Slashdot Mirror


Lego Logic Gates

Thud457 writes "LEGO Logic Gates - It's like Babbage, but with bricks. All the gates except XOR are here, and he goes on to develop a clocked flip-flop. While practical mechanical computers may be out, even at the nanotechnological scale, nanomechanical memory may be in. "

271 comments

  1. cool by mirko · · Score: 2, Interesting

    can one create robots with these or is it only to create simple electronic circuits ?
    I guess these could be combined with mindstorm, couldn't these ?

    --
    Trolling using another account since 2005.
    1. Re:cool by mirko · · Score: 4, Interesting

      I just realised : this doesn't use electricity, it's some mechanical representation of these electronic devices.
      I however think this still could be used in non electrical devices (quote any "infernal machine" here, like the one that get triggered after Indiana Jones picks a statuette)...

      --
      Trolling using another account since 2005.
    2. Re:cool by Anonymous Coward · · Score: 2, Insightful

      Anything too complex would likely need so much force applied to make its state change that the lego would break first.

    3. Re:cool by flumps · · Score: 5, Funny

      Have you seen the size of these things? Putting them in any kind of robot would be kinda stupid.

      "Here I am, brain the size of a planet...made out of lego.."

      I think you'd have more luck with the nano versions of these gates. But good luck trying eh :)

      --
      "So there he is, risen from the dead. Like that fella, E. T." - Father Ted Crilly
    4. Re:cool by fishbot · · Score: 1

      If the strength issue could be sorted then you'd have a great EMP proof robot for use in nuclear warfare.

    5. Re:cool by Anonymous Coward · · Score: 0

      If the strength issue could be sorted then you'd have a great EMP proof robot for use in nuclear warfare.

      And you could finally come out of your parent's basement as the conqueror of the world.

      You smell that? Do you smell that? Burning lego bricks, son. Nothing else in the world smells like that. I love the smell of burning lego bricks in the morning.

    6. Re:cool by nospmiS+remoH · · Score: 5, Funny

      I for one welcome our new EMP proof lego robot overlords.

      --
      !hoD
    7. Re:cool by Mysticalfruit · · Score: 2, Insightful

      The trick would be programming it. You'd need some way of storing information. My thinking would be to use a chain with links denoting one and zero. Then you'd just need a mechanism to read and write the chain.

      And you never thought that automata class was going to come in handy ;-)

      --
      Yes Francis, the world has gone crazy.
    8. Re:cool by PhuCknuT · · Score: 2, Insightful

      And with each logic gate being a couple cubic inches, they are going to be GIANT lego robot overlords!

      First thing that comes to mind is the giant bender statue belching flame and saying 'remember me'.

    9. Re:cool by LordEd · · Score: 1

      This could make parenting harder for non-tech parents.. "Mommy! My castle isn't doing my math homework!"

    10. Re:cool by Kehvarl · · Score: 1, Informative

      The trick would be programming it. You'd need some way of storing information. My thinking would be to use a chain with links denoting one and zero. Then you'd just need a mechanism to read and write the chain.

      I have just two words for you.
      Paper.
      Tape.

    11. Re:cool by marcansoft · · Score: 1

      I think building a Brainfuck interpreter out of this shouldn't bee too difficult.

    12. Re:cool by Anonymous Coward · · Score: 0

      Wow... Brain the size of a planet... Made out of Lego...

      Coooooool....

    13. Re:cool by kai.chan · · Score: 1

      I, for one, welcome the robot overlords whom can be taken apart by babies and children.

    14. Re:cool by Anonymous Coward · · Score: 0

      By automata class you mean The Diamond Age?

    15. Re:cool by fishbot · · Score: 1

      For the record, I was kinda implying that sorting the strength issue would involve making the blocks smaller, out of titanium or something.

      Although a world in which war machines are made entirely of Lego sounds pretty cool. You can imagine the hordes of 8 year olds sat in bunkers frantically trying to find a 4x2 block before the advance can begin!

  2. Lego-4 3 GHz, here we come! by Anonymous Coward · · Score: 5, Funny

    'Tis but a small step to a 64-bit processor with 2 MB of cache. Of course, the bricks might burn up upon power-up, but it's a small price to pay!

    1. Re:Lego-4 3 GHz, here we come! by Anonymous Coward · · Score: 1, Interesting

      If it's good enough for Google then it's good enough for me. Although I'm guessing that in order to make it completely out of LEGO it will require quite a bit more space. Like maybe a stadium?

      I would hate to see the instructions for it, and just imagine trying to find all the proper pieces. Wouldn't walk near it in the middle of the night either, man those things f'ing hurt.

    2. Re:Lego-4 3 GHz, here we come! by AndroidCat · · Score: 1

      I imagine that a 64 bit CPU "die" done with 100mm logic might be a tad large. Lemme finish my Lego calculator and I'll give you some numbers. I do predict that the gigabyte DIMMs will really blow blocks.

      --
      One line blog. I hear that they're called Twitters now.
    3. Re:Lego-4 3 GHz, here we come! by qray · · Score: 0

      You obviously haven't priced Lego's lately.

    4. Re:Lego-4 3 GHz, here we come! by Idarubicin · · Score: 4, Funny
      Hmmm...3 GHz...

      Since the push/pull distance for the gate outputs described is about one LEGO stud length, or 8 millimeters.

      At 3 GHz we have a cycle every 3e-10 seconds (~300 picoseconds), and so a mean speed of 8 mm / 300 ps, or nearly 25 million meters per second. That's nearly 10% of the speed of light, so you're going to have to watch out for relativistic effects. :D

      I don't know the mass of the LEGO axles used in the figure, but if they're 5 grams apiece, then each would have a kinetic energy of about 1.5 terajoules. That's about the same as the kinetic energy possessed by a Nimitz class aircraft carrier travelling at 300 knots.

      If you think P4s have heat dissipation issues....

      --
      ~Idarubicin
    5. Re:Lego-4 3 GHz, here we come! by Walt+Dismal · · Score: 1

      I, for one, welcome our new relativistic, heat-dissipating overlords!

    6. Re:Lego-4 3 GHz, here we come! by multiplexo · · Score: 1
      Nahhhh, you don't really have to start worrying about relativistic effects until you get up to about .9c. Time dilation at .1c is about .05 percent. There is a really cool calculator that will compute various relativistic effects for you.

      Of course as you point out the kinetic energy is a bit of a problem, but think of the great case mods you could do on a Lego computer!

      --
      cheap labor conservatives - they want to keep you hungry enough to be thankful for minimum wage.
    7. Re:Lego-4 3 GHz, here we come! by LizzyDragon · · Score: 1

      Guess someone will just have to build some lego heatsinks and fans for that computer . . .

    8. Re:Lego-4 3 GHz, here we come! by SySOvErRiDe · · Score: 1

      And soon we shall have a machine that uses pi rounded off to 3.

  3. No need for XOR by Anonymous Coward · · Score: 5, Informative

    XOR can be constructed by combining other gates. You acctually just need NAND-gates to be able to create any other gate or larger structure.

    1. Re:No need for XOR by ricotest · · Score: 3, Informative

      However, while NAND is the cheapest gate to make with silicon, it's quite a hefty Lego structure. So alternatives would be more desirable.

    2. Re:No need for XOR by LighthouseJ · · Score: 1

      NAND isn't even needed for XOR, just NOT, AND and OR:

      A XOR B = ((NOT A) AND B) OR (A AND (NOT B))

    3. Re:No need for XOR by Taladar · · Score: 3, Insightful

      With this being mechanical Gates suffering from friction using gates to emulate other gates isn't desirable.

    4. Re:No need for XOR by mausmalone · · Score: 1

      While you're correct, I think that you (and the parent post) are missing the concept of a NAND gate. First of all, all you need is NOT and AND. OR can be constructed with them. Secondly, a NAND gate is a type of electronic gate that can be modified to become either a NOT or an AND. This brings down the cost of hardware considerably because you're only producing one type of logic gate. After the NAND gates are placed on the silicon they're burned so that each is either a NOT gate or an AND gate.

      As far as I know (since the article is slashdotted), you can't even make a NAND gate from legos, but having both NOT and AND means you can apply the same logic in design. Your solution would be a pretty decent one, though, and totally implimentable given their resources.

      --
      -=-=-=-=-=
      I'd rather be flamed than ignored.
    5. Re:No need for XOR by Lord+Ender · · Score: 1

      I thought it was NOR that was used in silicon. Am I confused or something?

      --
      A slashdotter who didn't build his own computer is like a Jedi who didn't build his own lightsaber.
    6. Re:No need for XOR by LighthouseJ · · Score: 1

      You are correct, but I am speaking from the computer engineering curriculum and XOR's are very important to us because we use them alot and the increase in hardware cost and footprint for having to substitute I mentioned in the grandparent instead of a single XOR gate is unacceptable. However, yes, a NAND in a more general logic design is pretty useful.

    7. Re:No need for XOR by ScriptMonkey · · Score: 2, Informative

      If you are using static complementary CMOS logic, nands are cheaper. If you happen to be using an older logic family like NMOS or RTL logic, nor gates are cheaper. In any logic family, xor gates are quite a bit more complex than either.

    8. Re:No need for XOR by palad1 · · Score: 1

      I thought it was NOR that was used in silicon. Am I confused or something?

      Yes

    9. Re:No need for XOR by niktesla · · Score: 1

      Yes, there is an increase in cost and footprint, but it comes anytime you use an XOR since it is a much larger "gate". There really is no such thing as an "XOR gate" - all XOR's are made up of several gates in pretty much all the different IC technologies (NMOS, PMOS, CMOS, DTL, RTL, TTL, etc). The concept of an "XOR gate" just simplifies the schematics and function descriptions.

      --
      I've discovered a remarkable proof, but this margin is too small to contain it...
    10. Re:No need for XOR by Anonymous Coward · · Score: 0

      With this being mechanical Gates suffering from friction using gates to emulate other gates isn't desirable.

      Bill's doing what now?

    11. Re:No need for XOR by Anonymous Coward · · Score: 0

      Any logic that can be generated with NAND can be translated to NOR. The NOR gate was created with legos.

    12. Re:No need for XOR by Anonymous Coward · · Score: 1, Informative

      iirc in TTL nand was significantly cheaper due to the ability to use double base transistors

      i don't think it ever made much difference in mos technology

      in cmos the only difference between a nand and a nor is the P and N are the other way roung

    13. Re:No need for XOR by Anonymous Coward · · Score: 0

      They're the same thing, if you just change the interpretation of the signal levels.

    14. Re:No need for XOR by Anonymous Coward · · Score: 0

      NOT, AND and OR aren't needed, just NAND:

      A XOR B = (A NAND (A NAND B)) NAND (B NAND (A NAND B))

      Four 2-input NAND gates.

