pic: Ball Differential parts
 

Re: pic: Ball Differential parts
 

Clever... I never would have thought of that as a differential mechanism.

-q

Re: pic: Ball Differential parts
 

This is an "Old School" design that has worked beautifully for years in the RC Car world. Although, with RC 's, the motors output pinion would engage teeth on the outside of the ball ring. Then the left and right drive wheel would be attached to the outer plates through more gears. Compression force would be used against the outer plates to regulate the amount of slippage that occurred between the two output shafts.
(For those of you who are/were RC junkies, think RC-10);)

This is a great adaptation of that design. Well done!!

Re: pic: Ball Differential parts
 

My Rc car's got 2 of these but way way tinier with carbide diff balls.....
associated rc18t (cute little guy) with a mamba 8000kv brushless setup and a 3 cell lipo.. the ball diff slips a bit sometimes lol.

These have largely been used in the RC car world because its alot easier to make them tiny as opposed to the bevel gear type, the larger cars and trucks do use gear differentials though.

i kinda wanna drive this base you got going...

Re: pic: Ball Differential parts
 

I've got the original RC-10. :)

I've replaced almost every part over the years, especially the bearings.

Which brings me to my point, that you should replace those cheap sheet-metal OSH bearings with something better. Those things will deform quickly (especially if you apply enough pressure to the races) and not turn as well.

Re: pic: Ball Differential parts
 

The whole system is a test stand for development. With funding low preseason we are trying to build everything low cost just to see how it works. When we build the actual ball differential we will use harder bearings. Once we figured out how to put it together and did assemble it we found that it was easy to assemble. The most difficult was the ball guides that are welded to the shaft. We drilled the center hole in the steel and than drilled the holes around the center (they do not have to be perfect off center to work). We than welded the plate to the shaft and than lathed it out to make it round. We than threaded the rod. The rest we just had to slide spacers and bearings into place. We drumeled the keyway for the wheels.
To test how tight they have to be we will prop up the mecanums up with casters and than run with just the omni wheels. This should give us a good idea on the amount of slippage under pressure.