Pratap and Marcus,
My answer to rampant or random failures in the field is long but it's the best answer I've heard thus far. It comes from a paper that I wrote in 2008 that was mostly based on my emails between my former physics professor and friend, Dr. Henning Leidecker at NASA Goddard Space Flight Center and myself.
It explains why whiskering seems to be a reasonable explanation for a random small sample of sudden acceleration events. There are so many millions of cellphones in use, were EMI from a cellphone the cause, you'd think we'd have a lot more incidents. By that same token, were it a software bug, we'd see more events as well considering the millions of vehicles sold. There's something particularly nasty about tin whiskers which makes them, in my mind, and, I think, in Henning's mind more likely the cause agent. Either that or a bad electrical connection caused by brittle fracture or perhaps an equally subtle intermetallic effect.
(the complete paper is at this link: http://www.ewh.ieee.org/r1/newsletter/2008S-TW.pdf)
To quote Mark Twain: "It's not what you don't know; it's the things you know, that are not so, that really get you."
No one yet understands how to predict the whiskering proclivities of a given tin coating. The distribution of lengths is close to log-normal, and it is the median value of length that grows at a rate of 0.5 mm to 1.0 mm per year. Leidecker has gotten these values from a number of different reports on experiments dating from the 1950s onward to 2005 (and later). When the tin coating does grow whiskers, and not all do, they may grow only minimal ones. [3]
Some whiskers grow faster, some slower. Surface compressive stress seems to play a role, and humidity definitely does. For every datum that is reported about tin whisker growth, it sometimes seems that one can find a report of a contradictory datum.
There is a general consensus of opinion amongst the scientific community that temperature cycling greatly promotes growth, especially cycling above and below the 13.2ºC phase-transition temperature of tin. Some find faster growth around room temperature. Leidecker suspects that new whisker growth depends on a cascade of several events, and that these have opposite temperature dependences, and different net impacts, under different circumstances. All other things being equal, they probably grow faster in warmer conditions [3].
Whisker containment is not perfect with conformal coatings, but it is very good. Parylene lasts a few years, and then a tin eruption blows out a divot of it. Elastomers stretch a bit, then crack, and tear. Containment depends in part on inducing Euler buckling. [7]
To complicate matters, not all whiskered surfaces cause circuit malfunctions. Size, and geometry can increase risk more than six orders of magnitude.
When more than about 100 mV is applied across the metal part of the whisker (i.e., after the tin oxide layer is dielectrically ruptured), then enough current will flow to melt the whisker open, usually within a millisecond or less. Sometimes, this current event is so brief that it escapes being logged as a fault. Other times, the event is able to "latch" an enduring fault (as in alarm circuits), and then the trouble-shooter has difficulty finding where the now opened whisker was before the event. [e.g. Toyota's problem]
Not all whisker-induced failures can be identified. Very few analysts correctly identify whisker-induced problems. A professional failure analysis can run between $300 and $3,000 per job. Almost no broken commercial equipment is ever put through any such analysis. Rather, the failed unit is junked or refurbished without any assignment of the fault. It is characteristically only equipment used in tasks of high importance that gets any analytic attention. Sadly, only a very few analysts are able to correctly recognize whisker induced problems!
Does commercial-grade equipment have this problem? It is typically only the military and space communities that carry out the analysis that is necessary to locate the source of the damage. And then, only a few of the analysts are perceptive as to the real cause.
Not all cases of whisker-induced failures are reported! NASA has logged, in 5 years, 3 to 5 reports per month of tin whisker infestation that required urgent help (almost all reports are from non-NASA sources). Very few manufacturers have allowed NASA to document the problems in detail, or share results publicly due to fear of lost sales, warranty claims, punitive damages, injuries, and embarrassment. There is no desire to share solutions to problems with competitors.
"The hundreds of cases we have documented scale to roughly a few million to a few hundred million cases of whiskering problems over the last fifty years --- this seems about right to me," stated NASA's Leidecker [3]. He suspects that about 3% to 30% of electronics systems that are using pure tin plating are growing whiskers, that about 0.5% to 5% of the total are having shorts caused by these whiskers, that about 0.005% to 0.5% of the total are having the cause of these shorts correctly identified, and then about 0.00001% to 0.01% of the total are being publicly named.
=====================
The above is an argument, admitedly not yet backed up by evidence, as to why Toyota's incidents of sudden acceleration MIGHT be so few in number and MIGHT be due to tin whiskering.
NASA has offered to help Toyota; I can think of no-one more qualified and impartial to do the work than NASA who, by the way, just updated their website.
