[sdiy] uA726, heated pairs (slightly OT)

[metasonix]

>> ... but why the heck do you need a thermostated matched pair in a
>> missile?  I really found that kind of bizarre.
>They were part of all kinds of basic modules (multipliers, square roots
>extractors, expo converters, etc) on analog computers, used in automatic
>flght control.

And if you do some research on early missile development, you will find that
large rockets didn't become really reliable (and the US space program 
didn't get started) until mostly solid-state control and telemetry circuits 
were developed. 
Before that, they tried to use subminiature tubes, with questionable results 

Raytheon actually made a bizarre tube that performed square-law
conversion. Yes, for missile guidance analog computers. 
And National Union made analog telemetry commutator tubes. 
Literally a big (fragile) glass tube doing the function of a 4051 analog mux. 
There were also CRT-pickup dual tubes used to trace over a clear plastic 
graph and generate a control voltage, in kinescope fashion. 
You will see NONE of this stuff in old tube manuals.

A major reason why ceramic planar tubes were developed, and then 
semiconductors, was to make missile guidance systems more
reliable. Computers didn't count as a market until transistors dropped 
below $1 wholesale price, in 1958.
 
Consumer electronics didn't count for much in technological development. 
Tube TV sets were still being made by Sylvania until 1976. 
And in Russia until 1995. The only place transistors made a real
difference was in portable radios, until 1970 or so. 



uncle eric
metasonix.com