A new and effective drop test evolution to next-gen handheld applications

Abstract

Drop test performance is one of the key reliability parameters for characterizing electronic packages such as ball grid array (BGA) under mechanical stresses. It is especially important for a large BGA in a handheld device such as a smart-phone or tablet where the solder joint tends to crack under mechanical stress. This has gained more and more interest as mobile computing becomes dominant for next generation handheld applications. The JEDEC test board is widely used in the industry to calibrate the drop performance of solder joints in BGA and other area array components. One main drawback for this test method is in its test vehicle, which is a 132mmx77mm rectangular board populated a 3x5 array of BGAs. The stress and strain at each component behaves differently dependent to the distance of each component to the center and or screw holes used for mounting on the table. This makes statistical and quantitative analysis of the drop test results impossible unless a large number of sample sizes is used. This paper demonstrated two types of square board with 4 components mounted symmetrically on the top. This significantly improves the efficiency of drop test by making all data points undergo the same stress level so that a consistent drop and a good statistical analysis is feasible.

Crack monitoring and life modeling technique towards high thermal cyclic and mechanical reliability of fcBGA solder joint

“…The failure mode found by experiment is actually a combined failure mode where the solder joint may fail at either the trace/pad or within the solder depending on the pad/trace widths and the stress levels of those structures. An interesting finding from authors' previous work [6] shows that, contrary to earlier understanding, increasing the trace width does not necessarily increase the number of drops to failure. The reason may be due to the failure shift identified from the FEA study.…”

“…The mechanical shock test was performed per JEDEC 22B110 service condition B [5,6] with acceleration of 1500G and pulse duration of 0.5ms. The mechanical shock test was performed face down, i.e., -Z axis only, representing the worst case for this test.…”

“…The shock pulse is 1500G 0.5ms which is typically used for qualifying small BGAs in mobile applications [5,6]. For large BGAs (>23mm), a typical acceleration for the drop shock is only 140G or below.…”