A flexible approach to data collection for component test systems

Abstract

Yield improvements for VLSI devices require data collection and analysis during the manufacturing cycle. These data should be collected in a form that satig'ies the needs of those performing diferent functions within the factory. This paper describes an approach to data collection that focuses on collecting test data in a format customized to the users needs. Though implemented on an advanced VLSI test system, key elements of this approach can be applied to any data collection domain. IntroductionTo improve the manufacturing process of semiconductor components, data must be collected at each stage of the production cycle. These data must then be integrated for analysis to determine the root cause of component defects [ 1,2]. For these reasons, semiconductor manufacturers require data collection systems that streamline the process of integrating test data into existing computer integrated manufacturing (CIM) [3] systems. Advanced symbolic ATE programming (ASAP) [4] software executes in the UNIX (SUN OS) environment on a SUN4 workstation. This integrated test program develop ment environment provides X window system GUI (graphical user interface) [5,6] tools, a run-time environment, and background processes to support all phases of component test, including: Migration of existing test programs. Development of new test programs. Off-line simulation of test program executions. Characterization of IC components. Wafer sort [7,81 and final test. A key component of ASAP is the data collection package, a set of GUI tools and daemons (background processes), which configures and collects both datalog and summary data. Datalog data are collected during the test of a single device. These data include the start of test (SOT) and end of test (EOT) information, as well as the individual DC, functional, and search test results. The summary data are collected and summarized for an entire wafer or lot of device tests. These data consist of binning summaries for the hard bins (handler bins), soft bins (summary categories), and summary counters. Summary counters count either the number of times a given test has been executed, or the number of times the test has passed or failed. The data collection package provides a flexible interface to database management systems (DBMSs). The features include: 0 e e e 2.Modular design to allow for expansion.User specification of the format of the data collected. Ability to add custam data items to the data collected, without sacrificing the ability to format the data. Yield management to detect possible manufacturing processerrors.Graphical user interface to simplify performing the above tasks.