We've think that with seen solder shorts under BGAs after wave soldering the boards. In addition, it seems to be related to a certain soldermask configuration over the topside BGA vias.. THere is one vendor,A, that is applying soldermask after HASL, tenting the top of the via. THe vias on the bare board are filled with solder and soldermask. THis board has exhibited solder shorts after wave soldering when the vias were exposed to the wave. The next lot from the same vendor had kapton tape applied to the bottom of the BGA vias and no shorts were seen after wave solder.
The other vendor,B, applies the soldermask onto the top of the BGA vias before HASL. The holes are not filled with solder at the bare board stage, unlike vendor A's boards. THese boards never have exhibited the solder shorting after wave.
By the way I ran a temp profile with a thermocouple mounted unter the BGA and was a peak temp of 150 C ritght when the part was over the Lambda pot.
Has anyone any experience with this problem? Any insight into what requirements they have for via soldermask requirements?
Cuppla questions for ya ... Where are the shorts [above the mask or between the mask and the laminate]? On the board, what material is between the copper traces and the mask? What type of solder mask(s) used by your suppliers? What are they using to plug the vias?
Ima guessin� that the issue is not when the solder mask is applied, but how it was applied. Explanations are: * Some fabs have a tough time getting good adhesion of masks in high density areas. * LPI solder mask tents poorly.
This is compounded by the filled barrels, which are ... * Less reliable than unfilled barrels. * A primed pump full of solder filled all the way to the top of the board.