Alt-BEAM Archive

Message #11529



To: "'beam@sgiblab.sgi.com'" beam@sgiblab.sgi.com
From: "Vermeulen, Vincent" VermeulenV@health.missouri.edu
Date: Wed, 8 Mar 2000 16:44:22 -0600
Subject: [alt-beam] Servo resistor as feedback sensors


My first walker was the "Chiu" design (2 motor, 4 neuron microcore, 245
driver) but am having trouble with the legs drifting, especially when one of
the feet get stuck and the motor stops.

I was thinking of using the resistors that are mounted inside the Futaba
servo as feedback to the core so that eventually the legs come back
straight. The same loop should still allow turning behavior by having its
feedback be smaller that a "turn" feedback.

Anyone done that?

Thanks,

Vincent



11530 Wed, 08 Mar 2000 15:43:46 -0700 [alt-beam] implex, not feedback beam@sgiblab.sgi.com "John A. deVries II" implex was the word that Mark used originally, but somehow That Other Word
came into the conversation . implex implies (pun intended) a mingling of
stuff at one point (and it is hard to find a dictionary that will have a
definition for it.)

if you ignore the concept of power supply loading, there is still a hurdle
you need to overcome before you'll figure out what implex is doing. it
isn't the Nv neuron that is active NOW that is being affected. it is how
the NEXT Nv neuron's timing is changed / will be changed.

in any case, it isn't feedforward, and it isn't That Other Word. it is
feed-right-here.

im => in, towards; plex => join, entwine ... the intertwining of the
current requirements/voltage of the driven circuits with that which is
driving them. implicit sensing: no longer ignoring the affect the
driven has on the driver (and thus the next element in the chain).


of course, That Other Word will continue to be used and will continue to
confuse people. and one can't ignore the power supply loading / motor
noise arguments.


Home