Lead Free Soldier grows tin filaments which short out electronics.

[BrianHG]

I never knew, see here:
https://youtu.be/Q9zdt-rOB0Y?t=334
I guess military electronics are immune as they still use lead-tin solder.

[T3sl4co1l]

Tin whiskers are a problem which was blown way out of proportion -- I remember back in 2006 when people were panicking about RoHS.  It's over a decade later, and where's all the bricked cellphones?

Ah, well...

Figures that cellphones, and pretty much every other consumer electronic device, is bricked almost exclusively due to shitty software, doesn't it?  Or electrolytics, or batteries.

It's my understanding (but I haven't read about it in quite a while, mind) that lead-free is actually more reliable, when done correctly.  And it's worse, for this reason and others, when done poorly.  It's more of a quality differentiator among assemblers and their suppliers.  With a well tuned process, and precisely formulated alloys and fluxes, it's better; if any one of these is poor, the end product is poor.

If you expected to be able to hang on to that $50 television from five years ago, well... y'know?

Tim

[Gregg]

Tin whiskers are very real and certainly have the potential to damage electronics especially low voltage low current applications.  Zinc especially electroplated zinc rather than hot dipped galvanized metals exhibit similar metallic crystal growth from the surface.  Large data centers that use the area below a raised floor are prime candidates for the growth and spread of these whiskers after they break off of wherever they formed and are carried by the air conditioning air to things like servers and routers.  The cause of malfunctions due to these whiskers are likely to be overlooked because they are so tiny and if some piece of equipment malfunctions, a card or the whole thing is just replaced.  Conformal coating would go a long way to reduce the impact of tin or zinc whiskers, but even solder mask reduces the area that could be impacted.
Enclosed are a couple of pictures of tin plated bus bars in the main control cabinet of a Liebert brand 1500 KVA 480 volt 3 phase UPS system in a large data center in Silicon Valley.  The bus bus bars look fuzzy because they are covered with tin whiskers as long as 10mm; the bus bars are only 15mm apart between phases.  The bus bars were live at the time the pictures were taken and I couldn't get the light in a good place to really show up the whiskers.  These didn't cause any damage in this location probably because the potential was sufficient to vaporize and whiskers that touched and there was a lot of air movement to cool other components that would probably blow away any ionization caused by any whisker short circuits.  It was still pretty scary to see so many little tin filaments waving in the breeze with so little separation of 480 volts.  Shortly after the pictures were taken the center bus bar was wrapped with insulation during a maintenance shut down.   
 

[capt bullshot]

I don't think the whiskers at these bus bars are harmless.
I've seen arcing experiments with 400V high energy bus bars many years ago, they used a tiny copper wire to start the arc, which would have been very violent then it the power wasn't cut off within milliseconds. They were testing arc detection system based on fiber optics.
BTW - I've never seen tinned bus bars before, usually they are just made of copper with no plating.

[IanMacdonald]

The AF - numbered germanium transistors used to suffer this. They frequently used to fail for no apparent reason. Was a long time before the actual cause was found, which was apparently tin plating of the case.

[G7PSK]

I have come accros laptops that were malfunctioning and after brushing with a stif bristle brush worked correctly, no sign of dust but I could see tin whiskers on the board with a low power magnifying glass before brushing them off. 

[Mr. Scram]

I too suspect that many more failures are actually due to this. A phone does and people discard it, without ever realizing this was the problem. I see phones die all the time. Saying it's not a problem seems optimistic.

[kalel]

I am curious of the common failure mode of modern smartphones.

I always assumed it was moisture and physical damage, but how long do they last electronically and what fails there (if something does) when used regularly for many years.

[nctnico]

Tin whiskers are a problem which was blown way out of proportion -- I remember back in 2006 when people were panicking about RoHS.  It's over a decade later, and where's all the bricked cellphones?

I agree. I have seen tin whiskers once on a hand soldered board but after the alloys got sorted out the problem has dissapeared.

[R_Gtx]

For all matters pertaining to tin and similar metallic whiskers, the following NASA site is invaluable:

https://nepp.nasa.gov/whisker/

[NiHaoMike]

I am curious of the common failure mode of modern smartphones.

I always assumed it was moisture and physical damage, but how long do they last electronically and what fails there (if something does) when used regularly for many years.

Having used cheap smartphones and tablets for cryptocurrency mining for over 2 years, the most common failure modes are either the cheap Flash memory fails/corrupts or a voltage regulator fails. Of course, mining is a very stressful load. For phones that are actually used as phones, I'd imagine accidental damage would be the most common failure mode.

[tggzzz]

One example of whiskers found on a DMM's frame: https://www.flickr.com/photos/anachrocomputer/36125222306/

For all matters pertaining to tin and similar metallic whiskers, the following NASA site is invaluable:
https://nepp.nasa.gov/whisker/

Yes, especially http://nepp.nasa.gov/whisker/reference/tech_papers/2012-Panashchenko-IPC-Art-of-Metal-Whisker-Appreciation.pdf

[coppice]

Zinc especially electroplated zinc rather than hot dipped galvanized metals exhibit similar metallic crystal growth from the surface.

I don't know what factors stimulate the growth of zinc whiskers, but I've seen large equipment where most of the zinc coats parts look fine, but a few have an incredible network of fine whiskers stretching several centimetres through the air. There can be such a mass of super fine whiskers that it looks like fungal growth at first sight.

[coppice]

Tin whiskers are a problem which was blown way out of proportion -- I remember back in 2006 when people were panicking about RoHS.  It's over a decade later, and where's all the bricked cellphones?

