Fabricating Hardware From The Original Arcade Pong Schematics


This heavily populated PCB is a recreation of the original arcade version of Pong. That is an important distinction because the home version of Pong used a specialized chip to do much of the work. This is basically all stock logic, which explains the high component count. We wonder how many quarters it took just to pay for all 66 chips at the time?

[Pong74ls] was the person who took on this project. There is an original schematic available, but it’s incredibly crowded and rather difficult to figure out. Fortunately [Dan Boris] has already done a lot of the heavy work. He took the one-page nightmare and turned it into a sixteen page plan for building the original board (look for the schematic link under technical details).

Before the board could be laid out some redesign work was necessary. It sounds like some of the original chips are out of production and suitable replacements needed to be found. The board was then laid out in Eagle before sending the design off to a fab house. There was just one error which didn’t allow the ball to bounce when hitting a paddle while travelling downward. A couple of jumper wires fixed that right up!

[via Reddit]

[Original Reddit Post]

33 thoughts on “Fabricating Hardware From The Original Arcade Pong Schematics

  1. Nice work, but not completed yet or we weren’t shown it. Needs to be placed in free standing cabinet at comfortable bar stool height with a strong glass tap impervious to spilt beer,and booze

    1. The programs like that that I used could only order from the fab house that released the software. I might be thinking easy pcb, but I think perhaps they are the same. Eagle on the other hand you can export as you wish.

      1. One horizontal paddle traversing the bottom of the screen makes that more “Breakout” than “Pong”.
        The site seems like a good resource; I’ve now been nudged closer to taking the FPGA plunge

      1. It’s sort of an analog and digital hybrid computer that implements the physics of the game, and it does use voltage on capacitors as memory. The article explains it in detail:

        “The vertical velocity of the ball is represented by the voltage stored on the 1uF tantalum capacitor. This velocity is integrated by gate12c, and the resultant voltage, which represents the vertical position of the ball, is added to the frame ramp at the input of gate 12b, to produce a negative going edge representing the vertical position of the ball.”

        “The voltage on the 220uF electrolytic capacitor represents the horizontal position of the ball. This voltage is added to the line ramp at the input of gate 11b, producing a negative going transition at the output corresponding to the horizontal position of the ball. This is squared up by gate 11a, differentiated, and then squared up again by gate 11d. The width of the ball is determined by the differentiator time constant.”

  2. they should have just used an arm cortex A9 or a rhaspi back than -3-
    kidding ofc *21 and grew up on PS3*
    its a nice build and im a huge fan of logic circuits especially when they use logic to create an analog signal!

  3. That gave me cold shivers!

    I used to work on those, my first “real” job out of high school.
    Spent my days re-seating chips in sockets!

    If you want real horror, the original “Tank Battle” game had a board twice the size of “Pong”.
    Then everything shrunk and you could get “Pong” and “Tank War” on a AY-xxxxx 40 pin DIP.

  4. All I could think when looking at the boards is WHERE ARE THE IC SOCKETS! I help debug and fix things for people all the time and its so much easier when there are sockets and you simply pull out the dead chip and replace it.

    Awesome build tho. I always say I’m going to do something like that then I just read about people who go all they way and do it! Great job mate.

    1. What is the major cause you have found for integrated circuits failing? The only things I know of that kill them are poor voltage regulation, and trying to draw too much current through one due to improper circuit design. Beyond that I’ve never known one to ever fail. I’ve never seen a spring loaded socket be reliable though. They always fail. Machine turret sockets are more expensive than standard logic chips are too. But if for some strange reason I feel I need to use a socket in a design it is all I’ll use.

      If you really need to remove an IC you just cut off all it’s legs and clear the holes one at a time. There are no bonus points awarded for pulling out a damaged part intact. It is game over if you ruin the board though. Lots of amateur re-workers never understand this. I was taught by professionals though.

      1. I don’t think chip reliability from this era was as good as what it is now.

        Plus when you have so many, it increases the statistical probability that one will fail. I recall seeing a proposed supercomputer featured here built from simple logic chips. IIRC, it had something like 28,000 of them, with a MTBF of 22 hours!

        1. I have a lot of ICs from the 1970s I’ve never noticed them being any less reliable than today’s chips. Less capable sometimes, the old ones used a lot more current to operate, some I suppose have noise issues too. But if it works in a circuit, they generally keep working.

          You’re sure that 28,000 IC computer’s failures were all IC failures and not from other causes? I’d sooner believe solder joint thermal issues, or connector failures honestly. The high part count alone is what made it unreliable. I’m sure the folks that assembled it tried to do the best job that they could but I’ve seen what those old computers looked like and they were designed in situ.

          Back then you’d get a “new” computer from a manufacturer and you’d have to spend months rearranging things so it’d work! No one ever said anything about it because that was just how things were. If you were a geek you were just happy you had something to play with.

          So are you sure all of those failures were IC failures? Even if they were it could have been due to faulty design. But the odds of all of those failures being purely because chips were failing is vanishingly small. I’m sorry but I’d sooner believe it was roaches dancing in a conga line.

          1. If that was his first revision of the board and the first time populating it there is a good chance he wired something wrong and will burn up a chip. Things like that happen when your first designing something. Especially for a hobby project.

  5. wow
    just epic

    but i think even more epic is the repair…
    “just added a jumper” lol he probably means
    “spent days trying to find where in schematic the wrong signal would was”

    1. the original Pong board had Kynar wire trace “patches” all over the foil side.

      Like most games of that era.

      One interesting point, the Pong board used every single gate, except for 1 or 2 NAND gates of a 7400, just like the original Apple II board, which Woz developed to play Pong on.

  6. I heard a story about one of the early video games installed in a bar and the bar’s owner called the company that put the game in to report a problem. When the tech came to fix the problem he found the machine’s coin holder was so full no more coins could be inserted into the game anymore. It might have been Pong.

      1. so THATs why all the old hats ALWAYS check the darn coin box FIRST when theres a malfunction!!!!!

        i always wondered why. at the time it seemed like the least likely thing to go, i mean nowadays you’d be lucky to find 10$ or 15$ worth of coins

  7. Putting the part numbers and reference designators *under* the ICs on the silk-screen layer is a bad idea. Once the board has been populated with chips, none of these numbers are visable.

    While the old schematic–as scanned–is a bit unreadable, there is still much merit in having the entire plan laid out before you on a single page. Otherwise you are endlessly flipping back and forth between 16 pages of separate schematics–each offering insufficient context to grasp the entire design.

    But still, this is a fun project, and the original schematic, some 40 years old, is still educational, offering insights in logic design and compromise.

    1. Yeah, that is maddening when you have multi-page schematics. Even worse when you have nets that cross more than two pages and ZERO indicator as to what page it’s on.

      “OK, So C1A34 is on page 6 and 12. That’s good, but why does… Oh…. It’s on page 9 too? WTF? Then where does net B46AQ go? What the Hell? I’m missing a page? Oh, it’s back here on page 3… No wait, that’s B46AO…. Damn hand-written text.”

      It once took me nearly a week puzzling over where one net went off to before I discovered that A) it was in an entirely different book, B) it was mislabeled, and C) that that daughter card wasn’t even being used anyways. >:(

Leave a Reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.