Monday, January 18, 2010

Arrrrr, see?

Pirate radio...control.

I've been wanting to get my hands on a serious RC vehicle for some time now. Here it is:

It's a Team Associated TC4, 1/1oth scale 4WD remote-controlled car. Ignore the absurd-looking battery pack; I have no idea where that came from. I was lucky enough to get some drive time on the one nice day of weather we've had in the Boston area in the last two months. Conclusion: This thing is fun. I have literally zero RC experience (although I have driven robots) , but this seemed to be fairly easy to get the hang of, at least up to moderate speeds. Maybe it's the 4WD.

The car itself is mind-bogglingly impressive in mechanical detail. It's a shaft-driven 4WD with two ball differentials. There is no center differential, but the front and rear can be adjusted for more or less slip. It's got four-wheel independent suspension with adjustable springs and shocks. Lots of tiny screws, ball joints, linkages, all with tweakable settings. I don't even know where to start...

...Oh wait, yes I do: the motor.

This is the stock motor, a Reedy Radon brushed DC motor which has but one spec: 30,000 RPM. Thanks a lot. If there's one thing I hate about RC components it's the downright lack of real documentation. Coming from the world of robotic and electric vehicle components, it bothers me a lot to have a motor with no specs. So in case anyone else in the whole wide web is wondering, I decided to go ahead and measure the Reedy Radon using standard DC motor specifications: torque constant, back EMF constant, and winding resistance. Here ya go:
  • Motor: Reedy Radon
  • Type: Brushed DC
  • Turns: 17
  • Winding Resistance: ~69mΩ
  • Torque Constant: 0.00262 N-m/A
  • Back EMF Constant: 0.00262 V/(rad/s)
  • "kV": 3,650 RPM/V
(The resistance is only approximate because brushed motor resistance is sort-of hard to nail down. It depends a good deal on the brushes.)

And yes, if you use a fully-charged 7.2V NiCad battery pack, this nets you about 30,000RPM, which with the stock gearing is about 30mph. If you use a 9.9V A123 26650 pack, then you are just insane. And yes, this is quite a lot of punch for a three-pound car:

Cue gratuitous use of high-speed cameras.

It would be hard, then, for me to say something bad about this motor. In fact, I love brushed motors and for $23 this seems like a terrific option with plenty of power. But seriously do you think I could live with myself if I didn't at least try putting a brushless motor in it? The problem is, brushless motors for RC cars are expensive. Here are just a few examples:

CMS 36-4600 (Castle Creations, $100)
Velineon 3500 (Traxxas, $75)
Reedy 5000kV (Team Associated, $50)

Also, they are all specified by the wonderfully inverted "kV" rating. Higher kV means more RPM per Volt, so they go faster, right? Well...not really. That's only true if you're stuck with the voltage you're given. I'll take a low kV motor and a higher voltage any day. Conventional motor wisdom says this is almost always more efficient. In the age of lithium batteries, a 20V RC car is not unreasonable.

But still, I'm not paying $50-$100 for a motor that's easier to manufacture than the brushed one. And don't even get me started on controllers....RC controllers are amazing devices with unparalleled power density, but some of them are way overpriced. Interestingly, we live in a global economy and there are Inexpensive Chinese Brushless Motors (ICBMs, © 2010 Charles Guan, used without permission) on the market. I snagged this one for $15.95, minus my $2.99 Hobby King store credit (which I got for browsing but never buying anything). So what exactly does a $13 brushless motor look like?

Like this.

You're probably saying: "It has a reflective purple sticker and says 'HIGH TECH SUPER' on it. It must be a piece of garbage compared to those other motors." Well I don't know anything about those other motors because the RC companies don't publish any useful information about them. At least this motor is cheap enough for me to buy, ship, and test. Most importantly, I would like to know the torque constant and winding resistance so I can compare it to the stock motor. So I hooked it up to a dynamometer:


Actually, I later figured out that 3.2mm motor shafts fit nicely into a 1/8" dremel collet, so I could do some higher-speed testing. The procedure is really simple: spin motor, measure back EMF on oscilloscope.

It's definitely not a trapezoid (ideal BLDC) or a sinusoid (ideal BLAC). It's more like a sinusoid with fifth harmonic added in. Try it. It's a two-pole motor with who-knows-what-kind-of-magnetization, so the odd shape is not surprising. Rather than fret about it, I chose instead to just use the RMS measure and pretend it was a sine wave. Using some motor math tricks, I can get a decent torque constant estimate this way. I can also measure the resistance very easily. Here are the results:
  • Motor: Hobby King 13.5T RC Car Motor
  • Type: Sensored BLDC
  • Turns: 13.5
  • Phase Resistance: 16mΩ
  • Torque Constant: 0.00314 N-m/A
  • Back EMF Constant: 0.00314 V/(rad/s)
Okay, 16mΩ is low. I'm pretty much sold on that spec alone. That's where motor heating (I²R losses) comes from, so the lower the better. And the torque constant? That's for sinusoidal drive, and it's higher than the stocked brushed motor. More torque per Amp and less I²R in the same size means better motor.

What about kV, though? I didn't list it because it has no clear meaning in sinusoidal drive. But I can say that I think it's significantly lower than the stated value of 3,150 RPM/V. Meaning, more torque, but higher voltage required for high speed. That's fine for me, since I wouldn't mind using a 6s (19.8V) A123 26650 pack. And my controller won't even run below 18V.

Oh, right, I forgot to mention the other important difference. Unlike the three more expensive brushless motors of the same size, this one is sensored. It has integrated Hall effect sensors that detect the position of the rotor, same as the scooter motors and the majority of small EV BLDC motors. Not only does that work perfectly with my controller, but it will give better starting torque.

This is one insane deal of a motor for $13. As long as it doesn't explode or something. I know you won't believe me until I actually put it in the car and demonstrate, so I should get on that. Hopefully by then the snow will melt and I can drive again.

No comments:

Post a Comment