James W. Fuller

The execution of aggressive PBGA substrate yield learning in an existing PWB facility

As the market for PBGA products explodes and substrate facilities are designed, built, and brought on line, yield learning is vitally important. It is rare that a new product will be introduced at its steady state yield target, necessitating aggressive yield improvement planning. In particular, manufacturers who have converted portions of existing PWB capacity to PBGA product sets will find this to be true. In this paper, the authors articulate the significant challenges manufacturers face ramping up PBGA product. Complex logistics, multiple process flows, multiple customer requirements, aggressive delivery schedules, non-PWB defect mechanisms, non-functionally defined engineering specifications, and a paradigm shift in manufacturing philosophy complicate a product with great intrinsic manufacturing difficulty. This paper reviews in detail the challenges, philosophy and methodology employed to achieve dramatic improvement in PBGA product yields. The paper also includes suggestions for changes in business process procedures to ensure yield learning is engrained as part of any PBGA product introduction. A detailed system of matrix management that utilized process control as the foundation for yield improvement is included. The organization structure, review cycle, improvement road maps, yield tracking and data analysis are discussed in detail. Overall yield improvement results, along with several representative products, are generically shared to validate the philosophy and methodology employed.

Transition: lessons learned in the development and production of IBM's /spl mu/Laminate/sup TM/ PBGA

This paper describes challenges encountered in the development of IBMs /spl mu/Laminate/sup TM/ PBGA. The implementation of a new product line was based on the factories in place to ramp production with extensions to base technology. Lessons learned in the product development are focused to provide an introspective look at organization and approaches. Solutions to the challenges cited are specific to the IBM product line and structure of the Endicott and Bromont facilities, but are also inherent to the PBGA product itself and should be applicable to other tools and designs. Specific areas discussed include the product development process vehicle decisions, the laminate development and ramp, bond and assembly development, and concludes with recommendations from the experience.