ADC Error Compensation Using Output Powers for Nonlinearity Correction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Analog-to-digital converters (ADCs) suffer from signal degradation due to circuit non-linearities, which introduce harmonic distortion and intermodulation products, reducing the signal-to-noise ratio and undermining the noise-shaping benefits of sigma-delta modulators.
Innovation Solution
Implementing an error correction system using a neural network to compensate for ADC non-linearities by analyzing powers of the ADC output, specifically correcting polynomial errors through scaled versions of the output signal.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If noise shaping is used in sigma-delta modulators to shift quantization noise to upper frequencies, then signal-to-noise ratio in baseband is improved, but circuit non-linearities fold the shaped noise back into baseband through intermodulation, undoing the noise shaping benefits
Solution Approach 1:
The patent measures the intermodulation distortion products generated by circuit non-linearities and uses these harmful signals as training data to train a neural network. The neural network learns to predict and cancel these distortion products, converting the harmful intermodulation effects into useful information for error correction.
Solution Approach 2:
The system implements a feedback mechanism where the ADC output is fed into a neural network that has been trained to predict distortion products. The neural network's prediction is then subtracted from the original signal to cancel the intermodulation distortion, creating a closed-loop error correction system that continuously compensates for non-linearities.
2Measurement precision
If polynomial error correction is applied to compensate for ADC non-linearities, then signal accuracy is improved, but additional processing complexity is introduced
Solution Approach 1:
The patent performs preliminary action by training the neural network offline using measured distortion products from the ADC. During actual operation, the pre-trained neural network simply predicts and cancels distortion products without requiring real-time measurement or complex calculations, significantly reducing online processing complexity while maintaining high signal accuracy.
Data Source
AI summary
A device may include an input terminal configured to receive an analog input signal. A device may include an output terminal configured to output a digital signal x, wherein the digital signal x includes a digital approximation of the analog input signal. A device may include an error correction system connected to the ADC, the error correction system including a first input terminal configured to receive an Nth powered version of the digital signal x, wherein N is a whole number equal to or greater than two, wherein the error correction system is configured to: use the Nth powered version of the digital signal x to determine a correction value; and modify the digital signal x to generate a corrected digital signal by applying the correction value to compensate for analog-to-digital conversion errors occurring within the ADC.