Look Ma, No Hands!

Killing off this Blog

2008-01-02T17:27:00.001-08:00

I'm going to go out on a limb here (not really) and call this blog officially dead. The original 3WDM (since converted to a 4WDM) was stolen almost a year ago, and I haven't been able to get myself excited about re-building it. It may or may not come back in radically different form at some point in the future.

You might find some of the archived posts useful, so they'll stick around, but If you'd like to contact me about anything on this blog, please jump over to my personal web site.

Great Discussion of Hub Motors

2006-08-29T18:58:00.000-07:00

I just found a really good, comprehensive, and understandable discussion of hub motors (specifically for ebikes) on the UBC Electric Bike Club's site. Actually the whole site is worth reading.

Strain Gages have Landed

2006-06-29T18:28:00.000-07:00

I've just received a set of ten strain gages that I'm hoping to use to replace the force-sensing resistors that I've thus-far been using to measure the rider's (my) weight distribution.

The force-sensing resistors have worked quite well other than the fact that they tend to slide out from between the deck and the frame. Another issue is that the circuit I've been using to drive the servo amp doesn't have enough gain to be really effective.

The strain gages will go on the frame's crossbars, and I'll remove the mounting bushing at the center of the deck, so that the weight of the rider will be on the crossbars. The gages will measure the miniscule flexing of the crossbars and a pair of instrumentation amplifiers will amplify this measurement to a level where it can drive the servo amplifier.

It takes a good bit of finicky prep work to mount, hook up, and seal the strain gages, but I will post an update when I've got them installed, and another when I've got them driving the servo amp.

I'm (In)Famous!

2006-05-03T18:10:00.000-07:00

The good news is the 3WDM has made the front page of Skateboarding Sucks. The bad news is it's tagged as Stupid Shit and Useless Toys. Damn, skateboarders can be harsh!

Well, as they say, bad publicity is better than no publicity. And what is exactly is a skateboard if not a "Useless Toy"?

New Photos

2006-04-27T18:58:00.000-07:00

I have some new photos up on Flickr.

Similar Product

2006-03-14T16:59:00.000-08:00

Here's a similar three-wheeled electric skateboard, with hands-free control. As near as I can tell, the button in the middle is the "go switch" and the brake consists of stepping on the rear fender. Not as intuitive as the 3WDM, but admittedly a heck of a lot cheaper.

Looks interesting, but it's not something I'm ready to fork out $130 for.

New Remote Unit

2006-02-16T21:00:00.000-08:00

I replaced the previous keychain remote unit (about the size of a jumbo pack of cigarettes) with a new one not much bigger than a matchbox. The new one also has the advantage that it's a simple push-on push-off unit, rather than one designed to drive a power door lock switch.

(Note for new readers: The remote doesn't really do anything other than power up the electronics. Controlling the thing is strictly hands-free thanks to the weight sensing controller. But it's a little tricky to step onto the board when it's active, and bending over to flip a switch is just too undignified).

I bought the thing from All Electronics, a company in Van Nuys whose previous store (in downtown LA) I visited a few times when I was in college. The store is definitely worth a stop if you're in the neighborhood, and the catalog and web site are also worth a look if you're looking for fun, cheap electronic components and doodads.

Robot Marketplace

2006-02-09T14:31:00.000-08:00

I just placed an order with The Robot Marketplace, a company that sells parts for so-called combat robots (radio-controlled robot gladiators, basically).

I'm buying some "wet noodle" 12 AWG wire, which should be great for things like battery interconnects, along with a handful of Anderson PowerPole housings and contacts. I'll have some extras left over for whenever I get around to putting together the magic circuit board. Before I do that, however, I need to find a place that will sell me the PCB contacts in quantities of less than 100.

Test Ride

2006-02-08T19:06:00.000-08:00

It turns out the bigger battery pack was just what the doctor ordered: a slope that used to reduce the 3WDM to an embarrassing crawl is now hardly noticeable. Not that I'm about to take the thing up Alabama Hill, but the trip back from work should now be possible without the thing pooping out half way up F Street.

The first draft of the battery mounting consists of eight footman fittings with some webbing and associated buckles. I'm not particularly happy with it, since the batteries still move around much more than I'd like. Part of the problem may be the cam-style clasps I'm using, which depend on introducing a bit of slack into the webbing to stay closed. That, in turn, allows the pack to slide around along the length of the board, which just now disconnected one of the front weight sensors.

I also need to make the pack wiring a bit more reliable. The two major issues are semi-adequate insulation, and the fact that the wires I'm using are quite stiff. The problem with the latter is that vibrations will eventually cause something to fail if the wiring can't flex easily. This weekend I'll try and pick up some nice, floppy 10 AWG wire and a 10k Thermistor at the nearest electronics store (a tossup between Vancouver and Mount Vernon). I'm also thinking of re-doing the wiring on the ends of each stick in the pack by snipping open the heat shrink tubing, soldering on the wires, and putting on a new piece.

