Sensing QuadRover Heading

The Crimpster

I'm having an issue with the HM55B compass module on the prop control board mounted on the QuadRover robot. It appears that the engine is producing a magnetic field, likely from the magneto, that varies with the drive shaft's position. This appears to the HM55B compass as a rotating heading indication. We've tried moving the compass away from the robot and it seems to improve the situation, however we begin to loose signal integrity after about 1 foot of cable from the prop control board to the HM55B. My question is...

Is there a good way to shield this compass from the engine's magnetic field or is there a better way to reading a stationary heading on the robot?


Thanks for the help!

Sean KD0EHH

The Crimpster

Any input would be greatly appreciated.

Currently we're looking at using the GPS heading while the robot is in motion to give us a starting point for a stationary heading. As the robot turns in place, we can use a forward mounted accelerometer to measure the small acceleration the results when the sensor is moved along a circular path. This should allow us to turn the robot a number of degrees relative to the starting position. This has not yet been tested and we are sure to encounter problems with the accuracy of GPS data, noise floor of the accelerometer, sudden jerks in turns, and overall drift of position data and sensor data.

Another option may be to use sun tracking and a real time clock to calculate a fairly accurate heading, but only during the day.

Sean KD0EHH