D.I.Y. Electric Pedal Assist Unicycle

Logan Stafman   February 27, 2018
Over the past several months, my friend Jonathan and I built an electric assist unicycle. This design was based heavily on Justin Lemire-Elmore’s Electric Unicycle design (http://www.unicyclist.com/forums/showthread.php?t=119436) but is inferior in essentially every way. This album shows the build process in all of it’s painstaking glory. Neither of us had really machined before, or built a wheel, or knew anything about ebikes. HUGE thanks to Justin & Robbie at Grin Technologies for helping me out, and also to the guys in the shop who helped us with the machines.
The goal was to make a unicycle like Justin’s but ungeared. Instead of a throttle, like in his 2010 build, I decided to use up/down buttons as torque to allow my hands to remain free for balance. For some background, the reason this is difficult is because ebike hub motors have stators that don’t rotate, just like bikes have static axles that don’t rotate. On the other hand, unicycles don’t have a static axle that doesn’t rotate. As a result, the goal is to replace one of the bearings with a stator support that doesn’t spin and connects to the stator. This is the same principle by which Schlumpf hubs work.
Back to day 1. The motor arrives! How do I open it? I remove all the screws and find there is waterproofing silicon inside. I get basically nail polish remover and it does nothing. I get an undersized gear puller, too small and too wide of a hook to get into the motor anyways. Finally we use a chisel and the case starts to open.
The job is not over. We still can’t get the rest of the rotor out. This is a 2-pronged issue: The magnets of the rotor are keeping the rotor ring on, while the axle is friction fit to the side plate. We finally get it out with our bare hands, but it takes two of us (plus our feet). Let’s degrease the thing.
We have a stator! Removing the axle is not an easy feat though. The wires are threaded through the axle, then soldered to the Hall sensor board. We cut the wires. We use a press but it’s not working. We end up using a press with an extremely long handle and a hammer. It’s finally out. First things first, let’s remove the center. For protection, we tape up the outer rim of the stator. Then, using a drill press, we cut a 45mm hole in the stator.
We have a stator! Removing the axle is not an easy feat though. The wires are threaded through the axle, then soldered to the Hall sensor board. We cut the wires. We use a press but it’s not working. We end up using a press with an extremely long handle and a hammer. It’s finally out. First things first, let’s remove the center. For protection, we tape up the outer rim of the stator. Then, using a drill press, we cut a 45mm hole in the stator.
We bought a unicycle spindle; the plan was to machine off one of the flanges, a la Justin LE, then put some bearings around it and call it a day. Unfortunately, the spindle we bought was slightly concave, so it’s not going to work; we decide to make our own steel spindle with one flange. Since we won’t be attaching this to spokes, we don’t need to make spoke holes.
We 3d printed a mockup stator support just to see what we’re trying to do.
We 3d printed a mockup stator support just to see what we’re trying to do.
Before we can weld, we must remove the paint near the weld site on the stator:
We weld the spindle flange to the spindle shaft
We weld the stator support to the stator
We weld the stator support to the stator
We realize that our custom spindle isn’t going to cut it. Even simple square tapers are hard to machine correctly, and the whole thing is pretty heavy. We find that though it’s not a perfect fit, we can get a hub from UDC with square tapers meant for a smaller frame, then use spacers to make it fit. We machine off one of the flanges so it can fit through the hub; I take off a bit of my thumb in the process.
Time to modify the side plates! One needs an 85 mm hole to go around the 85mm x 45mm bearing that goes around the stator support.
The other side plate needs to have a bit of widening metal lopped off, then add bolt holes so we can bolt the hub to it.
stator support for the wires to get out…probably should’ve done this before welding but hindsight’s 20/20.
It’s time to BUILD IT! Not an easy task, as it turns out. First we get all of the components together.
First (unpictured), we widened the channel for the wires with a dremel. We also shave some off some of the supports on one side plate, as they get in the way of our bolt holes.
After re-greasing the rotor, we get it together! The bolts coming out the side will be cut off once we perfect the alignment with the unicycle fork.
BUT bad news…it’s rubbing and won’t rotate freely. Why?!? Even though I’m smiling here, I’m not happy, this motor won’t turn. I try many things over the next few months to try to get it to spin all the way around, but it simply won’t; the welded stator support is out of center. I could try sanding down the rare earth magnets but I’m not hopeful; I’m a bit lost on this project. Luckily, the motor I REALLY wanted to use all along was 50% off due to a slight manufacturing error, so I snagged one. It’s time to start over with this new motor.
We start by opening it, but are much more careful this time. The lightweight design of the new motor also makes it feel more fragile, and we’re paranoid that we may never get it to spin right. I come up with a new design that doesn’t require us to open the motor at all. Since this motor is a thru-axle motor, I only need to connect the stator to the unicycle fork by extending the thru-axle, which is much simpler than the first motor.
interface for a torque arm. My design idea is to use that interface to create the stator support. However, it will need to be as thin as possible to fit into the 100 mm dropouts of a unicycle.
A few hours in the shop and we’ve created the stator support connector. I email the folks at Grin and they’re happy to supply me with the design specs for the spline so I can recreate it via CNC. As I’ve never used CNC, Larry, who runs the shop, agrees to CNC it for us.
We come back a week later and oh no Larry has done the reverse of the cut we wanted…unfortunate misunderstanding!
Another few hours in the shop, and we’ve created another stator support, another week later and Larry has CNC’d the splined connection for us.
We drill bolt holes into the piece so it will connect to the motor. We also cut a groove for the wire to escape through.
We cut more holes in the bearing holder part of the stator support for the screws to access the holes through. It fits wonderfully!
the thru-axle and the disc-brake caliper, to fit into the 100mm dropouts. Were I to do it again, I would have left most of the disc-brake caliper to give us more material to bolt the flange of the spindle to, and bought a 127mm wide spindle and unicycle frame; as it is, we ended up using worryingly short bolts.
A new problem has arisen: the screw heads stick too far out of the flange and scrape against the bottom of the unicycle frame. We consider a few options
We decide to carve grooves for the heads of the screws and still have about 2mm of steel left.
with a bit thicker one and then it works.
We put it together and there is some rubbing. We used bluing to diagnose it, and it’s a combination of two things. First, our spindle-to-sideplate bolts are too long and rubbing against the stator. Second, the spindle widens by 1mm near the flange. We fix these things on the lathe.
We fit it in the fork!

 

 

 

 

 

 

 

 

 

 

 

 

The rim the unicycle shipped with is 36H while the hub is 32H, so I get a new rim after getting the runaround from some local bike shops. We learn to lace and true our first wheel!
We throw the inner tube/tire from the original rim onto the new rim. We throw some 95mm cranks onto the hub, and steal the pedals from the original unicycle as well. Not shown in the picture: the batteries. I finish setting up the Cycle Analyst and she works!
Me with the beast! This is just after a ride test in the dark. Will give it a bit more of a try tomorrow in the light. It’s heavy which makes it a bit hard to mount. I’m also terrified of dropping it which is making me scared to turn up the assist.
Jonathan showing off our finished product.
Original article posted by Logan Stafman on imgur.com HERE
« »