Issue link: https://iconnect007.uberflip.com/i/895789
68 SMT Magazine • November 2017 This experiment studied 10x10 mm sq., 15x15 mm sq., 17x17 mm sq., 19x19 mm sq. and 37x37 mm sq. package body sizes, and only the 37x37 mm sq. package body size did not re- sult in this solder separation failure mode with the mixed VIPPO and non-VIPPO BGA foot- print. However, additional work will be needed to better understand the influence of the pack- age body size on this defect and to determine where the cutoff, or threshold, body size should be defined. Although, the underlying root cause as to why the large body size tends to perform better is not clearly understood, it is hypothe- sized to be related to the package weight, BGA density and possibly the ratio and locations of the VIPPO-to-non-VIPPO solder joints with- in the BGA array. The larger 37 mm sq. FCBGA package is much heavier than the smaller over- molded BGA components included on this test vehicle and hence, may require more strain per solder joint to separate from the IMC. This 37 mm sq. package also has a higher BGA densi- ty due to its fine pitch (down to 0.7 mm pitch) and exhibits a lower ratio of VIPPO-to-non-VIP- PO pads within the BGA footprint as compared with the smaller BGA packages studied. Furthermore, the locations of the VIPPO pads are grouped together rather then being dis- persed among the non-VIPPO pads as seen with the smaller BGA components. This is illustrat- ed in Figure 14 below, showing the non-VIPPO pads connected to a plated through-hole (PTH) in green. Having fewer non-VIPPO pads sur- rounding the VIPPO pads may induce less strain within the VIPPO solder joints resulting in less likelihood for solder separation for the 37 mm sq. component. Further work in this area should be pursued in order to better understand the in- fluence of these parameters. VIPPO and VIPPO + Backdrill Test Vehicle A VIPPO pad with a certain level of back- drilling depth is expected to behave similarly to a non-VIPPO pad. Currently, besides the float- ing pad with traces, non-VIPPO pads are also defined as pads with microvias (layer 1 to lay- er 2 via) or skip-vias (layer 1 to layer 3 via). BGA footprint designs combining VIPPO and any of these non-VIPPO pads have been demonstrat- ed to exhibit the solder separation failure mech- anism. Therefore, BGA footprint designs with mixed VIPPO and VIPPO + back drilling to some threshold depth would be expected to behave similarly, exhibiting the same solder separation mechanism as seen with these standard mixed VIPPO and non-VIPPO BGA footprint designs. Therefore, the objective of this test vehicle is to understand what this backdrill depth threshold should be within a VIPPO and VIPPO + backdrill BGA footprint in order to prevent the solder sep - aration failure mechanism from occurring. Re- sults from this test vehicle will be used to pro- vide a guideline for future PCB designs with VIP- PO and VIPPO + backdrill BGA patterns. VIA-IN-PAD PLATED OVER DESIGN CONSIDERATIONS Figure 14: 37x37 mm sq. variable pitch BGA footprint vs. 17x17 mm sq. BGA footprint.