    15. Re:No need for XOR by dgatwood · · Score: 2, Interesting
      XOR should be easy in principle, though it would require springs, and I'm struggling to think of lego pieces with springs....

      Anyway, take a long brick, put it on a horizontal pivot. Use horizontal motion of other bricks to push either end of the brick in such a way that it trips some sort of switch. Put such a switch on either end. (If you don't want to use switches, find a way so that both ends can push simultaneously on something built like a space bar so when you push on either end, the center always goes down.)

      When one end of the rotating brick is pressed, it goes down. When the opposite end, ditto. When neither end is pressed, it does not move. When both ends are pressed, the force cancels out and nothing is able to move.

      Now find a piece that has some sort of spring tension to push both ends back to the neutral position when neither end is being pressed, and figure out a similar spring to serve as a buffer so that the failure to move the rotating bar doesn't cause the circuit upstream to wedge in weird ways, and you're done.

      --

      Check out my sci-fi/humor trilogy at PatriotsBooks.

  4. Bom proof by bain_online · · Score: 0
    They sure have an advantage of being immune to nuces, and lets face it they are lot more fun to work with than semiconductors.

    One question though will they run my fav. flavour of Linux? (assuming you can build a computer with these)

    --
    BAIN http://www.devslashzero.com
    1. Re:Bom proof by Ziviyr · · Score: 1

      They sure have an advantage of being immune to nukes.

      How much heat does a nuke put out?

      How hard is it to melt plastic again?

      --

      Someone set us up the bomb, so shine we are!
  5. Thats all very well but..... by JaF893 · · Score: 2, Funny

    Can it play .Ogg files?

    1. Re:Thats all very well but..... by Anonymous Coward · · Score: 0

      retard...

  6. Memory Error - brick not found, restart? (y/n) by flumps · · Score: 4, Funny

    ... just as long as we don't get a little lego man instead of that bloody paperclip I don't really care

    --
    "So there he is, risen from the dead. Like that fella, E. T." - Father Ted Crilly
  7. Will BSD run on it? by Anonymous Coward · · Score: 3, Funny

    Let us know when theres a NetBSD for it so we can build a LegoLAN

    1. Re:Will BSD run on it? by RangerRick98 · · Score: 1

      That'd be cool, but it would only be a matter of time before Netcraft confirmed the death of LegoBSD.

      --
      "You're older than you've ever been, and now you're even older."
    2. Re:Will BSD run on it? by julesh · · Score: 0

      ...LegoLAN ..and imagine a beowulf cluster of those!

      What, you mean like this one ?

    3. Re:Will BSD run on it? by Bob+McCown · · Score: 1

      I think the proper operating system for these is, of course, BrickOS

  8. Way to go, editors! by Zorilla · · Score: 2, Insightful
    --

    It would be cool if it didn't suck.
    1. Re:Way to go, editors! by lxs · · Score: 1, Funny

      Look at it as an open source approach to news.

    2. Re:Way to go, editors! by Anonymous Coward · · Score: 0

      Getting you comments from other posts.

    3. Re:Way to go, editors! by Zorilla · · Score: 1

      Not fair, he posted within less than a minute of mine independently. I was kinda pissed to see that right above mine.

      --

      It would be cool if it didn't suck.
  9. Mechanical Analogs by nyekulturniy · · Score: 4, Interesting

    There may be no obvious immediate use for mechanical analogs of digital circuits, when digital circuits are orders of maginitude faster than mechanical circuits, but dismissing the idea out of hand reminds me of old scientists telling the newspaper that "There's no need for flying machines here!"

    Besides, a steam-powered computer would be really fun to build!

    --
    Nyekulturniy... Proudly confusing readers and editors since 1981!
    1. Re:Mechanical Analogs by Viol8 · · Score: 2, Informative

      I guess the best use of mechanical digital circuits would be if you don't care about speed but want something almost indestructable. Build the cogs out of titanium with teflon coating for lubrication and who knows what the possibilities could be. Just think of the dangerous places you could send a robot built out of this stuff. Output could be an issue though unless you have the thing wave little semaphore flags and some guy with binoculars notes them down!

    2. Re:Mechanical Analogs by Ford+Prefect · · Score: 4, Informative

      There may be no obvious immediate use for mechanical analogs of digital circuits, when digital circuits are orders of maginitude faster than mechanical circuits

      If you scale things down a bit, mechanical 'circuits' can become a lot faster - and combined with the electrical properties of the components there might be an interesting hybrid some point in the future.

      Still, using Lego is just a little insane, and there is the minor problem of a '1' or a '0' slowly degrading into '0.5's further along the Lego logic chain... :-)

      --
      Tedious Bloggy Stuff - hooray?
    3. Re:Mechanical Analogs by roxtar · · Score: 3, Interesting
      There may be no obvious immediate use for mechanical analogs of digital circuits

      I think that they can be used to teach young children the basics of electronics. It will be fun as well as education.

    4. Re:Mechanical Analogs by ToteAdler · · Score: 1

      Still, using Lego is just a little insane, and there is the minor problem of a '1' or a '0' slowly degrading into '0.5's further along the Lego logic chain... :-) LEGO quantum compters now!!! Wow!

    5. Re:Mechanical Analogs by 4ginandtonics · · Score: 1

      "Besides, a steam-powered computer would be really fun to build!

      Build schmild... I would be really fun to operate!

      "More coal on the fire, Martha! I'm about ready to totally frag this noob!"

    6. Re:Mechanical Analogs by martingunnarsson · · Score: 1

      When my brothers physics professor first started studying physics his professor showed the class a transistor. "This is a transistor, it does this and that, but it will probably not be very useful." So, in 50 years we WILL have computers made out of Lego, right?

      --
      Martin
    7. Re:Mechanical Analogs by Anonymous Coward · · Score: 0

      I would be really fun to operate!

      That's not what she said!

    8. Re:Mechanical Analogs by 4lex · · Score: 1

      If you scale things really down, you arrive to quantum mechanics, and there things go really interesting (or will go, if we eventually get a scallable hardware candidate for quantum computing).

      By the way: a slash site on quantum computing: QubitNews. Join us and discuss the computational platform of the future :)

      --
      My journal. Mainly about freedom.
    9. Re:Mechanical Analogs by Anne+Thwacks · · Score: 2, Interesting
      My old college had a steam powered computer

      - an Elliot 803 powered by a 3 cylinder marine compound, driven from the central heating boilers (North London Polytechnic in Hollway Road)

      --
      Sent from my ASR33 using ASCII
    10. Re:Mechanical Analogs by Xiaran · · Score: 1

      It depends on what he means by not very useful. Early Ge transisters transisters were (Im told... Im not that old) extremely unreliable, prone to thermal noise problems and more expensive initally than the much more stable tubes. So maybe thats what he was referring to.

    11. Re:Mechanical Analogs by Chris+Mattern · · Score: 1

      It undoubtedly *was* what he was referring to. However, by saying "they *will* probably not be very useful," he apparently didn't think those problems would be effectively solved, in which he was, of course, spectacularly wrong.

      Chris Mattern

    12. Re:Mechanical Analogs by sporty · · Score: 1

      It's knowledge. If your componenets are limited to mechanical, you have an edge over someone else. The simplest I can think of, are locks. If you wished to have a two key system where both keys match, one of these mechanical gates would work well. No electricity required.

      --

      -
      ping -f 255.255.255.255 # if only

    13. Re:Mechanical Analogs by Anonymous Coward · · Score: 0

      That might not be so great -- teflon wears off and doesn't titanium weld itself together if it's not moved for a long time? That would be pretty bad if two of your cogs became the same piece of metal.

    14. Re:Mechanical Analogs by Tackhead · · Score: 2, Funny
      > Besides, a steam-powered computer would be really fun to build!

      Are you nuts? Valve's DRM system eats up 25-30 megabytes even when HL2 isn't running. Do you have any idea how much Lego costs these days? :)

    15. Re:Mechanical Analogs by LWATCDR · · Score: 3, Insightful

      "Still, using Lego is just a little insane,"
      actually it is very sane. It makes a great learning tool. I remember building simple logic gates with my radio shack electronics kit. This lets students build some simple logic "circuits" and see them work with there own eyes.
      Not to mention that mechanical skills in the US really seem to be going down hill. When was the last time you saw a kid build his own skateboard, tree house, or model airplane.

      --
      See my blog http://ilovecookes.blogspot.com/ for light hearted technical information.
    16. Re:Mechanical Analogs by merlin_jim · · Score: 1

      There may be no obvious immediate use for mechanical analogs of digital circuits

      Actually many people believe that rod computers are the way of the future. Miniaturized to an atomic scale, of course.

      If you google around you can find a paper from a few years ago that discusses one possible configuration for such a computer, right down to molecular structures and the thermal tolerances of the system!

      The only thing this guy DIDN'T figure out was how to build the bloody thing!

      --
      I am disrespectful to dirt! Can you see that I am serious?!
    17. Re:Mechanical Analogs by kfg · · Score: 1

      . . .old scientists telling the newspaper that "There's no need for flying machines here!"

      Name one.

      KFG

    18. Re:Mechanical Analogs by yorkpaddy · · Score: 1

      Exactly what I was thinking. I'm taking ECE 101 at college, how cool would those things be to take into a test. And yes I have memorized all the gates.

      --
      "brxref .k.p ,.by xprt. gbe.p.oycmaycbi yd. cby.nci.bj. ru yd. am.pcjab lgxlcj" don'
    19. Re:Mechanical Analogs by DLWormwood · · Score: 1
      When was the last time you saw a kid build his own skateboard, tree house, or model airplane.

      Since before when lawyers encouraged anybody and their dog to sue the "evil companies" to get the "settlement you deserve." That, and the various "home owner associations" which placed strict rules on property improvement and usage.

      --
      Those who complain about affect & effect on /. should be disemvoweled
    20. Re:Mechanical Analogs by operagost · · Score: 1
      Who do you sue if someone gets hurt building a tree house? Home Depot? Man, we used to make ours out of old scrap with rusty, bent nails and splinters hanging out. The roof was a rusted sheet of metal that really said "tetanus shot."

      And how about building the model airplane? That's pretty safe. Hmm, I guess the kid might glue his mouth shut or huff the glue. Guess we'll have to sue Borden or something.

      --

      Gamingmuseum.com: Give your 3D accelerator a rest.
    21. Re:Mechanical Analogs by HeyLaughingBoy · · Score: 1
      I think that they can be used to teach young children the basics of electronics

      Ummm. No.
      Logic gates are far from the basics of electronics. Basics of computers, maybe, but not of electronics.

