Electrical and Mechanical Measurements of Loudspeakers and Sound System Equipment. Tutorial to a new IEC Standard Project by Wolfgang Klippel,

Electrical and Mechanical Measurements of Loudspeakers and Sound System Equipment Tutorial to a new IEC Standard Project 2016 by Wolfgang Klippel, IEC Standard Project LOUDSPEAKER MEASUREMENTS, 1 Need for updated IEC Loudspeaker Standards (60268-xx) OBJECTIVES: Applicable to all kinds of modern audio devices (active, passive) Coping with any input signal (digital, wireless, ) Defining new measurement techniques (e.g. Rub & buzz test) Bridging manufacturing (QC) and system development (R&D) Providing comprehensive information in a shorter measurement time (e.g. directivity) Simplify interpretation (e.g. Root cause analysis) Increasing flexibility to consider particularities of the application (e.g. home, automotive, personal, professional, ) Avoiding redundancy with other standards (IEC, CES, AES, ALMA, ITU) Updating, merging of existing IEC standards (e.g. 60268-5) IEC Standard Project LOUDSPEAKER MEASUREMENTS, 2

Disclaimer This standard provides a physical evaluation of the sound system no fixed values for PASS/FAIL limits and quality grading no characteristics for assessing overall sound quality or preference of the audio system no modeling perceptive and cognitive evaluation of the reproduced sound quality by user Conclusions: This standard describes the general framework of the physical evaluation Further standards are required to consider the particularities of personal equipment, microspeakers, headphones, homeequipment, automotive, professional applications Perceptive evaluation requires a separate standard IEC Standard Project LOUDSPEAKER MEASUREMENTS, 3 How to organize the new standard? Problems: An overwhelming number of meaningful and important measurements and characteristics (not important for all users) Measurement of transducer parameters require access to the electrical terminals and diaphragm Two basic loudspeaker standards are required! Acoustical (output based) measurement (Part A IEC 60268-21) Applicable to transducers and systems System-oriented modeling Input-output transfer characteristics (distortion) no electrical and mechanical characteristics Important for end-user Electrical and mechanical measurement (Part B) Applicable to transducers and passive systems Access to internal state variables of the transducer Model based (lumped, distributed,...) Essential for transducer and system design, less important for end-user IEC Standard Project LOUDSPEAKER MEASUREMENTS, 4

SCOPE OF PART A (IEC 60268-21) ACOUSTICAL (OUTPUT BASED) MEASUREMENTS This International Standard applies to passive and active sound systems such as loudspeakers, headphones, TV-sets, multi-media devices, personal portable audio devices, automotive sound systems and professional equipment. The device under test (DUT) may be comprised of electrical components performing analogue and digital signal processing prior to the passive actuators performing a transduction of the electrical input into an acoustical output signal. The measurements presented here determine the transfer behaviour of the DUT between an arbitrary analogue or digital input signal and the acoustical output at any point in the near and far field of the system. This includes operating the DUT in both the small and large signal domains. The influence of the acoustical boundary conditions of the target application (e.g. car interior) can also be considered in the evaluation of the sound system. Note: This standard does not apply to microphones and other sensors. This standard does not require access to the state variables (voltage, current) at the electrical terminals of the transducer. Sensitivity, electric input power and other characteristics based on the electrical impedance will be described in a separate standard document IEC 60268-Xb dedicated to electrical and mechanical measurements. IEC Standard Project LOUDSPEAKER MEASUREMENTS, 5 SCOPE OF PART A ACOUSTICAL (OUTPUT BASED) MEASUREMENTS (IEC 60268-21) Properties of the black box depend on control parameters and stimulus Evaluation is based on evaluation of acoustical output digital audio stream Black box No access to internal states Near Field Far Field drivers control parameters (e.g. attenuation) IEC Standard Project LOUDSPEAKER MEASUREMENTS, 6

