DPRG List  

[DPRG] Self Balancing.

Subject: [DPRG] Self Balancing.
From: DeltaGraph at aol.com DeltaGraph at aol.com
Date: Mon Oct 17 07:06:01 CDT 2005


One source not mentioned in Earl's article is Rotomotion. You might want to 
check them out (www.rotomotion.com).

I used one of their gyro/accelerometer modules in my balancing robot. I did 
not use their controller, but instead integrated a modified version of their 
tilt.c program into my controller.
Tilt.c is a module which implements a Kalman filter that uses both the output 
of the accelerometer and gyro to give you a good estimation of robot tilt 

tilt.c runs on my microcontroller which is a Proto Mega32 from dalewheat.com.
The module outputs analog values so you must have a couple of analog input 

If your budget is limited, maybe start by using a potentiometer and feeler to 
sense the tilt angle of your robot rather than a gyro and accelerometer 
module. Of course to derive a rate of rotation you will need to differentiate the 
value from your pot. To make life easier. I would recommend subtracting a bias 
(zeroing out data) from robot at startup with the assumption that the robot is 
being held in a balanced position. (The pot idea was recommended to me by 
David Anderson, who created the first balancing robot, nBot,  in the club and 
mentored me in building my balancing robot, Tip.

Also, "divide and conquer" is an important strategy for producing a balancing 
robot in my opinion. For example, have some method of displaying tilt angle 
graphically or as a number before attempting to use this information with some 
kind of control algorithm.

Good Luck.

Ron Grant

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://list.dprg.org/pipermail/dprglist/attachments/20051017/c51e2881/attachment.html

More information about the DPRG mailing list