1551's encoder-controlled arm
 

This is the most complex thing we've ever coded... But thus far it appears to be easy to use and very precise!

http://www.youtube.com/watch?v=gTLMkUmhU2Y

Re: 1551's encoder-controlled arm
 

Looks very nicely controlled. That end screw "skewer" sticking out the front makes me a little nervous, though...

--Ryan

Re: 1551's encoder-controlled arm
 

That's a great color. Good work.

Re: 1551's encoder-controlled arm
 

I agree. if you have time, I would work on training your driver to protect that "skewer" from impact. This will both save you from potential penalties, and potential damage to your robot.

I would also be preparing spares for this, and figuring out what the easiest way to replace one of these was, in case it got bent beyond the ability to be screwed out.

on an unrelated note: nice programming! what are you using for feedback?

good luck!
-Leav

Re: 1551's encoder-controlled arm
 

[quote=Leav;1026498on an unrelated note: nice programming! what are you using for feedback?
[/QUOTE]

The title says encoders.

We're using three accelerometers (one on each part of the arm and one on the chassis) to control ours. I'm still planning for the last minute switch to potentiometers though :p

Did you use the jaguars' built in PID with the encoders? I'm going to try doing that for our wrist action, and I'm looking for some tips.

Re: 1551's encoder-controlled arm
 

A. Encoders with victors. No PID -- it's just not needed on these.

B. We have spares. Several of them. The entire robot is highly modular and easy to replace.

C. The screw should only ever extend out in skewer-like position when we are in scoring position for the middle row -- and thus with nothing in front of us. If the inspectors disagree, we have a plan in place (and a part half-made) to deal with it.

Re: 1551's encoder-controlled arm
 

You should do that right now. Accelerometers are terrible for this. You would have to integrate them twice, increasing the noise by more than a factor of two. The noise is quite bad already, and you will probably have a long cable run from the wrist to the controller, probably running next to some electrically noisy motors. If your accelerometer is analog (it probably is), this would make it impossible to use for this task. Even if it's digital, I doubt it can work at all.

Switch to encoders or potentiometers as soon as possible.

Re: 1551's encoder-controlled arm
 

We're getting around 5 degrees of accuracy. We're not actually measuring the acceleration of the arm (that's one of the problems) we're measuring the acceleration of something much bigger... gravity.