Electric Assist Unicycle at Maker Faire, this Sat. and Sunday

[nathan]

Justin, thanks so much for bringing the cycle over last night. Riding it up the hill on my street was the easiest climb of any hill anytime on any unicycle for me. It's a great creation. And it was great to meet you too. See you at U Games this summer!

---Nathan

[Justin LE]

Originally Posted by Chrashing

I hadn't noticed that the motor stator was fixed to the frame (your drawing clearly shows.) So now I realize the motor truly assist the rider. It is fascinating.

Sorry, that I have been so slow to catch on to this, and so quick to add confusion to this thread.

Hey, no worries and glad that you were able to make sense of it all in the end!

Justin

[Justin LE]

Originally Posted by dangerdog

Thanks corbin for the brief ride report but I would love to hear more about this creation.Justin,Corbin anyone please tell.

Ha ha, you'd think I'd given a mum order for people to keep buttoned up!

Anyways, the ride experience ended up being pretty much what I was hoping for, in that you could cruise along with the motor giving a huge assist behind the scenes, but not overtly affecting the pedaling and handling experience of the uni. What happens is that you get acclimatized to it, so you can zip along easily up a hill, then let go of the throttle thinking you don't need it anymore and fall flat on your face because you didn't realize just how much the motor was boosting until letting go.

The braking was really smooth, but unfortunately just a binary on/off lever rather than proportional control. So down a moderate hill it would peel off about 70-100 watts and put that into the battery and relieve your legs of braking duty, but on a steep hill it wasn't nearly enough regen torque to do it all electronically. But this is something that can be changed easily with a different controller.

Also, the throttle control was a lot more touchy and sensitive than you would want on a unicycle, so you had to be careful to ease on the throttle gradually or else the sudden kick from the motor could throw you off. The plan was to play with a capacitive filter on the throttle line which will give a better soft-start to the system, but didn't get a chance to do that yet.

Here are a few pictures from the fair. The flat parking lot busy with people wasn't the ideal testing grounds, no hills and not enough room to get it up to full speed, but those who were already familiar with big wheels had no issues:


Corbin


Trevor


Unknown

[scott ttocs]

Very Nice!

Will you be traveling anywhere else to put on demonstrations?

You could build a motor control circuit that only allows gradual changes to smooth out the transitions. It sounds like that is what you have in mind. I know my Prius has similar controls to keep me from punching the gas and wasting fuel.

Scott

[dangerdog]

Justin,by all accounts your power assist uni concept is a success.The mechanical part is working perfectly by the sound of it and you just have to make the control more user friendly.The choices for control electronics is almost endless and I'm sure you probably have the final control circuits already built in your head.Great job,I look forward to more developments.

john

[redwelly]

Wow. I started reading this thread thinking that it couldn't possibly work well enough to be useful, but Corbin and Nathan's comments amazed me.

That is a seriously nifty toy. Great work!

Next step: shiftable-on-the-fly gearing with handlebar controls and you're in business!

The machined bearing side where the cables show how even a non-electric techno hub could have (multiple) gears controlled by handlebar cable-shifters.

Sam

[ħǻřрέŗ]

Justin-

That is a fantastic project. Great work and congratulations on what appears to be a complete success. Do you happen to have the full specs on the motor you used? I was interested in the complete wiring diagram and the mechanical drawings of the rotor and stator if they're available someplace. I originally thought that you had wound the coils yourself when first looking through the thread (at which time I was stupendously impressed).

Did you design any part of the control circuit or modify an existing controller? Soft start features are usually just a simple time constant in the controller somewhere.

Again, that is a truly remarkable project. Thanks for sharing this with us.

[Justin LE]

Originally Posted by scott ttocs

Very Nice!

Will you be traveling anywhere else to put on demonstrations?

Not exactly. I'll be attending part of the U Games in Berkeley this July though, so if I still have this or newer prototypes then I'll bring them down.

If anyone is coming through Vancouver and wants to give it a try then feel free to give me a heads up! I've been taking it down to our regular wed. evening unicycle meet ups at Science World too.

You could build a motor control circuit that only allows gradual changes to smooth out the transitions. It sounds like that is what you have in mind.
Scott

Exactly, I actually have a fair bit of experience designing brushless motor controllers and have one that's mostly complete, just wasn't done in time for Maker Faire.

The idea for this is that it will have bidirectional torque control, so you can dial in exactly how much torque you want the wheel to produce (forwards or reverse) and the controller will adjust it's output to maintain this torque regardless of your speed. It should be more hands-off that way.

Regular ebike controllers are generally implemented with direct PWM control, so you have to keep adjusting the throttle up and down as your speed changes in order to maintain the right amount of power at the wheel. It takes a bit more concentration than is ideal.

[Justin LE]

Originally Posted by ħǻřрέŗ

Justin-

That is a fantastic project. Great work and congratulations on what appears to be a complete success. Do you happen to have the full specs on the motor you used? I was interested in the complete wiring diagram and the mechanical drawings of the rotor and stator if they're available someplace.

