Efficient Wooden Gear Systems

I’m thinking about a project that involves a large number (dozens) of wooden gears (think Easel gear generator app), all connected, driven by a motor. It doesn’t really do anything, just sort of a shape made of a big set of gears, so the whole thing’s in motion, but it’s not going anywhere.

I was planning to use 8x22mm skate bearings, press fit into a recess in the hub of the gears, on dowels fixed into the surface. I’m worried about the stability of that system and about the cumulative effect of friction from all of those gears. Does anyone have a more elegant solution?


yeah, look up wooden gear clocks also for some inspiration. I don’t think with bearings and being motor driven that friction should be a problem. As Angus said, precision of gears and gear spacing will be bigger concern.

If there’s enough initial torque on the input side, the cumulative friction along the drive chain should not be an issue. As mentioned above, if this is a low RPM scenario, then even better. Just make sure the axles are spaced properly to reduce slop where teeth mesh. Some people prefer plywood for dimensional stability, some prefer hardwoods for compression strength, some select woods that are intrinsically oily to add natural lubrication to the gears. As a mechanical engineer, I love anything gear-related. Please post pics of your progress/project.

Hey I’ve been working with gears on the xcarve in different iterations for about year. I came across this gear generator that is great if you need to control different aspects of them http://geargenerator.com which can help you get them really precise, and its nice for visualizing how they interact speed wise (which was key for me since I needed to slow down the speed between the crank shaft and the cam shaft so kids didn’t jam up my kinetic work). You can download them as svgs.

I do an initial MDF cut to make sure all my movements are what they need to be and play with it since I have to see stuff working to understand it (I wish i could conceptualize it better, but for me a bit of trial and error is the better route, if I try to do it by math I’ll mess it up).

After finalizing the design I do the final cut out of nice 3/4" plywood to avoid voids and because its prettier when finished. I made all of the components in my last piece from scrap I had accumulated around the studio, so even if you use nice wood you can get a lot of function out of a little material as long as you do your testing on something cheap first. After initial sand I use sanding sealer, do a 220 sand, then another coat and 220 sand and then screen print my graphics and urethane them 4/5 times (the work is designed to go outside so sealing is crucial. If I’m not against a deadline I’d probably do double that, but I also need the surface of the cams to be pretty slick to function reliably so the more coats the better)

Because I’m making cam shafts I have 1/2" aluminum rod going across the support box and through the components, and use marine epoxy to attach a collar with a set screw to the gears and cams (it has a bit more flex than normal epoxy and is nice and weather proof). I also use blue locktite in the set screws because kids are brutal as f*$&! with kinetic work.

here’s a link to the development of the work if you’re interested.

Very cool stuff - thanks for sharing!