People prefer vacuum tubes because of their particular non-linearities during amplification. A vacuum tube is just a very simple particle (electron beam) accelerator.
It's not a beam unless you work really, really hard to make it so. A typical filament in an ordinary vacuum tube for amplification is an omnidirectional emitter, the particles will be accelerated towards the charged grid of the tube.
Only in very specific 'cathode ray tubes' (which is how traditionally monitors and TVs would work) do the particles become a beam.
Very true, but for all intents and purposes as long as you get similar non-linearities, the actual design doesn't matter much. People are using neural networks for this now but I would prefer actual hardware. Bonus points if they can make the chip give off the nixie neon glow (maybe add some heavy metals for Cerenkov radiation?)
A vacuum tube in an amplifier is more of an electron attenuator than it is an accelerator. The electrons don't actually gain any energy in these vacuum tubes past the energy they have when emitted. They still travel at the same "speed" that they normally would around an atom, the only difference is that they are freed from the atom. Acceleration is not part of what an amplifier tube does.
> They still travel at the same "speed" that they normally would around an atom
Can you unpack "speed" there?
I figured the speed of an electron in an orbital was the same as the speed of the atom.
Certainly there's some energy associated with it being in that particular orbital, and if it takes on more energy then the region in which it might be measured will expand, but it's not like zooming around in there is it?
I am by no means an expert in particle acceleration, but I have worked with tube amplifiers extensively, specifically for audio (modifying and building tube guitar amplifiers).
The suggestion of the parent comment to use a nano-sized particle accelerator on a chip could somehow be useful for tube amplifiers just seemed ridiculous.
From your username I thought you might be an expert :). All I have is a small pile of physics classes, so I'm also not an expert. Still, I agree: ridiculous.
Just to nit pick your "same speed" comment though...
The tube you linked to indicates 250V at the anode. Electrons have a mass of 10^-13 kg. For such a tiny thing, that voltage is positively tremendous. Even in the tiny space of the tube I think that the electrons are accelerating from 0 when at the cathode to relativistic speeds at the anode. The tube is an accelerator, and the control grid opposes that acceleration.
Accelerating the electrons to even greater energies would just mean that more voltage is needed at the triodes to have any modulating effect on the signal. Also, the beam would be, well, a beam, rather than a... gust? It would either strike the anode or miss it--that's not modulating anything, that's just an on-off switch.
For simple thermionic electricity generation, sure; this is the effect in play when the first grid gains charge with no bias applied (and a near infinite load resistance). When a positive charge is applied to the anode, the electrons do accelerate although the effect isn't comparable to a magnetron, cyclotron or linac.
