Abstract: Aspects of the disclosed technology relate to techniques of activity coverage assessment. Transistor-level circuit simulation is performed for a circuit design under a set of test stimuli, which determines values of one or more electrical properties for each of circuit elements of interest in the circuit design. The one or more electrical properties are selected based on information of the each of circuit elements of interest, which comprises what circuit element type the each of circuit elements of interest belongs to. Based on the values of the one or more electrical properties, activity coverage information comprising information about which circuit elements in the circuit elements of interest are active or inactive under the set of test stimuli is determined.

Abstract: This application discloses a computing system to identify multiple views of cells in a circuit design for selective utilization during analog fault simulation of the circuit design. The views of the cells can include two or more of macromodel design views, schematic design views, or extracted design views that includes parasitic elements extracted from a physical layout of the circuit design. The computing system can prompt generation of multiple netlists, each netlist generated based on a different combination of the identified views of the cells in the circuit design, or a list of macromodels with pin accurate subcircuit wrappers, parse and organize the cells in each netlist or the list of macromodels, identify one of the cells to inject with a defect, and selectively simulate portions from a plurality of the netlists based, at least in part, on the identified one of the cells to inject with the defect.

Abstract: This application discloses a computing system to identify multiple views of cells in a circuit design for selective utilization during analog fault simulation of the circuit design. The views of the cells can include two or more of macromodel design views, schematic design views, or extracted design views that includes parasitic elements extracted from a physical layout of the circuit design. The computing system can prompt generation of multiple netlists, each netlist generated based on a different combination of the identified views of the cells in the circuit design, or a list of macromodels with pin accurate subcircuit wrappers, parse and organize the cells in each netlist or the list of macromodels, identify one of the cells to inject with a defect, and selectively simulate portions from a plurality of the netlists based, at least in part, on the identified one of the cells to inject with the defect.

Abstract: Aspects of the disclosed technology relate to techniques of activity coverage assessment. Transistor-level circuit simulation is performed for a circuit design under a set of test stimuli, which determines values of one or more electrical properties for each of circuit elements of interest in the circuit design. The one or more electrical properties are selected based on information of the each of circuit elements of interest, which comprises what circuit element type the each of circuit elements of interest belongs to. Based on the values of the one or more electrical properties, activity coverage information comprising information about which circuit elements in the circuit elements of interest are active or inactive under the set of test stimuli is determined.

Abstract: Aspects of the invention relate to techniques of defect injection for transistor-level fault simulation. A circuit element in a circuit netlist of a circuit is first selected for defect injection. Next, a defect is determined based on whether the selected circuit element is a design-intent circuit element or a parasitic circuit element. After the defect is determined, the defect is injected into the circuit netlist and then the circuit is simulated.

Abstract: Various aspects of the present invention relate to techniques of measuring delays between edges of signals of a circuit. Alternating signals, synchronous to a first clock, are supplied to a plurality of nodes of the circuit. First samples of a plurality of signals associated with the alternating signals are captured using a first capture clock, of which sampling instants are synchronous to a second clock. Second samples of the first samples are then captured using a second capture clock, of which sampling instants are also synchronous to the second clock. The captured second samples are conveyed via a shift register to a plurality of modulo counters. The measured signal delay includes a timing skew associated with the first clock and a timing skew of the first capture clock but not a timing skew of the second capture clock.

Abstract: Aspects of the invention relate to techniques of defect injection for transistor-level fault simulation. A circuit element in a circuit netlist of a circuit is first selected for defect injection. Next, a defect is determined based on whether the selected circuit element is a design-intent circuit element or a parasitic circuit element. After the defect is determined, the defect is injected into the circuit netlist and then the circuit is simulated.

Abstract: Various aspects of the present invention relate to techniques of measuring delays between edges of signals of a circuit. Alternating signals, synchronous to a first clock, are supplied to a plurality of nodes of the circuit. First samples of a plurality of signals associated with the alternating signals are captured using a first capture clock, of which sampling instants are synchronous to a second clock. Second samples of the first samples are then captured using a second capture clock, of which sampling instants are also synchronous to the second clock. The captured second samples are conveyed via a shift register to a plurality of modulo counters. The measured signal delay includes a timing skew associated with the first clock and a timing skew of the first capture clock but not a timing skew of the second capture clock.

Abstract: Aspects of the invention relate to techniques of defect injection for transistor-level fault simulation. A circuit element in a circuit netlist of a circuit is first selected for defect injection. Next, a defect is determined based on whether the selected circuit element is a design-intent circuit element or a parasitic circuit element. After the defect is determined, the defect is injected into the circuit netlist and then the circuit is simulated.

