If You Aren’t Making Your Own Relays…

We’ve all been there. Someone will say something like, “I remember when we had to put our programs on a floppy disk…” Then someone will interrupt: “Floppy disk? We would have killed for floppy disks. We used paper tape…” After a few rounds, someone is talking about punching cards with a hand stylus or something. Next time someone is telling you about their relay computer, maybe ask them if they are buying their relays already built. They will almost surely say yes, and then you can refer them to [DiodeGoneWild], who shows how he is making his own relays.

While we don’t seriously suggest you make your own relays, there are a lot of fun techniques to pick up, from the abuse of a power drill to the calculation of the coil parameters. Even if you don’t learn anything, we get the desire to make as much as you can.

In all fairness, using reed switches seems like skipping the hard part. We were hoping to see even the contacts being homebrewed. That could have more practical application since contact material is a vastly underappreciated characteristic of relays. A lot of the bad rap they get for high failure rates is due to the contact material being incorrectly specified. If you match the contact material to the application, relays are often quite reliable.

Did you or will you make your own relays? Tell us why in the comments. Meanwhile, if you want to build a relay computer, consider using the relays as multiplexers instead of switches. If you need inspiration, here’s one neat build.

45 thoughts on “If You Aren’t Making Your Own Relays…

    1. Or from the Lampoon’s parody “Deteriorata” (on youtube; a parody of Desirata, easily found on the net) “There will always be a great future in computer maintenance”
      To which I say I could still unjam and change the ribbon on an 029 keypunch.

      How about programming using plugboards? Never did that myself, however. That was how the old tabulating machines would be “programmed”.

      Nowadays programming software that runs on powerful Dell servers.

      1. I bow to your prowess in ribbon changing. Wiring boards weren’t hard, just tedious because you had a wire for each of 80 columns and if the wires were long enough you could reverse the columns

      2. There’s an old SF story about a man who is going to be executed because he can’t get a job. *Everyone* must have a job, or else. The thing is he had all the training to be an operator of the latest model computer but every company with that model computer turned down his application. The end of it it’s a clerical error. Someone entered the previous, obsolete, model of computer on his record. It’s discovered too late. The poor fellow has met the fate of the jobless and the people in charge of the department where the error was get quite nervous about their future.

  1. Sounds like the computer equivalent of the “Four Wealthy Yorkshiremen” sketch (on youtube) where they each try to outdo the others on how poor they were.
    Relays? I had to mine copper ore, smelt it, and make my own wire!
    Wire? You were lucky to have electricity! We had to use rocks for bits!

      1. So that explains why my dental hygienist calls the crusty stuff on my teeth “calculus”!
        I figured it was called that, because like the math discipline, it was “hard”.

  2. Someone mentioned the 029 keypunch ….

    Those things were entirely relay logic – with about a thousand reed switches in all.

    Electronic maintenance was:

    1. Identify the dead relay
    2. Identify the broken reed
    3. pull the relay
    4 remove the reed
    5 check the reed type (NO and NC)
    6 slot in a new reed
    7 push the relay back into its “socket”

    Not quite build your own relays, but they WERE in kit form.

  3. I fondly remember my first electronics book as a kid, which I inherited from my father, suggesting that to obtain a relay, one should get their hands on an electric bell and modify it. Those were very different times back then :-D

    1. My first electronics book was from an uncle, and it was on TV repair, not electronics as such. Tube TVs, The uncle worked for Ma Bell (“The Phone Company” and got several relatives jobs there, back when it was good solid work.
      And I would repair the family’s TVs.

  4. I made a relay as a lad using an old hacksaw blade (springy!) and insulated wire wrapped around a bolt. In this particular application it was wired as a buzzer for one of those pass-the-loop-along-a-wiggly-wire games.

    The big lesson, which still holds today, was that I found creating the game more interesting than playing it.

  5. Back in 1930 someone tried to make a mnemonic memory circuit using “stone knives and bearskins” . Since the NY 21st Street Mission building no longer exists this claim cannot be substantiated as if it never happened.

    1. I think somebody’s chakras need unblocking. Look around you. Industrial relays! You need to be surrounded by the calming aura of hand-made authentic anti-polar directional relays. Gold plated contacts and oxygen-free copper wire! Healing guaranteed.

    2. He now has a video up where he uses the relay to fix a very old nixie multimeter that was missing a relay.
      And it makes sense with such an old and beat-up and “modded’ device to not go out and find and buy a compatible relay, one that might not even have fixed the issue.

      And as he says, it’a about learning.

