Segmented DAC Array Sequencing for Low-Noise Linearity
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
High-resolution digital-to-analog converters face challenges in achieving highly linear characteristics and low quantization noise due to the complexity of components required for accurate conversion.
Innovation Solution
A digital-to-analog converter apparatus and method that involves resampling digital signals to lower resolution and higher sampling rates, using a DAC array with unit elements activated in a probabilistically disrupted cyclical sequence to contribute equally to the analog signal, and optional filtering to reduce noise.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If high-resolution digital signals are converted to analog signals using conventional methods, then the conversion accuracy is improved, but the device complexity and quantization noise increase
Solution Approach 1:
The DAC is divided into multiple unit elements (e.g., 8 unit elements for 3-bit resolution) that are activated in a cyclical sequence. Each unit element contributes a fraction of the total resolution, and their combined output achieves the desired high-resolution conversion without requiring complex high-resolution components in each element.
Solution Approach 2:
The unit elements are activated in a cyclical sequence over multiple sampling periods. This periodic activation pattern allows the system to achieve high-resolution conversion through time-averaging, where the analog output is the sum of contributions from multiple unit elements activated at different times in the cycle.
2Measurement precision
If conventional DAC methods are used with high-resolution components, then conversion precision is improved, but quantization noise and distortions increase
Solution Approach 1:
By segmenting the resolution across multiple unit elements activated cyclically, the quantization error from each low-resolution element is distributed and averaged over time, reducing the overall quantization noise in the analog output signal.
Solution Approach 2:
The digital signal is pre-processed with a digital filter before conversion to shape the quantization noise spectrum and push noise energy to frequencies outside the signal band of interest, thereby reducing in-band quantization noise.
3Volume of moving object
If current-source-based DACs are constructed with decreasing semiconductor line width, then device miniaturization is achieved, but linearity characteristics deteriorate
Solution Approach 1:
Each unit element uses simple current-source-based circuitry with relaxed linearity requirements, while the overall system achieves high linearity through the cyclical activation sequence and digital signal processing. This allows miniaturization of individual elements without compromising overall system linearity.
Solution Approach 2:
A digital filter processes the output signal to correct for non-idealities and maintain linearity. The filter compensates for deviations in individual unit elements, ensuring that the combined analog output maintains high linearity characteristics even when individual elements have poor linearity due to small semiconductor dimensions.
Data Source
AI summary
An apparatus and method for digital-to-analog conversion. A digital-to-analog converter includes a sampler for resampling a digital signal and a DAC array. The DAC array includes a sequencer, a unit element activator, and an array of one-bit DACs (unit elements). The unit elements are activated in a cyclical sequence, based on the resampled digital signal. Unit elements in the sequence may be skipped, based on a disruption probability. The disruption probability may be determined randomly, or pseudo-randomly. Output signals of the unit elements are summed or averaged to form an analog signal. The converter may include a filter to filter the analog signal.