2012 FRC Team 1717 Uncut

[jakemochas]

Originally Posted by MICHAELABICK

Yes magnetometer. But it lags, so you have to run it through a filter.

We did not use a magnetometer. We only used a gyro. Our implementation of field-centric swerve drive with a gyro can be seen below.

Originally Posted by Jared341

What gyro do you use for field-relative control?

Originally Posted by JamesTerm

Does it matter where the gyro is placed on the robot?

We ended up using a SparkFun gyro (product# 9094), but we were not satisfied with performance of this gyro with the cRIO’s analog interface. It is important that you mount the gyro in a spot on the robot that has little vibration because it is very susceptible to noise. I would recommend mounting it to the frame of your robot.

Originally Posted by JamesTerm

Thanks for the heads up and this was an idea we considered for last year's game (especially since most of the time the robot was facing you), but given what is said here... I feel there is a missing piece to the puzzle... A gyro can give delta's of angular acceleration but what can ensure the robot's heading is calibrated? Is there some point of reference that can be sensed dynamically while rotating?
I'll be anxious to hear how/if they solved that problem.

Originally Posted by JamesTerm

I have to ask... what language are you using to program?

How did you solve the problem for potential drift? Is there something you calibrated against to keep it in line? How reliable can it keep its heading? Does it matter where the gyro is placed on the robot?

The gyro measures angular acceleration, which we use to determine the orientation of our robot. When the robot is booting up, it calibrates the sensor by measuring the analog voltage when the robot is motionless and establishes that as a baseline. We used the C++ Gyro class provided by WPILib to integrate the analog channel and show us our current angle. We know the orientation of the robot at the beginning of the match and zero it then.

Throughout the season, we tried a variety of 500 degree/second gyros that were attached to the analog port on the cRIO. In the end, though, we were not satisfied with any of the gyros. The gyros had significant drift, asymmetrical behavior, and were extremely sensitive to collisions on the field. In about 45 seconds, the gyro would drift by as much as 60 degrees and the advantages of the gyro would be lost.

In order to solve this problem, the drivers rezeroed the gyro every time they lined up for a shot at the fender or the key (the Reset() function). For all of our shots, the robot faced directly forward, which gave the perfect opportunity to rezero. Between each shot, the gyro was good enough to keep the robot oriented in the correct direction. Also, we trained our drivers in a no-gyro mode (robot-centric) in the event of a gyro failure. This training paid off. Even though we had gyro calibration failures in each regional event, the robot appeared like it was operating normally to the outside observer.

In the end, we were not satisfied with the performance of the gyro. One of our goals for next year is to find a better gyro solution for our swerve drive system.