You said it, David, when the worst X was better than the worst Y, that's why I'm still a little suspicious about using glues where you have several watts per chip: people seem to know less about the adhesive process than the soldering one. Soldering is so established since decades, first Philips radio was soldered successfully nearly one hundred years ago, but glueing is not so mature a process, I don't mean for the well educated, I mean for the general production people, if I may use that expression without beeing accused for depreciation. Adhesives can occur with large variations on inside, with non-curing local volumes, bubbles and such things, and if you want to minimize the Rth by making a very thin line, you may get other troubles with thermal inflexibility (most gluemakers recommend some 25-100 microns). And the Tg must be high if the chip gets up to +125C. I know some have checked the Rth up to Tg and found little Rth decrease, but I am concerned about chip joints !
that go from a hard state to becoming Wrigleys in operating mode. On the other hand, Mercury is said to conduct heat well despite it's floating. One may have to change foot some day!
Rounding up with telling you that we will abandon glue in a missile project and go back to soldering, soldering seems more reliable and comfortable. I removed some goodlooking (from outside) components just to find that there was nearly no adhesive wetting under the termination, glue wetting nicely all around the the periphery, judged as good, but I'm not satisfied with that. The explanation was that the components were so heavy that they pressed most of the glue aside, and you got at 2-5 micron joint, which showed wrong chemistry/silver flake constitution..and hence wrong wetting. I got no applauds for finding that, people don't like to hear such storys, but it's another one. I have had some very interesting debating with Mike, Joyce, Paul and Carey in the matter, and it will probably go on up and then. The glue makers themselves don't look on it the same way, if I may say so. The OH shows are somewhat different from the floor world. I still wonder where Remo Cirone is, ple!
ase mail back if you see this. He was at Alenia/Roma before. Space glue specialist.
Ingemar
-----Original Message-----
From: David Whalley [mailto:[log in to unmask]]
Sent: den 12 juni 2000 17:31
To: [log in to unmask]
Subject: Re: [TN] Chip backside metal (poor thermal conductivity of
polymers)
Hi Steve,
>Regarding use of conductive adhesives - while it does seem rather silly to
>attach (a generally high-power device) with poor thermal and electrical
>conductivity polymers, we have done it (rarely) for those occasions that did
>not require sustained current handling capability. I believe they were more
>of a Commercial application than Military, however. Definitely not Space or
>Implantable applications. At least the silver presents a more chemically
>bondable surface than does pure gold, for example.
>
>We even used some of the silver-glass materials for a Mil application, but
>the die were only ~3mm square. Die were mounted on moly-tabs (spreaders) to
>facilitate re-work, if required. The tabs were adhesively bonded with
>traditional Ablebond 84-1 series adhesive - now that the thermal capacity
>and 'contact' area was increased by the tab.
>
>Steve Creswick
While materials like 84-1LMI do have a thermal resistance of more than 10
times
that of typical solder alloys, we have found that this does not neccessarily
impact significantly on the overall thermal resistance as a thinner bond can
be acheived and the die bond thermal resistance is also usually only a small
part of the overall thermal resistance of path the heat takes from junction
to ambient. In some trials a few years ago we found the average thermal
resistance of a chip bonded with 84-1LMI was slightly worse than for SnPb
solder, but the spread for solder was greater, with the worst soldered sample
actually being worse than the worst adhesive bonded sample!
David Whalley
##############################################################
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 Keach Sasamori at [log in to unmask] or
847-509-9700 ext.5315
##############################################################
##############################################################
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 Keach Sasamori at [log in to unmask] or
847-509-9700 ext.5315
##############################################################
|
