June 6, 2014

Spontaneous Hub Motor

A couple weeks ago, Michael's 2-hour electric scooter made me sad about my lack of small, practical electric vehicle.  As a result of the piles of motor stators and magnets lying around, my recent acquisition of some scooter wheels, and quality 1:00 AM logic , I came to the conclusion that I should build a hub motor to fix this problem.

The first problem I ran into was that there were no steel pipes of appropriate diameter from which I could make the motor can.  This bit needs to be steel for its magnetic properties.  However, the MITERS drawer of Absurd Round Things had some steel gears in it, which were similar diameter and also had a face width equal to the length of my magnets:

I think this makes me a terrible person...
I machined the giant gear into a thin ring on MITERS's new Big Lathe.


I carefully glued the magnets in place with thickened epoxy.  I went with a 16 magnet configuration, rather than the usual 14 for a 12-tooth stator.  This gives the motor massive cogging torque, but should also up the torque constant a bit.  Anyway, human inertia should smooth out the effects of cogging.


I then bored out a 125 mm scooter wheel and pressed it onto the can.


I made endcaps out of a big section of aluminum round.  I could not find a matching set of appropriately sized bearings, so they are mismatched.  The endcaps have a small lip around their edges, which fits around the steel can.  The outer portion of the lip compresses the tire, while the recessed portion clamps onto the can.



Next came an axle to hold the stator.  The stator is retained to the shaft by a bolt at its edge acting both as a bolt and a key, as well as epoxy.  The ends of the axle are squared off, since the axle-frame interface has to withstand motor torque.

This particular stator was harvested from a large copier, and is almost exactly 1/3 of a 80-100 "Mellon" stator  in size.


The endcaps are clamped together by six M3 screws around the edges:


The stator was wound with 2 parallel strands of <> magnet wire.  The 12 slot 16 pole motor configuration means an ABCABCABCABC winding pattern.  Some rough calculations gave me an estimate of 30 turns per tooth to get the torque constant I wanted for 40V operation.  I ended up having a bit extra space on the teeth, so I put on 35 turns per tooth, with room to spare.


Here's the completely wound stator.  I originally wye terminated it, but the motor controller I was testing with had trouble spinning up, so I switched it to delta for now.


Next things on the agenda: