Re: ATM 19th Nervous Breakdown

["Dave McCarter" <dmccarter@sympatico.ca>]

While I don't want to try to analize your mirror, I can suggest how to
proceed. I am currently figuring a BVC 16 inch f5.34, so my experience may
be closely relivant.

I do have a 12.5 inch lap with quite hard pitch which I used to get the
surface spherical. Then I made an eight inch lap with quite soft pitch and
carved channels into it with a sharp knife, making the outer facits into
points, so it is a star lap, with more surface contact in the center and
less to the edge.

I have the mirror face up and turning very slowly, about half a revolution
per minute. I squirt on a bit of water, a couple of cc of very fine rouge,
and start doing moderately slow circular to eliptical strokes, diameter
about four to five inches, in a counter clockwise direction across the width
of the mirror, from left to right. I allow the lap to overhang the edge of
the mirror a little more than an inch during each end of a pass. My three
mirror cleats are numbered so I can keep track of when I have gone three
times around, and it is time to wash off, dry, and let the mirror sit on the
stand for a minimum of two hours before I test. I use a 120 line per inch
Ronchi grid usually used outside of the RofC to evaluate the overall
progress and look for problems developing, and also foucault test with a
five zone mask for the hard numbers.

To me, a series of straight bands says a spherical surface, while bands
bending in at the edge says raised edge not getting any polishing. A band
that hooks out says it is a longer radius and I am probably not extending
the lap over the edge enough to keep the edge with the inner surface. Also
look for the diffraction edge to see if the edge is turned.

After 12 figuring sessions, the last at nine tonight, the mirror is
beautifully smooth and slightly under corrected with surface error of 0.078
waves. So I am closing in on the correct surface, and pray that I will be
wise enough to stop in time! Maybe two more sessions tomorrow.

We star tested the mirror last night, more to check that the poles were the
correct length (they were 1.25 inches too long) when it had a surface error
of slightly over 0.1 waves, and the images were quite good, but star testing
revealed the under correction.

Dave