Effective package-on-package warpage DOE design with analytical method

Abstract

Package-on-package (PoP) technology is widely adopted in today's processor construction in mobile applications, such as those used in smart phones and tablets. One of the crucial steps in a successful PoP structure implementation is to control the PoP bottom package's warpage. Due to the short life cycle of such a package, usually one or two years, PoP packaging development cannot afford large scale, time consuming and complex design of experiments (DOE's) to identify an optimized package bill of materials (BOM) to meet the package's warpage requirement. It is, therefore, in need of a simple but predicative analytical model that could make use of available warpage data to provide guidance to design an effective and efficient DOE plan for a new package's development or for an existing package's warpage improvement. Based on current mobile applications, a typical PoP package will have a ratio of its total thickness to body size in the range of 1:20, or 0.05. Such geometry can fit the PoP package into the category of a multilayered plate structure when dealing with its mechanical behaviors. This paper uses the classic theory of asymmetric layered composites to analyze how a PoP package warpage changes under the SMT process. It is discovered that, instead of the absolute values, the difference in linear coefficient of thermal expansions (CTE's), the relative temperature change, the mechanical properties ratio as well as the thickness ratio between various layers are the keys to control the PoP package's final warpage. The theory can explain why a bare substrate without any die attached shows substantial warpage under SMT reflow conditions, and why substrates with exactly the same BOM but procured from different vendors show different warpage performance. With this method, PoP package development engineers can understand the physics and mechanics of the thermal warping process of PoP package under the SMT environment. An example was given to demonstrate how this method can be applied to actual product development to reduce material cost and cycle time.