Automatic bicycle hub gear changer using servo

[felamaslen]

I'm new to electronics and servos, but I have an idea for a project: use a servo motor and PICAXE (or Arduino/similar) micro controller to automatically shift the hub gears on a bicycle depending on the current speed.

The first step is, of course, for the servo motor to be able to actually have enough torque to shift to the next gear. Any recommendations on how to do this (e.g. what motor to use, whether I need gearing etc.)? I'm thinking of having the gear cable wrapped in tension around some sort of spinner mounted to the motor, so that as the motor turns, it pulls or relaxes the gear cable, thus changing the gear.

I then need to be able to program the micro controller to automatically shift to the correct gear, depending on the speed. The speed will be measured through the use of a reed switch and magnet, located on the fork and front-wheel spoke respectively. I have no idea what the best/only approach would be to finding the speed like this (well, I do in principle of course, but not in practice).

These are my main questions for now, but more will follow as these are answered.

Thanks
Fela

 

[Raven Luni]

The best approach would be for the servo to operate on the gear cable - but even then it would still have to be pretty strong. Anything below 10kg/cm torque is going to be a disappointment. I tried making an animatronic tail a while ago with some 4kg servos pulling cables and that wasnt enough.

 

[(*steve*)]

You could perhaps use a stepper motor driving a worm gear, but as a cyclist, that would present 3 issues to me:

1) gear changes might be slow.
2) I don't want gears changing without warning
3) I don't want gears changing at inappropriate times.

Those may not be an issue for you.

 

[CDRIVE]

I don't like the concept of an AT on a bike either. It's very hard on the chain and shifter when you're torquing the peddles hard while shifting. I always ease off the peddle force a bit before I shift. How are you going to do this with an auto shifter?

Chris

 

[(*steve*)]

It would actually be quite cool if combined with an electric motor driving through the gears.

The motor could be throttled back during gear changes, and gear changes determined by torque demand (as measured by average current to the motor).

Of course, you (again) would not want to be pedalling at the same time.

There is some interesting additional magic,because the time taken to change gears depends on the combination of road speed and gear selected as well as how quickly you can move the change mechanism.

 

[BobK]

I think you should go for a torque converter.

Bob

 

[CDRIVE]

For what it's worth there's a Trek Cruiser that has a three or four speed, not sure which, auto transmission built into the rear hub. The chain drives a single rear sprocket. I don't think there's any electrics associated with it.

Chris

 

[felamaslen]

It looks like the biggest issue you've pointed to is pretty fundamental to the system - that I would need to know when the shifting happens so I can step down the RPM (pedals, that is) a bit.

What about simply building a manual transmission using a servo motor, and then possibly thinking of making it automatic?

A thought that just came to mind is a clutch, but I don't know whether that is even possible. Obviously cars need clutches (mostly) to change gears, so that is an issue.

 

[(*steve*)]

Well, you don't actually *need* a clutch to change gears in a car once it is moving (although a car manual transmission isn't designed for this)

The whole idea is to reduce the torque sufficiently so that you can unmesh and re-mesh gears without damaging them.

This doesn't apply exactly to a bicycle, but in general the entire gear change thing will happen better under reduced torque. A clutch wouldn't work because the gear change requires that the pedals still be moving. The torque is reduced by the rider not pushing so hard...

 

[felamaslen]

This doesn't apply exactly to a bicycle, but in general the entire gear change thing will happen better under reduced torque. A clutch wouldn't work because the gear change requires that the pedals still be moving. The torque is reduced by the rider not pushing so hard...

Actually, since this is a hub gear, I believe the chain needn't be moving in order to change the gears. I could be wrong on that though. I'll see once it arrives in the post (should be this Tuesday hopefully ).

 

[CDRIVE]

Actually, since this is a hub gear, I believe the chain needn't be moving in order to change the gears. I could be wrong on that though. I'll see once it arrives in the post (should be this Tuesday hopefully ).

Are you saying that the hub has a single sprocket and the shifting is is accomplished with gears internal of the hub?

Chris

 

[felamaslen]

Are you saying that the hub has a single sprocket and the shifting is accomplished with gears internal of the hub?

Chris

Yes, exactly. That's what I meant by "hub gear", although admittedly I could have been clearer.

 

[CDRIVE]

Yes, exactly. That's what I meant by "hub gear", although admittedly I could have been clearer.

Not your fault. I would imagine that they're not all that common. Besides, I've read many cycling posts where people are calling sprockets gears.

Chris

 

[flippineck]

I recall having these type of gears on a kids bike years ago. There was a dinky little chain that snaked out of the end of the axle, turned 90 degress to face forward, sliding over the bevelled bore of the axle, and then connected via a tiny turnbuckle to a cable pull which went to a thumb controller on the handlebars.

I recall it did take a substantial thumb push on the lever, in order to pull on the chain, to click down the gears. I can't quite remember, how upshifts were accomplished - might have been sprung retraction.

https://bicyclehub.co.uk/hub-gears

 

[Harald Kapp]

This is an old thread from 2013. I'll close it now.