Abstract: A device having a capacitive sampling structure that allows for removal of sampling noise can be implemented in a variety of applications. Noise cancellation can be achieved by storing on an auto-zero capacitor a scaled replica of kT/C noise by a mechanism of correlated sampling. In an example embodiment, a set of switches can be arranged such that, in switching, scaled thermal noise, generated in an acquisition phase in which a voltage signal is input to an input capacitor structure, is captured on an output capacitor structure and, in a conversion phase, the captured thermal noise is cancelled or compensated from an output of the output capacitor structure.

Abstract: A device having a capacitive sampling structure that allows for removal of sampling noise can be implemented in a variety of applications. Noise cancellation can be achieved by storing on an auto-zero capacitor a scaled replica of kT/C noise by a mechanism of correlated sampling. In an example embodiment, a set of switches can be arranged such that, in switching, scaled thermal noise, generated in an acquisition phase in which a voltage signal is input to an input capacitor structure, is captured on an output capacitor structure and, in a conversion phase, the captured thermal noise is cancelled or compensated from an output of the output capacitor structure.

Abstract: Distortion in a combined sample and hold circuit and multiplexer can be reduced by dividing the sample and hold circuit and the multiplexer up into main and compensation signal channels, and considering the total error signal that arises during an acquire phase across both the switches of the multiplexer and the input switches of the sample and hold stage as a single error signal that has to be compensated. This compensation is then achieved by causing the same error voltages to be induced in both the main and compensation channels of the whole MUX and sample and hold circuit, such that errors can be made to cancel, thus improving the performance of the stage.

Abstract: Distortion in a combined sample and hold circuit and multiplexer can be reduced by dividing the sample and hold circuit and the multiplexer up into main and compensation signal channels, and considering the total error signal that arises during an acquire phase across both the switches of the multiplexer and the input switches of the sample and hold stage as a single error signal that has to be compensated. This compensation is then achieved by causing the same error voltages to be induced in both the main and compensation channels of the whole MUX and sample and hold circuit, such that errors can be made to cancel, thus improving the performance of the stage.

Abstract: A sigma delta analog-to-digital converter (ADC) circuit comprises a capacitive gain amplifier circuit having a first input to receive an input voltage and a second input; a loop filter circuit connected to an output of the capacitive gain amplifier circuit; a sub-ADC circuit including an output and an input connected to an output of the loop filter circuit; and a digital-to-analog (DAC) circuit including a DAC input connected to the output of the sub-ADC circuit, and a DAC output connected to the second input of the capacitive gain amplifier.

Abstract: An input stage to an analog to digital converter (ADC) includes at least one sampling capacitor (SC) for sampling an input signal in acquire phases, a capacitive gain amplifier (CGA) for providing the input signal to the SC, and bandwidth control means. The bandwidth control means is configured to ensure that the SC has a first bandwidth during a first part of an acquire phase and has a second bandwidth during a subsequent, second, part of said acquire phase, the second bandwidth being smaller than the first. In this manner, first, the input signal is sampled at a higher, first, bandwidth allowing to take advantage of using a high-bandwidth CGA to minimize settling error on the SC, and, next, during a second part of the same acquire phase, the input signal is sampled at a lower, second, bandwidth advantageously decreasing noise resulting from the use of a high-bandwidth CGA.