From: "Jerry" 
To: "'Andy Saulietis'" , 
Reply-To: "Jerry" 


Keep in mind that the HST had the advantage regarding determining the
surface error that the original test setup that produced the error could be
determined. Had that not been the case could they have determined the
surface error by focal image alone. I don't think so.

Jerry

-----Original Message-----
From: owner-atm{at}shore.net [mailto:owner-atm{at}shore.net] On Behalf Of Andy
Saulietis
Sent: Friday, July 04, 2003 4:12 PM To: atm{at}shore.net
Subject: Re: ATM Robo-Foucault, the "Fresh Start" is in progress! :)


Mel et al write:

>  Briefly: Cover the mirror with a mask containing an grid of small holes.
>  Take two images at a known spacing.  Analyze the spots and viola - a
>  complete surface profile.
><<<
>
>I've done some thinking along these lines and discussed it with Jim
Burrows.
>He pointed out that the diffracion effects between the holes are
significant
>and suggested ignoring the focusing light from the holes and instead
>concentrating on the diffraction pattern.  Perhaps Jim can discuss this
>further.


Jim & I also discussed doing this test with a CCD camera, it's just a
variation of the Hartmann test that was (is?) used for
*large* mirrors as a proof test, they used large glass plates for the out of
focus images with spacings measured in feet, to get the minimum spot size
the system was capable of producing, usualy an arc second or so. Very few
large telescopes were actually diffraction limited.

We found this test doesn't work well with small apertures because the
distance between the two out of focus images is too small to measure the
slopes of the rays accurately enough to extrapolate the shape of a mirror
to the precision needed for figuring.

Jim also worked on a scheme using overlapping out of focus mask images to
determine relative distances between spots by interferometry, this looks
more promising but it's not mathematically trivial & doesn't test the
entire mirror, only one chord of it.

We're now discussing taking out of focus images of bright stars, similar to
the Hartmann test, but with no mask. The two images would be scaled and
subtracted, yielding an intensity pattern that could be used to deduce the
shape of the mirror..Suiter's descriptions of the star test are an
empirical eyeball method of doing this. Jim says that this no-mask image
technique was what was used to analyze the HST optica after the figuring
error was discovered, and the data was used to design the corrective optics
for that telescope..this seems to have worked fairly well..:)

I'm not an expert on this type of analysis but will be using my large
format CCD camera to generate test images for Jim & others to play
with.





Andy Saulietis
ISS Enterprises
PO Box 19
Mayhill NM 88339


World's first astronomy land development:

http://www.pvtnetworks.net/~iss/starend1.htm
505-687-3067 voice/fax, call voice or e-mail to set up e-mail: iss{at}pvtnetworks.net

--- BBBS/NT v4.01 Flag-4