I wrote a paper a few years ago in response to problems that IBM and
Seagate had run into in the early 90s. It was republished by Carlos
Barton of Rogers Corp in the 1997 IPC Flex Conference proceedings. In
our testing, we found that some types of adhesive (the competitor's of
course) acted as "solid state" electrolytes for electrolytic metal
migration in flex. We actually documented the dendrites growing along
the field lines independent of any interfaces where ionic contamination
might have been expected, effectively disproving IBM's theory about the
cause.

For materials prone to this problem the incubation period is very short
(a few tens of hours), and then the performance continues to degrade
slowly but steadily. Hint, look for adjacent dull/shiny traces as an
early indicator of anode/cathode cell establishment.

In some of our samples we discovered a fuse effect where the traces
briefly shorted out, and then the dendrite burned through. In these
areas the adhesive actually charred, and I suspect that under the right
conditions we could have started an adhesive fire.

Andy Magee
Flex Guru - Consulting
[log in to unmask]
(937) 435-3629

Glenn wrote,

Has anyone seen copper migration within a flex?

I have a test running where I believe this has occurred, but until I get
some
SEM and EDX work done, nobody seems to believe me.  The conditions are
just right for electrochemical migration, 85 C/85 %RH, 20 volt bias on
60-mil
space, ionic contaminate (maybe, residual etch chemistry?).  I see
dendrites
after 168 hours under test.  The traces are fully covered by the
polyimide
cover layers, but I know that the adhesive has an affinity for moisture.

Besides my calls to the fab house and material supplier, I thought I'd
ask the knowledge base here.  How about it?  Any experience you can
share?  I
have photos of the affected area (although not very good) that I can
share.

Thanks for your help,
Glenn

##############################################################
TechNet Mail List provided as a free service by IPC using LISTSERV 1.8c
##############################################################
To subscribe/unsubscribe, send a message to [log in to unmask] with following text in
the body:
To subscribe:   SUBSCRIBE TECHNET <your full name>
To unsubscribe:   SIGNOFF TECHNET
##############################################################
Please visit IPC web site (http://www.ipc.org/html/forum.htm) for additional
information.
If you need assistance - contact Gayatri Sardeshpande at [log in to unmask] or
847-509-9700 ext.5365
##############################################################