[vcf-midatlantic] vacuum tube computers
evan at vcfed.org
Mon Dec 4 13:56:40 EST 2017
> My view is, as a BSEE educated in the 1970's about both analog and
> digital circuits, who saw analog computers in use in academia and in the
> factory, who worked on one or two of them in my working lifetime. The
> hardest thing about explaining analog computing, is that it's almost
> TOTALLY EXTINCT today. It's hard to think or talk about something,
> outside your experiences. And preservation of extinct computing - and
> explaining it in it's own context - matters to me.
> One can explain "voltages not bits", in any number of ways, appropriate
> to audiences and settings. "Temperature" might be a good example.
> There's still mercury thermometers, mechanical thermostats and
> mechanical thermometers. They use the expansion of metals in an ANALOG
> way, to represent "temperature". The Weather Channel, gives it some
> number. But we feel "hot" or "cold" without numbers in our heads. A
> point to consider: the physical world is not lists of numbers, it's
> events and materials and the flow of time.
> "Differential equations" is jargon used by engineers and scientists. But
> it comes down to actions in systems over time. How fast does water boil
> on the stove? Or freeze if put outside in the winter? How long does it
> take your car to go from zero to 60? How long does it take a
> baseball/football in flight to be caught? These are ANALOG values that
> change with time. And, there are analog circuits (components wired up)
> which can represent these rates-of-change calculations. Then you feed
> those circuits the analog values as voltages, and watch what they do
> over TIME.*That* is what an analog computer is about. Circuits and
> components, wired for one computation, using time and voltages to
> represent values - no "bits", no digital (except for numeric results or
> I myself, don't tell people "this is too complicated to explain". I give
> them*something*; it will provide some impression and place to start;
> further consideration will be their choice to make.
> So. I hope the two explanations above, of "voltages not bits" and
> "differential equations", are helpful. I hope the notion of "extinct
> computer preservation" gets some attention. Otherwise, I'm not going to
> debate some ultimate description or how-to-talk-to methods. I've
> explained why, and I've offered my own views.
Excellent answer. Thanks Herb.
I'll throw in another example: volume knobs. Many (most?) car stereos
still have knobs you turn to adjust the volume. It may be easy to tell a
kid that's analog -- a directly relationship between mechanical movement
of the knob vs. increase/decrease of the volume. Whereas if you use a
software control to adjust the volume, even if it the interface is a
slider or knob, that's digital.
More information about the vcf-midatlantic