Hi. I'm new to the forum, so I apologize in advance if I'm breaking any rules.
A friend and I are making an art project that is basically like one of those "Test Your Strength" games at a carnival, where you hit a pneumatic thing with a sledge hammer as hard as you can and send a ball up a tube to ring a bell at the top. The difference is, we're trying to do it with an accelerometer built into the striking device, sending the data wirelessly to a controller which will light up some LEDs based on how hard the player strikes. For the sake of the game, we can't attach anything to the object being struck, only to the striking object (for reference, picture a foam LARPer sword hitting a moving person.)
We have successfully gotten the accelerometer to talk to the controller, although our first iteration wasn't wireless. We're taking the average acceleration across all three axes and using value = sqrt(x^2 + y^2 + z^2) as our reading. So we can successfully find max acceleration, but what we really need to do is find periods of sudden deceleration, and we don't know how to do that.
We don't actually need to measure force itself, just force relative to a hard strike or a soft strike, which we can calibrate to once we know those values. But we're having trouble figuring out how to deal with things like backswing and bounce, and figuring out what measurements to actually take in order to find a period of sudden deceleration, and determining when to take the measurements. Do we start a timer when we detect acceleration above a certain threshold and then wait for the motion to change suddenly? If so, how do distinguish between an actual impact and someone just flicking it back and forth?
Has anyone worked on something like this? Is it even a good idea to use an accelerometer for this, or is there some way of using a force-sensitive-resistor that's actually accurate enough and tough enough to withstand repeated impacts?
Thanks for any help!