If you could get your hands on an old analog computer...

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

Consul

Well-known member
Joined
Jun 3, 2004
Messages
1,653
Location
Port Huron, Michigan, USA
... what would you do with it?

The physics department at school has two old analog computers that have not seen any use for at least 25 years. It's probably been a good bit longer, seeing as they're both vacuum tube devices. The math and science head has offered to let me borrow one of them on an extended loan so I can play around with it for a while.

I was just wondering if anyone knows a thing or two about them, and what I might be able to do, sound-wise. I'll probably be in a position to do some sampling, so I'm hoping to get some interesting sounds out of it for that purpose. He's told me they can handle modeling oscillating systems.

As an aside, he also has an ancient HP all-in-one type of digital computer with a monitor, printer, and not one, but two high-speed tape decks. ;) It runs CP/M.
 
I guess I should point out that this particular analog computer was designed to allow for modeling of physics labs and experiments for learning. As I already mentioned, it should be capable of modeling oscillating systems.
 
It's probably full of nice, audio-worthy, long-life dual-triodes like E188CC's and such..

http://www.gyraf.dk/tmp/Turing_computer.jpg
or
http://img.photobucket.com/albums/v172/gyraf/Suze_1955.jpg

Unless, off course - it's a mechanical-analogue computer like these:

http://www.gyraf.dk/tmp/Babbage_computer.jpg
or
http://img.photobucket.com/albums/v172/gyraf/Suze_1936.jpg

..in which case you have a nice supply of mechanical parts.. :)

Jakob E.
 
Consul said:
... what would you do with it?

The physics department at school has two old analog computers that have not seen any use for at least 25 years. It's probably been a good bit longer, seeing as they're both vacuum tube devices. The math and science head has offered to let me borrow one of them on an extended loan so I can play around with it for a while.

What modules does it have? General purpose ones I've used usually had several modules (integrators, summations, nonlinear operators) and patch panel. They looked
something like this:http://archive.computerhistory.org/resources/text/Systron_Donner/SystronDonner.3300.1968.102646232.pdf.

That should fit right along modular synth.

cheerz
urosh
 
Since you are in Michigan, its probably a Heathkit EC-1?

vs-heathkit-ec-1-analog-computer.jpg


17 tubes implementing 9 operational amplifiers. 5 of the 17 tubes were used in the power supply stabilization circuit. There are some mechanical relays that transfer the inputs so that you can balance the amplifiers before activating the problem solution.

 
There was a larger version of that Heathkit, with a sloping panel, that Chris Shelton* and I dragged from Astronomy to the UCLA Electronic Music Studio back in the late '60's.  Chris was about the only person interested in getting it running, but I think he did manage to get a few setups to oscillate in various integrodifferential ways, and hooked it up to the big Moog as well, with due precautions (the analog computation standards in the tube era were +/- 100V signals, usually).

And yes, it did heat the room quite rapidly, which got warm enough most of the time as it was.  When I quit and Jim Cooper replaced, he figured out who to call in Facilities to get them to flip a particular switch and turn on the A/C, a trick that had eluded the professors and students (and me).


*among many other things a pioneer investigator in adaptive optics for astronomy.
 
That was probably a Heathkit H1 with 15 amplifiers. They also supplied a grill that would attach to the top and allow you to cook lunch.  :D

heathkit-analogrechner.jpg


This unit, originally introduced in 1956, was actually used by our space program.
 
Yes, that's the baby.  Quite a beast, definitely from the halcyon days of Benton Harbor---days not to be seen again I'd wager.* 

I think I still may have one of the short banana-plug jumper cables I abstracted, with very deteriorated rubber insulation.

I think the machine disturbed Cooper, and IIRC he had it removed.



*I knew Heath was going down when they insisted that a free "ad" extolling the company and its products be immediately removed from Stewart Brand's first Whole Earth Catalog (no dirty hippie longhair freaks tolerated in Michigan buster!!).
 
Well, I only get the smaller one, but hey, they've been sitting around for 25-30 years and maybe I can put it to good use. The other point is that there are two of them, and I might be able to finagle borrowing the second one later on for some more advanced stuff. Like I said, I just want to make some sounds and try to sample them. It's called suffering for my art. ;D
 
bcarso said:
*I knew Heath was going down when they insisted that a free "ad" extolling the company and its products be immediately removed from Stewart Brand's first Whole Earth Catalog (no dirty hippie longhair freaks tolerated in Michigan buster!!).

And yet, Paia manages to live on. Those damned dirty hippies are the ones who became the next generation of tinkerers and makers. :)
 
Consul said:
And yet, Paia manages to live on. Those damned dirty hippies are the ones who became the next generation of tinkerers and makers. :)

Paia may live on but not John Simonton RIP...

