Data Storage Tool Display Human Input Protocol Math Motor Control Sensor Signal Generation & Processing Speech & Sound Fun View all Objects


Quick Links
Contact Us
Forums
Support

Register
Login

This object is provided under license. By downloading this object, you are agreeing to the terms of the MIT License.

Kalman Filtered IMU via 5DOF Accelerometer/Gyro

This object uses an ADC to take readings from a 5DOF Accelerometer/Gyro unit and send them through a Kalman filter. This filter maintains an accurate angle, gyro rate and gyro bias based on the noisy readings. Hook up the MCP3208 ADC to the 5DOF circuit like the object shows. Then start up the filter. After that use the reading for balancing a robot or flying an aircraft.

update: Fixed a few bugs and set the correct stackspace (with a little extra)

update: Added 2 axis. You now use these functions: get_angle_pitch, get_angle_roll, get_rate_pitch, get_rate_roll, get_q_bias_pitch, get_q_bias_roll

Version:
  1.1

Type:
  Object

Attributes:


Author:
  Jwood
  jtw.programmer@gmail.com

Categories:
  Math
  Sensor

Downloads:
  6,291

Last Updated:
  2008-03-14



User Reviews:

  • Milt Fenner commented:
    I used your object with a few changes. The Accelerometer I used was a H48C and the rate gyro was a LPR530AL. I got them interfaced to the Propeller board just fine and was able to look at the acceleration and rate gyro values on a TV monitor. The values looked good but with some noise on them. I then ran the Kalman filter with the proper inputs. When I looked, after resetting the Propeller, at the Kalman output of 'get_rate_Pitch' and 'get_rate_Roll', sometimes it looked correct but at other times it looked like the Kalman filter output would go to an extremely large value. When it looked correct I could move the sensor system and the Kalman filter would track the rate gyros. What would cause the Kalman filter to runaway sometimes?
    August 3, 2010 7:53 p.m.


  • jim commented:
    Incredible information and work. Thanks for the effort and sharing. You have just convinced me to go to the Propeller Technology. I'll see if I can figure out the downloading and circuit mechanics as I have been experimenting with autopilot, memsic, gyro, and others on BS2. Keep up the good work!
    May 26, 2010 11:14 p.m.




  • Comments on this Object:


    Post a comment:




    Please contact obex.support@parallax.com with comments or questions.

    Copyright 2011 by Parallax, Inc.