Abstract:
A method and apparatus for testing complex nonlinear binary circuits by applying a bilevel signal pattern, particularly a random pattern, to both a plurality of inputs in the circuit being tested and to a corresponding plurality of inputs in a reference simulation of said circuit, and for comparing corresponding outputs from the circuit and the simulation. The apparatus includes means for converting the signal pattern which is to be applied to the simulation to a three level signal pattern in which two levels represent the two levels in the bilevel pattern and the third level represents an indeterminate binary circuit state. The reference simulation is adapted to receive a three level pattern input and to provide a three level output. Means for applying the three level signal pattern to the reference simulation include means for applying a third level signal to a given simulation input during a change between first and second level signals being applied to said input. Sensing means determine which portion of the simulation output remains at the third level, particularly after a change between any of the first two levels at one or more of the simulation inputs; the existence of the third or indeterminate level indicates that the change produced race conditions critical to the portion of the output at the indeterminate level. Inhibiting means prevent that portion of the simulation output at the indeterminate level from being compared with the output of the circuit being tested. This presents an improper rejection of a good circuit because of ''''compare failure'''' caused by the critical race conditions.
Abstract:
A method and apparatus for testing complex nonlinear binary circuits by applying a bilevel signal pattern, particularly a random pattern, to both a plurality of inputs in the circuit being tested and to a corresponding plurality of inputs in a reference simulation of said circuit, and for comparing corresponding outputs from the circuit and the simulation.