It's a very common style of hub motor with a 205mm rotor ID that came out a few years ago and is manufactured from dozens of companies in china. We get our directly from Nine Continent. The one I used in this project we've named the 2807, and there are more details here:
http://www.ebikes.ca/store/photos/M2807F26.jpg
http://www.ebikes.ca/store/diagrams/M2800F.pdf

and you can see the torque, power, and efficiency curves by selecting the NineCont 2807 motor option with the simulator page:
http://www.ebikes.ca/simulator/

Again, that is a truly remarkable project. Thanks for sharing this with us.

Hey, well thanks for all the encouraging feedback so far! I wasn't quite sure if people here would be excited or carry a lot of reservation about the idea. Certainly we've saw a lot of resistance initially from the traditional cycling community when it comes to embracing electric assist bikes, but mostly they've come around.

Justin

[kb1jki]

Originally Posted by Justin LE

Hey thanks, it's actually my 3rd attempt, so I've had some time to think it over and refine the process a bit. First was giraffe style electric unicycle back in 2006, which had a freewheeling chain drive between the pedals and a hub:

[/IMG]

Nice stator! did you hand wind it? Hey, I think the cranks on your giraffe version were recalled for breaking in '98... eyeball it for stress fractures on the crank arm itself. It was a shimano altus, or the model before it, can't remember.

[Justin LE]

So I had a chance to ride the electric unicycle on some round trips here in Vancouver on the weekend, and was surprised at how little net energy it draws from the pack when one rides at just a comfortable pace. That means using the motor on the uphills, and having the motor behave as a brake on the downhills, and riding unassisted and unhindered on the flats.

For the latest trip, which didn't involve any super steep hills but wasn't totally flat either, these are the details:

Trip Distance = 9.44 km
Max Speed = 23.1 kph
Avg Speed = 16.2 kph
Positive Amp Hours = 0.496 Ah
Regen Amp Hours = -0.238 Ah
Net Energy Usage = 1.1 Wh/km

By comparison, a regular electric bicycle typically uses 8-10 Wh/km of energy (but also goes at a faster speed). As well, the amount or energy returned to the battery from braking on the downhills was half again what was used on the uphills, while on an ebike this figure is usually around 5-10%.

It's not too surprising since on a bike you only use the brakes when you are coming to a stop, while on the uni it was useful on any downhill grade more than about 2%.

At that rate at just 1.1 Wh/km, it means that even with this fairly tiny 5Ah battery pack I'd need to go over 160km before depleting it. Crazy.

-Justin

[dangerdog]

That is a great set of numbers Justin.What is the weight penalty and do you think a smaller motor would be possible.

john

[Justin LE]

Originally Posted by dangerdog

That is a great set of numbers Justin.

Yeah, it's been surprisingly consistent. This evening I rode it from our shop down to the Science World unicycle meetup and back.

Main St. to Science World

The way there is pretty much all downhill, and so I only used the brake (no throttle) and put 0.239 amp-hours into the battery. On the way back, I didn't touch the brake and only used the throttle, and consumed 0.458 amp-hours. Averaged 18 kph. Round trip was 7.4 km and the net statistics were again 1.1 Wh/km with the regen energy recovered on the downhill being about half what was used to go back up again.

What is the weight penalty

Right now the motor portion of the wheel is about 4kg heavier than a regular unicycle hub, and the battery, battery mount, and electronics add about 2kg. So it roughly doubles the original unicycle weight. I can still hop with it, but it's a bit clunky for that.

and do you think a smaller motor would be possible.

For sure it would be possible, but the efficiency wouldn't be as good, especially in a big 29" wheel that is turning at comparatively low RPMs. That's why I used one of these larger diameter motors for this project. In a 20" wheel then there are some smaller direct drive hubs that would have been well suited and lighter.

You could also use an internally geared motor concept and make something lighter still with the same torque, but usually there is a lot drag in that arrangement if the freewheel is locked (as you would need to do regen) and this would make it less fun to ride when you aren't using the assist.

Justin

[Justin LE]

Hey, is there a chance that anyone here has either videos or pictures of the electric uni assist uni being ridden either at the U games or the vancouver muni weekend? If you do and are willing to share, please send me a PM or an email to justin - at - ebikes.ca

It turns out that a camera man from CNN was at the ElectraFest electric vehicle show here in Vancouver on Sunday and wanted to do a small piece on this, and asked if I had any other images or clips. But asides from the original build pictures I don't have very much.

Thanks, Justin

[uniShark]

From CNN:

Video at: http://ireport.cnn.com/docs/DOC-491972?hpt=C2

iReport — By now we're all familiar with the various forms of electric vehicles, including some bicycles. Taking this to it's logical endpoint, Justin Lemire-Elmore has developed and tested the worlds first electric unicycle. You'd be making a big mistake if you thought this was an attention getting novelty toy because Justin is about to smash several transportation records by being the first person to cross the United States from Canada to Mexico on a unicycle, while using a power source and motor that is not much bigger than your average home electric drill.