Re: ATM 11.2" F/2 Primary testing questions :)

["Ken Hunter" <kb7h@onemain.com>]

Did you forget that you can also do a double pass autocollimation test
with a flat?

Ken Hunter

----- Original Message -----
From: <Aplanatic@aol.com>
To: <atm@shore.net>
Sent: Monday, December 31, 2001 5:24 PM
Subject: Re: ATM 11.2" F/2 Primary testing questions :)


>
> Hi James:
>
> There's nothing like a null test for fast mirrors, IMHO.  So, I
would use one or more of the following tests:
>
> 1) Ross Null Test.  This requires an on-axis light source, a lens, a
ray tracing program and accurate placement of the nulling lens with
respect to the pinhole or slit and the KE.  Most likely use can find a
suitable lens from Edmunds Optical or Mels Griot for this application.
To ray trace the situation start with a point source, pass it through
the lens, reflect from the mirror and pass the light back through the
lens to the focal point.  The light source can be moved longitudinally
so as to bring the focal point back to the light source, and the lens
can be moved longitudinally to get the best overall null across the
aperture of the mirror.  Residual spherical aberration in this test
(if significant) can be calculated and removed during data reduction.
One should use a monochromatic light source for this test, but I have
found that a filtered red LED does a passable job.
>
> 2) Asymmetric Dall Null Test.  As far as I know, I'm the only one
that has ever used this one.  Place a point source as far away as
practical from the mirror.  In my case, I'm able to get the point
source about 30 meters from the mirror by opening my garage door and
placing the light source across our street on a neighbor's fence.  By
ray tracing, find a lens that when placed in front of the KE causes a
null for your paraboloid.  You will need a secondary diagonal mirror
in front of the primary in order to inspect the reflected light, and
this diagonal will prevent you from measuring the central area of the
mirror, usually not much of a problem.
>
> 3) Not Quite a Null Test (NQNT).  This is identical to 2) above, but
without the nulling lens.  By calculation or ray tracing, you can
determine the residual longitudinal aberration for a point source not
quite at infinity, and then do some data reduction to find the surface
profile of the mirror under this test condition.  Because the light
source is much farther away than in the conventional Foucault test,
the zones are considerably easier to read and the longitudinal
aberration is quite a bit smaller.  I've used this test on a 10" f/3.3
paraboloid at 30 meters with good results.  Once again, you will need
a diagonal to inspect the reflected light.  The inverse of this test
is also interesting, where the light source is placed close to the
mirror and the reflected light is inspected as far away as possible.
I've never tried this, but it should work in principle.
>
> 4) Point Source at Infinity.  The obvious one.  In your case, the
point source must be very far away for high accuracy, probably not an
option.
>
> 5) Secondary Collimation Telescope.  Use a known good paraboloidal
mirror to form a collimated beam that is equivalent to a point source
at infinity (after correctly focusing it).  This is my favorite test
of all.  I have several carefully figured mirrors in the f/6 to f/8
range that I use for this purpose.  The only significant problem with
this test is the care required in collimating the setup.  Doable, but
not trivial.
>
> Anyway, best of luck.  I've tried figuring optics below f/2.5 and
find it nearly impossible.  But, I have only limited patience.
>
> Dave Rowe
>