      Lightbulbs, batteries, and capacitors can teach the basics of electronics just fine, with a transistor or two thrown in.
    22. Re:Mechanical Analogs by Anonymous Coward · · Score: 0

      "Nobel price"???

      I thought slashdot was bad...

    23. Re:Mechanical Analogs by Eraser_ · · Score: 1

      Such as teaching your kids:

      Lego AND toybox = 1;
      Lego NOT Vaccuum_Cleaner = 1;
      Lego NOT LodgedInMyFoot = 1;
      It goes on...

    24. Re:Mechanical Analogs by ultranova · · Score: 1

      Or just build a titanium case for a normal chip. It would also act as a Faraday's cage, nullifying any EMP pulses...

      --

      Forget magic. Any technology distinguishable from divine power is insufficiently advanced.

    25. Re:Mechanical Analogs by Viol8 · · Score: 1

      Except it won't stand 2000C heat.

  10. Dudududupe by Repugnant_Shit · · Score: 3, Funny

    Next, CmdrTaco will build a duplicate article out of Dupe-lo blocks!

    1. Re:Dudududupe by iainl · · Score: 1

      Does this mean Roblimo can then do another dupe?

      Lego recently launched Quatro - double the size again, for kids too young even for Duplo. I'm definitely buying a boxful for my baby once he's old enough (there is enough Duplo and Lego still around from when I was tiny).

      --
      "I Know You Are But What Am I?"
    2. Re:Dudududupe by HiQ · · Score: 1

      for kids too young even for Duplo
      Oooh, my! Where will this end? Lego for the still unborn babies?

    3. Re:Dudududupe by Zorilla · · Score: 0

      Yeah, Technic sets that you can build coathangers with.

      * runs away from karma police *

      --

      It would be cool if it didn't suck.
    4. Re:Dudududupe by Anonymous Coward · · Score: 1

      Make a NAND out of Duplo, then I will be impressed.

    5. Re:Dudududupe by psergiu · · Score: 1
      --
      1% APY, No fees, Online Bank https://captl1.co/2uIErYq Don't let your $$$ sit in a no-interest acct.
  11. Testing the new server by peterprior · · Score: 1

    "The previous bandwidth issues should be solved now that I have a new server." ...or not...

    1. Re:Testing the new server by Anonymous Coward · · Score: 0

      Probably the new server has a Lego processor...

    2. Re:Testing the new server by Anonymous Coward · · Score: 0

      "The previous bandwidth issues should be solved now that I have a new server." ...or not...

      Actually, seems ok now.

      But honestly, couldn't much of the bandwidth issue been avoided by not making the main page one huge page filled with dozens of images?

  12. This was just plain mean by Anonymous Coward · · Score: 1, Informative
    From the (now Slashdotted) page accessed via MirrorDot:

    Before you read on, if you have had trouble accessing this page then please accept my apologies. The previous bandwidth issues should be solved now that I have a new server.

    1. Re:This was just plain mean by MrNonchalant · · Score: 4, Informative

      MirrorDot appears to be down at the moment. In the meantime I mirrored it here. Unfortunately my web host isn't that generous so I'll probably take it down very shortly. Those that want to mirror the mirror can grab a zip file copy here

    2. Re:This was just plain mean by MrNonchalant · · Score: 1

      Okay, just pulled it. The page now redirects to Coral Cache which has a copy of it.

    3. Re:This was just plain mean by lordDallan · · Score: 1

      Here is the Google cache of the page. The Coral Cache was going very slow for me.

  13. from the page.... by eobanb · · Score: 1, Funny

    "The previous bandwidth issues should be solved now that I have a new server"

    Not for long.

    --

    Take off every sig. For great justice.

  14. Re:Let me guess by Anonymous Coward · · Score: 0

    For your sig: "Ad" is short for advertisement, not "Add", unless I'm missing some sort of joke here.

  15. Re:Clusters, anyone ? by Anonymous Coward · · Score: 0

    But... But you are me !! Look, we have the same name !!!

    -- Anonymous

  16. WOW by oliverthered · · Score: 1

    Not that I could have read the article and worked that one out.

    On the simplest scale you only need or and not gates.

    A not gate can be constructed with one transistor and an or gate with two diodes.

    It would be nice to see a modern day version of an old valve computer, build from transistors and diodes not ic's.

    --
    thank God the internet isn't a human right.
    1. Re:WOW by Anonymous Coward · · Score: 0

      On the simplest scale you only need or and not gates.

      But, because NAND gates are much easier to produce, other gates are usually made up of a combination of NAND gates. A NOT gate is just a NAND gate with the same signal going to both inputs, for example.

    2. Re:WOW by Anonymous Coward · · Score: 0

      If you ever did VLSI design you would know you use nand & nor.

    3. Re:WOW by Anonymous Coward · · Score: 0

      When taking the mickey out of other people's posts, make sure you've got your facts right. D'oh.

    4. Re:WOW by oliverthered · · Score: 1

      The 'only' reason for noting the gate (adding a transistor) is to bring the voltage back up again.

      Or and Not gates are 'simpler', as demonstrated by the lego, electronics is the 'complicated' bit.

      --
      thank God the internet isn't a human right.
    5. Re:WOW by oliverthered · · Score: 1

      How are nand gates easier to produce out of diodes and transistors?

      --
      thank God the internet isn't a human right.
    6. Re:WOW by bsd4me · · Score: 1

      If you find an old version of TI's TTL Databook, you will see the schematics for the various devices. You should be able to find scanned PDF versions on TI's website.

      IIRC, 74LS00 was the quad NAND, 74LS08 was the quad AND, 74LS02 was the quad NOR, 72LS32 was the quad OR, 74LS04 was the hex invertor, and 74LS86 was the quad XOR.

      --

      (S(SKK)(SKK))(S(SKK)(SKK))

    7. Re:WOW by Ashtead · · Score: 3, Informative
      I'd say that depends on the particular kind of logic circuit. DTL and original TTL were natural for NAND-functions, just add another emitter in the input stage for TTL, or equivalently, another diode for DTL to add more inputs. A one-input gate would be the NOT gate. NOR-functions would be achieved by parallelling transistors in the second stage of the gate, hence adding another input then meant adding an another entire input stage also. This would be more space-consuming an thus more expensive.

      For ECL, the situation is the opposite, since here everything is mostly transistor stages connected in parallel, as well as generally available true and complementary outputs, so the NOR or OR functions are the most common. Add another transistor in parallel for each additional input, and outputs can be tied together in some cases, forming more OR-functionality. With inversions, the necessary AND and NAND functions may be generated as per DeMorgans theorem.

      The situation for NMOS, PMOS and RTL are similar to the one for ECL: transistors in parallel for the basic NOR function are generally preferred to transistors in series for the NAND function.

      In CMOS circuits, NAND and NOR are about the same in complexity, it is a matter of parallel-connecting the P-channel transistors and series-connecting the N-channel transistors for a NAND function, and vice versa for the NOR function.

      Here is some information about the internal connections of RTL, DTL, TTL, ECL and CMOS circuits.

      --
      SIGBUS @ NO-07.308
    8. Re:WOW by MindStalker · · Score: 1

      You can produce a factory full of OR and AND gates, or you can produce just NAND gates. The NAND gates are more expensive, but from a manufacturing standpoint its more streamlined to make just one gate.

    9. Re:WOW by stupidfoo · · Score: 0

      Uh ... look out! Fancy models numbers and scheme code (in the sig)... all in one post!

      AAHHHHHHHHHHHHHHHHH!!!!!!!!!!!!!!

      And here I am pretending that everyone is sayings "NADs" instead of "NAND"

    10. Re:WOW by oliverthered · · Score: 1

      So, nand gates are harder to make from transistors and diodes than or and not gates, they're just easier to use.

      looking at youe link the first gate 'DL, diode logic' is the simplest(funny that it being the simplest and at the top) only requiring diodes and resistors.

      And look no nand or nor....
      Diode Logic only permits the OR and AND functions.

      Now read the next bit....

      Diode Logic suffers from voltage degradation from one stage to the next.

      Which is why you not the or gate, making things simpler to use....

      A not only requires , it's so simple they've included it a whole 3 pages before the rest of the logic gates.

      Thanks for the link, it's been a while and I couldn't remember the names for TL,DTL,TRL
      , kinda cut of your nose to spite you face though!.

      --
      thank God the internet isn't a human right.
    11. Re:WOW by oliverthered · · Score: 1

      a not gate, 1 transistor 1 resistor. (TL)
      a or date 2 diodes 1 resistor. (DL)

      No need for DTL or TTL, all a lot easier than making a nand gate.

      Why have you only quoted low switching IC's to me.
      I don't want IC's I just want transistors and diodes.

      --
      thank God the internet isn't a human right.
    12. Re:WOW by bsd4me · · Score: 1

      Why have you only quoted low switching IC's to me. I don't want IC's I just want transistors and diodes.

      My pre-submit post said that I don't remember transistor and diode circuit design that well. I know those datasheets have schematics, and I remember going through them in college to identifying what transsitor/diode did what.

      --

      (S(SKK)(SKK))(S(SKK)(SKK))

    13. Re:WOW by Ashtead · · Score: 1
      In fact, that passive OR gate in the first figure is rather similar in function, though not in construction, to these LEGO gates of the article, since they are also passive, and that the OR function is the simplest one. And since they are passive there is a limit to how many stages can be connected before the signal is degraded and lost.

      Otherwise, I thought it more worthwhile to talk about the kinds of gates actually being useful in arbitrary large systems. As for whether NAND or NOR is the preferred gate, it depends on the logic family more than anything else, since everything can be made out of either all NAND (as for TTL) or all NOR (as for ECL) gates.

      For these LEGO gates, it seems that the OR is the simplest function, though internal inversions only require adding an extra spur gear, so NOR, NAND and AND become only slightly more expensive, and all of these become less expensive than the standalone inverter.

      This suggests using a mix of the inverted and non-inverted inputs of the basic OR gate as required so as to save on standalone inverters.

      All it would need in addition for arbitrary large systems would be some means of reconstituting the signals, that is, some kind of active device.

      --
      SIGBUS @ NO-07.308
    14. Re:WOW by oliverthered · · Score: 1

      it's easy....
      npn gives a notted output (pve on the collectors nve on the emitter), that's an inverter or not gate.

      Or gates are easy too, just think getting the two inputs and binding them together.

      You have to use diodes so that the current only flows in one direction, otherwise you'd get a short across the inputs.

      The problem with tieing the two inputs together is that it's 'passive' so slowly the amount voltage levels will drop due to resistance in the circuit.

      So, what they do it to add not gate (which is active) to the output of the or gate and create a NOR gate.

      This is very basic logic gate stuff..

