How long does the pi last?

[gladoscc]

Assuming it's handled correctly, not overclocked, etc, how long would the pi last before breaking? A few months of 7x24 usage?

[WereCatf]

Assuming it doesn't get damaged by moisture, it should last something around tens of years.

[gladoscc]

Wow, 10s of years of it constantly running?

[rurwin]

Assuming a no moisture, and assuming it was not writing any data anywhere, then it should last forever.

In practise that might be as small as a century or two.

If it is writing data anywhere, then the storage medium will wear out first, and that could be anything from tens of years up to centuries.

If it is kept in a drawer and never used then maybe as little as ten or twenty years. It will last longer if it is used, and longer still if it is never switched off.

In practise, dirt, moisture and accidents will probably be the downfall of most of them. Even normal atmospheric damp will corrode it eventually. But if you care for it properly, then decades would be perfectly achievable.

[rurwin]

If you want to know how long it will run before something causes an error, then probably five to ten years. Then surface corrosion on one of the sockets will cause a fault. You would have to power-off and plug and unplug a few times to clean it.

[WereCatf]

doesthenameappear said:


Wow, 10s of years of it constantly running?



I do not understand why you assumed it would last less than that. I have a 80286 in the closet that is still perfectly functional and before I got it it was in constant use for almost 10 year straight. RPi has no moving parts so the first part to die would be storage media, all other components should easily last atleast 10 years, likely much, much longer.

[bredman]

As an example, I have a similar ARM5 machine which has been running 24/7 for almost 4 years as a file server in my attic.

Because it has a battery backup, it has kept running even during power cuts. As far as I remember, I have not restarted it for at least 3 years now. I probably would have to crack the password if I wanted to log in now.

I expect the hard disks or power supply or battery to fail before the computer does.

Before you ask, I know it is still running because it takes a dejavu backup of all my computers every day.

[mole125]

The other type of failure that could occur is that of some of the solder joints, this is only likely to happen however either if it regularly experiences high levels of vibrations/shocks or repeated extreme cycles of fast heating and cooling - the variations in expansion coefficients of the different materials places mechanical stress where they come in contact with each other.

My guess would be that the demise of most RPi will be for the following reasons:

1. Getting stepped on/squashed

2. Being lost in the back of the drawer and forgotton about

3. Getting hidden among rubbish and being thrown out by mistake (it's so small it'll be easy to happen when not plugged in)

4. Poor power supplies frying the board.

5. Poor soldering burning out chips on the board.

[rurwin]

Somewhere around number 1 or 2 on that list should be:

Having coffee or coke poured onto it.

[Jongoleur]

mole125 said:


My guess would be that the demise of most RPi will be for the following reasons:

1. Getting stepped on/squashed

2. Being lost in the back of the drawer and forgotton about

3. Getting hidden among rubbish and being thrown out by mistake (it's so small it'll be easy to happen when not plugged in)

4. Poor power supplies frying the board.

5. Poor soldering burning out chips on the board.



And in the intended school environment, "The dog ate it..."

[RaTTuS]

Sometimes the electronics will fail - due to bad manufacturing it all depends ...

I've had some motherboards that died after a few years [capacitors leak, a retention clip de-soldered itself from heat and the heatsink poped off and killed the chip ] , mostly they last longer than the tasks they are used for ...

I had a P90 running for 10+ years as a printer server - it was switched off when the printers died and we no longer had any parallel port printers...

my most recent failures where a collection of 8port gig switches - worked fine 24x7 for 5 years then stopped working - they will not recognise any connections to them  - of if they do then they are flagged as 10Mb instead of 1Gb- [I've got 8 of them broken 5 will not work at all and 3 partially work EDIMAX e8-5800P's ] - they all stopped within 3 months of each other .... YMMV

but generally - they will last longer than you need ;-p

[adlambert]

My 1970s Casio calculator is 100% functional, as is my Sinclair ZX81. I also have a 1970s digital LED watch that is working fine.

My 2011 Samsung Android phone packed up after 5 months (not water damage).

[Burngate]

Where I worked there were ~50 video disc recorders, each with 10 scsi raid drives, working 24/7 with almost no problems (except software crashes) for several years.

Power to the building was cut by a JCB

When the power came back, about 50% had one or more non-functional drives. After replacing those, they carried on for a further number of years before being replaced for operational reasons.

