Re: [ATM] The zonal Foucault test is free of inherent correction bias

[Michael Peck <mpeck1@ix.netcom.com>]

At 10:22 PM 3/10/2005, Ken Lowther wrote:


>Quoting Michael Peck <mpeck1@ix.netcom.com>:
>
>>the atm list who haven't followed recent threads on ATM_FREE, a 
>>hypothesis has been advanced that zonal Foucault tests are biased towards 
>>undercorrection, at least for relatively large, fast mirrors.
>
>Isn't this why we tried the 'round robin'?  I can see differences by f 
>ratio, but why would size matter?

Just to add a bit to Vlad's comments -- as I recall the mirror round robin 
mirrors were all about the same size and f/ratio -- I think around 6-8" and 
around f/5. I can see the possibility of systematic effects of total OPD on 
test outcomes, although the evidence presented over on ATM_FREE seems 
pretty thin to me so far. Again from memory I don't think the round robin 
testers had any systematic biases in their estimates of overall correction, 
but that doesn't bear directly on the hypothesis that was put forward.

Part of what I'm doing is just trying to replicate 60 year old results from 
the professional literature. Linfoot worked out the complete diffraction 
theory of the Foucault test in a series of papers that were published in 
MNRAS and a book in the '40s and early 50s. That work is irrelevant to 
current professional practice, but it's of some relevance to amateur 
telescope makers who for the most part are unaware of it. The FFT was 
unknown at the time and of course ordinary civilians had no access to 
electronic computers, so his approach seems a little old fashioned to us -- 
he actually had to work through some math and perform integrals using 
pencil and paper. Given a basic knowledge of the physics it's fairly easy 
to reproduce his work numerically, which I don't think has been done 
thoroughly before.

Vlad is also right that a simulation strictly speaking only tells you 
something about the simulation.  The only thing this simulation was 
intended to test was whether diffraction effects within mask openings have 
systematic effects on estimated surface profiles. The answer I got was 
negative. That doesn't necessarily mean that real world testers couldn't be 
fooled by diffraction effects in systematic ways, or that one or more of 
dozens of other possible causes of error migh produce systematic effects.

Mike Peck

------
Michael Peck
mpeck1@ix.netcom.com

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