      Also: if you prefer buying stuff you are perhaps on the wrong site, try apple.com instead :)

    3. It lets you do things you would need a much larger number of common SPDT relays to do, because you can design it with a large number of poles and even multiple coils if you can fit them. That lets you express all sorts of complex logic in minimal form. You might even be able to get it to balance such that coil A cancels coil B’s field, which gets the function A XOR B just from your inputs. Then you can take however many copies of that value you need, entirely isolated from each other, by putting more than one reed in the coils. And the choices are/were normally open, normally closed, or change-over reeds. Might be awfully fiddly, admittedly, but still. There’s a potential for cool uses.

  6. Reminds me of the time Rinoa mentioned in one of her videos about how she would use to make relays out of old speakers and paper clips. The idea of a (then) little girl going around asking for old speakers to “build a computer out of” just sounds interesting.

  7. SSD? We had to use spinning rust.
    Spinning rust? We had to use dvd.
    Dvd? We had to use CD rom.
    CD rom?We had to use giant SD card?
    Giant SD card? We had to use Zip disk.
    Zip Disk? We had to use floppy disc.
    Floppy disk? We had to use cassette tape.
    Casette tape. Etc Etc. Magnetic bubbles, delay lines, rows of switches, magnetic core

  8. I like DiodeGoneWild very much this new video is also appreciated. I have not yet made a relay but I have once built a solenoid to pull a metal core into it. I actually did 3D print the housing the way that Diode made fun of. But it had to have a specific size that I could not find scrap material for so I think it was not overengineering. I made the coil using a very similar technique with a little electric drill. The wire was new instead of salvaged – less hacker points again. I am just not consistent enough with my hobby to gather proper material.
    I entertained my children with making the coil and it was fun. They even helped holding the pencil holding the wire so I could be less of a juggler. The end result was fine electrically but the movement of the metal core was not dependable enough due to friction. I have abandoned the project before optimizing the result though.

  9. That paperclip “computer” book shows how to make Binary Coded Decimal rotary switches from things like wooden thread spools.

    It’d be nice to see someone take that design and make it a proper *automatic* computer. The design in the book is essentially a fancy abacus that uses electricity to operate a row of light bulbs. Years ago I was going to follow the book and build it for fun – until I understood that it required manual manipulation of every step.

  10. I am an elevator mechanic and this is a little off topic but as for the OP comment about contact material and such. I don’t know much about reed switches in a vacuum tube but I have had the pleasure of working on 100 year old armor open faced relay logic controllers and metallurgy is a dead art. Whatever silver alloy used on oem contacts before they went out of business lasted decades with zero maintenance pushing hundreds of volts DC. Unfortunately due to lack of preventive maintenance, my boss opted to replace all the contacts with aftermarket silver plated copper contacts. In two months it was terminal and had to be replaced for a new “better” modern controller that has given me a horrific experience with reliability. So what I guess I’m saying is companies have long since realized reliability isn’t profitable and it has been so long since things were built to last that the technology in most cases has been lost.

    1. It is now upon us to save as much knowledge as possible, so the society doesn’t have to start from zero once we realize that newer is not always better. Possibly somewhere some original spare contacts have survived for analysis? Perhaps some old contacts in the dust at the bottom of the elevator?
      @all: If you happen to have the chance to take parts of old technology (or documentation thereof) out of the scrap pile, do it. Once lost it is lost forever. Empty space as such is worth nothing, one can always generate new empty space by throwing things out when it is needed.

  11. I know you had something to relay, but in my current frame of mind I have some resistance; perhaps something will excite my mortal coil and I’ll be able to field some questions. The core of the project seems solid though I’d feel better with an ironclad guarantee if the conductor of your project can deliver.

  12. I remember Hollerith card readers. We use to call them card eaters. One card out of place, and the main frame would crash. One card eaten, and the main frame crashes. Yeah, dems were the good ol days.

  13. Back in the 1970s, my (then not yet) wife re-wire-wrapped the entire backplane of a Data General minicomputer that came with defective wire-wraps. Remember wire-wraps? You made a circuit by wrapping actual wires (with insulation and everything!) around gold-plated posts. They connected up the ferrite core memory bits with each other and the CPU.

    And before that I had a colleague from Arizona State who built his own memory from a long closed loop of coax delay line with a regenerator at the end of the loop.

    I never fiddled with hardware myself, but I wrote a complete OS on our PDP-9 by toggling in a program to read the teletype and then typing in a program to read & write high speed paper tape… after that it was easy!

    1. At IBM, we spent at least one lifetime figuring out how to make wirewrapping work in 100% humidity 100%of the time. Only seemed to work with gold on gold. Ever try to order #26 in booku colors with 98% gold and 2% silver in 500’ spools?

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.