ADC Nonlinearity Testing Using Counter-Based Code Histograms
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Characterizing analog-to-digital converters (ADCs) is challenging due to noise sensitivity in measurement setups, particularly affecting differential non-linearity (DNL) and integral non-linearity (INL) metrics, which are crucial for ADC performance but difficult to measure accurately, especially in high-speed applications.
Innovation Solution
A counter-based measurement circuit that stores counts of digital codes outputted by the ADC during non-linearity testing, rather than raw code data, to reduce noise interference and data volume, using a decoder to increment corresponding counters for each code, allowing for precise DNL and INL characterization with reduced transistor count and noise immunity.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If raw code data with metadata (e.g., timestamps) is stored for each ADC output, then complete characterization information is preserved, but noise interference increases and data volume becomes large
Solution Approach 1:
The patent extracts only the essential counting information from the complete raw data, storing only the number of times each code appears rather than the full raw code data with metadata. This extraction process removes unnecessary information that contributes to noise and data volume while preserving the essential characterization data needed for DNL and INL measurements.
Solution Approach 2:
Instead of storing actual raw code data, the patent creates a simplified copy in the form of count values that represent the frequency of each code. This copying approach maintains the statistical information needed for analysis while eliminating the noise and large data volume associated with storing complete raw datasets.
2Measurement precision
If raw code data with metadata is stored for each ADC output, then complete characterization information is preserved, but data volume becomes large
Solution Approach 1:
The patent extracts only the essential counting information from the complete raw data, storing only the number of times each code appears rather than the full raw code data with metadata. This extraction process removes unnecessary information that contributes to noise and data volume while preserving the essential characterization data needed for DNL and INL measurements.
Solution Approach 2:
The patent discards the raw code data and metadata after extracting the count information, then recovers the necessary characterization data from these counts. This approach eliminates large data volumes while maintaining the essential information needed for ADC non-linearity analysis.
3Ease of manufacture
If conventional measurement setups are used for ADC characterization, then standard measurement procedures are followed, but noise sensitivity prevents accurate DNL and INL metrics
Solution Approach 1:
The patent replaces the conventional mechanical/electrical measurement setup with a counter-based digital system. Instead of using complex analog measurement equipment that is sensitive to noise, the invention uses digital counters to directly tally code occurrences, substituting a noisy analog measurement system with a robust digital counting system that is inherently more precise and noise-immune.
Data Source
AI summary
In some examples, a device includes an analog-to-digital converter (ADC) configured to receive an analog signal and output digital codes based on values of the analog signal. The device also includes a plurality of counters, where each counter of the plurality of counters is configured to increment in response to a respective digital code outputted by the ADC.