View Single Post
  #5   Spotlight this post!  
Unread 09-02-2011, 13:42
thefro526's Avatar
thefro526 thefro526 is offline
Mentor for Hire.
AKA: Dustin Benedict
no team (EWCP, MAR, FRC 708)
Team Role: Mentor
 
Join Date: Aug 2006
Rookie Year: 2005
Location: New Jersey
Posts: 2,599
thefro526 has a reputation beyond reputethefro526 has a reputation beyond reputethefro526 has a reputation beyond reputethefro526 has a reputation beyond reputethefro526 has a reputation beyond reputethefro526 has a reputation beyond reputethefro526 has a reputation beyond reputethefro526 has a reputation beyond reputethefro526 has a reputation beyond reputethefro526 has a reputation beyond reputethefro526 has a reputation beyond repute
Send a message via AIM to thefro526 Send a message via MSN to thefro526
Re: driving straight

Quote:
Originally Posted by CidTeach View Post
Is this a mechanical issue, or just inherent to the motors (one spinning its natural direction and the other "technically" spinning in reverse).

anyone else notice this, are there corrections for this?
Some DC motors have a slight bias to them, as in they spin slightly faster in one direction than the other.

Every FRC Robot I've ever driven has had some sort of drift to one side, some are worse than others. The problem can be magnified by having more weight on one side, a more efficient or less efficient transmission on one side, etc.

The easiest way I know of to fix this is in software, basically you take the speed of the slower motor, and tune the speed of the faster motor to match this one. This is usually done using an encoder or similar sensor to sense wheel speed and then fed through a loop that equalizes the speed, and can be really accurate if done correctly. You can also use a gyro, where a heading is taken from the joysticks and then that heading is compared to that of the robots and then is worked out from there, one side is sped up or slowed down accordingly. Personally, I played with scaling in 2007 where I found the approximate speed difference between the two sides, and then scaled the signal of the faster side to match the slower side - it worked okay, but it isn't really the right way to do things.
__________________
-Dustin Benedict
2005-2012 - Student & Mentor FRC 816
2012-2014 - Technical Mentor, 2014 Drive Coach FRC 341
Current - Mentor FRC 2729, FRC 708
Reply With Quote