From: "Jerry" 
To: "ATM List" 
Reply-To: "Jerry" 




-----Original Message-----
From: owner-atm{at}shore.net [mailto:owner-atm{at}shore.net] On Behalf Of Nils
Olof Carlin
Sent: Sunday, May 11, 2003 5:17 PM
To: Vladimir Galogaza; ATM shore
Subject: Re: ATM Robo-Foucault, Image intensity, and Changing knife edge
readings


Vladimir wrote:

> Not in the caustic ( Gaviola Platzek)) test.  Object of the test is
to
> determine longitudinal
> and lateral position of the source images from left and right parts
of the
> zone.
>  Not simultaneously but consecutively.
> This is considered as advantage of the caustic test.  ( Disadvantage
is
> extra
> precision required in determination of the lateral KE position.)

And of course the possibility that the whole setup will move by a few
microns between readings! But a robotic setup will certainly decrease the
risk for that, as compared to hand-driven XY "dingbats".
>
> If robo is capable of this extra precision than it could be that
> caustic test is more suitable for robotic testing.

Doing a Gaviola-Plarzeck caustic test means that the zones can be wider,
needing less "depth of focus" and with narrower peaks,
potentially more accurate readings. The zonal width of a Couder mask for
Foucault is less, the farther from center the zone is, for precisely this
reason. With G-P caustic, you can even use equal-width zones (not practical
with Foucault). Also, the diffraction peaks are separated, and do not cause
much interference as is the case if you let them overlap. Thus, with a CCD
chip at the zonal caustic, you can measure the separation of the spots to
an accuracy that is potentially a fraction of a pixel, without worrying
about the lateral resolution of the dingbat. You can only use one zone at a
time because of interference at the caustic plane, but with robotic
technique, you could make the mask variable in size and position. I believe
a robotic setup of G-P caustic is feasible and well worth investigating (at
least for mirrors that are not slow or small) - and since it involves
position of  diffraction peaks, rather than relative intensity of mask
openings, it should avoid problems of detector non-linearity etcetera.

Nils Olof


Vladimir and Nils

Concerning Foucault vs. Caustic and Robo I was being somewhat narrow in my
thinking. I am thinking more on a test sufficiently accurate for most
amateur mirrors. Not the larger faster behemoths some are making these
days. So I was sticking to Foucault in my comments and recognize that
astigmatism or off axis testing will cause the beams not to cross on axis.
But in Foucault one assumes they do and that is what one looks for in the
test.
        I would like to see James succeed in the Robo-Foucault without a
mask. To be able to just place a mirror into a simple test apparatus and
push a button and have a reasonably precise output of its quality would be
real nice and very useful to many ATMs.
        A mask is a complication. You have your mirror making class and a
number of people want to test their mirrors and Richard Schwartz shows up
with a 7 5/32" mirror and no mask. No problem! Or you have a mask but
a zone of interest on the mirror is at the boundary of mask zones. Off
hand, I can not think of a way to do a caustic test without a mask.
        I think that most likely James is going to find a good way to find
the Foucault test "null" automatically without a mask. I would
consider him successful if he can make it do as well as an experienced TM
testing manually. I'm betting he will do better than that.
        I'll still use my eyeball.... if for no other reason than to avoid
the "Hubble error".


What is the "poor man's caustic test"? I am not familiar.


Jerry

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