Issue link: https://iconnect007.uberflip.com/i/291105
28 The PCB Magazine • April 2014 fore testing (to establish a base line), after pre- conditioning (a simulation of assembly and re- work) and at the end of test. We then compare the measurement after preconditioning and at the end of test to those original readings. A -4% change or greater indicates significant material damage and a cross section is processed to con- firm or refute this finding. The major problem is the implementation of lead-free soldering that requires assembly temperatures of 260°C. The FR-4 material is at its limit to withstand the heat when exposed to 260°C. The Z-axis expansion is at its highest at this temperature, putting extra strain on the copper interconnects. The most common type of failure of a ro- bust interconnection is a barrel crack that oc- curs in the central zone of the plated through hole (PTH). When tested using thermal cycling to 150°C, this is a wear-out type of failure that happens over hundreds of cycles (500+ cycles). Surviving 500 cycles without any significant increase in resistance is considered a robust coupon. In a weak coupon, the failure before 350 cy- cles may relate to a process problem, with the most common problem being thin copper plat- ing. With thin plating, the barrel cracks may still be the cause the failure but it would be fail- ing in less than 350 cycles. The PTH may also fail for corner cracks or interconnection separa- tion. Weak buried vias fail typically for barrel cracks in the center zone of the structure simi- HDI PWB RELIABILITY continues Figure 1: Barrel crack due to metal fatigue. Figure 3: corner or knee crack. Figure 2: interconnect separation. Figure 4: lifted copper cap on buried via.