View Single Post
  #6   Spotlight this post!  
Unread 12-03-2008, 02:25
Guy Davidson Guy Davidson is offline
Registered User
AKA: formerly sumadin
FRC #0008 (Paly Robotics)
Team Role: Alumni
 
Join Date: Mar 2005
Rookie Year: 2005
Location: Ra'anana, Israel
Posts: 660
Guy Davidson is a splendid one to beholdGuy Davidson is a splendid one to beholdGuy Davidson is a splendid one to beholdGuy Davidson is a splendid one to beholdGuy Davidson is a splendid one to beholdGuy Davidson is a splendid one to beholdGuy Davidson is a splendid one to beholdGuy Davidson is a splendid one to behold
Send a message via ICQ to Guy Davidson Send a message via AIM to Guy Davidson Send a message via MSN to Guy Davidson
Re: Gyro Repeatability, Expected Drift ?

I'm not a fan of using the gyro for absolute angular position, not even over the short fifteen seconds of Hybrid mode. I do believe, however, that there are solutions that will let you do what you want with relative ease, at least over fifteen seconds:

A) Only use what the gyro actually gives you, the angular velocity. For example, for the relativly simple task of driving forwards, don't compare the angle to zero, compare the rate of change in the angle to zero. That way you don't have to deal with the integration.

B) In order to make turns, that do require the angle measurement, more accurate, you can look at the angle relative to where you're now. I'll give you an example. Say you've been driving roughly straight, and your angle is now 2 degrees. If you want to turn 90 degrees, set your target to 2+90 = 92, instead of to ninety. Now lets say you turned your about 90, drove straight again, and due to some drift, you're now around 110 degrees. You want to turn ninety again, so you should probably set your target to 110+90 = 200, instead of 180. I imagine this might work better.

Any thoughts?
__________________