how weird are the electronics actually used in nuclear weapons?

[coppercone2]

I mean if a standard engineer that maybe saw a taste of low end military was actually presented with a circuit, would we go 'holy crap that belongs in the art of electronics as its own chapter' or would it be more like 'they dotted their i's and crossed their t's'?

[raptor1956]

Timing is key so someone familiar with DSO front end design might appreciate the design.  I would suggest, however, that we avoid talking in any greater detail on this subject lest someone notice...


Brian

[retrolefty]

While no first hand knowledge, I've read enough to think that the electronics would fall into two categories:

1. Security, including safeties, interlocks, authentication, etc

2. Critical timing requirements for the detonators. Nuclear weapons require very high speed and consistent timing to insure the implosion is uniform enough to ensure critical mass is accomplished.

From wiki:

Krytrons and their variations are manufactured by Perkin-Elmer Components and used in a variety of industrial and military devices. They are best known for their use in igniting exploding-bridgewire and slapper detonators in nuclear weapons, their original application, either directly (sprytrons are usually used for this) or by triggering higher-power spark gap switches. They are also used to trigger thyratrons, large flashlamps in photocopiers, lasers and scientific apparatus, and for firing ignitors for industrial explosives.

 I did have military training and experience servicing teletype and some crypto equipment in the late 60s. Teletype was old tech even back then and I would classify it as mostly electrical rather than electronic. Crypto was a different animal and depending on the specific model the electronics was quite impressive at the time. The KG-13 was the newest design I saw and was very cool in design and construction, all solid state. The other I saw was the KW-26, which was older vacuum tube based and a great heat generator, kept the comm centers quite toasty.

[ajb]

Probably quite "weird" in the same way that all stuff that has to cope with severe mechanical stresses, have a long shelf life, and be 99.9999% reliable is.  Extreme attention to mechanical details, possibly unusual substrates/stackups, hiRel components throughout, etc.  I wouldn't imagine that the trigger mechanism is all that strange, there's definitely an art to getting timing right, but I wouldn't think it would qualify as "weird".  The associated guidance/fuse systems are probably more interesting in that respect.

[coppercone2]

but visually compared to stuff like russian missile electronics you can get off ebay, is it going to look much different? will unheard of amplifier topologies and stuff be used for the hi-rel reasons?

I figure they will put the best possible parts in the coolest packages and with the best strain relief and everything, but if you just look at the circuit diagram are you gonna think "why the hell....?"


i imagined like very wide input ranges and stuff like that, i.e. power supply that is 5V nominal ends up running fine off 500V.



did this ever happen to you with a teletype retrolefty? (watch last 1/3rd or so).
 

[CatalinaWOW]

Having spent a career in this I can offer some fairly accurate general observations.

1.  Engineers in defense have a wide range of skills just like everywhere else.  It shows in the designs.  A specific example of the low end I can mention is an amplifier circuit that had four inversion stages in series.  It was part of a control loop and there kept being confusion over the phase.  Successive engineers, rather than understanding what went before just added another inverter instead of removing one.  By the time I got involved there wasn't time to remove them (and go through all of the environmental qualification) before field test so it just went out that way.  But on the other end you occasionally come across some brilliant topologies or uses of parts.

2.  The vast majority of circuits can trace their pedigree to app notes.

3.  Environmental requirements do drive design a lot.  Other industries like automotive, space and oil field instrumentation have tough environmental requirements also, which is an aid in getting semiconductor parts that meet tough requirements, but are different as they are different from each other.

4.  Much military equipment is expected to sit in storage for many years and then work immediately when unboxed and turned on.  It turns out that that is very different from, and often more difficult than operating for the same number of years.

5. The defense marketplace is small, and currently not socially popular.  Thus you either make do with whatever the commercial world produces and do all of the characterization in your environment yourself, or pay the bill for a specialty manufacturer doing very limited production quantities.  If you use a commercial part and characterize it outside its commercial specifications you will likely have to start over every time they change a process (which often happens with no warning).  This leads to design choices that would seem strange in the commercial world.

[coppercone2]

what do you mean by non socially popular?

i assume a bunch of series inverter to extend bandwidth with specific impedance in mind?

[Brumby]

Society - that is, people - don't look upon defence suppliers with warm and fuzzy feelings.  A lot of the active social voice is happy to decry them as predators on human misery and destruction - and the silent majority is not showing signs of disagreement.

The patriotic reaction to war is not what it was 3 or 4 generations ago.  Certainly there is still a strong support of one's country - but increasingly these days the processes behind the scenes are held in contempt and cynicism abounds.

[coppercone2]

their actually effected by the minor protests we put up sometimes after camel strike #60434?

I know its bad in some countries like portugal when I spoke to some marine special guards but in general I thought it was a fart in the wind. Especially for nuclear. Also some places like Malta's airbases in the 19x0's (def not nuclear free) but now?

[T3sl4co1l]

Nothing very remarkable really, in terms of modern capabilities of timing or logic, or general reliability (the above discussion of military requirements notwithstanding).  The only standout is peak power of the firing pulse (its energy delivered in such a short time), which is much higher than you can do with most transistors.

