Browse Prior Art Database

Systems Approach to Diagnostics

IP.com Disclosure Number: IPCOM000083142D
Original Publication Date: 1975-Apr-01
Included in the Prior Art Database: 2005-Mar-01
Document File: 4 page(s) / 55K

Publishing Venue

IBM

Related People

Narasimha, MS: AUTHOR

Abstract

Because of the degrading influence of the rework process on LSI (Large-Scale Integration) modules, there are restrictions on the number of times they can be reworked. Therefore, the traditional approach of diagnosing and correcting one defect for every test-rework cycle is inadequate and a practical approach for diagnosing multiple defects is imperative. In addition, multiple defect diagnosis improves the tester utilization and the module manufacturing cycle time. The following is a systems approach to this diagnostics problem.

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 43% of the total text.

Page 1 of 4

Systems Approach to Diagnostics

Because of the degrading influence of the rework process on LSI (Large- Scale Integration) modules, there are restrictions on the number of times they can be reworked. Therefore, the traditional approach of diagnosing and correcting one defect for every test-rework cycle is inadequate and a practical approach for diagnosing multiple defects is imperative. In addition, multiple defect diagnosis improves the tester utilization and the module manufacturing cycle time. The following is a systems approach to this diagnostics problem.

The key to the approach lies in first detecting a defect and pereceiving the sphere of its influence, and then continuing application of tests addressing the domain of circuits outside this sphere to possibly detect another independent defect. This process is repeated until all the independent defects are identified. An effective realization of this procedure can be achieved by:
(a) Exploiting the implication of the LSI design rules

and
(b) Imposing a suitable test sequence organization to facilitate defect circumvention and continued testing.

A probe can then be used to converge upon a net directly affected by each defect, referred to as a terminal net. The set of terminal nets corresponding to the identified independent defects can be further analyzed for diagnosing physical defects at these nets (such as shorts or opens), or the components associated with the nets as defective.

Erroneous perception of a Sphere of Influence (SOI) larger than warranted by a defect, reduces the potential of the number of independent defects that can be diagnosed. The major thrust of the design rules is to realize a design with mostly combinational logic circuits, surrounded by primary or latch inputs and outputs (I/O's). These latches are appropriately clocked and are connected as scannable shift registers, to provide surface visibility of these internal storage elements at the I/O's.

The rules offer a higher level of commonality across different product designs than the primitives AND, OR and INVERT circuits. The logic is separated into three identifiably different categories viz., a) data path logic (combinational), b) scan path logic, and c) clock path logic. The SOI of a defect determined by exploiting this separation and based on the intent of the failing test will be optimal. There are three types of tests viz., 1) Prepower tests. 2) Shift register tests. 3) Stuck-fault tests for the combinational logic. Prepower Tests. A condition such as a short to a power or voltage net, could render powering up of the module hazardous for the product and the tester. The prepower tests are intended to detect such conditions in addition to other problems at the product I/O's viz., missing terminators, shorts or opens on I/O nets. When a hazardous condition is detected, the module power-up is prevented and the problem can be reported for correction. But if no hazardous condition is witnessed, then the...