Higher Order Harmonic Signature Analysis for Loudspeaker Defect Detection
Loudspeaker assembly faults, such as a rubbing voice coil, bent frame, loose spider, etc., have traditionally been detected using experienced human listeners at the end of a production line. Previous attempts to develop production measurement systems for on-line testing typically analyze only low-order harmonics for the primary purpose of measuring total harmonic distortion (THD), and thus are not specifically designed to detect defective rub, buzz, and ticking sounds. This paper describes a new method wherein the total energy of high-order harmonics groups, for example, 10th through the 20th or 31st through the 40th, are measured and analyzed. By grouping high-order harmonics and resolving their respective total energies, distinct signatures can be obtained that correlate to the root cause of audible rub and buzz distortions (Temme, 2000). The paper discusses loudspeakers tested with specific defects, as well as results of a computer-based electroacoustic measurement and analysis system used for detection.
Authors: Dan Foley, Dr. Robert Celmer, Benjamin Sachwald, James Anthony, Tony Pagliaro, Shane Thompson
Presented at the 117th AES Convention, San Francisco, 2004


During loudspeaker production, particles may become trapped in the loudspeaker motor and voice coil vicinity, resulting in a distinctive defect that is easily heard, but difficult to detect by traditional test and measurements. We found that a Sine Sweep Stimulus followed by a High Pass Filter and RMS Envelope Analysis efficiently detected Loose Particles and Rub & Buzz defects. The remaining problem is how to reduce the effect of background noise, and get more reliable results. Statistical descriptors such as Crest Factor, Skewness, and Kurtosis are first investigated. Experimental results are given and the different tools are compared. New enhancements are described that increase effectively the overall immunity to background noise and discrimination of the method.
During the loudspeaker manufacturing process, particles may become trapped inside the loudspeaker, resulting in a distinctive defect that is easily heard but difficult to measure. To give a clearer view of the problem, Time-Frequency maps are shown for some defective loudspeakers. Based on this analysis, a reliable testing procedure using a swept-sine stimulus, high-pass filter, and RMS-envelope analysis is presented. Further possible enhancements and applications of the method are listed.
This landmark paper by Angelo Farina discusses the benefits of the Log Sine Sweep (also known as Frequency Log Sweep or Log Chirp), one of the test signals available in SoundCheck.
Although Virtual Instruments have been part of SoundCheck as long as most of us can remember, back in 1998 this was a huge innovation. This paper introduces the concept at a time when computerized audio measurement was a novel concept.