Direct PDM Signal Mixing for Low-Area Beamforming Circuits

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current Digital Signal Processing (DSP) systems that use Pulse Density Modulation (PDM) inputs require complex Sigma-Delta designs and significant silicon area and processing power to convert PDM signals into multiple-bit samples for processing, especially in applications like beam-forming systems with multiple digital microphones.

Innovation Solution

A device and method for mixing multiple PDM bit streams that includes input logic, an error accumulation circuit, an error correction circuit, and an adder, allowing for the selection and combination of PDM bits during each clock cycle with adjustable delay and coefficient application, reducing the need for complex hardware and processing power.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Ease of operation

If PDM signals are converted into multiple-bit samples using dedicated Sigma-Delta hardware, then signal processing capability is improved, but silicon area and processing power consumption increase significantly

Engineering Contradiction:
Improvesignal processing capabilityVSAvoidsilicon area
Core Design Contradiction:
Ease of operationVSArea of stationary object

Solution Approach 1:

The patent extracts the essential mixing function from complex Sigma-Delta conversion hardware and implements it directly in the PDM domain using simple logic circuits. Instead of converting PDM to multi-bit samples and then mixing, the system performs mixing directly on PDM bits using an adder and logic gates, eliminating the need for complex Sigma-Delta decimation filters while maintaining beam-forming functionality

Inventive Principle:
Principle #2Taking out (Extraction)

Solution Approach 2:

The patent creates a simplified logical model of the mixing operation that operates directly on PDM bit streams. By copying the mathematical mixing operation into the discrete PDM domain using logic circuits (XOR gates, AND gates, and an adder), the system achieves the same signal processing effect without requiring physical conversion to multi-bit samples

Inventive Principle:
Principle #26Copying

2Ease of operation

If PDM signals are converted into multiple-bit samples using dedicated Sigma-Delta hardware, then signal processing capability is improved, but processing power consumption increases significantly

Engineering Contradiction:
Improvesignal processing capabilityVSAvoidprocessing power consumption
Core Design Contradiction:
Ease of operationVSUse of energy by moving object

Solution Approach 1:

The patent extracts the essential mixing function from complex Sigma-Delta conversion hardware and implements it directly in the PDM domain using simple logic circuits. Instead of converting PDM to multi-bit samples and then mixing, the system performs mixing directly on PDM bits using an adder and logic gates, eliminating the need for complex Sigma-Delta decimation filters while maintaining beam-forming functionality

Inventive Principle:
Principle #2Taking out (Extraction)

Solution Approach 2:

The patent replaces the mechanical/computational process of Sigma-Delta conversion and multi-bit arithmetic operations with a logical system operating directly on PDM bits. The mixing operation is performed using logic gates (XOR, AND) and a simple adder that operates at the bit level, substituting complex computational mechanics with simpler logical operations that consume less power

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

3Measurement precision

If complex Sigma-Delta designs are used for PDM input processing, then processing accuracy is improved, but device complexity increases

Engineering Contradiction:
Improveprocessing accuracyVSAvoidhardware complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent extracts the essential mixing function from complex Sigma-Delta conversion hardware and implements it directly in the PDM domain using simple logic circuits. Instead of converting PDM to multi-bit samples and then mixing, the system performs mixing directly on PDM bits using an adder and logic gates, eliminating the need for complex Sigma-Delta decimation filters while maintaining beam-forming functionality

Inventive Principle:
Principle #2Taking out (Extraction)

Solution Approach 2:

The patent changes the operational parameters of the mixing system by working directly in the PDM bit domain rather than converting to multi-bit samples. By operating at the binary pulse level with simple addition and logic operations, the system achieves the same signal combination effect with reduced hardware complexity while maintaining the statistical properties of the mixed signal

Inventive Principle:
Principle #35Parameter changes

Data Source

PatentUS8937515B2Device and method for direct mixing of pulse density modulation (PDM) signals
Publication Date: 2015.01.20 DSP GROUP
  • US8937515B2 patent drawing
  • US8937515B2 patent drawing
  • US8937515B2 patent drawing

AI summary

A device for mixing multiple (N) pulse density modulated (PDM) bit streams of a bit rate, the device comprises an input logic, an error accumulation circuit, an error correction circuit and an adder of more than N bits; wherein the device is arranged to output an output PDM bit stream that represents a mixture of the multiple input PDM bit streams; wherein the output PDM bit stream comprises a plurality of output PDM bits, wherein a certain output PDM bit of a plurality of output PDM bits that form the output PDM bit stream is generated during a certain clock cycle; wherein the input logic is arranged to select, during each fraction of the certain clock cycle, a current bit of a selected PDM bit stream, wherein different PDM bit streams are selected during different fragments of the certain clock cycle; wherein the error accumulation circuit is arranged to store intermediate values during a first fraction till a penultimate fraction of the certain clock signal and to store a last value during a last fraction of the certain clock signal.