Interesting compact high ratio transmission design

[s_forbes]

Those of you into Rubik's Cubes may know about this guy; Oskar is a puzzle designer that makes crazy cool stuff on Youtube. Here's his channel.

A few weeks ago he posted this simple three piece transmission, with a reduction of 27:1. It works on a similar concept to harmonic gearboxes. This seems like a design that might work for some mechanisms on the robots we build, particularly for something that needs a very large reduction and doesn't see a lot of torque. A similar design could probably be printed and interfaced to a motor quite easily, giving a cheap and quick gearbox for those teams with 3d printers.

If that's not enough, here's a variation on that concept that goes up to 11,373,076:1.

[GeeTwo]

It reminds me more of a two-stage cycloidal gearbox, though without the pegs and oversize holes in order to minimize the wobble - the wobble is just left there.

This could probably be driven by a belt or chain that was actively tensioned or some sort of CV joint. Either way, I would attempt to minimize the mass of that wobbling inner gear.

The 11M:1 gearbox is interesting, as well. The output torque is obviously limited by the strength of the material, implying that the efficiency is going to be exceptionally small. To put that gearing in perspective: to get a single revolution per day from the output, you'd need to drive it at 7898 rpm.

[Hugh Meyer]

His last line is the best! "This is a solution looking for a problem." Interesting design.

-Hugh

[carpedav000]

Originally Posted by Hugh Meyer

His last line is the best! "This is a solution looking for a problem." Interesting design.

-Hugh

Definately see some drivetrain use here. To prevent being pushed backwards/forwards, have something clamp down on the yellow part to lock up the whole thing.

[asid61]

Neat! I couldn't see this being used for FRC though, maybe for big arms that require large reductions. The issue would be finding internal gears with a tooth difference of 1 (unless you offset the internal gears and used a spur as an output).

[Richard Wallace]

This looks similar to the Pittsburgh Torque Multiplier Lugnut Remover.

[wgardner]

Originally Posted by s_forbes

A similar design could probably be printed and interfaced to a motor quite easily, giving a cheap and quick gearbox for those teams with 3d printers.

It's like the 3D printed gearbox shown at this link, but with more teeth.

[Clem1640]

This is a harmonic drive, also known as strain wave gearing. It provides very high reductions can provide very high torque and does not suffer from backlash.

This technology is used and available commercially.
http://harmonicdrive.net/

It is distinctly different from Cycloidal gearing, but the two do share some attributes. Primarily, both advance the output by one tooth of an annular gear for every full rotation of the input shaft.

[Foster]

Originally Posted by Richard Wallace

This looks similar to the Pittsburgh Torque Multiplier Lugnut Remover.

Thanks for the link and the blast from the past. I just spent 20 mins following all the threads back. It does look like it would be good for arm movements.

The printed one seems like it would have lots and lots of chatter. There is one post that talks about having a finer pitch so it's a lot smoother movement.

Ahh the things you learn here on a lazy Sunday morning

[GeeTwo]

Originally Posted by Clem1640

This is a harmonic drive, also known as strain wave gearing.

No, harmonic/strain wave has all three gears concentric in the same plane, with a flexible intermediate gear, forming a single stage of reduction. This unit has none of those features:

• all of the pieces are nominally rigid
• the drive gear unit is not concentric with the two annular gears
• it is two stage - the inpututput gear ratio is the product of two stages (or it can be more physically viewed as the difference between the two ratios (for a<b, 1/a - 1/b = 1/ab, or if you swap a and b, you get -1/ab).

This drive can also be used to generate a rotation in the same direction as the input, or the opposite direction, by switching which of the two annular gears is fixed and which is rotating.