Scholarlink

<p>Delta-sigma data converters oversample the input signal and perform noise shaping to produce a high-resolution digital output. Many existing methods implement a continuous-time delta-sigma modulator (CTDSM) for multi-channel ADCs, but these systems often suffer from flaws such as inadequate state resetting, which leads to inter-channel crosstalk and linearity issues. To overcome these limitations, a novel Hybrid Stochastic Divider Delta-Sigma Modulator is proposed to enhance multi-channel ADC performance. This approach integrates a Hysteresis Split Source Comparison Quantizer and a Distributed Stochastic Quantized Neural Network to automatically reset states, effectively reducing crosstalk across channels and eliminating linearity issues using Linear Ensemble Half-band Filtering. Furthermore, existing CTDSM control mechanisms struggle with poor performance due to threshold-based methods, resulting in quantization errors, slope overload distortion, and high latency. To address these challenges, a Robust Feedback Compression Controller is introduced, optimizing multi-ADC operation by mitigating high latency through a Tunable Column Bit Compression mechanism. Additionally, a Scalable Extrapolation GAN Controller is employed to predict and correct quantization errors, improving speed and efficiency.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Hybrid Stochastic Divider-Based Delta-Sigma Data Convertor Using Linearity Enhancement of Multi-channel ADC with Robust Control Mechanism

  • 期刊论文

  • 发表日期：

    2025年5月21日

查看全文
  • G. Jyothish Chandran,
  • P. R. Nishanth,
  • P. V. Ashwin

摘要

Delta-sigma data converters oversample the input signal and perform noise shaping to produce a high-resolution digital output. Many existing methods implement a continuous-time delta-sigma modulator (CTDSM) for multi-channel ADCs, but these systems often suffer from flaws such as inadequate state resetting, which leads to inter-channel crosstalk and linearity issues. To overcome these limitations, a novel Hybrid Stochastic Divider Delta-Sigma Modulator is proposed to enhance multi-channel ADC performance. This approach integrates a Hysteresis Split Source Comparison Quantizer and a Distributed Stochastic Quantized Neural Network to automatically reset states, effectively reducing crosstalk across channels and eliminating linearity issues using Linear Ensemble Half-band Filtering. Furthermore, existing CTDSM control mechanisms struggle with poor performance due to threshold-based methods, resulting in quantization errors, slope overload distortion, and high latency. To address these challenges, a Robust Feedback Compression Controller is introduced, optimizing multi-ADC operation by mitigating high latency through a Tunable Column Bit Compression mechanism. Additionally, a Scalable Extrapolation GAN Controller is employed to predict and correct quantization errors, improving speed and efficiency.