      Transistors can also be used to produce and gates the or gates are then notted in TTL producing the NAND gate mentioned in the earlier post.

      Basically I'd like to see a computer build using only Diode logic (or gates) and Single NPN Transistors (not gates) instead of IC's, Logo or Valves. Since it is the easiest 'simplest' form of logic.

      Everyone can understand NO and everyone can understand some, you could even build a computer out of chimp or small children.

      --
      thank God the internet isn't a human right.
    15. Re:WOW by Anonymous Coward · · Score: 1, Funny

      Everyone can understand NO and everyone can understand some, you could even build a computer out of chimp or small children.
      Creating a computer out of small children would be a very challenging task, due to their unique logic properties which I've outlined briefly below:

      When placed at the beginning of a circuit, a small child acts as a logical inverter -- a "yes" input becomes "no", and a "no" input becomes "yes".

      However, when placed at the end of a circuit, the same child acts as a NOR gate -- all output is "no", usually with a corresponding increase in amplitude. This output is usually looped back on itself creating a recursively louder "NO!" output.

    16. Re:WOW by operagost · · Score: 1

      Actually, the LS means "low-power Schottky". There are often High speed, Low power, and standard Schottky versions, but LS is the most common.

      --

      Gamingmuseum.com: Give your 3D accelerator a rest.
    17. Re:WOW by Kiryat+Malachi · · Score: 1

      In CMOS technologies, NAND is actually the simplest full-coverage logic gate. 4 transistors. 6 for an AND, or OR gate. NOT is a 2 transistor gate (see CMOS inverter); however, you can't build other gates out of NOT gates.

      NAND gates are easier in the most common IC logic processes, which is why most logic processes base off of NAND.

      DL, DTL, RTL, are all fun and everything, but pretty useless in the real world, and TTL has mostly been superceded by CMOS. Real circuits are based on NAND gates, or on larger-scale AND-OR-INVERT combination gates.

      --

      ---
      Mod me down, you fucking twits. Go ahead. I dare you.
      (I read with sigs off.)
    18. Re:WOW by Kiryat+Malachi · · Score: 1

      I'd like to note one thing about your statement: NAND is cheaper in terms of transistors than OR or AND in a CMOS process. This is why most CMOS-based chips are NAND-based.

      --

      ---
      Mod me down, you fucking twits. Go ahead. I dare you.
      (I read with sigs off.)
    19. Re:WOW by Kiryat+Malachi · · Score: 1

      Wrong.

      Show me how to build an AND or OR gate in CMOS with less than 4 transistors. I will happily show you how to build a 4 transistor NAND gate. (roughly halfway down the page.)

      --

      ---
      Mod me down, you fucking twits. Go ahead. I dare you.
      (I read with sigs off.)
    20. Re:WOW by mOdQuArK! · · Score: 1

      Use transmission/pass-gates. Of course, then you have signal integrity problems...

    21. Re:WOW by Kiryat+Malachi · · Score: 1

      Hey, that's cheating.

      (Smartass.)

      --

      ---
      Mod me down, you fucking twits. Go ahead. I dare you.
      (I read with sigs off.)
  17. How long before someone ports linux to them? by Viol8 · · Score: 1

    Well , kind of inevitable really.

    1. Re:How long before someone ports linux to them? by Tony+Hoyle · · Score: 1

      Shall we get the 'Can you imagine a beowulf cluster of these' posts out of the way to.

      And let's not forget the timeless 'In soviet russia, Lego builds *you*'.

    2. Re:How long before someone ports linux to them? by swv3752 · · Score: 0

      And of course: In Korea only the old build Legos.

      --
      Just a Tuna in the Sea of Life
    3. Re:How long before someone ports linux to them? by awolk · · Score: 0

      No, it's
      In Korea, electronic computers are only for old people

  18. From the article: by IO+ERROR · · Score: 1, Funny
    Before you read on, if you have had trouble accessing this page then please accept my apologies. The previous bandwidth issues should be solved now that I have a new server.

    That is, until /. arrived....5:30am and the site's already running slow. How much longer you think it will hold up?

    --
    How am I supposed to fit a pithy, relevant quote into 120 characters?
    1. Re:From the article: by advocate_one · · Score: 1

      5.30 am... yeah, you yanks aren't awake yet... this is an European slashdotting... go team...

      --
      Donald 'Duck' Dunn: We had a band powerful enough to turn goat piss into gasoline.
    2. Re:From the article: by goneutt · · Score: 0

      This is why my "favorites" list has a section for "check after the /. ing is over"

      --
      Bacardi + slashdot = negative karma.
  19. Clever, but not exactly practical by f(Root) · · Score: 1

    It's quite clever how you can do this mechanically. It's a reminder of the old mechanical code-breaker computers. However, imagine trying to make a 'pocket' calculator out of these. It would be massive. Computers the size of rooms again...

    Someone should make a Lego computer, and then run a CAD program on it, just for irony's sake.

    --
    Programmers never get old.. They just can't C as well.
    1. Re:Clever, but not exactly practical by HyperChicken · · Score: 3, Insightful

      Yeah, most "hacks" are indeed pointless. That doesn't make them any less fun.

      --
      Free of Flash! Free of Flash!
    2. Re:Clever, but not exactly practical by Ziviyr · · Score: 2, Funny

      Someone should make a Lego computer, and then run a CAD program on it, just for irony's sake.

      I'd be happy with an BF interpreter made out of Lego...

      --

      Someone set us up the bomb, so shine we are!
    3. Re:Clever, but not exactly practical by Decimal · · Score: 1

      How would you power it, though? You've got mechanical resistance on each gate, and you hook 1,000 or so of those together, you'd need a lot of force on one end to move the rest of them.

      --

      Remember "Bring 'em on"? *sigh
  20. Imagine by Anonymous Coward · · Score: 2, Funny

    .. a Bionicle cluster of these ...

  21. "Apple iBricks Launch delayed by goldcd · · Score: 3, Funny

    and projected power decimated."
    A spokesman said Steve would only use the white bits of Lego.

  22. NAND gates... by Gopal.V · · Score: 1, Funny

    The road to hell is paved with NAND gates.

    -- /usr/bin/fortune

    1. Re:NAND gates... by greenguy · · Score: 1

      Huh. I thought the road to hell was paved with Bill Gates. Or at least by him.

      --
      What if I do the same thing, and I do get different results?
  23. As in Bill? by Phidoux · · Score: 1

    Oh crap!

    1. Re:As in Bill? by Anonymous Coward · · Score: 0

      Yes, I had thought it was some new initiative from MS.

      Any parallel between the difficulty in prising apart two 2x1 Lego plates and prising apart IE from Windows?

  24. Competition by anum · · Score: 2, Funny

    How long will we have to put up with this LEGO monopoly? Will Megablox answer this challenge? That upstart Knex, perhaps?

    Back in my day we had to build our mechanical analogs of digital circuits out of Tinkertoys! And we liked it! What's this new fangled "plastic" stuff anyway. Wood! that's the way to go.

    On a slightly more serious note: If we had built some of these in my CS3?? class instead of just diagraming them on paper I might have paid more attention. But I doubt it.

    --
    I don't think, Therefore I'm not.
    1. Re:Competition by Ziviyr · · Score: 1

      What's this new fangled "plastic" stuff anyway. Wood! that's the way to go.

      Kids these days, back in my day we didn't have the luxury of wood, all we had was volcanic ash that we molded after mixing it with the bodily secretions we used to drag ourselves around on!

      --

      Someone set us up the bomb, so shine we are!
    2. Re:Competition by berbo · · Score: 1

      Well back in my day, we solved problems with bricks the old fashioned way, and we didn't need no stinking 'logic'. We used them to beat up people who disagreed with us.

    3. Re:Competition by fishbowl · · Score: 1

      "What's this new fangled "plastic" stuff anyway. Wood! that's the way to go."

      Do you know the tragic early history of the Lego company?

      There is a good reason they were driven to innovate plastic toys.

      --
      -fb Everything not expressly forbidden is now mandatory.
  25. OT: Question for Slashdotters by rcs1000 · · Score: 1

    I would love to play with logic gates like these. Trying to create early computers, or just a simple "addition machine" with a series of logic games would be of immense intellectual interest.

    *But* physical stuff is a pain; has anyone put together a program that allows me to simulate basic electronic gates, and "build" these things. It would help me understand a lot better how computers' internal logic work.

    Thanks,

    Robert

    --
    --- My dad's political betting
    1. Re:OT: Question for Slashdotters by a24061 · · Score: 4, Informative

      If you mean software to simulate building digital circuits out of gates, look at TKGate.

    2. Re:OT: Question for Slashdotters by Ford+Prefect · · Score: 0

      *But* physical stuff is a pain; has anyone put together a program that allows me to simulate basic electronic gates, and "build" these things. It would help me understand a lot better how computers' internal logic work.

      The 'real thing' is actually surprisingly easy - with a power supply, a breadboard, logic chips and some LEDs, you can build your own flip-flops, counters and suchlike.

      I did a short course in digital microelectronics at university, and it was great fun. I was busy building my own, non-course-required, multi-digit adder when I ran out of components, and had to get back to the actual course material. If you've got any knowledge of programming and Boolean logic, the basics are pretty easy - I found it far easier than the course on analogue stuff involving impedance, resistance and whatever, i.e. lots of complex numbers. I've forgotten most of that already. :-)

      You can probably get pre-organised logic kits from Maplin (UK) or Radio Shack (US) - failing that, I'm sure other Slashdotters will know where.

      --
      Tedious Bloggy Stuff - hooray?
    3. Re:OT: Question for Slashdotters by mewphobia · · Score: 1

      what's wrong with using the contructs in most programming languages? | & ^ etc.

    4. Re:OT: Question for Slashdotters by Anonymous Coward · · Score: 0

      One option would be to use vhdl and design everything from the ground up, without worrying about the wires running everywhere. you can build registers, have timing signals, and all types of adder fun.

      if you want to play with things on the actual gate level, pencil and paper will be the cheapest solution. its a PITA to sit at a breadboard without having a good block diagram.

      Trolling some universities computer engineering websites for notes may help you as well.

    5. Re:OT: Question for Slashdotters by Anonymous Coward · · Score: 0

      and there is also Klogic for building gate circuits, etc. Found here:Klogic home page

    6. Re:OT: Question for Slashdotters by Anonymous Coward · · Score: 0

      Back in college (late 70's) we simulated logic gates (AND, OR, NOT, NAND, NOR) by writing them as FORTRAN functions, then passing variables as function arguments. From those function calls, D, T and JK Flip-Flop functions were created, also callable, and so on. As tedious as it might sound, it was fun.

      Today, you could do the same in C or Java.

      The best part of course would be not having to punch cards, submitting to a System\370 waiting hours for the printout.

