void on BGA Ball due to via on BGA pad

Subscribe to Thread

To whom this may concern, I recently ran into the problem wi... - Aug 15, 2002 by robertnguyen

We baked the board first and print 2 mil water soluble paste... - Aug 16, 2002 by Dason C

You have a very bad situation. It�s tough for the gas, flux... - Aug 16, 2002 by davef

robertnguyen

#21187

void on BGA Ball due to via on BGA pad | 15 August, 2002

To whom this may concern, I recently ran into the problem with vias on pads of BGA when place BGA and reflow gas trap on vias escape and result in ball having void greater than 50% of BGA diameter. The BGA pad is measured at 20 mils and the vias measure at 10 mils. These vias are drilled only from layer 1 to layer 2. Some vias can be seen with a microscope, other was capped or partially cap through a HASL process. So would anyone has any suggestion on how to resolve this issue would send me an e-mail I would be greatly appreciated.

Rgds,

Robert N

reply »

Dason C

#21199

void on BGA Ball due to via on BGA pad | 16 August, 2002

We baked the board first and print 2 mil water soluble paste at the BGA location and reflow. Wash and baked before the production run.

reply »

davef

#21200

void on BGA Ball due to via on BGA pad | 16 August, 2002

You have a very bad situation. It�s tough for the gas, flux material, er whatever to escape when the BGA is sitting on top of it and the blind via is blocking it from the other side. Obviously the vias should have been: * Placed on the edge of the pad and not near the center of the pad. * Plugged and plated over.

Since they weren�t, someone has to pay and apparently THAT someone is you.

It�s interesting that this board is HASL, given the high density design of the board that necessitated via in pad.

It�s tough to get real optimistic about this situation. Ordinarily, the approaches to reach a half-way tenable situation are: * Playing with your reflow profile [ie, reduce peak; soak will not help, providing it�s over 150 seconds; etc]. * Selecting a less voiding solder paste [ie, larger paste-powder sizes, decreased paste solvent volatility, lower paste metal content, etc]. * Using other good stuff on voiding talked about in the fine SMTnet Archives.

� but given this board is HASL, try fluxing and reflowing the bare boards to fill the via with solder. Then print, place, reflow, as in good practice �

reply »