This project is currently in development. Stay tuned for updates!
¶ Overview
¶ Motive
Before I moved to Seattle, Unicycling around was always pretty easy, as Minnesota is not very hilly. Although it's possible to unicycle around on even Seattle's steepest hills, it can be pretty exhausting. Having some electic assistance will also allow for the use of shorter cranks, which are desirable for the long, flat sections of a ride.
¶ Prior Work
- Justin Lemire-Elmore's Ultimate Commuter Unicycle (Made by Grin Technologies)
- Logan Stafman's Electric Assist Unicycle (Based on the above project)
Both of these projects use a modified bicycle hub motor, which requires extensive modification to both the motor and the wheel/frame of the unicycle being converted. I want to find a solution that provides similar functionality, but requires minimal changes to the wheel/frame of the unicycle being converted.
¶ Requirements
- Must opperate at a speeds up to 8 m/s (29 kph).
- Must be able to produce at least 400 N of linear force against road at stall.
- Must be able to supply and regenerate at least 600 W of mechanical power (peak).
- Inverter must use a torque control loop rather than the typical speed control.
- Battery must have enough energy to assist an 80 kg person up a vertical 200 meters.
- Must tolerate rain/wet conditions
- Must not interfere with normal, human-powered riding when electronics are inactive.
¶ Design
¶ High Level
To achieve an easily installable design, a "rim drive" configuration will be pursued. The inpiration for this design is the VELOSPEEDER, a driven rim electric assist device for bicycles. (see below)
This configuration allows for minimal modification of the existing frame, wheel, and hub. It also allows for the use of a disk brake, which would otherwise interfere with a hub-based motor design.
¶ Motor Selection
For reference, the higher-power version of the Velospeeder appears to use a 3135 sized motor.
Assuming we'll be using a motor on both sides, A good choice of motor is a 3542 sized, 1000 KV outrunner like this one. This is definitely on the larger end at 750 W each, should look into smaller options. One advantage of going with an oversized motor though is not haveing to worry as much about excessive axial force though, as the bearings will be larger.
Could also go with a "2216" sized motor like this one. This motor is the one I've chosen to go with for now. It can source 0.25 Nm of torque, and has an 880 KV.
¶ Battery
From requirement 5, we can calculate that our battery must have at least 156800 J, or about 43.6 Wh of energy. Assuming we use a 4S Lipo battery pack, that means we need at least a 3 Ah (3000 mAh) pack. This is actually a pretty common size. Assuming a 60C discharge, the battery easily meets requirement 3 for discharge (up to 2.6 kW), but the suggested 1C charge rate puts us at only 43 W of regenerative braking power. May have to look into alternate battery chemistries for improved regenerative braking performance.
This battery Looks about perfect. 5000 mAh, 4 cell Lipo, 1480 W constant discharge, 370 W charge. It doesn't quite meet (3), but I don't think there's an easy way to get better regenerative performance without just adding more batteries.
¶ Motor Controller
The most relevant off-the-shelf products for this are electic long-board/skateboard controllers. Lots of them have dual controllers built in.
By far the best controller I've found so far is the Flipsky Mini FSESC4.20. It's a 50 Amp controller with an insane amount of software customizability, and it even has a CAN interface that could talk to a custom system controller board.
