It’s easy to get caught up in the excitement of creation as we’re building our latest widget. By the same token, it’s sometimes difficult to fully appreciate just how old some of the circuits we use are. Even the simplest of projects might make use of elements that were once a mess on some physicist’s or engineer’s lab bench, with components screwed to literal breadboards and power supplied by banks of wet-cell batteries.
One such circuit turns 100 years old in June, which is surprising because it literally is the building block of every computer. It’s the flip-flop, and while its inventors likely couldn’t have imagined what they were starting, their innovation became the basic storage system for the ones and zeros of the digital age.
If you were given the task of designing a computer at a time when computers weren’t really even a thing, how would you start? How would you take a collection of vacuum tubes, passive components, and a precious few germanium diodes and engineer something to sell to customers looking for an “electronic brain”?
Where there’s a paycheck, there’s a way, and computer archeologist [Ken Shirriff] laid his hands on some old IBM hardware that tells us a lot about how engineers thought in the earliest days of the computer industry. The gear is a pluggable module from IBM, one of hundreds that once went into their Model 705 computer from the mid-1950s. The particular module [Ken] has is a 5-channel contact debouncer, or in Big Blue’s mid-century parlance, a “Contact-Operated Trigger.” It was used to debounce five of the many, many mechanical contacts in the machine, both buttons and relays, and used eight dual triode tubes to do it. Other modules with the exact same footprint formed the flip-flops, inverters, buffers and clocks needed to build a computer.
[Ken]’s analysis of the debouncer is a fascinating look at what was possible with the technology of the day, and the fact that it led to a standardized framework for generic modules that were actually hot-swappable with what essentially was a zero insertion force plug was quite a feat of engineering. And as a bonus, [Ken] and friends actually got the module up in running in the video after the break.
As useful as computers are, most of them have all the design charm of a rubber doorstop. Oh, for the heady early days of computing, when vacuum tubes ruled, hardware was assembled by hand, and engineers always wore a tie.
Looking to recreate an elegant bit of computing hardware from that more civilized age, [updatebjarni] built a reproduction of a 1948 IBM TR-2 flip-flop module — 1,250 of which once formed the memory of the IBM Model 604 Calculating Punch. Admittedly more of a high-speed adding machine than a computer, the 604 is still an important piece of computing history, and [updatebjarni]’s scrap-bin reproduction of the field-replaceable module served as part of a computer history exhibit.
With a single 6J6 double triode tube nestled inside a bent aluminum frame, the goal was to reproduce the appearance of the original TR-2 module, and so the passive components wired up point-to-point style below the tube socket were chosen for their vintage look. That’s not to say the flip-flop won’t function. Although [updatebjarni] hasn’t tested it, he’s built other functional flip-flops from vintage components before, so this one should work too. Only 1,249 left to build and he’ll have enough for a working 604.
If you like this kind of build, you should probably check out some of our Vintage Computer Festival coverage. VCF East in April was a huge success, and VCF West is coming up in August in Mountain View. Hackaday will be well represented there, so stop by.
Psst… Wanna make a canning jar diode? A tennis ball triode? How about a semiconductor transistor? Or do you just enjoy sitting back and following along an interesting narrative of something being made, while picking up a wealth of background, tips and sparking all sorts of ideas? In my case I wanted to make a cuprous oxide semiconductor diode and that lead me to H.P. Friedrichs’ wonderful book Instruments of Amplification. It includes such a huge collection of amplifier knowledge and is a delight to read thanks to a narrative style and frequent hands-on experiments.
My well worn copy of Instruments of Amplifications
DIY point-contact semiconductor transistor
Friedrichs first authored another very popular book, The Voice of the Crystal, about making crystal radios, and wanted to write a second one. For those not familiar with crystal radios, they’re fun to make radios that are powered solely by the incoming radio waves; there are no batteries. But that also means the volume is low.
Readers of that book suggested a good follow-up would be one about amplifier circuits, to amplify the crystal radio’s volume. However, there were already an abundance of such books. Friedrichs realized the best follow-up would be one on how to make the amplifying components from scratch, the “instruments of amplification”. It would be unique and in the made-from-scratch spirit of crystal radios. The book, Instruments of Amplification was born.
The book includes just the right amount of a history, giving background on what an amplifier is and how they first came in the electrical world. Telegraph operators wanted to send signals over greater and greater distances and the solution was to use the mix of electronics and mechanics found in the telegraph relay. This is the springboard for his first project and narrative: the microphonic relay.
