Formative Events

This is the latest in a series of posts about the key events that shaped my life. Search for “formative events” to see them all.

When I first started getting into electronics, the general pattern of my activities was to build a design I found in a book. At the time, hobby electronics stores carried lots of inexpensive paperback books full of designs for various things, mainly based on vacuum tubes (remember, this was the 1960s). In these books you could find designs for power supplies, audio amplifiers, radio receivers and transmitters, and more exotic things like portable radios (very challenging with power-hungry tubes), light-sensitive switches, etc.

From old dead TVs (scrounged from the then-abundant TV repair shops) I’d scavenge the parts needed to build the designs from these books. Not just the obvious electronic components, but even the chassis — I’d strip an old TV chassis down until it was just the steel or copper (if it was a good TV) chassis with tube sockets and lug strips. Ah, lug strips — many of you, even if you’re into electronics, won’t know about these. Back in the 1960s, TVs weren’t made with printed circuit boards — instead, all the components were wired directly together, point-to-point, using either their leads or short pieces of wire. And the “lug strips” were the standard device to do this. They were just short strips of insulating material with a series of “lugs", which were small hoops, like croquet wickets, which you could wrap the ends of leads or wires around, and solder them securely.

Anyway, I’d start with such a stripped-down chassis, and then build my project directly onto it. Looked like hell, but if I did it right it would actually work.

One such project is the subject of this post. It was (or should have been) a very simple project — I wanted to make a regulated, variable 200 to 300 volt B supply (the high voltage supply used for the plate supply in a tube-based design). This would be the basis for other projects I wanted to build, and it would let me build just one regulated supply instead of several.

An aside for those who are electronically inclined: back then, a regulated supply was much more difficult to build than it is today. There was no such thing as the convenient integrated circuits of today, or even zener reference diodes — we had to build the whole darned thing from scratch, and then calibrate it ourselves. When I build an electronic project today, I still marvel at (by today’s standards) the simple little voltage regulators we can get so easily, or the very sophisticated switching regulators — either would have been sheer magic in the 1960s!

The design I’d chosen, out of one of those books of designs, used a couple of the most common tubes I’d find in TVs or radios: the 5U4 dual high-voltage rectifier, and the 12AU7 dual triode amplifier. It was a simple, straightforward design. I found the parts, I built it, and turned on the power. A bright blue arc formed across the top of the 5U4 (which was one of those big, chunky tubes not much smaller than an ordinary light bulb), and there was a sizzling sound — and a bit of smoke. The arc worked its way down the length of the tube, and there was a lould “pop", and then the fuse blew. Meanwhile, my voltmeter (which was monitoring the output) had pegged at something higher than 300V. In other words, the whole thing was a spectacular failure.

Since the 5U4 had disintegrated so spectacularly, I suspected a bad tube was the culprit. I had plenty of 5U4s, so I just popped another one in, replaced the fuse, and tried again. Same exact results. Next suspicion was that despite my careful check and re-check, I’d wired something wrong. So I checked again, against the circuit diagram in the book. Nope, everything was right.

Now what?

For a while, I was stumped. This had never happened to me before. Sure, I’d had projects that had failed to work — but every time the problem was either a mistake in my construction or a bad component (a common problem with scrounged components). This time, however, everything looked like it was correct. And it was such a simple circuit, compared to some other projects I’d built!

So I started troubleshooting in the only way available to me. I couldn’t turn the thing on and start measuring things, because it would self-destruct within seconds. So I started analyzing the circuit, trying to get a clue why the 5U4 would be arcing over. I worked backwards from the symptom (the arcing). Arcing over must mean the voltage was too high. Where does the voltage come from? The transformer. How could the transformer put out too much voltage? Well, this I could test — I unhooked the transformer from the circuit, and tested it in isolation. Sure enough, it was putting out 700 volts instead of the 350 volts called for by the circuit diagram. How could the voltage be exactly double? Ah…if the transformer’s dual secondaries were wired in series instead of parallel, that would do it — but the circuit design clearly showed a series wiring. Could the designer have made a mistake?

A mistake. Yes, the designer made a mistake. I rewired it in parallel, and flipped the power on — my meter showed 250V, exactly what it should have. No arcing, no sizzling, just the warm glow of the tube filaments and the appropriate hum of the transformer.

This silly little incident, though, was the first time it ever dawned on me that I might be able to design a circuit myself, instead of using these books. All the reading I’d done about electronics was focused on understanding how things worked, not designing them. I had a default assumption that only people with years of schooling could possibly understand how to design an electronic circuit. But in the course of troubleshooting that darned power supply, it dawned on me that I was figuring out how to make it work — how to design it. And after that, I built a series of projects that were of my own design — the first real engineering I ever did. The first design I still remember: a simple unregulated high-voltage supply. I had no purpose for it; I just wanted to see if I could do it. It was laughably simple: a full-wave rectifier with an RC filter on it. And even that simple thing didn’t work right the first time — I’d missed a few decimal points when computing the size of the capacitor, and my power supply was horribly noisy (the dreaded “hum” of the vacuum tube days). But it was my design, and eventually I figured out what was wrong and fixed it. I’ve been designing things, with slightly improved success, ever since.

It all started with that mistake.