Please create an account to participate in the Slashdot moderation system


Forgot your password?
Hardware Hacking Intel Build

A Replica of the First 4004 Calculator 63

mcpublic writes "For the 37th anniversary of Intel's 4004, the world's first off-the-shelf, customer-programmable microprocessor, vintage computer enthusiast Bill Kotaska has successfully built a replica of Busicom's historic 141-PF printing calculator using vintage Intel chips. Decades before the ubiquitous 'Intel inside' sticker, Japanese calculator maker Busicom introduced the first product ever built around an Intel microprocessor. Bill's homebrew replica includes a rare Shinshu Seiki Model-102 drum printer and runs firmware extracted from the original Busicom ROMs. Schematics and photos of his re-creation are available at the unofficial 4004 web site, along with Tim McNerney's new PIC-based emulator of the Model-102 printer. The site includes the Busicom 'source code,' 4004 details, interactive simulators, and other goodies for students, engineers, and computer historians." We discussed the 36th 4004 anniversary project here last year.
This discussion has been archived. No new comments can be posted.

A Replica of the First 4004 Calculator

Comments Filter:
  • by gowen ( 141411 ) <> on Sunday November 16, 2008 @06:26AM (#25776249) Homepage Journal

    ... you would have just got a "4004 Not Found" error.

  • by Anonymous Coward on Sunday November 16, 2008 @06:45AM (#25776321)

    This criminal mind has misappropriated proprietary copyrighted code by the Japanese company Busicom. If he can't wait until 70 years from the death of the author, i.e. until year 2100 or so, jail is too good for him. I hope they throw him to a bunch of radioactive mutated lawyers.

    • Back in the late-60's and early-70's, when the Busicom 141-PF calculator software was written, United States copyright law was very different, you needed to explicitly mark a work with a copyright symbol, and register it with the U.S. Copyright Office. Nowadays everything is automatically protected by copyright law. Back then it was not. There was no copyright on the Busicom binaries, so this code is free-and-clear. The re-created "source code" was written without access to the original Busicom source c
    • by Tolkien ( 664315 )

      Oh, this was supposed to be funny? Posting to retract my -1 Flamebait.

      Apparently I need to be recalibrated.

  • You mean (Score:3, Funny)

    by Whiteox ( 919863 ) on Sunday November 16, 2008 @07:15AM (#25776431) Journal

    You mean I shouldn't have thrown mine out in the trash?

    • by drmpeg ( 1408305 )
      I still have my 1975 Intel data book.
      • by Whiteox ( 919863 )

        That's going back some time.

        You should see my collection! One day I'll photograph it and stick it up on a webpage.

        Bill Kotaska's site mentioned 'chip collectors'. I didn't think they rated. Learn something every day I suppose.
        I've got a few too. I've got a eeprom burner (16 pin) that I should fire up one day....

        I wonder if there is a special retirement home for people like us? Not only should they give me a room, but a workshop and storage facilities as well.

        • The big question is: do you have the '386 DX Microprocessor Programmer's Reference Manual'??

          I have two copies. Linus references it in one of his syllabuses of the books he used to create Linux. It's very much extremely rare in this day and age.

          Of course, I also have multiple versions of Isis [] on original diskettes, with original documentation (and original Intel hardware to run it on)

          Also an original printed CP/M-80 manual (those were rare even when they were current- very few people actually bought CP/M

          • by Whiteox ( 919863 )

            Your Isis 225 MDS sounds interesting. Probably as useful as my RS Mk II without any drives except a cassette port.
            I have a lot of docs/software stored under the house, as well as old equipment stored in other locations not at my house. They are hard to get to though. I really need to get rid of the wife and kids - they use up too much space and take up too much of my time.
            Some I remember having include a perspex boxed original Multiplan (Apple //), original IBM PC with about 4 manuals. CP/M - yes but I'm no

          • Does an 8.5x11 format MOSTEK 6502 manual count?
            How about a vintage 1981 iAPX86 manual on onionskin?

  • Looks complicated (Score:3, Informative)

    by troll8901 ( 1397145 ) <> on Sunday November 16, 2008 @07:17AM (#25776433) Journal

    Looks complicated. I would have a very difficult time coming up with such a polished work.

