Column-grid-array (CGA) versus ball-grid-array (BGA): board-level drop test and the expected dynamic stress in the solder material

Abstract

Board level drop test is considered with an objective to develop a physically meaningful analytical predictive model for the evaluation of the expected impact-induced dynamic stresses in the solder material. Ball-grid-array (BGA) and column-grid-array (CGA) designs are addressed. Intuitively it is felt that while the application of the CGA technology to relieve thermal stresses in the solder material might be quite effective (owing to the greater interfacial compliance of the CGA in comparison with the BGA), the situation might be quite different when the PCB/package experiences dynamic loading. This is because the mass of the CGA joints exceeds considerably that of the BGA interconnections and the corresponding inertia forces might be substantially larger in the case of a CGA design. The numerical example carried out for rather arbitrary, but realistic, input data has indicated that the dynamic stresses in the solder material of the CGA design are even higher than the stresses in the BGA interconnections. This means particularly that the physically meaningful drop height in board-level tests should be thoroughly selected and that this height should be different, for BGA and CGA designs.