I know this isn’t exactly the place for FTC discussion, but I just wanted some feedback on this “gearbox” my FTC team is using on our robot from the people who actually design gearboxes.
So in the picture, you can see three AndyMark Neverest gearmotors on the left. They’re all mounted to an 1/8" aluminum plate, and each have a spur gear fixed to the shaft. The other end of the shaft is supported by a bearing in a piece of aluminum plate parallel to the first. All three motors drive another large gear, which sits on top of the assembly.
Is there anything inherently wrong with how these motors are being driven? Would there be a better/more efficient way to make this gearbox, or just any advice on improving it?
Hmm… Interesting way to do this. I would not design this gearset in this fashion, but I can’t say for sure that there is anything wrong with it. Typically, I would have the three input gears mesh with the one output gear, not an input meshed with an input meshed with an input meshed with an output. But, like I said, I can’t say for sure that this is wrong in any way. Though, wouldn’t a bag CIM with a 100:1 planetary reduction give comparable speed and torque to your current solution with much less weight and complexity? Just a thought.
BAG motors are not legal for FTC, though you could remove the motor from the AM gearbox and put it into a versaplanetary gearbox wth a 100:1 reduction.
Addressing the original question, I don’t believe that this setup would cause you any issues as long as you keep in mind that the middle motor will have to spin opposite to the top and bottom motor for this gear train to function. As long as you mind that I think you’ll be okay. If you don’t mind me asking, what sort of mechanism is this gear train powering? Using three motors on one mechanism when you’re limited to 8 motors seems a bit excessive. Perhaps you could use a larger reduction to accomplish the same task?
For the OP: With the power levels of FTC motors, I don’t think so.
In FRC, doing this is generally discouraged because CIM motors are powerful enough that putting the force of three of them on one pinion can cause that pinion to break or the motor shaft to bend.
As for this being overkill-- it might be. Or the team might be going for a very fast hang with a heavy robot. I don’t know if the OP’s team has done the math, but you can calculate the power/motors needed to lift your robot using the JVN calculator and inputting the NeverRest motor specs.
Ah, I see. Sorry, I posted before checking the FTC rules. But I’m sure there is a motor/gearing option that gives you the speed and torque you want with a simpler solution, and less motors.
NeverRest 20/40/60 (all the same motor with a different gearbox attached)
Tetrix 12v (almost complete garbage)
Matrix (also pretty bad, but I don’t have a ton of experience with them)
VEX 393 (completely underpowered here)
A variety of servos (not what we’re looking for)
If they’re 40s in there, they might be able to lift with 1 or 2 60s depending on their robot’s weight, but that also might be slower than they want. It’s a pretty limited motor pool for FTC, especially compared to FRC. I don’t doubt they might be able to crunch some numbers and come up with a “better” solution, but we don’t know what numbers they’ve already run and exactly how fast they need to pull up how much weight.
As for improving it: While this might be fine, generally I prefer to place the motors as few stages away from the shaft they’re driving as possible. If you can, I’d second the motion to drive the winch gear with each of the motors directly next to it instead of driving them through each other. This is generally good practice for gearbox design.
It looks like you might have small bearings in the gearbox plate-- this is good, keep it up. For something that gets a lot of stress, you want to avoid cantilevered shafts, so supporting both sides is good.