My customer is seeing a problem with the LMZ12001 what appears to be related to the internal bond failures. Customer wants to know if TI will change this part to make it more tolerant of normal reflow process temperature curves.
Following is a description of the current layout:
This is a four layer .062 PCB, with the LMZ12001 on the top layer.
Under the LMZ12001 there is a 336 x 211 mil pad for the ground / thermal connection. Within this pad there are 8 16 mil holes as two columns of 4 holes. The next layer down is a ground plane, with no relief. All 8 vias are tied completely to this plane. The fourth layer is the bottom / 'solder' side. It has a 250 by 350 mil pad connected to the 8 vias. But not connected to anything else on that side of the boards.
Can someone validate if this looks good?
Customer is open to changing the drill file for this board. Instead of 8 rather small holes customer could either make these holes larger or use fewer holes and make them much bigger. Any recommendations on this?
The logic behind this might be that in the reflow soldering process the cool down to the LMZ12001 might be faster to the underside if there was more solder / metal under the part.
This seems like a question best answered by a TI process engineer. Could someone advise?
Customer failure rate is over 10%.
