Hey Kepler, let me know if you prefer PM, but I have some questions about operating the friction drive design with a torque-simulation controller like a S06s bmsbattery controller. The main issue I'm finding is that the controller has a very gentle ramp on the torque, so in turn it doesn't really want to climb into the tire. I assembled my bike a while ago, and got it tuned pretty nicely so that it is so *almost* touching normally, and the slightest touch goes into the wheel. When I lift the back wheel, I have no issue getting it to operate as designed and "climb" into the tire.
However, when I'm riding on the bike, the motor doesn't climb into the tire well enough because of that gentle slope. It slips and falls back. I tuned the spring to the point where it is overbalanced, and that can work, but then it doesn't really ever fall away from the wheel.
One solution I have is perhaps the only solution with the given controller I have. If I setup the motor so it's always touching the wheel as I''m not killed by the drag of such a small motor, but I can use it for assist. This is a little defeatist because afterall one of the most beautiful parts of this kepler drive is the fact you get a "drag free" motorless operation.
Another solution might be to get a VESC controller, or a different controller that doesn't sacrifice having sine-wave operation for normal use, but allowing me to use a different more aggressive current ramp to start the engagement process.
Maybe a third solution would be to go down the road of some of your alternative designs, and add a brake cable to force engagement into the tire.
What do you think? Any advice? I am also considering modifying the S06s controller to have a higher current draw. It's something like a measly 17A atm. I can swap the mosfets and double that, and will probably get a more aggressive ramp that might work for me.