Re: [ATM] Ain't got no shperometer

["Stephen Koehler" <s.c.koehler@gmail.com>]

Dave,

> I'd really like to hear more about setting an optical tester up to test a
> mirror mounted in a scope.  I've heard of this, but not much.

I assume you are talking about a quantitative tester.  (For
qualitative tests we have the star test, null Ronchi and null knife
test using a star source, of course.)

For a mounted telescope, you are limited to doing the test at focus,
thus you have to supply approximately collimated light or you won't be
able to reach the point of focus through the telescope's focuser.
Working at this position, you won't have much focus shift for
performing a quantitative Foucault test.  This leaves something like
interferometry.

To do standard interferometry, your source is usually near the point
where you capture the interferogram (the focuser).  Thus, you would
need a full-size flat mounted beyond the end of the telescope tube to
return the light.  This can be done, but has the practical
disadvantage of having to perform the test with the telescope
horizontal, which is not its normal operating mode, and may introduce
extraneous aberrations.

An alternative is to use a remote source (star or artificial source)
and use one of the algorithms for reconstructing a wavefront from
defocused star images.  (This is what was done to quantitatively
diagnose the problems with Hubble after the mirror escaped the lab and
was located in a very inconvenient place.)  There's free software for
doing one variant of this on the astrosurf site.  Search for "projet
roddier".  My thinking is that this is the most convenient way to do a
quantitative test of a complete telescope system, but more work needs
to be done on the software end.  A disadvantage of this approach is
that a remote source must necessarily travel through a large amount of
air, so air currents can disturb the results.
-- 
Steve Koehler
_______________________________________________
ATM mailing list http://www.atmlist.net/