[ATM] Spherometer

[mikell at optonline.net (Michael Lindner)]

On Tuesday 09 March 2004 09:26 pm, Robert Turner wrote:
> Can someone point me in the direction of some resources on making myself a
> spherometer? I am interested in trying to make an achromat...

Before you settle on a design for a spherometer, decide what you need it to 
do. To what tolerance do you need to measure the curve? Take the sperometer 
formula for the kind of spherometer you're considering (ring, 2 ball, 3 ball, 
etc.), put in the high and low tolerance values, and you'll see what sort of 
precision in meansurement you'll need (which will depend on the diameter of 
the spherometer).

For instance, I made a 3 ball spherometer out of a pulley with a 1/2" bore. 
The pulley was placed on a shaft on a jig on my drill press, and I drilled 3 
holes the same distance from the center. socket head machine screws were put 
through the holes, and the sockets on the screws were filled with epoxy and 
topped with 1/4" ball bearings, to make the feet. Then I inserted a 3/8" 
brass bushing into that, and put a dial indicator through the bushing. The 
pulley set screw holds the dial indicator in place.

Before I started, I used the formula for a 3 ball spherometer, the diameter I 
planned to make it, and the curve I wanted to generate, and determined that I 
only needed to measure +/- 0.001" on the dial indicator (which saved me quite 
a bit of money).

Also, for a reference, I took a small plate glass mirror blank, and one side 
was flat enough. I also took a spherical mirror with a known radius (measured 
by foucault test) and used it to calibrate the spherometer. With 2 known 
curves, or in this case a flat and a curve, one can correct the constants in 
the spherometer formula for any manufacturing defects.

-- 
Michael Lindner
http://www.starastronomy.org *** http://home.att.net/~mikel
http://www.atmsite.org *** http://www.atmlist.net