Segmented Delta-Sigma ADC for Faster, Area-Efficient Conversion

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing ADC architectures, such as SAR and flash ADCs, are inefficient in terms of area and power, while conventional delta-sigma ADCs have longer conversion times, posing challenges for memory systems as they scale down in size.

Innovation Solution

An n-split delta sigma ADC architecture that splits the input analog signal range into n+1 sub-ranges, assigning an N-bit digital code to each sub-range and performing a delta-sigma operation using reference voltages, resulting in a more area-efficient and faster conversion process.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If conventional delta-sigma ADC architecture is used, then high resolution is achieved, but conversion time becomes too long

Engineering Contradiction:
ImproveADC resolutionVSAvoidconversion time
Core Design Contradiction:
Measurement precisionVSLoss of time

Solution Approach 1:

The patent divides the ADC conversion process into multiple parallel sub-ranges (e.g., 8 sub-ranges for 3-bit segmentation). Each sub-range is processed simultaneously by dedicated comparator circuits, transforming a sequential high-resolution conversion into parallel operations. This segmentation reduces the number of conversion steps required while maintaining overall resolution through digital reconstruction of the segmented results.

Inventive Principle:
Principle #1Segmentation

2Speed

If flash ADC architecture is used, then conversion speed is improved, but area efficiency deteriorates

Engineering Contradiction:
Improveconversion speedVSAvoidADC area
Core Design Contradiction:
SpeedVSArea of stationary object

Solution Approach 1:

The patent segments the full-scale voltage range into multiple sub-ranges, each handled by a smaller number of comparators. Instead of requiring 2^N-1 comparators for an N-bit flash ADC, the segmented architecture uses (2^k-1) comparators per sub-range where k < N, with multiple sub-ranges processed in parallel. This dramatically reduces the total comparator count and circuit area while maintaining conversion speed through parallel operation.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent transitions from a single-dimensional sequential comparison approach to a multi-dimensional parallel processing structure. By organizing comparators into multiple sub-range blocks that operate simultaneously and combining their outputs through digital logic, the system achieves flash-like speed with segmented architecture, effectively adding the dimension of parallel processing to reduce area requirements.

Inventive Principle:
Principle #17Another dimension (Dimensionality change)

3Area of stationary object

If SAR ADC architecture is used, then area efficiency is improved, but conversion time increases

Engineering Contradiction:
ImproveADC areaVSAvoidconversion time
Core Design Contradiction:
Area of stationary objectVSLoss of time

Solution Approach 1:

The patent implements parallel segmented conversion where multiple sub-range comparisons occur simultaneously rather than sequentially as in SAR ADC. Each segment performs its conversion in parallel, and the results are combined digitally, eliminating the iterative sequential search process of traditional SAR ADC while using comparable area-efficient circuitry.

Inventive Principle:
Principle #1Segmentation

Data Source

PatentUS12425040B2Area-efficient and moderate conversion time analog to digital converter (ADC)
Publication Date: 2025.09.23 SANDISK TECHNOLOGIES LLC
  • US12425040B2 patent drawing
  • US12425040B2 patent drawing
  • US12425040B2 patent drawing

AI summary

Systems and methods for converting an input analog signal to a digital representation thereof. A method includes determining an input analog signal voltage range of the input analog signal, and splitting the input analog signal voltage range into n+1 sub-ranges, n being a number of splits in the input analog signal voltage range. The method also includes assigning a respective N-bit coarse digital code i to each sub-range. The method also includes identifying the input analog signal with a corresponding sub-range, the corresponding sub-range having respective digital code i. A delta-sigma operation is performed on the input analog signal using upper and lower reference voltages of the corresponding sub-range that the input analog signal is identified with, to produce the digital representation.