The Wheelbox V2 - Planetary Falcon Gearbox

I love it.

What does the weight on this look like?

1 Like

Solidworks is saying 2.03 lbs not including the motor, which is another 1.1. My confidence in the accuracy of that is pretty low though.

1 Like

This design would be perfect for this year’s control panel, although the Falcon 500 is probably a bit too much and you wouldn’t want it to break against it.

Are you sure the thin section bearings would be able to hold that? If they’re the same bearings people use in swerve drives, they aren’t great for radial loads.

What’s your source on that? They’re typically rated quite high for both axial and radial.

I know I saw numbers somewhere at some point, but Slim Section Bearings simply rates them as “poor.”

Will look for actual numbers to back it up. There’s a good chance it’s fine in this application though, as “poor” could still be 100+ lbs per bearing.

Looking at the silverthin listing for the common size swerve bearing, it looks like the x-contact bearing will be fine in this application. Apparently x-contact bearings have the same radial load capacities as normal radial bearings of the same size.

Thank you @asid61 for calling me out on this one.

1 Like

This is something I’ve been thinking about doing for a while! Awesome design.

1 Like

Looks like a really great design.

We were working toward something like that too, in January. Knew we wouldn’t be the only one.

We built the 'motor in the wheel" in 2019 that actually put mini-cims inside a cylinder-shaped wheel, so in 2020 this seem the next logical step. But ours is the “Neo in the wheel”. Falcons are rare and we’ve never seen one in our shop.

We built two with 3d printed parts. Intention is to switch to metal gears when available.
Performance was amazing. No belts, tiny space, tiny weight. And thinking about six–or even eight wheels…!!

But after fiasco of motor-in-the-wheel development during build season making our 2019 robot often immobile, kids didn’t want to develop while building this year.

We plan to work on in the the spring. Oh yeah, there is no spring in robotics this year…
Plan to work on it sometime…?


For bearings, we use 6813-ZZ, 65mm id, 85mm od, can purchase on ebay for about $6 each if you wait for china shipping.

Similar in size to my car’s back wheels. No chance of breaking that on a robot.

This is really cool and detailed concept.

What’s the rationale behind driving the wheel via the carrier plate, instead of driving the wheel through the sun gear? I would think the design could potentially be made much simpler, though I haven’t thought much about the loading either way.

Smells like in-hub motors on EV racecars.

Which gives me some ideas. :slight_smile:
Can we have an offroad racing game next year?

Using the planet carrier as output provides a little extra reduction over using the ring gear as the output. (This is part of why planetaries can provide high reductions in small packages… I’d argue that if the output or input isn’t the planet carrier, you’ve just got a glorified spur gearbox)


Using the ‘planets’ static comes from last year’s design…the axles in them held the mini cim.
If you made them spin, your design wouldn’t be simpler.

Currently reduction is like 61/9 so good for 14-15 ft/ sec. The 3D printed ring gears worked without a problem last year…no failures of that part. Printed in CF-20. Would be under even less load in 3-planet design above.

My original design was two stages, with the final stage having a fixed carrier and the final ring gear was the wheel. It ended up being very complex, and as ThaddeusMaximus said it was basically just a glorified spur gearbox. Gearboxes like this one where the carrier drives the wheel is used in industrial robotic applications for drive wheels. I basically just gave that design an FRC treatment.

I realized after I posted this, that because the wheel plate spins at the same speed of the wheel, it would be easy to add a pulley mounted to that plate, and drive another set of wheels. You could then use just 4 motors to drive 4 traction and two omni wheels or something like that.

This is true until you account for the fact that the ring gear has to be smaller to accommodate the bearing and extra material outside of it. With a driven ring gear you can build much closer to the size of the wheel, so the ring gear will have more teeth, mostly offsetting this advantage.

Really any planetary is a “glorified spur gearbox” - you’re using a planetary to get more tooth contact / less backlash and thus can get away with smaller gear teeth to make everything more compact. I get what you mean in the sense that you don’t gain the extra tooth count from the sun gear, but that’s not really the reason someone would build a wheel like this.

That’s what I’m saying - lock the planets. The OP is doing so in the above wheel which adds a ton of complexity. Your team’s example is what I had in mind.

I imagine that almost all of the complexity in your design last year came from the gearbox being two stages, and not from the ring gear being driven.

You’re right, the main reason the other one was so complex was the dual stages. Your suggestion of using the ring gear to drive the wheel is how I originally started this project last year. I seem to remember having some difficulties with the size of the ring gear relative to the size of the bearings I was using, and getting all the fasteners to fit where they needed. I was also using a larger pinion instead of the 8 or 9 tooth. I should revisit it though, as what you are describing does seem much simpler.

Reminds me of an old design I did before we even had minicims. I think you’ll like putting the ring gear in the wheel better than driving it off the planetary carrier. You can get plently of reduction, especially now that we have 8t pinions.

1 Like

Really cool stuff. I also offer Homage to Pwnage, having seen there’s before we did one. But was that 2014? Thought it was 2012…?

Also note, our design originally started this year as a swerve module with a 5-inch colson built in. But that didn’t seem the right thing once we saw the game.

This topic was automatically closed 365 days after the last reply. New replies are no longer allowed.