Summary What is new in IEC 60268-21? updating measurement techniques using new stimuli (chirp, multi-tone complex, burst) ( Comprehensive ) physical evaluation of the acoustical output A single value (Umax or SPLmax) rated by the manufacturer to calibrate the rms value of the stimulus Assessing large signal performance (considering heating, nonlinearities) complete assessment of the 3D sound field radiated by the loudspeaker in an anechoic environment (near and far field) physcial measurement of impulsive distortion in the time domain to assess rub & buzz and other loudspeaker defects IEC Standard Project LOUDSPEAKER MEASUREMENTS, 7 Relationship to related standards projects electrical & mechanical Measurements IEC 60268-XX Microspeakers IEC 63034, TC100-2683 NP Standard Method of Measurement for Powered Subwoofers ANSI/CEA 2010 Standard Method of Measurement for IN-Home Loudspeakers ANSI/CEA 2034 IEC Standard Project LOUDSPEAKER MEASUREMENTS, 8

SCOPE OF PART B ELECTICAL AND MECHANICAL MEASUREMENTS u terminals i Electro- Mechanical Conversion F coil Voice coil x coil Mechano- Acoustical Conversion p( r c ) radiator s surface x( r c ) Air Load This International Standard applies to transducers but also to passive and active sound systems such as loudspeakers, headphones, TV-sets, multi-media devices, personal portable audio devices, automotive sound systems and professional equipment where the electrical input terminals and the surface of the radiator are accessible by an electrical or mechanical sensor. The standard describes only physical measurements which assess the transfer behaviour of the device under test (DUT). This includes operating the DUT in both the small and large signal domains. The influence of the acoustical boundary conditions of the target application (e.g. car interior) can also be considered in the physical evaluation of the sound system. The standard does not assess the perception and cognitive evaluation of the reproduced sound and the impact of perceived sound quality. IEC Standard Project LOUDSPEAKER MEASUREMENTS, 9 IEC 60268 PART B Electrical and Mechanical Measurement LIST OF CONTENT: Measurement of the electrical signals at the terminals (u, i) Electrical characteristics (input impedance, power, ) Efficiency, sensitivity, Lumped parameters (TS, other linear, nonlinear Coil and magnet temperature, thermal parameters Mechanical characteristics and distributed parameters (cone) Long-term testing Time varying parameters (aging, fatigue, ) Climate impact IEC Standard Project LOUDSPEAKER MEASUREMENTS, 10

Not covered in Part B sound radiation Sound propagation Room Interaction Room Interaction Room Interaction p(r1) Audio signal Amplifier Crossover EQ u(t) x(t) Electromechanical Transducer Mechanoacoustical Transducer (Cone) Sound radiation Sound propagation Room Interaction Room Interaction Room Interaction p(r2) sound field i(t) Sound radiation Sound propagation Room Interaction Room Interaction Room Interaction p(r3) Radiation and propagation of sound into 3D space Black box modeling of DSP, crossover, amplification Linear and nonlinear distortion in the output signal IEC Standard Project LOUDSPEAKER MEASUREMENTS, 11 Exploiting A Priory Information from Physics and Psychoacoustics Grey Model parameters Black Box state variables Input variables Structure Output variables Model 1. Structure, Relationship, Operators (e.g. equivalent circuit) 2. Parameters (e.g. moving mass M ms,...) 3. State variables (e.g. displacement x,...) IEC Standard Project LOUDSPEAKER MEASUREMENTS, 12

Normal Measurement Condition Mounting of the DUT and acoustical loading (baffle, clamping in free air, coupler, horn, plane wave tube, ) Acoustical environment (full free space, half space, free air, target application, ) Unwanted electrical, mechanical or acoustical signals (e.g. noise) The DUT is acclimatized to the normal ambient conditions Additional cooling periods are required Test signal (stimulus) with specified properties (spectrum, duration, etc.) at specified rms value Attenuators, equalizers, dynamics and any other active control elements shall be set to their normal position Measuring equipment suitable for determining the wanted characteristics (accuracy IEC Standard Project LOUDSPEAKER MEASUREMENTS, 13 Electrical signals at the terminals (u, i) Measurement voltage and current Four wire sensing Peak and rms values Maximum Input Voltage Characteristics Rated noise voltage Short term maximum noise voltage Long term maximum noise voltage Rated sinusoidal voltage Related to maximum input power and rated impedance IEC Standard Project LOUDSPEAKER MEASUREMENTS, 14

