Good morning All -
Recommend your first steps be to isolate cause. Yes, entrained molecular
moisture is a possibility, but there are several others as well. This could
well be a materials problem, a materials handling problem, a laminate cure
problem, etc., etc.
Get back to your board supplier with the problem -- microsections of the
failed items will help isolate exact position within the laminate structure.
Chances are that the board fab house has preserved compliance test specimen
from those panels. TMA or DSC testing will go a long way toward identifying
whether entrained moisture is the problem. A low Tg reading is a good
indication of moisture (water or other) or incomplete cure of the resin.
The board supplier should also be able to help, from their experience base.
Baking boards isn't necessarily the wrong thing to do, but it costs time and
money, and for boards with a very thin tin/lead coat, may impair
solderability by increasing the intermetallic layer. I'm a strong believer
in buying boards that are ready to use upon receipt, even if that includes
having them packed in sealed vapor barrier bags by the fabricator. After
all, we dry pack plastic encapsulated microcircuits don't we? What's wrong
with drypacking boards?
If you have to bake, then a temperature around 105-115C for two to three
hours should be sufficient. Once again, it is well to support this time and
temperature by having DSC or TMA samples run on a few of the before and
after specimen to determine if you actually moved the Tg back up to that of
the original laminate suppliers data sheet.
Regards - Kelly
-----Original Message-----
From: Ramsey's <[log in to unmask]>
To: [log in to unmask] <[log in to unmask]>
Date: Wednesday, October 13, 1999 7:51 PM
Subject: [TN] Board Delamination . . . Boards too old?!
>Should I start out with the excuse? We put in this new ERP system. Y2K
>and all. We must have got something wrong, and fell asleep at the wheel.
>Any way, we don't make enough boards for a couple of months. Suddenly,
>we wake up to real demands (as opposed to ERP demands).
>
>We think (or feel) we need A LOT OF BOARDS FAST.
>We ask . . . no DEMAND! lots of boards on a four week turn to meet our
>"new" requirements. That was February.
>
>In late August, (there is no way we could have stuffed that many boards
>in two or three months) we started to see some board delamination, after
>ref low usually but sometimes after wave solder.
>
>These are four layer boards with mixed through hole and SMT on both
>sides. In our process, they see three temperature excursions, bottom
>side reflow, top side reflow and wave solder.
>Defect rate is about 1 in 50 boards. Delaminations are between layers
>one and two or layers three and four, they range is size from half an
>inch to over two inches. Big blisters the kind you can push up and down.
>
>The reflow profiles are a straight ramp at one to two degrees Celsius
>per second to reflow. The conveyor travels at 65 cm per min. Reflow for
>35 seconds. The boards see a maximum of 209 degrees Celsius, then cool
>at three to four degrees per second.
>
>Our stockroom is air conditioned but has no humidity control, worst case
>for a few weeks the boards saw 25 degrees C and 70% humidity. Most times
>the stockroom is 23 degrees and 50%. We are having no solder problems,
>we use an RMA paste 89% solder by weight.
>
>Our supplier, who has my sympathies, says that the boards must have
>absorbed moisture over the summer, and should be baked prior to use.
>
>These boards, when stuffed, are pretty expensive, containing eight 20
>mil pitch DSPs that run at 100 MHz. There is plenty of Flash RAM, PLD
>sockets, and glue logic. 96 pin DIN connectors . . .
>
>Will baking the boards prior to using them reduce our fall out? Should
>we scrap the boards? We have another 300 bare boards made by the same
>supplier only two weeks later (early March) . . . scrap them too?
>
>If Bake, how long should we bake? At what temperature? Do we need to
>ramp to that temperature?
>If Scrap, is our supplier in any way responsible for the defects?
>
>Gentlemen. Let's get ready to rumble. Oh, and uh, feel free to edit all
>but the relevant stuff out if you choose to quote this message in your
>reply.
>
>##############################################################
>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
>##############################################################
##############################################################
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
##############################################################
|
