Window Motors Breaking!

[roeiy]

Hello, our robot is almost finished. As you can see, we have a lay-up robot that uses an elevator to place a magazine directly in front of the basket, then release three balls in to it. To tilt the magazine we use 2 regular window motors connected to each side of the magazine. However- we encounter a repeating problem: one of our window motors keeps breaking upon rotation. So far we have lost 3 motors to it. Another problem is the gear: The motor stalls whenever we try to lower the magazine down to a horizontal postion and we can only do it by giving it a bit of momentum.
Have you ever encountered these kind of problems before?

http://www.youtube.com/watch?v=E7UvG...ature=youtu.be

thanks.

[EricH]

What are you controlling the motors with? Jaguars, Victors, or Spikes?

[the man]

Sounds to me like a gear reduction issue. Try increasing the reduction and see if the problem persist. Or some thing is not designed right.

[roeiy]

we power the window motors with victors because we know they don't work well with jaguars, so i think this isn't the problem.

about the gear, this is a window motor so the gear reduction is quite high anyway... and we don't really have time or resources to change the gear ratio now.

any other ideas?
thanks

[Taylor]

Looks like a pretty severe moment arm you've got there. Perhaps counterbalance with some weight on the other side of your arm or a passive-assistance method like surgical tubing may help. Asking a window motor to keep all that weight level may be a lot.

[Tom Line]

I suspect you have a combination of not enough torque (not enough gear reduction) and a problem with the back drive pins when you are trying to lower the platform.

[roeiy]

hi,
we thought about passive assitance, but we are not sure in which direction...
the thing is, the window motors can turn the ball holder with no problem, even with 3 balls in it.
on the way up, they need like 60% power, and they lift it in no time.
the problem is on the way down...
when we give them something like 10-15% power, they stall :S.
when we give them 20% power, the ball holder flies down, crashes against its restrainer, and then one of the motor breakes...
they may also break when we give them too much power on the way up, and the ball holder crashes against its restrainer on the way up...

so we figured out we should move the ball holder much slower, but it won't move any slower... giving the motors 5% less power makes them stall, 5% more power and they turn too fast.

help!

[beiju]

Are you stopping the motor manually or via sensor? It could be that your reaction time isn't enough, and your window motors are straining to continue after your ball holder hits a mechanical stop. Try adding a microswitch (they're really easy to mount and wire, and I can help if you have any questions) and programming whichever button lowers the system to stop working if the switch is pressed.

And of course, keep changing your power level in decreasing increments until you get the level that works. One year, we had a system that only worked properly with a very specific power level that was carried out to four or five significant figures, but once we found that level it worked beautifully and didn't need changing even at the competition.

[Nirvash]

If you have not removed the locking pins from the motor, that could be causing your issues on the way down. When we used these motors last year, while saw some similar issues with the pins in.

[Ether]

Originally Posted by beiju

One year, we had a system that only worked properly with a very specific power level that was carried out to four or five significant figures

How were you able to control the power level to four or five significant figures?

[beiju]

Originally Posted by Ether

How were you able to control the power level to four or five significant figures?

It was FTC, and I believe we were using ROBOT-C.

[Ether]

Originally Posted by beiju

It was FTC, and I believe we were using ROBOT-C.

I've never been involved in FTC. They use power supplies and motor controllers that are accurate and controllable to four or five significant figures?