ATM : Re: Astigmatism

["Dan Chaffee" <dchaffee@blitz-it.net>]

John Sherman wrote:

>1)  Astigmatism cannot be seen with the Foucault test or the Ronchi test,
  >       unless it is horribly severe.

What do you consider horribly severe? Can you give us some rms values
based on your findings in terms of the threshhold of detectability with
foucault
or ronchi?

>And here is how you can prove to yourself that I am correct. Make your own
>astigmatism!! Last night I performed yet another experiment. This time I
>used an 8" f/8 128" roc paraboloid mirror. This is an excellent Richard
>Fagin mirror known to not have astigmatism by using the eyepiece
>Suiter-style star test. Using a pinhole light source and a Spot tester,  I
>set up with the spot at the 70% zone. When there is less than one inch
>lateral distance between source and spot the ring shaped shadow of the spot
>was round, showing no detectable astigmatism in the mirror or the setup. At
.two inches lateral separation the ring is very elliptical shaped. This is
>the visual manifestation of simple astigmatism, you can see it first hand.
>At three inches the astigmatism is so severe that an image was not
>obtainable with the 2D Spot tester. Yet with the Ronchi tester I still saw
>normal lines at each orientation. Since this is a paraboloid mirror, the
>lines are slightly curved to match the curvature of the mirror (with a
>spherical mirror I would see perfectly straight lines). Even at six inches
>lateral separation between the light source and the Ronchi grating the
>Ronchi still showed normal lines at each orientation. No matter how I
turned
>the gratings (I have two, 100 lpi & 150 lpi), astigmatism was not visible.

The relative insensitivity to apparant astigmatism resulting from
increasing lateral from ronchi/knife edge to source seems to me
a bonus, since it allows for astig on the glass to be differentiated
from that due to misalignement along the x axis.

>Using a slit light source instead of the pinhole might make some small
>difference, but the slit has to be aligned with the grating. If they are
not
>aligned or if something is not square to the optical axis then you might
see
>a ying-yang effect because of that.

Wouldn't you want the grating perfectly aligned with the slit for
the ronchi test regardless of the aberrations being detected?

>Even though the Ronchi tester makes measurements at every point on the
>mirror, it is not a 2D test. Why do so many people insist on using a
Ronchi?

Because it allows the eye to do something it does very well, that being
the detection of  subtle changes in contrasty patterns. It is very effective
for
seeing zonal errors, especially a turned edge. Any sphere I have verified
as a null with ronchi and foucault has not shown any sign of astigmatism
in a star test at high power unless the orientation of lines/ knife edge
shadow with the mirror in question was slightly tilted with respect to the
slit. An early 150mm f8.3 mirror I worked star tested with slightly less
than the amount of deformation comparable to Suiter's illustration of
3/8 wavefront astigmatism, yet showed obvious astigmatism in
both foucault and ronchi tests; hardly what I would call "severe".
I know what caused the stig, a clumbsy attempt at small lap
work at the edge in one quadrant of the mirror and the aberration
was not global.

>It is pre-Hitler technology, found in dusty old books.

This has nothing to do with whether or not it is effective.
The star test has been around for ages too, along with  birefringence,
double bypass, and so on. Does that invalidate them as  worthwhile
tests? With all due respect, it's one thing to validate your own variation
of a foucault test, another to attempt to invalidate other tests that do
have merit.

Dan Chaffee