MASH Delta-Sigma Modulator for Quantization Error Cancellation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Delta-sigma modulators face challenges in reducing side effects caused by quantization errors, particularly in high-speed applications, due to mismatch design issues between analog and digital components.
Innovation Solution
A multi-stage noise shaping (MASH) structure is employed to shape quantization errors between stages, utilizing noise shaping quantizers and digital cancellation filters to attenuate noise and improve signal transfer functions, thereby reducing the impact of mismatch errors.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a single-stage delta-sigma modulator is used, then the device complexity is low, but the measurement precision and resolution are insufficient for high-speed applications
Solution Approach 1:
The patent divides the single high-order modulator into multiple cascaded lower-order modulators (first modulator and second modulator). Each modulator operates at a lower individual order but collectively achieves high-resolution conversion through the cascade arrangement, reducing the complexity burden on any single stage while maintaining overall precision.
Solution Approach 2:
The patent transitions from a single-stage architecture to a multi-stage cascaded architecture, adding the dimension of stage multiplication. This dimensional change allows the system to achieve higher resolution by combining multiple modulation stages rather than relying on a single complex stage.
2Measurement precision
If quantization error suppression techniques are applied, then the measurement precision improves, but the device complexity increases due to additional components
Solution Approach 1:
The patent converts the harmful quantization error from the first modulator into a useful shaping function. By deliberately allowing the first quantization error to pass through and be shaped by the second modulator's noise transfer function, the system transforms an error source into a mechanism for error suppression, achieving high precision without additional complex error correction components.
Solution Approach 2:
The patent implements feedback mechanisms where the output of the first modulator feeds into the second modulator, and the quantization errors are fed back through the noise transfer function to be shaped and suppressed. This feedback loop allows continuous refinement of the conversion precision without requiring external correction circuits.
3Measurement precision
If high-order noise transfer functions are used to suppress quantization errors, then the measurement precision improves, but the manufacturing precision requirements become more stringent
Solution Approach 1:
The patent segments the high-order noise transfer function into multiple lower-order stages. Instead of implementing a single high-order function that would require extremely precise component matching, the system distributes the noise shaping across multiple cascaded modulators, each with simpler transfer functions that are easier to manufacture with standard precision tolerances.
Solution Approach 2:
The patent changes the architectural parameter from a single high-order system to multiple low-order systems. This parameter change allows each individual modulator to be manufactured with standard precision requirements while the collective system achieves the equivalent performance of a high-order design, effectively decoupling performance from manufacturing difficulty.
Data Source
Figure 1
Figure 2A~2B
Figure 3A~3B
AI summary
To convert a first stage input to a digital output, a delta-sigma modulator (40), an analog-to-digital converter and an associated signal conversion method based on an MASH structure are provided. The analog-to-digital converter includes the delta-sigma modulator and a sample and hold circuit. The delta-sigma modulator (40) includes a first signal converter (41), a second signal converter (43) and a digital cancellation logic (45). The first signal converter (41) converts the first stage input to a first converted output. The first signal converter (41) shapes a first stage quantization error to generate a second stage input. The first stage input and the second stage input are analog signals. The second signal converter (43) converts the second stage input to a second converted output. The digital cancellation logic (45) generates a digital output according to the first converted output and the second converted output.