Offset calibration of analog-to-digital converters using a spectrum analyzer
Approaches provide for calibrating high speed analog-to-digital converters (ADCs). For example, a calibration signal can be applied to parallel ADCs. The output of the parallel ADCs can be analyzed using a gradient-based optimization approach or other such optimization approach to determine optimized gain error calibration data to compensate for gain mismatch in and between individual parallel time-interleaved ADCs and to determine time-offset calibration data to compensate for timing errors in and between individual parallel time-interleaved ADCs. For example, once a calibration signal is applied to an ADC, the output of the ADC can be analyzed to determine a spectrum of the calibration signal. One or more images (e.g., phasors) of the spectrum can be determined and used to determine initial values of the optimization. Thereafter, the optimization approach can be utilized to determine optimized gain error calibration data and optimized time-offset calibration data, which can be stored and/or used to calibrate individual time-interleaved ADCs.
1. A computing system, comprising:
a computing device processor;
a memory device including instructions that, when executed by the computing device processor, enables the computing system to:
receive a calibration tone to a set of analog-to-digital converters (ADCs), the set of ADCs associated with an initial pre-determined value of gain error and an initial pre-determined value of time-offset;
determine an ADC output spectrum based at least in part on the calibration tone, the initial pre-determined value of gain error, and the initial pre-determined value of time-offset;
determine a first set of image phasors and a set of mismatch image phasors based at least in part on the ADC output spectrum;
determine a first image power based at least in part on the first set of image phasors;
determine the first image power does not satisfy an image power threshold;
determine gain error calibration data and time-offset calibration data based at least in part on image power gradients with respect to the set of mismatch image phasors;
apply the gain error calibration data and the time-offset calibration data to determine a second set of image phasors;
determine second image power measurements based at least in part on a second set of image phasors; and
apply the gain error calibration data and the time-offset calibration data to an output of the ADCs based at least in part on whether the second image power measurements satisfy the image power threshold.
2. The computing system of claim 1 , wherein the instructions, when executed by the computing device processor, further enables the computing system to:
determine the second image power measurements fail to satisfy the image power threshold; and
determine updated gain error calibration data and updated time-offset calibration data based at least in part on image power gradients with respect to mismatch sets of image phasors, wherein gain error and time-offset determination iterates until respective image power measurements satisfy the image power threshold.
3. The computing system of claim 1 , wherein the instructions, when executed by the computing device processor, further enables the computing system to:
determine the second image power measurements satisfy the image power threshold;
receive analog inputs to the set of ADCs; and
apply the gain error calibration data and the time-offset calibration data to an output of the ADCs.
4. The computing system of claim 1 , further comprising:
a signal generator configured to generate at least the calibration tone;
a switch configured to operate in at least one of a first position and a second position, wherein the switch, when operating in the first position, is configured to couple an input source with a set of ADCs, and wherein the switch, when operating in the second position, is configured to couple the signal generator with the set of ADCs;
a gain correction component configured to apply the gain error calibration data;
a time-offset correction component configured to apply the time-offset calibration data; and
a time-offset and gain estimator configured to determine one of the gain error calibration data or the time-offset calibration data.
5. The computing system of claim 4 , wherein the time-offset and gain estimator further includes:
a spectrum analyzer component configured to determine the ADC output spectrum;
an image power component configured to determine a total image power of at least the first set of image phasors; and
a gradient-based optimization component configured to optimize the gain error calibration data and time-offset calibration data based at least in part on an image power gradient with respect to a corresponding set of mismatch image phasors.
6. A computer-implemented method, comprising:
obtaining a signal during a first mode;
providing initial gain error calibration data and initial time-offset calibration data;
using a gradient-based optimization technique to determine optimized gain error calibration data and optimized time-offset calibration data based at least in part on the initial gain error calibration data, the initial time-offset calibration data, and the signal; and
applying the optimized gain error calibration data and the optimized time-offset calibration data during a second mode.
