[windell747]

Hi, I was just reviewing a electronics design course textbook and I realize that all the integration that has been mentioned thus far has been implemented in software. How about performing the integral in hardware by cascading two op amp circuits? The first circuit in the chain would be an inverting amplifier(to get rid of the nagative voltage produced by the integrator) with the resistor values the same and the second circuit would be an integrator with the RC close to 1. This would theoretically produce a voltage that is proportional to the current orientation. The op amp could be reset to give small displacements in between measurements. This way we don't need to perform a Riemann sum and instead evaluate exactly.

Any forseen problems? Maybe some filtering would be required...but maybe this will be a little simpler. Maybe it would be advantageous to integrate the accelorometer data twice as well. For this we could just cascade two integrator circuits.