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

NV Core/Nu Core - God I'm still a newbie!

Expand Messages
  • nakazoto_fc
    Howdy all! So, I ve got another walker planned but before I get on with the fun part of building the thing, I really wanted to wrap my head around nv/nu cores.
    Message 1 of 3 , Mar 12, 2012
    • 0 Attachment
      Howdy all!

      So, I've got another walker planned but before I get on with the fun part of building the thing, I really wanted to wrap my head around nv/nu cores. I figured the best way to do this was to grab a 74HC14 (which was shipped halfway across the world for me) and tinker around. The good news is, I haven't sent it up in smoke yet! The bad news is, I clearly haven't a clue what I'm doing.

      I've built the Wilf's PNCFree core several times using different resistors and caps to get different time values and I got it sending a signal in a nice awesome circle, just like I wanted. The problem I was running into with this (aside from the fact that I couldn't get it to do what I wanted it to) was that it would only work if I used white LEDs instead of the standard red LEDs. I ignored this little hiccup and soldiered on. Getting a single process to loop around was fairly straight forward, but not what I needed for my project at all. What I really need is multiple processes.

      My goal is to have a process like this:
      0011
      1001
      1100
      0110
      0011
      etc.

      So after a lot of reading, I was thinking the nu core could help me out here. So after some more experimenting (using this standard diagram: http://www.beam-wiki.org/wiki/Image:NuRing_-_ascii_graphic.GIF), I got some interesting results. Firstly, I'm not entirely sure how you hook up LEDs to this circuit to see how the signal is propagating. At first, I hooked them up in between the output lines and the return feedback into the first 3m resistor. This worked only when at least one of the LEDs was connected with its polarity opposite the rest. After some tinkering around with the LEDs polarity, I discovered I could get a whole slew of interesting LED patterns, none of which were useful.

      So, I figured it was time to ask for clarification on multiple process nu cores. I can't seem to wrap my head around it, so any insight at all on how to tackle multiple processes, what actually goes on inside an nu/nv core (in layman's idiot terms) or just some random finger pointing and laughing at the new guy would be greatly appreciated!

      Thank y'all very much!

      Cheers
      David
    • nakazoto_fc
      Alrighty! Time to reply to my own message! I think I may have figured it a bit out. I m not entirely sure how I missed this page
      Message 2 of 3 , Mar 13, 2012
      • 0 Attachment
        Alrighty!

        Time to reply to my own message!  I think I may have figured it a bit out.  I'm not entirely sure how I missed this page (http://www.beam-wiki.org/wiki/The_Basic_BEAM_Neuron_Circuit ) but it helped clear up a huge amount of my confusion on the basic circuit.  That page is a very excellent read! 

        As I said, I think I may have made a little bit of headway today!  I was thinking that instead of mucking about with multiple processes, I should really focus on trying to extract everything I can out of a single process before I move on to something even more complicated.  My goal was to get two LEDs actively lighting up in the following pattern.

        LED1 - LED2
        ON-OFF
        ON-ON
        OFF-ON
        OFF-OFF
        etc.

        I knew that this should have been easily possible with the standard single process quadcore but for some reason I just couldn't get it to work.  So I started from scratch again today and decided to start with an NV core before thinking about the (in my eyes) more difficult NU core.  So, after reading around a lot I kind of based an idea off of this page (http://www.beam-wiki.org/wiki/Servo_Neuron_%28Droidmakr%29  ) and came up with the following:



        Schematic:



        Now this worked absolutely perfectly!  This got my LEDs blinking exactly how I wanted them to!  I used massive resistors to slow the time constant way down so I could visually see the overlap of the two LEDs.  This got me thinking though.  Wouldn't it be possible to build this exact same thing only much simpler with a non-inverting buffer using a NU bicore setup?  So, just as an idea to toss around, I drew this up:



        If you made the time constant of each one long enough, in my perfect little world, the first neuron would go high on the output This would slowly rise the input of the second neuron until it tripped.  Now, both neurons would be outputting high.  The first neuron would time out shortly thereafter sending it's output low.  The second neuron would then be working its way towards time out and then just before time out it would trip the first neuron on again, sending its output high.  This cycle would hopefully repeat giving me the exact LED overlap that I'm looking for.

