i have built an IMU of my own design into my homebrew RC system on a fixed wing airframe.
i am using a MMA7260 accelerometer on an ebay "prototype PCB" along with some rate Gyros for the IMU which is proving quite robust while gliding.
the problems start when i power the motor on.
the accelerometer readings become highly erratic with a gain in output voltage from the MMA7260 on average.
the physical distance between the motor/speed controller is around 300mm.
the accelerometer is powered by a 3.3v regulator on my RC board which is in turn powered by the BEC in the motor controller.
the problem starts to present when the motor is drawing anything more than a few hundred mA but gets worse with more power draw.
i have eliminated mechanical vibration by testing the system with the accelerometer mounted outside the airframe.
i've tried adding smoothing capacitors on the power to the accelerometer and associated microcontroller but have not yet tried any thing else to test/clean up the power supply.
i'm looking for advice as to the most likely cause of this interference before i start ripping it all apart searching for the root cause.
magnetic / power / something else i have not considered?
some details on the components i am using:
my home brew RC system: http://sites.google.com/site/mrdunk/
the ebay MMA7260 accelerometer "prototype PCB: http://cgi.ebay.ie/3-Axis-low-g-Accelerometer-MMA7260-prototype-PCB-/350410579842?pt=UK_BOI_Industrial_Automation_Control_ET&hash=item5196191f82
motor controller: http://www.giantcod.co.uk/hobbywing-pentium-18amp-brushless-motor-speed-controller-p-40.html
motor: http://www.giantcod.co.uk/a22089-2200kv-outrunner-brushless-motor-p-405122.html
i have no experience with accelerometers so any pointers on where to start troubleshooting this would be appreciated.
dunk.
You need to be a member of diydrones to add comments!
Replies
i solved this problem.
the issue was nothing to do with the accelerometer and everything to do with EMF from the power wires affecting the signal wires from the accelerometer.
in a stroke of blind good luck, the combined error after all my IMU mathematical calculations was directly proportional to the current draw which i was already measuring so i was able to fix the problem in software by simply subtracting (k*current_draw) where k is a constant, calculated experimentally.
a little more info here:
http://www.rcgroups.com/forums/showthread.php?t=1336424#post16542715
dunk.