Non-rollover safe code for velocity calculation

developement
1

Hey all,

Still working on optimizing my loop time. Running into some code segments from the library:

 /** get current angular velocity (rad/s) */
inline float Sensor::getVelocity() {
    // calculate sample time
    float Ts = (angle_prev_ts - vel_angle_prev_ts)*1e-6f;
    if (Ts < 0.0f) {    // handle micros() overflow - we need to reset vel_angle_prev_ts
        vel_angle_prev = angle_prev;
        vel_full_rotations = full_rotations;
        vel_angle_prev_ts = angle_prev_ts;
        return velocity;
    }

Here’s a bit to catch overflows in the microsecond timers. The timestamps are defined as longs however:

long angle_prev_ts=0; // timestamp of last call to getAngle, used for velocity
long vel_angle_prev_ts=0; // last velocity calculation timestamp

If these would be defined as uint32_t’s, like the microsecond timer, the overflow would correct itself through unsigned integer math. That would be a nicer solution, right?