User transparent self-calibration technique for pipelined ADC architecture
AssigneeTexas Instruments Incorporated
InventorPaul C. Yu
Life of this patent
Estimating the term…
Abstract
A user transparent self-calibration technique for an analog to digital converter is described. The technique can correct for capacitor mismatch error with minimal additional power consumption. This is done by generating a calibration signal, one for each capacitor whose calibration is desired. The signal is interleaved with the input signal, and digitized by alternating with the input signal digitization using capacitor arrays.
Claims31
A method for performing analog to digital conversion, comprising the steps of:
- in first stage of an analog-to-digital converter:
- sampling an analog signal;
- digitizing said analog signal to produce a digital output;
- generating a residue signal relative to said digital output;
- and
- generating a calibration signal relative to the capacitor mismatch of said first stage;
- wherein said generating steps are interleaved in a manner such that the throughput of said analog-to-digital converter is not decreased.
A method for performing analog-to-digital conversion, comprising the steps of:
- performing a first stage of analog-to-digital conversion on an analog input signal in an analog-to-digital converter;
- producing a residue signal;
- generating a calibration signal;
- in successive stages of said analog-to-digital converter:
- digitizing said residue signal to obtain a coarse digital value;
- and
- digitizing said calibration signal to obtain a coarse digital calibration value;
- using said coarse digital calibration values to correct said coarse digital values;
- and
- producing a calibrated digital output from said analog-to-digital converter;
- wherein said calibration signal represents the capacitor mismatch of a capacitor used to generate said residue signal and said residue producing step and said generating step are interleaved in a manner such that the throughput of said analog-to-digital converter is unchanged.
A method for converting analog signals to digital signals, in a converter, comprising the steps of:
- sampling an analog signal in an analog-to-digital converter to produce a sampled signal;
- alternately generating a residual signal from said sampled signal and a calibration signal in at least the most significant bit stage of said analog-to-digital converter;
- alternately digitizing said residual signal and said calibration signal to obtain digital values in successive stages of said analog-to-digital converter;
- and
- correcting the output of said digitized residual signal with said digital results;
- wherein during said sampling step, a substitute capacitor is switched in place of a capacitor generating said calibration signal and said calibration signal represents the capacitor mismatch of a capacitor used to generate said signals and digitization of said residual signal and said calibration signal is interleaved in a manner such that the sample rate of said analog-to-digital converter is maximized.
An analog-to-digital converter, comprising:
- a first stage of an analog-to-digital converter connected to receive and sample an analog input and produce a coarse digital representation of said analog input, a first residue signal, and a first calibration signal which represents the capacitor mismatch of a capacitor used to generate said residue signal;
- and
- one or more subsequent stages of said analog-to-digital converter, connected to receive a residue signal and a calibration signal from an immediately preceding stage and produce coarse digital representations of said residue signal and said calibration signal and successive residue and calibration signals;
- wherein said first stage production of said residue signal and said calibration signal is interleaved such that the throughput of said analog-to-digital converter is not decreased.
Dependencies are inferred from the claim text.