Rub & Buzz and High Order Harmonic Analysis Tutorial

Rub and buzz measurement is a type of harmonic analysis often used in audio and electroacoustic applications, focusing on higher order harmonics, typically greater than 10th order.

Overview of Rattle Noise and Rub & Buzz

A variety of consumer products have the potential to create undesirable mechanical noise that may be unacceptable or irritating to the consumer including:

• Automotive parts
• Compressors (refrigerator, freezer)
• Computers (blower activated)
• Electric drives
• Hearing aids/headsets
• Loudspeakers
• Transducers in phones and mobile phones
• Televisions

Rattle noise or rub and buzz are part of a class of non-linear, irregular, impulsive and unwanted distortion effects, which are not normally found with regular units under test (UUTs), rather they are typically produced by mechanical and structural defects in the UUT.  Due to short pulses of very low energy traditional testing and analysis methods like RMS-FFT and Total Harmonic Distortion (THD) are not useful.  The images below show typical effects in the signal:

1) Glitch in Signal

2) Steepness

3) Absolute (Steepness)

Case Study - Loudspeaker Defect Detection

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. But can unique defect signatures be determined for fault-diagnostic purposes?  Initial experimental results indicate that this can be achieved and that measuring in the ultrasonic range (> 20 kHz) is a key factor in identifying these signatures.  This case study 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.  S

Rub & Buzz Signal Characteristics of a Loudspeaker

Why focus on high-order harmonics?

THD is dominated by 2nd and 3rd harmonics, which rarely correlates with audible distortion, that is what we really hear.  For example, consider this speaker buzz below.  The speaker buzzes are typically an impulse train which result in a harmonically-rich spectrum.

Experimental Process

1. Loudspeaker samples with known defects that produce audible distortion when excited with a sine wave were obtained
2. Speakers were visually inspected to determine root cause(s) of failure
3. Speakers were manually swept through the audio range (20 Hz to 1 kHz) to identify excitation frequencies that cause audible distortion
4. Harmonic energy up to 100 kHz was analyzed using swept-sine excitation

Measurement Setup

• Each speaker was suspended using rubber bungee cords to isolate vibration
• The measurement microphone was placed in near field as shown below

Results

Visual Inspection for Defects

Speaker #	Surround	Cone	Voice Coil	Spider	Dust Cap	Loose Particle
1	None	None	None	None	None	Yes - Inside Voice Coil
2	None	Yes - Creased	None	None	None	Yes - Inside Voice Coil
3	None	Yes - Hole	None	Yes - Cut	Yes - Missing	Yes - Behind Spider
4	None	Yes - Creased	None	None	None	None
5	None	Yes - Creased	None	None	Yes - Dented	None
6	Yes - Dented	None	None	None	None	None
7	Yes - Dented	Yes - Hole/Creased	None	None	None	None
8	None	Yes - Dented	None	None	None	None
9	None	None	None	None	None	None
10	Yes - Dented	Yes - Creased	None	None	None	None

 

Analysis

• Speaker #9 was used as a control
  • No visual or audible defects
  • Below 0.01% Rub & Buzz for 40 – 100 Harmonics (results shown in graph below)

• Other speakers were compared for similar defects
• Looked for trends with the same harmonic grouping

Loudspeaker Defect Correlation

Surround

• Harmonics 1-10: Speakers 6, 7 & 10
• In Fundamental Range: 600-1300 Hz

[+] Enlarge Image

Dust Cap

• Harmonics 11-20: Speaker 3 & 5
• In Fundamental Range: 100-400 Hz

[+] Enlarge Image

Loose Particle

• Harmonics 31-100: Speaker 1
• In Fundamental Range: 100-150 Hz

[+] Enlarge Image

Spider

• Harmonics 31-100: Speaker 3
• In Fundamental Range: 400-550 Hz

[+] Enlarge Image

Cone

• Harmonics 61-70: Speaker 2, 3, 4, 5, 7, 8, & 10

[+] Enlarge Image

Conclusions

Using rub and buzz analysis provide a better method than traditional analysis.  The analysis is not only repeatable, but it is also adaptable for different speaker models.  Also critical is acquiring data into the ultrasonic range and the harmonics in this range provided insight into defects that otherwise would have been hard to detect including spider defects.  Additionally, not only did the higher order harmonics definitively show single defects in loudspeakers, they also helped characterize defects in speakers with multiple defects.

This case study was originally presented by Dan Foley, Vice President, Listen Inc., at NIWeek 2005.

Rub and Buzz Analysis VI - Example Code

Add this rub and buzz measurement VI into your application to calculate the rub and buzz percentage of your audio system components.  The VI utilizes THD and harmonic analysis data from the THD VI in the Sound and Vibration Measurement Suite.  Users can configure the minimum and maximum harmonic to calculate rub and buzz percentage from.

Related Links

• NI Sound and Vibration Portal
• Audio Test Technical Library
• Audio Test Case Studies

Special thanks to Birger Schneider Director, MicroLEX for background information in this tutorial.

Reader Comments | Submit a comment »