[ATM] Hartmann Testing

[Ted Cohen]

Bravo Jim - that is excellent - letting the computers do all the work means
you can measure the whole mirror in a snap. Use of the webcam is very
clever. Since I'm not using a webcam at this point, I have to manually
measure each zonal reading, which is very time consuming. My method only
measures one chord across the mirror at a time, while your method measures
the entire mirror in two or three unique sets of measurements. Now I see why
you guys are testing at the ROC - to get the whole image of the mirror onto
the imaging chip while using an incoherent light source spread out over the
whole mirror. Very interesting. Thanks for sharing that. 

Ted.
-----Original Message-----
From: atm-bounces at atmlist.net [mailto:atm-bounces at atmlist.net] On Behalf Of
Jim Burrows
Sent: Sunday, May 11, 2008 8:37 PM
To: ATM List
Subject: Re: [ATM] Hartmann Testing

At 2008-05-11 15:04 -0700, Mark Cowan wrote:

>I've been working off and on at developing what I've called a "scanning
>Shack-Hartmann" test...
>This test will work at ROC on (I'm optimistic with reason) almost 
>any asphere.
>That it's a complex and unproven approach I acknowledge, the thread 
>I'm linking
>to describes pretty much everything so far.  My exploration of the
necessary
>accuracy looks promising, however.  :)

At 2008-05-11 12:29 -0700, Ted Cohen wrote:

>I've adapted the Hartmann test so that it can be performed theoretically at
>any distance from the mirror.
>However, I don't yet know if my setup will be accurate
>enough when figuring, so take this with a grain of salt. I'm using
>interpolation based on calculated slope measurements. I don't know if this
>applies to what you're doing, and I don't claim that it is an improvement,
>or is more promising, or anything else. Rather, it's a reflection of my
>interest in what somebody else is doing that is very similar to what I'm
>doing.

Hey, you two (or anybody else thinking about 2D non-interferometer 
tests), take a look at
         http://home.earthlink.net/~burrjaw/public/hart2d.zip 
(2006-09-18, Win32, 1.13 M)
It's a fairly well tested method - compared very nicely with 
interferometer results on the Mirror Round Robin, I've used it on my 
RC cass project (including - modified to do a Hindle-Hartmann test - 
on the secondary (star test coming up)), and, as mentioned before, it 
found 20 nm bumps on a thin mirror while sitting on a cell.

The general idea is to put a checker board mask on the mirror, place 
point source and webcam (lens removed) so that you get a full-chip 
image.  The program processes the image to find the shape of the 
mirror surface using the FFT - iterate the surface shape until the 
theoretical image matches the actual image.  Some of the processing 
is a modification of the Misell iteration used to find the Hubble 
primary errors after it had been stupidly chucked up into orbit.

         -- Jim Burrows
         -- http://home.earthlink.net/~burrjaw
         -- mailto:burrjaw at earthlink.net
         -- Seattle N47.4723 W122.3662 (WGS84)

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