Making Radio Chips for Hell | Hacker News Reader

Making Radio Chips for Hell

spectrum.ieee.org

94 pointsrbanffy28 comments

These chips are not really for Hell, as Hell is much hotter than 500°C, see Healey, T (1979). A Refutation of the Proof That Heaven is Hotter Than Hell. The Journal of Irreproducible Results, 25(4), 17-18

Well that felt like a tease. They're using silicon carbide and... That's all the detail she wrote.

SiC and GaN are being used for a fair number of parts these days for RF applications. It seems that the novel part of this paper is that the device worked up to 500C. The device worked for 3 hours with some degradation, likely due to Al electromigration from the metallization.

The acutal paper is here: https://ieeexplore.ieee.org/document/8345290/

I was going to say my employer made this stuff, but they are getting out of the business. Must not be profitable enough.

https://aerospace.honeywell.com/en/products/navigation-and-s...

SOI is an excellent RF process, and is what would be needed to get the rest of the radio on there.

I wonder if simply cooling isn't an option in those applications? I guess that'd depend on the chips power output. Otherwise you could just surround it in a really good insulator and pump residual heat out somehow (maybe can peltiers withstand such high temps using multiple stages?)

Cooling only works if you have somewhere to put the extra energy. Peltiers are inefficient to run and generate a ton extra heat. It's best to build something that can run with as little support needed.

Indeed, cooling would be quite bulky and require lots of extra energy. But for an environment like Venus cited (instead of small sensors), I think it should be a good option. The Carnot efficiency to cool from a Venutian environment (462C) down to 50C (reasonable chip temp) is a surprisingly good ~60%. In practice you might get 15% (? plus insulation leaking), that should still allow cooling a chip of less than ~1 watt of tdp without too much hassle inside a robot.

I thought >300C power MOSFETs in SiC were already commercially available? But I might be confused.

Very neat. Dealing with a thermal problem in a design right now at work; wish these were available for an 802.11 radio!

The issue is that your 802.11 radio needs digital logic, whereas this is just a dumb mixer. Which is, conceptually, just a dual-gate MOSFET run with a bias voltage to exhibit linearity and often without the body diode, in order to permit AC inputs. I do hope for SiC based VLSI, but this won't happen any time soon. One could probably run them pretty hot though, possibly without special cooling fans or such, or at temperatures that allow radiating the generated heat away into space, as long as it's not needed/used for anything better. This would also allow the recycling of waste heat from electronics for industrial processes and generating steam, for which we already know pretty well how to pipe it across a city in a city-scale central heating system.

Absorption cooling (often using ammonia/water as a mixture) could be powered by this and provide sufficiently low temperatures for airflow-free cooling of e.g. hard discs or other components where the temperature would kill them.

I think the VLSI is coming sooner than you expect! There are already teams trying to build a 4 bit microprocessor. And when they get the layering and packaging working, scaling is going to go a LOT faster than Moore's law did because so much of the technology from silicon will just be reusable.

[1] http://www.imapsource.org/doi/abs/10.4071/2016-HITEC-242?cod...

[2] https://www.lpi.usra.edu/vexag/meetings/archive/vexag_14/pre...

Thanks. Any predictions? How long until we get a 486 in SiC?

Didn't realize this was silicon carbide - didn't read original article too closely. I used to work with a guy who did a lot of formative work with SiC.

SiC is used in the field for HVDC inverters, as it can be driven harder and the cooling is less critical. The difference seems to be that they got an RF mixer, while we currently don't use lithography finer than what I can do with a ruler, masking tape and an exacto knife.

OT, but the site doesn't load in firefox for me.

Hmmm. Fine in Firefox (59.0.2) for me.

it work fine with ff 60

$ cat reduce-clickbaitiness.sed

s/Hell/Venus/

And it's portable to every major news outlet.

I think it's a concise and appropriate title. Making electronics that will work on Venus is essentially designing for hell, with all the heat, pressure, corrosives, and who knows what else. And 14,000 Gs of acceleration? Wow. I've never heard of that happening in hell.

This one weird trick lets a microchip survive temperatures of 500C! Intel and heat sink manufacturers HATE this! That would be clickbait.

I would have been more interested if the title had said Venus.

Venusians go to Hell for vacations.

They're talking about a variety of hostile, high-temperature environments, not just Venus. "Hell" is a bit cheesy, but it's a totally reasonable expression for "hostile, high-temperature environments."

Venus is like Hell though right? Super hot, acid, high pressures? I guess still click-baity though?

Sure, but "Hell" could equally apply to underground on Earth or the surface of basically any of the planets of the solar system. The conditions on most of them would kill anybody within seconds in pretty gruesome ways. Swapping that one word makes the headline much more accurate, interesting, and information rich.

I really did not expect this to trigger the HN crowd so much. Changing this word would objectively make the headline more accurate.

Hell Town's hot, but according to the Wikipedia article has record highs of 42C in the summer - rough, but bearable. Not sure about the acidity, though.

...why

If offered a choice between Hell and Venus, I'd pick Hell. At least the commute is shorter. Also lots of friends already there.

To be fair it's a pretty good article despite the title.

Reply on news.ycombinator.com