Coarse-SAR ADC Architecture With Offset Error Correction

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Successive approximation register analog to digital converters (SAR ADCs) face challenges in high-speed applications due to inherent slowness and high dynamic power consumption, which is exacerbated by the need for error correction cycles, and combining SAR ADCs with coarse ADCs introduces complexity and area/power issues.

Innovation Solution

A SAR ADC system that incorporates a coarse ADC to resolve initial bits, followed by SAR ADC for fine resolution, with error correction using a predefined offset to simplify the process and reduce complexity, eliminating the need for additional capacitors and complex digital implementations.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Speed

If a SAR ADC operates at high frequency to achieve 100 MSPS throughput, then speed is improved, but dynamic power consumption increases proportionately

Engineering Contradiction:
ImprovethroughputVSAvoiddynamic power consumption
Core Design Contradiction:
SpeedVSUse of energy by moving object

Solution Approach 1:

The patent divides the 10-bit conversion process into two segments: a coarse ADC that resolves the first 2 bits (requiring only 4 comparators) and a SAR ADC that resolves the remaining bits. This segmentation allows the SAR ADC to operate at lower frequency since it only needs to process 8 bits instead of 10, thereby reducing dynamic power consumption while maintaining 100 MSPS throughput

Inventive Principle:
Principle #1Segmentation

2Speed

If a coarse ADC is used to resolve first few bits, then SAR ADC speed requirements are relaxed, but device complexity increases

Engineering Contradiction:
ImproveSAR ADC operating frequencyVSAvoidconversion process complexity
Core Design Contradiction:
SpeedVSDevice complexity

Solution Approach 1:

The patent merges the coarse ADC and SAR ADC into a unified conversion architecture where the coarse ADC output directly feeds the SAR ADC. The capacitor array is shared between both converters, and the conversion process is integrated such that the coarse conversion and fine conversion occur in a coordinated manner, reducing overall system complexity despite the dual-converter approach

Inventive Principle:
Principle #5Merging (Combining)

3Measurement precision

If error correction cycles are added to SAR ADC, then measurement precision is improved, but speed decreases and dynamic power consumption increases

Engineering Contradiction:
Improveconversion accuracyVSAvoidoperating speed
Core Design Contradiction:
Measurement precisionVSSpeed

Solution Approach 1:

The patent performs error correction in advance by incorporating a predefined offset (0.5 LSB) during the coarse conversion phase. This preliminary error correction eliminates the need for subsequent error correction cycles, allowing the SAR ADC to complete its conversion in fewer steps and maintain high operating speed while achieving 10-bit precision

Inventive Principle:
Principle #10Preliminary action

Data Source

PatentEP3090488B1Combining a coarse ADC and a SAR ADC
Publication Date: 2023.08.02 TEXAS INSTRUMENTS INC
  • EP3090488B1 patent drawingFigure 1~2
  • EP3090488B1 patent drawingFigure 3
  • EP3090488B1 patent drawingFigure 4

AI summary

In described examples, a successive approximation register analog to digital converter (SAR ADC) receives an input voltage (102) and multiple reference voltages (105). The SARADC (100) includes a charge sharing DAC (110). The charge sharing DAC (110) includes an array of MSB (most significant bit) capacitors and an array of LSB (least significant bit) capacitors. A zero crossing detector (116) is coupled to the charge sharing DAC (110). The zero crossing detector (116) generates a digital output ( 118). A coarse ADC ( 122) receives the input voltage ( 102) and generates a coarse output (124). A predefined offset is added to a residue of the coarse ADC (122). A successive approximation register (SAR) state machine (120) is coupled to the coarse ADC (122) and the zero crossing detector (116) and generates multiple control signals (126). The control signals (126) operate the charge sharing DAC (110) in a sampling mode, an error-correction mode and a conversion mode.