Re: [ATM] Mirror cell for an 8" mirror

[Mark Holm <mdholm@telerama.com>]

Corona Australis wrote:
> I'm making a 3 point mirror cell for an 8" mirror. I was wondering and
> also hoping someone here could tell me what would be the preferable
> distance from the mirrors centre to position the three points.
> 
> The mirror is an F/6, 204mm neat in diameter and is of a thickness of of 
> 23mm.
>

The optimum for this mirror, assuming plate glass, and a 1.52 inch 
diagonal is 39.7%.  This results in a 3.05 nm RMS deformation, which is 
good enough.

Fixing the supports at 60% results in a deformation of 5.87 nm, starting 
to be a bit high, not because it would be a problem if the real 
deformation were that large, but because it gives little margin for 
construction tolerances.

Ross Sackett wrote:

> While the worst-case error pointing at the zenith is a
> little greater than the PLOPtimized cell at ~40%, the
> mirror is much less likely to get hung up on your edge
> supports as you collimate it, which is a common
> problem with 3 point cells with the pads closer
> together.  

Not sure I understand this statement.  The amount of tilt needed to 
collimate a mirror is the same regardless of where the adjustments are 
located.  The problems with adjustments closer to the mirror center are: 
1. adjustment becomes more sensitive and more sensitive to slop. (use 
fine thread screws.)  2. The structure isn't as inherently stiff. (I 
wouldn't worry about it on a mirror this light.)

Ross Sackett also wrote:

> If you really want to complexify it and toss in more
> points, PLOP says 6 points will support it better than
> 9. 

Agree with this 100%.  9 points is traditional, but wrong.  Kriege and 
Berry wrote an excellent book, but they wrote it before Plop.

In my not so humble opinion, a K&B style tailgate isn't such a hot 
design for an 8 inch mirror.  It's main strengths (lightness, stiffness 
and relative ease of construction for the lightness and stiffness 
obtained) really only become important at larger sizes.  One of it's big 
weaknesses is putting the collimation adjustments at the same radius as 
the mirror supports.  When you accept the fact that a 3-point cell is 
optimal for 8-inch and smaller mirrors and that the optimum is at 40% 
radius, you need to start thinking about decoupling the mirror support 
function from the alignment function.  Think about it for a while.  It 
is actually much simpler than it sounds.

-- 
Mark Holm
mdholm@telerama.com

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