Building a Waineo-type grinding machine
One of the projects that has been on my "someday" list was a rebuild of a lens/mirror grinding machine. I built a Hindle type machine about 12 years ago, which would grind blanks to about 12" in diameter. When we moved, it stayed with the old house, and after a few years of non-use, the motor refused to start again. I toyed with the idea of rebuilding another Hindle, just scaled up 50%, but after researching various designs, I decided to abandon the Hindle type, and build a Waineo style machine, named after the late Tom Waineo. (See also Sky and Telescope, November 1995)
The basic machine described by Tom Waineo was designed to do glass to about about 10". It was driven by a surplus gearmotor, and had one speed for both turntable rpm and strokes per minute. The machine was, however, very compact, about 1/4 the size of my Hindle. Some searching on the web showed others had taken this simple design and scaled it down to handle small lenses and mirror, and scaled it up to handle 20"+ optics. I decided to build a Waineo sufficient to do 20" glass, adding some small modifications for varying the turntable and stroke speeds, as well as the ratio between these two speeds.
Some of the web pages about the Waineo mentioned the sloppiness inherent in the center spindle design. I decided the turntable of my machine would be supported on a bearing, with shaft and a V-belt pulley for power transmission. Changing the pulleys would allow slowing down the turntable and stroke speed for large elements, but keeping the higher speeds available for the more common 8" to 12" projects.
A bit of work on a spreadsheet gave me a good set of pulley sizes, and a few sketches started to gel the general outline for the design. I started at the turntable axis and worked my way out for the details.
I had a 2 1/4" thick by 20" disk built of 3/4" plywood, used in a long-ago-and-quickly-abandoned attempt at enlarging my Hindle. My original thought was to add a 1" shaft to ride in a central bearing, but that changed very quickly, as life took another one of those turns....
A 32" disk of glass had been offered for sale to another buyer, but he wasn't interested. The seller then asked me, and I jumped at the chance. Based solely upon his "in a dark warehouse" examination, I purchased what turned out to be a 32" disk of plate glass, 4.0" thick at the center, and with a 102" convex radius. This disk is destined for "X the Unknown", a 32" Cassegrain.
This instantly changed my thinking about the machine. It became clear that I needed either two sizes or one convertible grinder. Since it took me years after my last one died to get started on a new machine, I dismissed building two as wildly unrealistic.
I decided to build a basic core machine for up to 20", with a single 1/4hp gearmotor providing both stroke and rotation power, but with the option of adding an auxiliary stroking assembly, driven by a separate motor for very large disks. That implied a very sturdy central support, much stronger than the small bearing I had intended to use.
Some rummaging in the bearing box turned up a 5.75" ID tapered roller
and race. The bearing had at least one roller with some obvious wear, but as I
would be using it at 60 rpm or less, and at only a few percent of its rated load
capacity (probably in the tens of tons), I thought it more than sufficient. I
initially thought of turning a wooden axle to ride in the huge bearing, but then
I came across a surplus 7" diameter by 11" cylinder of aluminum, so I
cut that down on my Shoptask to fit the bearing. 
Also gathering dust in the garage was a 2.5'x14'x14" block of UHMW (ultra high molecular weight) plastic, purchased with the intent to make large lens cells. A large section of this block was sacrificed to the grinding machine project, and I machined it to hold the bearing race and the cut away the remaining ID for the axle. UHMW was very easy to work with, and didn't create a cloud of sawdust like wood would.
After completing the bearing block, it was fastened underneath the tabletop with six 1/2"x9" carriage bolts (that oughta' hold it!); these bolts extend through the block and form the supports for a small shelf, which holds the bottom bearing, a 1.125" tapered roller. Rather than try to get the location perfectly on-center, I drilled 3/4" holes for the 1/2" bolts, and adjusted the bottom bearing shelf to match the exact centerline of the main bearing. By using opposing pairs of bolts and lock washers, I can adjust the bottom shelf up and down to vary the load on the bottom bearing as well.
Since I was using belts, I need some method of adjusting the tension, as the belts will stretch a bit over time. I decided to mount the gearmotor (a Bison 1/4 HP, 47 RPM) on a hinged door (the tips of the hinges are just visible in the photo). I adjust the tension by the incredibly hi-tech method of a soft wood wedge; tap the wedge with a hammer, and that will increase the tension on the belts. Crude, but very effective. Here's a shot of the interior of the machine, showing the Bison motor and belts and such.
Next, I mounted a saw blade mandrel (salvaged from the old Hindle) on the edge of the table, using prefab steel strips. The bottom end of the mandrel holds the 6" drive pulley, and the top a 3" solid aluminum idler pulley, upon which I mounted a steel strip, drilled with a set of holes to vary the stroke length. The holes hold an axle bolt, which connects via a 1/2" rod end bearing to a very large turnbuckle (from McMaster-Carr). The opposite end of the turnbuckle holds another rod end bearing, and it connects to the stroke arm in one of a number of holes. These allow some additional variation of the stroke length as well.
The hinged stroke arm was built from the largest hinge I could find at Home Depot. To minimize slop, I used one with a plastic liner on the hinge axle. I bolted a 2x4 to the hinge, and drilled a hole directly over the center of the table, into which I inserted a length of 5/8th threaded rod, with an acorn nut on one end, filed round as Tom Waineo recommended.
The rotating table was covered with 2" strips of Velcro, to facilitate changing the configuration of the table top. By using plywood rings or V-notches and plastic sheet, I can instantly change from one size to another. This will also minimize the cleanup required in between grades of grit.
The entire machine was painted in glorious "Shoptask Yellow", and an initial shakedown with a 24" disk of flint glass, weighing over 170 pounds proved to me that this machine will take whatever I can throw at it. Using an 20" pitch lap with 40 lbs of weight, the machine easily took 8" strokes, with the belts only slipping when the lap started to dry out. Tom Waineo's simple, and elegant design works very well in this size, and even larger I suspect.....
Here is the entire machine.... For scale, the turntable is 20" in diameter.
Here's a view of the saw mandrel and driving arm.
Here are some Quicktime files (no audio, sorry...) of the machine in operation, grinding (tool on top) a 12" disk of CerVit. As you can see, I normally cover the tabletop with wet newspaper while grinding (but I've removed some of it here to give a better contrast with the parts of the machine). This makes cleanup between grades fast and easy. Wet newspaper is also sufficient to hold the disk to the turntable. The Velcro is only required during polishing when the frictional forces are much higher.
Driving disk, turnbuckle arm, and stroke arm
Here's a picture of the tool and the mirror being ground in the videos...
Any questions? I'll be glad to answer, just send me an email.