초록 ▼

An interactive diagnostic methodology and apparatus that enables a specimen (microelectronic device) to undergo detailed behavioral examination while allowing the Diagnostician to incrementally proceed from a set of symptoms to a formal diagnosis. Diagnostic tactics are predicated on comparing measu

An interactive diagnostic methodology and apparatus that enables a specimen (microelectronic device) to undergo detailed behavioral examination while allowing the Diagnostician to incrementally proceed from a set of symptoms to a formal diagnosis. Diagnostic tactics are predicated on comparing measured behavior to predicted behavior. The methodology allows coupling to a variety of behavior-level, switch-level (state), and circuit-level (node) simulators in determining predicted behavior. The architecture allows a variety of beam-probing apparatus to be supported in acquiring nodal waveforms and logic state images from a specimen as a means of determining measured behavior. This measured behavior can be graphically compared to the predicted behavior for study and analysis by the Diagnostician to aid in the formulation of diagnostic tactics. The ability to easily couple a wide variety of simulation techniques and beam probing apparatus is achieved by viewing simulation and measurement as generic functions, which can be executed concurrently, from an architectural standpoint. To accelerate nodal navigation and model extraction, an object clustering data structure is used that exploits apriori continuity relationships between objects, symbolically representing the specimen\s design. The importance of defining stimuli, for exercising the specimen under various measurement and simulation conditions, is fully recognized with an integrated capability for the interactive composition and modification of stimuli patterns. The apparatus architecture provides for network coupling to the CAD environment for access to information concerning the specimen\s design and in performing simulation activities to support diagnostic tactics.

대표청구항 ▼

Apparatus for diagnosing abnormal behavior in a mal-functioning integrated circuit specimen comprising in combination: noncontact probe means for acquiring voltage waveforms to be measured at selectible circuit nodes within the specimen; stimulator means; simulator means computer means including dis

Apparatus for diagnosing abnormal behavior in a mal-functioning integrated circuit specimen comprising in combination: noncontact probe means for acquiring voltage waveforms to be measured at selectible circuit nodes within the specimen; stimulator means; simulator means computer means including display means for composing, modifying and displaying input stimuli patterns for application to the stimulator means the simulator means; said stimulator means utilizing said input stimuli patterns from the computer means for dynamically exercising the specimen to output voltage waveforms to be measured; said simulator means storing a mathematical model of the integrated circuit specimen, as designed, in the form of object clustering data structure enabling concurrent production of corresponding predicted waveforms and waveforms to be measured due to simultaneous stimulation of the specimen and the simulator means; said display means further exhibiting pictorial representation of the specimen layout, as described in the mathematical model, to allow an operator select a node on the layout where comparison of measured and simulated waveforms from the mathematical node and corresponding specimen node are to be made; said simulator means outputting predicted behavioral characteristics of the circuit specimen as predicted waveforms applied to the display means along with any output waveform of the noncontact probe means acquired during said exercising; and said display means also enabling pictorial comparison of said predicted and measured waveforms to permit identification of circuit errors causing abnormal behavior in the specimen. A method for determining the cause of mal-function in an integrated circuit, comprising the steps of: noncontact probing for detecting the magnitude of an analog waveform at a selected node site on the circuit; storing a mathematical model of the integrated circuit based on an object clustering data structure for predicting analog and digital waveform; storing analog measured and predicted waveforms; storing digital stimuli waveforms; simultaneously applying selected stimuli to the circuit observed by the noncontact probing and to the stored mathematical model underlying the simulation; displaying a response to the stored output of the noncontact probing and to the stored simulated output of the mathematical model to portray separate waveforms for each to permit determination of whether or not the waveforms agree or disagree; and steering the noncontact probing to additional node sites for exploring circuit behavior, incrementally until the node site indicating a mal-function by waveform disagreement is identified.