    7. Re:OT: Question for Slashdotters by Anonymous Coward · · Score: 0

      Another good choice might be Altera Quartus II (used to be Maxplus II). It's a pretty comprehensive software package and it's free.

    8. Re:OT: Question for Slashdotters by AllUsernamesAreGone · · Score: 1

      Just a note...

      Maplin Electronics is getting to be a bit of a joke these days. Four or five years ago I'd have agreed with you, but the range of components they have is dropping year on year to be replaced by standard consumer electronics.

      However, in the UK RS Components now supplies individuals as well as companies and their range of stuff is vast. You don't get the dead tree catalogues unless you've got a commercial account (which, I believe, still requires a minimum monthly order) but if you register online you can get at all the details of the products on the website and there's no minimum order size so it's great for hobbyists.

    9. Re:OT: Question for Slashdotters by mozingod · · Score: 1

      We use the Altera MAX 7000S in our digital systems classes. The software it uses, Quartus II. is very easy to use and can do pretty much anything you'd ever need. All you have to do is wire up the board once, then download your projects there. 16 dip switches, 4 push buttons, 2 seven segment displays, 16 leds, and pin outs for everything. Plus PS/2, VGA ports, and a bunch of voltage/temp/IO sensors. Nifty little device...

    10. Re:OT: Question for Slashdotters by Ford+Prefect · · Score: 1

      Maplin Electronics is getting to be a bit of a joke these days. Four or five years ago I'd have agreed with you, but the range of components they have is dropping year on year to be replaced by standard consumer electronics.

      Admittedly I haven't bought components from Maplin for years - Tandy would have been another suggestion, but I was recently surprised to learn that they'd disappeared from the UK ages ago.

      I usually get my stuff from a titchy little electronics shop in Derby called R.F. Potts. They don't appear to have any kind of internet presence (beyond random posts in tech mailing lists from people saying how great it is) - but it's exactly the sort of place you'd go to if you needed a sync-on-green graphics card, stepper motor, obscure rechargeable battery or giant obsolete monitor. As a result, my knowledge of other electronics suppliers is a bit lacking, and I always think other stores should have assistants who clamber up ladders and rummage through wooden drawers to find that precise shape of transistor heatsink you were looking for...

      Okay, RS has more stuff, but it's nowhere near as much fun. :-)

      --
      Tedious Bloggy Stuff - hooray?
    11. Re:OT: Question for Slashdotters by scampiandchips · · Score: 1

      The programm i have been using for a p/t distance learning course is called OrCAD and happily runs all the basic logic gates, along with some of the more common IC's available. The demonstration version is free but only lets you have up to 60 componenets, but that should be more than adequate for most mini projects. The program (like any) takes a little while to get used to though.

      cheers
      steve

      --
      There are things we know we don't know and things we don't know we don't know. - Donald Rumsfeld
  26. Speed? by the_Twisted · · Score: 3, Funny

    At how many (K)hz will this thing operate when it's finished? How many FLOPS will it achieve? Could you overclock it by adding oliveoil to the gears?

    1. Re:Speed? by Agret · · Score: 1

      The only possible way to overclock it would be to have a pirate standing in the middle....it wouldn't make it any faster but the coolness level is through the roof and with the lego background it should be easy to add a pirate from the pirate set. Maybe I shouldn't cross the streams though....

      --
      Have you metaroderated recently?
    2. Re:Speed? by Taladar · · Score: 0

      That would be case-modding, not overclocking ;)

    3. Re:Speed? by PhuCknuT · · Score: 1

      I think the only possible prefix to that Hz is m (as in milli).

  27. This post was lifted verbatim from Metafilter by Jude+T.+Obscure · · Score: 5, Interesting

    Have a look - it's identical.

    1. Re:This post was lifted verbatim from Metafilter by Jude+T.+Obscure · · Score: 1

      That's maybe why it stuck in my memory - it's not the usual mefi fare in these politics-soaked days. Plus it was a very odd sensation reading it on the front page and being certain I'd seen the very same post somewhere yesterday.

    2. Re:This post was lifted verbatim from Metafilter by AndroidCat · · Score: 1

      So that's why that guy kept e-mailing me for time-machine parts. He wanted to use it to scoop Slashdot.

      --
      One line blog. I hear that they're called Twitters now.
  28. So what if it is identical by RedLaggedTeut · · Score: 1

    Maybe the author submitted to two sites?

    --
    I'm still trying to figure out what people mean by 'social skills' here.
    1. Re:So what if it is identical by TheOnlyCoolTim · · Score: 1

      I only submitted to Metafilter.

      Tim

      --
      Omnia vestra castrorum habetur nobis.
    2. Re:So what if it is identical by Thud457 · · Score: 1
      Freakin' AWSOME!!!! BoingBoing bites, doesn't get attribution right, either!

      If you're the original poster of the story to MetaFilter, I would just like to apoligize for hijaaking your writeup.

      And the damn trackback info is wrong.


      I was only : A) jerking around the slashdot editors, B) trying to bring something cool and nerdly to a wider audience, something slashdot's been kind of light on lately.


      Please feel free to resubmit the exact same story for the obligatory Taco duplicate story post. (Which would be extra ironic in this case...)

      --

      the preceding comment is my own and in no way reflects the opinion of the Joint Chiefs of Staff

    3. Re:So what if it is identical by TheOnlyCoolTim · · Score: 1

      It's no biggie. Just change up the text next time.

      This is the second time I've done something like this. The first was when I submitted Don't Copy That Floppy to bytemonsoon, someone linked it on slashdot, and it spread to be much bigger than it was before...

      Tim

      --
      Omnia vestra castrorum habetur nobis.
  29. I don't think this will work by Anonymous Coward · · Score: 0

    So, how does he address the issue of fan out? As far as I can tell, if changing an input will cause a computer built from mechanical lego to do N transitions, then the mechanical force for all these N transitions would have to be applied at the input.

    This doesn't scale and will hit the limits very soon.

    1. Re:I don't think this will work by julesh · · Score: 1

      So, how does he address the issue of fan out? As far as I can tell, if changing an input will cause a computer built from mechanical lego to do N transitions, then the mechanical force for all these N transitions would have to be applied at the input.

      This doesn't scale and will hit the limits very soon.


      He doesn't. But it would be trivial to add force amplifiers.

      This page is about torque amps, which aren't quite what you want here, but the design might serve as a starting point. :)

  30. Re:Warning: obvious joke by Proud+like+a+god · · Score: 1

    Someone, look if you dare but mod that Troll.

  31. Friction Losses and Gain Stages by Fleetie · · Score: 4, Interesting

    There is much friction inherent in these, and as shown, no "gain stages" to overcome these losses. So the "fan-out" would be appalling as implemented here. This would preclude their use to build anything other than the simplest logic constructs. However, I think it would not be too hard to add "gain stages" to act as "buffers", which could, for example, use falling weights to act as "supply rails" to increase "fan-out", thereby facilitating construction of far more complex circuits. Martin

    --
    "Absorbing your worst..."
    1. Re:Friction Losses and Gain Stages by Anonymous Coward · · Score: 0

      And for the English speaking amongst us??

      "Your ideas are intriguing to me and I wish to subscribe to your newsletter." -HJS

    2. Re:Friction Losses and Gain Stages by oliverthered · · Score: 1

      Ok, I'll try to explain it in the simplest terms possible.

      find something.
      put it on the carpet.
      try to push it, and judge how easy it was.
      find something else.
      put it on the carpet inline with the other thing.
      try to push it .....
      repeat until it gets hard/impossible to push the things along the carpet.

      Try to work out a way to make it easier to push the things along the carpet, say by standing them on end like domino's.

      --
      thank God the internet isn't a human right.
    3. Re:Friction Losses and Gain Stages by famebait · · Score: 1

      I think a more realistic version would have each gate draw power individually from a main rotating shaft, whichc would double as clock.
      Much like a machanical cash register.

      It might also be possible to come up with even simpler designs if the gate is powered: the swithching could depend on pushing unstable systems in one direction or the other. This would obviate both the need for precisely metered responses, and reduce the needed power on the inputs, thus improving fan-out.

      --
      sudo ergo sum
    4. Re:Friction Losses and Gain Stages by Anonymous Coward · · Score: 0

      "How can 59,054,087 people be so STUPID? God help us all."
      Because they didn't all go out, buy a gun (you can still do that in the US i think) and take out the bush voters!

    5. Re:Friction Losses and Gain Stages by oliverthered · · Score: 1

      A main rotating shaft would have torque problems.

      A shaft a few meters long (made out of that Lego shaft stuft) would probably be at least half a turn out of sync along it's length, maybe more.

      You could defiantly use a 'cam' type system though, a bit like a sewing machine.

      --
      thank God the internet isn't a human right.
  32. A couple problems. by PenguiN42 · · Score: 4, Interesting

    As another poster mentioned, there's no gain in these devices, so after a few stages of friction loss and imperfections in the mechanisms, the whole thing will lock up. Electronic gates have inherent gain, and thus are resistant to noise and slight differences between gates.

    Another problem is the way his clock works -- the clock has to go to zero before the set or clear bits can change. This won't happen in a real circuit -- generally everything changes just after the clock rises. One solution is some sort of two-phase system, where alternate flip flops use the rising and falling clocks, but I'm not sure how much this would limit the circuits you can build.

    He mentions that "It is possible to build an edge detector for the clock signal. It requires a few more NAND gates. The advantage of doing this is that it no longer matters when the clock signal goes back to 0 and the indeterminant state is avoided." But I want to see it in action before I believe it.

    Giving the gates gain may be possible, too, but it would require powering each gate, either with electrical power or some sort of funky mechanical setup.

    --
    The following sentence is true. The preceding sentence was false.
    1. Re:A couple problems. by RPI+Geek · · Score: 3, Informative

      Giving the gates gain may be possible, too, but it would require powering each gate, either with electrical power or some sort of funky mechanical setup.

      When I read the title of this article, I tried coming up with a design in my head of how I would do this, and my idea is to make the clock pulse out of a rotating shaft with cams (easy enough for Lego pieces). The cams are what would actually power the pieces, thereby eliminating the effects of backlash and fanning. The only problem comes with a lack of torque, but it would be easy to simply tie in more motors (or more hand-cranks) to get more power down the line. I think that this is a more elegant solution than having a person manually slide a shaft back and forth for the clock pulses.

      --

      - "Nobody came out that night, not one was ever seen. But Old Man Stauf is waiting there, crazy sick and mean!"
    2. Re:A couple problems. by Mad+Bad+Rabbit · · Score: 1
      As another poster mentioned, there's no gain in these devices, so after a few stages of friction loss and imperfections in the mechanisms, the whole thing will lock up.

