Actually, what we all are trying to do with the RC-based crank setups is a more powerful, more efficient and lighter version of the Cyclone 1000W setup. Although my Cyclone setup has almost too much low-end torque, and a top speed faster than I'd want to go, it is not without some problems. First of all, it has a ridiculously small 6T motor sprocket that the chain has a hard time getting around. The result is all kinds racket is generated, especially under load. Not very stealthy.
The second problem is that Cyclone puts a cheap freewheel in the crankset that after awhile adds to the noise pollution problem and wears out pretty quick. Finally, because I changed out the 36T rear drive chainring that comes on the Cyclone crankset, with a 24T "granny" gear, the pedals end up being pretty useless, even in 3rd gear on the Nexus-3 hub. I was trying to match the gearing of what Matt (1000W here...) did on his chopper/cruiser bike, because of the performance he was getting. At the time, nobody else really had a Cyclone 1000W setup, so I decided to try and match what he did to get a performance baseline. Actually, I had a hard time believing that you could get this much power, which seemed by his video and his descriptions to be similar to what I was seeing with a 72V/65A x5 setup, with less than half the number of watts. I now know it is absolutely possible.
I now have more lower end torque, and a higher top speed than the 72V 5303 setup that was on this bike previously, and I could do it with a lower voltage (48V...) which simplifies the setup. What I should have done was leave the 36T chainring alone. It would still have more than enough torque in 1st to get my generous butt up any hill, but would also return some functionality to the pedals.
There is another thread we have here on freewheel cranks. Basically, there are two options (well three, counting the Cyclone crankset...). One is that you use a high-end trials-type front freewheel cranksset, which is basically nothing more than a high quality White Industries ENO freewheel with special crank arms that screw into the FW threads. What they typically do is have a fixed 12T cog on the rear hub, and an 18T FW at the crank. Anyway, what we are doing is using a 22T ENO FW, with the trials crankarms, and then make an adapter that bolts to the 22T FW and then has a standard 110mm BCD mountain bike chain ring pattern, so that we can have lots of chainring choices to achieve the right gearing. The second option for getting a freewheel into the crankset, is to use a IPS crankset from some tandem bike setups. This is used sometimes, to allow the second rider on a tandem to not have to pedal. Tandem purists will cringe at the thought (
), but it is good for setups tor kids and/or wives who just want to go along for the ride. Anyway, again the IPS crankset is nothing more than crankarms that screw into a regular freewheel, which is attached to a standard 110mm BCD spider. the nice thing aout this setup is that you don't have to make the adapter, but the downside is that like the Cyclone crankset, they use cheap 16T freewheels. The solution for both, however, is to replace the cheap FW with an $80 ENO version.
The Cyclone 1000W motor, at about 10-11 pounds, is about as heavy as mid-size hubmotor (Clyte 40x, etc...). Matt's eDrive with the AF 3210 is way less than half that weight, so although I saved a ton of weight switching to the Cyclone setup from the uber-heavy 5303 hubmotor, there's plenty more room for improvement. This is what made me come full-circle, and take a fresh look at what could be done with the larger RC-based motors, which are a lot more efficient and a lot less weight. Around the same time, Matt had already done his recumbent setup, using the insanely expensive Predator, and that gave me even more confirmation that we were on the right path, and RC-based systems could be made to work quite well in an ebike application.
The reason for Matt's drive is that most large RC motors have a kV (rpm/volt...) that is a lot higher than something like the Cyclone 1000W motor (kV of about 70...), so we need a reduction. The smaller Cyclone motors (350-650W...) are also higher rpm motors so they add a planetary gear reduction. Anyway, most big RC motors are designed to spin a big prop directly and so the kV ends up being around 130-150. In order to match the Cyclone 1000W motor rpm, assuming the same 48V source, all you would need is a 2:1 reduction, but I wanted to not have to use such a small motor sprocket, so on mine, the reduction is about 4:1.
Anyway, my plan is to replace the Cyclone 1000W setup on this bike, with the 3210-based setup. I began fabricating the mounts I need to do this yesterday. What I'm going to do first today, however, is put back on the "stock" 36T Cyclone chainring, in place of the 24T, and then hook up my Eagletree eLogger so I can get a performance baseline of the Cyclone setup. Then I'll take out the Cyclone motor and controller, and install the 3210/eDrive assembly, and do the same tests.
-- Gary