Particle accelerator on a chip hits penny-size | Hacker News Reader

Particle accelerator on a chip hits penny-size

spectrum.ieee.org

125 pointsthunderbong18 comments

The particle accelerator itself is on a chip. The laser that powers it, and the electron gun that injects the electrons, are not.

The smallest and simplest practical particle accelerator are pyroelectric crystal linear accelerators. They are typically cylindrical rods of lithium tantalate (or niobate) cut so their pyroelectric face is exposed on the ends (~1 to 2 cm long). Put one of these in mild vacuum (~10^-4 torr) and use a resistor to periodically heat it up and let it cool down. During the temperature changes moving charge will be exposed on the crystal's pyroelectric faces. This exposed charge creates voltage gradients which can be used to accelerate ions or electrons (alternately) up to 200-300 keV, about 10x the peak speeds of the linked laser accelerator (over ~5x longer distance).

People have used them as linear accelerators with deuterated plastic targets to accelerate deuterium ion gas residuals and achieve toy-scale desktop fusion. https://en.wikipedia.org/wiki/Pyroelectric_fusion They were also used as a portable x-ray source in a commercial x-ray fluorescence spectroscopy kit called "Cool-X", https://www.amptek.com/internal-products/obsolete-products/c...

If you powered the resistive heater with a battery the device could literally be only a few cm long and maybe a cm radius. Very easily held in the palm of the hand. It's not a penny, but it's the full device and it's energy storage.

Where'd you learn all of this?

It's just an amateur interest for fun. I think I heard about it first in the late 2000s in some press release. I've been following the academic literature since, ref: http://erewhon.superkuh.com/library/Physics/Pyroelectric%20A...

The new vacuum tubes for audio will be amazing

How does this enable vacuum tubes for audio?

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.

https://en.wikipedia.org/wiki/Cathode-ray_tube

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).

https://en.wikipedia.org/wiki/12AX7

https://www.effectrode.com/knowledge-base/the-12ax7-tube/

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.

Can a (purpose built) chip simulate tube output so their alleged benefits can be had for cheaper?

Fascinating! We are closer to ST TNG "dermal regenerator" and let's hope that much earier before XXIV century https://memory-alpha.fandom.com/wiki/Dermal_regenerator

Seems a good alternative for a hobby like telescope for astronomy.

Does this mean we'll have handheld Ghostbusters-like guns soon?

Reply on news.ycombinator.com