Steering angle sensor- best ideas!

[MichaelDrost]

I've been pondering this for some time now.

Here's what I thought of: Single-axis analog gyroscope for $8.54 from e-bay, which I'll probably stick somewhere on the steering wheel.

If that fails, I'll ducttape/tie-rib it to the steering shaft. Logic dictates (my logic at least, which could mean I am missing something obvious somewhere) that calculating actual angle should be relatively simple.

I could use the GPS signal (averaged) to get some correlation between the two signals. If nothing else, I'd get excellent tracking of driver input for under $9,-.

Simple, easy and cheap. Not sure how accurate yet, but we'll find out. Wiring will be the next challenge, but I'm pretty sure I've got that sorted as well.

I'd love to hear anyone's thoughts on this.

Michael

[brentp]

This is really interesting; the primary concern is how much drift there would be over time.

I wonder if there could be a zero trigger at steering wheel neutral that would reset the zero point and continuously re-calibrate the sensor every time the wheel passes through center point. Of course, dealing with multiple turns lock-lock could be challenging.

[MichaelDrost]

I was actually thinking about using the GPS heading for that. A bit blunt, but testing should reveal just how correlated the two are.

[toga94m]

I'd think the RCP yaw sensor would be a good indicator of not-turning. Zero yaw (at speed) = steering wheel is at center, or you've locked the front brakes and are sliding straight toward the tire wall...

[MichaelDrost]

Now why didn't I think of that!

[dsm2002]

Has anyone consider an rotary optical encoder with two channel quadrature output instead of a pot for a steering angle sensor ? A few years ago I played around with one on an arduino course - a Bourns ENS1J encoder (http://www.mouser.com/ds/2/54/ENC1J-26993.pdf) - and got it to record output.