[Lynbarn]

Not quite on topic, but my mother had to replace her chest freezer just before Christmas. We found the receipt on the back of it - she had bought it in 1970, and had been in continuous use (barring brief periods for 4 house moves) throughout!

[Burngate]

One of the more common failures we had was electrolytic capacitors dieing.

In switch-mode power supplies, the high frequency of the switch (relative to linear PSUs) means that smaller capacitors are needed, but they still carry the same ripple current, and generally have higher internal resistance. This implies they get hotter, and dry out quicker, so reducing the life of the supply.

Elsewhere, the life of a capacitor is very dependant on its temperature: if the equipment runs hot the capacitor will dry out quicker. Modern equipment uses less power than old stuff, so gets less hot, so lasts longer.

Our Pi uses very few watts: it won't get warm unless something near it gets warm, so its life should be longer than you need.

[Burngate]

rurwin said:


Somewhere around number 1 or 2 on that list should be:

Having coffee or coke poured onto it.



My laptop keyboard died after a couple of months.

It wasn't coffe or coke - it was cider

[ArborealSeer]

WereCatf said:


doesthenameappear said:


Wow, 10s of years of it constantly running?


I do not understand why you assumed it would last less than that. I have a 80286 in the closet that is still perfectly functional and before I got it it was in constant use for almost 10 year straight. RPi has no moving parts so the first part to die would be storage media, all other components should easily last atleast 10 years, likely much, much longer.


Yup. We still have some old 386s in our production dept here that are used daily. They have gigantic handmade custom cards in their ISA slots that are made from prototype board type stuff. They run a DOS based pascal app to 'flash' our hardware devices at a point during the build process. They were here before I joined in the mid 90s, and I believe date back quite a few years before that. So I wouldn't be suprised if they aren't turning 20 soon. The disks may have been replaced here n there, but outside booting their sw is running from the network. They run co-ax ne2000s into a switch that does co-ax->10meg that then connects to the same gigabit switch in our server room everything else does.  Our pick n place machine is also used regularly. Its powered by a 486 and dates from the late 90s too.

Good PC hardware will go on and on..

[rurwin]

My first computer was an MK14.

The edge connector was behind the voltage regulator, which had a large piece of metal as a heatsink. After some time the continual movement broke one of the three legs on the regulator. It was of course, the middle one, the only one that could destroy the board. I only noticed when the paper dot on the ROM started to send up smoke signals. It was an unregulated 12V power supply, so I assumed that I had fried everything.

So skip forward several months, I finally got around to sending it off to the manufacturers for repair. And a few weeks later I got it back. In fact the only thing that had been bust was the ROM. I hooked it up and laboriously typed in a program. It worked! I called over my sister to have a look.

She leant over the back of the arm-chair, holding a mug of tea...

It didn't work immediately after, but I took it to university anyway, and the journey must have broken the suger-based short-circuit because it worked perfectly until it was obsoleted by an Acorn Atom.

[grumpyoldgit]

I would imagine that the first things to go would be the connectors. With just a bare board and constant pulling, I suspect it won't be long before someone manages to yank something off the board. A properly designed casing will cushion this sort of wear and tear. Has anyone mentioned death by pet yet? As soon as I sit down at the computer the cat insists on walking all over the keyboard and generally getting in the way.

[johnbeetem]

Burngate said:


One of the more common failures we had was electrolytic capacitors dying.


These days, the things I worry about most are:

1.  Electrolytic capacitors, as Burngate mentions.  If a chip runs hot, the electrolyte boils away over time.  This is especially of cheap aluminum can caps.  There are more expensive versions for high-temp applications, but manufacturers often use the cheap ones prefering that consumers need to buy new product every few years instead of every few decades.  Failure mode is that voltage is no longer well regulated and non-reproducible errors occur, followed by hard failure.  Mitigation is to not design products that run hot.  Use ARM instead of x86, for example.

2.  Ball grid arrays and other lead-less packages ("lead" as in "Leeds").  These are nasty packages from a reliability standpoint.  All packages used to have pins that went through the PC board or "gull wings" where one end connected to the chip and the other was soldered onto the PC board.  Leads are a great way to absorb the thermal expansion difference between the IC package and the PC board: if there is a mismatch, the lead flexes and the solder joints remain intact.

