Hi, Ioan
I have had the "opportunity" to see several failures due to voiding in
previous jobs where I worked on military products. One of the most
memorable was a circuit board that was part of a device that was
launched; it was shot out of a gun.
To qualify the device, several were assembled, inspected, tested, etc.
very carefully. Upon the gunshot, the device failed. During failure
analysis several cracked solder joints were found on a particular SMT
1208 capacitor which was used several times on the assembly. One of
these that actually caused the electrical failure came out of its solder
joint on one end and caused an open circuit condition. A very intensive
analysis was undertaken to find the reason why this particular capacitor
part number's solder joints and no other solder joints on the assembly
cracked. There were many other similar SMT chip caps and resistors, as
well as much larger components. None displayed any stress related to the
gunshot. The failed capacitor that caused the open circuit was at a 45
degree angle with the board. The anode was still soldered to the pad,
but the cathode was up in the air, fully out of the solder joint. The
top half of the solder joint was still soldered to the end cap, the
remainder of the solder joint was still soldered to the pad.
While much analysis was being done, X-rays, SEM analysis to determine
impurities, XRF to determine robustness of the surface finish, etc. and
other attempts at root-cause determination, etc., a second unit was
built up and was gunshot. This time all of the processes had been
double-checked. Not only did the second assembly fail the same way, but
the same capacitor fully fractured on one end and caused the open
circuit again, in the same location on the assembly, and again it was
sticking up in the air exactly as it had done on the first assembly.
Talk about perfect defect replication! At this point I was called in to
the factory to see if I could determine what the problem was.
Upon examining the X-rays of the two failed assemblies, I noted two
things. The first was that there was a small solder void on every single
one of the capacitor's set of solder joints, the second was that it was
always on the cathode end of the capacitor, never the anode. When
looking at the end cap's configuration I was puzzled, as both appeared
to be the same. So I assumed that some type of outgassing of the
component occurred on the end that was voiding. This was not the case,
however. When I was finally able to get some loose parts to examine, no
outgassing was observed up to 230 C. in a special chamber. After the
outgassing test, I looked for evidence of venting under a microscope.
What I found was a very small .015" hole on the bottom of the cathode
termination. This hole had been overlooked by those who had looked at
the parts before me, and I did not see it when working on the failed
boards. It was not shown or noted on the component print, as it was felt
to be "insignificant" by the component engineer.
This hole provided a means of trapping air and flux during assembly and
was the cause of the voids seen only on the cathode end. These voids
were not large, perhaps between 15 and 25% of the flat area of the
solder joint formed between the termination and the pad. However, it was
enough to cause a solder joint fracture, and because this particular
failing component was near the center of this board, the board flex was
higher than those near the edges. When the capacitor was replaced with a
different vendor that had no hole, the boards all passed gunshot
testing.

I have also seen voiding cause solder fractures on interposed pin
components. When examining solder joints that underwent thermal cycling,
those solder joints with voids were much more likely to develop cracks
and fail electrically than those that had no voiding. Fixing the voiding
problem by changing to a different solder paste and or a slightly longer
time above liquidus temperature (TALT) fixed the cracks.

In a previous job, BGAs that had failed in the field and were returned
for failure analysis were found to have voids in the failed ball(s). In
nearly every case, if the BGA failed at all, it was noted that voids
were present. When we looked at good units returned from the field for
other problems, such as a cracked flex circuit, no voids were seen when
we examined the BGAs or they were extremely small. This assembly did not
see any abnormal physical stresses, and its normal service environment
was typically that of a laptop computer, it was essentially a field
computer for the military, so it saw occasional instances of extreme
heat and cold and some drop shock.

I know of several studies that have stated that voiding will help
prevent a crack from propagating completely through the solder joint as
it provides a stress relief mechanism, and several other papers that
take an opposing view.
I am no Werner Engelmaier, but one thing I have noted is that
microsections of solder joints from assemblies where some of the solder
joints show voids typically have a much coarser grain structure than
those that do not, and in that case all of the surrounding solder joints
also display this coarser structure, even in the solder joints without
voids. Solder joints microsectioned from assemblies where a different
solder paste was used that was formulated to resist voiding, and a
slightly longer TAL during the reflow profile was used, produced a much
tighter grain structure and no voiding. Whenever this was accomplished
there were few, if any, fractures.

That much I know for sure.

-----Original Message-----
From: TechNet [mailto:[log in to unmask]] On Behalf Of Tempea, Ioan
Sent: Monday, October 17, 2005 8:03 AM
To: [log in to unmask]
Subject: [TN] Voids in BGAs, again

Hi Technos,

I know this has been debated at least once a month, but I will not ask
what is the acceptability.

What I want to know is if anybody had the ocasion to actually see
failures related to voiding. How did you realize they were due to
voiding? Also, do you have any dirty pictures of the issue?

Thanks,
Ioan

---------------------------------------------------
Technet Mail List provided as a service by IPC using LISTSERV 1.8e 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 1.8e
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
-----------------------------------------------------