6

In 1972, when I was 16, I was sent on a one-week computing course at the local technical college in Worcester (UK). I was one of a group of girls who were intending to take Maths A-level the following year.

One of the interesting things I remember was being shown how to use an 'analog computer' to solve quadratic equations. They were open circuit boards, 12-15" square, and you plugged in cables with banana plugs to set up the equation to solve. I don't remember much else about them, but I have often wondered whether they had a real world use, and if so, when and why they became obsolete.

Does anyone remember these?

[I have to say that that week put me off computers for ten years, which I spent learning librarianship, hating that, then becoming a 'fine wirewoman' for a lift control company. I did eventually get into computing via that route, and took a degree in Computer Systems Engineering in 1993.]

  • Never seen one of those before. The old warships used to use some form of mechanical computer for working out gun elevations based on bearing and distance. Don't know whether those were considered analog or not – cup Oct 1 '20 at 11:41
  • Ken Shirrifs last blog explains how an analog multiplication chip works. righto.com/2020/09/how-to-multiply-currents-inside.html – Patrick Schlüter Oct 1 '20 at 11:43
  • Not early '70s, but do early 80's count? – Jon Custer Oct 1 '20 at 12:50
  • The question "Does anyone remember these" is off topic but this post should not be closed because it contains a salvageable and useful question, which is the sentence starting with "I have often wondered whether ...". – OmarL Oct 1 '20 at 12:57
  • 2
    You can still see them and still buy them new - They're just very small niche market products: analogparadigm.com/products.html – tofro Oct 1 '20 at 17:36
5

I'm a few years younger than you (I was 13 in 1972), but I remember them. My college roommate had an old (and very crusty) Heathkit ES-400 analog computer in his family's barn back home. I don't think it would have been capable of being powered up without a lot of cleaning and restoration.

Heathkit ES-400

One summer during high school (1975, IIRC), I also spent a week at a "summer camp" at the US Naval Academy, where they showed us a much larger rack-mounted analog computer.1 As I recall, they were demonstrating its use in simulating vehicle suspensions — a kind of "hardware-in-the-loop" simulation, in which they had a real wheel rolling over real terrain. A sensor converted the wheel's vertical position into an analog signal, and the computer then simulated the vehicle response, allowing parameters such as sprung weight, spring and damper constants, etc. to be "dialed in".

Analog computers were also used in real-time industrial process control — the sort of thing for which you'd write a PID loop on a microcontroller today.


1 Actually, now that I think about it, this may have been a hybrid computer, with both analog and digital sections. Something like this (from here).

  • Thank you so much! The question has vexed me for nearly 50 years... – SM. Oct 1 '20 at 12:21
  • Just a small correction, ES-400 is the order number for the cabinet/front panel. The whole system never had a dedicated name, as Heathkit was just selling ... well,.. kits :) It was announced as 'Heath Inexpensive Analogue COMPUTER Kit' and referenced in it's manuals as 'Heath electronic analogue computer'. It did not have any digital parts. 'Just' up to 15 amplifiers - which is the important part as they are needed to set up operations. Still got one ... memories :) – Raffzahn Oct 1 '20 at 19:10
5

Yes. Not only do I remember analogue computers, but I have used one.

My Leeds University computer science course in the mid-1970s had an analogue computer programming module.

The computer was 2 or 3 racks, if I recall correctly. I think lethal voltages were involved, but that could be faulty memory.

Programming involved patching together your 'equations' from various computing elements: inductors, capacitors, and operational amplifiers. Results were viewable on an oscilloscope display, or could be sent to a graph plotter.

AFAIK, the equipment was more-or-less obsolete even then. The course module existed because the computer did. The 'why' they became obsolete seems straightforward: because stored-program digital computers became available and capable of solving the same sort of things by numerical methods. Programming was easier (didn't involve manual plugging), programs could be saved without having a room full of plugboards, programs could be easily communicated to others, several people could use the computer at the same time, etc.

I cannot so far find a photo online, unfortunately.

At about the same time, the UK electronics hobbyist magazine 'Practical Electronics' published plans for the Practical Electronics Analogue Computer in several monthly installments, but that was beyond my skill and budget at the time (and still is beyond my skill).

3

Original analogue computing was the only way to solve complex math. Already 100+ years ago major ports had huge analogue computers to calculate tide tables.

It wasn't until the late 1960s /early 70s that digital computers became a sensible tool in for complex calculations, and even then they were still way too slow for many questions/usages. Especially when it was about solving in real time, like for simulations etc.

New analogue computers were designed way into the 1980s.

And just recently a Startup was founded in Germany to build analogue/dgital hybrid processors using kind of an FPGA alike structure to allow programming of analogue configuration on the fly.

  • 1
    I think you might be conflating "analog" with "mechanical". They certainly did not have electronic analog computers in 1920. – Dave Tweed Oct 1 '20 at 14:15
  • 2
    @DaveTweed - 100 years may be optimistic for electronic analog computers, but 90 is likely pessimistic. Westinghouse started working on electronic analog computers at some point in the late 20s, and AIUI had working systems before 1930. – occipita Oct 1 '20 at 15:26
  • @DaveTweed Wait, analogue computing can not be made with gears? Interesting PoV. A bit off, but interesting. – Raffzahn Oct 1 '20 at 18:40
  • Well, it's clear that the OP is asking specifically about electronic computers, on circuit boards. – Dave Tweed Oct 1 '20 at 18:46
  • @DaveTweed I have a hard time to see it. She reports abut having used such and asking for analogue computers in general. Beside, even restricting the question to electric doesn't put history out of scope. – Raffzahn Oct 1 '20 at 18:53
1

I haven't seen one ever, but we did still study them in school, in the last year of my electronics bachelor degree, which was 1988-1989. This was in the course "Operational Amplifiers".

  • Operational amplifiers as sum and difference amplifiers
  • Operational amplifiers using a diode to build logarithmic and exponential amplifiers
  • Taking these circuits together, it becomes possible to implement analog multipliers and dividers (input -> log -> sum or difference -> exponential -> result)
  • Integration and differentiation are simple operations with operational amplifiers
  • Taken all together, it becomes possible to build circuits which implement various equations

And, yes, we made (theoretical) exercises designing such systems. I suppose it was a way to give more diverse exercises on operational amplifiers.

On second thought, I think I might have seen such a setup in one of the electronic labs at school, kept a bit as a curiosity, a think that could be wired up manually.

Not the answer you're looking for? Browse other questions tagged or ask your own question.