From: "CSC" 
To: "Atm" 
Reply-To: "CSC" 


Great test setup!  I am also fascinated by this laser star test on spheres.
Here's my 2cents.

You're not at sphere yet, as shown by Foucault.

Looks like KE from the right, and inside focus, as you say.

You're undercorrected, with the center and edge high, and a low 70% zone,
classic oblate spheroid, the inverse of a paraboloid or prolate spheroid
(ellipsoid).  This is likely from the edge of the lap stopping and changing
direction in this region, causing greater friction and wear there. Needs
longer strokes, etc. or whatever your favorite method is for parabolizing.

The edge looks good, based on the diffraction ring that shows in the dark
half of the mirror.

No astigmatism showing in the star test.

Fairly smooth, as there are no major "ray like" streaks or lumps
in the rings.

The star test is most useful comparing inside and outside focus images. I
suspect that the extrafocal image shows minima (lines) that are not as
crisp. Ideally the contrast in the rings is equal on both sides.

Try a 30% obstruction on the mirror, like a 2.5" paper disk, taped on
the mirror and observe when the shadow appears in the diffraction image,
both inside and outside.  The distances should be equal.  If it appears
sooner inside, then it's undercorrected.

What is your magnification?  Go as high as you can, and as you get closer
to the sphere, look at the star test images closer to focus. The test gets
extremely sensitive when there are only 2-3 rings in the diffraction image.
I use about 700x.

Read Suiter's book on star testing and keep in touch.  I had briefly tried
to get star test images of this process, but   The testing you're doing
will
get you to 1/40 wave.

Congratulations,

Colin






-----Original Message-----
From: owner-atm{at}shore.net [mailto:owner-atm{at}shore.net]On Behalf Of Ray J
Sent: Monday, March 17, 2003 12:14 AM To: Peter
Cc: atm{at}shore.net
Subject: Re: ATM Making & Testing a Perfect Sphere



Peter,

I am working on a the same thing, testing a sphere, except I just want to get
the hang of testing.  I thought I had a good sphere but what happened is that I
got better at testing and the mirror got worse and worse.  I put some pics at
http://www.j-engineering.com/ATM

Some rings in the star test are different from the others.  I have not figured
out what they mean other than that something is not quite right.  For me,
enclosing the light path was a must.  Quality knife edges and an 3000-7000+
mcl
LED's also helped.

To answer your question, I used Sixtests with b fixed to 0.0.  FigureXP does the
same thing with Desired Deformation set to 0.0  My knife edge positions
were close together but different with practice.  For me, the range was
0.008", but I
will test again with more zones.

Knife edge with masks, software, the knife edge straightness near null,
then just looking for shadows at null, is the general testing route I'm
taking.

If someone could tell me what the star test is telling me, give me a shout.

8" F/6.3, Ray

Peter wrote:

> Hi,
>
> After sending my 1st email I realized that all I need is a method to
> evaluate the Foucault results. I'm struggling with the math involved to
test
> the sphere.
>
> I've used the Foucault test in combination with Tex for porabolized
mirrors.
> Is there a program like Tex which will allow me to test spherical accuracy
> Vs parapolic accuracy?
>
> If not would someone please give me a reference to the applicable formula.
>
> Thanks,
>
> Peter

--- BBBS/NT v4.00 MP