A Transistor for Sound Points Toward Whole New Electronics

https://spectrum.ieee.org/topological-transistor-acoustic

by graderjs • 4 years ago

102 35 4 years ago

spectrum.ieee.org

gene-h• 4 years ago

Controlling the transistor requires an intermediate step converting the sound to heat. In my book, this isn't really an 'acoustic' transistor. Phononic transistors have been proposed though[0]. Phononic transistors have been built, albeit with some magnets, but merely for their nonlinear mechanical properties[1]

[0] https://arxiv.org/abs/1509.00679

[1]https://www.pnas.org/content/114/18/4603

omeze2• 4 years ago

I recently read “To The Digital Age” (would recommend) which covers the history of the transistor and I left with a few takeaways:

- New transistor architectures can matter a lot (MOSFETs are why we enjoy the benefits of Moores law, BJTs wouldnt have gotten us here)

- It took an entire industry decades to mobilize the tooling, manufacturing, art, and science of a single architecture (MOS) even with related tools for BJTs

- Once an architecture shows promise, even giants like IBMs and its infinite funding cant fight against the tides

- Even when some people know an architecture is promising, organizational politics can prevent work on it

So, new architectures are always interesting but they have a huge slew technology transfer problems. But if theyre better in some way, they can find a niche and over time (decades) may turn into the next big thing. But I wouldnt hold my breath.

afandian• 4 years ago

Is that ISBN 9780801868092?

omeze2• 4 years ago

Yea, thats the one

mwattsun• 4 years ago

This documentary by PBS claims William Shockley was an ego maniac who stole a lot of the credit for the transistor from his subordinates John Bardeen and Walter Brattain

https://www.youtube.com/watch?v=U4XknGqr3Bo

aksss3• 4 years ago

This is getting to be a trope - "what you knew is wrong, the guy who is generally given credit was actually an evil bastard who oppressed these other team members who did the lion's share of the work."

I'm not sure we need sacred cows and I have no dog in a fight for credit for transistor, mouse, DNA, electricity or anything else, but the constant demand for deconstruction - "everything you know is wrong" - gets a bit wearisome.

hulitu2• 4 years ago

it is suspected, Potential, scientists at Harvard have now invented and simulated

So another hype. I thought IEEE are serious people eveb if they charge a lot of money for their publications. You can simulate _anything_. All you need is a simulator and a model. Reality is a bit different.

mikro2nd• 4 years ago

Do you suppose that modellers are unaware of the differences?

hulitu2• 4 years ago

I don't think they are not aware. But for fame, they ignore it.

cracrecry• 4 years ago

In particular you need to know what to simulate, because when you simulate you leave lots of things outside the simulation.

The only thing that proves that you made a good work ignoring those things is prototyping on the real world.

Without that it is just vaporware.

hughrr2• 4 years ago

God no they are a shit show. Having been an EE years ago I stopped reading any journals and research. 99.9% of it is garbage which is engineered to sell attention rather than a better future.

Anything worth giving attention to turns up in EDN when a product comes to market after someone took it serious enough to actually invest in it. Then you pay attention.

My favourite recent thing is chip scale atomic clocks: https://www.microsemi.com/product-directory/clocks-frequency...

You can actually buy them!m

Edit: the worst garbage is actually IET’s E&T Magazine. That’s horse dung.

ChrisMarshallNY2• 4 years ago

You can actually buy them!

For $5,200.

I wonder how many they sell. They would be useful in things like rangefinding and positioning (and maybe encrypted packet comms).

Sounds like a military market.

pwr-electronics• 4 years ago

I didn't really understand IEEE until I realized that they're more of an academic society than a professional society, despite what their marketing suggests. The academia mill is a weird way to fund training engineers for industry, and IEEE is just a forum for it.

SOMA_BOFH• 4 years ago

very interesting and innovative. great work.

Animats• 4 years ago

A heat-activated switch for sound? Is this extremely slow, or what?

cousin_it• 4 years ago

I think the idea is they hypothesize a new way to build electronic transistors, and the equations it relies on happen to have an exact counterpart in sound, so they built a sound-based one as a kind of scale model.

graderjs7• 4 years ago

I thought that was part of it, but with s/electronic/photonic/g

dontcare007• 4 years ago

Wow, I wonder if topological materials exist for heat? It would be cool to have a super efficient heat exchange or deliver heat to a distant source...

A Transistor for Sound Points Toward Whole New Electronics

Related Stories