Tuesday, January 5, 2010

"The problem with your blog is, I can't understand any of it." -Ken Colton

Happy New Year! Time for my first post of twenty-oh-ten. My New Year's resolution is to work on my blog post explanations a little, if for no other reason that so that my brother (who made my awesome site banner^^^) can follow along when not reinventing the internet. Shouldn't be that hard...I just need to stop talking about software, right?

I'm back in Cambridge and back to work, getting ready for another fun-filled year of 2.007: Design and Manufacturing I (OpenCourseWare from last year for those of you working on your FREE MIT education). Although the stardard box of raw metal and plastic stock is pretty much unchanged, we have some new and exciting additions to the kit this year. For one, everyone gets there very own small fireworks Lithium Polymer battery and charger thanks to dirt cheap overseas manufacturing. There's a whole new line of servos and gearmotors this year, too, which means that I may finally be free of the curse of the Tamiya 72001 Planetary Gearbox Kit.


How about Project Updates?

Well I must admit I've been spending most of my project-time on making this microcontroller do impossible things that it wasn't designed for, like simultaneous field-oriented control of two motors. You might think I'm wasting my time, but allow me to motivate this with a scenario:

You're stranded on an island after a plane crash, but after hours of walking the shore you find a boat. This boat happens to be powered by two brushless motors, and it happens to come with all the batteries, power electronics, and maps and charts you'd need to make it home by dinner.

Motivated now?

But there's just one problem. Your boat is missing a controller. Your laptop, cell phone, and iPod were all destroyed in the crash. But wait, in your back pocket:

Yes, a TI-83.

That's the type of controller we're talking about here. A graphing calculator. Not even a new one. That's roughly the processing power equivalent of the MSP430F2274. But hey, it's all you got. At least it's better than an Arduino. Oh, and also, it isn't sufficient to control one motor, because clearly the boat will go in circles. So now you see why this is important. And if you're already putting in the effort, you might as well make it quiet.

Well if I was stranded on an island, I would have starved to death, because I've been working on the software for this thing for two weeks now (not including the holiday break). Software, though, is a thankless job that produces literally no tangible evidence of progress, which is part of the reason why my recent blog posts have degraded to pretty much indecipherable gibberish (that I assure you is cleverly-constructed and efficient code) .....

BUT NO MORE. Next, I will show an actual, measurable result. And it's more than just that the motor runs more quietly...that's a nice bonus, but it's not the ultimate goal of the upgrade. Look for a post by the end of the week.

And how about the other projects? Well, I plan to do some serious work on the axial motor this month. Even if it's doomed to failure, I'd like to get there quickly. First step is to make some combination hub adapters / bearing retaining plates. Then, spacers that will give the rotors at least some small chance of being aligned to each other. Then, assembly and the single-tooth test, which is about the most boring way to test a motor ever but gives me all the information I will need and costs nothing to do.

Those are the two big ones. Might throw in a brand new mini-project along the way...depending on how things go. 2010 begins.


  1. hello... hapi blogging... have a nice day! just visiting here....

  2. great work!, it is posible to control a scooter motor BLDC with trapezoidal conmutation without make noise?

    1. BLDC motors with trapezoidal back EMF, driven by square wave voltage, will always have some commutation noise. How bad it is depends on a lot of things. A fractional-slot motor (e.g. the common 12t/14p) will usually be quieter than an integer-slot motor. Anything that rounds off the corners of the back EMF, such as using a slightly larger air gap, will also make it quieter, at the expense of some peak torque.

      It's possible to drive a motor with trapezoidal back EMF with sinusoidal voltage, which will usually make it much quieter. But sinusoidal back EMF with sinusoidal voltage drive is the way to go if you want it to be virtually silent.