First, there were the cogwheels.
Then came the electric relay.
Then came the vacuum tube.
Then came the transistor.
Then came the integrated circuit.
Integrated circuits are fast, miniaturised, cheap and consume few
energy. They are arrays of transistors printed on a chip of silicium
and wired through photolithographic techniques.
Why not print arrays of vacuum tubes?
Cathodes and grids can be printed on a piece of ceramic and be wired a
way similar to transistor arrays. Then, for example, an electron canon
can be aimed at the vacuum tubes chip in order to make the little tubes
work. That would resemble a little television tube with the electron
canon at one side and the surface of cathodes and grids on the other
side,
instead of the screen. The electron canon's purpose is not to aim at
one pair of cathode and grid at a time, rather spray fast electrons
towards all cathodes and grids globally. Each grid will modify the path
of the electrons it
receives according to its logical state.
This is a drawing of a possible NOT gate:
If the grid has a level of 0V, it will let the electrons go straight
towards the cathode. That will make the cathode receive the electrons
and become electrically negative.
If the grid has a negative level, it will repulse the electrons hence
modify their path. They will no more hit the cathode. So the cathode
will remain at 0V.
When the cathode becomes negative it will repel the electrons too. What
matters is that it needs to be strongly more negative than the grid in
order to be no more hit by electrons. That's the condition in order to
be able to speak of "amplification".
A drawing of a NOR gate:
An integrated vacuum tubes circuit would thus be an array of such gates
with the wiring underneath. It would be encapsulated in a little vacuum
shell, facing a rudimentary electron canon.
Some advantages of integrated vacuum tubes are the following:
The ability to resist extreme temperature, maybe above 2,000 K
(land a permanent probe on Venus, security devices...).
The ability to resist strong radioactivity (nuclear power plants,
remote planets...).
They can be made out of materials less fragile than common
semiconductors.
I cannot estimate if they would be more or less expensive than
semiconductor devices. In order to make them have a long lifetime it
will be necessary to include inside the vacuum shell some device to
maintain a good vacuum (for example a little ion accelerator that would
launch the gasses away through a thin wall).
For special purposes the electron canon can be replaced by a beta-rays
radioactive source. The circuit would then work tens of years with no
external electric power source.
Electron canon devices can have other applications for modern usage. A
CD that would have a thin surface of aluminum read and written by an
electron beam would have a storage capability tens of thousands times
greater than a normal laser diode CD. Just put the CD into the electron
CD-drive, wait a second till a strong vacuum pump does it's job, then
you may freely read and write the aluminum surface. The electron canon
may be a little device moved above the CD surface just like a laser
diode is. It may also be a much bigger device, with no moving parts,
reading the surface of a stamp CD a way similar to an electron
microscope. In that case the aluminum surface must contain "calibration
zones", in order to allow the electron beam to move around a relative
way and not having to move with absolute coordinates (expensive).
No tests have been made of the content of this page. An integrated
vacuum tube is buildable for sure and probably will match at least part
of the advantages mentioned. Maybe the way described here to conceive
it is not the best. Anyway what matters is the principle of an array of
printed vacuum tube elements.