Maximum Input and Output Signal selected input ~ stimulus u Sound pressure output amplifier, equalizer, ect.) transducer Idea Using only one single value meaningful for engineering, marketing, final user Rating of the maximal amplitude by manufacturer based on design, target application, evaluation Rated value can be applied to input and output DUT will not be damaged by the broadband test stimulus in a frequency range defined by manufacturer Rated maximum input value u max Good for DUTs with a single input and constant transfer function between input and output Not meaningful for active systems Rated maximum (output) SPL max Universal approach for passive and active systems Can be applied to any input channel Can cope with gain controllers, equalizers, limiters, ect. IEC Standard Project LOUDSPEAKER MEASUREMENTS, 15 How to rate u max and SPL max? The manufacturer has freedom to rate u max and SPL max but should consider the following requirements: final target application (rated frequency range, spectrum of typical program material, evaluation point, ect. ) DUT can reproduce the stimulus at SPL max for any time without damage sufficient sound quality for the particular application acceptable regular nonlinear distortion (harmonic + intermodulation) low compression of the fundamental (heating, mechanical limiting, protection) effective frequency range corresponds with the rated frequency range no rub & buzz or any other defects Durability of the loudspeaker shall be tested by a 100 h test using this stimulus IEC Standard Project LOUDSPEAKER MEASUREMENTS, 16

How to verify the maximum amplitude? 1. Defining a test value u test (based on information from customer, marketing or development) 2. 100 h test with the stimulus at test value u test 3. Measurement of characteristics defined in the data sheet 4. Assigning the test value to the rated maximum reference value u max =u test, if the DUT is not damaged and the within the stated specification Repeat the test with a lower test value if the evaluation was not successful IEC Standard Project LOUDSPEAKER MEASUREMENTS, 17 Should the manufacturer rate u max or SPL max? Maximum input value u max is preferred in the development and (end-of-line) testing of transducers and passive systems Maximum output value SPL max is preferred when Using multiple input channels (digital, analog) Using different parameter settings of the audio device (gain, equalizer,...) describing the physical limits and performance of the audio device comparing competitive products Generating useful information for the end user IEC Standard Project LOUDSPEAKER MEASUREMENTS, 18

Amplitude Adjustment of the Input Signal based on SPL max rated by manufacturer V Note: u max depends on selected input channel, setting of control elements (gain, equalizers, ect.) Evaluation point Broad-band stimulus in rated frequency band u ~ u ~ max f start f end Comparator Characteristics stated by the manufacturer Rated frequency band defined by f start and f end Rated maximum sound pressure level SPL max Evaluation point r e (distance, angle, ) Properties of the stimulus used during calibration (multi-tone or pink noise with shaping, ) Objectives of the calibration process Fast determination of the maximum input value u max based on SPL max Using u max for the calibration of other test stimuli Full flexibility for using any input channel of the active system (analogue, digital, ect. ) IEC Standard Project LOUDSPEAKER MEASUREMENTS, 19 Input Electrical Power Definition of Input Electrical Power Real input power Power dissipated in DC resistance R e Power dissipated at nominal impedance 1 P REAL ( t) u( t ) i( t ) d T ~ 2 P ( t) i ( t) R ( t) RE P u~ 2 ( t) / Z E, N ( t) E N Maximum Input Power Characteristics Rated maximum noise power (power handling capacity) Short term maximum noise power Long term maximum noise power Rated sinusoidal power P u ~2 / max max Z N IEC Standard Project LOUDSPEAKER MEASUREMENTS, 20

