Charge Redistribution ADC Noise Filtering by Multi-Sample Averaging
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Charge redistribution analog-to-digital converters are sensitive to noise in analog signals, leading to degraded image quality due to random signal fluctuations, and increasing resolution to mitigate noise degrades the image definition.
Innovation Solution
The solution involves recording multiple values of the signal at each sampling instant and averaging them to filter out noise, using a network of capacitors with a geometric progression series of capacitances and switching circuits that allow capacitors to float, effectively reducing noise sensitivity while maintaining resolution.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If the resolution of the converter is increased to mitigate noise, then noise sensitivity is reduced, but image definition is degraded
Solution Approach 1:
The patent divides the conversion process into multiple independent conversion cycles, where each cycle converts a portion of the signal. By segmenting the analog signal into multiple samples and converting them separately, the system reduces noise sensitivity through averaging while maintaining image definition through controlled charge redistribution.
Solution Approach 2:
The patent employs periodic sampling and conversion cycles, where the analog signal is sampled at multiple time points and each sample is converted in sequential periods. This periodic action allows noise to be averaged out across cycles while preserving the essential signal characteristics for maintaining image definition.
2Measurement precision
If multiple samples are recorded and averaged to filter noise, then noise sensitivity is reduced, but conversion time is increased
Solution Approach 1:
The patent maintains continuous charge redistribution operations across multiple conversion cycles. Instead of completing one full conversion before starting the next, the system continuously redistributes charges from previous samples during the conversion process, ensuring that useful action (charge redistribution) continues without interruption and reducing idle time.
Solution Approach 2:
The patent performs preliminary charging of capacitors during the sampling phase before the actual conversion begins. By pre-charging the capacitive network with the sampled signal, the system prepares the charge distribution in advance, so that when conversion starts, the heavy lifting of charge redistribution is already partially complete, reducing the time needed for the actual conversion process.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This approach filters noise from the signal, improving the accuracy of digital conversion and maintaining image quality by averaging multiple samples, thus enhancing the robustness of the charge redistribution analog-to-digital conversion process.
Implementation Method 1
Charge redistribution converters include capacitors which are charged in a first sampling step. Then in a second digitization step, the electrical charges are transferred from one capacitor to another.
Data Source
AI summary
An N-bit type charge redistribution analog-to-digital conversion device includes an input terminal configured to receive an input signal and coupled via a line to an output terminal. The output terminal is configured to be coupled to a comparator. The device further includes three reference potential sources of different values and a network of capacitors, where a first terminal of each capacitor is coupled to the line, and where a second terminal of each capacitor is coupled to switching circuit configured for coupling the second terminal of each capacitor to one of the reference potentials.