Hi Dave,
For once your insight has deserted you.
You are right, that the failure location is the same whether you have SMD or
NSMD. However, it is not the failure location per se, but the
time/stress-to-failure that determines whether there is a reliability impact.
And there is no question that SMD SJ fail first, provided of course that you
have good wetting.
Below s a solder joint where because of misregistration one side is SMD and
the other is NSMD. The consequences of the SMD-stress concentration are
obvious. Depending on the severity of loading, NSMD SJs will give fatigue
lives of 1.25 to 3 times longer than SMD SJs, everything else being equal.
[Unable to display image]
As I have indicated in my previous e-mail, I am in agreement with you
regarding the pad geometry.
On the other hand, it is not so much "the minimal cross-section area of the
solder joint is the real culprit," as is the difference between the
cross-section area near the interface relative to the cross-section area for
the rest of the ball. In other words, a column is always better than a ball
of equal height.
You write: "We have even attempted to force a stress riser generated crack in
BGA solder joints by "dimpling" the center of the solder balls on a BGAs
outer row and then
subjected them to thermal cycling. The solder joints still cracked at the
solderball/component pad interface!" Unfortunately, because of time/money
constraints people do "Highly Accelerated Tests." In many cases these tests
are not accelerating the dominant product damage mechanism(s), but create
earlier failures by making another damage mechanism or loading scheme
dominant. Also unfortunately, in most cases that does not result in a dif
ferent failure mode/location. Thus, improper conclusions are drawn regarding
product behavior that are not justified. Your observation is a case in point.
If you do T-cycling [forget about T-shock altogether, and be careful about
bend testig] at loading conditions similar to what actually happens with
product, you will come to different conclusions. I have run T-cycle studies
with more than 100,000 cycles than ran for over 3 years and where not
everything failed even than. Nowadays nobody can afford to do this--this was
only possible in the Bell Labs which ceased to exist around 1990. But having
to do things 'quick and dirty' [emphasis is on 'dirty'] should not encourage
the industry to draw unwarranted conclusions--thiis can be very dangerous.

Werner Engelmaier
Engelmaier Associates, L.C.
Electronic Packaging, Interconnection and Reliability Consulting
7 Jasmine Run
Ormond Beach, FL  32174  USA
Phone: 904-437-8747, Fax: 904-437-8737
E-mail: [log in to unmask], Website: www.engelmaier.com

---------------------------------------------------
Technet Mail List provided as a free 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/html/forum.htm for additional
information, or contact Keach Sasamori at [log in to unmask] or 847-509-9700 ext.5315
-----------------------------------------------------