      Maybe mechanical gain could be obtained from a pulley-weight-escapement mechanism, like the one from the Lego grandfather clock ? Each state-change of an output rod might allow the escapement to turn a notch and provide extra power (so the input rods aren't under any extra load).

      --
      >;k
    3. Re:A couple problems. by Yobgod+Ababua · · Score: 1

      Yeah... the first thing I thought was that he needs a unity gain buffer to help with fan out issues.

      The second thing I thought was that this sort of mechanical logic would prove much easier using an asynchronous handshake (One example description at: http://caltechcstr.library.caltech.edu/312/) than a standard clock pulse.

      Using cams or a rotation-based system helps the power distribution, because you can always add additional 'power rails' as the circut grows... your first gate doesn't need to push through everything else.

      The lack of torque is usually addressed through a simple gear-down at the motor.

      Now that this has got me thinking, I'm wondering about circuts that are powered by rotation, but store their signals through the movement of rods (either back and forth or up and down). *sigh* and my lego gears -were- all neatly packed away in the closet...

  33. The Tinkertoy Computer That Plays Tic-Tac-Toe by rmathew · · Score: 3, Interesting
  34. Logic gates... by Anonymous Coward · · Score: 0

    Try to combine some and make a binary clock, instead of some old grandfather clock.

    I would wear it !

  35. Slippage and friction... by tonywestonuk · · Score: 1

    .... Means, like what already has been said, that after a few gates, the movement (and therefore the on/off state) deteriorates to nothing. However, this is exactly what would happen in an electronic circuit, if said circuit was built like this (Inputs and outputs only, no Power). Now I'm wondering, if it is possible to add a third rotating input to the gate, for power, so the logical output was of the same (or better quality) of the inputs....

    1. Re:Slippage and friction... by Lumpy · · Score: 1

      no but using springs would. the toggle state could be built so that the output has two helper springs that resist state change unles they are acted upon by a reset lever. trigger the gate and the output lever get's snapped into position with the help of the spring and the only mechanical advantage over that spring is the reset bit that is sent before every state change.

      add the input to also have a spring to help avoid partial detection and you are closer.

      problem is that you now need a stronger clock to hit the reset every cycle. on all your gates, if an input is no longer being pushed then reset the output as in the case of an Or gate.

      lots of cams and gears. you certianly will need a continous movement to act as the amplification to reset the springs used for bounce and hysterisis reduction.

      --
      Do not look at laser with remaining good eye.
  36. Obligatory SF references by dpilot · · Score: 1

    Brings to mind the book, "Souls in the New Machine" by Sean McMullen. (http://www.sfsite.com/08a/soul62.htm) Don't want to say more without SPOILERS, but it envisions a computer in the future where processing elements are slaves, and there's a darned good reason why it's not electronic.

    Also on the fringe, one story from "Tales of the Flying Mountains" by Poul Anderson. The computering technology on a warship is TEMM - Thermionic Emission Micro Miniaturized, selected for its radiation hardness. It turns out that when you microminiaturize, you don't even need thermionic emission. Take a look at FEDs - Field Emission Devices.

    --
    The living have better things to do than to continue hating the dead.
    1. Re:Obligatory SF references by WhiteDragon · · Score: 1

      I've read Souls in the Great Machine, and its sequels, The Miocene Arrow and Eyes of the Calculor, and I thought they were all three really good. If you have ever read Dune, Butlerian Jihad (approved by the estate of Frank Herbert), you will find some other interesting human-powered computation. Also, The Difference Engine by Gibson and Sterling is a great book about mechanical steam powered computing. (i.e. if Mr. Babbage had successfully built the Analytical Engine, the industrial revolution and the information revoultion could have occured at the same time.)

      --
      Did you mount a military-grade, variable-focus MASER on an unlicensed artificial intelligence?
    2. Re:Obligatory SF references by dpilot · · Score: 1

      Never knew there were sequels to Great Machine. I had borrowed it, and hadn't really remembered the author's name, until looking it up for the grandparent post. I did enjoy it, though mostly for the concepts. Some of the characters were a bit overdone.

      As for the Dune stuff, they just invoked the Really Smart People clause in the name of avoiding thinking machines. At the same time, they didn't really do that much to make them interesting. For contrast in a short story, Isaac Asimov introduced a person/profession called a "Mnemonic" in a short story. Partly training, partly breeding, (and today's parlance he probably would have used genetic engineering) these people soaked in the knowledge, counting on free association and such for useful information to come out when needed. The person/profession was new, and not fully explored as of the time of the story. The idea was that a machine couldn't know what it knew - in other words, stored information was just bits - but a person could.

      On the interesting side, these people thus trained would up being minux-6-sigma social incompetents as a side effect of their breeding/training. It lent an interesting flavor - what good is it to know so much if nobody wants to listen to you when you "do your job?"

      And of course, if we're going to talk about computing with human elements, how can we leave out the Drummers, from "The Diamond Age."

      --
      The living have better things to do than to continue hating the dead.
    3. Re:Obligatory SF references by WhiteDragon · · Score: 1

      Aye, I read that Asimov story back in the day, and it was pretty good. Of course the drummers were really more like hosts to nano based computation in Diamond Age though. Still, a great book. I personally liked it better than Snow Crash.

      --
      Did you mount a military-grade, variable-focus MASER on an unlicensed artificial intelligence?
    4. Re:Obligatory SF references by WhiteDragon · · Score: 1

      Sorry to reply to myself, but Diamond Age also had some nano-mechanical computers, in the spirit of the article on lego logic gates.

      --
      Did you mount a military-grade, variable-focus MASER on an unlicensed artificial intelligence?
  37. Needlessly overcomplicated gates. by Kickasso · · Score: 1

    Cog wheels? Rack and pinion? Totally overengineered.

    1. Re:Needlessly overcomplicated gates. by multipartmixed · · Score: 1

      No kidding.

      With push and pull as 1 and 0, I could make a D-type flip-flop from a piece of scotch tape.

      --

      Do daemons dream of electric sleep()?
  38. An 'Nor' gate by tonywestonuk · · Score: 1

    Ok, off the head design .... (I'll have to get some Lego).

    1) Rotateing axle power input into gate, with medium cog rotateing freely...

    2) Axle for logical output... with medium sized cog near to power input cog, but not connected

    3) Power input is coupled to logic output via fairly loose rubber band - output will turn in unison with power input, with no other parts added.... both input/output cogs will be turning same direction

    4) Add one logical input.... This will be connected via rubber band to a mechanism to shove a cog inbetween the power input, and output cogs, engaging, and forceing the output axle to reverse, as cog drive will not slip unlike the rubber band drive. If the logical input is rotateing one way, the cog will engage, If the input rotates in other direction, cog will be pulled from between power input, and output, and so both will rotate in unison again

    5). Second logical input added in same way.

    Damn, I'm getting some lego now, to try it out!!!

  39. Re:My question by AndroidCat · · Score: 2, Funny

    He could build a Lego vacuum-cleaner made of Lego, but that's like canabalism. Before he knew it, he'd have a Legoid-lifeform eating his other projects and growing stronger.

    --
    One line blog. I hear that they're called Twitters now.
  40. Nothing new under the sun by Ancient_Hacker · · Score: 1

    My wife's 1881 pump-organ has a patent for its arrangement of knobs that control the various sound-modifying apertures. They didnt have names for these concepts at the time, but looking it over with a modern eye, one can see it includes AND gates, OR gates, and open-collector wired-OR pullups, all made out of wooden rods and leather straps.

    1. Re:Nothing new under the sun by Rob+the+Bold · · Score: 1

      Got any pics of that?

      --
      I am not a crackpot.
    2. Re:Nothing new under the sun by XanC · · Score: 1

      My mind boggled trying to picture your wife with an artificial heart built in 1881.

    3. Re:Nothing new under the sun by shrubya · · Score: 0, Offtopic

      No, he said she has a "pump-organ". This is a 19th century euphemism. It is a tool for pumping up a particular organ.

      To spell it out in modern terms ... a PENIS ENLARGER. The wooden rods and leather straps should make much more sense now.

  41. falling weights? by oliverthered · · Score: 2, Insightful

    Why not drive the clock at various stages and take some power out of it.

    --
    thank God the internet isn't a human right.
  42. great pics of Hillis' original TT machine... by museumpeace · · Score: 3, Interesting

    here are some nice pictures of Hillis' tinker toy tic tac toe machine...It predates his work on the Connection Machine and Thinking Machines Inc.
    Goes to show you how strong the mind can become with a little exerecise in logic. Other posters are right about how limited the potential circuits are with lossy elements but all the the same, kudos to Lego for hoping that at least some of us consider thinking a form of recreation.

    --
    SLASHDOT: news for people who can't concentrate on work or have no life at all and got tired of yelling back at the TV.
  43. If he built a switch... by Otto · · Score: 1

    Electrical gates have an inherent gain in that they all use transistors, with the inputs to the gate being inputs to the transistors, and the outputs of the gate being whether or not the transistors flip on, allowing power to come from a power source to the output of the gate. This is why electrical gates all require an always on power input, sort of thing. But it avoids the whole friction/power loss problem.

    If he built a switch (kinda like a transistor) out of LEGO and then used a motor to provide the power to the mechanical action of the devices, he could avoid the problem he's having with the lack of gain on the devices (after 3 or 4 devices in a row, the whole mechanism stops due to wobble room). But then it would be partly electrical, I grant you, since you gotta have a powered input of some type.

    With some cleverness about arrangement, you could make the whole setup only require one motor, although using two or three would be more likely for ease of construction.

    --
    - Give a man a fire and he's warm for a day, but set him on fire and he's warm for the rest of his life.
  44. Before one gets too enthused... by Ancient_Hacker · · Score: 1
    Before you get all enthused about nanotech mechanical computers, pls consider:
    • The mass of a gate goes down as the cube of its linear dimension, while effects like friction and surface tension and static electricity only go down as the square. So if you make a lego gate 1000 times smaller, it only weighs a billionth as much. That's good. But it has only a millionth the surface area, which is bad, proportionally. It is 1000 times more bothered by friction, surface tension, and static. That's very bad. You end up with a gate that is hard to move (because of friction), tends to stick in one position (due to surfaqce tension), and is very likely to flip or stick (due to static electricity).
    • As others have mentioned, there's no amplification, so one gate can't effectively drive mroe than one other gate (and with difficulty at that because of all the friction).
    • There's nothing like WIRES that can carry mechanical motion without lots of mechanical inertia. So you can only build small clusters of gates if you want any speed at all.
    • Mechanical gadgets you may have noticed WEAR OUT after a few million operations.
    • Mechanical gadgets have a relatively high error rate. It's hard to move something small a few million times without it glitching at least once or twice due to dust, vibration, cosmic rays, microbes,viruses, or brownian motion.
    • At best you're maybe talking about gate propagation times in the low milliseconds, while electronics is working in the middle picoseconds. That's NINE orders of magnitude slower. Slower than a diesel Vega. Slower than a Turing machine emulating a Cray. Really slow.
    1. Re:Before one gets too enthused... by multipartmixed · · Score: 1

      > There's nothing like WIRES that can carry mechanical motion
      > without lots of mechanical inertia.

