Correction of non-linearities in ADCS
View Patent ↗Techniques for calibrating non-linearities of ADCs are described, which can be applied whether or not the non-linearities change with frequency. When the non-linearities do not change (are static), the frequency of a calibrating signal is first estimated coarsely in a calibration mode, then a fine estimate is determined using the coarse estimate. These estimates are then used to predict the sinusoidal signal using a linear predictor. A Look Up Table (LUT) containing corrections to the ADC is derived from this result. The LUT is then used in a normal operating mode to correct the output of the ADC. In a case where the characteristics of the non-linearities of the input signal are dynamic and thus change with frequency, a frequency spectrum of interest is broken into several regions. In each of these regions, a frequency is identified and used as a calibrating signal to generate the corresponding LUT. During normal operation of the ADC, in a first method, the bin corresponding to dominant frequency of the signal is identified using a short-length FFT. This bin is used to select the appropriate LUT for operating on the output of the ADC to provide the calibrated output. In a second method used when dynamic input is expected, a single LUT is developed using the averages values from the LUTs determined from the various regions.
1. A method for correcting errors in an Analog to Digital Converter (ADC) comprising:
applying a test signal s(k) to an input of the ADC, thereby producing an ADC output signal x(k) as a discrete time sampled signal;
estimating a frequency of the test signal from the ADC output signal;
predicting further samples, s′(k), of the discrete time sampled signal from the ADC output signal; and
determining a set of ADC correction values from the predicted further samples s′(k),
wherein estimating a frequency of the test signal further comprises:
(a) determining a coarse frequency estimate, F in c from the ADC output signal x(k); and
(b) determining a fine frequency estimate F in ƒ from the coarse frequency estimate, and
wherein determining a fine frequency estimate further comprises:
(c) filtering the ADC output signal with a notch filter, the notch filter tuned to a notch frequency that depends on the coarse frequency estimate; and
(d) detecting a power output from the notch filter.
2. The method of claim 1 wherein the step of determining a coarse frequency estimate further comprises:
counting zero crossings in the ADC output signal, x(k), over a predetermined time interval.
3. The method of claim 1 wherein the notch filter is a second order all pass digital filter.
4. The method of claim 1 wherein filtering the ADC output signal further comprises:
selecting a set of frequencies depending on the coarse frequency;
searching for an optimal frequency within the set of frequencies by the further steps of:
tuning the notch filter to two or more selected frequencies within the set of frequencies; and
comparing a detected power level of the notch filter at the selected frequencies; and
selecting the optimal frequency as the fine frequency estimate.
5. The method of claim 1 wherein the step of predicting further comprises:
filtering selected samples of the ADC output signal, x(k), spaced by a distant sample spacing value, L, with a recursive digital filter.
6. The method of claim 5 wherein the recursive digital filter is a two coefficient, linear prediction filter.
7. The method of claim 1 wherein the correction values are determined by minimizing a mean square error between s(k) and s′(k).
8. The method of claim 1 additionally comprising:
dividing an expected input frequency range into a number of frequency regions;
applying a test signal for each of the frequency regions; and
determining a set of correction values for each of the frequency regions.
9. The method of claim 8 additionally comprising:
averaging the correction values for each of the frequency regions, to determine a single set of correction values.
10. The method of claim 8 additionally comprising the step of:
during an operational mode of the ADC,
detecting a dominant frequency in the ADC input signal, s(k); and
selecting one of the sets of correction values for a frequency region containing the dominant frequency.
11. The method of claim 8 wherein the ADC correction amounts for the number of frequency regions are stored in a corresponding number of Look Up Tables.
12. The method of claim 1 wherein the ADC correction values are stored in a Look Up Table.
13. An apparatus comprising:
a sample and hold circuit, connected to receive a test signal, s(t), and output a discrete time sampled signal s(k);
an Analog to Digital Converter (ADC), connected to receive the discrete time sampled signal s(k) and output an ADC output signal, x(k);
a frequency detector, connected to receive the ADC output signal and to provide a frequency estimate;
a linear prediction filter, connected to receive the frequency estimate and the ADC output signal, to provide a set of predicted signal samples, s′(k); and a Look Up Table (LUT), storing correction values for the ADC, the correction values determined from the predicted signal samples s′(k); and
wherein the frequency detector further comprises:
a zero crossing detector, connected to receive the ADC output signal and to provide a coarse frequency estimate;
a tunable notch filter, tuned to a notch frequency that depences on the coarse frequency estimate; and
a power level detector, connected to the tunable notch filter.
14. The apparatus of claim 13 wherein the linear prediction filter is a recursive linear digital filter that operates on samples of the ADC output signal spaced by number, L, of samples.
15. The apparatus of claim 13 additionally comprising:
a plurality of frequency detectors, for providing a plurality of frequency estimates for a plurality of test signals, each test signal having dominant frequency content in a frequency region;
a plurality of linear prediction filters, connected to receive the plurality of frequency estimates and the ADC output signal; and
a plurality of Look Up Tables, for storing correction values corresponding to corrections to be applied for input signals in corresponding frequency regions.
16. A computer program product comprising:
a computer readable storage medium having stored thereon a set of computer readable instructions embodied in computer readable program code that, when executed by a processor, causes the processor to correct the operation of an Analog to Digital Converter (ADC) by:
receiving samples of an ADC output signal x(k) produced as a discrete time sampled signal in response to a test signal s(k) applied to the ADC;
estimating a frequency of the test signal from the ADC output signal samples;
predicting further samples, s′(k), of the discrete time sampled signal from the ADC output signal; and
determining a set of ADC correction values from the predicted further samples s′(k)
wherein estimating a frequency of the test signal further comprises:
(a) determining a coarse frequency estimate, F in c from the ADC output signal x(k); and
(b) determining a fine frequency estimate F in ƒ from the coarse frequency estimate,
wherein determining a fine frequency estimate further comprises:
(c) filtering the ADC output signal with a notch filter, the notch filter tuned to a notch frequency that depends on the coarse frequency estimate; and
(d) detecting a power output from the notch filter,
wherein filtering the ADC output signal further comprises:
(e) selecting a set of frequencies depending on the coarse frequency; and
wherein searching for an optimal frequency within the set of frequencies further comprises:
(f) tuning the notch filter to two or more selected frequencies within the set of frequencies; and
(g) comparing a detected power level of the notch filter at the selected frequencies; and
(h) selecting the optimal frequency as the fine frequency estimate.