        I haven't actually tried to build this yet as I'm exceedingly skeptical it will actually work.  So, here's my question: What did I miss?  Instead of ever reaching timeout will both neurons stay high all the time?  Will nothing happen at all because it's a broken design?  Will the whole thing go up in smoke (in which case, I'll totally build it right now to watch the fireworks, haha)?

        At any rate, I'll keep experimenting and reading as much as I can!  Maybe some day I'll wrap my head around this insanity!
        Thanks guys!

        Cheers
        David



        --- In beam@yahoogroups.com, "nakazoto_fc" <nakazoto_fc@...> wrote:
        >
        > Howdy all!
        >
        > So, I've got another walker planned but before I get on with the fun part of building the thing, I really wanted to wrap my head around nv/nu cores. I figured the best way to do this was to grab a 74HC14 (which was shipped halfway across the world for me) and tinker around. The good news is, I haven't sent it up in smoke yet! The bad news is, I clearly haven't a clue what I'm doing.
        >
        > I've built the Wilf's PNCFree core several times using different resistors and caps to get different time values and I got it sending a signal in a nice awesome circle, just like I wanted. The problem I was running into with this (aside from the fact that I couldn't get it to do what I wanted it to) was that it would only work if I used white LEDs instead of the standard red LEDs. I ignored this little hiccup and soldiered on. Getting a single process to loop around was fairly straight forward, but not what I needed for my project at all. What I really need is multiple processes.
        >
        > My goal is to have a process like this:
        > 0011
        > 1001
        > 1100
        > 0110
        > 0011
        > etc.
        >
        > So after a lot of reading, I was thinking the nu core could help me out here. So after some more experimenting (using this standard diagram: http://www.beam-wiki.org/wiki/Image:NuRing_-_ascii_graphic.GIF), I got some interesting results. Firstly, I'm not entirely sure how you hook up LEDs to this circuit to see how the signal is propagating. At first, I hooked them up in between the output lines and the return feedback into the first 3m resistor. This worked only when at least one of the LEDs was connected with its polarity opposite the rest. After some tinkering around with the LEDs polarity, I discovered I could get a whole slew of interesting LED patterns, none of which were useful.
        >
        > So, I figured it was time to ask for clarification on multiple process nu cores. I can't seem to wrap my head around it, so any insight at all on how to tackle multiple processes, what actually goes on inside an nu/nv core (in layman's idiot terms) or just some random finger pointing and laughing at the new guy would be greatly appreciated!
        >
        > Thank y'all very much!
        >
        > Cheers
        > David
        >
      • David Buckley
        ... Why and how can this happen? Its input is held high by the second gate, there is nothing that could cause it to change. DAvid ... From: nakazoto_fc To:
        Message 3 of 3 , Mar 13, 2012
        • 0 Attachment
          >  The first neuron would time out shortly thereafter sending it's
          output low
          Why and how can this happen? Its input is held high by the second gate, there is nothing that could cause it to change.
          DAvid
           
           
          ----- Original Message -----
          Sent: Tuesday, March 13, 2012 9:13 AM
          Subject: [beam] Re: NV Core/Nu Core - God I'm still a newbie!

           

          Alrighty!

          Time to reply to my own message!  I think I may have figured it a bit out.  I'm not entirely sure how I missed this page (http://www.beam-wiki.org/wiki/The_Basic_BEAM_Neuron_Circuit ) but it helped clear up a huge amount of my confusion on the basic circuit.  That page is a very excellent read! 

          As I said, I think I may have made a little bit of headway today!  I was thinking that instead of mucking about with multiple processes, I should really focus on trying to extract everything I can out of a single process before I move on to something even more complicated.  My goal was to get two LEDs actively lighting up in the following pattern.

