Placment of NavX Micro

So this last year my team started using the NavX Micro to aid in autonomous. They ender up velcroing it down to make it easier to swap between the pratice bot.

This year we are buying two which will allow us to bolt it down permanently. My question is with the magnetometer be adversely affected if it is mounted between the gearbox CIM motors?

Probably not, as long as the robot is powered up significantly before the match. The magnetometer would be adversely affected if there was significant current running through those motors, but the navx calibrates itself before the match, while no motors should be heavily energized. It should be near the center of rotation, however if you’re using rotation measurements.

The NavX will calibrate itself to all of the static magnetic fields on the robot, in other words when the robot is at rest. Problem is there may be considerable dynamic fields generated on your robot, like when motors fire up. Before you make any irreversible decisions, I suggest taping a compass on the location you are considering and then turning on various nearby motors to see how much deflection of the compass needle you get. Try this with the robot placed facing north, south east and west.

Protect your NavX

Printed those the day you posted it.

Does anyone have an ELI5 explanation for this? We’ve thought back and forth on it a bit.

Putting any sensor near the center of rotation (and/or center of mass) seems to be a good idea. In terms of acceleration sensors, I definitely buy it. As far as gyroscopes though, in theory is every point on the robot undergoing the same rotation when the robot changes pose angle. This old thread from FIRST Forums would seem to confirm that line of thought.

Is the NavX doing additional sensor fusion which makes placement important for rotation?

1 Like

Admittedly I’m not extremely knowledgeable in this area, but AHRS do use a fusion of accelerometers, gyroscopes, and magnetometers to determine heading. MEMS gyroscopes are prone to accumulating error, and accelerometers are prone to noise, so an AHRS algorithm tries to fix both predictions by combining them in a special way.

For what it’s worth, this recommendation is also found in the NavX Best Practices. The error would probably be pretty small in practical terms but still there, depending on your goals.

In 2015, our NavX was located in the corner of our robot, and we were able to pull off this auto:

Glad to know it’s not a big issue! I always assumed since it’s on the best practices page it would be a major pain to adjust for. Thanks for posting!