Complex Electrical Input Impedance Z E ( f ) U( f ) / I( f ) Definition Ratio between complex voltage spectrum and current spectrum (transfer function of a linear system) Condition Sufficient spectral excitation small signal domain (distortion THD < 1 %) Single value Characteristic: Rated input impedance (based on minimum electrical impedance) IEC Standard Project LOUDSPEAKER MEASUREMENTS, 21 Mechanical Measurements u terminals i Motor F coil Voice coil x coil Radiator p( r c ) radiator s surface x( r c ) Air Load Vibration x(t,r c ) at arbitrary point r c on the radiatior surface Non-destructive, non-contact measurement without additional load optical sensor measured in the direction of the normal vector Dynamic measurement required (full audio band) Scanning technique provides sufficient spatial resolution Forces are difficult to measure ( x, v, a) Measurement of displacement x provides dc-component generated dynamically by transducer nonlinearities IEC Standard Project LOUDSPEAKER MEASUREMENTS, 22

Positioning of the Radiator n r Reference plane and normal vector r r,ref z The reference plane with the normal vector n r shall be used to define the reference axis and the reference point r r,ref. Reference point o r O y The reference point r r,ref shall be a point on the radiator s surface cutting the reference plane. The position of the reference point r r,ref shall be specified by the manufacturer. Rated conditions used to describe the geometry and position of the radiator in the coordinate system x Orientation vector The orientation vector o r defines the orientation of the radiator within the reference plane and the direction of azimuthal angle =0 in spherical coordinates. IEC Standard Project LOUDSPEAKER MEASUREMENTS, 23 Mean Voice Coil Position u terminals i Motor F coil Voice coil x coil Radiator F( r c ) radiator s surface x( r c ) Air Load Mean coil displacement x coil averaged over coil x coil L x( t, rc ) dr N 1 ( t) x( t, rc, L N 0 ) n n1 Measurements at 4 points x x x Voice coil x Radiator s surface IEC Standard Project LOUDSPEAKER MEASUREMENTS, 24

[mm/v] [mm] Voice Coil Position 5 overload Voice coil displacement 00:15:30 Xpeak Xdc Xdcmax Xbottom Upper boundary overload KLIPPEL 4 Dynamic DC displacement X DC 3 2 1 0 displacement x rel rest position of the voice coil -1-2 Coil s rest position X 0-3 -4-5 Lower boundary 0 100 200 300 400 500 600 700 800 900 t [sec] Absolute voice coil position is determined by x ( t) x ( t) X 0( t, DUT ) abs x rel rel Voice coil displacement ( t) x ( t) x ( t) AC DC Voice coil rest position X 0 depends on time t and the device under test (DUT) DC displacement X DC depends on the audio signal and transducer nonlinearities Only AC component x AC generates sound pressure output IEC Standard Project LOUDSPEAKER MEASUREMENTS, 25 Displacement Transfer Function displacement voltage u Motor F V Vibration X(r) F(r) Radiation soundfield near far field field H X, U ( f, r ) X ( f, r ) / U( f ) r r Log-Model creep H x (0) 10-1 Magnitude of transfer function Hx(f)= X(f)/U(f) Q ts > 1 H x( fs) Qts H (0) Total quality factor x 10-2 12 db/octave 10-3 compliance dominant f s mass dominant 10 20 50 100 200 500 1k 2k 5k 10k Frequency [Hz] IEC Standard Project LOUDSPEAKER MEASUREMENTS, 26

[Ohm] Small Signal Lumped Parameters based on linear modeling Electrical Parameters DC resistance R e (Tv) Lossy inductance (Wright, Leach, LR2,...) Electrical representation of the fundamental resonator (capacitor C MES, inductance L CES and resistance R ES ) Relative lumped parameters resonance frequency mechanical quality factor, electrical quality factor, total quality factor Mechanical lumped parameters moving mass, stiffness, compliance and mechanical resistance IEC Standard Project LOUDSPEAKER MEASUREMENTS, 27 Interpretation of the Electrical Input Impedance Z e (jw) Electrical Impedance at the Terminals Z e Re(TV) ( j w ) ZL(j w) voltage U ( j w ) I ( j w ) current 70 60 50 40 30 20 Magnitude of electric impedance Z(f ) Measured Fitted KLIPPEL Z L (jw) i 10 R e u Cmes Lces Res 0 1 2 5 10 20 50 100 200 500 1k Frequency [Hz] f s,q es,q ms f s Resonance Frequency 1 2 1 C M ms ms 1 2 1 C L mes ces Electrical Quality Factor Qes C ms R e 2 Bl 2f s 2f C s mes R e Mechanical Quality Factor Q ms 1 2f sc C R 2f ms ms s mes R es IEC Standard Project LOUDSPEAKER MEASUREMENTS, 28