      Actually, *that* particular problem could be addressed by making use of hydraulic fluid lines.

      Actually, that might also solve the amplification problem, the same as the master cylinder / brake cylinder in a typical piston (disc) brake system. Although 1LOTD tells me I'm missing something, because that sounds suspiciously like free energy. :)

      It *does* mean at the very least that buffers could be built -- "power brakes".

      --

      Do daemons dream of electric sleep()?
    2. Re:Before one gets too enthused... by Anonymous Coward · · Score: 0

      Several of these issues are addressed in the literature, actually.

      Friction, for example, just doesn't work the same with atomically-precise parts at a microscale. There are apparently quite a few good, low-friction designs for rotating bearings, sliding surfaces, and so on. Many of these have been run in detailed simulation; others have actually been built and measured. (Richard Feynman actually discussed the feasibility of using unlubricated surfaces in nanomachines almost 50 years ago, and concluded that it could work pretty well.)

      Don't automatically assume that your ordinary engineering knowledge applies at this scale. There's already a lot of scientific literature in this topic.

  45. Quatro and Quadro? by doodlelogic · · Score: 1

    If Lego really are building a system called Quatro, might they not run into trouble with Quadro - the system that lets you build your own climbing frames, slides and other life-size kit?

  46. wasteful by mennucc1 · · Score: 1

    I think that starting from logic gates to accomplish things in Lego is a waste. There are usually more legoish ways to do the same thing.
    EG, when I was very young I assembled a LEGO calculator that would compute simple additions; unfortunately I do not remember how it worked, but I am pretty sure it did not use gates (at the time I did not know what gates where, I did not even know what " digital" meant).
    ps: if you do not believe me, ask confirmation to my parents: they are the ones bragging this memory around; I personally had at a certain point completely forgotten the Lego calculator.

    1. Re:wasteful by Anonymous Coward · · Score: 0

      I don't belive you. Could you please send me the names and email address of your parents?

  47. What I first envisioned by RealProgrammer · · Score: 1

    before I read the Fine Web Page in Google's cache, was a logic board with individual gates housed on standard lego pieces. You'd build your circuit by placing the lego pieces on a lego board.

    I'm not sure how the traces between the pieces would work, but legoCAD would be fun, like those 10-in-1 electronics kits from Radio Slack.

    --
    sigs, as if you care.
  48. Sorry by Anonymous Coward · · Score: 0
    All the gates except XOR are here, and he goes on to develop a clocked flip-flop.

    Sorry, but Kerry Labs already patented that.

  49. Re:Let me guess by Anonymous Coward · · Score: 0

    I think you're supposed to send him a pair of numbers, and he'll put the total in his signature.

  50. *rolls eyes* by ClosedGL · · Score: 1

    Wake me up when the obligatory Linux installation has completed.

  51. OT - word rant by Anonymous Coward · · Score: 0

    you haven't priced Lego's what? that's a possessive apostrophe.

    they're called "Lego" or "Lego bricks" if you must, but not "Legos" and certainly not "Lego's".

    Lego. Plural and singular. Same f*cking word! Why is this so hard for /.ers to grasp?

    1. Re:OT - word rant by Anonymous Coward · · Score: 0

      Capitalization--it's what we do to the first letter of a sentence. You spout about possessives vs. contractions but cannot even handle basic, first grade skills such as capitalization yourself. Why is it so hard for retards like you to grasp?

      Does it make the parent's post any less readable or understandable? Do your literary skills make your post any more or less understandable?

  52. All we need is 48million of these gates!! by jhandel · · Score: 1

    with just 48 million of those and 165 million Square Miles we could make a model of a Pentium 4...

    The only question left is what is the kinetic force needed to run a single cycle through or 165 million square mile processor?

  53. In the interests of full disclosure: by Thud457 · · Score: 1
    http://www.metafilter.com/mefi/37826

    damn, can't post annonymously!!!!

    Yeah, I was being a smartass. It's still cool though.

    --

    the preceding comment is my own and in no way reflects the opinion of the Joint Chiefs of Staff

  54. Here's an honest query: by Thud457 · · Score: 1

    If QM effects come into play at a certain threshold of scale, do both mechanical and electronic implementations of the same gates give the same results under QM? That would be weird.

    --

    the preceding comment is my own and in no way reflects the opinion of the Joint Chiefs of Staff

    1. Re:Here's an honest query: by 4lex · · Score: 1

      I think it's weirder than you might expect: a fully new computing paradigm arises with quantum computing, without regard to your hardware, be it electrical, magnetic, mechanic or raw photons. New logic gates (how about sqrt(NOT) such as (sqrt(NOT))^2 = NOT?), new algorithms (e.g. Quantum Fourier Transform) and a new complexity class (BQP)...

      --
      My journal. Mainly about freedom.
  55. It's a fair cop... by Thud457 · · Score: 2, Informative
    Yeah, I was being a jackass and seeing if the editurs were awake.

    It's still cool though.

    --

    the preceding comment is my own and in no way reflects the opinion of the Joint Chiefs of Staff

  56. Priceless...... by bigmike_f · · Score: 1

    Number of logic gates in UNIVAC XXXX Number of legos needed for UNIVAC XXXX Being able to outsource to preschools Priceless Or something like that.

  57. I don't think it would work... by Se7enLC · · Score: 1

    It looks like a great idea, but you'll notice that there are no movies. It is because only certain state changes will work. There is no clear way for it to go from an output of 1 back to an output of zero:

    Take a look at the And gate: 0 * 0 = 0: Notice that you can see through the section between the yellow cross-bar and the gear slides. This is because there needs to be some play in the system between the 0*0 and 0*1 cases. It is clear from the photos that the gear slides are *pushing* the crossbar forward, so that when both inputs are forward, the output will be one. The question is, what pulls the crossbar back? Is there a hidden rubberband? Some magical force? Or did they not consider the case of the transition from 1*1=1 to 0*0=0?

    I could very well be missing something (frankly, I hope I am missing something, this seems like far too obvious a thing for the designer to miss). Without a better picture of the design, I can't see what it would be, can somebody fill me in?

    1. Re:I don't think it would work... by Se7enLC · · Score: 1

      In the bad-nature of replying to my own post, take a look at 1 NOR 1=0. You can see that the crossbar is still diagonal from the previous state. If the previous state had been an output of 1, the output would still be 1

    2. Re:I don't think it would work... by Anonymous Coward · · Score: 0

      there is a fixed cross bar at the other end, so moving any input back to 0, in the case of the AND gate, pushes the whole mechanism back and hence the output goes to 0.

      The diagonal position of the moving cross bar is irrelavent - it's sole job is to ensure that, in the case of the AND and NOR, the output changes only when both racks are moved forward. Similarly it dictates the logic of the other gates too.

      I can provide a video if you would like.

  58. Re:degrading into '0.5's... by Anonymous Coward · · Score: 0

    You can fix that by using the logic result to open/close an air/steam valve. That way the value remains at a "constant" level depending on the air/steam pressure.

  59. Neal Stephenson ... Prophet? by H0ek · · Score: 2, Funny

    The work depicted in TFA seems awfully similar to the rod logic found in Neal Stephenson's book Diamond Age. This begs the question, is Neal a prophet? Has he foretold the future? Will I be getting my pizza in thirty minutes or else the Mafia comes and apologizes personally?

    Only time will tell. Until then, I have to thank this lego-builder for making my life more interesting. Now I know what I'm gonna do for Christmas, play with Legos and make logic gates!

    Lord, I'm such a geek.

    --
    H0ek
    Think you're smart? Prove you've got brains!
  60. Exact dup of Metafilter post by WasteOfAmmo · · Score: 0, Flamebait
    Perhaps when a submitter copies a story word for word from another site they should credit the other site.

    If nothing else perhaps the the submission page should contain a "source" field. It may be too much to ask the editors to catch this sort of thing (seems they can't even eliminate dups from their own site; sigh) but a quick look at the more popular geek sites (automated or otherwise) might be in order.

    Merlin.

  61. This is sort of cool, but... by pclminion · · Score: 3, Insightful
    Although I find this cool from a purely geek standpoint, and while I realize that the guy probably doesn't intend this as a practical device, I can't help but wonder... WHY use a binary representation?

    With electronic computers, binary makes sense. A capacitor is either charged, or not charged. A transistor is either conducting, or not conducting. It's HARD to make electronic devices with some fixed number of states other than two (let's disregard analog computation, with its infinite number of states, for now).

    Yeah, this thing is like Babbage's machine in the sense that it computes mechanically, but Babbage's machine wasn't binary. It's EASY to make multi-state mechanical devices.

    We shouldn't let our current computer technology make us too narrow-minded when designing new computer technologies. Binary representation is no Holy Grail, it's merely a convenience in the world of semiconductor electronics.

    1. Re:This is sort of cool, but... by nixterino · · Score: 1

      But a capacitor is analog - it can store charge in quanta of 1 electron. If you set a threshold, then it can represent a 1 or a 0.

      Same with a transistor - it's not either ON or OFF, but can vary completely between those two states in an analog fashion.

    2. Re:This is sort of cool, but... by multipartmixed · · Score: 1

      > It's EASY to make multi-state mechanical devices.

      Yes, but his representation for digits were the concepts "push", and "pull". It's a LOT easier to build a mechanical device which knows the different between "push" and "pull" than it is build one that knows about 10 different dowel positions.

      --

      Do daemons dream of electric sleep()?
    3. Re:This is sort of cool, but... by pclminion · · Score: 1
      Same with a transistor - it's not either ON or OFF, but can vary completely between those two states in an analog fashion.

      I understand that, but analog behavior is not conducive to precise computations. It's too susceptible to noise and other influences from the surrounding environment.

      There's a reason analog computers are no longer used. Digital is where it's at, and my point was that digital not not necessarily mean binary.

    4. Re:This is sort of cool, but... by Big_Breaker · · Score: 1

      Digital transistors are actually designed to minimize the operating regime in which the output varies with the input. They are designed to be completely on or completely off for the greatest range of inputs possible.

    5. Re:This is sort of cool, but... by maxpuppy · · Score: 0

      pclminion is more right than wrong but transitors are not simply on or off. Transitors are used in radios and amplifiers to achieve gain. Digital transitors have two "states" which are levels of output one of which can trigger another transitor and one of which can not. "Bad" transitors do not have the two states within the acceptable parameters. They can leak -- some gain or have insuffient output.

