Browse Prior Art Database

Method of Applying, Monitoring and Recording Multiple Test Criteria

IP.com Disclosure Number: IPCOM000110277D
Original Publication Date: 1992-Nov-01
Included in the Prior Art Database: 2005-Mar-25

Publishing Venue

IBM

Related People

Curtis, BA: AUTHOR [+4]

Abstract

Described is a method of applying, monitoring and recording multiple test criteria for the efficient simultaneous characterization of performance envelope data for multiple units under test.

This text was extracted from an ASCII text file.
This is the abbreviated version, containing approximately 25% of the total text.

Method of Applying, Monitoring and Recording Multiple Test Criteria

       Described is a method of applying, monitoring and
recording multiple test criteria for the efficient simultaneous
characterization of performance envelope data for multiple units
under test.

      The multiple test criteria method provides an optimized
approach for estimating multiple boundary conditions along multiple
independent unidirectional vectors where each test is applied
empirically and simultaneously to all units under test.  Through the
use of this optimized testing method, stress data can be obtained
quickly, thereby enabling the user to evaluate design capabilities.

      In prior art, manual and/or exhaustive testing has been used to
determine the operating space or the performance envelope for the
unit under test.  This envelope was determined by empirically
applying test criteria to the unit under test while various stress
parameters were applied.  Mapping the pass or fail response of the
unit under test was then performed in order to obtain an estimate of
the pass/fail boundary of the product.  Typically, this procedure was
performed manually on one unit at a time so as to apply one or more
test criteria sequentially.  As automation became evident, the manual
one at a time approach became obsolete.  Software was developed which
exhaustively maps the response of each unit under test at each stress
condition.  However, this approach was considered inefficient since
the brute-force algorithm led to many meaningless applications of the
test and was time consuming.  Also, this approach lacked the means to
obtain separate data for different test criteria.  For example, if
separate performance envelopes were being obtained for a product
running test criteria A and results were also required for test
criteria B, the entire test had to be run separately for each of
these test criteria.

      The implementation of the multiple test criteria method
consists of the disclosed algorithms used to control automated test
equipment.  The algorithms make use of the stress vector concept
where the vector involves a combination of stresses.  A vector may be
thought of as a stress parameter where the controlled setting is
under the control of an algorithm.  The vector is incremented in
predetermined steps by the algorithm, and all chosen criteria are
applied until all units under test fail their criteria.

      The following list outlines the features provided in the
multiple test criteria method:
      -   Selectable Number of Pass/Fail Retries - The user is able
to select any number of pass or fail events which must occur in a row
in order for the result to be considered valid.  This allows the
filtering of intermittent or spurious results.
      -   Selectable Combination of Search Types - The user may
select one of several search algorithms which include linear
searches, binary searches and a combination of linear/bin...