A digital to analog converter is constructed using a neural network layer. The converter has inputs for receiving parallel bits of a digital input signal and an output for outputting an analog signal which is based on the digital input. Connecting the input and the output is a neural network layer which is configured to convert the parallel bits into an output analog signal that is representative of the digital input signal. The neural network may be hardwired and the synapses may rely on memristors as programmable elements.

In accordance with an embodiment, a method for monitoring a data converter configured to convert data using a calibration determined by a calibration data record includes calibrating the data converter in order to determine a corresponding multiplicity of time associated calibration data records at a multiplicity of different times; and determining a state of the data converter based on comparing at least one of the multiplicity of time associated calibration data records with a comparison data record.

An N-channel interleaved Analog-to-Digital Converter (ADC) has a variable delay added to each ADC's input sampling clock. The variable delays are each programmed by a Successive-Approximation-Register (SAR) during calibration to minimize timing skews between channels. An auto-correlator generates a sign of a correlation error for a pair of ADC digital outputs. SAR bits are tested with the correlation sign bit determining when to add or subtract SAR bits. First all pairs are calibrated in a first level of a binary tree of mux-correlators. Then skews between remote pairs and groups are calibrated in upper levels of the binary tree using auto-correlators with inputs muxed from groups of ADC outputs input to the binary tree of mux-correlators. The binary tree of mux-correlators can include bypasses for odd and non-binary values of N. Sampling clock and component timing skews are reduced to one LSB among both adjacent channels and remote channels.

A successive approximation register analog to digital converter includes a charge injection digital to analog converter (DAC) circuit, a comparator circuit, and a control logic circuitry. The charge injection DAC circuit includes capacitors that sample input signals to generate first and second signals and charge injection circuits that selectively adjust the first or the second signals according to enable signals and decision signals. The comparator circuit compares the first and second signals to generate the decision signals. The control logic circuitry controls a circuit of the charge injection circuits to adjust the first and the second signals during an initial phase, in order to adjust a switching sequence of the circuit according to the decision signals corresponding to the initial phase, and generates the enable signals according to the decision signals and the adjusted switching sequence during a conversion phase to generate a digital output.

A successive-approximation register analog-to-digital converter (SAR ADC), a correction method and a correction system are provided. The SAR ADC generates an original weight value sequence according to multiple original weight values. The SAR ADC converts an analog time-varying signal to establish a transforming curve corresponding to the original weight values. In addition, the SAR ADC generates an offset value sequence according to an offset of the transforming curve, uses the offset value sequence to correct the original weight value sequence to generate a corrected weight value sequence, and uses multiple corrected weight values of the corrected weight sequence to improve linearity of the transforming curve.

A calibrating device can mitigate the static mismatch error of a digital-to-analog converter (DAC), and includes a digital code generating circuit, the DAC, an analog-to-digital converter (ADC), a filter circuit, an indicating circuit, and a statistical circuit. The digital code generating circuit generates a digital code of N digital codes. The DAC generates an analog signal corresponding to one of N signal levels according to the digital code. The ADC generates a digital signal according to the analog signal. The filter circuit generates a gradient value according to the difference between the digital code and the digital signal. The indicating circuit generates a selection signal according to the digital code. The statistical circuit learns from the selection signal that the gradient value is corresponding to a K.sup.th digital code of the N digital codes, and determines whether the K.sup.th digital code should be adjusted according to the gradient value.

An analog to digital converter (ADC) device includes ADC circuits, a calibration circuit and a skew adjusting circuit. The ADC circuits convert an input signal according to clock signals, to generate first quantized outputs. The calibration circuit calibrates the first quantized outputs to generate second quantized outputs. The skew adjusting circuit includes an estimating circuit and a feedback circuit. The estimating circuit analyzes the second quantized outputs to generate detection signals, wherein the detection signals are related to time difference information of the clock signals. The skew adjusting circuit outputs the detection signals as adjustment signals, wherein the adjustment signals are configured to reduce a clock skew of the ADC circuits. The feedback circuit analyzes the detection signals generated by the estimating circuit, to generate a feedback signal to the estimating circuit, wherein the estimating circuit is configured to adjust the detection signals according to the feedback signal.

A method for calibrating currents includes performing a first sorting operation on a plurality of first current sources according to current levels generated by the first current sources, performing a second sorting operation on a plurality of second current sources according to current levels generated by the second current sources, determining a first switching sequence for the first plurality of current sources according to a result of the first sorting operation, and determining a second switching sequence for the second plurality of current sources according to a result of the second sorting operation and the first switching sequence. The plurality of first current sources have a same target current value, and the plurality of second current sources have a same target current value.

A successive-approximation register analog-to-digital converter (SAR ADC), a correction method and a correction system are provided. The SAR ADC generates an original weight value sequence according to multiple original weight values. The SAR ADC converts an analog time-varying signal to establish a transforming curve corresponding to the original weight values. In addition, the SAR ADC generates an offset value sequence according to an offset of the transforming curve, uses the offset value sequence to correct the original weight value sequence to generate a corrected weight value sequence, and uses multiple corrected weight values of the corrected weight sequence to improve linearity of the transforming curve.

An N-channel interleaved Analog-to-Digital Converter (ADC) has a variable delay added to each ADC's input sampling clock. The variable delays are each programmed by a Successive-Approximation-Register (SAR) during calibration to minimize timing skews between channels. Each channel receives a sampling clock with a different phase delay. The sampling clocks are overlapping multi-phase clocks rather than non-overlapping. Overlapping the multi-phase clocks allows the sampling pulse width to be enlarged, providing more time for the sampling switch to remain open and allow analog voltages to equalize through the sampling switch. Higher sampling-clock frequencies are possible than when non-overlapping clocks are used. The sampling clock is boosted in voltage by a bootstrap driver to increase the gate voltage on the sampling switch, reducing the ON resistance. Sampling clock and component timing skews are reduced to one LSB among all N channels.