The microphonic relay example shown on the right places a speaker facing a microphone; the speaker is the input with the microphone amplifying the output. He uses a carbon microphone salvaged from an old telephone headset, housing everything in an enclosure of copper pipe caps, steel bar stock, nuts and bolts mounted on an elegant looking wood base. All the projects are made with simple parts, with care, and they end up looking great.
If you have ever had a go at building a tube-based project you will probably be familiar with the amount of metalwork required to provide support structures for the tubes themselves and the various heavy transformers and large electrolytic capacitors. Electronic construction sixty years ago was as much about building the chassis of a project as it was about building the project itself, and it was thus not uncommon to see creative re-use of a chassis salvaged from another piece of equipment.
This morning we stumbled upon a rather nice solution to some of the metalwork woes facing the tube constructor courtesy of [Bruce], who built his tube audio amplifier on a chassis made from a cake tin and with its transformers housed in decorative display tins.
The circuit itself is a straightforward single-ended design using an ECL82 triode-pentode on each stereo channel, and comes courtesy of [Nitin William]. The power supply is on-board, and uses a pair of silicon diodes rather than another tube as the rectifier.
It’s true that [Bruce] has not entirely escaped metalwork, he’s still had to create the holes for his tubes and various mountings for other components. But a lot of the hard work in making a tube chassis is taken care of with the cake tin design, and the result looks rather professional.
We have something of a personal interest in single-ended tube amplifiers here at Hackaday, as more than one of us have one in our constructional past, present, or immediate futures. They are a great way to dip your toe in the water of tube amplifier design, being fairly simple and easy to make without breaking the bank. We’ve certainly featured our share of tube projects here over the years, for example our “Groove tube” round-up, or our look at some alternative audio amplifiers.
There is a rich history surrounding the improvisation of electronic components. From cats-whisker foxhole radio detectors using razor blades through radio amateurs trying antique quartz lenses as crystal resonators and 1950s experimenters making their own point-contact transistors, whenever desirable components have been unavailable the ingenuity of hackers and makers has always sought to provide.
In an age when any component you might wish for is only a web browser and a courier package away, you might think there would be no need for such experiments. But it is in our curious nature to push the boundaries of what can be made without a factory at our disposal, so there are still plenty of ingenious home-made components under construction.
One such experiment came our way recently. It’s a few years old, but it’s a good one. [Nyle Steiner, K7NS] made a working triode without any form of vacuum, instead its medium is a flame. He’s demonstrated it as a rectifier, amplifier, and oscillator, and while it might not be the best triode ever it’s certainly one of the simplest.
In a traditional vacuum triode the current flows as electrons released from a hot cathode and are able to cross the space because there are no gas molecules for them to collide with. The flame triode has an abundance of gas, but the gasses within it and its immediate surroundings are also strongly ionized, and thus electrically conductive. Flame ionization detectors have exploited this phenomenon in scientific instruments for a very long time.
A roaring flame might not be the most practical thing to keep in your electronic equipment, but [Nyle]’s experiment is nonetheless an impressive one. He’s posted a video showing it in action, which you can see below the break.
It all began with a cheap Chinese rotary vane vacuum pump and a desire to learn the witchcraft of DIY vacuum tubes. It ended with a string of successes – a working vacuum chamber, light bulbs, glow tubes, diodes, and eventually this homebrew power triode and the audio amplifier built around it.
[Simplifier]’s workshop seems like a pretty cool place. It must have a bit of an early 20th-century vibe, like the shop that [John Fleming] used for his early work on vacuum tubes. Glass work, metal work, electronics – looks like [Simplifier] has a little bit of everything going on. True to his handle, once [Simplifier] had a cheap but effective vacuum rig he started with the easiest projects – incandescent and gas discharge lamps. Satisfied that he could make solid electrical and physical connections and evacuate the tubes, he moved on to diodes and eventually triodes. The quality of the tubes is pretty impressive – stray gasses are removed with a bake-out oven and induction-heated titanium getters. And the performance is pretty solid, as the video below reveals.
Very impressive overall, and it’s not just the fact that he’s building tubes from scratch – we’ve seen that before. What shines here is that specialized equipment is not needed to make working and reliable tubes – just a MAPP torch, simple hand tools, and a low-end vacuum rig. Anybody could – and probably should – give this a try.
