0

Do electronic components have an expiration date?

How long does it take to components to go bad if we store them at room (normal) temperature and humidity?

What components go bad sooner? (resistor, capacitor, transistor, ...)

What are the best conditions for components storage?

Conditions of maintenance maybe are written in the datasheet, but is expiration date coming from the datasheet too?

Vladimir Cravero
  • 16,007
  • 2
  • 38
  • 71
musefian
  • 139
  • 2
  • 12
  • 3
    related: [Do ICs have a shelf life?](http://electronics.stackexchange.com/q/27499/7036) [What's the shelf life of ceramic & tantalum capacitors?](http://electronics.stackexchange.com/q/79931/7036) – Nick Alexeev Aug 15 '14 at 21:37
  • 1
    https://www.youtube.com/watch?v=U-3_2UtTMtA – Ignacio Vazquez-Abrams Aug 15 '14 at 22:01
  • 2
    Also related: [Should I discard my inventory of old electrolytic capacitors?](http://electronics.stackexchange.com/q/56474/2028) – JYelton Aug 15 '14 at 22:17

4 Answers4

2

Semiconductor parts I buy from Future sometimes come in hermetically-sealed antistatic bags, with silica gel inside.

Each come with a warning about humidity. (I will try to find one.) I believe the warning is parts should be used within six months once the package is opened.

I believe the issue is not about the silicon deteriorating. It is about moisture being absorbed by the package which may cause failure during the surface mount soldering process.

I have no reason to believe it is unique to the parts I buy from Future, and may apply to other parts with similar package properties.

gbulmer
  • 10,040
  • 20
  • 29
2

As well as the well discuss by others moisture ingress issues there can be solderability considerations.

Oxidation of the "pins" occurs over time - the rate depends on the finish used and base conductor material. Old components can become almost impossible to solder reliably with standard methods and may need special fluxes and different temperature profiles or even mechanical cleaning if value of the component and desperation levels are suitably high. The same applies to PCBs.

Russell McMahon
  • 147,325
  • 18
  • 210
  • 386
1

I think about the only component that I've had a problem expiring are electrolytic capacitors. The dielectric compound dries out so this changes or even kills the capacitance. Some times the expiration condition isn't a date. Flash memory often advertises a 100,000 write cycle as expiration. As far as a date? I don't think I've ever seen one, not to say there isn't one. My guess is that its far enough in the future the device will be obsolete. That being said I still have an ICOM radio from the 80's another from the 90's and use them regularly too. Today's electronics might not last as long though. I know that a lot of people were bummed when they took lead out of solder, but there is always a trade off; Right? I mean longevity Vs environment....

Kyle
  • 9
  • 1
0

There ARE expiration dates, but they're seldom printed on the devices. Occasionally you'll find a manufacture date on solid-state components.

The old standard in field engineering manuals used to be an expected service life of no more than ten years from original component manufacture until failure. Lots of exceptions, but this is EXPECTED service life.

Today's components generally have a shorter expected service life, because they're more heavily integrated.

The problem is that silicon in general (and the silicon compounds in specific that we use for solid-state electronics) is not a true crystalline substance - it's a PLASTIC substance. It flows or slumps, very slowly, over time. This can be witnessed easily by visiting any building with very old window glass - the glass is always thicker at the bottom than it is at the top.

In solid-state devices, the silicon can't slump very far before the component's ratings are compromised and the device fails. Much more so with greater integration, up to today's practical limits (which approach molecular transistor sizes). It takes VERY little slump in a five-micron transistor before the transistor fails under normal operating conditions.

And... it doesn't matter whether the device is under power or not - "slump never sleeps". Unpowered equipment lying on a shelf is just as likely to experience silicon-slump failure as well-designed equipment always powered up... which is why a "new old stock" replacement for an automotive computer is as likely to experience failure as the computer it's bought to replace.

ALL devices have SOME finite life, no question of that... especially if installed in poorly designed equipment and operated under conditions hostile to electronic devices. Passive devices (resistors, capacitors, inductors, etc), provided they're in well-designed equipment and operated well within their limitations, tend to last longer than active (silicon-based solid-state) devices, though, and gallium-based active devices tend to last better than silicon-based devices.

We'll all see some anecdotal responses here: "But I have a stereo that's thirty years old" and the like. Sure, there ARE exceptions. Sometimes we get lucky. How many of us are using computers older than about three years? Not many. How many 1990s vehicles do we see on the roads? We see them often in the junkyards, mechanically sound and sheet metal in good condition, but electronically not worth repairing. How old is your television? Your microwave oven? Your watch? Your camera? It all goes away, and every year the new equipment will go away earlier.