7. The computer-implemented method of claim 6 , further comprising:
using a signal generator to generate the signal, the signal associated with at least one characteristic, the at least one characteristic including a determined fundamental frequency.
8. The computer-implemented method of claim 6 , further comprising:
using a switch to operate in at least one of a first position in a calibration mode or a second position in a normal mode, the first mode corresponding to the calibration mode, the second mode corresponding to the normal mode, wherein the switch, when operating in the first position, is configured to couple an input source with a set of time-interleaved ADCs, and wherein the switch, when operating in the second position, is configured to couple a signal generator with the set of time-interleaved ADCs.
9. The computer-implemented method of claim 6 , wherein using the gradient-based optimization technique includes:
determining whether a value of a total image power satisfies an image power threshold; and
iteratively adjusting gain error calibration data and time-offset calibration data based at least in part on the value of the total image power when the value of the total image power fails to satisfy the image power threshold.
10. The computer-implemented method of claim 6 , further comprising:
receiving the signal at a set of analog-to-digital converters (ADCs); and
generating a digital signal, the initial gain error calibration data, the initial time-offset calibration data, and a value of a total image power based at least in part on the digital signal.
11. The computer-implemented method of claim 6 , further comprising:
using a spectrum analyzer component to determine a power spectral density of a digital signal, the power spectral density including one or more image components and a main component.
12. The computer-implemented method of claim 11 , further comprising:
using an image power component to determine a total image power of a set of image power components of the power spectral density.
13. The computer-implemented method of claim 12 , further comprising:
using a gradient-based optimization component to optimize gain error calibration data and time-offset calibration data based at least in part on a gradient of the total image power.
14. A computing system, comprising:
a signal generator configured to generate a calibration tone;
a switch configured to operate in at least one of a first position and a second position, wherein the switch, when operating in the first position, is configured to couple an input source with a set of time-interleaved ADCs, and wherein the switch, when operating in the second position, is configured to couple the signal generator with the set of time-interleaved ADCs;
a set of time-interleaved ADCs that include a number of parallel ADCs, the set of time-interleaved ADCs associated with an initial pre-determined value of gain error and an initial pre-determined value of time-offset;
a spectrum analyzer component configured to determine an output spectrum of the set of time-interleaved ADCs based at least in part on calibration tone, the initial pre-determined value of gain error, and the initial pre-determined value of time-offset when the switch is operating in the first position;
an image power component configured to determine total image power and mismatch image phasors based at least in part on the output spectrum;
a time-offset and gain estimator configured to use the mismatch image phasors to determine one gain error calibration data and time-offset calibration data;
a gradient-based optimization component configured to optimize the gain error calibration data and the time-offset calibration data based at least in part on a gradient of the total image power;
a gain correction component configured to apply gain error calibration data; and
a time-offset correction component configured to apply time-offset calibration data.
15. The computing system of claim 14 , wherein the time-offset and gain estimator are further configured to:
determine a total image power of a set of image power components of the output spectrum;
determine whether a value of a total image power satisfies an image power threshold; and
iteratively adjust gain error calibration data and time-offset calibration data based at least in part on the value of the total image power when the value of the total image power fails to satisfy the image power threshold.
16. The computing system of claim 14 , further comprising:
a computing device processor;
a memory device including instructions that, when executed by the computing device processor, further enables the computing system to:
use the time-offset and gain estimator to determine optimized gain calibration data and optimized time-offset calibration data during a first mode of operation; and
use the gain correction component and the time-offset correction component to apply the optimized gain calibration data and the optimized time-offset calibration data to an output of a set of time-interleaved ADCs during a second mode of operation.
17. The computing system of claim 16 , wherein the instructions, when executed by the computing device processor, further enables the computing system to:
iteratively adjust gain error calibration data and time-offset calibration data until at least one threshold is satisfied.
18. The computing system of claim 14 , wherein the calibration tone is a predetermined signal, the predetermined signal associated with a predetermined fundamental frequency.