AVR-Chat

By the sounds of it an accelerometer is not the solution you want - a
wheel encoder would work much better, I think. The basic idea is, you
have a disc with a pattern of black stripes on it, or slots cut in it,
that selectively blocks light. By counting the number of pulses, you
can figure out how far you've turned. By proper placement of the light
sensor / emitter & using two of them, you can figure out direction of
rotation using quadrature decoding. The resolution on this is limited
only by the number of slots / stripes you can print / detect.

-Randy

On Tue, Dec 2, 2008 at 2:34 PM, Richard <richardt.bradshaw@tesco.net> wrote:
> Hi everybody I have been hanging around here for a while, great forum.
> Do any of you have any experience with accelerometers? I am thinking of
> a project using an accelerometer to measure angles. The idea is this:
> the accelerometer will be on the outer edge of a wheel. As the wheel is
> turned the position of the accelerometer will move then using a bit of
> grigonometry I can calculate the angle that it has moved through. My
> question is how much accuracy can I get. I will be looking to measure
> fractions of a degree on the wheel aprox 4 inches or 100mm in diameter.
> Doing the maths gives a movement of about .88mm/degree so I would like
> to be able to detect movement of about 0.2mm is this realistic?
> Cheers Rich
>
>