Folks,
I talked with several kind and knowledgeable folks since Friday. It is
obvious to me, as it must be to you, I have much to learn about CAF and its
direction.
I have condensed my replies into the following and am now ready to take my
medecine or just get shot at though hopefully missed:
xxxxxxxxx,
First let me thank you for inviting me into your worlds. That makes you a
big part of mine as I do understand the need for effective reliability
testing and appropriate test methods. You, Werner, and some other fine folks
have drilled that into me. I do understand your concern with various test
methods being considered. However, you understand mine as considering why
the stated phenomenon should be a concern.
I clearly understand electrical shorting, whether at/in hole walls or
through innerlaminar planes and MLB structures, is not desirable, to say the
least. My biggest problems relate to:
1) Drilling holes as small as those spaced on 10.5 mil centers and who does
that or would want to with other circuit alternatives available. Even at 25
mil spacing, these would become buried vias obviously not found on the
surface especially a BGA pattern as there are so many problems associated
there with this issue. However, I do understand possible use in the BGA
device substrates themselves.
2) I've never doubted there could be hollow glass fibers. I've always
believed the phenomenon to be very rare as 7 and 9 micron diameter
filaments, drawn from a molten silica bath, almost makes that impossible.
Also, these glass fibers should be drawn under extremely well managed vacuum
processes for Class 3, and other, applications where cost is not the primary
issue though extreme reliability, under extreme operations, is.
3) Hole "drilling" for such small diameter holes as mentioned in 1, above,
should not be considered using mechanical processes. Laser drilling of such
holes for micro/blind, micro/buried, and micro/interfacial can really only
be effected efficiently using laser or plasma methods, depending on
materials used.
4) As the Navy, as it always does and well much of the time, spends very
much time and money on such issues, I'm sure it must recognize the type
boards, no matter their class, should be considered only with MCM's or
device substrates (BGA?s as one example) that is a whole other world from
"regular" Class whatever types. Also, with the increasing ability to use
ASIC's with down to .18 micron line widths on fairly "standard" BGA ball
spacing (to .5 mm), we really don't need such a plethora of such small
diameter holes in such thin substrates no matter their construction ? as
paint thin dielectrics, etc.
5) As I asked on the TechNet forum, who really uses low resin content
materials, as 7628, anyway in anything over a really cheap four layer board?
Using 2113, and other higher percent to glass ratio materials, removes most
of the problem being discussed either at hole wall junctures or those inner-
plane.
6) The thing with extremely thin, unsupported "paint" or resin systems is an
ongoing issue as well it should be. One of the reasons we created rigid/flex
circuitry was to take advantage of very thin polymide film and adhesives in
conjunction with foils. I have built, as have others, purely rigid MLB's
using these materials and the CAF problem was never a problem, for obvious
reasons, though other problems presented themselves as they continue doing.
7) I am in no way against CAF reliability studies or it's impact, positive
or negative, on HDI systems that have few, if any, reliability data. It
possibly could aid in finding that data but I know not how. My concern is
spending too much time on that and not enough on finding better interconnect
technology where CAF would not be a factor. This point seems a target rich
environment to me as well as doing reliability studies associated with the
environment.
8) I understand how glass fracturing can and does occur especially in not
well managed drilling process conditions as poor feeds and speeds effecting
hole "punching" instead of cutting ? while admitting this process cannot be
managed perfectly anyway. Also, too wear can be a part of the poor process
management equation. Too many hits, etc. but still, these are nearly micro
holes and should not, most of the time, be mechanically drilled.
9) Micro delamination is micro-voiding most often except when micro-voids
propagate. IPC-6012, for "normal" board performance and qualification
requirements clearly indicates the "old time" requirements for void sizes in
the laminate evaluation area. It is concerned as will with performance
during thermal stress for qualification purposes. It also points to
pre-conditioning specimens before subjecting them to thermal stress and
shock conditions. This is analog to baking, or otherwise pre-heating, boards
before soldering operations while elevating and otherwise removing moisture
in the MLB structure anywhere, supposedly.
10) And this is far as I can go now is, again, all the thermal cycles boards
undergo during their fabrication and assembly cycles. FR-4, at about a .2%
hygroscopic level, is pretty "dry" anyway so I really have a hard time with
the CAF issue here. Plus, after so many fabrication, assembly, and
performance cycles ? over many years, doesn?t the salt and bromide issue
kind of dry up? Well, the bromide issue doesn?t as it is inherently part of
the resin system and is required to put out the fires. Polyimide has no
bromine or fire retardant problems.
Again, I am not being argumentative. Simply, I don't get it. However, folks
far more knowledgeable and astute than I have really dug into the dirt, not
being derogatory here, to find this stuff. My hat?s off to them for doing it
but what?s the alternative to more studies when the facts are in and we need
new materials and processes to turn them into initially acceptable quality
and long term reliability product.
Again, why aren't we finding alternatives to the processes and parts instead
of going down a path of finding CAF testing methodology. We have, you have,
the proof CAF exists. So, lets find a process or material, or whatever that
is CAF free, for lack of better terms. One thing of that I am sure is my
book has lot's of room for upgrading over the years to come.
I?m really not even trying to stir the pot here. I?m really just want to
learn why realizing I have a lot of it to do,
Earl
---------------------------------------------------------------------------------
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
---------------------------------------------------------------------------------
|