The main updates include augmentation of the Photo Gallery and Anecdotes pages.
http://nepp.nasa.gov/whisker/photos/index.html and http://nepp.nasa.gov/whisker/anecdote/index.html
[note in particular anecdote #9 - how tiny whiskers caused major arcing damage to a 500V power bus]
Texts that teach newcomers about ways to make systems more reliable do not mention the dangers of whiskering as strongly as they should. A few allude to whiskering, usually as rare without distinguishing between rarely happening, and rarely publicly documented.
A typical company, selling parts with pure tin coatings that are occasionally causing a short, will continue this practice. They will promptly replace any one of their parts that the customer can show has shorted as a result of a whisker. And buyers of these parts will point to this prompt replace policy, and to the lack of a publicly documented problem with the use of pure tin coatings, to support the choice of purchasing these relatively inexpensive parts in favor of more expensive parts with whisker-free coatings. No one is charged with tracking injuries or deaths that result from this practice.
Aye, there's the rub.
Sincerely,
Bob Landman
H&L Instruments, LLC
-----Original Message-----
From: Leadfree [mailto:[log in to unmask]] On Behalf Of Marcus L. Thompson
Sent: Monday, March 01, 2010 3:19 PM
To: [log in to unmask]
Subject: Re: [LF] Expert recreates Toyota sudden acceleration
Hello Pratap.
I probably am ignorant of a salient issue here, but doesn't the failure mode which you mention generally manifest itself as partial- to whole-run? In that case, shouldn't we be seeing rampant (as opposed to
random) failures in the field?
Relatedly, is this type of failure included in the matters-of-study within the Manhattan Project bailiwick?
And, indeed, it is perhaps a valid point that IPC may have something valuable to add in this investigation. After all, it seems as though Dr. Leidecker has tossed his hat into the ring here...
--- Previous Message in This Thread ---
Subject: Re:[LF] Expert recreates Toyota sudden acceleration
From: Pratap Singh <[log in to unmask]>
To: [log in to unmask]
Date: Monday, March 01, 2010 11:21:58 AM
> Hi Werner,
>
>
>
> Thanks.
>
> The CAF and general metal migration in PCB has been described in
> details on pages 68, 86, 223 and 225 in Failure Modes and Mechanisms
> in Electronic Packages' co-authored by me and Viswam. The other terms
> that have been used by various authors are - dendritic growth,
> conductive migration, conductive corrosion migration, filament
> formation, Electromigration etc. These make and break mechanisms are
> exhibited by Ag, Au, Pb, Cu, and Sn commonly used in PCB assembly and
> packages. My comments were more to the point that in addition to Tin
> Whiskers (a LF issue), other mechanisms are available to cause a
> shorting failure. We all need well documented failure analysis of
> parts from a failed Toyota. Should IPC take a lead and provide a helping hand to Toyota to get to the bottom of these events?
>
>
>
>
>
> pratap singh
>
> ___________________________________
>
> Tel/Fax: 512-255-6820; Cell: 512-663-8903
>
> email: [log in to unmask]
>
>
>
>
>
> ______________________________________________________________________
> This email has been scanned by the MessageLabs Email Security System.
> For more information please contact helpdesk at x2960 or
> [log in to unmask]
> ______________________________________________________________________
>
> ----------------------------------------------------------------------
> ---------Leadfee Mail List provided as a service by IPC using LISTSERV
> 1.8d To unsubscribe, send a message to [log in to unmask] with following
> text in the BODY (NOT the subject field): SIGNOFF Leadfree To
> temporarily stop/(start) delivery of Leadree for vacation breaks send:
> SET Leadfree NOMAIL/(MAIL) Search previous postings at:
> http://listserv.ipc.org/archives Please visit IPC web site
> http://www.ipc.org/contentpage.asp?Pageid=4.3.16 for additional
> information, or contact Keach Sasamori at [log in to unmask] or
> 847-615-7100 ext.2815
> ----------------------------------------------------------------------
> ---------
> ----------------------------------------------------------------------
> --
>
> Hi Pratap,
> The copper migration' you describe has a name_'conductive anodic
> filament (CAF)'.
> Werner
>
>
>
>
>
>
>
>
> -----Original Message-----
> From: Pratap Singh <[log in to unmask]>
> To: [log in to unmask]
> Sent: Fri, Feb 26, 2010 8:21 pm
> Subject: Re: [LF] Expert recreates Toyota sudden acceleration
>
>
> Chuck,
>
>
>
> I addition to Tin Whiskers, other possible failure cause can be Copper
> Migration. This phenomenon also behaves like a whisker that is it will
> create a short, then burn out and grow again. Copper migration has
> occurred on PCB surface and internally (PTH to plane). I know it is
> not politically correct to say any thing against LF solder caused tin whisker thinking.
> There are few other mechanisms in PCB assembly that can cause internal
> surface shorts. Tin whisker is just one of the many.
