Hi Steve
I'm sure you find a good lab close to you. Finding the short is no big deal with an infrared camera. One has only to be careful about applying the power. It's best done with a curve tracer used for semiconductor analysis because with that you can easily control voltage and current. 
If the short is not visible from the top of the PCB because its too deep, a microsection (a little bit away from the via, getting closer to the signal step by step), leaving the contacts to the via intact, and proceeding as above will show the conductive path. The rest is standard FA.


Have a nice day

Günter

Günter Grossmann
Scientist
Empa
Swiss Federal Laboratories for Materials Science and Technology
Überlandstrasse 129
8600 Dübendorf
Switzerland
Tel +41 58 765  4279
Fax +41 58 765 6954
[log in to unmask]
www.empa.ch

Diese Mitteilung ist ausschliesslich für die als Empfänger aufgeführte Person(en) oder Firma bestimmt. Sie kann persönliche und/oder vertrauliche Informationen enthalten. Sollten Sie diese Mitteilung irrtümlicherweise erhalten haben, bitten wir Sie, uns umgehend per E-Mail zu informieren, sämtliche Ausdrucke zu vernichten und die E-Mail-Dateien zu löschen.
This message is intended for the exclusive use of the individual(s) or entity to whom it is addressed. It may contain information that is privileged and/or confidential. Should you have received this message erroneously, please notify us immediately by e-mail, destroy any printouts and delete all electronic files contained in this message.

