[ATM] holographic null test on round robin mirror C

[Nils Olof Carlin]

matt tudor wrote:




> With all the talk about the holo[g]raphic null test, what about a more
modern
> version of a Hartman test ? All it takes is a mask in front of a ccd
camera
> . A precision metal mask can be made for very little money . The rest is
> calibration and software .

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?

> There are very few ATM attempts at Hartman testing, yet in the
professional
> world Hartman started being used a century ago.

The only one I know of is Jim B - apparently successfuly these days but not
without a lot of effort, I presume. And there are diffraction issues that
are not easily dealt with - I believe if you use "shutters" to open one zone
at a time and record its spot position, you may avoid interference between
zones and still have the images close to each other for accurate centroid
locations.

One variety would be a 2-dimensional Platzeck-Gaviola caustic test with some
moving apertures and measurements from the CCD. Just thinking loud....


> Modern varieties like Shack
> Hartman are the rage, lenslet arrays can be had from a number of companies
> for not too much money. CCD's of sufficient resolution

> The great advantage of Hartman variety tests is that they're not sensitive
> to local seeing the way interferometry is, they collect data over the
whole
> mirror not only along a diameter, and combine the best of both worlds . A
> Hartman mask with a few hundred subapertures would be equivalent to an
> interferogram in the number of datapoints sampling .

TANSTAAFL

Nils Olof