          LED1 - LED2
          ON-OFF
          ON-ON
          OFF-ON
          OFF-OFF
          etc.

          I knew that this should have been easily possible with the standard single process quadcore but for some reason I just couldn't get it to wo rk.  So I started from scratch again today and decided to start with an NV core before thinking about the (in my eyes) more difficult NU core.  So, after reading around a lot I kind of based an idea off of this page (http://www.beam-wiki.org/wiki/Servo_Neuron_%28Droidmakr%29  ) and came up with the following:



          Schematic:



          Now this worked absolutely perfectly!  This got my LEDs blinking exactly how I wanted them to!  I used massive resistors to slow the time constant way down so I could visually see the overlap of the two LEDs.  This got me thinking though.  Wouldn't it be possible to build this exact same thing only much simpler with a non-inverting buffer using a NU bicore setup?  So, just as an idea to toss around, I drew this up:



          If you made the time constant of each one long enough, in my perfect little world, the first neuron would go high on the output This would slowly rise the input of the second neuron until it tripped.  Now, both neurons would be outputting high.  The first neuron would time out shortly thereafter sending it's output low.  The second neuron would then be working its way towards time out and then just before time out it would trip the first neuron on again, sending its output high.  This cycle would hopefully repeat giving me the exact LED overlap that I'm looking for.

          I haven't actually tried to build this yet as I'm exceedingly skeptical it will actually work.  So, here's my question: What did I miss?  Instead of ever reaching timeout will both neurons stay high all the time?  Will nothing happen at all because it's a broken design?  Will the whole thing go up in smoke (in which case, I'll totally build it right now to watch the fireworks, haha)?

          At any rate, I'll keep experimenting and reading as much as I can!  Maybe some day I'll wrap my head around this insanity!
          Thanks guys!

          Cheers
          David



          --- In beam@yahoogroups.com, "nakazoto_fc" <nakazoto_fc@...> wrote:
          >
          > Howdy all!
          >
          > So, I've got another walker planned but before I get on with the fun part of building the thing, I really wanted to wrap my head around nv/nu cores. I figured the best way to do this was to grab a 74HC14 (which was shipped halfway across the world for me) and tinker around. The good news is, I haven't sent it up in smoke yet! The bad news is, I clearly haven't a clue what I'm doing.
          >
          > I've built the Wilf's PNCFree core several times using different resist ors and caps to get different time values and I got it sending a signal in a nice awesome circle, just like I wanted. The problem I was running into with this (aside from the fact that I couldn't get it to do what I wanted it to) was that it would only work if I used white LEDs instead of the standard red LEDs. I ignored this little hiccup and soldiered on. Getting a single process to loop around was fairly straight forward, but not what I needed for my project at all. What I really need is multiple processes.
          >
          > My goal is to have a process like this:
          > 0011
          > 1001
          > 1100
          > 0110
          > 0011
          > etc.
          >
          > So after a lot of reading, I was thinking the nu core could help me out here. So after some more experimenting (using this standard diagram: http://www.beam-wiki.org/wiki/Image:NuRing_-_ascii_graphic.GIF), I got some interesting results. Firstly, I'm not entirely sure how you hook up LEDs to this circuit to see ho w the signal is propagating. At first, I hooked them up in between the output lines and the return feedback into the first 3m resistor. This worked only when at least one of the LEDs was connected with its polarity opposite the rest. After some tinkering around with the LEDs polarity, I discovered I could get a whole slew of interesting LED patterns, none of which were useful.
          >
          > So, I figured it was time to ask for clarification on multiple process nu cores. I can't seem to wrap my head around it, so any insight at all on how to tackle multiple processes, what actually goes on inside an nu/nv core (in layman's idiot terms) or just some random finger pointing and laughing at the new guy would be greatly appreciated!
          >
          > Thank y'all very much!
          >
          > Cheers
          > David
          >

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