As we approached a toll booth on the way to the motor factory, Kenny rolled down the window and Lee said something to the attendant; she waved us through past the policeman standing stiffly by the gate. Kenny laughed. I asked what Lee had said to the attendant. He had said “we know the guy,” which was a bluff. “Guanshi?” I asked. “Yes! You are learning Todd Fahrner!” “Does that work in Beijing?” They both laughed, “No! no! too big!” This is one use of an official-looking black car in the Chinese countryside.
We arrived at a home in a village. One moment we were in a courtyard greeting the owners, and in the next we were on the “factory floor,” in a room with about ten workers. Women sat at a table winding the motor stators by hand, and young men and teenage boys busied themselves with the final assembly and testing. And there were my monkeys!
After a little smiling, waving, and gestures that everybody should relax in my strange green-eyed presence, I checked the paint quality and confirmed that the now-anticorrosion-treated axles fit the mount that I had sent previously as a quality checker: very good. Next we broke out the reference controller and explained how the motor connectors needed to be assembled to match. We put a motor on a test stand to complete this. Unfortunately, the motor cables had been terminated before applying the final shrink-wrap to the wires, meaning that the connectors would need to come off first. Somebody wasn’t looking at the whole picture. Fortunately the Anderson connectors we’re using permit reassembly without waste, unlike the molded sort they are replacing.
As at the controller shop, I was eager to show the people how these motors were being used, the whole picture. Out came the laptop and all huddled round. There had been curiosity about the application because these motors normally serve as the hubs of bicycle wheels, laced up directly with spokes, but mine have the spoke hole flanges machined off. It’s always a surprise for people used to the motor performance of the typical configuration to see how well Stokemonkey performs. In the normal application, the motor is forced to turn at the speed of the wheel, and because the desirable RPM range of an electric motor is far smaller than the range of bicycle wheel speeds, performance is mediocre under many conditions — conditions that may be considered abusive but nonetheless commonplace, especially in China.
These motors are fundamentally similar to those of many bicycle hub motors in China, but Kenny does not sell into the rather glutted and oppressively regulated domestic market. His niche is instead the small enthusiast export market, where his small-run production capability gives him the flexibility to make many models and variations meeting diverse needs.
These motors are elegant in their simplicity. There are no internal moving parts, just massive copper coils arranged around the axle, with neodymium magnets set in the motor case turning on sealed bearings. The mass of the windings and general “inside out” configuration of the motor produces good power at low rotational speeds, meaning that there’s no need for mechanical speed reduction before coupling with standard bicycle drivetrain parts. This makes them quiet and maintenance-free. The mass of the stator lets the motor absorb quite a bit of heat, too, for abuse tolerance. My product takes these wheel motors out of the wheel to let them operate continually at their best speeds, and relieves them of the axle loads they’d encounter as wheels, leaving me with a large surfeit of robustness. I’m not worried about the motors failing, pretty much ever. All failures to date in development have resulted from the vulnerability of the external wires, connectors, and electronics, and most of our recent design activity has been to address these areas. If you’re thinking about buying a Stokemonkey, you should know that the basic technology has the most extensive field testing and toughness proven in the largest practical bicycle market on earth; it’s not some rolling science experiment.
The stator cores are made from punched steel wafers resembling flowers, laminated together, welded and machined smooth at the axle interface. They’re then wound with varying gauges of copper wire to support the desired torque and speed characteristics, and baked with a resinous material to prevent shorts. The axles are milled and pressed into the stator cores in a separate heavy machining facility nearby that we visited, where also the stator wafers are punched. I biked there, actually, as I was learning that the Brompton-folding demonstration was a great icebreaker and general introduction to what kind of weird outsider they were dealing with.
The machine shop was of rustic construction, housing extremely heavy, mid-century-looking implements. Bikes parked in one corner. This shop makes all kinds of parts for the various other manufacturing operations nearby. I told Kenny that the rough timber roof of the place would make it a historic treasure in America; I’m not sure he understood it the right way.
Kenny was a little concerned that my photos of the process would make people worry that everything was too low-tech. It’s true that the large number of hand steps involved, the lack of programmable machinery and so on can introduce a level of variance into the product that, if uncontrolled, is trouble. I was there to control it. The variably machined and pressed axles of the last batch that caused me so much trouble were due to the fact that, indeed, the machinists’ main gauge is the number of turns of a hand crank advancing the stock, and my lack of explicit tolerance guidance did me in. I needed tighter than usual tolerance. Problem solved by me sending a mount and giving dimensions followed by a range of permissible variance; otherwise they’d guess at permissible variance.
Otherwise, I think low tech is fine, often best, when low tech is available and the nature of the thing is simple, as these motors are physically. There is no such thing as a low-tech computer, so give me the high tech. But I’ll take a hand-built bicycle, a low-tech sweater, a letterpress book, mouth-blown glass, etc. over the alternatives any day, at least as long as the craftspeople know what they are making, hence my eagerness to show Stokemonkey being used. Sometimes high-tech can offer only the most illusive benefits to the user — the manufacturer’s attempt to sell his cost-cutting mass-production methodology as an implication of superior quality. Hand-wound motor stators are fine. Robots tend to build things that only robots can fix (though more likely they will be discarded), and the same can be said of most any level of tooling and specialized worker qualification.
I’m pleased to report that the stuff I source in China is not made with prison or child labor, unless you count the teenage sons of the factory manager handling some of the final assembly, right there in their own home, like another household chore. I didn’t ask about pay; I’m sure it’s very low by Western standards, but I saw no evidence of compulsion beyond filial piety (Ã¥ÂÂ?), and those obligations tend to flow both ways; I couldn’t bring myself to press the matter. Certainly the working conditions are far more pleasant than others I saw in China, no worse superficially than the minimum wage jobs I’ve held, and quite a lot better than some of what I went through in the US Army. Cleverchimp is also a family business run from our home, so what we’ve got here is a global cottage industry, with components flowing from a huge bicycle culture in decline under automotive ascendancy to a tiny one in resurgence after a century of automotive insanity.
I admire them. Whatever you may think of the fact that I get these motors from a country with low labor costs and under the conditions indicated, I admire them. And the truth is, I know of nobody making suitable alternative motors in any country at any cost. Commencing such manufacture from scratch is unthinkably beyond my means, and to what end? Meanwhile, Stokemonkey’s highest cost components are manufactured in Portland, Oregon.