The 10 or 20 year electronics project

[mzzj]

I would say this idea of 20 years project starts to go dangerously close to something nasa engineers does. Systems with multiply failsafes, etc... In fact thats probably how it should be treated..  2 decades starts to be long enough time frame that ageing effects of components and materials starts to play big role on desing. Not to mention few other things i cant event dream about...

But nasa has one trick that might help to get this started, use the oldest tech you can find. Chips that are no so highly integrated, etc... Those now retired spaceshuttles had really really old tech in flight computers.

I cant even imagine how current plastic capsulated few nanometer line width microcontrollers would even work for 20 years from now. Thermal cycles, oxygen creeps in, etc... Would be interesting to find out. Anyone got time machine? 

Just look around your(parents)  home and You will probably find plenty of 20 years old electronics. Anything low-powered and chances for 30´+ years lifetime  are pretty good.
Of course everything depends on the required reliability level. 5% failure rate for 20 years sounds pretty easy and  0.00001% failure rate per 20 years would be freaking scary.

Even NASA is(was) running ordinary modern laptops on ISS (and shuttles)  but not for mission critical systems!

[SeanB]

Did you see the way NASA handles failure on those laptops? Pop out a spare from storage, plug it in and carry on, with the old one kept as spare parts until the next time a Progress module is near full, when they put them in for incineration over the Pacific. Bit of a problem for the guy on the ground who signed for it, at audit he says - went up on to ISS and was destroyed over Pacific, sorry no idea when.

[Harvs]

In Oz at least you can look around the streets and you'll see plenty of cars 25yrs old that are EFI.  There's not too many every day electronic devices that put up with as harsh a life as those in a car's engine bay being driven to work and back every day.

[Psi]

He's a cool project with long term design requirements.

An underground file server...

You build the thing inside an industrial IP68 metal box (that wont rust/corrode over 10 years) Maybe use a proper time capsule or something.
Inside the box your mount...
- Low power PC like a raspberry pi  (maybe even a second backup one)
- Reliable array of data storage (usb / ssd or server grade hdd etc..)
- Supercapacitor based backup power supply. Should be able to get an hour or so running a pi on supercaps.
- watchdog reset system with the ability to disconnect faulty peripherals that may die and prevent the PC from starting.
- 802.11n Wifi router plus a backup.
- Some sort of passive cooling system, eg all heat sources bolted to the metal box
- The only physical connection to the box is 230V power
- LED light and webcam inside, just for kicks.

Then you dig a big hole and bury the thing in your backyard so the top of it is ~10cm below ground level. Run underground power cable to it and fixed wire that to the house somewhere. (Ideally you'd use a radioisotope thermoelectric generator to power the thing, like the mars rover has, but getting one might be a little hard  )

The goal is to have something in your backyard that will run for 10+ years without physical maintenance. You'd control and talk to it over wifi.
You can store all your *ahem* legally purchased TV shows on it plus backup your files in case your house burns down.

[Griffin]

@Psi that is actually a nice idea. Build it with the idea that it will become the worlds longest running file server. I think that will be a worth file project. Cast the thing in solid potting just before you put it in the ground.

[ddavidebor]

Less things you use less things can broke

Computer broke everytime

Old-school 74xx electronic is more rugged

[JPL]

that's right, my digital clock based on TTL circuits made as my school project in 1978 is working continuously since that time and without any problem. Also every day I still use my HP9815 calculator.  It is working including tape cartridge. Let me one question: where will be PCs with xxx GHz Pentium of today in year 2050 ? 

[ddavidebor]

Inthe trash

[mzzj]

How about 20 year constant-on LED flashlight?

I just checked and old crappy Cree XR-E led will produce enough light to read a book couple of alphabets at a time with 10 uA.
Even without time for my eyes to adapt to the darkness I was able to see alphabets on roughly 4 cm² area.

With good quality 2 Ah Lithium battery you get roughly 2 decades of light 

No need for unreliable switches, you can just cover the led with a plug when you don't need light.

[ddavidebor]

You should use something like 30-50w led with 0.5w only, so extremely underpowered...

I think it will be more durable.

[mzzj]

You should use something like 30-50w led with 0.5w only, so extremely underpowered...

I bet it won't run 20 years non-stop with single small battery

[JoeyP]

That sounds like a good idea. Maybe something as simple as a perpetual clock powered by scavenged RF power - though AM radio stations will probably be entirely replaced by the internet within 20 years.

I bought a digital clock for $0.50 at a garage sale more than 30 years ago, which has been running continuously since that time (9V battery back-up during power outages - but no display on backup).

The end of an era! The clock that I mentioned in this previous post, purchased for $0.50 in ~1982, has finally died! Amazingly, the main chip (TI TMS3450NL) can still be purchased NEW at DigiKey more than 40 years later! But at $25.72/ea they make my $0.50 clock look very expensive. I figure at an average cost of $0.0125/year I got my money's worth out of it, so I gave it a decent burial  Guess it's time I bought something a little more modern, maybe even from this century