I sadly can't report any new progress on the mini ATmega board due to travel and busyness, but I have recently made some progress on a new project:
This is a first attempt at making a brushless hub motor from (almost) scratch. Its inception began when I found an intact old copier in the school electronics dump area. Remembering this really great Instructable, I decided to rip it open in search of some motor stators. I got a medium brushless and small brushed motor out. After much work removing the brushless motor casing:
 |
| Stator |
With this crucial component in hand (for free), I decided that there was no excuse for not attempting to rebuild a hub motor around it. For anyone not familiar with brushless or hub motors, I'd highly recommend looking at the link above for a more thorough introduction. As far as my purposes are concerned, a small hub motor like this is essentially an outrunner electric motor that is compact and robust enough to be mounted as the hub of a wheel. It skips the mechanical commutator found in brushed DC motors in favor of a winding that uses a sensor-based or sensorless brushless motor controller. This direct drive application usually comes with a loss of torque because there is no room for any kind of gearbox.
The theory behind the motor's mechanical construction is much simpler than the electrical side, so I started with that. I modeled the stator and an approximation of the space taken up by the coils first. For the stator shaft, I decided to cheat with a two part piece. Instead of one shaft going through the stator's center, I modeled a half on each side with holes for bolts to compress the stator between them. This obviously wouldn't work if the motor was expected to have the weight of a rider on top of it, but this little motor is never going to be ridden on anyways. This allowed me to avoid the challenge of finding a way to constrain the motor on a slide-in shaft. I'd never had the opportunity to do any work on a metal lathe before, so there is an ongoing theme of cutting corners to make the machining easier throughout this project.
Next was fitting the magnets. The magnet ring runs along the circumference of the stator, and is carefully fitted to the inside of the motor can. This is the part that will (optimistically) experience rotation. While SuperMagnetMan didn't have very ideally sized neodymium magnets for fitting around the stator, I found that by using smaller magnets, in sets, I could get around a 73% volume fill (the percentage of the motor ring volume occupied by magnets). This isn't a great fill, but I decided that I would rather lose some torque than pay for custom magnets. The magnets gave me the ID of the motor casing, which gave the OD of the motor's end caps. The end caps house the bearings that allow rotation about the shaft and connect to the motor can. In another ease-of-fabrication design compromise, I decided not to go with a threaded end cap design. Instead, the end caps slide inside the motor can, and are connected through many tiny screws tapped radially through the motor can.
The yellow thing represents the shielded bearing. I bought longboard bearings inexpensively from VXB bearings (which were really nice for their price). Now with the finishing touches:
The motor can and screws are the only parts of the assembly not made of aluminum. The motor can needs to be a ferrous metal in order to contain the magnetic field created by the magnets. A few metal and hardware orders later, and I began a "learn as I go" policy to making the parts.
 |
| Notice the spare pieces for when I mess up |
I'm using a Harbor Freight 7x10 mini metal lathe with accessories from littlemachineshop.com for most of the operations here. Experienced machinists please forgive the quality of the parts seen below.
 |
| Drilling the starter hole for an end cap |
 |
| End cap with bearing |
 |
| Axle piece (bolt holes not yet drilled) |
 |
| Axle test fit |
I still have a lot more to make before the pieces can come together, and it seems very possible that I will not finish before college begins again. In that case, this project will gather dust for a few months before it can be tested using Metal Scooter's controller and battery.
Speaking of the electronics side of the motor, I'll go over my thoughts about the windings. To again heavily paraphrase many pieces of the excellent Instructable referenced above, this kind of electric motor is terminated in three wires that strategically connect the coils around the stator. By alternating the flow of electricity between these wires with the right timing, each coil can attract nearby magnets and produce rotation. By knowing the physical characteristics of the motor, it is possible to estimate the torque produced through the following equation (not accounting for the air-gap or magnet fill inefficiencies):
Torque = 4 * teeth per phase * # of wire turns per stator tooth * magnetic field strength * stator length * stator radius * current through motor windings
Just for fun, I decided to find the torque needed to push me up a 15 degree slope and back solved to find the amps that would needed to be drawn. The answer was unfeasibly many, given the gauge of wire that would be needed and the physical capacity to fit that many turns in the tiny stator. While it might be possible to get going on a flat surface, it's just another reason that this motor isn't going to see use in a vehicle.
Without a torque goal in mind, I intend to just use whatever magnet wire is most practical for winding around the stator. I also have to keep the diameter of the axle exit hole in mind, because it places a definite limit on the max gauge of the three wires. I'll document how it works out when I get to that stage.
No comments:
Post a Comment