September 19, 2007
about 5 years ago
Refer to the schematic. You must power via the JST 2 pin connector unless you close the solder bridge SJ1
Would it be a valid approach to remove the zero-g (1.65 V) for all three axes?
-That is the idea, however the accelerometers used on this IMU are relatively cheap and thus tend to drift. The voltage bias you need to remove may drift as well.
How could the gravitational force be removed? Say, for a human motion analysis application?
Do we need to use angular data from the gyros to calculate the actual orientation of the sensors, and then compute the contributions of the gravitational vector to remove it from each axis?
-The gyros give you your angular rate, so you would have to integrate those values (after adjusting for the bias) to get your angular position. If you want to do complete 6DOF tracking (position + orientation) GOOD LUCK! It is very difficult to track position with this module, since 1) gravity is tough to subtract out once you are moving and 2) acceleration is integrated twice to get position. HUGE errors will add up after a short time
Would you kindly give an overview of such estimates?
We are seriously stuck on this issue.
-I would not recommend using the accelerometers for tracking position on any small scale, but they would work fine if you are not moving (in a particular axis) and you want to determine your gravity vector.
about 5 years ago
If you look at the schematic, it looks like if you close SJ1 it charges the battery
about 6 years ago
I believe the gyro axes are mislabeled. Pitch is about the y-axis and roll is about the x-axis. It looks like the code has is mislabeled as well. If all the channels are enabled it should be:
Forgot your password?
No account? Register one!