Re: [midatlanticretro] Analog vs. Digital
- On Mon, Nov 8, 2010 at 8:34 PM, Bill Degnan <billdeg@...> wrote:
I have just mailed you the final prototype Apple IV analog computer. It is
based entirely on the ENIAC and the family dishwasher per your suggestion.
You were right that it would work. Please accept this as my gift of
thanks to you.
-------- Original Message --------
> From: "Evan Koblentz" <evan@...>>
> PS - Degnan has a letter from Steve Jobs saying the Apple IV was going
> to be analog. ;)
I kinda think the The Woz would be more apt to pull off a stunt like this, and even make you get fooled ;)
- Analog computers were used for a lot of flight control systems as well. A friend spent his first years out of college working on stability control systems for helicopters... it was formed from analog computers. Back in the late 70s, analog was king for a lot of applications, and those newfangled microprocessors were just starting to be looked at, but people doubted their ability to do what the analog machines did. Then again, my friend spoke of tests where they'd be hovering around 100' in the air, the pilot would key the radio and the 'copter wold suddenly drop 20'. He said they had scopes and voltmeters packed into the cargo area and had to try debugging with the sound of the engines right next to them. Lots of yelling.
> > What did the ENIAC machine code listing look like?Probably a lot like programming IBM tabulating equipment: either detailed switch and plugboard diagrams, or some shorthand/symbolic notation that an experienced operator would understand (such as saying to connect things to an adder/accumulator without specifying what one, leaving it up to the operator to use the one that's still working and uses the shortest wires).
Somewhere I have lab assignments where I used an analog computer like the one MARCH has. I don't think I specified every wire, just a diagram that rather directly translated to the spaghetti-o-banana cords. Just as printed material is not quite the same as handwritten (script or block printing), the symbols, notation and shorthand may always need some translation or understanding on part of the person performing the programming.
As to digital vs. analog: one of my cherished possessions is a little plastic pouch from GE Computers: an abacus and slide rule to demonstrate analog vs. digital (since GE made both kinds of computers in the 50s).
Digital means discrete values as in digits. The abacus (or other counting beads) naturally work in base 10 but also support base 16 (as required for certain currencies long before computers). Babbage's machines, adding machines, tabulating machines, even ENIAC were decimal (base-10), whether in gears or electronic circuits. My understanding is that ENIAC used "ring counters" instead of binary since they were well understood, and perhaps helped detect errors since only one output is valid at any time (whereas binary requires parity or other error checking). It would be interesting to find any records of ENIAC "common failure modes and how they were detected". We're so engrained with binary computers that it's hard to remember there are other ways to accomplish it. That's where retro groups such as the Neonixie folks come in with re-creating old technology such as neon trigger tubes for implementing ring counters the real "old fashioned way" without binary anything.
Analog means continuous waveforms, usually visualized on plotters or oscilloscopes. Analog computers use integrators, differentiators, amplifiers, op-amps, attenuators and such to implement equations. The stuff of early synthesizers! A friend had a lamp with a strange curvy-wavy base. It was actually part of an analogue computer for cannon firing: the surface described an equation. I guess rotating and moving it in/out chose the fuze-delay.
Analog is also like slide rules and needle-meters: the accuracy is in the skill to read it to the next decimal place.
- I've got a lot of boards without the handles. What's a current part number that works, preferably from a place like Digikey? Surplus places are just too hit-or-miss.
- On Thu, Nov 11, 2010 at 10:03 PM, Bob Applegate <bob@...> wrote:
I've got a lot of boards without the handles. What's a current part number that works, preferably from a place like Digikey? Surplus places are just too hit-or-miss.
You know what, this is something I had way in the back cobwebs of my mind all along and never looked into this. I could use a lot more too. So I took some measurements of these sitting on my cards. They appear to follow a somewhat standard convention, a 0.1" hole at 0.25" from the side and top edges. And I found them to be very inexpensive at Mouser instead, between $0.50 to $0.75 - links are below. In general, these only work when your card has the chamfer edge.
Although, not all of the S-100 cards follow this, many of the early ones I have here have components butting up to the edge of the corners, so it's impossible to install them. Then others have the mounting holes to close to the edge with more components sitting very close, and no chamfer edge to be practical.