With a lead-less package, you're hoping the solder flexes if there is a thermal mismatch.  The problem is, solder doesn't flex: it cracks.  After enough thermal cycles, a few of those hundred of balls start to have poor connection.  Ah, but which ones?

Solder ball fractures are of great concern for industrial temp products that live in the desert where the daily thermal cycle can be pretty bad.  Outer space can be even worse: you may have a satellite that's behind Earth part of the time and fully exposed to the Sun part of the time.

To mitigate, leave your device on all the time to avoid thermal cycling.  However, then your electrolytic caps boil away.  Sigh.  I've heard of hackers having success reflowing the balls on video game CPUs using a heat gun.  Only recommended as an act of desperation.

3.  Tin whiskers: this is a relatively new failure mode and not well understood.  Some materials, particularly tin, sometimes engage in strange crystal growth in the form of tiny filaments -- called whiskers -- which can create shorts between PC board traces or IC pins.  Until recently, tin whiskers were rare because the lead (Pb, pronounced "led") in the solder made the tin behave itself.  However, nowadays you have RoHS requirements that restrict use of lead and tin whiskers are starting to become a serious problem.  For example, they were recently implicated as a cause for unexpected Toyota acceleration.  So would you prefer to die of lead poisoning from poorly-disposed-of electronics or in a car crash?  Would you prefer to have a moderate amount of somewhat toxic lead-based electronics in a landfill or huge quantities of less toxic new electronics in a landfill?

[monthos]

Burngate said:


One of the more common failures we had was electrolytic capacitors dieing.


Not sure if this is still a problem, but not too long ago the chinese manufacturing market was known to have had shipping capacitors which would fail in a short order of time. This was a problem across all hardware types, TV, PC monitors, computer motherboards/videocards/power supplies you name it, it was a problem with the solution in them and not any particular application.

That is something that actually worries me a bit, but the pi is a small low power device so it should have quite a bit less(honestly I never did not count them on the pictures), so coupled with this fact that I beleive that capacitor sourcing problem has been resolved I do not worry about it.

Quality built, low stress (This mostly means environmental) systems that are properly cared for will outlive their usefulness.

[abishur]

If it makes you feel better/brings you some relief, the components were very carefully chosen.  The r-pi is by no means a cheap Chinese build (no offense), I remember towards the beginning they were saying that Gert or Jamesh had been very obsessive when it came to the quality of the components.  I forget the exact time frame they expected on the r-pi so don't hold me to this, but I *believe* it was over 5 years.

[dh04000]

How to make the R-Pi last FOREVER!

Step1: Put R-Pi in desiccator(http://en.wikipedia.org/wiki/Desiccator) for a week or two to dry out the R-PI

Step2: On a dry day! Quickly spray polyethylene on the R-Pi with plugs in ports to avoid polyethylene getting in the usb, ethernet sd card slot, hdmi slot, ect (http://www.grainger.com/Graing.....-Gun-3WFT7)

Step3: Let the R-Pi dry out, plug it in and never, EVER unplug it! Use a backup power supply to make sure it never turns off. Also keep in dry environment, preferably with desiccating beads in an air tight container.

Step4: Profit as the R-Pi lasts for the next few hundred years.

[ArborealSeer]

John Beetem said:


2.  Ball grid arrays and other lead-less packages ("lead" as in "Leeds").  These are nasty packages from a reliability standpoint.


see : http://www.nvidiadefect.com/

18 months or so ago I had a 2 year old HP laptop with the BGA problem, didn't know the cause at the time so put it to one side and bought a shiny new vaio. Found about about the problem and paid a bloke in scotland with the right kit around 70 quid to fix it and add a heatsink. Result: Two working laptops

[monthos]

Please no disrespect, by no means am I stating that! Like I said I beleive that problem has been resolved in the manufacturing process for the most part. But also at the time it was not just a cheap chinese build problem as in cheap no-name companies, it was affecting many large names as well, hurting their own reputation which is why it gained some press and caused giving extended warranties and such.

That is why I brought it up, its food for thought, since the OP asked about how long it would last, its entirely reasonable to bring up past issues that major companies had regarding quality control.

Also stuff happens, the best manufacturing in the world can produce 99.99% perfect units, and that 1% can have an issue that is missed since it works well at testing and only fails a few months to a year down the line. In my case, I hope that its just a bad solder job somewhere I can see/work on it and I can get back to work, which has been the case for me at least some of the time.