My query concerns BGA devices with Eutectic solder spheres.

As the spheres collapse during reflow the extent to which they spread
latteraly is controlled by the solder resist. Hence manufacturers recommend
resist defined lands. At PCB fab there is a tolerance on the accuracy of the
resist registration. This means that the diameter of the resist window must
be bigger than the land by at least 3 thou (mil)and probably 6 thou. If the
spread of the sphere is controlled by the resist then the SMOOTH SURFACE of
the solder sphere will come to rest on an EDGE (corner) of the resist.(The
sphere lying across the gap between land and resist). This edge is going to
make an indent or irregularity in the smooth surface of the sphere and is
said to be a potential weak point in the joint.
It would therefore be desirable to have a one to one resist window i.e. no
border around the land, with the resist and land being the same height. The
solder would spread until it touched the resist and not be deformed by it
since it was on the same level.

Is there any solution to his scenario?

Is one to one solder resist achievable?

Is this potential weakness just a necessary evil which we have to live with?

If the resist defined lands are created by overlapping the resist on top of
the copper pad there should be problems with accurate locating of the lands
and the same indent into the sphere will occur.

If the lands are defined to say 28 thou (mil) by the resist then this means
copper pads are 22 - 25 thou (mil). The general practice seems to be small
pads with reasonable sized resist window.
Are these small pads sufficient contact for long term reliability,
especially with modern high thermal output processors?

What about stencil aperture, is it enough to print paste on the 22 - 24
diameter pad?

What are the effects of printing paste at a wider diameter than the pad but
within the resist window?

Thanks in advance

##############################################################
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.
For the technical support contact Dmitriy Sklyar at [log in to unmask] or 847-509-9700 ext.311
##############################################################