Feedback DAC Element Ordering for Higher Sigma-Delta ADC SNDR

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Sigma-Delta Analog-to-Digital Converters (ADCs) face significant nonlinearity due to feedback from Digital-to-Analog Converters (DACs), leading to limited system performance, with existing solutions increasing power consumption, noise floor, or degrading loop stability.

Innovation Solution

Optimizing the ordering of DAC elements by determining the optimal sequence through a series of analog-to-digital conversions and storing this ordering in a lookup table to minimize cumulative offset currents, thereby improving Signal-to-Noise plus Distortion Ratio (SNDR) without affecting control loop stability.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If feedback DAC is used in Sigma-Delta ADC, then digital output can be converted back to analog form for comparison, but significant nonlinearity is introduced that limits system performance

Engineering Contradiction:
ImprovelinearityVSAvoidsystem performance
Core Design Contradiction:
Measurement precisionVSReliability

Solution Approach 1:

The patent changes the ordering parameters of DAC unit elements to optimize linearity. By systematically permuting and selecting the best ordering sequence of unit elements, the cumulative offset errors are minimized, improving the overall linearity of the feedback DAC without changing the fundamental feedback conversion mechanism.

Inventive Principle:
Principle #35Parameter changes

2Measurement precision

If existing solutions are applied to reduce nonlinearity, then linearity improves, but power consumption increases

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

Solution Approach 1:

The patent performs preliminary characterization of DAC unit element offsets during manufacturing testing. The optimal ordering sequence is determined in advance and stored in a lookup table, eliminating the need for real-time calculations or additional active correction circuits that would consume power during normal operation.

Inventive Principle:
Principle #10Preliminary action

3Measurement precision

If existing solutions are applied to reduce nonlinearity, then linearity improves, but noise floor increases

Engineering Contradiction:
ImprovelinearityVSAvoidnoise floor
Core Design Contradiction:
Measurement precisionVSObject-generated harmful factors

Solution Approach 1:

The patent optimizes the ordering parameters of DAC unit elements to minimize cumulative offset errors. This parameter optimization reduces distortion components that would otherwise appear as noise in the signal band, thereby lowering the effective noise floor while improving linearity.

Inventive Principle:
Principle #35Parameter changes

4Measurement precision

If existing solutions are applied to reduce nonlinearity, then linearity improves, but loop stability degrades

Engineering Contradiction:
ImprovelinearityVSAvoidloop stability
Core Design Contradiction:
Measurement precisionVSStability of the object's composition

Solution Approach 1:

The patent determines the optimal unit element ordering in advance during manufacturing and stores it in a lookup table. This preliminary action allows the system to use the pre-optimized ordering without requiring real-time adjustments or additional feedback mechanisms that could destabilize the control loop.

Inventive Principle:
Principle #10Preliminary action

5Measurement precision

If DAC unit element ordering is optimized, then SNDR improves by up to 15 dB, but additional testing and lookup table storage are required

Engineering Contradiction:
ImproveSNDRVSAvoidtesting and storage requirements
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent performs the complex optimization work during manufacturing testing rather than during product operation. The lookup table storing optimal ordering sequences requires minimal additional storage and can be implemented using standard memory elements, making the manufacturing complexity acceptable in exchange for significant performance improvement.

Inventive Principle:
Principle #10Preliminary action

Data Source

PatentUS10298257B1SNDR improvement through optimal DAC element selection
Publication Date: 2019.05.21 NXP USA INC
  • US10298257B1 patent drawing
  • US10298257B1 patent drawing
  • US10298257B1 patent drawing

AI summary

A method for Signal-to-Noise and Distortion Ratio (SNDR) improvement through optimal Digital-to-Analog-Converter (DAC) element selection includes randomizing an order of a plurality of unit elements of a DAC, wherein each of the unit elements is controlled by a respective one of a plurality of digital inputs of the DAC. The plurality of digital inputs is sequentially asserted over at least a subset of a full set of the digital inputs to generate a plurality of analog values of an output of the DAC. A first SNDR of the DAC is measured from the plurality of analog values. A maximum SNDR, corresponding to an optimal order, is determined from the first SNDR and at least one previously measured SNDR. The optimal order of the unit elements of the DAC is stored in a memory to define connections between the digital inputs and the respective unit elements based on the optimal order.