BEST PRACTICES FOR DESIGNING IN QUALITY AND RELIABILITY

Over the years system manufacturers have continually placed requirements on suppliers of PEMs for demonstrating ever increasing high levels of quality and reliability, which in many ways exceed those imposed by the military (see Appendix Tables A2 and A3). Demands for AOQs (Average Outgoing Quality) of less than 20 ppm and failure rates of 10 FITs (Failure in Time) are not uncommon today, with these values expected to decrease an order of magnitude before the end of the decade. In order to meet these challenges, Total Quality Management (TQM) systems are employed by Intersil throughout all phases of product development, manufacture, and service. Within the scope of TQM, new technologies are developed by the Technology (Process) Development System (TDS), which is overlapped by a system called Applying Concurrent Teams to the Product-to-Market process (ACT-PTM). In brief, these systems are manifested in well defined, multidisciplinary project teams, which are given full responsibility for successful project execution from early product development to end customer service. This results in greater first pass success of new processes and products, reduced cycle time, enhanced quality and reliability, and continuous improvement. A full description of these systems is given in reference [26].

The principles for Building in Reliability (BIR) are used during the early stages of technology/product development to ensure wear-out and special cause mechanisms are eliminated from useful product life [27,28]. Figure 9 illustrates the paradigm shift made in the mid-1980's with respect to the BIR concept. The old paradigm (energy down stream) with Reliability involvement at the end of the development cycle led to qualification failures, recycle of design and three to four years to introduce a new technology. The BIR paradigm (energy up stream) with early Reliability involvement beginning at the concept facilitates the `build-in' process, increases first pass success and has reduced cycle time by a factor of four. Two important benefits of BIR are reliability inputs into the design/layout groundrules and definition of reliability critical process node parameters. Appendix Table A4 gives a brief listing of some of the best practices employed.

Verification of reliability is achieved through rigorous characterization/qualification testing at both the wafer and package level and then through the use of continuous production reliability monitoring (Appendix Table A3). SPC methods are employed throughout the manufacturing processes. Each manufacturing line (both wafer fab and package assembly) has a critical node list to assure quality and reliability are maintained and improved over time (examples are provided in Appendix Tables A5 and A6). These methodologies apply to the development of all product technologies, regardless of whether the end-use application is commercial, industrial, or military. Some of the tools/methods employed are listed below.