Ball posession: Roller (ball magnet) or vacuum?

[ernie.]

What I was wondering was, if you were to use a ball magnet/roller would the roller have to spin at the same speed as the wheels? What's the math behind it?

[MrForbes]

Roller needs to spin so that it's surface is turning at least twice as fast (relative to the robot) as the robot is driving, so it can hold the ball as the robot drives backwards.

The actual rotating speed (rpm) depends on the diameters of the roller and the wheels.

[Alex Cormier]

Originally Posted by Vince

I spy, with my little eye, two dog toys....

How well does that work? We are currently using one window to a pvc roller with 35 teeth on the window and a 12 tooth on the pvc. It is a bit too slow for us to really grab the ball well.

[Vince]

Originally Posted by Alex Cormier

How well does that work? We are currently using one window to a pvc roller with 35 teeth on the window and a 12 tooth on the pvc. It is a bit too slow for us to really grab the ball well.

As of right now it is fairly effective but we are still maximizing the effectiveness.
The model that you see in the picture there is a 36:16 reduction from the denso motors with 2.5" diameter Vex wheels. We found that while driving backwards at full speed we cannot hold the ball as well as we desire, so we have changed the ratio to 48:10, to give us about 400 rpm. This may change again before our regional (week 5) but after a bit of testing we will know for sure.

[cziggy343]

for the florida regional, we had a shop vac that was powered by a fisher price motor. but when we get to palmetto, we are planning on testing our ball magnet to see if that works any better

[KrazyCarl92]

We used long rollers with latex flaps at BAE, which did not work well at all. We changed to vacuums for our week 2 regional at WPI and we were unparalleled in our ball possession. However, the ball had to be centered as we drove into it in order for us to gain possession. We were able to go around anybody once we got possession, using our swerve drive at full speed. Now we may be changing to something that's even better . I'll hesitate to call it a roller but it's along those lines...

[waialua359]

I'd look at the lack of rigitidy (softness) in addition to gripiness and speed of rollers.
I also saw a lot of FP and gearboxes for the rollers.
We used a CIM with a 2:1 reduction and adjust the maximum output via programming. We used 60% in SD, but for some reason had to step it up to 80% in AZ to achieve the same result.
Go figure.

[Curt Henderson]

Check out 1501's ball handling abilities. We are using a two wheeled system and it's performance far exceeded our own expectations.

http://www.vimeo.com/10396844

[Ether]

Originally Posted by Curt Henderson

We are using a two wheeled system...

It looks like you have three wheels :-)

Seriously though, did you mean that only 2 of the 3 wheels are driven? Are they omni or mecanum?


~

[Chris is me]

Originally Posted by Ether

It looks like you have three wheels :-)

Seriously though, did you mean that only 2 of the 3 wheels are driven? Are they omni or mecanum?


~

Two wheeled ball magnet.

[Curt Henderson]

We have a three wheeled Kiwi drive system for the robot and a two wheeled Dribbler System for holding the balls while we move around the field.

[Ether]

Originally Posted by Chris is me

Two wheeled ball magnet.

Right. I got so caught up watching the fascinating video that I lost the context.

~

[Ether]

Originally Posted by Curt Henderson

We have a three wheeled Kiwi drive system for the robot and a two wheeled Dribbler System for holding the balls while we move around the field.

I'm not familiar with Kiwi drive. Does that use omni wheels?


~

[Curt Henderson]

Yes it does. Kiwi is a three wheeled drive utilizing omni wheels. We used the A-M Nano gearboxes direct driving the omni wheels.

[Ether]

Originally Posted by Curt Henderson

Yes it does. Kiwi is a three wheeled drive utilizing omni wheels. We used the A-M Nano gearboxes direct driving the omni wheels.

Did you guys home-brew your own kiwi drive algorithm, or is there a good-quality tested library routine available somewhere?


~