Alt-BEAM Archive
Message #11270
To: beam@sgiblab.sgi.com
From: "Dennison Bertram" dibst11+@pitt.edu
Date: Thu, 2 Mar 2000 05:44:04 -0800
Subject: [alt-beam] Nice stryder!
Hey dave, I was pouring over the solarbotics site, (Daily habit you know)
and I caught a glimpse of your Stryder robot. Very nice! I had a few
questions about your's. I know you said you ran into difficulty getting it
to walk right. So far, I've built four stryder clone's and only one was
succesfull. The first one I built right after I first saw stryder and I
tried doing it with a microcore. (ooh!) I followed the topology of walkman,
since I figured it ought to be simmilar. Didn't work. The bot dance in
circles. The second attempt was just a poor atempt, using busted servo
motors, and again the microcore. The third was the charm and produced a
lovely bot (albeit overcomplicated) which has since meet it's own demise
upon being shipped with me to college. While I was here, I built another
one, but it suffered from mechanical problems. (I think, now I relate it
more to the noise you mentioned and lack of a 'clock' bicore) What do you
think were the greatest problems in getting your Stryder to work? My feeling
is that the robot itself is very complicated in terms of mecahincial design.
That's one reason why I was surprised at the leg orientation on your
stryder. It's so hard to get things to measure up and be perfectly
semetrical. My third stryder was a succes I think inpart because I used a
neat method for supporting the Motors at 30degree angles, (it is 30 degrees
from the horizontal right) Also I used 245's for the motor drivers so I
think I eliminated most of the motor noise problem. So I was curious as to
your experinces building your stryder. For me the design has been
surprisingly frustrating, yet when the robot finnaly works, it's quite a
rewarding bot. Anyway, I was interested in your thoughts as I was thinking
if I get this fellowship I'll be concentrating a little on this type of
design as it seems to lend itself fairly well to a variety of layered
control experiments. It may even be the ideal platform for starting to
implement the 'horse and rider' control system, while at the same time this
robot also finnally starts to indicate that cpu's may be not be needed as
soon as some seem to think.
Dennison
Home