[Ohm] Equivalent Circuit of a electrodynamical transducer operated in free air Electrical domain representing voice coil Mechanical domain representing mechanical elements including air load Electrical dc Resistance R e Impedance describing lossy inductance Z L(f) Driving force current i Bli U Voltage Blv Back EMF Bl V=dx/dt velocity M ms C ms(f) R ms(f) -1 Moving mass compliance Losses h res(f) residual admittance Force factor mechanical admittance (Mobility, Fi) Type Analogy IEC Standard Project LOUDSPEAKER MEASUREMENTS, 29 Lossy Inductance Z L (jw) measured curves fitted by an ideal inductance Re ZL(f) U i Blv Bl Bli V=dx/dt Mms Cms(f) Rms -1 L e 1 Parameter only Large deviation limited use 6dB/octave 10 1 Phase (fitted) 100 90 80 70 [deg] 90 degree 0.1 Magnitude (measured) Magnitude (fitted) Phase (measured) 40 1 2 5 10 20 50 100 200 500 1k Frequency [Hz] IEC Standard Project LOUDSPEAKER MEASUREMENTS, 30 60 50

Models for Electrical Impedance Z L Eddy currents cause a lossy inductance Re ZL(f) i Bli Leach U Blv Bl V=dx/dt Mms Cms(f) Rms -1 Z L (jω)= K (jω) n ; ω= 2πf LR-2 (shunted inductance) Z L (jω) = L e jω + (R 2 L 2 jω ) / (R 2 + L 2 jω) Z L (jw,x) L e L 2 a) Wright Z L (jω)= K rm ω Erm + j (K xm ω Exm ) R 2 b) LR-3 (shunted inductance) and more (e.g. Thorborg) IEC Standard Project LOUDSPEAKER MEASUREMENTS, 31 L e L 2 R 2 L 3 R 3 c) Identification of Mechanical Parameters We need more information about the mechanical system Known Perturbation of Mechanical System (traditional technique) Direct Measurement of a Mechanical Signal Requires second measurement with additional mass or enclosure Based on impedance measurement No mechanical sensor required Time consuming Problems with mass attachment, box leakage Requires mechanical (acoustical) sensor (e.g. Laser) Only one measurement (fast) Driver in free air or in enclosure Reliable and reproducible data Can be applied to tweeters IEC Standard Project LOUDSPEAKER MEASUREMENTS, 32

Perturbation Method: Sealed Test Box Technique: A second measurement is performed while a known air stiffness K air is added to the suspension Advantages: simple technique C ms is measured primarily S d Air volume V box generating an additional stiffness K air Problems: depends highly on precise value of effective radiation area S d residual air volume (inside the transducer) can not be considered requires sealed diaphragm cannot be used to measure mechanical mass without air load Time consuming IEC Standard Project LOUDSPEAKER MEASUREMENTS, 33 Perturbation Method: Added Mass ring of clay used as added mass M add Technique: 1. In a first measurement the resonance frequency f s of the transducer is measured 2. In a second measurement the resonance frequency f m of the transducer is measured while a known mass M add is added to the cone. 3. The moving mass may be calculated by two methods M ms M add f s f m 1 2 Assumption: C ms is constant Advantages: Simple technique M ms is measured primarily M ms M add Qem f s 1 Q f es m Bl is constant Problems: cannot be applied to tweeter and microspeakers Time consuming Mechanical Resistance or stiffness are assumed as frequency independent parameters IEC Standard Project LOUDSPEAKER MEASUREMENTS, 34

