> Only the edge (about 1/2 inch all around) seem to be in contact.
>
> Still with #80, I have tried TOT, MOT, 1/3D centered, 1/3D W, 1/2D
> centered, 1/2D W, about 8 - 10 wets of each. Pencil testing after
> every 5 wets. Nothing seems to change.

Now I realize that there may be others in this group who have more
experience than I, but from my experience I would say that there are
two things going on here.

1 - Rough grinding using the strokes you name above tends to create a
surface that is deep in the center of the mirror. An over exagerated
example of the shape is the end of an egg. The radius of curvature
in the center of the mirror is longer than at the edge of the mirror.

2 - You have a turned "UP" edge at this stage. Depending on the
extent of the turned "UP" edge, this can be desirable at the end of
80 grit, and can easily be ground out by the natural grinding process
as you progress with smaller and smaller grits.

I have ground or helped others grind a little more than 50 mirrors.
I ALWAYS use a small (3" diameter) spherometer to test the mirror at
the end of each grit. The criteria I look for when testing the
mirror to determine if the mirror is ready for the next grit...

1 - No pits / scratches remaining from the previous grit size.
2 - No astigmatism around the circumference of the mirror.
3 - Spherical or slightly turned up edge on the mirror. I.E. no
turned down edge.

To perform these tests ...
1 - Pits / Scratches is a visual test. You look at the surface with
an eyepiece and a bright light source grazing across the surface of
the mirror. The texture on the glass should be consistent without
any larger than normal pits or glints of light on the surface of the
mirror. If pits / scratches exist, continue grinding. A few small
pits are acceptable, especially for your first hand ground mirror.
If you extend the last three to five wets of each grit, this helps
reduce the pits A LOT making the next grit go much faster.

2 - Astigmatism - The measurement on the spherometer should be the
same all the way around the mirror. To fix astigmatism, continue
grinding making certain that you are properly rotating the tool and
mirror in opposite directions every few strokes. Also make certain
that your grinding support is perfectly flat and not flexing the
mirror while grinding.

3 - Spherical surface - Set the spherometer at the center of the
mirror. Note the measurement. Now slide the spherometer out to the
edge of the mirror watching what the needle on the dial indicator is
doing. It will most likely vibrate through the end of 220 grit, this
is O.K. What you are looking for is the same measurement all the way
across the mirror. If you have a turned up edge, the dial indicator
will show a small increase in measurement. I.E. The pin of the
indicator will be moving AWAY from the indicator. A Turned down edge
will have the pin moving TOWARDS the indicator. All indicators are
different, so press the pin by hand and watch which direction the
needle moves. The absolute measurement is not important here, just
how much the needle moves. I try to have the surface between
spherical and .0005" turned up edge by the end of 80 grit, .0003" by
the end of 220 grit, .0002" at 320 grit, <3u by the end of 25 micron,
and <2u by the end of 12 micron. The first measurements are done by
a spherometer using a 1/10,000 dial indicator. The last two
measurements use a dial indicator having a resolution of 1u or
1/1,000 of a mm. This is about the same as .00004", smaller than
what is measurable by the first spherometer, but is completly useless
on grits larger than 320 since the texture on the glass causes the
needle to swing a full 1/4 turn on the dial.

When I grind and polish a mirror using these tests, I always get a
spherical surface the first time tested. When I teach other first
time mirror makers, they sometimes get a perfect sphere, but most
often have a gradual curve up on the mirror which can be polished out
to a sphere with minimal polishing time. Sometimes if they have been
using a stroke that is too long, they get an overcorrected surface
which takes a while to polish out, or possibly require going back to
25 micron.

For more information on how to make an accurate spherometer
inexpensively, see my web site http://home.attbi.com/~john4937 follow
the menu to Mirror Grinding then Spherometers. By buying Ebay dial
indicators, it cost me about $20 to build a spherometer. You need
less than 1mm or 1/16" of travel on the dial indicator. However you
do need an indicator with a resolution of .0001" or .001mm. Other
indicators like the popular ones with only .001" resolution will not
have enough accuracy to show you what is going on with the surface of
the mirror.

John R. Zeigler
Vice President
University of Utah Astronomical Student Society