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


I don't use the laser light for Foucault.  It creates so many interference
effects, it's difficult to use. I use an ordinary slitless tester.

Below I describe using the laser light for Ronchi, which is VERY interesting.

By star test, I mean an artificial star test at COC.  It is interpreted the
same as in Suiter's book, "Star Testing Astronomical Telescopes",
which inspired this test.  Also see "Star Tests" by W.T. Welford
in "Optical Shop Testing" edited by D. Malacara.
Note that testing on a real star (infinite distance) is the null test for a
paraboloid.
Star testing at the COC is the null test for a sphere.  On the bench, the
f# is doubled, making it extra sensitive.

The monochromatic light from the laser makes the patterns very distinct,
and absence of seeing conditions makes it possible to test at extremely
high magnifications.  I did calculate how many waves of defocus I was
using, but don't have the calcs here to determine how far I had to move the
eyepiece. It was .010" to get 2 waves of focusing abberation or
something like that. Star testing is extremely sensitive at small defocus,
and with only 1-2 waves of defocus, (corresponding to 1-2 rings in the
defocused image) any slight error in the wavefront clearly shows very
different patterns.

I did use the laser source with a 200 lpi Ronchi and it is WILD!  The
diffraction with a monochromatic source creates bands WITHIN the dark
Ronchi bands, (let's call them internal diffraction bands) and by expanding
the image of one Ronchi line to nearly cover the mirror, it becomes a most
elegant test for straightness of the bands and hooks and crooks at the
edge. The internal diffraction bands are superbly sharp as opposed to
blurry with monochromatic light. Also, while regular white light Ronchi
creates diffraction effects which make reading the edge difficult due to
the overlapping images, this Monochromatic Ronchi test uses this
diffraction to advantage. Internal diffraction bands are sharp and clear to
the edge.  Any TDE or zone shows much more clearly.

I don't know who else uses it. Let's call it the Monochromatic Ronchi test.
Perhaps it's an illusion, but it appears to reliably show very small edge
defects.  Anyone care to analyze this effect?  Is it sort of a Shearing
interferometer?

Colin



-----Original Message-----
From: J-Engineering [mailto:Ray{at}J-Engineering.com]
Sent: Monday, February 10, 2003 1:51 PM To: CSC
Subject: Re: ATM Tesing a spheriod?


Wow, what a nice reply!

Do let me ask some dumb questions.

Do you then almost always use the laser diode as the source for the foucault test?
Then, since it is a point source, diffraction shadows will be visible?  My slitless
has no diffraction lines that I can see.

By star test, you mean a real star?  Or are you looking at the reflection
of the diode
at COC with an eyepiece?

I will look more closely at Peters site.  I am pretty sure I will set one up.

Thanks much,

Ray

CSC wrote:

>         I recommend the knife edge test, and carefully measure the
longitudinal
> motion (spherical abberation or zonal error) necessary to get the
wavefront
> error.  I used this test when I was pursuing "perfect" spheres for
flexing.
>
>         I found the point at which I could detect general shadow motion
inside
> focus, then measured the distance to where I could detect shadow motion
> outside focus. I also made sure that any slight zones (and I mean really
> faint at the end of figuring) were also extinguished at this knife
setting.
> Since there were no other defects visible on the surface within this
range,
> I assume that all errors were within this wavefront difference.
>
>         According to Diane Lucas, High Priestess of Telescope Making,
(didn't know
> we had one?), the knife edge travel (longitudinal abberation) for 1/4 wave
> WAVEFRONT error is:
>
> assume stationary light source
>
> longitudinal abberation=16* wavelength*f^2/y^2
>
> wavelength is .000022"
> f is the focal length of your mirror
> y is the radius of your mirror
>
> My 8" f/7 would be:
>
> LA=16*.000022*(56^2)/(4^2)
>
> LA=.069"
>
> a moving light source would be half that.
>
> I found no error in the final figure within .007" travel, using a moving
> source, so I calculate I was within:
>
> ..007/(.069/2)= .2  of the above formula, or 1/20 wavefront pv
>
> I also used a laser pointer with the lens removed as a point source of
> monochromatic light.  Using a 4mm orthoscopic eyepiece to view the
> diffraction image, I watched the results of the figuring as it showed in
the
> star test.  The slight astigmatism was isolated by rotating the mirror,
and
> was found to be due to the lateral offset in the test rig.  Could have
been
> designed out.  With a diverging beam, an uncoated surface and high
> magnification, the diffraction image was not too bright.
>
> At 711x, I figured to the point where I could not tell any difference in
the
> star test. You would not believe how beautiful this test is. It is
> supposedly accurate to 1/40-1/60 wavefront under lab conditions. Any
> over/under correction shows obvious patterns, and coupled with the
foucault,
> it is easy to get a superb sphere.
>
> If you setup the laser as above, you might also try Peter John Smiths
> Shearing interferometer.   Simply place a vertical plane of glass to
> intercept the returning rays from the laser, and by shearing the wave
front
> (folding it back on itself), you can see true interference bands which
will
> be perfectly straight when null (sphere).  See it at
>
> http://www.users.bigpond.com/pjifl/page15.html
>
> Let me know if you need any more info.  I think making a great sphere and
> flexing it is THE way to a fine scope.
>
> Colin
>
> -----Original Message-----
> From: owner-atm{at}shore.net [mailto:owner-atm{at}shore.net]On Behalf Of
> Ray{at}J-Engineering.com
> Sent: Sunday, February 09, 2003 7:29 PM
> To: atm{at}shore.net
> Subject: ATM Tesing a spheriod?
>
> Hi All,
>
> When testing a spheroid,  how can you quantify the figure?
>
> As far as I can tell, the edge is good.  I used an oversized lap which
> seemed to work rather well.
>
> If I set things up just right, I can get the Ronchi lines (grating /
> grating) to bend just a bit.  Hardly noticeable.  It seems to only work
> at the few positions where the diffraction images combine or merge just
> right. How can I tell how far off the surface is?
>
> In a slitless knife edge test,  at ROC, I can just eek out a few shadows
> just before the whole image gets too dark.
>
> I'm working on getting a feel for using these tests and trying to
> associate some kind of number to what I see.
>
> (8", f/6.3)
>
> Thanks for any help or web links,
>
> Ray

--
Ray Jurevicius, PE
Jurevicius Engineering, Inc.
New Berlin, Wisconsin  USA

Metal Forming Simulation
http://www.J-Engineering.com
(262) 782-7073

--- BBBS/NT v4.00 MP