I’ve been fighting with my robot’s drivetrain for a few weeks, and to be honest, I’ve been losing the battle against traditional belt systems. In FTC, we have to deal with very tight space constraints, specifically fitting everything inside a 43mm U-channel. My goBILDA motors are fast at 5000 RPM, but turning that into 435 RPM at the wheel using belts is a headache. Either the gear ratio is too small, or the pulley is so big it hits the floor. Plus, belts tend to strip whenever the robot hits a wall. I wanted something better, so I started working on a 3-phase cycloidal drive. It’s a completely different way to move power. Why I’m using a 3-phase cycloidal drive , Traditional cycloidal drives are cool but fragile. They usually rely on small internal pins and bearings that are hard to line up. This “inside-out” 3-phase design is different because it uses an epicycloid curve to grab pins on the outside of the plates. This gets rid of those tiny, weak pinholes that break in 3D printed gearboxes. The other cool thing is that it uses three plates, each staggered 120 degrees apart. This keeps the plates supported and stops the mechanism from getting stuck in dead zones where it would otherwise lock up. It’s a smooth way to get a lot of torque in a small package. getting the CAD right hasn’t been easy. I’ve been testing different pin counts and eccentricity settings. I’m aiming for an 11-pin setup to get a 10:1 ratio. If you push the eccentricity too high, the gear profile loops over itself in the software, which means it won’t work in real life. Finding the right balance between high torque and making sure the part can actually be manufactured has been a big learning process.