Dual-Modulator Noise Shaping for Low-Noise Sigma-Delta DACs
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing sigma-delta digital-to-analog converters face challenges in reducing noise energy, particularly due to high power consumption from increased oversampling rates, increased noise energy from higher filter orders, and the need for large capacitors in low-pass filters to maintain signal-to-noise ratio, which complicates circuit design and increases costs.
Innovation Solution
A noise shaping circuit comprising two modulation units and an analog high pass filter, where the filters have specific transfer functions and relationships, allowing for the generation of an overall analog output signal that shapes noise spectrum with various slopes, reducing noise energy and increasing signal-to-noise ratio while minimizing circuit area.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If the oversampling rate is increased to increase the resolution of the sigma-delta DAC, then the resolution is improved, but the power consumption increases
Solution Approach 1:
The patent divides the single high-order modulation process into multiple cascaded modulation stages. Each stage processes a portion of the quantization noise, allowing the system to achieve high resolution without requiring excessive oversampling rates, thus reducing power consumption while maintaining measurement precision.
Solution Approach 2:
The patent employs feedback mechanisms where the output of each modulation stage is fed back to influence subsequent stages. This feedback loop enables the system to refine the signal progressively through multiple stages, achieving high resolution with moderate oversampling rates rather than requiring extremely high single-stage oversampling that would consume excessive power.
2Measurement precision
If the order of filter is increased to increase the resolution of the sigma-delta DAC, then the resolution is improved, but the energy of the out-of-band noise increases
Solution Approach 1:
The patent segments the filtering function across multiple cascaded modulation stages, where each stage applies a moderate-order filter. This segmentation allows the system to achieve the equivalent of a high-order filter effect while each individual stage generates limited out-of-band noise, preventing the noise energy accumulation that would occur with a single high-order filter.
Solution Approach 2:
The patent converts the quantization noise, which is traditionally a harmful out-of-band signal, into a useful shaping element. By deliberately shaping the noise spectrum through multiple modulation stages, the system pushes quantization noise to specific frequency bands where it can be effectively filtered, transforming the harmful noise into a controllable and manageable component that actually aids in achieving high resolution.
3Object-generated harmful factors
If the number of bits within the quantizer is increased to reduce the out-of-band noise, then the out-of-band noise is reduced, but under limited quantization bits the noise energy remains high
Solution Approach 1:
The patent segments the quantization process across multiple stages, where each quantizer operates with a limited number of bits. By distributing the quantization function across several stages rather than using a single high-bit quantizer, the system achieves effective noise reduction without requiring an impractically large number of quantization bits in any single stage, thus balancing noise reduction with device complexity constraints.
4Measurement precision
If the resistance of the resistor in the LPF is decreased to achieve high signal-to-noise ratio, then the signal-to-noise ratio is improved, but the capacitor requires large capacitance to maintain the corner frequency, increasing the circuit area
Solution Approach 1:
The patent segments the filtering function across multiple cascaded modulation stages, where each stage incorporates its own filtering elements. This segmentation allows the use of multiple smaller capacitors distributed across different stages rather than requiring a single large capacitor in a traditional low-pass filter configuration, thereby achieving the required signal-to-noise ratio while significantly reducing the total circuit area.
Data Source
AI summary
The present application provides a noise shaping circuit including a first modulation unit, configured to generate a first digital output signal according to a first digital input signal, the first modulation unit comprising a first quantizer; a first subtractor, coupled to an input terminal and an output terminal of the first quantizer, configured to generate a first quantization noise; and a second modulation unit, configured to generated a second digital output signal according to a second digital input signal, wherein the second digital input signal is related to the first quantization noise; wherein the noise shaping circuit generates an overall analog output signal according to the first digital output signal and the second digital output signal.