A Half Century Ago, Better Transistors and Switching Regulators Revolutionized the Design of Computer Power Supplies

Not strictly about (retro) computers, but still interesting. Article by Ken Shirriff on the history of the switching power supply and current power supply technology.

It also gives some context to the claim by Steve Jobs that the Apple ][ introduced the switching power supply. Following is from the article:

Power supplies are not without ardent champions, however, including one that might surprise you: Steve Jobs. According to his authorized biographer, Walter Isaacson, Jobs had strong feelings about the power supply of the pioneering Apple II personal computer and its designer, Rod Holt. Jobs’s claim, as reported by Isaacson, goes like this:

Instead of a conventional linear power supply, Holt built one like those used in oscilloscopes. It switched the power on and off not sixty times per second, but thousands of times; this allowed it to store the power for far less time, and thus throw off less heat. “That switching power supply was as revolutionary as the Apple II logic board was,” Jobs later said. “Rod doesn’t get a lot of credit for this in the history books, but he should. Every computer now uses switching power supplies, and they all rip off Rod Holt’s design.”

Jobs’s claim is a big one, and it didn’t sit right with me, so I did some investigating. I discovered that, although switching power supplies were revolutionary, the revolution took place between the late 1960s and the mid-1970s as switching power supplies took over from simple but inefficient linear power supplies. The Apple II, introduced in 1977, benefitted from this revolution but didn’t instigate it.

2 Likes

Useful things those switching PSUs. I just built a 2 channel vacuum tube mic preamp that I powered with a switching PSU. People in audio circles are often afraid of these, and often for no good reason.

1 Like

Notably, suitable power supply technology became available only in the late 1920s. (E.g., until then vacuum tube radios had to be powered by three types of batteries for each of the major voltages, which was somewhat messy to maintain. Therefor, solid-state detector radios were popular. Suitable power supplies really started the electronic age and made tubes a practically viable technology.)

This was still new technology when computers became a reality. (Wacky power supplies were really the reason for using binary, as opposed to ternary or decimal, since stable voltage levels were still a luxury.) And switching power supplies were still cutting edge in the 1970s…

1 Like

Ken references thyratron tubes in that blog post. You can see them in the PSU in this Curious Marc video on his teletype restoration.

2 Likes

A relative of the thyratron is the mercury arc rectifier, as seen powering trams/streetcars. Good visuals on these!
https://edisontechcenter.org/MercArcRectifiers.html (video within)
https://en.wikipedia.org/wiki/Mercury-arc_valve

Another sort of power supply I rather like the idea of - a bit steampunk - is the motor generator set, where an early computer would have a stabilised power supply at a convenient frequency using a motor-flywheel-generator combination.

As a couple of examples, they get a mention in the contexts of Remington Rand’s UNIVAC II and IBM’s 709 machines:



(Photos from Ed Thelen’s site, linked above.)

Oh, and the Cray 1 too:
http://gordonbell.azurewebsites.net/craytalk/sld062.htm
“The Cray 1 had two of these motor-generators to generate the 3 phase 400 cycle power that was fed to the power supplies located around the base of the computer. Seymour claimed it was the world’s most expensive “love bench”.”

2 Likes

I was thinking of this one, too! However, I was too lazy to look it up… :wink:

This is still the preferred technology for coupling power networks.
Remember, when the network frequency had been dropping in some parts of Europe last year? This was, because the networks were just barely able to keep up after som hick-ups with cheap power in the post-Yugoslavian region. Motors in the coupling stations weren’t able to keep the flywheels at speed and so frequency dropped below the standard 50Hz (which apparently equals to revolutions per second of those flywheels).

P.S.: Related technology: precision motors driving plate capacitors to generate a wave form, e.g., for oscilloscope sweeps.

See also synchronous or DC governed motors in Teletype machines, providing the timebase for asynchronous serial encoding and decoding.