>
>
>
>
>
> pratap singh
>
> ___________________________________
>
> Tel/Fax: 512-255-6820; Cell: 512-663-8903
>
> email: [log in to unmask]
>
>
>
> ______________________________________________________________________
> This email has been scanned by the MessageLabs Email Security System.
> For more information please contact helpdesk at x2960 or
> [log in to unmask]
> ______________________________________________________________________
>
> ----------------------------------------------------------------------
> ------
> ---Leadfee
> Mail List provided as a service by IPC using LISTSERV 1.8d To
> unsubscribe, send a message to [log in to unmask] with following text in
> the BODY (NOT the subject field): SIGNOFF Leadfree To temporarily
> stop/(start) delivery of Leadree for vacation breaks send:
> SET
> Leadfree NOMAIL/(MAIL)
> Search previous postings at: http://listserv.ipc.org/archives Please
> visit IPC web site http://www.ipc.org/contentpage.asp?Pageid=4.3.16
> for
> additional information, or contact Keach Sasamori at [log in to unmask] or
> 847-615-7100 ext.2815
> ----------------------------------------------------------------------
> ------
> ---
>
>
> Now that a plausible train of events has been demonstrated, i.e. a
> intermittent short circuit as seen in previous links to demonstration
> videos, perhaps it would be beneficial to ask what could cause a short
> circuit to exist that would clear itself when the circuit is subjected
> to an impact stress? Having seen long tin whiskers in person, and how
> they did indeed short out circuitry, I am willing to 'suspend disbelief'
> and accept tin whiskers as a possible cause of these failures once we
> have evidence of tin whiskers growing in the units, but not before.
>
>
> Chuck Patten, PMP
> Engineer Lead Project
>
>
> -----Original Message-----
> From: Leadfree [mailto:[log in to unmask]] On Behalf Of Marcus L.
> Thompson
> Sent: Friday, February 26, 2010 9:39 AM
> To: [log in to unmask]
> Subject: Re: [LF] Expert recreates Toyota sudden acceleration
>
> "Doing whatever we can to call attention to this possibility in the
> public eye is crucial."
>
> Because the public has essentially no knowledge of this new and
> particularly insidious failure mode for electronic systems (tin
> whiskers), it's important that we, as a community with superior
> knowledge of same, present the issue for examination.
>
> I support taking a tack at this forum which seeks to discuss and
> tangibly understand whether or not lead-free is a fitting vector for
> these failures; as any such discussions can only serve to deepen our
> understanding of the phenomena as a whole. I believe that we all can
> benefit from any dialog which further elucidates this issue as it
> affects real-world products and systems.
>
> As others have rightly said, silence is what brought the world
> lead-free
>
> in the first place. Shall we continue here???
>
> Just my 2 cents on a Friday morning --
>
>
> ...
>
>
> --- Previous Message in This Thread ---
> Subject: Re:Expert recreates Toyota sudden acceleration
> From: Pete Houwen <[log in to unmask]>
> To: [log in to unmask], Marcus Thompson <[log in to unmask]>
> Date: Friday, February 26, 2010 7:04:22 AM
>
>
>> callling attention to unfounded "possibiilities" will be damaging to
>>
> any attempts
>
>> to educate on the dangers of lead free solder. Crying wolf will only
>>
> destroy
>
>> credibility.
>>
>> There is NO data, evidence, indications, etc. whatsoever that the
>>
> Toyota
>
>> problems are in any way related to tin whiskers or lead free solder.
>>
> All
>
>> implications of electronic failure are EMI and software related. Not
>>
> even sure
>
>> why this is a topic of discussion on the LeadFree forum.
>>
>> Pete
>>
>>
>>
>>
>
> ______________________________________________________________________
> This email has been scanned by the MessageLabs Email Security System.
> For more information please contact helpdesk at x2960 or
> [log in to unmask]
> ______________________________________________________________________
>
> ----------------------------------------------------------------------
> -- -------Leadfee Mail List provided as a service by IPC using
> LISTSERV 1.8d To unsubscribe, send a message to [log in to unmask] with
> following text in the BODY (NOT the subject field): SIGNOFF Leadfree
> To temporarily stop/(start) delivery of Leadree for vacation breaks
> send: SET Leadfree NOMAIL/(MAIL)
> Search previous postings at: http://listserv.ipc.org/archives Please
> visit IPC web site
> http://www.ipc.org/contentpage.asp?Pageid=4.3.16 for additional
> information, or contact Keach Sasamori at [log in to unmask] or
> 847-615-7100
> ext.2815
> ----------------------------------------------------------------------
> --
> -------
>
> ______________________________________________________________________
> This email has been scanned by the MessageLabs Email Security System.
