[MichaelGolden]

Accuracy is definitely costly!

I'm thinking about using the 90:1 worm gearing from a 6" rotary table.
With 2000 microsteps that would be 90 x 2000 = 180,000 steps per revolution of the table.. 360 degree / 180,000 = 0.002 degree per microstep

For the trunion mount I would use a 72:1 worm gear from a 4" rotary table.
With 2000 microsteps that would be 72 x 2000 = 144,000 steps per revolution of the clamping surface.. 360 degree / 144,000 = 0.0025 degree per microstep


With the trunion table surface widest point at 3" radius from the center.
3.1416 x 6 = 18.8496" perimeter... 18.8496" / 144,000 steps = 0.0001309" movement around 3" radius from the center
At 0.5" radius. 3.1416" perimeter... 3.1416" / 144,000 steps = 0.00002181666" movement per microstep

So I would assume put the chucks center as close to the pivot point, in order to keep the movement of the workpiece very small. But if I was to use the table without the chuck my step movements should still be manageable.

Any thoughts?

[mcphill]

But your "accuracy" is calculated not taking any backlash in to account. Any worm driven rotary WILL have backlash, or it will be bound up and not be able to be turned. You should consider a different technology for the rotary if precision is important to you. As suggested earlier, I would look at Harmonic Drive components...

[MichaelGolden]

I understand. And I began looking into harmonic drives. Not sure where I would get them and at what cost. I did a search and they seem very costly. Anyone know of a distributor for this application?

But realistically, I dont need +-15 microns and wouldnt expect it. Especially without going with a heavier machine to begin with.

I've read some success with cnc converted worm gears. You have to adjust them. And binding would be the biggest concern.

Some people have said to have about 0.002-0.005" of backlash with the same gearing. And they say it can be adjusted and mostly eliminated. Lapping the gears is also an option. I think its worth a try for the cost.