RE>New BGA Via Soldering                     12/11/95

Bob,

Some very good points regarding Via in Pads for BGAs.
 
You are quite right about the problems seen with secondary reflow of topside BGA joints during wave solder processing.  The via array on the bottom of our boards gets masked with Kapton tape for protection against this.

Like you, we've seen problems with HASL'ed boards experiencing the stencil being lifted locally off the pad surface during pasting of the second side.  Just one more reason why we went to Bare Copper!

The use of "low" or "high" temperature balls is dictated more by reliability rather than by processing concerns.  High temperature balls will allow the device to have greater standoff, which improves the compliancy of the solder joints.  This is important to increase resistance toward fatigue mechanisms, which are much worse with device substrates which expand at different rates relative to the PWB.  CBGAs and TBGAs need this higher standoff as a result, but PBGAs don't -- therefore, they can use the low temperature (i.e., eutectic) solder balls.

Heat dissipation isn't that much of an issue with us because our devices don't generate a lot of heat.  Still, I'd be interested to see either data or calculations which shows the differences between the two approaches  
 
I've never heard of using a solder fountain to remove BGAs.   Very interesting approach.

Finally:  thanks very much for your kind words on our products.

Greg Bartlett
Mercury Computer Systems
Chelmsford, MA