Comment by theandrewbailey | Hacker News Reader

Comment by theandrewbailey | Hacker News Reader

Parent Full thread

theandrewbailey

Several years ago, I read about Americans capturing Soviet-era jet fighters. They were analyzing it and finding all sorts of deficiencies compared to their own fighters, but noted its vacuum tube electronics were resistant to EMPs.

> Perhaps the most stunning discovery was the fact that most of the MiG-25’s avionics were based on vacuum tubes—not solid state electronics. This was considered woefully outdated for a top-of-the-line military jet in the 1970s, but the vintage system had its advantages. The vacuum tubes were more temperature-tolerant than modern avionics were, which allowed the MiG to fly without weighty environmental controls in the avionics bay. Plus, the tubes allowed for quick and easy maintenance at Russia’s primitive airfields, and the antiquated system would better withstand the circuit-frying power of an electromagnetic pulse created by a nuclear blast.

https://www.popularmechanics.com/military/aviation/a38083125...

> Plus, the tubes allowed for quick and easy maintenance at Russia’s primitive airfields,

I don't buy the maintenance angle.

Modern solid-state avionics are swappable LRUs. Like lego, you plug and unplug them. Just as easy as tubes, in fact the exact and precise same. Plus you can drop them and they don't shatter.

Techs aren't out in the field desoldering surface-mount components and swapping capacitors, the are unplugging and plugging modules the same as unplugging and plugging in a vacuum tube.

I also don't buy the EMP angle. The metal that aircraft avionic components are encased in also serves as EMP shielding. Not that that matters, the effects of thermonuclear EMP effects are so poorly understood that anyone claiming to be able to predict them is a liar. An EMP in the upper atmosphere may fry well-shielded components 1 meter away from unshielded components if the unshielded components are disconnected and at a different angle to the blast than the shielded ones, and vice-versa. Every real-world analysis of EMP effects is punctuated by how seemingly random the effects are.

An analysis of the effects of STARFISH PRIME on the streetlights of Honolulu notes how random and unpredictable the outages were, and this is on unshielded long-wire circuits (ideal for soaking up EMP radiation).

Did High Altitude EMP Cause the Hawaiian Streetlight Incident? http://ece-research.unm.edu/summa/notes/SDAN/0031.pdf

USSR was using tubes because it was technologically far behind the West (it was still producing cars that were between 30 to even 50 years behind the West) and was under heavy sanctions. Soviet bloc made chips were actually less reliable and that's one of reasons why commie bloc TTL based computers were so unstable as they had either to import components from the West (see story of Karpiński and his K-202 architecture) if they wanted components that they were allowed to order them in the 1st place or smuggle them (which would not work for large scale) or use inferior in quality components produced within the bloc. Vacuum tubes were much harder to fuck up while production – hence more reliable than commie bloc TTL or LSI. How do I know? I am from one of countries that been in that bloc.

This is exaggeration and oversimplification. From American/German high quality tubes.

I've deal with old soviet tubes. They where extremely specific, You even couldn't just place new tube into device, but need to make some manipulations with it before turn on (run it in special device on low currents).

And unfortunately, soviet tubes unreliable, many (not all), work very few hours. So MiG-25’s where for soviets, like SR-71 for US - very expensive in making and in support.

For example, known fact, MiG-25’s could for tiny time make more than 2.5Machs, but after this 100% engines go to landfill.

Yea, just look at how much redundancy in satellites and space probes we need to operate within the range of Sun's heliosphere. Outer space outside it would be even more killer to current electronics. Even if we switch to silicon carbide (which could possibly still remain harder to manufacture process) stuff it still would be a factor to consider.

> Even if we switch to silicon carbide (which could possibly still remain harder to manufacture process)

In the 1980s I got a copy of the latest Cricklewood Electronics catalogue that had a whole page devoted to these newfangled SiC blue LEDs. Blue! Actually properly blue, too!

They were 40 quid each, and indeed the text box where it gave the price said something like "£40 - yes, really, yes these are one of the most expensive components we have ever stocked".

And for the past 10 years, almost every cheap electronic device has been permanently glowing blue because of them.

EDIT: aren't modern blue LEDs based on GaN?

Silicon Carbide LEDs were the first true blue (not blue-violet) ones, but they were horribly inefficient.

Doped GaN made blue LEDs viable, and LEDs in general more available.

Yah, the first ones were a very different shade of blue and really expensive.

The inventor of the blue LED came to our school for a talk like 5 years ago. IIRC he won a Nobel prize?

He got the Nobel Prize because the blue LED + phosphor enabled the use of cheap, energy-efficient LEDs for general illumination, whereas previously LEDs had been used mostly for indicator lights.

<insert Cmdr Scott or Lt Cmdr Data bafflegab>

sounds like the [possibly-apocryphal] store of the invention of red-black trees

during the space race, the Americans threw more computing power at the problem

the Soviets didn't have the computers to use, so they threw math at it

sometimes simplest is bestest!

Reply on news.ycombinator.com