Digital correction techniques for data converters
View Patent ↗A method and apparatus is disclosed to compensate for impairments within a data converter such that its output is a more accurate representation of its input. The data converter includes a main data converter, a reference data converter, and a correction module. The main data converter may be characterized as having the impairments. As a result, the output of the main data converter is not the most accurate representation of its input. The reference data converter is designed such that the impairments are not present. The correction module estimates the impairments present within the main data converter using its output and the reference data converter to generate corrections coefficients. The correction module adjusts the output of the main data converter using the corrections coefficients to improve the performance of the data converter.
1. A composite analog-to-digital converter (ADC) for converting an analog input to a digital output, comprising:
a main data converter configured to convert the analog input to provide main digital samples, the main data converter having an impairment;
a reference data converter configured to convert the analog input to provide reference digital samples, the reference data converter not being impaired by the impairment; and
a correction module configured to:
quantify the impairment according to a model,
determine a solution to the model to provide a correction parameter, and
adjust the main digital samples based upon the correction parameter to compensate for the impairment to provide the digital output.
2. The composite ADC of claim 1 , wherein the reference data converter has a second impairment that is different from the impairment.
3. The composite ADC of claim 1 , wherein the reference data converter is further configured to sample the analog input at a different sampling rate or resolution than at which the main data converter samples the analog input.
4. The composite ADC of claim 3 , wherein the correction module is further configured to align the main digital samples and the reference digital samples in time.
5. The composite ADC of claim 1 , wherein the correction module comprises:
an estimation module configured to:
map one or more main digital samples from among the main digital samples and one or more reference digital samples from among the reference digital samples onto a coordinate system to develop the model, and
determine the solution to the model based upon a statistical relationship between the one or more mapped main digital samples and the one or more mapped reference digital samples; and
a compensation module configured to adjust the main digital samples based upon the correction parameter to provide the digital output.
6. The composite ADC of claim 5 , wherein the coordinate system comprises: a Cartesian coordinate system.
7. The composite ADC of claim 5 , wherein the compensation module comprises:
an adaptive filter configured to adapt its impulse response by updating an adaptive filtering coefficient in response to the correction parameter.
8. The composite ADC of claim 5 , wherein the model comprises:
a difference equation with an unknown parameter, the solution of the model representing a solution for the unknown parameter.
9. The composite ADC of claim 5 , wherein the statistical relationship comprises:
a line of best fit between the one or more mapped main digital samples and the one or more mapped reference digital samples;
a quadric polynomial of best fit between the one or more mapped main digital samples and the one or more mapped reference digital samples; or
a cubic polynomial of best fit between the one or more mapped main digital samples and the one or more mapped reference digital samples.
10. A composite analog-to-digital converter (ADC) for converting an analog input to a digital output, comprising:
a main data converter configured to convert the analog input to provide main digital samples;
a reference data converter configured to convert the analog input to provide reference digital samples; and
a correction module configured to:
map one or more main digital samples from among the main digital samples and one or more reference digital samples from among the reference digital samples on a coordinate system,
determine a statistical relationship between the one or more mapped digital samples and the one or more mapped reference digital samples,
determine a solution to a model of an impairment based upon the statistical relationship, and
adjust the main digital samples based upon the solution to compensate for the effect of the impairment to provide the digital output.
11. The composite ADC of claim 10 , wherein the coordinate system comprises:
a Cartesian coordinate system.
12. The composite ADC of claim 10 , wherein the correction module comprises:
an adaptive filter configured to adapt its impulse response by updating an adaptive filtering coefficient in response to the solution.
13. The composite ADC of claim 10 , wherein the statistical relationship comprises:
a line of best fit between the one or more mapped main digital samples and the one or more mapped reference digital samples;
a quadric polynomial of best fit between the one or more mapped main digital samples and the one or more mapped reference digital samples; or
a cubic polynomial of best fit between the one or more mapped main digital samples and the one or more mapped reference digital samples.
14. The composite ADC of claim 10 , wherein the main data converter is configured to convert the analog input at a first rate to provide the main digital samples, and
wherein the reference data converter is configured to convert the analog input at a second rate that is different from the first rate to provide the reference digital samples.
15. The composite ADC of claim 14 , wherein the correction module is further configured to interpolate or average between samples of the digital samples to align in time the reference digital samples and the adjusted main digital samples.
16. A method for converting an analog input to a digital output, comprising:
converting, by a composite analog-to-digital converter (ADC), the analog input to provide main digital samples;
converting, by the composite ADC, the analog input to provide reference digital samples;
mapping, by the composite ADC, one or more main digital samples from among the main digital samples and one or more reference digital samples from among the reference digital samples on a coordinate system:
determining, by the composite ADC, a statistical relationship between the one or more mapped digital samples and the one or more mapped reference digital samples,
determining, by the composite ADC, a solution to a model of an impairment based upon the statistical relationship; and
adjusting, by the composite ADC, the main digital samples based upon the solution to compensate for the impairment to provide the digital output.
17. The method of claim 16 , wherein the mapping comprises:
mapping the one or more main digital samples and the one or more reference digital samples as one or more points on a Cartesian coordinate system to develop the model.
18. The method of claim 16 , wherein the adjusting comprises:
adapting an impulse response of the composite ADC by updating an adaptive filtering coefficient in response to the solution.
19. The method of claim 16 , wherein the determining the statistical relationship comprises:
determining:
a line of best fit between the one or more mapped main digital samples and the one or more mapped reference digital samples;
a quadric polynomial of best fit between the one or more mapped main digital samples and the one or more mapped reference digital samples; or
a cubic polynomial of best fit between the one or more mapped main digital samples and the one or more mapped reference digital samples.
20. The method of claim 16 , wherein the converting the analog input to provide the main digital samples comprises:
converting the analog input at a first rate to provide the main digital samples, and
wherein the converting the analog input to provide the reference digital samples comprises:
converting the analog input at a second rate that is different from the first rate to provide the reference digital samples.