SAR ADC Non-Uniform Quantization for Lower Power Conversion

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

SAR ADCs consume excessive power due to the need for greater precision in quantization, as they typically assume equiprobable data values, leading to increased bit requirements and power consumption.

Innovation Solution

Implementing non-uniform quantization in SAR ADCs by adjusting voltage levels based on data distribution, allowing for fewer quantization values and reducing power consumption.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If uniform quantization is used in SAR ADC, then the ADC can handle equiprobable data values, but power consumption increases due to greater precision requirements

Engineering Contradiction:
Improveability to handle equiprobable data valuesVSAvoidpower consumption
Core Design Contradiction:
Adaptability or versatilityVSUse of energy by moving object

Solution Approach 1:

The patent applies local quality by using non-uniform quantization where different regions of the signal range have different quantization densities. Specifically, the quantization intervals are adjusted based on the actual data distribution, with finer intervals where data points are more concentrated and coarser intervals where data points are sparse. This resolves the contradiction by maintaining high precision only where needed rather than uniformly across the entire range, thereby reducing overall power consumption while still accurately representing equiprobable data values.

Inventive Principle:
Principle #3Local quality

Solution Approach 2:

The patent changes the quantization parameter from uniform to non-uniform distribution. By modifying the quantization interval sizes based on data statistics and distribution characteristics, the system adapts the precision requirements dynamically. This parameter change allows the ADC to use fewer bits for quantization in regions with sparse data while maintaining high precision in regions with dense data, thus reducing power consumption without sacrificing the ability to handle equiprobable data values.

Inventive Principle:
Principle #35Parameter changes

2Measurement precision

If greater precision in quantization is used, then measurement precision improves, but power consumption increases

Engineering Contradiction:
Improvequantization precisionVSAvoidpower consumption
Core Design Contradiction:
Measurement precisionVSUse of energy by moving object

Solution Approach 1:

The patent applies partial action by providing high quantization precision only partially - specifically in the regions where data actually occurs with high probability. Instead of uniformly applying high precision across the entire dynamic range, the system concentrates precision resources where they are most needed based on data distribution analysis. This partial precision approach maintains effective measurement precision for the actual signal while avoiding the excessive power consumption that would result from uniform high-precision quantization across the full range.

Inventive Principle:
Principle #16Partial or excessive action

3Use of energy by moving object

If non-uniform quantization is implemented, then power consumption reduces, but device complexity increases due to offset control circuit

Engineering Contradiction:
Improvepower consumptionVSAvoidcircuit complexity
Core Design Contradiction:
Use of energy by moving objectVSDevice complexity

Solution Approach 1:

The patent applies preliminary action by pre-calculating and storing the non-uniform quantization offset values in lookup tables before the actual conversion process. The offset control circuit retrieves these pre-computed values based on the current conversion stage, avoiding the need for complex real-time calculations during operation. This preliminary preparation of quantization parameters simplifies the runtime operation and reduces the complexity of the control circuit while still enabling power-efficient non-uniform quantization.

Inventive Principle:
Principle #10Preliminary action

Data Source

PatentUS12489455B2ADC architecture with non-uniform quantization
Publication Date: 2025.12.02 APPLE INC
  • US12489455B2 patent drawing
  • US12489455B2 patent drawing
  • US12489455B2 patent drawing

AI summary

An ADC with non-uniform quantization is disclosed. A SAR ADC includes a comparator configured to generate, during successive clock cycles, respective comparison signals based on an analog input signal and a corresponding offset voltage. The SAR ADC further includes a digital logic circuit configured to store values of the respective comparison signals for corresponding ones of the successive clock cycles and further configured to output a digital value after a specified number of successive clock cycles. An offset control circuit is configured to, for a current one of the successive clock cycles, generate the corresponding value of the offset voltage based on one or more of values of the respective comparison signals from previous clock cycles. Respective offset voltages for successive clock cycles are spaced such that the SAR ADC implements non-uniform quantization values.