The concept of the switching power supply goes way back. A crude form was used in early (tube) car radios – the mechanical vibrator – to raise the 6 or 12 volts to the hundred+ volts needed for the tubes.
But in the age of vacuum tubes there was little incentive to use switching power supplies. They really come into their own in low-voltage applications(1). Plus, tube circuits used hundreds to thousands of watts, so losses in the power supply were of little concern. Vacuum tubes don’t make good switchers anyway.
When transistors came along, most circuits still used at least 12 volts and still used a fair amount of power, only mildly regulated, if at all.
Digital ICs that often ran on 5 volts at dozens of milliamps each and needed very well regulated supplies began to make switchers relevant. At around the same time, the transistors, diodes, capacitors, and magnetic core materials needed started becoming available. But switching regulators were still expensive and difficult to design relative to a big honking transformer and linear regulator. Around 1970 or so they started becoming common on larger, fairly expensive equipment where the cost became reasonable. For “cheap” personal computers the size of a small filing cabinet (think Altair and other bus type systems) they still didn’t make much sense. And the power supply likely would have cost as much and been harder to design than the computer.
I’m not really a fan of Jobs, but I have to agree with most of what he said. As far as I know he never claimed Apple invented the switcher as some people have stated. But he did claim that Apple innovated personal computers by using a switcher. And that they did. How many personal computers used switchers before the Apple ][? None that I know of. Even after the 2, most home computers and many “business” computers still used linear supplies. Not until the IBM PC came out did switchers become common in personal computers.

1: A linear supply regulates the voltage essentially by using a variable resistance to waste extra voltage as heat. The AC input has to be dropped from 120/240 to something “close” to but above the needed DC. The rectified AC from the transformer then has to be filtered, keeping the dips above the minimum needed to regulate. That minimum is typically 2 to 3 volts. The difference between the dips and peaks is called “ripple.” The 2 to 3 volts plus ripple gets wasted. A one amp supply will waste at least 2 watts, no matter what voltage it outputs. For a 1 amp, 12 volt supply that provides 12 watts that isn’t too bad. For a 5 volt 1 amp supply it’s at least 40% waste. Typically it will be more like 60%.

EDIT: Here is a description of a vibrator power supply.

That’s an amusingly noisy thing to need in association with a radio! It’s also unearthed a memory for me: somewhere in my world of electronics kits, the Book of Experiments, and an inherited 1956 Gadget Annual, and other miscellaneous reading, I came across some sort of buzzing design which produced a high voltage by using the relay’s own coil as the inductor whose current is interrupted. I was a bit fascinated at the time with electric shocks, and wanted to make a shocker - fortunately, I never did manage to, with that approach, so far as I recall.

It strikes me that the good old fashioned HT coil as found in the petrol engine might be related.

2 Likes

I did build a shocker once (because I have a younger brother) - wish I could remember the place I saw it - although my memories are that it might have been a Ladybird book… Anyway, inside an old Swan Vesta match box, a nail cut to length, 2 thumb tacks, a lot of enameled copper wire and a battery (probably a paperclip or 2 too). So push the drawer in and made contact battery through the coil. Then to open it, you grasped the side of the drawer which I’d fitted the thumb tacks into - these were connected over the coil. you pull the drawer out and as you do so, the broke the circuit… Which resulted in a bit of a jolt through your finger and thumb holding the drawer…

I’ve had a google and can’ find anything, but it might be lurking there - and there are some Ladybird books with electrics and motors, so who knows.

-Gordon

2 Likes

Typically a magneto. Works by spinning a magnet, so doesn’t need field coils which require a current source to make them work, so used in things without a battery - e.g. lawnmowers, chainsaws, boat outboard motors, and a lot of light airplane engines where you want to be able to survive failure of the low voltage supply…

Also old telephones - lift the handset, crank the lever - to generate a signal to the operator…

-Gordon

1 Like

And lets not forget the legacy that these old switching PSUs leave us with… X2 capacitors that needs changing!!!

-Gordon

I vaguely recall Nolan Bushnell writing something about the switching supply of the Apple II in “Finding The Next Steve Jobs: How to Find, Keep, and Nurture Talent” . Unfortunately I don’t have a digital edition so searching for that at the moment would be time-consuming. :disappointed:

You can read some of that book, and search all of it, here:
https://books.google.co.uk/books?id=SJ3pk_phancC
(But I don’t see anything about power supplies - maybe a different book?)

1 Like