Now that we understand the pieces we can buy to assemble the whole system, lets evaluate some possible routes.
We have two types of motors to choose from:
- motor integrated into a wheel hub
- stand alone motor, connected to a wheel by a chain
Hub motors are more expensive and heavy and give us less control over gearing, but they don’t need a transmission. The added cost might justify the added simplicity. Me? I’m cheap. So bring on the cheap motors.
Just a Motor
A stand alone motor needs to attach to a wheel somehow. The first problem is that they generally spin at 3 to 4 thousand RPM. One mile is 2437 revolutions of a 26″ wheel. So to go about 30 miles an hour, or half a mile, we need to spin at about 1220 rpm. This is almost a 4:1 reduction, which is about the maximum that normal bicycle sprockets will let us accomplish. Typical motor sprockets have about 11 teeth. Then our wheel sprocket needs almost 4 times as many teeth, or 44. Look at the table below; the largest rear sprocket I have is 36 teeth. Ok, well, this bike will go a bit faster than 30mph if the motor has enough power at those RPM.
Below is the number of teeth on three bikes I own:
Ratios
Front Rear Largest Front/Smallest rear Smallest front, largest rear
Burning Man 28t – 48t 14t – 28t 3.4:1 1:1
Recumbent 36t – 50t 14t – 28t 3.5:1 1.29:1
Miyata Road 40t – 52t 13t – 36t 3.8:1 1.11:1
Next problem. We can attach the motor to the rear wheel, but the cranks a bicycle rider pedals are also connected to the wheel. Hmm. We can get a double sided hub, but this is expensive. We can dedicate one sprocket on the rear to the electric motor also. When the rider doesn’t pedal, the cranks will freewheel.
We could also attach the motor to the sprockets on the cranks, but actually this is a problem. It would be better, because then you can have the motor assist your pedaling, which I find superior to using a throttle (but this is a matter of taste). However, a human pedals at up to 200rpm. So, to even be humanly possible, we need to reduce the speed by almost 20 times. Normal gears do not allow this. We must use a $150 planetary gear reducer or some concotion of multiple gears.
Finally, my original idea was just to use front wheel driver. Spread the forks a little and put a sprocket on the front wheel. This allegedly performs worse on wet roads; but it’s cheap and easy. Motor mounting may be a problem though, since the front of the bike is using for steering!
So front wheel drive is how I’m going to do this, unless I’m going to rig up a rear sprocket somehow without a double sided hub.
Serial Hybrid
One other idea that no one seems to think highly of is the series hybrid bicycle. This would have the rider pedal a generator which supplies power to the motor and batteries. The benefits are no chain to rust and wear out and an always optimal pedaling cadence. The down side is that generators are not efficient. Bicycling Science, 3rd Edition has a short section on replacing the chain in bicycles. The book cover several “chainless” ideas: hydraulics, ratchet and cable, steel “tape” with holes, and the serial hybrid, and some other technologies . As the book points out, drawing a diagram with efficiencies of each component, the efficiency of the serial hybrid (50-80% by their reckoning) is not as good as a well designed and maintained bike, but, if well designed itself, does exceed the average rusty crap bike on the street.
IHPVA’s site lists the serial hybrid bicycle as one of the top 10 bad HPV ideas, citing that it can’t possibly approach the efficiency of a chain drive. But the IHPVA people are interested in going really really fast. I’m interested in not having to pedal very hard. So I’m going to investigate this and report back. Several people on the Endless-Sphere forums were kind enough to help me find cheap motors to use as generators. At a cost of roughly $70, I’m going to evaluate two approaches: a geared 250W series DC motor, and a 300W BLDC Kollmorgen with bridge rectifier.
Bicycling Science brings up one more important point: that the efficiency of a good serial hybrid exceeds a crap bike may or may not justify the system; the authors note that additional requirements can make the system more attractive. Examples include recumbents where the chain can be a problem, small folding bikes where the chain can’t fold, limiting portability, or where pedaling cadence can only fall into a small range (perhaps a person with an injury or disability or a strange build).
I would like to point out that you can save on SOME costs if you assemble from scratch. You only need one sprocket on the pedals, rather than three, which is cheaper. You also don’t need as many gears, or maybe any at all, on the rear. If you go with a hub motor as the only propulsion, you don’t need sprockets or a derailleur at all– which, besides the frame, add the most cost to a bike.
So I’ll see if I can work some of these other ideas into the bike to make this idea work.