Abstract: A circuit and method provide built-in measurement of delay changes in integrated circuit paths. The circuit includes a digital shift register to access multiple paths, and may be implemented in digital boundary scan to test I/O pin delays. Synchronous to a first frequency, the circuit applies an alternating signal to the paths and samples the paths' output logic values synchronous with a second frequency that is asynchronous and coherent to the first clock frequency. The shift register conveys the samples to a modulo counter that counts the number of samples between consecutive rising or consecutive falling edges in the signal samples from a selected path. Between the two edges, the path or a path characteristic is changed, and the resulting modulo count after the second edge is proportional to the change in delay. The circuit can compare the count, or the difference between counts, to test limits.

Abstract: A circuit and method provide built-in measurement of delay changes in integrated circuit paths. The circuit includes a digital shift register to access multiple paths, and may be implemented in digital boundary scan to test I/O pin delays. Synchronous to a first frequency, the circuit applies an alternating signal to the paths and samples the paths' output logic values synchronous with a second frequency that is asynchronous and coherent to the first clock frequency. The shift register conveys the samples to a modulo counter that counts the number of samples between consecutive rising or consecutive falling edges in the signal samples from a selected path. Between the two edges, the path or a path characteristic is changed, and the resulting modulo count after the second edge is proportional to the change in delay. The circuit can compare the count, or the difference between counts, to test limits.

Abstract: A circuit and method provide built-in measurement of delay changes in integrated circuit paths. The circuit includes a digital shift register to access multiple paths, and may be implemented in digital boundary scan to test I/O pin delays. Synchronous to a first frequency, the circuit applies an alternating signal to the paths and samples the paths' output logic values synchronous with a second frequency that is asynchronous and coherent to the first clock frequency. The shift register conveys the samples to a modulo counter that counts the number of samples between consecutive rising or consecutive falling edges in the signal samples from a selected path. Between the two edges, the path or a path characteristic is changed, and the resulting modulo count after the second edge is proportional to the change in delay. The circuit can compare the count, or the difference between counts, to test limits.

Abstract: A method and built-in circuit are described for testing direct current (DC) parameters of the input and output pins of a circuit by testing the transition time interval for rising and falling voltage transitions. When the voltage transition is for an integrated circuit (IC) pin having a known capacitance, which can include off-chip capacitance, the magnitude and direction of current at the pin can be determined. The method enables testing an IC via a test access port (TAP) comprising a subset of the pins of the IC, for example in conformance with the IEEE 1149.1 boundary scan test standard. For sufficiently small current magnitudes, such as leakage current (IIL and IIH), the technique can use only on-chip circuitry to sample a pin voltage at time intervals after an output transition is generated at the pin, the time intervals pre-determined to be less than the transition time interval.

Abstract: A method of testing phase locked loops (PLL) and a testing circuit comprising the steps of applying a normal stimulus signal whose frequency is within the lock range of the PLL to the input of the PLL, substituting the normal input stimulus with an alternative signal derived from an internal feedback of the PLL, adding or deleting one or more cycles from the alternative signal and observing the response of the PLL to the alternative signal. Variations of the method allow for determining Gain-Bandwidth product, lock range, lock time, Bit Error Rate, Jitter and other parameters which can then be compared with predetermined values to determine whether the PLL is properly functional.

Abstract: A circuit and method is described whose objective is built-in self-test (BIST) for analog-to-digital converters (ADCS) and input logic gates of an integrated circuit. The technique converts the switching point voltage, or logic threshold, into a binary-encoded digital value which can be compared to upper and lower limits to decide pass or fail. Every clock cycle, the output of the ADC is compared to a digital output value, and if the output is larger than the reference a logic 0 is output, otherwise a logic 1. This series of ones and zeroes is fed back to an analog low pass filter connected to the ADC's input, and also to a digital averaging circuit which counts the number of ones in a constant interval. The number of ones is linearly proportional to the switching point voltage.

Abstract: A circuit and method is described which generates a high frequency clock signal whose frequency is accurate enough to be used for testing other circuitry, yet the circuit can be described using a hardware description language so that it is suitable for logic synthesis and automatic layout. The technique uses a plurality of programmable ring oscillators and means to select and enable one of the ring oscillators. The output frequency is measured relative to that of a lower frequency reference signal, and when the output frequency is incorrect, a different ring oscillator is selected or the present ring oscillator's frequency is changed. Circuitry is included to prevent glitches at the output of the clock generator when the frequency is changed, regardless of how the ring oscillators are constructed.