People were panicking because the early adopters were already seeing whisker problems after just a few months. This wasn't the first time serious problems with tin whiskers had occurred. It caused a lot of problems in the 1960s. I guess metallurgists in the 2000s had forgotten the lessons from that era.

It's my understanding (but I haven't read about it in quite a while, mind) that lead-free is actually more reliable, when done correctly.  And it's worse, for this reason and others, when done poorly.  It's more of a quality differentiator among assemblers and their suppliers.  With a well tuned process, and precisely formulated alloys and fluxes, it's better; if any one of these is poor, the end product is poor.

I'm not sure what was done, but the problem seemed to become much less of an issue after a couple of years. Maybe the metallurgists relearned 1960s skills, and modified the alloys which are used.

[Ian.M]

There's another wonderful failure mode for solder alloys that are nearly all Tin.  Prolonged exposure to low temperatures can cause a phase transition into brittle, non-metallic, alpha-form grey tin, with a 27% volume increase which will cause it to flake off or crumble into a powder!

See http://www.indium.com/blog/tin-pest-still-a-forgotten-concern-in-lead-free-assembly.php

Edit: Wikipeda has the alpha and beta allotropes confused in their tin-pest article.

[BrianHG]

There's another wonderful failure mode for solder alloys that are nearly all Tin.  Prolonged exposure to low temperatures can cause a phase transition into brittle, non-metallic, beta-form grey tin, with a 27% volume increase which will cause it to flake off or crumble into a powder!

See http://www.indium.com/blog/tin-pest-still-a-forgotten-concern-in-lead-free-assembly.php

Watch the beginning of the video in my opening post.  You will see a video clip of this happening....  It is cool....
https://youtu.be/Q9zdt-rOB0Y?t=90

[chris_leyson]

I've only ever seen one tin filiament or whisker. Had to find out out why a hermetically sealed relay in some medical equipment had failed. Cut the top and found a huge tin whisker growing out of a solder joint, it was several mm long and you could see it with the naked eye if you got the illumination just right.

Ian.M posted an interesting link about Tin alloys at low temperatures but I've seen unusual things happen at higher than nornal temperatures. Had a poorly designed legacy PCB with a SOT89 transistor that ran so hot it would eventually fail short circuit, I can only guess the package temperature but it must have been at least 200C or higher and sitting there for quite some time judging from the delaminated pads and underlying FR4. Anyway, in some cases, try as I might I could not get the solder to melt even with the iron set to 400C. For some reason the melting point had shot through the roof. A bit off topic but it may be related to Tin alloy solders, don't remember seeing a similar problem with leaded solder.

[Ian.M]

That's quite common - copper diffuses into the solder, altering the composition and rapidly raising the meting point once 0.9% Cu is exceeded in the SnCu system.

More interesting would be the potential side-effects of the tin pest transformation.  I believe grey tin is a semiconductor, and that would give rise to the possibility of a Schottky diode junction forming in a joint degraded by tin pest. 

[jimdeane]

Tin whiskers are a problem which was blown way out of proportion -- I remember back in 2006 when people were panicking about RoHS.  It's over a decade later, and where's all the bricked cellphones?

My Nokias from the 1990s and early 2000s still work just fine when I want to plug them in (batteries are weak) and play Snake.

My recent smartphones have had a useable life (treated extremely well) of 15-18 months. I never even exceed the lifespan of the battery any more. I'm really hoping my S8+ lasts longer than that, but I'm not betting my life on it.

I can't say for sure it's tin whiskers, but I suspect that or BGA failure, either of which would not be nearly as likely with real solder.

[ajb]

I can't say for sure it's tin whiskers, but I suspect that or BGA failure, either of which would not be nearly as likely with real solder.

Anecdotally, bga failures seem to be a pretty big issue.  There was the XBox Red Ring of Death some years ago, that was apparently due to BGA failures (hence some people "fixed" their systems by deliberately overheating them or baking them in ovens to partly reflow the problem joints).  My partner and I both had our phones (one a nexus 5X, by LG, the other a Nexus 6P by Huawei) fail within the same week recently.  Apparently both phones are known to be susceptible to assembly failures, although I don't know if it's clear that these are failures occur between parts of the SOC or between the package and the board.  When consumer electronics push the limits of mass manufacturing there are bound to be issues.  On the other hand, the sheer volume of production is such that even  rare problems can wind up happening thousands of times, so it's difficult to draw firm conclusions without access to the actual production and failure numbers.

[Mr. Scram]

Anecdotally, bga failures seem to be a pretty big issue.  There was the XBox Red Ring of Death some years ago, that was apparently due to BGA failures (hence some people "fixed" their systems by deliberately overheating them or baking them in ovens to partly reflow the problem joints).  My partner and I both had our phones (one a nexus 5X, by LG, the other a Nexus 6P by Huawei) fail within the same week recently.  Apparently both phones are known to be susceptible to assembly failures, although I don't know if it's clear that these are failures occur between parts of the SOC or between the package and the board.  When consumer electronics push the limits of mass manufacturing there are bound to be issues.  On the other hand, the sheer volume of production is such that even  rare problems can wind up happening thousands of times, so it's difficult to draw firm conclusions without access to the actual production and failure numbers.

BGA failures could well be caused or exacerbated by the solder used. Less flexible or more brittle solder will sooner lead to cracking.

[wraper]

I can't say for sure it's tin whiskers, but I suspect that or BGA failure, either of which would not be nearly as likely with real solder.

Anecdotally, bga failures seem to be a pretty big issue.  There was the XBox Red Ring of Death some years ago, that was apparently due to BGA failures (hence some people "fixed" their systems by deliberately overheating them or baking them in ovens to partly reflow the problem joints).

In vast majority of cases this is chip internal problem with bumps under the die and heating temporarily rises it from the grave.