Thread: Why you WANT to use a Gyro this year
View Single Post
  #2   Spotlight this post!  
Unread 13-01-2010, 10:12
Jared Russell's Avatar
Jared Russell Jared Russell is offline
Taking a year (mostly) off
FRC #0254 (The Cheesy Poofs), FRC #0341 (Miss Daisy)
Team Role: Engineer
  
Join Date: Nov 2002
Rookie Year: 2001
Location: San Francisco, CA
Posts: 3,078
Jared Russell has a reputation beyond reputeJared Russell has a reputation beyond reputeJared Russell has a reputation beyond reputeJared Russell has a reputation beyond reputeJared Russell has a reputation beyond reputeJared Russell has a reputation beyond reputeJared Russell has a reputation beyond reputeJared Russell has a reputation beyond reputeJared Russell has a reputation beyond reputeJared Russell has a reputation beyond reputeJared Russell has a reputation beyond repute
Re: Why you WANT to use a Gyro this year
Quote:
Originally Posted by JesseK View Post
Bluabot has a suspension system and a 50lb drive train that's meant to conquer the bumps like they're a Mars rover with 120 amps to spare while always staying upright. Yet on a cross-field escape from Redabot's defensive moves, Bluabot doesn't turn enough before trying to climb the bump, thus in the ascent Bluabot flips over without even climbing... How could the team have easily prevented this?

They could have programmed in a Gryo and PID control. Bluabot's drivers could have pushed a button that quickly re-centered the robot's true heading to a vector orthogonal to the climb. In plain English terms, with most robots this would be the same heading as the robot's starting orientation, or 90 degrees from it, so long as that heading points parallel to the sides of the field (aka straight down the field).

Sure, it's not a perfect protection since the driver could still try to climb before the bot is done turning; yet if it prevents your team's robot from tipping over just once this season I believe it's an idea worth evaluating. More probable is that drivers will constantly use it to save time by auto-aligning to the bumps and then gunning it up/over. Tunnel-goers may see benefits too, but it might not be as useful as those who want to do both.

All it takes is a couple of ounces and a couple of hours!
FWIW - you can also solve this problem by using the (linear) accelerometer in the kit. When your robot is stationary, you know the total acceleration sqrt(aX^2+aY^2+aZ^2) should be 1g. When your robot is on flat ground, aZ (vertical axis) should equal 1g and the other two axes should each be 0g. When your robot is pitched up the ramp (driving perfectly perpendicular to the ramp), aX (your "roll") should be 0g, and aY and aZ will have the square root of their squares add to 1g (depending on the degree of pitch). You can monitor aX to ensure there is no "roll".

Granted, this would require that you stop the robot momentarily, but it would have a significant advantage over the gyro - it's measurement doesn't drift over time (since you are reading the raw accelerations and not integrating them).