Log in

View Full Version : pic: 3D Printed Gyro Mount for PDB

Mr. Lim
31-01-2014, 14:44
One of our senior students, David Ferris, designed this very simple mount for our gyro.

It just so happens that our PDB was placed right near the axis of rotation of our robot, which is the ideal place for a gyro.

We didn't have many options for where to mount the gyro, until we notice the raised divider on the PDB.

http://i.imgur.com/8smC8pGl.jpg
http://i.imgur.com/nx84fVUl.jpg
http://i.imgur.com/VNlAGmPl.jpg

It was printed on the Cube printer we purchased for 300 FIRST Choice credits.
AllenGregoryIV
31-01-2014, 15:11
It just so happens that our PDB was placed right near the axis of rotation of our robot, which is the ideal place for a gyro.

That's pretty cool, I like the idea of using that space. Does putting it at the axis of rotation make that much of a difference? (http://electronics.stackexchange.com/questions/38431/does-a-gyro-need-to-be-placed-at-the-center-of-rotation) I assume your chassis is pretty rigid.
cgmv123
31-01-2014, 15:59
Is it held on by friction or is there some other fastener? (Be mindful of R64)
magnets
31-01-2014, 16:04
There isn't a benefit to getting the gyro on the center of rotation. You can put it anywhere on the robot. If I have a gyro in the center, and a gyro on the outside that are both facing forward, and I spin the robot, both gyros will always be at the same angle. When the gyro is on the outside corner, the rotation of the gyro remains the same, but an additional circular path is added. The gyro measures angular rate, which is not influenced by any x/y/z motion of the robot, only rotation around a certain axis.
Alan Anderson
31-01-2014, 16:23
The gyro measures angular rate, which is not influenced by any x/y/z motion of the robot, only rotation around a certain axis.

In reality, there is indeed a slight sensitivity to acceleration by the yaw rate sensor. It's usually not enough to worry about in the context of an FRC match. However, if you've managed to get a precise enough calibration and to minimize the temperature effects, you might notice a very small error in the heading if the sensor is not at the center of rotation.
magnets
31-01-2014, 16:26
In reality, there is indeed a slight sensitivity to acceleration by the yaw rate sensor. It's usually not enough to worry about in the context of an FRC match. However, if you've managed to get a precise enough calibration and to minimize the temperature effects, you might notice a very small error in the heading if the sensor is not at the center of rotation.

True, but I would be really, really surprised if the acceleration of the robot spinning was noticeable compared to the acceleration the robot faces when it hits something (or is just driving) in terms of messing with the gyro.
gpetilli
31-01-2014, 16:36
There isn't a benefit to getting the gyro on the center of rotation. You can put it anywhere on the robot. If I have a gyro in the center, and a gyro on the outside that are both facing forward, and I spin the robot, both gyros will always be at the same angle. When the gyro is on the outside corner, the rotation of the gyro remains the same, but an additional circular path is added. The gyro measures angular rate, which is not influenced by any x/y/z motion of the robot, only rotation around a certain axis.

I believe they are still essentially a spinning disk (abet very small). I would think the additional centrifugal forces on the disk would make it less accurate. Wont you get a "modulated" output from the gyro if it is off center of rotation and therefor having another circular path?
magnets
31-01-2014, 16:50
I believe they are still essentially a spinning disk (abet very small). I would think the additional centrifugal forces on the disk would make it less accurate. Wont you get a "modulated" output from the gyro if it is off center of rotation and therefor having another circular path?

Modern gyro's don't have a spinning disk in them.

If the gyro's reading depended only on angular rate, then no, you would not get a "modulated" output. Having "another circular path" will do nothing to the angular rate of the gyro. The angular rate of a rotating body is the same at every point.

The second circular path (not it spinning on it's own axis, but it rotating around the center of the robot) is really the thing just moving in x and y in a circle. Any acceleration that the gyro feels does mess it up, but only a little. Drift/temp variances are much worse. The gyro feels a centrifugal force as the robot spins because the speed in the x and speed in the y of the gyro change as the robot spins, and acceleration (and the forces that go with them) is just change in velocity. The acceleration of your robot stopping quickly is worse than the circular acceleration. If the robot is providing a force on the robot, either the centripetal force as the robot spins, or the contact force as the robot stops quickly, the gyro will feel the same effect.

If you don't believe me, look at a quadrotor. It doesn't always have the gyro in the center.