THE STORY OF A FACETING MACHINE

I began faceting in my teens, my first machine being little more than a few lumps of scrap metal that spent a few lunchtimes in the school metalwork shop. It was crude and very wobbly but it worked (or at least could be persuaded to work sometimes).
The head itself could now be tackled. My initial plan was just to screw it down and get cutting, but I realized that if I did that, it would probably stay like that for the next 20 years. There were a lot of bits that were too flexible and poorly designed, so a fair amount of rebuilding was done. Now with a decent workshop, what took me 6 months on the Perris was finished in a fortnight's spare time. The original index wheel was recut, since I previously couldn't cut gears and had to rely on a finely divided disc and graticule. The mounting fork was stiffened up and a surplus micrometer head and dial gauge mounted on the cheater.
I cut about 20 stones on it. Then I was lucky enough to find a lathe that I could afford - a Perris 'Modelmaker', forerunner of the Cowells lathes, and having learnt a bit more engineering, I set to build a proper machine.
Unfortunately the Perris lathe only has a centre height of 1.5" and takes 8" between centres, so I couldn't work with big lumps of metal, and lots of parts were a compromise. However, it was finished, it was a huge improvement on Mk I, and I was very pleased with it.
About this time I went to university, and the machine was mothballed. Life overtook me and it languished for the best part of 20 years. Ultimately I began to accumulate the sort of workshop in which I could build a decent machine, so the old one was taken out of storage for an overhaul.
The main problem that set me back was the lapping unit. Mk. I and II both used a reduction gearbox taken from inside a photocopier. It was disastrously unsteady and after 20 or so stones the bearings were totally worn out. I spent some time toying with designs with roller bearings, variable speed motors, and the like, and was just about to get started when I had a remarkable piece of luck. A friend asked me if I had a use for the motor from a rotary evaporator. If you've not seen one of these, they consist of a motorized unit which rotates a flask full of liquid under vacuum so that the liquid can evaporate from the surface. They're used in laboratories all the time. When I got hold of it, it was complete with speed control and gearbox, and the "output shaft" was actually a conical socket running in preloaded bearings. Due to the nature of the work it was used for, it was designed to cope with heavy off-axis loads. The top speed is about 350 rpm, a little slow, but acceptable. All that was needed was to make up a master lap on a tapered shaft which could drop into the conical socket. The moral is, if you ever see one of these things in the skip, get hold of it quick.
In this picture you can see the adjustable bearing about which the fork swings, and the return spring for the cheater, which is the 14g steel rod that presses down onto the brass fork near the dial gauge. The gauge, of course, doesn't have a strong enough spring to provide the return action itself. The only problem with this arrangement is the bearing about which the fork swings; despite some close-tolerance work, it's still a weak point.
Here you can see the rotary unit mounted on the baseplate. The protractor is finely engraved. I haven't gone digital yet! The scale was drawn in MacDraw, printed on inkjet acetate, and attached with cyanoacrylate.
This view looks down into the conical socket. The index gear has rectangular slots and the index pin is lapped to a close fit. I can't detect any shake in this assembly. The numbers in this case were put on with Letraset.
The socket ran fairly true and since the lap and shaft were turned between centres, the finished unit had a TIR of about 2 thou at the edge of the master lap. The master lap itself was made from a disk-shaped piece of stainless that was rescued from a skip. Don't know what it was, but it was just right The index arrangement can be seen more clearly from the back
The index assembly is built to detach and reposition simply by retracting the right hand cone bearing. This is a great help on those occasions that you can't get the lighting just where you want it.
The master lap is half an inch thick, more at the centre. Here you can see it mounted on the taper shaft. The dial gauge resolves .001 mm so cheater repeatability is excellent.
Looking down onto the master lap. It weighs about 5 kilos and provides a very smooth movement.
Of course it isn't finished. When you start building these things, you just can't stop. There are still a few ghosts of the old machine, in particular the sleeve which runs up and down the column is captive on the screw, so there's a lot of twiddling to change angle.
An unlatching half-nut is needed here. Also the sleeve was too big to bore on the Perris, so I used a bit of tube with a bore that was slightly too big and it actually runs on 6 teflon screws.
That can all be changed when time permits (or when a big lump of metal comes along). The total cost was probably less than £100 over the years.
A 6 inch copper lap in place.
. This stone has been on the machine for the past 20 years. When I finish it, will it be a record?