    • Re: (Score:2, Funny)

      by Smivs ( 1197859 )

      I don't remember these calculators, but having RTFA and STFP (seen the f*****g picture), I can't believe anyone could get away with selling something which apparently consists of a couple of techy-looking boxes joined together by wire!

    • Re:Looks complicated (Score:5, Interesting)

      by sam0737 ( 648914 ) < minus cat> on Sunday November 16, 2008 @07:59AM (#25776595)

      The architecture diagram is actually so simple...each rectangle there is representing at most ~30 transistors.

      Take a random rectangle of the current whatever chips architecture diagram, even for the simple one likes microcontroller, each rectangle is more complicated than the whole 4004 diagram there.

      The final project of 2*14 weeks (semester) IC design course could easily be as complex as the 4004.

      I have to admit it's like rocket science 35~40 years ago though. I actually admire that they could actually come up with that...imagine that they could actually be using pencil and ruler to draw the schematic and layout.

    • Re:Looks complicated (Score:5, Informative)

      by TheRaven64 ( 641858 ) on Sunday November 16, 2008 @08:39AM (#25776745) Journal
      The 4004 was around 2300 transistors, which was close to the limit of the fabrication technology at the time (if you read about how it was created you'd be amazed at how primitive it seems - you couldn't quite do it in your own home, but it's not far off). With a modern HDL designing something like the 4004 would be trivial, and even designing it a gate at a time is not hugely difficult.
      • If you're going to emulate it in programmable logic, why not just emulate it in software? If you're just going to emulate it in software.... well, you know what I mean. Use a modern HDL to do modern stuff. Unless you've got legacy code you want to run in it's original form. This guy is running the legacy code. On legacy hardware, even.

        • Who said anything about programmable logic? You can use an HDL when designing an ASIC and for something that small you could get it fabbed on a really old process very cheaply in quite a few places.
    • by slowbad ( 714725 )
      You would really need to be a Bomar Brain [] to do such work.
  • Linux (Score:4, Funny)

    by mcnazar ( 1231382 ) on Sunday November 16, 2008 @08:06AM (#25776621)

    yes.. yes.... but does it run Linux?

  • I'm afraid he's going to have a lot of trouble finding printer cartridges for that thing.
    • by WWWWolf ( 2428 )

      I'm afraid he's going to have a lot of trouble finding printer cartridges for that thing.

      Eh, I know you're kidding, but if it's anything like many other old printing calculators, it's probably just got an ink pad or ribbon or something other good ol' well-understood stone age technology.

      I've often wondered about our fancy new printer cartridges - how did these bloody things become so complex. But then again, I'm old enough to remember an age when thingies were actually user-serviceable. Back in the day, my father could fix everything, and now he's just as puzzled about some of these newfangled

      • how did these bloody things become so complex.
        because we moved away from impact printing to inkjet printing due to it's far better handing of images that had more than one bit per color channel.

        but inkjet printing requires ink to be delivered in a very pricise and bubble free manner. That means either integrating the printhead in the cartridge (HP style) or having relatively complex systems for ensuring bubbles don't block the printhead (epson style)

  • Link to article (Score:2, Informative)

    by howardd21 ( 1001567 )
    The link posted is to the main site 4004 site; the actual project article is here: []
  • Funny how TFA talks about wire wrap boards giving projects a "vintage" look. I saw, as recently as 2004, an Augat wire-wrap board being used as a part of a PhD student's research work. This isn't so bad in itself, except it had about 20 ECL logic chips, carrying 80 MHz signals. There were runt pulses and false triggering all over the place. I replaced it with a single Xilinx Coolrunner II.

    • by Intron ( 870560 )
      For ECL we preferred wirewrap. You can twist two wires together and have a differential controlled impedance line as short as possible. Parallel printed circuit lines were the ones that picked up crosstalk.
  • After looking at the photos, I recognized the printer as being exactly like one I got as surplus around 1977. I adapted it to work with my Dad's Commodore PET through the parallel port. It had a spinning drum covered in raised numbers and symbols, and solenoid hammers for each column. By firing the hammers at the right time as the drum spun, I could make it print any number. I wonder what ever happened to it...

  • He used a 2716 eprom in the re-creation. That was a part NOT available in the time of the 4004. He should have used a 1702 eprom. These parts are not THAT rare, though he would have needed 8 of them to replace a single 2716.

  • Aren't those chips well past their Use by date.

Can anyone remember when the times were not hard, and money not scarce?