Abstract:
Methods and systems for time interleaved analog-to-digital converter timing mismatch calibration and compensation may include receiving an analog signal on a chip, converting the analog signal to a digital signal utilizing a time interleaved analog-to-digital-converter (ADC), and reducing a blocker signal that is generated by timing offsets in the time interleaved ADC by estimating complex coupling coefficients between a desired digital output signal and the blocker signal utilizing a decorrelation algorithm on frequencies within a desired frequency bandwidth. The decorrelation algorithm may comprise a symmetric adaptive decorrelation algorithm. The received analog signal may be generated by a calibration tone generator on the chip. An aliased signal may be summed with an output signal from a multiplier. The complex coupling coefficients may be determined utilizing the decorrelation algorithm on the summed signals. A multiplier may be configured to cancel the blocker signal utilizing the determined complex coupling coefficients.
Abstract:
An asynchronous successive approximation register analog-to-digital converter (SAR ADC), which utilizes one or more overlapping redundant bits in each digital-to-analog converter (DAC) code word, is operable to generate an indication signal that indicates completion of each comparison step and indicates that an output decision for each comparison step is valid. A timer may be initiated based on the generated indication signal. A timeout signal may be generated that preempts the indication signal and forces a preemptive decision, where the preemptive decision sets one or more remaining bits up to, but not including, the one or more overlapping redundant bits in a corresponding digital-to-analog converter code word for a current comparison step to a particular value. For example, the one or more remaining bits may be set to a value that is derived from a value of a bit that was determined in an immediately preceding decision.
Abstract:
Systems and methods are provided for jitter improvement in serializer-deserializer (SerDes) transmitters. One or more adjustments may be applied in SerDes transmitter circuitry to reduce jitter in a serial output of the SerDes transmitter circuitry, which may occur as a result of processing of input data. Applying the one or more adjustments may comprise use of dummy data. The dummy data may be configured to generate corresponding dummy current pulses which may in turn be used in controlling supply variations occurring during processing of the input data and/or generation of the serial output. The dummy data may be configured to generate the dummy current pulses such that they are applied along with current pulses corresponding to the input data. The dummy data may be adaptively set or adjusted based on the input data. The use of the dummy data may be selectively turned on or off.
Abstract:
An asynchronous successive approximation register analog-to-digital converter (SAR ADC), which utilizes one or more overlapping redundant bits in each digital-to-analog converter (DAC) code word, is operable to generate an indication signal that indicates completion of each comparison step and indicates that an output decision for each comparison step is valid. A timer may be initiated based on the generated indication signal. A timeout signal may be generated that preempts the indication signal and forces a preemptive decision, where the preemptive decision sets one or more remaining bits up to, but not including, the one or more overlapping redundant bits in a corresponding digital-to-analog converter code word for a current comparison step to a particular value. For example, the one or more remaining bits may be set to a value that is derived from a value of a bit that was determined in an immediately preceding decision.
Abstract:
Systems and methods are provided for jitter improvement in serializer-deserializer (SerDes) transmitters. One or more adjustments may be applied in SerDes transmitter circuitry to reduce jitter in a serial output of the SerDes transmitter circuitry, which may occur as a result of processing of input data. Applying the one or more adjustments may comprise use of dummy data. The dummy data may be configured to generate corresponding dummy current pulses which may in turn be used in controlling supply variations occurring during processing of the input data and/or generation of the serial output. The dummy data may be configured to generate the dummy current pulses such that they are applied along with current pulses corresponding to the input data. The dummy data may be adaptively set or adjusted based on the input data. The use of the dummy data may be selectively turned on or off.
Abstract:
Methods and systems are provided for controlling signal processing outputs. In signal processing circuitry, searching through a plurality of quantization levels for a quantization level that matches an analog input, and when the search fails within a particular amount of time, adjusting at least a portion of an output of the signal processing circuitry. The adjusting comprises setting the at least portion of the output to a predefined value. Setting the output, or portions thereof, may comprise selecting between output of a normal processing path and output of a code generation path configured for handling search failures. Timing information may be generated for use in controlling generating of the output of the signal processing circuitry. The timing information may be used in measuring per-cycle operation time during the search through the plurality of quantization levels.
Abstract:
Receiver architectures and methods of processing harmonic rich input signals employing harmonic suppression mixers are disclosed herein. The disclosed receivers, mixers, and methods enable a receiver to achieve the advantages of switching mixers while greatly reducing the mixer response to the undesired harmonics. A harmonic mixer can include a plurality of mixers coupled to an input signal. A plurality of phases of a local oscillator signal can be generated from a single local oscillator output. Each of the phases can be used to drive an input of one of the mixers. The mixer outputs can be combined to generate a frequency converted output that has harmonic rejection.
Abstract:
Methods and systems are provided for asynchronous successive approximation register (SAR) analog-to-digital converters (ADCs) that utilize preemptive bit setting decisions. In particular, such SAR ADC may be operable to, when a failure to determine a valid output decision for each comparison step occurs, set one or more remaining bits, up to but not including one or more overlapping redundant bits in a code word corresponding to the comparison step, to a particular value. The value may be derived from a value of a bit determined in an immediately preceding decision. The failure may be determined based on dynamic and/or adaptive criteria. The criteria may be set, e.g., so as to guarantee that a magnitude of a difference between an analog input voltage to the SAR ADC and analog output voltage of a digital-to-analog converter (DAC) used therein is within overlapping ranges of voltages corresponding to the overlapping redundant bits.
Abstract:
Receiver architectures and methods of processing harmonic rich input signals employing harmonic suppression mixers are disclosed herein. The disclosed receivers, mixers, and methods enable a receiver to achieve the advantages of switching mixers while greatly reducing the mixer response to the undesired harmonics. A harmonic mixer can include a plurality of mixers coupled to an input signal. A plurality of phases of a local oscillator signal can be generated from a single local oscillator output. Each of the phases can be used to drive an input of one of the mixers. The mixer outputs can be combined to generate a frequency converted output that has harmonic rejection.
Abstract:
A phase locked loop may be operable to generate, utilizing a frequency doubler, a reference clock signal whose frequency is twice a frequency of a crystal clock signal and is keyed on both rising and falling edges of the crystal clock signal. The phase locked loop may enable usage of both rising and falling edges of the crystal clock signal, based on the generated reference clock signal. The phase locked loop may perform an operation of the phase locked loop based on the enabling. The phase locked loop may perform a phase comparison function, based on both rising and falling edges of the crystal clock signal. By utilizing a sampled loop filter in the phase locked loop, the phase locked loop may eliminate, at an output of a charge pump in the phase locked loop, disturbance which is associated with duty cycle errors of the crystal clock signal.