Jan,
Actally, there are many solders within range of termperatures which you
specify. However, practically, I am not so sure I would recommend any to
you without knowing the requirements of your system.
You probably already knew that MOSFET manufacturers normally attach their
devices to copper slugs using high lead alloys such as Sn10Pb88Ag2,
Sn10Pb90, Sn5/Pb95 or other variations incorporating Indium as well. Check
data books or web sites for International Rectifier, XYIS, and others. Each
of these alloys has the benefit of being very ductile, absorbing the thermal
coefficient of expansion mis-match between the high expansion copper slug
and the low expansion silicon chip - allowing the device to survive many
power cycles. On the other hand, most of these alloys melt at 300+°C and
need to be reflowed in a hydrogen or reducing atmosphere. Thermal
conductivity is so so, but the joint isn't more than 25-50um thick, so it
becomes relatively negligible, if it is void free.
By going to any number of SnPb, SnAg, or SnPbAg alloys you will gain the
decreased liquidus temperature which you desire, but give up much of the
ductility of the traditional high lead alloys. However, since you are
currently using AuSn, you don't have a very ductile system in the first
place - the redeeming grace is that the composite thermal expansion of the
DBC substrate (~7-10ppm/°C - depending on foil to ceramic ratio) is much
closer to the silicon (2-3ppm/°C) than copper (17-18ppm/°C). Net result
after device attachment, the device is under compression - better than the
opposite!
If your substrate were aluminum nitride (4-5ppm/°C) the structure could be
even better. But, not knowing your design etc. this may not be suitable to
your application. I believe that Brush Wellman (and maybe Stellar) did
offer the effective equivalent of DBC on AlN.
In either case, if you choose to go to a Sn, Pb, or In bearing alloy go
lightly on the gold plating, unless it is required elsewhere for wire
bonding.
PbIn or other alloys. Daughter-boards alleviate some of these problems, but
may create others. There should always be a nickel plating layer between
the copper and the gold.
The traditional Pb/In alloys work well for this application, but it sure
helps to have hydrogen or reducing atmosphere reflow capability. Fluxes do
a great job of creating voids! Scientific Sealing Technology vacuum reflow
equipment also is a big benefit, but again, thou shalt not use flux!
Depending upon the circuit, in some fault conditions, the junctions can get
rather hot - hot enough to melt Sn63! Consider what safeguard you require
in the event of a fault?
Review your solder preform manufacturer's selection of alloys for one within
your temperature range. Consider your system requirements, and any
re-qualification liabilities. If you still wish to proceed, buy some ribbon
stock of the alloy/s you wish to evaluate. Cut them to size manually, and
attach some devices. Acoustically, or x-ray scan the interface for voiding.
Do some die shear and power cycling of the devices.
To conclude - even though the AuSn causes you repairablity problems, please
attack the problem with your eyes wide open. Sometimes, low temp solders in
situations like this can be too good to be true. Be careful!
Steve
At 11:11 AM 3/30/00 +0200, you wrote:
>Hi guys and girls!!
>
>I have a question that I think many of you can answer.
>Today we solder FET chips with Au/Sn solder onto a DBC substrate with Au
surface.
>The soldering area of the FET is Chromium/Nickel/Silver.
>
>I want to replace the Au/Sn solder with "something else".. I don't want to
glue the chip to place, only solder.
>The FET working temperature reaches up to 120degC.
>
>Any ideas on what kind of solder to use?
>__________________________________________________
>Jan Merstrand
>Microwave MCM, Design/Production Support
>Ericsson Microwave Systems AB. Defence Manufacturing
>Office address: Bergfotsgatan 2, SE-431 84 Mölndal
>Phone: +46 31 747 0725, Fax: +46 31 747 3515
>
>##############################################################
>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
##############################################################
|
