Manish, I've seen similar phenomenon when processing CSPs on our cards. In our case, we were placing the components over the opening to in-pad blind vias. Due to poor drilling and plating processes by our board suppliers, volatiles were outgassing from the board substrate (through holes in the via wall plating) and blowing up the balls (pretty cool to see in cross-sections). Our suppliers have improved their processes since then.
If you're placing your BGA's over vias, check that the plating is fairly robust and consistent (full coverage of the walls) and that there is an alternate path for the volatiles to escape through.
I hadn't thought about it, but Joe's suggestion makes a lot of sense too (thanks Joe!).
| Hi Everybody | | | | | Is anyone aware of the possible causes that may lead to voids in the solder bumps after assembly. We assembled some dice on thin substrates and after assembly, void formation in the solder bumps were observed. | | | | | We have already tried baking the laminate and subjecting the laminate to reflow prior to assembly, but that does not seem to help. Does anyone has any insights on impurities in gold plating on the substrate affecting void formation. | | | | | | Any suggestions would be highly appreciated. | | | | | | Regards | | | Manish | | Manish, | I don't believe that the voids are coming from the laminate. Most of the time these voids may be caused by "overprinting" the pad and the amount of the pad covered by the component. I have seen this especially on BGA packages. What happens is that as the paste reflows the remaining, un-burnt off flux that is caught beneath the joint boils and "blows off" thus releasing trapped gas and causing visable voids during X-Ray. | | -Joe | |