Mornin' Michael and All -
Sounds as if you're using a CIC or similar core, so I'll make that
assumption and go on.
The following are lessons learned with two different designs: a pair of 12
layer polyimide boards on a .050" CIC core; and a pair of 14 layer polyimide
boards on a .06" CIC core with .006" CIC innerlayers. Both were in a
military avionics application. The assemblies were double side surface
mount.
In that day and age (early/mid 1980's) CIC .06" thick was roll annealed,
then flattened by the factory. Fabricated cores and thermal frames had a
bad habit of curling during the laminating process. Problem solved by the
fabricator reannealing to a special process provided by the material
supplier, prior to detail parts fabrication.
We encountered two bonding problems:
- Bond line ruptures during the soldering process - worked with the board
fabricator and laminates manufacturer to select an alternate material;
- Adhesive failures at the core surface - changed from nickel plate to
oxide treated copper in the bonding zone.
The board fabricator followed the laminate manufacturers guidelines for
temperature and pressure. This did include running some instrumented
specimen to confirm that the process was operating correctly. We also used
a bit of special fixturing and slightly more conformer material than normal.
Initial development boards produced some shorts to core, in spite of a
planned .005" minimum bond line. We resorted to filled vias in these
boards, with an overcoat of SR1000 solder mask. (I would probably choose a
different material if I were doing it today.)
The first units were soldered in a batch vapor phase, which was almost
impervious to the high thermal mass of the assembly, but getting the second
side soldered without dropping first side parts off was a real challenge.
(None of the chip bonders at that time were acceptable to our customer, for
various reasons.) We finally found an IR unit, by a now defunct
manufacturer, that would process the boards, and switched the process over.
I would think some of the newer IR's would handle these boards, with some
careful profiling.
The net result was that we qualified the designs and built 3-4000 each of
these assemblies - successfully. Process engineering costs were higher than
normal due to the problem solving efforts. First pass yields were lower due
to perceived imperfections in solder - grain texture from long liquification
periods as a result of their high thermal mass.
My approach to this design today would be to consider using a low expansion
laminate and minimum thickness foils to produce controlled low expansion
boards. These boards would be "soft" bonded to a thermal core after
assembly.
One last suggestion - once you have fabricated your first few board/core
assemblies, have some testing done at several locations on the assembly to
get a true reading of what the surface expansion characteristics are. We
found that they vary considerably dependent on the amount of underlying
copper.
Regards - Kelly
First, your soldering processes will impose a lot of shear stress on this
bond line due to dissimilar thermal expansion during soldering. The
presence of the core itself may try to do some bizzare things to your
soldering process, due to its very high thermal mass.
-----Original Message-----
From: Michael Becker <[log in to unmask]>
To: [log in to unmask] <[log in to unmask]>
Date: Thursday, November 04, 1999 2:48 PM
Subject: [TN] Bonding PWB's to Centerline Heatsinks
>I am planning on bonding Polyimide PWB's to centerline heatsinks using an
>Epoxy B-stage material. Does anyone have experience with the process used
>to do this. I am interested in Temperature, pressure, time information. I
>am planing on using an AlliedSignal product called A11.
>The purpose of this bondline is to constrain the PWB during temperature
>cycling. I am looking to get a CTE of less than 10 over the temperature of
>-55 to 125 C
>
>Any information would be helpful
>Thanks,
>
>Michael Becker
>Mechanical Engineer
>Raytheon Systems Company
>Baltimore, Maryland
>410-583-4036
>[log in to unmask]
>Instant Messanger: mbecker999
>
>##############################################################
>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
##############################################################
|
