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

Re: A Bot with Peripheral Vision

Expand Messages
  • dkbovaird@aol.com
    My apologies to both Martin and Clifford, but I had to redraw it...hopefully I didn t mess it up. Lotta diodes. looks like another good app for what started as
    Message 1 of 12 , Feb 16, 2002
      My apologies to both Martin and Clifford, but I had to redraw it...hopefully
      I didn't mess it up. Lotta diodes.

      looks like another good app for what started as a head circuit.

      dave.

      --- In beam@y..., "R. Martin Keen" <keenerd@j...> wrote:
      > I'm not so sure about this one. It seems like a clever way to get around
      > the lockup that SC-head pvores suffer from. Toggling the pdiodes bias
      > voltage might cause some odd stuff. Interesting stuff with the diodes, I
      > can't wait for the write up. I'm guessing with is a variant of the diode
      > based SC-head, not the negative feedback SC-head. For those who find the
      > original cryptic, I included an easier to read schematic. Sorry about
      > the eraser marks. Took me a few tries to get it right.
      >
      > -Martin
      >
      > THE resource for beginners
      > http://beamlinks.botic.com
      > "The best of everyone's BEAM"
      >
      >
      > On Fri, 15 Feb 2002 15:53:42 -0800 (PST) Clifford Boerema
      > <droidmakr@y...> writes:
      > >
      > > Hello all,
      > >
      > > I wanted to share a circuit that results in
      > > a bot with a form of Peripheral Vision.
      > >
      > > It's a Master/Slave form of the SC-Head.
      > >
      > > The photo-diodes should be positioned across
      > > the front of the bot with the ones marked
      > > Far Left and Far Right mounted at a greater
      > > angle off center than the ones marked left
      > > and right.
      > >
      > > I'm putting together a write up on the
      > > circuit and will post it to my site as
      > > soon as I can.
      > >
      > > Hope you find it interesting.
      > >
      > > So long,
      > > Droidmakr
      > >
    • droidmakr
      ... No appology needed. In fact thanks. You did a great job and I do not see any errors. My schematic may make it easier to assemble and wire but yours makes
      Message 2 of 12 , Feb 17, 2002
        dkbovaird@a... wrote:
        > My apologies to both Martin and Clifford,
        > but I had to redraw it...hopefully
        > I didn't mess it up. Lotta diodes.
        >
        > looks like another good app for what started
        > as a head circuit.
        >
        > dave.

        No appology needed. In fact thanks. You did a
        great job and I do not see any errors.

        My schematic may make it easier to assemble and wire
        but yours makes how the circuit works more clear.
        Would you mind if I use it when I post my write-up
        on the circuit?

        Which I think will be ready after one more proof
        reading. That is unless i decide to rewrite a part
        of it again LOL.

        So long,
        Droidmakr
      • Wilf Rigter
        Thanks Martin, Your sketch is easier to understand but the layout drawing is valuable during construction. Huge thanks to Cliff for this wonderfully elegant
        Message 3 of 12 , Feb 17, 2002
          Thanks Martin,

          Your sketch is easier to understand but the layout drawing is valuable
          during construction.

          Huge thanks to Cliff for this wonderfully elegant circuit creation.

          Let me take a stab of how it will respond to its environment:

          The behaviour of this circuit depends on the location of the PDs. From the
          title it would appear that all 4 Pds face forward but the 2 inner PDs face
          directly forward and the outer 2 are angled to the left and right. Then the
          effect is that the robot turns towards a light that is in directly in front
          by turning off one motor and pivots around the stopped wheel. But if there
          is a brighter light on the periphery (front left or front right) then the
          bot reverses one motor and the robot rotates around its center towards the
          light. So light in front only = small adjustment and brighter light to the
          side = rapid large adjustment. This effect of a strong peripheral light
          response mimics natural behavior.

          The tactiles switches behave even more strongly: if a switch is closed then
          the bot turns away unconditionally. If both switches are closed the robots
          reverse straight back regardless of light level.

          In fact the behavior is even more complex and interesting as the light
          conditions themselves change with turning, etc. All in all, inmho this
          circuit is destined to become a classic.

          Kudos to Cliff the droidmakr!

          wilf
        • Wilf Rigter
          Grrr. Scooped again! Right on Bruce. Wilf ... From: Bruce Robinson To: beam@yahoogroups.com Sent: 2/16/02 2:40 PM Subject: Re: [beam] A Bot with Peripheral
          Message 4 of 12 , Feb 17, 2002
            Grrr. Scooped again! Right on Bruce.

            Wilf


            -----Original Message-----
            From: Bruce Robinson
            To: beam@yahoogroups.com
            Sent: 2/16/02 2:40 PM
            Subject: Re: [beam] A Bot with Peripheral Vision

            Clifford Boerema wrote:
            >
            > I wanted to share a circuit that results in
            > a bot with a form of Peripheral Vision.
            >
            > It's a Master/Slave form of the SC-Head.

            Hi, Cliff. Very nice design.

