|
|
|
![]() |
|
|||||||
|
||||||||
A little something we cooked up for an upcoming team demo. Yes, it does follow a green light, and yes, that is a rubber band gun on the back.
06-21-2007 08:58 PM
Jeff 801this is a way better rubber band firing mechanism
http://www.mocpages.com/moc.php/22386
06-22-2007 05:08 PM
dbell
pretty nice.
are all 6 wheels powered?
how many drive motors?
06-22-2007 07:09 PM
s_forbes4 motor, all wheel drive, 1:1 ratio. The drive base was taken from a previous vex bot made to play our summer vex camp game.
) Our programmer finally got a hold of easy c, so he is having fun learning how to use it.
06-27-2007 12:03 PM
Doug GYou can use the camera without connecting it to a TTL port on an RC? How do you feedback "pan" info to the Vex controller? I take it the drive platform is the camera's "pan" system. I have to admit I don't do a whole lot of camera programming, so please elaborate, I'm a bit confused.
06-27-2007 12:56 PM
TheOtherGuy
I found out how to connect the camera to the microcontroller via this post. You basically connect the TTL port on the camera to the tx and rx ports on the vex microcontroller (which are on the same side as the ana/dig inputs). The camera does a lot of processing itself, and it returns a bunch of variables, including the pan angle for the pan servo and the center of the mass. Using EasyC Pro made getting the camera to work a breeze, all you have to do is make some unsigned characters and assign them to whatever the camera returns.
The pan system is only half of the drive system, you need the tilt to figure the disance from the object. The farther away, the lower the angle, the closer, the steeper the angle.
We got it so it'll drive arcade mode to drive up to the light, so it combines the two values and figures the value for each motor. We just made the rubber band gun fire when the robot was happy with it's position.
That's pretty much all there is to it! I found out vex doesn't work too well with the camera because the motors don't provide as much control as you want (ie adding a pid loop made the robot stop short of it's goal, and then 5 seconds later JUMP forward several inches. We never quite got it to accurately tell the distance...)