Hi. I'm trying to figure out how to measure the rotation of a bike wheel. Not just it's speed, but the position it is at at any give moment.
I'd like to mount a bar of leds (and an arduino + sensor + batteries) inside a bike wheel, then track the rotation of the wheel. I'll use this rotation value to update the leds to create a persistence of vision display.
I know I can use something like a hall sensor to tell every time the wheel finishes 1 rotation, but I'd like more resolution than that, as I image the fluctuations of speed when cycling would make the resulting image jerk around alot.
My best idea so far is to use >1 hall sensor mounted on different spokes of the wheel, with a magnet attached to the forks. I think I'd need quite a few though and it could be cumbersome. Any better ideas?
look at quadrature encoders. very simple, very rugged.
you could print or paint your wheel for the reflective white and black needed.
I think a disk brake would offer the smaller diameter and flat surface that would make it work.
you did not mention how precise you needed.each ring would be one input pin, and double your precision.
first ring, half white, half black, you know you are on half the wheel or the other half. next rings, quarter of rotation, next 1/8 and so-on.,
you can have two sensors on the inner most ring, highest count, one will always precede the other, so you know the direction of rotation.
Hi Dave. The problem I can see with that is the encode would need to be mounted on the frame of the bike, but I need it to be mounted in the wheel with the rest of the electronics, unless I use wireless communication (which maybe isn't a bad idea)
I terms of resolution I guess maybe 8 steps would be the minimum..
Hall sensors mounted on spokes is going to present a wiring problem; at the least, slip rings would be required. Perhaps a better solution would be something like this. I am imagining one mounted to the forks and aimed at the rim of the wheel. The rim could be blackened out except where needed to get the resolution you desire. The linked device, being driven at 38khz, would eliminate any problems that may be caused by ambient light. - Scotty
mrboni:
My best idea so far is to use >1 hall sensor mounted on different spokes of the wheel, with a magnet attached to the forks. I think I'd need quite a few though and it could be cumbersome. Any better ideas?
A Hall sensor or reed switch strikes me as the simplest approach - with the magnet on the fork, you could add as many sensors to the wheel as you want.
mrboni:
I know I can use something like a hall sensor to tell every time the wheel finishes 1 rotation, but I'd like more resolution than that, as I image the fluctuations of speed when cycling would make the resulting image jerk around a lot.
Quite the contrary - fluctuations in speed on a bicycle will be quite minimal - because I very much doubt whether you intend this for a racing bike where the rider is standing to pedal and the cycle surges with every step.
So long as the rider is actually sitting on the seat, his inertia will regulate the speed and a single sensor will suffice.
Have you watched all the (many!) YouTube videos of this system in operation? I haven't looked at them in such detail, but they seem to be quite stable (consistent with the relatively slow POV in any case) and I do believe they use a single synchronising sensor.
The message would be - build it with one sensor to start with, get your PLL sorted out, and go from there.
Henry_Best:
How do you plan to get your output from the wheel mounted electronics to where the rider can see them?
The "output" is optical - a POV display to be seen by others.
(Not - should anyone challenge - to be confused with the other "POV". )
You could get fine-grained measurements of a wheel's position by using a small coil mounted on the brake assembly. The coil would have to be positioned so that the spokes were the closest metal to the coil. This coil would then be fed with a relatively high frequency current, and you could then detect the presence of individual spokes by detecting the voltage fluctuations caused by eddy currents in the spokes - the same principle as a metal detector.
Getting the signal from the frame to the wheel could be done with RF or with a pair of inductively coupled coils near the axle, one fixed to the frame and the other to the wheel.
@ Paul__B - thanks, a really useful answer. And no I haven't seen any of the videos...
Whaddaya know, ladyada do a kit
edit: from the ladyada page - "With one spoke, total persistence at 15mph. Two spokes, 10mph. Three spokes, 7mph. (Assuming a mountain bike wheel). Looks cool even if not completely persistant."
)