pic: 114 2 speed Coaxial Crab Drive

My team members say that I’m crazy… But after cooking this design up in a couple days, I think I may finally Agree with them.

Two speed, dog shifted, Coaxial Crab. Low speed: 5. High speed: 10. Entire weight (minus steering chain): 45 pounds.

Questions, comments, and critique welcome.

I like it. Not bad at all. Whats the demensions of the chassis gonna be.

I am by no means an expert, but isn’t the general recommendation to have at least 120 degrees of chain wrap on each sprocket? I know that what you did is probably done to reduce weight (eliminating the need for 2 other sprockets), but that one sprocket on the corner looks like it has only about 90 degrees of wrap.

If it’s not a problem (maybe someone else can answer that for me), then you can leave it alone. But if you decide you want to change it, you wouldn’t necessarily have to add the two more sprockets. You could just add a roller that would press the chain inward, creating more contact with the sprocket teeth.


Other than that, looks pretty cool. I like the placement of the motors… it’s space efficient and keeps the CG low to the ground.

Also, what’s with all the 6 motors? Is that a little excessive?

Why not just have the motors that turn the wheels between the wheels they are turning and have idler sprockets if needed.

maybe it time for timing belt- get it time-timing belt or what Doc.(john)gutman said up above

I’m sensing a little confusion…

the globe motor is the steering motor. It is not connected to the drive sprockets. The blue gearboxes are the gearboxes that drive the wheel. The steering chain has not been added as of yet.

This Crab drive is coaxial, which makes programming a tiny bit easier, and also makes the electronics guys love me more than they already do.

It looks like the “put the power to the wheels” drive transmission is modular, in the sense that it can fit on any of the 4 sides of the chassis. If this is so, good thinking! If your swerve capabilities ever give way, this gives you the option of changing your robots aspect (wide or long) without too much hassle. I’ve seen a bunch of teams attempt swerve drive, just to have something break midseason and have to revert to a skid steer base. Hopefully you’ll have enough time to work all the bugs out and get this thing rolling for next season.

Everything else looks pretty solid, but I would recommend a some kind of chain roller/tensioner as Jaine posted.

This is the precise reason behind designing the frame and gearbox as it is. If need be, we could change the bot into a 4 wheel tank drive, with minimal work required. As well as this, once the game challenge comes around, this design is very modular, to allow the best possible game manipulator to be added with as much space as needed to the frame.

As for experience doing a Crab, this is building off of our Triple Play Crab drive, which worked quite well. This design allows for easier programming (that goes along with a fun new control and OI scheme), easier machinability, and a lighter weight system. Did I mention it has a 2 speed?

Love this idea, I have some sick and twisted designs in my head that look like this. One question I have for you, ever think of making the four wheels orient in a square rather than a rectangle? That way you’d have a zero degree turning radius, and would be able to turn faster with more precision. That gearbox idea was similar to what I have thunked up, the only difference is you’ve got yours to cad allready. But I DO LOVE THE DESIGN!

How are you going to tension the chain off of the globe. Are you planing on using one or two chains. I’d like to see a picture.

Keep up the good work.

Keep posting.

For the tensioning of the steering chain (the one that will be attached to the globe) I have a set idea: Two 1/4" plates, with two delrin rollers, running the chain to the steering sprocket. The two rollers will form a triangle with the globe sprocket, making a nice tensioning system, and giving the sprocket nice wrap.

Will post picture upon CAD completion of that system

This is exactly what I have implemented into my own swerve design. I have had some bad experiences with Delrin chain tensioners, so I am wary of this approach. I may decide to use idler sprockets instead.

Have you considered timing belts? What are the pros/cons of using them.

(yeah, I’m too lazy to make a real chain in inventor (black-thing = chain))



I’ve used timing belts before… And had bad experiences with them. When doing a rapid change in direction, The belt tends to skip… aka bad stuff, bad noise, bad wear.

So chain seems like the better option… for this application, that is.

I can’t recommend idler sprockets enough, they’re easy to work with, reduce your efficiency losses, and aren’t too heavy. Here is a picture of our implementation of them on a u-bolt, this design can easily be mounted on some aluminum angle and can be used sideways (although the chain might be a bit more likely to fall off, but some sort of spacer on the u-bolt could solve that problem, if it occurs) a simple nut can adjust tension, or they can even be spring loaded, as I suggested here .