Abstract: An amplifier may include two or more pulse-width modulators (PWMs) controlling respective sets of switches to produce an amplified version of a source signal. The clocking for the amplifier may be controlled to delay signal processing within the PWMs relative to one another in time, thereby providing an effective time offset between the absolute moment in time of the edge transition of the controlling signals to the respective sets of switches. The PWMs may include a decrementor that counts down to zero from the next PWM duty-cycle value when a new data sample is detected, beginning a new count when the next sample is present. The PWM output may correspond to the counter value, outputting a pulse when the counter value is nonzero. A “data-sample-ready” signal may be decoded from a master counter, which may be clocked based on the high speed PWM clock, and the delay mechanism may be based on adjusting the decode value to determine when the PWM should initialize to the next data sample.

Abstract: An amplifier may include two or more pulse-width modulators (PWMs) controlling respective sets of switches to produce an amplified version of a source signal. The clocking for the amplifier may be controlled to delay signal processing within the PWMs relative to one another in time, thereby providing an effective time offset between the respective edge transitions of the controlling signals provided to the respective sets of switches. The PWMs may count down to zero from the next PWM duty-cycle value when a new data sample is detected, beginning a new count for each new sample, with the PWM outputting a pulse when the counter value is nonzero. A “data-sample-ready” signal may be decoded from a master counter, which may be clocked based on the high speed PWM clock, and the decode value may be adjusted to determine when the PWM should initialize to the next data sample.

Abstract: An amplifier may include two or more pulse-width modulators controlling respective sets of switches to produce an amplified version of a source signal. A positive DC-offset based on the source signal may be applied to the pulse-width modulator controlling one respective set of switches, and an equal value negative DC-offset may be applied to the pulse-width modulator controlling the other respective set of switches, to provide an effective offset between the respective points in time of the rising/falling edges of the different pulse-width modulated control signals. The addition of alternating positive and negative DC-offset values doesn't affect the output load, and doesn't degrade the signal. The DC-offsets may be added at a frequency selected to be beyond the signal baseband, and the value of the small input signal level may be determined using an RMS level comparator or similar measurement technique.

Abstract: An amplifier may use pulse-width modulators controlling respective sets of switches to produce an amplified version of a source signal. A phase locked loop in the amplifier may generate a differential clock signal. A first processing element operating according to a first supply voltage may generate a PWM signal representative of the source signal, and also generate a clock enable signal corresponding to the differential clock signal. A second processing element (PE2) may receive the differential clock signal, the PWM signal, and the clock enable signal, and level shift the PWM signal and the clock enable signal to operate according to a second supply voltage, and may generate a resampling clock signal from the differential clock signal according to the level shifted clock enable signal. The PE2 may provide a PWM output signal representative of the source signal by resampling the level shifted PWM signal with the resampling clock signal.

Abstract: Integrated circuits are fabricated with mounted discrete capacitors. Bond pads and land pads are fabricated on a semiconductor wafer. Discrete capacitors are mounted on the semiconductor wafer with flexible adhesive. The flexible adhesive accommodates a difference in thermal expansion between the discrete capacitors and the semiconductor wafer. The land pads are electrically coupled to the electrodes of the discrete capacitors. The semiconductor wafer is separated into multiple semiconductor dice. The semiconductor dice are mounted in respective packages. The bond pads on each semiconductor die are electrically coupled to the interconnect terminals of the respective package.

Abstract: Approaches for analyzing a power grid of an integrated circuit are described. In one embodiment, a method includes selecting at least one portion of the integrated circuit to be analyzed. A power grid model corresponding to the integrated circuit is retrieved from a database, and a first simulation of the programmable integrated circuit is performed. The first simulation generates a respective waveform of an electrical characteristic over time for each connection of a component within the selected portion to voltage supply or voltage ground. A simulation is performed of the power grid model using the respective waveforms as input stimulus for each connection in the selected portion.