> -----Ursprüngliche Nachricht-----
> Von: TechNet [mailto:[log in to unmask]] Im Auftrag von Steve Gregory
> Gesendet: Sonntag, 10. April 2016 02:11
> An: [log in to unmask]
> Betreff: Re: [TN] Recommendations for Failure Analysis / Root Cause of PCBA
> Failure...
> 
> Hi George,
> 
> Those little young-un's can be a handful huh? I'm sure you're doing your job
> as all grandparents do and spoiling him to death (GRIN). Glad you're enjoying
> him!
> 
> So I just want to give a little more information about the board. The was
> returned to us conformal coated, so what you may have mistaken as flux
> residues is actually conformal coating. The board finish is actually Sn/Pb HASL.
> We assemble the SMT with a Sn62/Pb36/Ag2 water soluble paste and is
> cleaned in a DI inline cleaner after reflow. The through-hole is stuffed and
> wave soldered, again using a water soluble flux. It is again cleaned in the
> inline cleaner. The board is inspected at the end of the reflow oven, and
> after wave solder, as well as going through AOI. Any touch-up that is done
> requires that the board go through the inline cleaner again, so I'm confident
> that the board is clean when we ship it to our customer.
> Unfortunately we don't electrically test the assembly, the test set-up is a
> custom deal that our customer doesn't want to part with, so they do the
> electrical test and conformal coat the assemblies.
> 
> The red arrows were placed by our customer, and the probe marks that you
> see in the solder joints were done by what I assume were their test
> technicians with a DMM to isolate the fault. What isn't readily apparent in the
> photos is a slight darkening around the via, and you can see that the via looks
> like it blew a solder ball out that you see on the annular pad If you look
> carefully in the photos you can see tiny solderballs down inside the via, and if
> you you look at the via pad surface and compare the other via pad surfaces in
> the photo, you can see that the arrowed via did over-heat.
> 
> The other assembly came to us with several IC's on the board with lifted pins
> which I assume were done trying to track down where the short was.
> This board also had a blown tantalum capacitor (which I didn't post a photo
> of), but the capacitor was installed with the correct polarity as you could still
> see the polarity stripe on the cap and see that it was installed correctly. So
> that was evidence in itself that voltage came from the wrong direction to that
> cap...
> 
> So I've been reading everything I can find about CAF, and trying to
> understand the problem better. I think that probably that's what is
> happening given that it's taking time for the failures to occur.
> 
> Steve
> 
> On Fri, Apr 8, 2016 at 5:34 PM, George Wenger <[log in to unmask]>
> wrote:
> 
> > Hi Steve,
> > I haven't been on TN for some time (my 3 year old grandson has been
> > taking up much of my time).  Greg Munie had forwarded your TN post and
> > Dave Hillman's reply and I sent an email repl to Greg that he
> > suggested I post on TN so here it is:
> >
> > From: "Greg Munie" <[log in to unmask]>
> > To: [log in to unmask]
> > Sent: Friday, April 8, 2016 12:39:07 PM
> > Subject: FW: [TN] Recommendations for Failure Analysis / Root Cause of
> > PCBA Failure...
> >
> >
> >
> > Dis looks like a job for George . .
> >
> > From: [log in to unmask] [mailto:[log in to unmask]]
> > Sent: Friday, April 08, 2016 3:20 PM
> > To: Greg Munie <[log in to unmask]>
> > Subject: Re: [TN] Recommendations for Failure Analysis / Root Cause of
> > PCBA Failure...
> >
> >
> >
> > Dis would be a job for George if George still worked in a Lab like he
> > had in Lucent or Andrew but doesn't have that now.  I believe that
> > Dave Hillman is correct that with two PCBAs with shorts there are many
> > labs that could most likely find the root cause for the failure.  I
> > didn't see Steve's resopnse to David's question about if Steve
> > believes the problem to be an internal short.  I know what I would do
> > if I were doing the failure analysis and the first thing would be to
> > ask Steve many more questions about the PCBA and when the failures are
> > detected before I did any physical analysis of the two PCBAs because
> > the analysis is destructive and if you don't find the root cuse you're
> > out of luck
> >
> > After reading Steve's email and looking at the two photos a couple of
> > things come to mind.  The short could be an internal PCB short or it
> > could be an external PCBA short.  From the photograph the surface
> > finish on the PCB appears to be ENIG and it also appears that there
> > isn't a cleaning process done after PCBA solder assembly process.  My
> > first questions would be what solder paste is used for the assembly?,
> > Is the PCBA a double side surface mount reflow assembly?, is there a
> > wave soldering assembly performed and if there is what flux is used?
> > I'd also ask for PCB layer drawings and an electrical circuit diagram.
> > I would do a high resolution scan of the top and bottom sides of both
> > of the failed PCBAs and inspect both failed assemblies in detail under
> > a microscope. Before doing any cleaning of the PCBAs or cross sections
> > of the area that is shorted I would systematically begin removing
> > surface mount components that are soldered across the 12V power. I'd
> > carefully examine the bottom of each removed component and the PCB
> > surface from which it was removed for any debris, dendrites, or shorts
> > and also measure the removed component to see if it is shorted as well as
> re-measure the via to see if the PCBA is still shorted.
> > An alternative to doing doing a component by compnent analysis is to
> > do an agressive cleaning and drying process on the PCBA and then
> > measure the via to see if the short is still present.  The clening
> > process I would us is to do ultrasonic cleaning in a terpine cleaner
> > like EC-7 followed b ultrasonic cleaning in IPA to remove any terpine
> > residue and then do a long (4 to 8
> > hour) bake at 100C to drive off any moisture that might be traped
> > under components or in component flexure cracks and then check the via
> > to see if it is still shorted.  If it isn't then the problem is an
> > external short rather than an internal PCB short.
> >
> >
> >
> > Since the failures are occuring after 111 to 472 days field exposure I
> > believe there is a strong possibility that the shorts are external and
> > are occuring under or within components rather than internal to the PCB.
> >
> >
> >
> > If the cleaning and component removal doesn't eleminate the short than
> > I would do careful transmission xray imaging around the via that is shorted.
> > Actually I would do this x-ray imaging of the shorted via before doing
> > the cleaning and componet removal just to have the images.  The x-rays
> > may not be able to see if there is a short but I would still do them.
> > If the x-rays son't show anything and the via is still shorted I think
> > I would begin a very slow and careful lapsectioning of the via and
> > examine each of the PCB layers for shorts as the laping progresses through
> the PCB.
> >
> >
> >
> > Sorry for the long winded answer but I hope it helps Steve find the
> > root cause for his problem.  Feel free to share my response with Steve
> > or Dave or any lab Steve might work with.  I haven't made any
> > recommendations for an external lab because there are so many.  There
> > are deffinately some labs out there that I wouldn't use because I'm
> > not sure they really have the capabilityy and all they want is your
> > money.  The one lab I've actually never used but I know their
> > capability and would certainly reccomend is George Westby's Universal
> > Instruments Lab in binghamton, NY.  Although the are really good the
> > one drawback is they are expensive.  I think their minimum is $1000.
> >
> > Regards,
> >
> >
> >
> > George
> >
> >
> >
> > George M. Wenger
> > Failure Signature & Characterization Lab LLC
> > 609 Cokesbury Road, High Bridge, NJ 08829
> > (908) 638-8771 (Home) (732)-309-8964 (Cell) [log in to unmask]
> >
> 
> 
> __________________________________________________________
> ____________
> This email has been scanned by the Symantec Email Security.cloud service.
> For more information please contact helpdesk at x2960 or [log in to unmask]
> __________________________________________________________
> ____________
> --
> 
> 
> This email and any attachments are only for use by the intended
> recipient(s) and may contain legally privileged, confidential, proprietary
> or otherwise private information. Any unauthorized use, reproduction,
> dissemination, distribution or other disclosure of the contents of this
> e-mail or its attachments is strictly prohibited. If you have received this
> email in error, please notify the sender immediately and delete the
> original.

______________________________________________________________________
This email has been scanned by the Symantec Email Security.cloud service.
For more information please contact helpdesk at x2960 or [log in to unmask] 
______________________________________________________________________