JR

PS: Heathkit was buried by low cost assembly automation, not marketing... The rules changed for that game...
 
> I have a slide rule...

Next to the PC monitor?

> what would you do with it?

It is 9 entire inverting op-amps, WOW! Any low-speed circuit which can be built with nine '741s or 2.25 LM324s, with all nine "+in" pins grounded, can be built with this box. Change all the resistors 10 times higher: all 10K become 100K.

The inverting-only amps foil a lot of modern thinking. But hey, common-mode errors are zero!

One amp with 2K input and 1Meg NFB resistor is a 54dB mike amp. Heck, do eight. Then amp #9 can have eight 100K input resistors and one 1Meg NFB resistor for an 8-in 20dB active summing amp. It may be only 1KHz response and 20dB noise figure, but the concept is right. Put 9V into the summer, see 90V out, then put your tongue on the summing node and prove that it is "zero voltage summing". (But if the system clips, your tongue will retract like a roll-up shade.)

It looks cool but it has less functionality than $0.69 of chippery and two 9V batteries.

Oh, the meter; and are the top-knobs semi-precision coefficient pots? The pots are useful when you do the cannon problem: given a target range and elevation, what angle do you set your gun at, how big a shell, how many bags of powder? One integrator sims gravity (which would be adjustable if you were doing multi-planet war), one is gun angle, one is shell mass, one is bags of powder. With a 9-amp computer you could maybe sum a tailwind also.

The manual is probably a lot more mind-growing than the console.
 
Well, the idea isn't for it to be good. ;) If nothing else, I'll try setting up a few interesting physics problems on it and see how well it does. If I can make some interesting sounds, I'll sample them. I'm not expecting a live performance instrument.

And as it should turn out, he does have the manual.
 
PRR said:
> I have a slide rule...

Next to the PC monitor?

> what would you do with it?

Nah, it's sitting in a junk box. It's a cheap plastic one, not a good one. But it worked adequately at the time it was bought by my employer (early '70s).

It was actually pretty useful back in the day for circuit design. The significant digit precision was adequate to select 5% capacitors and resistors for filter circuits et al. It was good discipline to keep track of the decimal points in your head..

I wrote a (digital) computer program run on my H-11 (using tab to print frequency response graphs) to help design cascaded multi-pole filters in the late '70s and never looked back.

JR
 
> a cheap plastic one, not a good one

I've used some "very fine" sliderules. But I use a vintage cheap pearwood which may have cost $2. My spare (yes I keep a spare) is the universal $0.99 Sterling Plastic Ten. (I may have a Stirling Six somewhere.)

For what they are good for, I don't lust for a "better" rule. I could have my Dad's; he was in the racket and owns a very fine wood rule with lush leather holster, and a mildly collectable Aluminum rule (one of the first that didn't shed its paint). Wood rules did tend to stick in summer in Missouri and New Jersey (and Mississippi?), the Al rule was supposed to cure that, and did.

They are not seriously more accurate than the Sterling. The Stirling has scales I never use, and the aluminum monster has even more scales I'd never use. When a design-team circulated around each others' drawingboards, the fast-draw hip-holster made sense; today if I had a team we'd throw "drawings" through the wire so the rule may as well sit at the PC.

I think the idea of a "good" sliderule comes from sliderule salesmen, encouraged by the few engineers who had way too much income, no non-work friendships, and no way to show-off except cufflinks and sliderules.

I admit I don't have any sliderule HERE at home; at work it lays near the main PC.

> actually pretty useful back in the day .... adequate to select 5% capacitors and resistors...

One problem can still be done faster on slip-stick, unless you happen to have a for-purpose program (yeah, there's a Java applet somewhere).

You need a voltage divider, say 3.45:1 voltage ratio, 2.45 resistor ratio. Total resistance must be a certain magnitude but final impedance is free. What Standard Resistor Values do you grab?

Set "2.45" above "1", scan the scales. (Of course you know the 1, 1.5, 2.2, 2,7, 3.3.... series.)

It took me over a minute to find 56:22 on the calculator, another half minute to find that 68:27 is trivially better. It is a 13-second breeze on the slipstick. 62:15 would be better but 62K was not a common value when slipsticks ruled the world.

Working on a breadboard power amp, the slipstick was doing "Vrms squared /8" or "Vpk/1.4, squared, /8" much faster than the finger-smash torture machine. A ten-inch stick can usually be read to 2% precision, which is much more than is warranted when the output is a 7% THD roundy-wave.

Another thing: you don't reset a sliderule. For some reason my TI-30Xa is going stupid between uses and needs three "C" pokes to get sane for use. (The battery is barely over 10 years old....)
 
Back
Top