    6. Re:This is sort of cool, but... by LanceUppercut · · Score: 2, Informative

      Ternary "Setun" machines developed in Moscow State University in the 50's are well-known examples of non-binary computers.

    7. Re:This is sort of cool, but... by Teancum · · Score: 1

      Actually, the first electronic computer, the ENIAC, was a decimal (base-10) computer rather than a binary computer. It turns out that it creates many more hassles when you have to (in this case with a decimal computer) wire each number seperately. In other words, each number in the computing base has to have a seperate wire, or 10 wires for each digit.

      With a binary circuit, you only need 1 wire, which is either on or off. And to make all of the numbers from 0 to 10, you only need five wires instead of ten. Really this is more than a mere convenience, but a real practical issue that would substantially affect the complexity of any computing machine that does not use binary calculations. Even with the Babbage machine, there is a substantial increase in complexity by using connections that were not binary. I'm certain that if mechanical computers were to continue from what Babbage had done instead of the development of electronic computers we have today, they still would have gone binary just due to the practical nature of how everything was interconnected.

      Some "alternatives" to pure binary can include trits (trinary bits... made infamous with the programming language INTERCAL) where a single wire can have a positive, negative, or zero voltage. While this keeps the number of wires down, there are some practical engineering issues in trying to nail down detection circuits to keep the three states seperate, designing practical building blocks for trits can also be difficult. Also, maintaining a neutral zero voltage can be a royal pain in the behind when designing a circuit.

      An alternative to the base-2 system is also the base-minus-2 number system. This really is just a reworked binary system and even there the computational logic is a little more complex, as there are two carries and two borrow bits to a simple adder circuit. Base-minus-2 is a good way to shake up the thinking of a digital logic design class, however.

    8. Re:This is sort of cool, but... by Anonymous Coward · · Score: 0

      Actually, ENIAC used bit-simple binary coded decimal and ring counters. Binary doesn't always mean base-2.

      Bob S.

  62. Legos? pfff by Tablizer · · Score: 2, Funny

    I found a new device, the vacum tube, to be about 60% cheaper and faster than Legos. They are gonna make me rich, I tell ya, rich!

  63. I swear I've seen this... by alphax45 · · Score: 1

    before. On BBSPOT.com, about a month ago. I'm way too lazy to look it up, but if anyone else wants to/remembers go for it

    --
    K Man
  64. Speaking of mechanical computers... by sgage · · Score: 4, Interesting

    When I was a kid back in the 60's, I had a toy mechanical computer called 'Digicomp'. It was a funky conglomeration of springs-and-rods-and-plastic-things that you built from a kit, and programmed it by putting little pieces of tube over various tabs to affect the flip-flops.

    To operate it, you pushed a sliding thing in and out (a clock cycle). You could add and subtract and multiply and divide in binary, albeit rather small numbers. Hard to describe this thing, but it was very cool!

    Actually, the following year I got Digicomp II for Christmas, which ran by letting a stream of marbles flow through it by gravity, and these marbles toggled the flip-flops. Very cool again!

    These toys came with excellent little books on Boolean Algebra, and sure taught me a lot about the interface between binary math and physical things. I felt right at home when I started programming 6502's and Z-80's in machine language back in the day...

    Are these things, or anything like them, still around?

    - sgage

    1. Re:Speaking of mechanical computers... by amjacobs · · Score: 2, Interesting

      I did a little bit of searching and haven't been able to find anyone who still has one. I did find this site which has the digi-comp 1 implemented in knex.

  65. logic gates aren't clocked by Anonymous Coward · · Score: 0

    Yeah, flip-flops and latches and things are clocked.

    But an AND or OR gate doesn't take a clock, nor need a clock. Except for propagation delay, the output is always indicative of the input.

    1. Re:logic gates aren't clocked by RPI+Geek · · Score: 1

      But they ARE powered, and they COULD be clocked. It would be possible to do the equilavent of powering the logic gates, but it would add more hardware and this is a way to power them using already-necessary hardware. Electronic logic gates have a delay between changing the inputs and the outputs changing, so why not clock it? I suppose that aanother solution would be to evaluate the slop on the mechanical gates to determine the maximum length of gates that could be be passively run, and then for every 5th gate, say, clock it to get rid of that slop. Think of it as repeating a degrading signal.

      --

      - "Nobody came out that night, not one was ever seen. But Old Man Stauf is waiting there, crazy sick and mean!"
  66. No beowulf cluster reference yet! by iluvcapra · · Score: 1
    $ curl "http://science.slashdot.org/article.pl?sid=04/12/ 15/0226220&tid=222&tid=159&tid=14" | grep beowulf

    Nothing. I'm very shocked.

    ...

    howaboutabeowulfclusterof ZAP! ^D

    --
    Don't blame me, I voted for Baltar.
  67. Not true. by autopr0n · · Score: 1

    A device could be built with lots of small "force amps" around it, as the initial force disipates through the system, it could be boosted.

    --
    autopr0n is like, down and stuff.
  68. Well. by autopr0n · · Score: 1

    You could pre-set RAM memory. Looks like you forgot your circuitry class.

    --
    autopr0n is like, down and stuff.
  69. Well by autopr0n · · Score: 1

    Actualy, you can calculate the speed of the thing by figuring the speed that the compression of the rod is transfered through it. If you're pushing it fast enough, the rod will act more like a spring then a rod, assuming it dosn't melt or catch fire or something.

    --
    autopr0n is like, down and stuff.
  70. Wonko Really Is Sane by DLWormwood · · Score: 1
    Who do you sue if someone gets hurt building a tree house?

    Who don't you sue?

    And how about building the model airplane? That's pretty safe.

    The post made me think of something like this...

    Hmm, I guess the kid might glue his mouth shut or huff the glue. Guess we'll have to sue Borden or something.

    The U.S. is now officially a culture that slaps "Warning: HOT" on coffee cups. We have now trumped Douglas Adams' toothpick.

    --
    Those who complain about affect & effect on /. should be disemvoweled
    1. Re:Wonko Really Is Sane by LWATCDR · · Score: 1

      Actually it is better than that. Atlanta bread has this warning on it's cups.
      "Warning hot drinks served hot"

      --
      See my blog http://ilovecookes.blogspot.com/ for light hearted technical information.
  71. Oh ye of the modern age.. by Kwil · · Score: 1

    ..ever heard of a pendulum?

    --

    That Jesus Christ guy is getting some terrible lag... it took him 3 days to respawn! -NJ CoolBreeze

  72. City on the Edge of Forever by serutan · · Score: 1

    Didn't Spock build a time machine out of Legos at some point? Or was it Wesley Crusher?
    Well if they didn't they should have.

  73. Old news by LanceUppercut · · Score: 1

    This is old news. At least a month old. Actually, I can't get rid of a feeling that this was already published right here, on Slashdot.

    1. Re:Old news by k512-arch · · Score: 0

      i actually think i saw this half a year ago...

  74. degradation by Anonymous Coward · · Score: 0

    The degradation is pretty easy to handle: just add an elastic and cam such that the output rod tends to "snap" from 0 to 1.

  75. The term is "logically complete" by Anonymous Coward · · Score: 0

    There is more than one set of gates that is "logically complete". That means that you can use gates from that set to synthesise any others that may be needed. Simple examples are "and", "or", and "not" and "nand". The nand set is particularly elegant since you only need one type of gate, althogh of course you may need examples with many different numbers of inputs. You can use a karnaugh map to find an implementation that would use "and" and "or" (and "not") and then substitute nand gates for all of the and and or, and it will work just fine. Since nand was a simple and economical gate in TTL, this was a very practical approach way back then. (A nand gate in TTL was built around a single multiple emitter transistor.)

  76. actually by argStyopa · · Score: 1

    I think these are wonderful. Talk about a brilliant way to get 9 and 10 year olds interested in the basic concepts of computing. I have a sneaking suspicion that I'm going to have to be buying bulk brick packs soon...

    Nice job!

    --
    -Styopa
  77. Hum... by SWTP_OS9 · · Score: 1

    This reminds me of the analog computers thay were used for about 1/2 of the last century. They caculated tables etc. Mostly balistic problems.

    I like the idea of binary and Lego. Is very intersting.

  78. One word... by cr0sh · · Score: 1
    Friction.

    Basically, when you are implementing mechanical adding machines (and even logic) - ultimately you need to overcome friction. You wouldn't believe the lengths to which Babbage (and his engineer/draftsman Clement) went to overcome this issue. Litterally months and years were spent on this problem - and the designs which came out of the work were very beautiful (especially the spiral carry mechanism).

    Push-pull logic has this same issue, too - I believe Konrad Zuse had to come up with some interesting designs on his mechanical machines as well.

    Another issue with gear-based mechanisms is that of tooth lash - that is, the small amount of movement between the teeth as the gears mesh - this slight amount of movement can cause premature wear and failure, excess friction, and inaccuracy - this too, was something Babbage had to work very hard at overcomming in his systems, and came up with a very elegant locking mechanism for his machines.

    Interestingly enough, friction played a large part in the various differential analyzers created in the early part of the 20th century - knife edge wheels rotating on glass discs. Very tough to transmit the miniscule torque - what was done was a form of servo mechanism/mechanical assist with motors - very similar to that of power steering in an automobile. Even so, it wasn't enough, and these systems tended to get "out of tune" and produce wrong answers at a whim, or at least inaccurate results. However, they were a little more forgiving, being an analog system (and, interestingly, they drove a two-axis pen plotter which drew a graph to show the progress/answer to the simulation being run)...

    --
    Reason is the Path to God - Anon
  79. Need a dropping weight for that. by Otto · · Score: 1

    Pendulum wouldn't work without a weight of some sort attached to it to override the friction incurred from the motion of the machine. Same as an old grandfather clock, sort of thing. You gotta have that falling weight to keep the pendulum in motion or eventually the clock stops from friction losses.

    But yeah, a pendulum hooked up to a gear with a dropping weight, same as in one of those old clocks, could provide back and forth horizontal motion. It could also act as a clock input if needed, and eliminate some design problems. It'd cause other problems though, I'm sure.

    --
    - Give a man a fire and he's warm for a day, but set him on fire and he's warm for the rest of his life.
  80. Re:Speaking of mechanical computers... (Digicomp) by htf · · Score: 1

    There are enthusiasts at the "friendsofdigicomp" Yahoo! Group. They have a couple of emulators, pdfs of the docs, some info about the original designers (and the patents), and a guy who fabs replacement parts.

    The various Digicomps and related toys show up on eBay periodically.

    - htf