Ouais! Quelques réponses:

Many substances can sublime, although not metals, as a general rule. I 
was thinking of the sublimation of flux residues. It has nothing to do 
with electromigration. It is the same as evaporation but directly from a 
solid phase to a vapour phase, without passing through a liquid phase. 
Some flux activators can sublime, even at ambient temperature.

For the anecdote, I have seen one truly catastrophic sublimation 
failure. It was a high-power, high voltage PCB which was soldered with a 
"no-clean" flux. As the assembly warmed up, the flux residues sublimed 
and condensed on big fuseholders, which must have been the coldest 
parts. Eventually, the build-up was sufficient that it flashed over from 
one to a neighbouring fuseholder and an arc started. As this was the 
unprotected side of the fuse (of course!), the current was sufficient 
that some 3 mm wide tracks of 70 µm copper simply fused and spattered 
molten copper all over the place, causing more high-power short-circuits 
in a chain reaction. I have never seen such a mess! Nothing to do with 
your case, of course!

I would not expect a dV/dt of <1E11 V/s to be significant.

Oxyde pulvérulent translates to powdery oxide, a not very scientific 
term if it can apply to any salt! The gelatinous nature probably means 
that it is a component of the flux (or, rather, the chemistry of the 
solder paste or flux. These may contain thixotropic thickeners to 
facilitate stencil printing without slumping; these would normally 
evaporate during soldering but there may have been an excess).

Having seen the later photos, I feel reasonably sure the green salts are 
probably copper compounds (possibly nickel, but less likely), suggesting 
a chemical attack. I doubt whether they are solder mask pigment/dye 
(I've seen this once on some PCBs where the mask was not fully 
polymerised and, on wave soldering, made a horrid green mess by reaction 
with hot flux).

En bon français, bon weekend :-)

Brian



Sylvain Kaufmann wrote:
> Hello technetters,
> Sorry for the subject but that's what came to my mind when I faced this case :) I never saw this previously, even when I burnt 2kW/chan amps in rock shows the result was not so catastrophic!
> 
> I would be very glad to have your thoughts about some pictures I sent to Steve.
> http://stevezeva.homestead.com/files/G_n__fonct_pignard_03_crop.JPG
> 
> http://stevezeva.homestead.com/files/G_n__fonct_pignard_closeup_16.jpg
> 
> Additional pictures has been sent to Steve this morning (relatively to my location).
> 
> Here is the case:
> One customer from us came up with one function generator with its case open; he opened it because it stopped working and had a bursty odor (smell?) and he wanted to see what was this odor and if he could change a fuse and let it work again. When he saw what was on the board, he came up to me to take pictures and have little explaination of what went wrong on this board. The customer bought this generator one year and a half ago, so it's still under guarantee, but he wanted to protect himself against a manufacturer's complain that could have said: "you did a short on the output, this repair doesn't fall into guarantee scope, so you pay for it" (now that he opened the case, maybe the guarantee is allready gone, but whatever). This generator has not been used a lot, understand it's not a generator that is installed in an automatic test setup that runs 24h a day.
> 
> The facts:
> - Presence of flux residues on both sides
> - Presence of several different copper oxydes (I don't know their english correct names, but the green ones, the green gelatin like ones, the red ones)
> - PCB's temperature increased up to at least 300°C or so
> - Generator was not working in a harsh environment; it was lying on a table in engineer lab, so, except if there was some fog from the open window near to the table, this generator never saw external humidity :)
> 
> My guess:
> - The manufacturer/assembler did a rework/touchup on a QFP on the top side of the PCBA; to perform it, the operator used a very active flux that he spread around the component he wanted to rework; some of flux went on the other side through the PCB vias.
> - Then the generator worked properly for about one year and a half, but during this time, some aggressive compound eat gently the copper barrel in several vias, leading to a increase of the barrel electric resistance, up to the time the resistance was so high that the dissipated power elevated the temperature to about 300°C.
> - Or this could be the result of electromigration and a hard short-circuit that happened on the board.
> - This part of the design has a control loop and power was first provided by control electronics up to the time this power could not be dissipated anymore and the PCB broke and so the generator stopped working
> 
> My concern:
> - As I see this happening without other factors than ones related to assembly (no humidity, no high voltage, no thermal shocks, no vibrations), I wonder what are the circumstances that lead to such destruction; is it only the flux chemistry? If I see some green gelatine like copper oxydes in some vias on one of our board, should I consider that this extreme end could happen to my product? Is it one isolated case (PCB destruction with risk of fire) or is it how it will allways end? Etc etc
> 
> Thank you for your inputs and have a nice Friday
> sly
> 
> S. Kaufmann
> 
> ************************************************************
> Sylvain Kaufmann
> CERN / European organisation for nuclear research
> Site de Meyrin
> 1211 Geneva 23
> Switzerland
> Dpt: TS-DEM-WS (int mail J06600)
> Phone: +41 22 767 37 02
> Fax: +41 22 766 87 77
> E-mail: [log in to unmask]
> Web: www.cern.ch
> http://cern.ch/dem
> ************************************************************
> 
> ---------------------------------------------------
> 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
-----------------------------------------------------