Loading ...
Sorry, an error occurred while loading the content.

Re: H.A.W.C.

Expand Messages
  • connor_ramsey@ymail.com
    ... I use a two-Nu reversing circuit. The reverse neuron is attached to the forward neuron through a kill connection. When the output goes low, the
    Message 1 of 3 , Sep 16, 2012
    • 0 Attachment
      --- In beam@yahoogroups.com, "Amit" <amitjones101@...> wrote:
      >
      > Sounds cool. Can't wait to see it when it's built.
      > When I designed my bot I just think of the behaviours I want and kind of copy/paste the components/neurons to do exactly that in and tweak them a bit. I'm not big on the paring down of neurons. I probably have an extra IMx or bicore or two I could get rid of. So I can't really say much for expected behaviour other than what I know I designed it to do. Emergent behaviour? I'll see when I finish it.
      >
      > It's Smart to have the robot reverse and turn, or some combination of that. Reversing straight back doesn't work out to well.
      > The way I've designed that behaviour into my robot is to use Nu neurons for the the tactiles, and the circuit is set up so that one of the turning neurons has a longer activation period then the other. So when the bot reverses, it reverses then turns or turns and reverses, something like that.

      I use a two-Nu reversing circuit. The "reverse" neuron is attached to the "forward" neuron through a kill connection. When the output goes low, the "forward" neuron's output goes high, and makes the phase shift between the motors double to 180'. However, when the rear "forward" sensor is triggered, implying the robot has backed into something, the "forward" neuron's output goes low again, and the since the "reverse" neuron times out before the "forward" neuron in any case, the robot will continue to go forward until encountering another object. It's really a monostable circuit with a long delay, backing into another object simply cuts the pulse shorter.

      The side tactiles, however, are connected to the rear motor bicore directly. I only did this because the forward bicore already had to deal with information from the photo-bicore, and I don't want it to become overwhelmed.

      Also did I mention I plan on shoking the motor bicores with a high frequency monocore? I think shoked bicores would be more interesting.
      So that's 2 master bicores, a photo-bicore, two Nus, an IMX inverter, and a monocore. And later a solar engine. And eventually I even plan on maybe adding a righting feuture to the IMX: if the robot flips over, the IMX receives a sensor input to triple the phase shift to 270'. An idea by Tom Gray(?), that I read a while ago on a post. I haven't quite figured out how the IMX is going to do that, but I won't worry about it until H.A.W.C. is up and working. Then it's time for upgrades, yes?

      You got any ideas for it?
    Your message has been successfully submitted and would be delivered to recipients shortly.