[Raytrace] Dall Null for 22" f4.5

[John D. Upton]

Robbin,

At 06:48 AM 9/2/2002 -0700, Robbin Palmer wrote:
>I set up a Dall Null test in OSLO Lt per John's instructions and found
>0.15 wave error in the wavefront using an f = 300 mm, d = 66 mm plano
>convex lens. Then I tilted the mirror 1.7 degrees as it would be in a
>real setup to avoid interference of the lens with the light reflected
>from the mirror and bring the focus over to the knife edge, reoptimized
>and refocused. The wavefront error jumped to 20 waves! Did the same
>thing for 0.5 degree, error of 1.8 waves.  Looks to me like I can't use
>this test.  Am I doing something wrong?

     You have not done anything wrong in the setup.  OSLO-LT is 
unfortunately telling us exactly what was asked.  You are seeing the 
effects of astigmatism in the test since it is being done off-axis.  The 
wave-front error being reported includes the light from the whole mirror.

     There are two ways around this dilemma.  The first is to use a 
beam-splitter so that the Null Test may be conducted completely on 
axis.  This is often not possible if we want a simple setup for amateur use.

     The other way around the problem is to simply ignore it as much as 
possible.  When the return image is either over or under the pinhole but 
off axis, the astigmatism will appear along the vertical axis of the 
mirror.  Since the knife edge cuts into the beam 90 degrees from that axis, 
we can train our eye to ignore the astigmatic effects and concentrate on 
the horizontal axis only.  Imagine a mask on the mirror that only shows a 
swath 50 mm high across the center of the mirror.  We watch this band for a 
null condition as the knife edge cuts into the returning beam.  We can 
actually train our eyes to pay attention to that part of the mirror without 
using a mask.  You might want to try a mask until you are comfortable 
concentrating on the centered horizontal null area only.

     To see how large or small a mask (or just visual swath) might need to 
be, try this in OSLO-LT.  Open your Dall Null file and click on the 
Aperture Radius button for the mirror.  Select "Special Aperture Data 
(X)."  For "Type", select Rectangle.  For "Action", select Transmit.  Now, 
enter the aperture data.  For X min and X max, enter the radius of your 
mirror, -279.4 and +279.4, respectively.  For Y min and Y max, enter -25.0 
and +25.0, respectively.  Close the dialog.

     You now have a mask across the center of the mirror which is 50 mm 
high.  Refocus and check the wave-front error.  You should find some more 
reasonable numbers now.  You can play with the height of the mask to see 
how large or small an area you need to watch when nulling the mirror with 
your off-axis setup.  If the numbers are still unacceptably high for you, 
you will have to use a beam-splitter or choose another test.  Note that the 
Ross Test could be a little better, but will still be off-axis for many setups.

Sincerely,

John D. Upton
Georgetown, TX
<http://www.atm-workshop.com/>     "The ATM's Workshop"
<http://hometown.aol.com/RonWin20/> "Ronchi For Windows Home Page"