ADC Calibration Using Multi-Stage Error Extraction and Trimming
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional error testing and calibration methods for analog-to-digital converters are inadequate for multi-stage and multi-error calibration, and their efficiency is low, particularly in high-speed and high-precision converters, due to limitations in extracting errors across multiple conversion stages and gain errors.
Innovation Solution
A method involving initial performance testing, pre-trimming, error extraction, soft trimming, and hard trimming is employed to calibrate analog-to-digital converters, including configuring input signals and output modes to extract and correct multiple errors, with adaptive parameter extraction to reduce sample points and improve efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If conventional error testing and calibration methods are used, then the testing process is simple, but multi-stage and multi-error calibration cannot be performed
Solution Approach 1:
The patent segments the calibration process into distinct stages (initial performance test, pre-trimming, error extraction, soft trimming, hard trimming) and separates different error types (weight mismatch error, gain error) for targeted extraction and correction. This segmentation enables multi-stage and multi-error calibration while maintaining process manageability through structured division of tasks.
Solution Approach 2:
The patent performs preliminary actions by conducting initial performance tests and pre-trimming before main error extraction. This preliminary calibration removes obviously defective circuits early, reducing the burden on subsequent multi-stage calibration processes and enabling more complex error extraction to be performed efficiently on pre-screened converters.
2Measurement precision
If conventional error extraction method based on INL error is used, then the extraction process is straightforward, but sample points increase exponentially with precision improvement
Solution Approach 1:
The patent extracts specific error components (weight mismatch error and gain error) separately from the overall conversion process. By taking out and addressing each error type independently through dedicated extraction methods, the system achieves high-precision calibration without requiring exponential increases in sample points, thereby maintaining testing efficiency.
Solution Approach 2:
The patent changes the calibration approach by switching from traditional INL-error-based methods to parameter-specific extraction methods that target weight mismatch and gain errors. This parameter change enables efficient extraction of multiple error types without the exponential sample point increase that would be required by conventional precision improvement methods.
3Speed
If high-speed manufacturing process is adopted, then the converter speed increases, but gain errors increase due to insufficient gains
Solution Approach 1:
The patent implements feedback by extracting gain errors through systematic measurement and then applying corrective trimming based on the extracted error data. This closed-loop feedback mechanism compensates for gain inaccuracies introduced by high-speed manufacturing processes, allowing the converter to maintain both high speed and accurate gain performance.
Solution Approach 2:
The patent replaces physical manufacturing precision improvements with digital error extraction and correction mechanisms. Instead of relying solely on manufacturing processes to achieve accurate gains, the system uses digital calibration to extract and correct gain errors, enabling high-speed operation without sacrificing gain accuracy.
Data Source
AI summary
A method for calibrating an analog-to-digital converter includes the following steps: conducting an initial performance test and judgement on the analog-to-digital converter; if the initial performance test succeeds, performing a pre-trimming and judgement on the analog-to-digital converter; if the pre-trimming succeeds, performing an error extraction on the analog-to-digital converter, obtaining errors of conversion stages of the analog-to-digital converter; performing an error soft trimming and test on the analog-to-digital converter according to the errors of the conversion stages; and if the error soft trimming and test of the analog-to-digital converter succeed, performing an error hard trimming and test on the analog-to-digital converter according to the errors of the conversion stages.