Direct Parameter Identification using an optical laser sensor Laser triangulation sensor Technique: In addition to the voltage and current also the voice coil vibration (e.g. displacement) is measured by using an optical sensor Advantages: Fast (one step technique) Simple to use Bl is measured primarily Most precise results Can be applied to most transducers Problems: Optical problems (angle, surface) Coil displacement is not axialsymmetrical IEC Standard Project LOUDSPEAKER MEASUREMENTS, 35 Pure lumped mechanical parameters measured in vacuum Re ZL(f) i Bli electro-dynamical transducer operated in air U Blv Bl V=dx/dt Mms Cms(f) Rms(f) -1 hres(f) Re ZL(f) radiation resistance + turbulences i Bli U Blv Bl V=dx/dt Mmd Cmd(f) Rmd(f) -1 Mair Cair Rair hres(f) air leaks pure mechanical elements measured in vacuum moving air mass air Cavities IEC Standard Project LOUDSPEAKER MEASUREMENTS, 36

[ m m / N ] Mechanical Compliance C md (f) creep factor or C MD ( f ) C min 1 log 10 f / f min 2 1 f / f min M echanical com pliance (driver in vacuum ) C o m p lia n c e C m d ( f) fm in EFFECT: compliance increases to lower frequencies 0,0 0 1 0 0 0,0 0 0 7 5 Minimum c omplianc e C md0 C md (f d ) C md0 (Ritter) C md (f d ) CAUSE: viscoelasticity of the material CONSEQUENCES: more displacement than predicted by traditional modeling 0,0 0 0 5 0 0,0 0 0 2 5 0,0 0 0 0 0 creep factor or describes relative increase of compliance per decade f min 10 1 10 2 10 3 10 4 F r e q u e n c y [H z ] f d IEC Standard Project LOUDSPEAKER MEASUREMENTS, 37 Relative Small Signal Parameters Transducer operated in air (#=s), enclosure (#=c) or vacuum (#=d) (2nd order system) Resonance frequency f # Total quality factor Q t# Electrical quality factor Q e# Mechanical quality factor Q m# Additional Resonantor (4th order system) Resonance frequency f p of the additional resonator (port, passive radiator) Quality factor Q p of an additional mechanical or acoustical resonator IEC Standard Project LOUDSPEAKER MEASUREMENTS, 38

Acoustical Lumped Parameters small signal parameters based on a linear model Mechano-acoustical coupling function S d (f) Nominal effective radiation area S d (f=fs) Acoustical load impedance Z AL (f) describes the air load on the radiator s surface and effect of the acoustic system (port, enclosure, horn) versus frequency. Lumped parameters of the acoustical load (e.g. vented box) IEC Standard Project LOUDSPEAKER MEASUREMENTS, 39 Effective Radiation Area S D Electrical domain Mechanical domain Acoustical domain Re ZL(f) qa=sdv i Bli psd qb ql U Blv Bl Mmd Cmd Rmd -1 p Sd Cab V=dx/dt Ral Map S D is an important parameter of the lumped parameter model describes coupling between mechanical and acoustical domain determines the acoustical output (sensitivity, efficiency) affects the precision of the lumped parameter measurement if the test box perturbation technique is used (M ms, Bl, K ms, C ms ) Precise Measurement of S D is important! IEC Standard Project LOUDSPEAKER MEASUREMENTS, 40

Effective Radiation Area Sd 1. Geometrical Definition based on surround geometry Easy to use Applicable to woofers (surround area is much smaller than cone area) 2. Acoustical Definition Based on voice coil displacement and acoustical output Required for headphones, microspeakers IEC Standard Project LOUDSPEAKER MEASUREMENTS, 41 A Good Approximation? Calculation of Effective Radiation Area S D based on measured diameter Assumption: displacement decreases linearly over the surround displacement in constant in the inner part d i d SD d 4 2 i di d0 4 2 2 d 4 2 2 do d i do di d i 3 d 0 R. Small: less than 1% error if 0.8 d 0 < d i ) IEC Standard Project LOUDSPEAKER MEASUREMENTS, 42

