Browse Prior Art Database

Test Data Compression for Logic Circuit Tester

IP.com Disclosure Number: IPCOM000083020D
Original Publication Date: 1975-Mar-01
Included in the Prior Art Database: 2005-Feb-28
Document File: 3 page(s) / 57K

Publishing Venue

IBM

Related People

Chia, DK: AUTHOR [+2]

Abstract

Test data compression techniques will reduce testing data and tester memory size, while at the same time help speed up the testing process.

This text was extracted from a PDF file.
At least one non-text object (such as an image or picture) has been suppressed.
This is the abbreviated version, containing approximately 53% of the total text.

Page 1 of 3

Test Data Compression for Logic Circuit Tester

Test data compression techniques will reduce testing data and tester memory size, while at the same time help speed up the testing process.

The error pattern (E ) for test pattern i (T(i)) is obtained by the exclusive OR of T(i) with its preceding pattern (T(i-1)), i.e., E(i) = T(i) V T(i-1) where T(i) and T(i-1) are in binary representation.

This run length coding scheme encodes the error patterns in two steps: first, count the number of 0's in a run between successive 1's in the error pattern E (since minimum changes between adjacent patterns implies more 0's than 11s, This maps the error pattern into intermediate run-length counts by the rule given in Table 1. The second step assigns binary coded words to run-length counts. The proposed run-length code is shown in Table 2.

It is quite possible that the error pattern in the coded form may actually require more bits to describe than the number of bits in the test pattern itself. In the worst case, it may take up to four times the number of bits required to describe an uncoded test pattern. To prevent possible worst-case data explosion, a noncompress mode of operation will allow the test pattern to be stored, instead of the Sun length coded error pattern. As shown in Table 2, code word 9 is used for indicating noncompress mode of operation.

In addition, the described run-length code has the capability of distinguishing error patterns ending with a 0 from those ending with a 1. With the single or minimum number of input change restriction, the probability that an error pattern ending with a 0 is greater than ending with a 1. For error patterns ending in 0, the encoding will be the run-length code, indicatin...