pic: low profile 3 CIM gearbox



Single speed 3 CIM gearbox with the lowest height possible. The idea of putting the output shaft in front of the CIM was inspired by 1296’s 2016 drivetrain gearbox. Weight is just over 2lbs without motors and over about 10.5lbs with gearbox. Theoretical adjusted speed is 12.66 ft/s with 4" wheels, but I can easily switch out gears to get 14.8 ft/s with 6" wheels.

CAD: https://workbench.grabcad.com/workbench/projects/gc8-CFjo2IDNWgzEcFg8GkyKWCYQy5hLaEExJhO07p7Mt5#/space/gczLctrCPkmGsjfu2z3XF1MF91d6eJfBQ7sPCuomoAPGnm

Edit: hyperlink here. I don’t know how to hyperlink in CD photos directly.

I really like this gearbox profile. Do you usually cut this out on a water jet and drill the holes separate? If it is 100% water jet cut, how do you account for the stream taper when pressing bearings and whatnot into the side plates?

Once could always have the holes cut small and drill/ream them after the fact.

Does the middle CIM have a CIMcoder on it?

This is what we do with our water-jetted plates.

I believe he cuts his gearboxes with his team’s tormach cnc mill.

Wouldn’t it be easier to just support the tip of the output shaft instead of having two bearings on it inside the gearbox?

Thanks. This is just a summer CAD thing I wasn’t planning on making, so I assumed I had all the resources in the world :stuck_out_tongue: . If it came down to it though I would probably use a CNC mill because of all the tapped holes and countersinks in this gearbox.

Yep :wink: . I did modify the housing a bit just to make this gearbox work, but thanks sharing the base CAD file! It was conveniently in native Solidworks format too so I could directly modify the features. :]

Yes, but that cantilevers the gear. And cantilevered gears, like the one on 1296’s 2016 gearbox or the one in vexpro’s ballshifter gearboxes, tend to shred gear teeth, which to my understanding is because of deflection. This can be solved by steel gears, but I thought that not cantilevering the gear was a more elegant solution. Either solution will probably work, so I picked the one I liked better and was more fun to CAD. I don’t know for sure, but I feel like not cantilevering the gear is better over the long run of the gearbox.

It is a bit more complicated than cantilever = shredded and not cantilevered = OK. It depends on how well supported the shaft is when cantilevered, how far the load is from the shaft, etc. I’d be inclined to think deflection from the loading being placed on both sides of the gear might sort of cancel out? Regardless this approach is probably fine, just be careful to make sure the inner bearing is very carefully aligned. If bolt holes are the only thing constraining that bearing in place, that might get tricky to line up as you assemble the gearbox. Especially if you are running this with a third bearing in the drive tube / plate / whatever.

Considering you didn’t pocket the plates at all, I might even go down to 3/16". That way, a bit of the bearing on the output shaft will stick through the plate and allow you to pilot the whole gearbox correctly. But then you’d need to put clearance holes as the 3/8" bearings would stick through as well. Always a tradeoff.

Not necessarily, as the gear in question (if I’m looking at the same thing) is on the output shaft. If it were in a WCD type application, you’d have plenty of deflection. If there were something else supporting the output shaft or had it connected to two or more things evenly distributed (no direct driven wheel) then you might be OK.

You have a cycle in your gear mesh that will not be guaranteed to have full contact all around. There is even the possibility that teeth won’t line up and the gearbox will bind, but with your tooth counts, the symmetry of the shaft placements, and enough backlash this is unlikely. However, I’m not sure what sort of effect an incomplete mesh can have on wear. Likely you are just transferring all the torque through one line to the output. You are certainly not going to balance it through both gears driving the output gear unless you have much much much finer control over the backlash.

In that vein, then, the pinion on the middle CIM does not have much engagement with the cluster gears. It seems this is due to space constraints, but whatever you can do to get full face width contact will help.

I really like the footprint. I’d suggest freeing the output shaft from the gearbox, but keeping the bearing that holds the backside of the shaft in the floating plate. This way the shaft can be integral to the drive rails (chain in tube, wheels and all), and the gearbox bolts on, piloting on with the bearing, with no need to ever remove the output shaft. The drive base should roll around real nice with the gearbox removed, instead of awkwardly missing that middle wheel and leaving chains dangling (or alternatively pulling the shaft out of the gearbox and leaving the gear and shaft collar loose). And gearbox removal and replacement is a breeze.

Edit: one last thing, you could swap the floating plate with the middle CIM screws, this way the floating plate is more inline with the forces applied to the output shaft by the driving gears. Gives those standoffs more gear clearance too.