Review of test methods for zero PPM performance

Abstract

PurposeMarket demands, especially within the automotive sector, are pushing towards increased product complexity and performance with zero ship parts per million (PPM) requirements. To achieve both quality and performance goals very stringent requirements are being placed on the test manufacturing solution. These requirements lead to conflicts between cost, performance and quality. The purpose of this paper is twofold: first, to investigate the conflicts that exist between quality, performance, and cost, and second, to review current practices and techniques being used in tests to minimise ship PPM.Design/methodology/approachIn the paper a test process development flow chart is presented, along with a review of current methods being used for both defect screening and performance testing. The relationship between test coverage and ship PPM is investigated using established yield models. The cost in terms of gross margin degradation of yield loss at final test to extensive screening and aggressive limits is modelled.FindingsThe paper finds that to maintain ship PPM very high levels of test coverage are required – typically test coverage needs to be > 98 per cent. The cost of yield loss to this testing typically matches on a percentage point basis gross margin degradation. Reviewing current test methods shows the need both for extensive defect‐screening techniques for the defective portion of the population, and for optimised guardbanding techniques for performance testing. Weaknesses that exist are the absence of a model to predict outgoing PPM, and the conservative nature of existing guardband techniques for performance testing.Originality/valueThis is a review paper and it serves to highlight both the weaknesses in current practices, and areas where improved models are required.