Here’s a model of a Fisher Price gearbox I designed today for use with a potential arm design. It uses two sets of Toughbox gears and the CIMple Box gears in order to save as much money as possible. Right now the gearbox with motor weighs just under 5 pounds.
With the Planetary the box has a 780:1 reduction. With a 16:54 sprocket reduction that raises the ratio to 2633:1. That should be enough to drive an arm with 7 lb load at a comfortable 45 degrees per second.
This box is relatively easy to machine, especially if you take out the pocketing (though that brings the box closer to 6 pounds). Combine that with the recycled gears and recycled AM Planetary, and the only cost to Shaker would be the small amount of 1/4" plate and 3/8" hex stock, which is really nothing compared to an unreliable set of Banebots gearboxes with multiple extra large sprocket reductions. The weight isn’t the best, but those gears could be cheesed to bring the box to 4 pounds if we’re desperate.
I like that. What type of sprockets do you plan on uses when attaching that to the arm?
I know H was thinking about using a Giant sprocket which adds bulk to the system. This is slightly more complex but it will still be easy and it will be way cleaner. If we use Al box for our vertical post , we could mount it directly to the beam, like the drive train.
Nope, none at all. The 1/4" is heavily pocketed on the inside and is mainly used as a shim to prevent that bearing right next to the Planetary from sticking out. I figured that worse comes to worst, a little overkill on such a vital part of the robot wouldn’t hurt.
This is designed to interface with a 16 tooth hex sprocket to a 54 tooth sprocket, both available in 25 chain size on AM’s website. I think if I did a bit more work we could squeeze a stage of reduction out of the gearbox by using a larger sprocket reduction - I can’t figure out how to right now but that could save a lot of weight. Either way, this puts less load on the chain by doing more “work” in the gearbox.
I don’t see why this couldn’t be integrated into the side of the arm - the only thing is that the 50 / 56 tooth gears are more than 2 inches wide so they’d stick out.
Pocketing the middle 50 tooth gears brings this box down to about 4.2 pounds.
No idea if the lightened gears can, but I would be very surprised if the .4 inch steel gears from AndyMark could not handle the load the FP puts on them, considering people lifted robot-size loads with those gears last year.
To be safe I’d personally forego lightening the final gear even though it’s the biggest one.
Hopefully the arm can be balanced with surgical tubing in all states but it’s “reverse dunk slamma jamma mode”. For the rest of it, the arm is a surprisingly light load on the motor and running it at ~5% power or so to hold it up should be very easy, especially with balancing reducing the already light load further.
So a nice and easy way to reduce weight is to replace most of the gears with 50 tooth AL cluster gears. I assume for the first two reductions this won’t compromise gearbox strength - but how can I check on the third? I swore I learned how to do this in Strength of Materials but it’s escaping me right now.
For high reduction arms, depending on how they are loaded, it’s possible the motor never even comes near stall torque, so aluminum may work fine.
Also, if you space your plates 1.125 apart, you can make this same gearbox with ZERO custom shafts, and the use of no spacers, by utilizing the toughbox and cimple box gears. It’s a trick we’ve been using for three seasons now.