To: imasloth{at}bigpond.com, atm{at}shore.net
From: mdholm{at}telerama.com
Reply-To: mdholm{at}telerama.com


John,

We need a couple of more items of data:

1. Is your tester fixed source or moving?  If the light source stands still
while the knife edge moves toward and away from the mirror, then it is
fixed source.  If the light source moves toward and away from the mirror
the same as the knife edge, then it is moving source.

2. When you have the knife edge adjusted to the first zone (zone 0 in the
list you posted) measure from the center of the mirror to the knife edge
and light source using a tape measure.  Measure to about 1 mm accuracy. 
(It is a good idea to put a piece of adhesive tape over the end of the tape
measure so it doesn't scratch the mirror.)  This measurement is easiest if
you can get someone to hold the end of the tape measure to the mirror while
you make the reading.  If your light source and knife edge are not the same
distance to the mirror, report the average of the two numbers.  This
measurement goes into the X bias and Distance entries of Sixtests.

I tried the numbers in Sixtests using 2044 mm as an estimate for X bias and
Distance.  Since I don't know if you are using fixed source or moving, I
did both.  If you have a fixed source, the mirror is undercorrected.  That
is, it is about 1/2 way between a sphere and a parabola.  Compared to a
parabola, it is high in the center and at the edge and low at 60 mm radius,
rising slowly toward 80 mm radius, then rising steeply toward the edge.

If you have a moving source, then the mirror is slightly overcorrected with
a Strehl ratio of about 0.84.  This is already about as good as many
commercial mirrors, but it is probably worth while for you to push for a
Strehl ratio above 0.90  The improvement in image quality is significant.

You should download and start using Sixtests.  You won't get much farther
without being able to do quantitative analysis of your Foucault numbers,
and Sixtests or one of the other Foucault analysis programs is the smart
way to go.  You can find Sixtests at
http://home.earthlink.net/~burrjaw/atm/atm_math.lwp/atm_math.htm

Also, you are close enough to a parabola that testing in more than 4 zones
is a good idea.  Make a second mask (use the intermediate mask in
CouderMask). Test first with one, then, without disturbing the tester or
mirror position, change to the second mask and continue testing.

Believe me, there are good theoretical reasons for using more than 4 zones,
even though lot's of atm's will call me bad names for saying so.   I know
testing 7 or 8 zones is tedious.  I know many good mirrors have been made
with 4 or even three zone testing, and some with no quantitative testing at
all. Unless you are severly math handicapped, getting the numbers and
interpreting the results is the most certain way to go.  It is also the
easiest way to let other atm's evaluate your progress and give figuring
advice.  (Using Sixtests or another of the Foucault analysis programs
removes most of the math from the problem.  You plug in numbers and get a
graph that shows your mirror's shape.)

When you send the two items of data mentioned above, I will post on a web
site, a Sixtests analysis of your mirror, so that you can see exactly what
it looks like.  I will show screen shots of each of Sixtests' pages.

Mark Holm
mdholm{at}telerama.com

--- BBBS/NT v4.01 Flag-5