Browse Prior Art Database

Weighted Syndrome Sums Approach to VLSI Testing

IP.com Disclosure Number: IPCOM000048468D
Original Publication Date: 1982-Feb-01
Included in the Prior Art Database: 2005-Feb-08
Document File: 2 page(s) / 76K

Publishing Venue

IBM

Related People

Barzilai, Z: AUTHOR [+4]

Abstract

A new technique for self-testing VLSI (very large-scale integration) combinational logic employs one or more weighted syndrome sums, rather than a collection of syndromes, as the reference for comparison in test mode, to implement self-testing with a low hardware overhead.

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 2

Weighted Syndrome Sums Approach to VLSI Testing

A new technique for self-testing VLSI (very large-scale integration) combinational logic employs one or more weighted syndrome sums, rather than a collection of syndromes, as the reference for comparison in test mode, to implement self-testing with a low hardware overhead.

Previous VLSI testing approaches have employed syndrome testing. The notion of syndrome testing is based on counting the number of binary ones realized by a Boolean function and comparing this number to the fault-free count. It is assumed that the VLSI circuit has been designed to be syndrome-testable to assure that all faulty syndromes will differ from the fault-free one. In a typical VLSI environment, the implementation of syndrome testing may require the use of hundreds of syndrome reference words. The use of a weighted syndrome sum enables the dramatic reduction in the number of references needed for syndrome testing.

In order to implement weighted syndrome sum testing, a VLSI chip is partitioned into R macros, where a macro is a multi-input, multi-output combinational circuit having, for example, n inputs and m outputs. Weighted syndrome sum testing is used to test each macro of the chip.

For an m output macro, let approaching S equals (S(1),S(2),...,S(m)) be a vector representing the output syndromes of the macro, and let the function g equals (see article) w(i)Z(i) be a linear combination of the variables Z(1), Z(m) with coefficients w(1), w(2),..., w(m),...