Prediction of Theoretical Limit for Test Data Compression

Abstract

Reduction in time for testing high volume of test data can be accomplished through test data compression. The number of pins available in Automatic Testing Equipment (ATE) plays a major role in deciding the achievable compression ratio. Number of pin limits the maximum number of scan chains loading in parallel. One of the method to overcome this pin limitations and to achieve increasing compression level is applying more number of scan chains that decreases the volume of test data which needs to be encoded in the pool of test vectors. But serially applying more test cases increases the testing cost. This increased test cost nullifies all the benefits of the achieved test data compression. In this paper, a Probability based Entropy Dependent Predictive Technique (PEDPT) for envisioning the most extremely acceptable data compression and the testing cost of a given scan design, is proposed. This will permit the designer to choose appropriate technique based on design considerations, the data available, computational unpredictability and the achieved compression rate.