At 03:20 PM 6/18/2001 -0400, [log in to unmask] wrote:
Hi Steve
Nice Blisters!
Microsections could tell you at what interface the delam occurred and
that
might give you a clue.
Yes. Had to go through this several times recently. Lately
we've seen some problems with basic oxide coverage. There's a whole
host of possible causes. In the past I've seen delam problems due
to contaminates on innerlayer, undercured epoxy allowing chemical attack
from the holes, prepreg moisture/aging in storage, and several more
obscure failure modes.
I agree with Susan. Do a vertical section first to determine where
it failed (metal to dielectric bond or interdielectric
failure). If you determine that it's a dielectric to metal failure
you'll want to do a horizontal section to view it from that aspect.
You may have something as obvious as poor oxide . That's the
obvious and simplest failure mode to check out. If there's a
problem with the oxide you'll be able to see it through the FR4 as you
grind through the preceding layer. How the oxide looks will give
you some clues about the nature of the failure (whether it's chemical
attack in from holes or whether there's poor oxide coverage). If it's
chemical attack due to undercure or pinkring you'll want to do some cure
testing. If due to contaminates then off to a lab for analysis of
the failed spot.
If there's an interdielectric failure then you need to have an entirely
different conversation. But leave that worry for after doing a
vertical section -- likely you'll find the failed interface is on an
oxide to prepreg layer.
I would
include holes in the sections where you can.
Concur
Wedge voids can
lead to such blisters if moisture is trapped at the interface.
And also even if there is no moisture the wedge void may be an indicator
of chemical attack in from the drilled hole along a bond interface.
good luck,
Dwight