Abstract: An amplifier may include two or more pulse-width modulators (PWMs) controlling respective sets of switches to produce an amplified version of a source signal. The clocking for the amplifier may be controlled to delay signal processing within the PWMs relative to one another in time, thereby providing an effective time offset between the respective edge transitions of the controlling signals provided to the respective sets of switches. The PWMs may count down to zero from the next PWM duty-cycle value when a new data sample is detected, beginning a new count for each new sample, with the PWM outputting a pulse when the counter value is nonzero. A “data-sample-ready” signal may be decoded from a master counter, which may be clocked based on the high speed PWM clock, and the decode value may be adjusted to determine when the PWM should initialize to the next data sample.

Abstract: An amplifier may include two or more pulse-width modulators controlling respective sets of switches to produce an amplified version of a source signal. A positive DC-offset based on the source signal may be applied to the pulse-width modulator controlling one respective set of switches, and an equal value negative DC-offset may be applied to the pulse-width modulator controlling the other respective set of switches, to provide an effective offset between the respective points in time of the rising/falling edges of the different pulse-width modulated control signals. The addition of alternating positive and negative DC-offset values doesn't affect the output load, and doesn't degrade the signal. The DC-offsets may be added at a frequency selected to be beyond the signal baseband, and the value of the small input signal level may be determined using an RMS level comparator or similar measurement technique.

Abstract: Predicting simultaneous switching output noise of an IC device is described. User input is obtained. The user input includes: an identification of an input/output bank of an integrated circuit die; an identification of a device package substrate to which the integrated circuit die is to be attached; and an identification of input/output interface to be used by the input/output bank. A noise factor and an impedance are selected responsive to the user input. The noise factor is multiplied with the impedance to provide a result. The result, which is output, is a prediction of the simultaneous switching output noise of the integrated circuit device.

Abstract: The relative timing of triggering switching events in a circuit block of an IC device is dynamically adjusted in response to fluctuations in device's supply voltage to minimize clock jitter caused by supply voltage noise. A control circuit monitors supply voltage fluctuations, and in response thereto dynamically phase-shifts a clock signal that triggers the switching events so that the switching events occur during relatively quiet time intervals in which fluctuations in the supply voltage are minimal.

Abstract: A method for the prediction of simultaneous switching output (SSO) noise that may be generated by one or more signal conduction paths within an electrical system. Electrical disturbance waveforms are first recorded for each signal conduction path that may be affected by the electrical disturbances. Next, principles of superposition are utilized to coherently combine each of the electrical disturbance waveforms in the time domain to generate the predicted SSO noise waveform that is imposed upon the affected signal conduction path. The electrical disturbance waveforms may be produced either by using bench measurements performed on an actual integrated circuit, by simulation, or by a combination of simulation and bench measurements.

Abstract: Induced crosstalk is predicted for the input/output pins of a programmable logic device. Signal edge rates for the input/output pin are determined from selected interface protocols for the input/output pins. For each pair of the input/output pins, a first coupling coefficient specifies a coupling between the pair of input/output pins within a package for mounting the programmable logic device to a printed circuit board. A depth is input for each via coupled to an input/output pin by the printed circuit board. From the via depths, a second coupling coefficient is determined for each pair of the input/output pins that satisfy a separation criterion. For each of the input/output pins, a predicted value of an induced crosstalk is determined from the first and second coupling coefficients for each pair that includes the input/output pin and another input/output pin, and from the signal edge rate of this other input/output pin.

Abstract: Transient processing mechanisms for power converters. Error generation circuitry in a power converter may generate an error signal based on the difference between a power converter output voltage and a reference voltage. Transient detection circuitry may detect whether the error signal exceeds at least a first threshold. If the first threshold is exceeded, timing control logic may generate a low band correction pulse to adjust the power converter output voltage, and thereby adjust the error signal to a level within the first threshold. If the error signal exceeds a second threshold, the timing control logic may generate a high band correction pulse to adjust the power converter output voltage, and thereby adjust the error signal to a level within the second threshold. The timing control logic may initiate a low band blanking period following the low band correction pulse and high band blanking period following the high band correction pulse.