Let's be real, they originally built these things with 1940s tech after all.

Tim

[coppercone2]

Nothing very remarkable really, in terms of modern capabilities of timing or logic, or general reliability (the above discussion of military requirements notwithstanding).  The only standout is peak power of the firing pulse (its energy delivered in such a short time), which is much higher than you can do with most transistors.

Let's be real, they originally built these things with 1940s tech after all.

Tim

keep in mind the weapons have different generations and weird stuff like tactical odd shapes, bunker busters, etc. who knows what kind of horrors were developed on a drawing board and possibly require different active technologies. given how many they made i am sure they were trying everything they could to make em smaller and cheaper and lighter etc.

[coppice]

There's nothing very special about the circuitry in nuclear weapons. However, missiles must withstand high G forces, so the packaging schemes can be quite novel to deal with those physical stresses.

[CatalinaWOW]

Socially unpopular.  Several major vendors refuse to sell their products to defense firms.  For a specific old example think of Motorola's decision to stop selling parts used in land mines.

My comments weren't specific to nuclear weapons, but nuclear weapons have much in common with other systems.  Arming and fuzing (plus the extra layers associated with nuclear safety).  navigation, guidance, built in test, power supplies...

[EEVblog]

what do you mean by non socially popular?

Tell people at a dinner party you design nuclear or some other weapons system and see how well it goes down.

[coppercone2]

hmm they don't want middle men selling land mines (they get em right into your pocket?)

[coppercone2]

what do you mean by non socially popular?

Tell people at a dinner party you design nuclear or some other weapons system and see how well it goes down.

i think once your that far up the equipment chain you are scared to talk to people at dinner because the KGB might rip your head off no?

nuclear might be better because they don't test it on goats anymore? worst case is what some ameoba or something getting fried in a super secret subterranean test, so long your not in north korea. i hope so anyway....

[floobydust]

Anyone remember the story of a defense contractor (engineer) that set up a store across the street selling electronic components. There was so much paperwork and approvals to procure a component through the military channels, taking months so he set up (his store) with Purchasing and walked across the street to get the parts him and others needed.

I remember it because the story said some Radio Shack capacitors ended up in cruise missiles.

[james_s]

Not really the same thing, but at a previous employer my boss once went to Costco to buy one of our own products because somehow it was cheaper for my group than acquiring it through the standard internal channels. I don't know how that works but there were some pretty convoluted arrangements.

[VK5RC]

In a similar but different vein, one large organisation I worked for had it's own mail section, with two full time staff, it would still take on average 5 working days to hit the mailbox. Any time critical mail, you would buy the stamp and walk it to the box yourself.

[0culus]

There's a reason the really high speed 'scopes and related test equipment from US companies are on the ITAR in the US, and nuclear weapon triggering is pretty much that. At the component level, even "old school" stuff like krytron switching tubes are export controlled.

https://en.wikipedia.org/wiki/Krytron

[hans]

In some colleges, courses about philosophy of engineering have become mandatory. Not only to teach the ethics in academia, like fraud and citing your sources, but also keeping in mind for what causes your inventions can be used.

We have had presentations from some military companies making some really cool systems with very sophisticated information technology in them. But in the end it is a military system, either used to determine defense or offense tactics. That can have ethical pros and cons , depending on which side you fantasize more with I guess.

Likewise, oil/fossil fuel companies have become incredibly non-popular to work for (i.e. internships etc.) in recent years. Having research groups that actively do work on green energy systems doesn't "help" neither.

[CJay]

There's a reason the really high speed 'scopes and related test equipment from US companies are on the ITAR in the US, and nuclear weapon triggering is pretty much that. At the component level, even "old school" stuff like krytron switching tubes are export controlled.

https://en.wikipedia.org/wiki/Krytron

That's just one reason, there are many others, some of which are obscure and non obvious without a little knowledge of how certain military technologies work.

[coppice]

There's a reason the really high speed 'scopes and related test equipment from US companies are on the ITAR in the US, and nuclear weapon triggering is pretty much that. At the component level, even "old school" stuff like krytron switching tubes are export controlled.

https://en.wikipedia.org/wiki/Krytron

Most of the ITAR restrictions, including scopes, relate to sensor technologies, not nuclear ones.

[james_s]

It would be pretty silly anyway, nuclear weapons have been around since the 1940s, plenty of less than friendly nations already have all the knowledge and tools they need to build them. Controlling exports on test equipment isn't gonna do a thing. Any entity with the budget and motivation to build a weapon will find ways around the export controls, there is no shortage of people willing to smuggle items for the right price.

[0culus]

Well, yes, the potential applications in the military space are quite broad, and the ITAR is fairly vague for this reason.

 

It would be pretty silly anyway, nuclear weapons have been around since the 1940s, plenty of less than friendly nations already have all the knowledge and tools they need to build them. Controlling exports on test equipment isn't gonna do a thing. Any entity with the budget and motivation to build a weapon will find ways around the export controls, there is no shortage of people willing to smuggle items for the right price.

I would assume that this is fairly obvious. I was commenting on the fact that a lot of this stuff is ITAR. Whether the ITAR actually prevents anyone from doing anything is another story.