TDHofstetter
  • 1,292
  • 6
  • 12
  • 1
    Glass in old buildings is tapered because they didn't have float glass when it was constructed. – Ignacio Vazquez-Abrams Aug 15 '14 at 21:54
  • 1
    Oh, and I have several computers older than 3 years *and* a car from the 90s. *runs* – Ignacio Vazquez-Abrams Aug 15 '14 at 21:55
  • 1
    Two bits of that anecdotal evidence. 8) If the tapering is because of the lack of float glass, why is the thickest always at the BOTTOM instead of random places? – TDHofstetter Aug 15 '14 at 21:57
  • 1
    Because the thicker part holds the rest up better than the thinner part. – Ignacio Vazquez-Abrams Aug 15 '14 at 22:00
  • 1
    That's a weak argument - the glazier SURELY didn't intentionally cut that precious glass specifically for that reason and carefully measure it so it'd be installed "right side up". – TDHofstetter Aug 15 '14 at 22:06
  • @TDHofstetter,tanks for your reply."ALL devices have SOME finite life, no question of that.."true,but human life maybe very Shorter than That! – musefian Aug 15 '14 at 22:15
  • @TDHofstetter of course they did that. Imagine you are building a church and you've got this bunch of glass pieces: would you put the thicker end on the bottom or on the top? The glass-is-a-very-thick-fluid thing was known to be false *years* ago – Vladimir Cravero Aug 15 '14 at 22:40
  • [The Corning Museum of Glass, "Does Glass Flow?"](http://www.cmog.org/article/does-glass-flow) debunks the myth about glass flowing. – gbulmer Aug 15 '14 at 22:45
  • @TDHofstetter - would you please provide a reference which supports "... the silicon compounds in specific that we use for solid-state electronics) is not a true crystalline substance - it's a PLASTIC substance. It flows or slumps, "? I believe analogy with glass is invalid because it doesn't "slump" – gbulmer Aug 15 '14 at 22:50
  • Also, because the silicon used in sold-state electronics overwhelmingly *is* monocrysalline material. Amorphous silicon is used only in certain specialized areas. – The Photon Aug 15 '14 at 22:54
  • @ThePhoton - Yes, I apologise, I forgot to mention that. I was certain it had to be crystalline, or it is impossible to keep defects to a low enough level; the defects cause horrific behaviour. Which is partly why the material costs so much. So would it be fair to suggest that everything after paragraph two is either incorrect or not relevant? As an aside, I have six computers at home, and they are all significantly older than 3 years. The parts that have failed are the hard drives and batteries. – gbulmer Aug 15 '14 at 23:00
  • To be fair, doping *does* make the structure not-quite-crystalline; forcing other atoms in there is bound to shift the silicon out of alignment. – Ignacio Vazquez-Abrams Aug 16 '14 at 00:54
  • Interesting - I'd never run across that article in the Corning Museum of Glass before (little wonder, it's one piece among trillions). That may be true - I'm not in a position to prove otherwise. Even if my seemingly viable cited cause turns out to be untrue, the fact stands that solid-state silicon devices have a more finite life than passive or non-silicon devices... be that from growth of zinc/cadmium hairs, or from stray gamma/beta irradiation, or from any other cause. It all DIES, at a younger age now than forty years ago. – TDHofstetter Aug 16 '14 at 01:05
  • @gbulmer, all the parts about silicon "slump" are, AFAIK, nonsense. There is truth that as device size decreases and as number of transistors per chip increase, MTBF decreases. AFAIK, of widely-used parts, the worst for longevity are electrolytic (aluminum and tantalum) capacitors. – The Photon Aug 16 '14 at 02:49
  • @ThePhoton - I knew about aluminium electrolytic capacitors. I'll have to go read-up on Tantalum. – gbulmer Aug 16 '14 at 10:05
  • @TDHofstetter - I had believed that mechanical components, e.g. potentiometers, 'cheap' switches and 'cheap' connectors started to fail before the majority of other parts. Are you saying "solid-state silicon devices have a more finite life than" those? "It all DIES, at a younger age now than forty years ago." Is that a comment on the trend for "planned obsolescence", which AFAICT is a natural response in a market-driven economy, or something that has evidence when that factor is removed? – gbulmer Aug 16 '14 at 10:11
  • Oh, no doubts at all - poor mechanical devices, or mechanical devices selected without due regard to circuit requirements, will tend to fail very early on... switches & connectors before potentiometers (which can USUALLY be cleaned & kept in service). – TDHofstetter Aug 16 '14 at 14:12
  • As to the "planned obsolescence" aspect, I don't think that's nearly as common as it's made out to be. I think we see vastly more "neglectful obsolescence", from casual or thoughtless design, and the early-demise failures are never fixed because they don't HAVE to be - the public never demands it. The public is fast losing its interest in repairable/maintainable merchandise, preferring cheap disposable merchandise despite the high turnover rate. Or BECAUSE of the high turnover rate - the public is fickle, and loves "new" stuff with "added features". – TDHofstetter Aug 16 '14 at 14:17
  • Even if glass cold flowed (which AFAIK is a myth), there is a big difference: Unlike common glass (which has been used for applications like hard disk platters where the tiniest amount of cold flow would be a disaster), the materials making up the bulk of a silicon chip are crystalline and not amorphous - plastic deformation would require rupturing atomic bonds and not just working against viscosity... electromigration in an operating state, or slow chemical changes from impurities, CAN be a real problem however. – rackandboneman May 06 '15 at 22:39