Differential Roller Screw

BTW, this is not patentable. Far too much prior art.

If grooves engaging the screw are circular, you get translation that is the thread pitch. If the inside is helically threaded, then there is a multiplier that approaches infinity as the female thread approaches the diameter of the screw. The roller offset must be half of the diameter difference. It is an epicyclic differential effect.

great man, but i keep playing and playing but i still can´t understand how it works...the excentric nuts compensate themselves while running?

It's a bit of a mind bender, yeah. The things that are eccentric are the bearings.  Imagine three rollers outside a cylinder, the middle one pushing against the other two. The cylinder can rotate freely, and the roller centerlines are (maybe eccentric is not the best word)... offset. Now use the inside of a bearing as the roller in the same way.

Thanks, guys. I've designed larger, but not made any yet. I did not tap the inside of the bearing, but steel sleeves that are pressed in.

Man amazing idea, you should pattent that...

Did you try higher diameters? Like a 3/4 inch rod or something?

I might make one of thoose for my cnc mill...

You gotta use a carbide tap to thread the bearings?

Interesting...

How did you tap the inner race? Is it not hardened steel?

any backlash?

Yes, backlash must be built in when using non-ground threaded shaft. More complexity comes in when using threaded ID rollers, as some mechanism is needed to keep the rollers synched with each other, or binding results. Circular grooved rollers do not have this downside, but then the pitch is not differential, but the same as the screw.