Grinding started off with MOT, pin off center aprox 3/4", with a wide, 40% stroke, 80 grit SC. I favor the maximum hogging formant, which digs a hole quickly, but, as we know, leaves different radius from center to edge.
I use a small (2"dia base) sphereometer similar to Texereau's (page 170), but with a 3" dia. clock face with divisions to .0001". With this you can check radius over small areas. My 8" F5 primary has a sagitta of .100", but this doesn't discribe the curvature of the mirror, only how deep the center is. When we see bubbles between mirror and tool, that shows lack of contact.
The sphereometer quantifies this, allowing for very accurate grinding. The 2" base should have a sag 0f .00625 to be exactly f5, the gage reads to .0001", and if calibrated on a known flat, can be accurately read to .00005" (some will argue that this is pushing reality, but we do it at work all the time.) In theory, with a +/- tolerance of .00005, focal length can be accurately guaged on the mirror to aprox 1/3".
I also use another type of sphereometer which has a 3/4" circle for the base, only it uses 3 equally spaced nylon rollers as contact points. The 1.5" dia barrel is marked off in .0001 increments also, which allows for very accurate measuring, out to aprox 1/4" from the edge. In the above f5 case, the sag in this case would be .00088".
Keep your eyes open at swap meets and garage sales for used dial indicators, since they're used in just about every machine shop, and often are very cheap, especially ones reading to .001", which is probably fine for most purposes. They're great for Foucult testers also.
Anyway, to even out the deep hole I've dug, which I purposely dug aprox 20% deeper than spec., flip to TOT, 180sc, 1/3 stroke slightly off center (1/2"), which usually brings to a sphere pretty quickly, and reduces the sag. The trick is to get sag to where you want and the pitting from 80sc gone at the same time, as IMHO, when you get to 240grit and below, you really have to work to change the radius significantly. But with a controlled stroke, final radius was fine tuned by alternating MOT/TOT and by 5u AO, sphereometer shows its' where it should be.
By the way, for the finer grits (I used 80/180 sc, 240 white AO, 20u/9u alum oxide lapping powders) I use another 1 armed bandit that has a variable speed DC drive on both spindle and eccentric, which allows for lots of variations in stroke, speed etc, and with a 500ml plastic wash bottle with the bottom cut off, and a hose with a pinch clamp on a piece of vinyl tubing attached to the top, which is now the bottom, to control the drip of the lapping solution ( 200ml Antifreeze <suspending agent> + 100ml H2O + 1 cup powder) on to the mirror/tool. So, you can just sit and watch the mirror turning, or in our case, go work on other stuff.
It'll be interesting to see how close actual foucus will be compared to theoretical from the sphereometer readings, after I put an initial shine on this thing.
Next: Pitch/Polish: This is where the real fun begins!
Bill Marriott Forest Knolls, Ca. USA btk@ix.netcom.com