cloakedvillain
Well-Known Member
Either an arduino reading a hall effect sensor on the chain ring, moving a servo that controls the throttle or one of those cable speedometers used on the chain ring instead of the wheel.
Heres an idea: pedal assist throttle.
- Thread starter cloakedvillain
- Start date
This is one of those ideas that sounds great in theory, but has a lot of flaws and obstacles in practice. Chief among which is the delay with hall effect chain rings and the assist coming in. Or the fact that even a 49cc 2 stroke is 6 or more times more powerful than a fit cyclist, which can manage about 200 to 300 watts sustained. Or the fact that many internal combustion engines don't often deliver power in a linear fashion like an electric motor.
Sidewinder Jerry
Well-Known Member
If you're trying to make a throttle based on crank cadence; it probably isn't going to work very well. The average human is most comfortable doing a cadence of 70-90. This is why cyclist use a gear they can comfortably maintain a cadence of 70-90. If a gear is too easy or too hard to comfortably maintain a cadence of 70-90 it means you're in the wrong gear and you need to shift up or down.
Chainlube
Well-Known Member
If space wasn't an obstacle a brake dyno could be used to transfer throttle to the carb. A dyno would probably die off when you hit a certain rpm. Just an idea.
cloakedvillain
Well-Known Member
I was looking at ebike pedal assist sensors and these act as potentiometers allowing a 0-10v analog signal to the controller, which then interprets it and again outputs an analog signal.
Ebikes.ca has a torque sensing bottom brack that I think would work well. This same system could be used to control a servo motor attached to the throttle on a 4 stroke. The power wont be linear but it wouldn't have as distinct of a difference in the power band Let's say a very low power engine, maybe a honda gx 25 or 35.
Ebikes.ca has a torque sensing bottom brack that I think would work well. This same system could be used to control a servo motor attached to the throttle on a 4 stroke. The power wont be linear but it wouldn't have as distinct of a difference in the power band Let's say a very low power engine, maybe a honda gx 25 or 35.
Attachments
DieselTech
Well-Known Member
- Local time
- 3:03 AM
- Joined
- Sep 18, 2021
- Messages
- 7,108
Check out the cub cadet xt1 riding mower with powermore electronic intellipower. It's a electronic servo that controls the throttle thru engine demand. As more engine power is needed the electronic servo/components, increases the throttle for the engine/cutting demands automatically. It's kinda a stupid setup, but it works. I think they might be using a map sensor for the engine demand controls.
GW's Motorized
Well-Known Member
Both of my cars have throttle cables. One is throttle body efi and the other is carbureted
cloakedvillain
Well-Known Member
I have been thinking about this a lot and to get a good idea of how well this would work beyond my write up would require prototyping. I can dream up programs and designs all day but without any real world testing it's just that: day dreaming.
My idea for a first prototype is a chain drive set up on the rear rack with the drive by wire controlled by a torque sensor. I can imagine problems where the engine isn't contributing to moving the system forward without a sensor to detect speed of the bike but a wide range in gearing on the human side of it may negate the need to account for that.
As the power curve on these particular models is relatively linear, torque alone and a wide range in gearing on the human side should work most of the time. When I would want more assist, is when I would be pedalling the hardest, opening the throttle more. At high speeds and very low speeds this may not work so well (high meaning above 20 mph, and low meaning less than 5mph) if I am spinning out on the pedal side and the engine side is just freewheeling it's just dead weight. But shifting to a higher gear so that the sensor picks up more torque, thus more throttle, should eliminate this problem. Again testing would help me figure this.
On a single speed bike this system may also work well. If you've ridden one up a hill, you know: You don't pedal at the same cadence but rather at a lower one with higher torque, usually putting in less power but maintaining the same effort. (At least that's what I do, some people just start standing on the pedals.) Mashing is harder than spinning after all. So when you need the assist the most, you get more throttle, then as you speed up you put in less torque, reducing the throttle and thus the assist. At higher speeds you do put some good torque into the pedals but most of your energy is into keeping pace. Again this is just day dreaming, only a real prototype would show real results.
If I were to pursue this I would have a 6v dynamo go into a regulator and a rectifier to top off a battery powering the electronic controls. Maybe in the real world careful use of the throttle would work better but, again that would require testing.
My idea for a first prototype is a chain drive set up on the rear rack with the drive by wire controlled by a torque sensor. I can imagine problems where the engine isn't contributing to moving the system forward without a sensor to detect speed of the bike but a wide range in gearing on the human side of it may negate the need to account for that.
As the power curve on these particular models is relatively linear, torque alone and a wide range in gearing on the human side should work most of the time. When I would want more assist, is when I would be pedalling the hardest, opening the throttle more. At high speeds and very low speeds this may not work so well (high meaning above 20 mph, and low meaning less than 5mph) if I am spinning out on the pedal side and the engine side is just freewheeling it's just dead weight. But shifting to a higher gear so that the sensor picks up more torque, thus more throttle, should eliminate this problem. Again testing would help me figure this.
On a single speed bike this system may also work well. If you've ridden one up a hill, you know: You don't pedal at the same cadence but rather at a lower one with higher torque, usually putting in less power but maintaining the same effort. (At least that's what I do, some people just start standing on the pedals.) Mashing is harder than spinning after all. So when you need the assist the most, you get more throttle, then as you speed up you put in less torque, reducing the throttle and thus the assist. At higher speeds you do put some good torque into the pedals but most of your energy is into keeping pace. Again this is just day dreaming, only a real prototype would show real results.
If I were to pursue this I would have a 6v dynamo go into a regulator and a rectifier to top off a battery powering the electronic controls. Maybe in the real world careful use of the throttle would work better but, again that would require testing.
cloakedvillain
Well-Known Member
I was researching the necessary components and its looking to be very very easy to build such a system. The torque sensor will put out a 1500 millivolt to 3 volt analog signal where the frequency is your cadence (how fast you pedal) and for analog servo motors the position is controlled by the frequency of the analog signal, strength by the voltage.
All I would need to do is figure out the strength of the return spring, and how much travel the cable can take up then select the servo based on that. The program would be something like this: output a 6v signal with the frequency set by the voltage of the torque sensor. Now there is more to it like setting the frequency range and factoring in cadence into the equation to determine frequency, safety switches to cut throttle when brakes are pulled, maybe making a variable set by a potentiometer as a coefficient of the throttle equation to have a tunable assist level without updating the program.
If I pursued this I would use a friction drive and a single speed bike to really focus my attention on the programming side but its not looking too difficult at the moment.
All I would need to do is figure out the strength of the return spring, and how much travel the cable can take up then select the servo based on that. The program would be something like this: output a 6v signal with the frequency set by the voltage of the torque sensor. Now there is more to it like setting the frequency range and factoring in cadence into the equation to determine frequency, safety switches to cut throttle when brakes are pulled, maybe making a variable set by a potentiometer as a coefficient of the throttle equation to have a tunable assist level without updating the program.
If I pursued this I would use a friction drive and a single speed bike to really focus my attention on the programming side but its not looking too difficult at the moment.
Similar threads
