Posted by Fraser Allan.
Student on team #188, Woburn Robotics , from Woburn Collegiate Institute.
Posted on 3/27/2000 8:45 PM MST
I just have a quick question about the programming the other teams use. Does changing the input from the joystic to make it a smooth curve rather than a jagged one reduce stress on the motors and prevent spinning your wheels? If so what method would you use, an incrementation that goes from the current levels of the motors to the joystic level, or do you divide the difference by a constant until that difference reaches zero.
I must admit I do not know the full extent of the capabilities of the code used, I was just wondering and would like any input people who understand the code better than myself (read: everybody) think.
-Fraser
Posted by Mike Gray.
Engineer on team HOT from Huron Valley Schools sponsored by GM Milford Proving Ground.
Posted on 3/28/2000 9:47 AM MST
In Reply to: Programming posted by Fraser Allan on 3/27/2000 8:45 PM MST:
Fraser,
Filtering the joystick variables will help machine performance, especially when things get a bit tense on the field. Its best to use a low-pass filter that emulates nature (exponential decay). This is much more intuitive to the driver than using a linear ramp.
Check the White Papers at www.innovationfirst.com for low-pass filter code and a simple method of creating a joystick deadband. Adjust the filter coefficients impirically rather than trying to determine the system Tau. For a 130 pound machine, start with 140. (coefficient .7). A smaller number will make the robot respond more slowly. The coefficient must never exceed 200…unless you thrive on crisis.