Abstract: Transient processing mechanisms for power converters. Error generation circuitry in a power converter may generate an error signal based on the difference between a power converter output voltage and a reference voltage. Transient detection circuitry may detect whether the error signal exceeds at least a first threshold. If the first threshold is exceeded, timing control logic may generate a low band correction pulse to adjust the power converter output voltage, and thereby adjust the error signal to a level within the first threshold. If the error signal exceeds a second threshold, the timing control logic may generate a high band correction pulse to adjust the power converter output voltage, and thereby adjust the error signal to a level within the second threshold. The timing control logic may initiate a low band blanking period following the low band correction pulse and high band blanking period following the high band correction pulse.

Abstract: Transient processing mechanisms for power converters. Error generation circuitry in a power converter may generate an error signal based on the difference between a power converter output voltage and a reference voltage. Transient detection circuitry may detect whether the error signal exceeds at least a first threshold. If the first threshold is exceeded, timing control logic may generate a low band correction pulse to adjust the power converter output voltage, and thereby adjust the error signal to a level within the first threshold. If the error signal exceeds a second threshold, the timing control logic may generate a high band correction pulse to adjust the power converter output voltage, and thereby adjust the error signal to a level within the second threshold. The timing control logic may initiate a low band blanking period following the low band correction pulse and high band blanking period following the high band correction pulse.

Abstract: An IC uses a tunable interconnect driver between a data source and a data destination to selectively slow down (“de-tune”) data signals. Data sent along relatively short paths are de-tuned to reduce power supply noise during synchronous switching events. In some embodiments, the tunable interconnect driver delays data transmission relative to an un-delayed signal path, in other embodiments, the slew rate of the tunable interconnect driver is selectively reduced.

Abstract: Transient processing mechanisms for power converters. Error generation circuitry in a power converter may generate an error signal based on the difference between a power converter output voltage and a reference voltage. Transient detection circuitry may detect whether the error signal exceeds at least a first threshold. If the first threshold is exceeded, timing control logic may generate a low band correction pulse to adjust the power converter output voltage, and thereby adjust the error signal to a level within the first threshold. If the error signal exceeds a second threshold, the timing control logic may generate a high band correction pulse to adjust the power converter output voltage, and thereby adjust the error signal to a level within the second threshold. The timing control logic may initiate a low band blanking period following the low band correction pulse and high band blanking period following the high band correction pulse.

Abstract: Transient processing mechanisms for power converters. Error generation circuitry in a power converter may generate an error signal based on the difference between a power converter output voltage and a reference voltage. Transient detection circuitry may detect whether the error signal exceeds at least a first threshold. If the first threshold is exceeded, timing control logic may generate a low band correction pulse to adjust the power converter output voltage, and thereby adjust the error signal to a level within the first threshold. If the error signal exceeds a second threshold, the timing control logic may generate a high band correction pulse to adjust the power converter output voltage, and thereby adjust the error signal to a level within the second threshold. The timing control logic may initiate a low band blanking period following the low band correction pulse and high band blanking period following the high band correction pulse.

Abstract: A packaged integrated circuit (“IC”) includes a substrate, an IC die, and a molded plastic lid. A test point standoff is electrically connected to the IC die and extends away from the surface of the package substrate through the molded plastic lid toward the top surface of the molded plastic lid. The top of the test point standoff is below the top surface of the molded plastic lid.

Abstract: A power converter including a hardware efficient control loop architecture. Error detection circuitry may generate an error signal based on the difference between a power converter output voltage and a reference voltage. An oversampling ADC may digitize the error signal. The transfer function associated with the ADC may include quantization levels spaced at non-uniform intervals away from a center code. A digital filter may calculate the average of the digitized error signal. A nonlinear requantizer may reduce the number of codes corresponding to the output of the digital filter. A proportional integral derivative (PID) unit may multiply the output of the nonlinear requantizer by PID coefficients to generate a PID duty cycle command, and a gain compensation unit may dynamically adjust the PID coefficients to maintain a constant control loop gain. A noise-shaped truncation unit including a multi-level error-feedback delta sigma modulator may reduce the resolution of the PID duty cycle command.