Loudspeaker Test Signal Generator with Frequency-Dependent Crest Factor

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Conventional test signals for evaluating loudspeaker performance, such as white noise and pink noise, have a constant and low crest-factor that does not accurately reflect the varying crest-factors of real signals like speech and music, leading to inaccurate performance assessments under real-life conditions.

Innovation Solution

A noise generator producing a composite test signal by combining broadband noise and randomly generated noise impulses, with separate equalization to achieve an average level and crest-factor that approximates real signals, including adjustable filters and gain controls to modify noise levels and impulse rates, ensuring a crest-factor that increases with frequency.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If conventional test signals (white noise, pink noise, sine sweeps) are used, then the test signal generation is simple and equipment complexity is low, but the crest factor is constant and low which does not reflect real signal characteristics, leading to inaccurate performance evaluation

Engineering Contradiction:
Improveaccuracy of loudspeaker performance evaluationVSAvoidcomplexity of test signal generation system
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The test signal is segmented into multiple frequency bands (e.g., octave bands), with each band having independently adjustable crest factor and average level parameters. This allows the system to complexly generate frequency-dependent test signals that accurately represent real signals, resolving the contradiction by making the complexity manageable through structured segmentation of the frequency spectrum.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The test signal parameters (crest factor and average level) are made dynamic and adjustable across different frequency bands, rather than being fixed as in conventional signals. This enables the system to adapt the test signal characteristics to match real-world signal conditions at different frequencies, improving measurement precision while maintaining operational simplicity through programmable control.

Inventive Principle:
Principle #15Dynamics

2Measurement precision

If the test signal has a constant and low crest factor, then the signal generation is simple and energy distribution is uniform, but it fails to approximate real signals where crest factor increases with frequency, resulting in inaccurate performance measurement

Engineering Contradiction:
Improveaccuracy of performance measurement under real conditionsVSAvoidsimplicity of test signal configuration
Core Design Contradiction:
Measurement precisionVSEase of operation

Solution Approach 1:

The invention changes the parameters of the test signal (crest factor and average level) as a function of frequency, allowing each frequency band to have optimized parameters that match real signal characteristics. This resolves the contradiction by enabling accurate performance measurement through parameter variation, while the systematic approach to parameter changes keeps the system relatively simple to operate.

Inventive Principle:
Principle #35Parameter changes

3Measurement precision

If broadband noise is equalized to reduce high frequency average level, then the signal better approximates real signals, but the crest factor at high frequencies becomes insufficient unless impulsive noise is added

Engineering Contradiction:
Improveaccuracy of high frequency performance evaluationVSAvoidcomplexity of noise generation and equalization system
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The invention merges two types of noise sources (broadband noise and impulsive noise) to achieve the desired test signal characteristics. The broadband noise provides the base signal with reduced high-frequency average level, while impulsive noise adds the necessary high-frequency crest factor. This combination resolves the contradiction by achieving accurate high-frequency performance evaluation through the synergistic merging of complementary noise sources.

Inventive Principle:
Principle #5Merging (Combining)

Data Source

PatentUS10841717B2Signal generator and method for measuring the performance of a loudspeaker
Publication Date: 2020.11.17 MEYER SOUND LABORATORIES INC
  • US10841717B2 patent drawing
  • US10841717B2 patent drawing

AI summary

A noise generator and method are provided for generating a test signal for measuring the performance of a loudspeaker over an operating broad band of frequencies ranging from low to high frequencies. A broadband random noise source is provided for generating broadband noise over the operating broad band of frequencies of the loudspeaker, and an impulsive noise source is additionally provided for generating random impulses of noise. The broadband noise and the randomly generated noise impulses are equalized to produce a composite noise signal having a desired crest factor as a function of frequency.