> For more information please contact helpdesk at x2960 or
> [log in to unmask]
> ______________________________________________________________________
>
> ----------------------------------------------------------------------
> ------ ---Leadfee Mail List provided as a service by IPC using
> LISTSERV 1.8d To unsubscribe, send a message to [log in to unmask] with
> following text in the BODY (NOT the subject field): SIGNOFF Leadfree
> To temporarily stop/(start) delivery of Leadree for vacation breaks send:
> SET Leadfree NOMAIL/(MAIL)
> Search previous postings at: http://listserv.ipc.org/archives Please
> visit IPC web site http://www.ipc.org/contentpage.asp?Pageid=4.3.16
> for additional information, or contact Keach Sasamori at
> [log in to unmask] or 847-615-7100 ext.2815
> ----------------------------------------------------------------------
> ------
> ---
>
> ----------------------------------------------------------------------
> ------
> ---Leadfee
> Mail List provided as a service by IPC using LISTSERV 1.8d To
> unsubscribe, send a message to [log in to unmask] with following text in
> the BODY (NOT the subject field): SIGNOFF Leadfree To temporarily
> stop/(start) delivery of Leadree for vacation breaks send:
> SET
> Leadfree NOMAIL/(MAIL)
> Search previous postings at: http://listserv.ipc.org/archives Please
> visit IPC web site http://www.ipc.org/contentpage.asp?Pageid=4.3.16
> for
> additional information, or contact Keach Sasamori at [log in to unmask] or
> 847-615-7100 ext.2815
> ----------------------------------------------------------------------
> ------
> ---
>
>
>
> ______________________________________________________________________
> This email has been scanned by the MessageLabs Email Security System.
> For more information please contact helpdesk at x2960 or
> [log in to unmask]
> ______________________________________________________________________
>
> ----------------------------------------------------------------------
> ------ ---Leadfee Mail List provided as a service by IPC using
> LISTSERV 1.8d To unsubscribe, send a message to [log in to unmask] with
> following text in the BODY (NOT the subject field): SIGNOFF Leadfree
> To temporarily stop/(start) delivery of Leadree for vacation breaks send:
> SET Leadfree NOMAIL/(MAIL)
> Search previous postings at: http://listserv.ipc.org/archives Please
> visit IPC web site http://www.ipc.org/contentpage.asp?Pageid=4.3.16
> for additional information, or contact Keach Sasamori at
> [log in to unmask] or 847-615-7100 ext.2815
> ----------------------------------------------------------------------
> ------
> ---
>
> ----------------------------------------------------------------------
> ---------Leadfee Mail List provided as a service by IPC using LISTSERV
> 1.8d To unsubscribe, send a message to [log in to unmask] with following
> text in the BODY (NOT the subject field): SIGNOFF Leadfree To
> temporarily stop/(start) delivery of Leadree for vacation breaks send:
> SET Leadfree NOMAIL/(MAIL) Search previous postings at:
> http://listserv.ipc.org/archives Please visit IPC web site
> http://www.ipc.org/contentpage.asp?Pageid=4.3.16 for additional
> information, or contact Keach Sasamori at [log in to unmask] or
> 847-615-7100 ext.2815
> ----------------------------------------------------------------------
> ---------
>
______________________________________________________________________
This email has been scanned by the MessageLabs Email Security System.
For more information please contact helpdesk at x2960 or [log in to unmask] ______________________________________________________________________
-------------------------------------------------------------------------------Leadfee Mail List provided as a service by IPC using LISTSERV 1.8d To unsubscribe, send a message to [log in to unmask] with following text in the BODY (NOT the subject field): SIGNOFF Leadfree To temporarily stop/(start) delivery of Leadree for vacation breaks send: SET Leadfree NOMAIL/(MAIL) Search previous postings at: http://listserv.ipc.org/archives Please visit IPC web site http://www.ipc.org/contentpage.asp?Pageid=4.3.16 for additional information, or contact Keach Sasamori at [log in to unmask] or 847-615-7100 ext.2815
-------------------------------------------------------------------------------
______________________________________________________________________
This email has been scanned by the MessageLabs Email Security System.
For more information please contact helpdesk at x2960 or [log in to unmask]
______________________________________________________________________
-------------------------------------------------------------------------------Leadfee Mail List provided as a service by IPC using LISTSERV 1.8d
To unsubscribe, send a message to [log in to unmask] with following text in
the BODY (NOT the subject field): SIGNOFF Leadfree
To temporarily stop/(start) delivery of Leadree for vacation breaks send: SET Leadfree NOMAIL/(MAIL)
Search previous postings at: http://listserv.ipc.org/archives
Please visit IPC web site http://www.ipc.org/contentpage.asp?Pageid=4.3.16 for additional information, or contact Keach Sasamori at [log in to unmask] or 847-615-7100 ext.2815
-------------------------------------------------------------------------------
|
