that disturbs the
> OA the least possible amount. 

If you do the collimation, whichever method you chose, the optical axis
will be defined by the optical center of the main mirror (fixed), and
the optical center of the secondary (the true one, or any other point
that you chose for centering collimation). Thus, moving the secondary
can only be useful to compensate for e.g a truss tube that cannot,
after mounting, give a reproducible position of the secondary - but I
am sure such a problem should be solved in a more direct way.

The main mirror should not, for the same reason, be allowed to move
sideways.
Compensating for small angular errors of the main mirror should be done
the usual way.

> One way to do this is to
> collimate the primary/secondary initially and then do the OA
> orthogonality calibration.

I agree

> Any subsequent collimation would
> be done by moving the entire secondary cage normal to the 
> axis of the primary (rather than tilting the primary). If
> the primary mounting is sufficiently rigid the collimation
> tweaking should result in the OA being returned to its original
> location.

I can see 3 options:
1) is to mount the main mirror rigidly in translation and angulation,
and move the secondary within the cage to put it at the optical axis,
2) main mirror as in 1), but a more normal cage that is moved to its
true position by adjusting the effective lengths of the trusses when
mounting (I believe this calls for 6, rather than 8, trusses - or else
let the 2 remaining hang loose until this is done).
3)ensure that the tube (truss or otherwise) is dimensionally stable,
and do the fine collimation needed at the main mirror support.

I suspect trying to compensate for both weakness in the tube and in the
main mirror cell will be an exercise in frustration.

One confounding factor, I think, is the ubiquitous secondary mount with
bolt and3 screws at the secondary holder, this places the diagonal at
the end of a long arm and puts very stringent demands on its rigidity.
I prefer mounting the secondary rigidly at the center of a 3 vane
spider, with the mirror almost level with the "T" cross, clamping the
outer ends of the vane and adjusting there - something closely related
to Rien's recent proposal. But I must confess I have never felt
motivated to actually try another, more conventional, approach ;-) 


> Any thoughts from the group on this subject would be welcome.

Those were mine!

Nils Olof