Cyclodial Drive

Status: completed
Progress: 100%

Build Log

As a good amount of my projects here, This is writen after I "finished" this project. so this is more of a post dev log.

I was watching a Youtube video by James Bruton about a Cycloidal disk drive for this amazing robot dog he built, and I fell in love with how the 2 gears mesh and look when in motion, so I set off to make one for my self!

When I started this project I had absolutely no knowledge about gears and how they worked. heck, I had zero mechanical engineering knowledge or experince at this level. So being my self, I spent a few weeks teaching my self how gears work and the math behind how they work. turns out its a lot more dificult than just guessing random numbers and hoping.
Here is an animation from Wikipidia of what it looks like with just a single gear, the one I made has 2 gears. we use 2 gears here with 180 degree offset to remove as much vibration as possible.

GIF

Now isnt that the coolest thing you have ever seen! wait untill you see what 2 gears look like!

GIF

At first I thought I needed to hand draw all of the gears in a sketch and go from there, but I found a plug-in for Fusion360 that does all of the manual work, I just need to feed it the correct numbers and it will make the gears I needed.
After watching a few videos and reading up on things, I set off to make my own. It took many attemps of making the gears, making them mesh correcty and then just when I thought I was done, I'd have to redesign everything all over again because it was ether too big, small, and or expensive for me. But after a good 10-20 redesigns I landed on somthing that I was happy with.
Looking back on it, There was absolutely a more cost effective way I could have made it. Who would have thought bearings get REAL expensive REAL fast! the entire thing was 3d printed besides the bearings and metal stand off's
given that this was my very first time making this I started desinging things in the wrong order. I just came up with numbers on the spot, I should have found some hardware that I liked and fit what had in mind. the body did not really matter beacuse it was 3D printed, filament was cheap enough it did not really matter. The bearings is what killed me.

Just like every enginer has said "it works in CAD" I thought I would just need to print off a few things and id be done! but nope, its never that easy.
Tolerances was another one of my killers. I was using what ever filament I had loaded up in the printer at the time, and because it was in the prototype stage it was my older filaments. The filaments were full of water and diffrent brands so one print could be too big by 0.1 mm and the same exact print a few days later would be smaller by 0.1 mm and as a result made my troubleshooting so much more difficult and confusing. after a while, I figured out it was the filament it self that was causing the issues.
After many reprints and attemps I got somthing I was happy with and was at a point I can call finished.