Thanks for the input, John.
Your knowhow in this matter is unquestionable. I really appreciate that you gave our problem a minute.
(In fact, not our problem, I do consulting for another big company).
Inge
PS. It's Friday, and I feel for a little 'Maxwell' test: John, what is the basic equation for the magnetic flux in a electromagnetic coil? As son-son-son of famous Maxwell, you must keep the magnetic tradition....he-he

-----Original Message-----
From: TechNet [mailto:[log in to unmask]] On Behalf Of John Maxwell
Sent: den 12 september 2008 01:44
To: [log in to unmask]
Subject: Re: [TN] exploding cercaps

Inge,
I have included some comments added to your speculations based on my experience and represent my observations only.

John

>Hi all,
>I have solved numbers of cercap issues, some easy tasks, others very
>qualified. Now, in order to find the right track asap, I thought ; why
>not utilize TN expertise? Alzo, my question:
>
>What is the failure mode most likely to be?
>
>FR4 backplanes, 4 mm thick, lots of pressfit connectors. Transient
>protection by means of ceramic capacitors, size approx 8x6x1mm (LxWxT).
>SMT 60Sn. The catastrophic failures ALWAYS start at the end of the cap,
>from small microfumaroles to half ceramic body vaporized. From tiny
>spots of soot on the FR4 to something that reminds of what happens when
>you are careless with your gas torch welder!  The board is warped,
>about 2 mm along 200 mm. The board is torqued down flat with several
>bolts. May not seem to result in a lot of tension, but ceramic
>capacitors are very sensitive to axial/bend forces. I have cut out
>samples about 25x100 mm, populated with 4-8 caps, then polished them
>until only half thickness remains, then bent the samples slightly
>('simulate' torque down). Out of a hundred, some caps developed a
>obvious crack CLOSE to the end metallisation (Ag;Ni). The cracks are
>always in parallel with the end metallisation and across whole width of
>cap body. EDS on the solder joints gives that the solder surface is
>irregular and grainy, and contains small amounts of both Silver and
>Nickel. The end metallisation is unusually thick with a 60-80 um Silver
>layer. On outside of this is some 2-5 um Nickel. The meniscus is
>normal, which means very 'fat' in my opinion (most of us don't realise
>yet, the benefit with meagre fillets). The failure occurs
>stochastically in field use. Forgot to mention that the BaTiO is beige
>and it's a 100k MLC. One and the same manufacturer.
>
>My speculations are:
>
>A. Mechanical force from curved/straightened board is root cause

(John Maxwell) Mechanical/flexure stresses are the most common root cause of MLCC failure but these are failures on a 4mm thick board.
This related to what Werner stated in is response, large fillets lead to failure similar to mechanical/flexure cracks.

>B  Anomalous metallisation give local current issues which create
>hotspots and finally sparks etc

(John Maxwell) At best a third order effect

>C. Solder metallurgy faulty, or fatigued, giving too high serial
>resistance which gives local hotspots etc

(John Maxwell) See item B above

>D. Shorts or delaminations not likely, as the failues are always at the end.

(John Maxwell) I agree

>E. Bad Silver adhesion not very common

(John Maxwell) Cap loss is the failure mode and not burning caps.

>F. Nickel with bad pretinning = high resistivity or loss of galvanic
>contact not very likely either.

(John Maxwell) See item B above


>I know some of you have had many working days without sleep, because of
>puffing cercaps.

(John Maxwell) When I was a consultant I viewed puffing caps as my house payment, retirement and tuition for my kids to go to school. As a user I did loose sleep.


>Thanks in advance
>
>Inge

---------------------------------------------------
Technet Mail List provided as a service by IPC using LISTSERV 15.0 To unsubscribe, send a message to [log in to unmask] with following text in the BODY (NOT the subject field): SIGNOFF Technet To temporarily halt or (re-start) delivery of Technet send e-mail to [log in to unmask]: SET Technet NOMAIL or (MAIL) To receive ONE mailing per day of all the posts: send e-mail to [log in to unmask]: SET Technet Digest Search the archives of previous posts 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
-----------------------------------------------------

---------------------------------------------------
Technet Mail List provided as a service by IPC using LISTSERV 15.0
To unsubscribe, send a message to [log in to unmask] with following text in
the BODY (NOT the subject field): SIGNOFF Technet
To temporarily halt or (re-start) delivery of Technet send e-mail to [log in to unmask]: SET Technet NOMAIL or (MAIL)
To receive ONE mailing per day of all the posts: send e-mail to [log in to unmask]: SET Technet Digest
Search the archives of previous posts 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
-----------------------------------------------------