Then I just need a better method of strapping the cells to the deck and I should have a good, solid pack.

New Magic Box

2006-02-07T12:21:00.000-08:00

I'm starting design work on a magic box (really more of a magic circuit board) that will integrate the features of the current hand-soldered circuit board, the main relay, the protection fuse, and the servo amplifier evaluation board.

It will have connections for the battery, motor phase windings, motor hall effect sensors, and weight sensors. It will also have a connection for an on/off switch (currently provided by the keyless entry receiver) and a dedicated connection for the battery charger. Finally I'd like to add outputs for the four status LEDs (currently on the evaluation board and not visible while riding) and a brake light output based on the state of the current monitor on the servo amplifier. I'm still deciding how much room for future expansion I want to put on it.

I'll probably use the PCB123 service to design and fabricate the board. Finally I'll conformal coat the whole thing to make it waterproof, and probably put some kind of a box around it to protect it from scrapes with the ground.

This should clean up the underside of the board a great deal and make the 3WDM that much more durable and reliable.

The Batteries Have Landed

2006-02-06T23:11:00.000-08:00

I am now the proud owner of 30 10Ah D-size NiMH Cells. Here's a picture of 10 of them:

I've spent the last couple hours soldering them together in sticks of six volts (5 cells). The solder tabs make this relatively easy, and the heat shrink tubing should enhance the conductivity by forcing the cells together:

Here's the nearly-complete pack, shown roughly where it will go on the (underside of the) board:

I still need to solder on the interconnect wiring, seal the ends up with silicone, and position the thermistor for the charger's temperature sensing. Plus I'll need to figure out how I'm going to affix the pack to the board. I'm leaning toward using the industrial-strength velcro that held the previous pack in place.

Finally it will need a full charge before I taking it for a test run around the block.

Just Bought Some Batteries

2006-02-01T18:00:00.000-08:00

I just ordered up 30 10Ah high-discharge-rate tabbed NiMH cells. The pack weight should be around ten pounds and they should fit neatly under the deck.

The present pack (five 7.2V 3000mAh packs) seems to sag a lot on modest slopes. The new pack should be at least three times stiffer under load (probably more, since the cells are designed for high current), and give roughly three times the range (which translates to around 10 miles, or about an hour of riding time).

Unfortunately I have to build up the pack myself. The good news is that I can configure them however I want, and they have tabs, which should make soldering them much less dicey. I'm leaning towards six sticks of five cells each, with three sticks nestled under each side of the deck towards the front of the board.

That should leave room at the rear of the board for the servo amp, op amps, and keyless doohickey.

Background

2006-01-27T16:55:00.000-08:00

Sometime in the fall of 1995 my friend Craig Meyer and I were somewhat impressed with a little electric scooter one of the Frosh had put together that he used to run around campus and get to class on. At that point you couldn't just go down to your local auto parts store and pick one up for $99, you actually had to build the thing.

We'd often bullshitted about various ridiculously elaborate ways of getting to class. Among our group of friends, bicycles, inline skates, and skateboards were popular (at one point some of my friends even started work on an motorized sofa). But bicycles had to be locked up (the Honor Code notwithstanding, ripping off bikes was popular pastime among the local ne'er-do-wells), inline skates had to be doffed (or you had to wear those silly little Metroblade™ slippers in class), and skateboards were a pain in the ass on the way back which was mostly up a gentle slope. And our Alma Mater's trademark unicycles were, well, just plain goofy.

After considering and rejecting things like self-balancing electric unicycles (well beyond our abilities) and blatantly copying the electric scooter (it'd been done), we considered and rejected an electric skateboard that responded to changes in weight distribution by accelerating or decelerating.

Fast forward a few years, and Craig and I were both working for a company called AeroVironment, with myself in beautiful downtown Monrovia, California (the hottest city in the L.A. basin), and Craig and hour-and-change away in beautiful downtown Simi Valley (home of a Denny's that closes).

AeroVironment had a subsidiary that made a very neat electric bicycle called the Charger Bike. The ChargerBike was unique in that it didn't have any sort of conventional throttle. Rather it used a torque sensor on the chain to match or multiply (or sub-multiply) your pedal effort with electric assist. This made it totally transparent to ride, except that it gave you superhuman strength (at least while pedaling it. For an hour or so before the battery went flat).

The fact that smart people at a real company were excited about making the control of an electric vehicle completely transparent got us excited again about the whole electric skateboard thing. It could be a ChargerBike that you could tote on the bus with you or stick in the trunk of your car (don't get me started—it's a long story). After a bit of legal yada yada to get permission to develop it independently and retain control over any results, we set about building a proof-of-concept model. By the fall of '98 we actually had a prototype that worked for about 30 seconds before it caught fire (literally).

