Quote:
Originally Posted by Scott Bahl
254 also helped us a lot in getting the rollers functional. We had been having alot trouble with getting enough grip on the ball until 254 showed us their magic material, which made all the difference. They also helped us tune the position correctly. Having 254's 'bot as an example was an enormous help.
-Scott
|
To add a little more to what Scott was saying:
What started out as a front/back single roller after 5 iterations morphed into what we have today. Front and back software clutch variable torque pinch rollers with pneumatic ball gates and a drivers station ball sensor. In theory the ball gates allowed the driver to close a gate so we couldn't intake more than one ball. The gates also centered the ball once captured to our small 3" kicker. The ball sensor allowed the driver to know he had captured a ball even if he could not see it. It took too long to center the ball which is why you didn't see us use our kicker too much. But that's another story.
What made our roller unique was the fact that our rollers are on the front and back of the bot. We use a Jag connected to a CIM and chained up the front and back rollers. We could grab a ball on the front or back of the bot by reversing the rotation of the CIM. We pulsed the speed of the top roller then we sensed current to the roller CIM to change the amount of torque applied once we captured a ball. We also tied in forward/reverse ground speed of the robot to the amount of torque applied and also the torque applied to the ball when we fired the kicker. This gave us our ball loft trajectory. We also used the circuit board in the KOP to add a LED light to the drivers station to sense ball captured. This was helpful when Scott couldn't see the ball.
It took use awhile to figure out how to program the CIM to our control loop. At SVR we found that crashing into a wall would reset the Jags and we would lose the ball and the bot would stall. The Jag has a 3 second default once a fault occurs. That can be changed to .5 sec. We also monitored the amps to the Jag and found out we were hitting a 80 amp spike when we hit hard. Once we figured out the Jags were resetting we changed out the drive train Jags to Victors and wrote in code to prevent the roller CIM to run full power.
The materials we used on the rollers was found by a lot of trial and error. We used tennis grip, duct tape, bicycle hand grip. I think the Poof guys finally showed use what they were using. Water proof electrical tape from Home depot(the magic material) We later changed the top roller material to a polyurethane tube that had a durometer of 30.
So in the end we had over 36 " of ball intake that could pick up a ball going full speed and stuff it into a goal. You will have to ask Scott about learning to drive backwards to stuff a ball in.
Here's a pic of our roller:
http://www.bmf.com/frc%20971/2010%20...s/DSC_7689.htm
Roy