Re: [ATM] holographic null test on round robin mirror C

["Nils Olof Carlin" <nilsolof.carlin@telia.com>]

I replied to Matt Tudor:

> >This may sound attractive - but you should not forget to estimate the
kind
> >of precision you are out to get. A tilt of one wavelength over a 2" zone
is
> >about 1/100 000 - you need to better this by a factor 10 at least - a few
> >microns over a ROC of a few meters! Or for that matter, the diffraction
peak from a zone, measured to a precision of some 1/10 of its width. I can't
> >imagine Shack-Hartmann doing that - or am I wrong?

Looking a little closer at it, I suspect I am! With one setup described, you
would get a diffraction disk something like 50 micron, or 10 pixels wide -
and it ought to be possible to pinpoint the centroid to much less than one
pixel width.

Hmm - interesting...

I don't know if you could do without an autocollimator flat, doing a
near-COC test - it shouldn't in principle be impossible.

What I believe this test does is separate the images from each "zone", thus
avoiding interference problems.

Another way that I guess might work is "scanning" the surface of the mirror
using a movable mask opening, determining the centroid position for a number
of known mask positions - easier perhaps to make at home, but requires a
stable setup not to induce drift.

Anyway, I do believe there are possibilities for practical full-aperture
alternatives to interferometry.

Nils Olof



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