ATM : Making and mounting newtonian diagonals

["Dan Chaffee" <dchaffee@gvi.net>]

My best results in making flats has been with Texereau's method
of polishing and figuring a relatively large piece of glass _before_
being cut up into diagonals. This way they are worked stress-free
which seems to be critical when working with plate glass.
The piece you choose to work needn't be cut into a circle;
an octogonal shape will suffice--just be sure to grind a small
bevel on both sides. 

Of course, the usual precautions of finding a suitable support
with some 'give' to it, like a rubber mat or an old mouse pad,
stand as with grinding thin primaries, they just should be 
proportionally thinner.

It is important that you determine whether the face you are 
going to attack is concave or convex. Choose the convex
side, as this has, in my experience, the best chance of moving
towards flatness. Fringe analysis is of course, essential--get 
use to it right off, since you will be constantly checking  fringes
when figuring. I use a stripped broken diagonal that is still of very 
good figure. If you don't have a reference flat, it's going to
be harder, because you will have to find two extra pieces
that are fairly close to flat that will allow you to solve with
simple equations which is concave/vex and by how much.
You can read about this technique on Jeff Baldwin's web
page at <http://members.gotnet.net/bald/fringe.htm> I think it's also
discussed in Tex.'s, Thompson's and Howard's books.

If the blank tests more than two waves of curvature across the 
expanse of the test flat (3-5 cm across),  you will want to grind 
with 320 using a full sized tool and work your way down to
9 micron. Flash polishng and testing as soon as the glass
will show fringes in flourescent light will tell you whether 
you are close. I polish with a full sized lap TOT and to a lesser
extent, MOT. It is much easier to approach flatness from a 
slight amount of convexity after grinding, than if the blank tests 
concave after grinding.  Keep in mind that the glass will change 
shape by the time polishing is complete, so this has to be factored
in while working. That's right, worry more and grind/polish
less:-) It will tend to move away from the amount of convexity 
there was in the beginning by the time it's polished,
so if it tests flat at the onset of polishing, it
likely won't be by the time you are ready to figure it. This
is assuming everything is done by hand. Not the case with
planetary polishing machines, but that's another ball game.

Things can and do change VERY fast with 
plate glass, especially as you get within 1/2fringe of flatness, 
so beware! On the other hand, plate can take one heck of a 
shine.

 Steve Fejes  wrote:

>Normally a secondary mirror can only be supported at the 
>edges - so this is like a primary mirror with 3 points
>of support at the edges

I have never supported my diagonals this way; three dabs
of silicone cement at the 70% zone have always worked
form me, even with relatively thin flats. I don't know why
this isn't normal(?).

D Chaffee