Thanks! Yeah it's quite tough, that was my 3rd try with the HT. I always end up hitting something when I get to the last guy!

Sweet, I am working on a helmet mounted break light for cyclists with the identical setup except I am measuring pitch rate with the gyro! I loved the video; I'll have to try that. I've had some luck implementing low pass filtering on the accels that really perform well for low freq stuff like head motion. I've posted some results here wyoinnovation.blogspot.com/2009/07/improved-break-light.html. You can also check out the schematics for my break light board here: wyoinnovation.blogspot.com/2009/07/sharkfin.html. Same same! I want to get some of these printed up. If you are interested in getting a few let me know.

wyojustin@gmail.com

Hey that is awesome. What do you need the gyro for though? Aren't accelerometers enough for measuring braking? I recently converted my bike to electric, I might look into making a accelerometer activated brake light like yours, probably without the turn signal functions though.

The bike mounted break light does not need the gyro. The next version will be helmet mounted which means that when the rider is looking down breaking will be measured in the z axis of the accel. When the head is level breaking is measured in the x axis. The head angle can of course be anything in between too. The gyro helps keep track of the head angle. The gyro tracks the head angle. The long term down reference is given by the accels and removes the gyro drift.

If you decide to build a break light be sure to check out the low pass filter. Initially I was using an average, just like you are using for your project. The problem was that every time I rode over bumps, the break light would come on. It turns out the sidelobes of the moving average filter are high enough to trip the break detector over bumps.

I'll let you know if we build a batch of boards in case you'd like to get a couple.