            By my quick analysis ...

            1) Normal behaviour is both motors running "forward".

            2) When the robot is pointing in the general direction of a light,
            increased light on one of the "normal" photodiodes causes the motor on
            that side of the robot to stop. This causes the robot to pivot about the
            stopped wheel and face toward the light.

            3) When "peripherial" photodiodes detect a difference between one side
            and the other, they over-ride the "normal" PD's. The PD that detects the
            brighter light reverses the motor on that side so the robot pivots
            sharply about its center (one wheel forward, the other reverse) to face
            more toward the light.

            4) When the robot bumps into something on one side, it over-rides all
            the PD circuits and reverses the motor on the OPPOSITE side, causing the
            robot to pivot sharply away from the obstacle.

            5) When both touch sensors contact an object, the robot backs up.

            All done with a single 74HC14 (the '240 being a motor driver).

            How's that for a back-of-the-envelope analysis?

            Bruce


            To unsubscribe from this group, send an email to:
            beam-unsubscribe@egroups.com



            Your use of Yahoo! Groups is subject to
            http://docs.yahoo.com/info/terms/
          • Bruce Robinson
            ... Now THAT doesn t happen often :) Bruce
            Message 5 of 12 , Feb 17, 2002
              Wilf Rigter wrote:
              >
              > Grrr. Scooped again! Right on Bruce.

              Now THAT doesn't happen often :)

              Bruce
            • dkbovaird@aol.com
              I d be flattered if you did use it, no problem. I think that layout drawings such as yours are probably more useful to the group in general and take longer to
              Message 6 of 12 , Feb 17, 2002
                I'd be flattered if you did use it, no problem. I think that layout drawings such as yours are probably more useful to the group in general and take longer to create since both the design as well as the hookups need to be thought out before the presentation. The layout drawing together with the schematic may help people who are trying to understand how to read a schematic. And a schematic definitely helps me understand what the circuit is doing.

                And for those wondering about the semantics of the whole thing, your use of the term "schematic" and my use of "schematic" are both valid descriptors of the drawings, just different conventions.

                Nice work, you might want to discuss "why all the diodes?" and "what if I don't use so many (or used more) in between the photodiodes?" since neither wilf nor bruce touched on this aspect.

                dave.

                In a message dated 02/17/2002 2:00:35 AM Pacific Standard Time, droidmakr@... writes:


                My schematic may make it easier to assemble and wire
                but yours makes how the circuit works more clear.
                Would you mind if I use it when I post my write-up
                on the circuit?


              • Wilf Rigter
                As a matter of fact, Cliff will remember earlier comments when discussing the Schmitt Comparator that apply here as well: the multiple Si diodes provide a
                Message 7 of 12 , Feb 17, 2002
                  As a matter of fact, Cliff will remember earlier comments when discussing
                  the Schmitt Comparator that apply here as well:

                  the multiple Si diodes provide a "dead band with memory" between one or two
                  Schmitt inverters switching state. They are used as a constant voltage drop
                  and can be replaced with LEDs, each of which has a higher forward voltage
                  drop providing the same function as 2 or 3 series diodes.

                  The nested (hierachical) SC architecture of Cliff's circuit and the use of
                  the PDs as diode logic makes for the compact and highly functional design of
                  the PeriPhoBot.

                  wilf

                  -----Original Message-----
                  From: dkbovaird@...
                  To: beam@yahoogroups.com
                  Sent: 2/17/02 11:30 AM
                  Subject: Re: [beam] Re: A Bot with Peripheral Vision

                  Nice work, you might want to discuss "why all the diodes?" and "what if
                  I don't use so many (or used more) in between the photodiodes?" since
                  neither wilf nor bruce touched on this aspect.
                • Bruce Robinson
                  ... More useful for building, less so for understanding. I personally like to redraw the circuit myself, whatever form it s in, because I find the act of
                  Message 8 of 12 , Feb 17, 2002
                    dkbovaird@... wrote:

                    > ... I think that layout drawings such as yours are probably
                    > more useful to the group in general ...

                    More useful for building, less so for understanding. I personally like
                    to redraw the circuit myself, whatever form it's in, because I find the
                    act of redrawing it helps me understand it better. But then my brain is
                    a little strange anyway :)

                    > Nice work, you might want to discuss "why all the diodes?"
                    > and "what if I don't use so many (or used more) in between
                    > the photodiodes?" since neither wilf nor bruce touched on
                    > this aspect.

                    Alters the hysteresis (or dead band) between left and right. Each diode
                    has a small voltage drop across it, so the points where the inverter
                    inputs tap off are at slightly different voltages. I haven't played with
                    this at all, but it seems useful to have the "peripheral" PD's with a
                    greater voltage difference between them than the "regular" PD's. They
                    also prevent a direct short through the PD's if they both happen to be
                    forward biased (e.g., both touch sensors tripped at once).

                    Bruce
                  Your message has been successfully submitted and would be delivered to recipients shortly.