To: atm{at}shore.net
From: "Matthew L. Brown" 
Reply-To: "Matthew L. Brown" 



Mark,
   I think we are working in parallel.  My little 4.25" f9 is at the same
spot: center 90% is nice and spherical, with a 3/16" TDE.  Ronchi is
hard to read with the overlapping images, so what I see matches your
description quite well.  TDE is very clear under Foucault, with a ROC about
0.5" longer than the central region.  Under Foucault, the TDE gives me
a darker region right along the knife-edge side of the mirror when the
knife is nulling the central region.  It doesn't look flat all the way out.
 I really know its a TDE because I can null that region by pulling the
knife back 0.5".

My computer setup is a little in disarray, but I'll try to put up matching
Ronchi and Foucault test images on my web page.  Anyone else have example
images on their web pages?

=Matt


>Mark:
>
>         I'm a stickler for edges and find that the Ronchi test can be
> difficult to
>read at the exact edge.  The grating causes diffraction of the light and
>creates overlapping images of the mirror.  The Ronchi bands seen in these
>overlapping images at the edge create a confusing image, looking like an
>outward flare and hook.
>
>         I suggest you take real close care in looking at the edge, and
> see if this
>overlapping image effect is what you're looking at.
>
>         Varying the grating spacing may make it less of a problem.  I
> prefer 50lpi,
>and careful observation.
>
>         Also, it makes sense to use the KE to determine the hooking of
> the band at
>the edge. Put it a half inch inside focus and cut the shadow 1/4 the way
>into the lightpath.  If you see the distortion at the edge, it's real.
>
>         While at a sphere, try the star test using a bright pinhole or
> underpowered
>laser pointer with its collimating lens removed.  Use diffract or Suiter's
>book on star testing to verify you have a sphere and a good edge.
>
>         Using the Ronchi with a monochromatic source makes for
> interesting ronchi
>bands which I think are worth looking at.
>
>Colin

--- BBBS/NT v4.00 MP