In April of 2000 I was a bit dismayed to find a patent granted that covered the idea pretty thoroughly. Both Craig and I were busy with other things and we again let it go dormant for a couple more years.

The company that was granted the patent was absorbed by ZAP Inc. and apparently they didn't pay maintenance fees on the patent, as it expired in April of 2004. I'm not sure if they have any recourse, but it appears that the technology is essentially in the public domain at this point.

F.A.Q.

2006-01-27T15:52:00.000-08:00

Q: What is a 3WDM?

A: "3WDM" is my own personal shorthand for the Three Wheel Dorkmobile™. It's an electric scooter I built about a year ago based on a Mongoose Slalom 132 mountainboard. I've replaced the rear truck with a small BMX bicycle fork, which holds a wheel with an in-hub electric motor. A battery pack and some control electronics are mounted to the underside of the deck.

Q: Why build your own electric scooter? Why not buy one for $99 at Wal-Mart?

A: This one's special. Unlike most scooters, electric or otherwise, there isn't any sort of handlebar: the remaining truck provides steering like on any other skateboard. Nor is there any sort of cheesy handheld throttle. Rather, a set of force-sensing resistors mounted between the deck and the frame detect the weight distribution of whoever is riding the thing, and direct the control electronics to accelerate or decelerate accordingly.

I couldn't find anything like this commercially available. If you find one, I'd love to know about it.

Q: So how do you ride it?

A: Very well, thanks. The basic idea is I set it on the ground, get settled standing on it, and use the keychain remote to turn it on. Then I shift my weight forward to speed up, and backwards when I want to slow down. If I keep leaning backwards it will begin to move in reverse, until I lean forward again. When I'm stopped, or nearly stopped, I use the keychain remote to turn it off.

Q: So sort of like a Segway™?

A: Yeah, sort of. Except the Segway (at least the one I tried) is a lot more sensitive to weight changes (some might say twitchy) because it would fall over otherwise. The 3WDM is more like snowboarding on pavement.

Q: How much did/does it cost?

A: If I had to build one from scratch that matches it's current configuration, it would cost around US$1500 in parts, and about 6 hours of labor (most of that spent soldering). I've spent a bit more than that if you include parts that I bought and later upgraded or didn't end up using, and of course countless hours designing and rebuilding it to get it to its current level of (cough) refinement.

Q: How much does it weigh?

A: The current weight is around 42 pounds (about 20 kilograms).

Q: How fast does it go?

A: The top speed is around 12.5 miles per hour (20 km/h).

Q: How far will it go on a single charge?

A: I haven't tested the new battery pack's range. Extrapolating from the old pack, I'd say around 10 miles on gentle slopes.

Q: Why three wheels?

A: Well, two wouldn't really be stable, and four means that you have to drive two of them on the same axle to prevent torque steer. That means having either two motors or a differential (if you want to be able to maneuver around tight corners). A single drive wheel (in the middle) avoids these complications, while the pair of wheels up front gives you stability and the ability to turn. Incidentally, it's much more difficult to ride backwards (which probably comes as no surprise to you vehicle engineers out there).

Q: Will you build me one?

A: No.

Q: Why not?

A: Lawyers. But seriously, it's not something I've tested enough to inflict on the general public. There is no redundancy in the propulsion—and more importantly—braking system, and I haven't really analyzed any of the myriad other failure modes. It becomes possessed by evil spirits when it gets wet, and sometimes turns itself off when riding over bumps. In short, it's not ready for prime time.

Q: Can you tell me how to build one?

A: I'm not going to write the 3WDM Cookbook if that's what you mean. But with basic electronics skills can look at the reference designs for the various parts and put them together with some simple op-amp glue circuitry.

Q: Have you thought about patenting the idea?

A: Yes, but someone beat me to it.

Q: Did you steal their idea?

A: Nope. A friend and I came up with the idea back in college and started developing it before the patent was issued. Still, we don't have any documentation that we were first to invent, so we couldn't invalidate the patent by any stretch.

Q: Don't you owe them money?

A: Hmm, ask a patent lawyer. When I originally built the thing their patent had expired due to not paying their maintenance fees. But it looks like they've had it reinstated since then. In any case, I think they'd have trouble showing damages.

Q: Why don't you mass-produce and sell them?

A: That would be fun. But to sell more than a couple of them I would have to cut the size, weight and cost in half without sacrificing range, speed or torque. Not to mention improving the safety, reliability, and manufacturability of the design to consumer-product standards. And that's not something I can do while holding down a day job, if in fact it can be done at all. Plus I'd have to acquire the relevant patent rights, and I don't have that kind of money.

Q: Couldn't you raise some VC money or something?

A: Probably not. It's not Web 2.0, and the only other guys doing this sort of thing haven't been a huge financial success.