Audio Feedback Instability Detection for Wider-Band ANR
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing audio systems with active noise reduction (ANR) features face instability issues due to feedback loops that can become unstable under certain conditions, leading to audible artifacts like tones or squealing, and conventional methods require conservative design to avoid instability, limiting system performance.
Innovation Solution
The system combines playback audio signals with feedback signals, filters them through various filters, and compares the results to detect instability, allowing for corrective actions without needing an inverse feedback filter, thus enabling more efficient noise reduction across a wider bandwidth.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If feedback microphones are used to pick up acoustic signals from the driver to form a closed loop system for noise reduction, then noise reduction performance is improved, but system stability deteriorates and may become unstable under certain conditions
Solution Approach 1:
The system performs preliminary action by detecting feedback instability conditions before they cause audible artifacts or system failure. The detection mechanism monitors the closed-loop system state continuously and triggers corrective actions (such as adjusting feedback filter coefficients or disabling the feedback path) preemptively, preventing instability from manifesting as harmful effects.
Solution Approach 2:
The system applies feedback by using the detected instability information to adjust the feedback path parameters. When instability is detected, the system modifies the feedback filter characteristics or gain to restore stability, creating a closed-loop control mechanism that automatically responds to stability conditions while maintaining noise reduction performance.
2Stability of the object's composition
If conservative design is used to avoid feedback instability, then system stability is improved, but noise reduction performance and bandwidth are limited
Solution Approach 1:
The system transitions from static conservative design to dynamic adaptive design. Instead of fixing feedback path parameters to ensure stability under all conditions, the system continuously monitors stability conditions and dynamically adjusts feedback filter coefficients, gain, or activation state based on real-time conditions, allowing optimal performance when stable and safe operation when unstable.
Solution Approach 2:
The system changes parameters dynamically by adjusting feedback filter coefficients, gain values, or bandwidth settings based on detected stability conditions. This allows the system to operate with higher gain and broader bandwidth when conditions permit, while automatically retreating to more conservative parameter settings when instability is detected, eliminating the need for permanently limited conservative design.
3Reliability
If feedback instability detection and correction mechanisms are added to the system, then system reliability is improved, but device complexity increases
Solution Approach 1:
The system achieves multi-functionality by integrating stability detection and correction capabilities into existing audio processing components. The same digital signal processor that performs noise reduction filtering also performs stability detection and parameter adjustment, eliminating the need for separate dedicated hardware or processing units for stability management.
Solution Approach 2:
The system applies self-service by automatically detecting instability conditions and correcting them without external intervention. The feedback path monitors its own stability and autonomously adjusts its parameters or disables itself when necessary, eliminating the need for external control systems or manual intervention to maintain stability.
Data Source
AI summary
A playback audio signal is combined with a feedback signal from a feedback microphone to provide a first combined signal. The first combined signal is filtered with a feedback filter to provide a driver command signal. The driver command signal is provided to an acoustic transducer for transduction to acoustic energy. The first combined signal is compared with the feedback signal to detect a feedback instability based upon the comparison.