Limits of the Approximation In headphone, micro-speakers, tweeters, compression drivers : Voice coil No constant displacement in the piston area No linear decay of displacement in the surround area IEC Standard Project LOUDSPEAKER MEASUREMENTS, 43 Effective Radiation Area S D Definition Radiator s surface replaced by Rigid piston v w, r ) ( c D Sc S ( w) v( w, r ) ds v coil c ( w) c (w) v coil S D (w) using mean voice coil velocity r coil q(w) v coil ( w) 2 0 v( w, r coil 2, ) d q(w) S ( w 0 ) Reading the absolute value at fundamental D S D resonance The effective radiation area S D is an important lumped parameter describing the surface of a rigid piston moving with the mean value of the voice coil velocity v coil and generating the same volume velocity q as the radiator s surface. The integration of the scanned velocity can cope with rocking modes and other asymmetrical vibration profiles. IEC Standard Project LOUDSPEAKER MEASUREMENTS, 44

Sd [cm^2] SPL [db] 80 70 60 50 40 30 Total Sound Pressure Level 2 10 103 104 Frequency [ Hz] Predicting the Acoustical Output at higher frequencies based on effective radiation area S D (f) Effective radiation Surface (Sd) 1000 900 800 Sd KLIPPEL 700 600 500 400 300 200 100 using effective radiation area S D (f) as a function of frequency f 0 Re ZL(f) 102 103 104 f [Hz] qa=sd(f)v F=Bli psd(f) i U Blv Bl MMD CMD RMD -1 p S d (f) v=dx/dt 0c RAR S ( f ) d radiated sound power, (SPL in a duct) Total Sound Pressure Level useful for transducers having high complexity of the mechanical vibration low complexity of the acoustical system (ka < 1) e.g. (in-ear) headphones, microspeaker application IEC Standard Project LOUDSPEAKER MEASUREMENTS, 45 Mechanical Distributed Characteristics u terminals i Motor F coil Voice coil x coil Radiator F( r c ) radiator s surface x( r c ) Air Load X ( f, rc ) H x( f, rc ) U( f ) Set of transfer functions between input voltage u(t) and displacement X(t,r c ) at arbitrary point r c on the radiatior surface IEC Standard Project LOUDSPEAKER MEASUREMENTS, 46

1078,1 Hz Mechanical and Acoustical Characteristics derived from Displacement Transfer function Scanning Vibrometer X ( f, rc ) H x( f, rc ) U( f ) Diagnostics On Cone vibration Accumulated Acceleration level AAL(f,r) Relative Rocking Level RRL n of the nth rocking mode Modal Expansion (Eigenfunction, natural frequencies, loss factor) Rocking Mode Parameters (Imbalances of mass, stiffness and Bl) Applications: Sound pressure prediction (system design based on measured vibration) Verification of FEA Optimization of modal vibration and sound radiation root cause analysis of rocking modes rub and buss IEC Standard Project LOUDSPEAKER MEASUREMENTS, 47 Accumulated Acceleration a( r c ) a( r c ) Integral of the absolute value of weighted cone acceleration p aa a ( r ) W a( r ) Sc cone s surface S c c ds c paa( ra ) AAL a p ( r ) 20log 2 o db 90 db 70 60 50 40 Accumulated Acceleration Level Rigid body modes with reference sound pressure p 0 and a useful scaling W to comparable ALL with SPL output 0 W 2 r r a c ra r c r a 30 100 1000 10000 f [Hz] IEC Standard Project LOUDSPEAKER MEASUREMENTS, 48

Relationship between AAL(f,r a ) and SPL(f,r a ) 90 db 70 60 50 40 AAL Rigid body modes Rigid body mode SPL Acceleration level Total sound Pressure level 30 The Rayleigh Integral is a useful approximation for the sound pressure output. The definition of the AAL(f,r a ) corresponds with Rayleigh Integral but neglects the phase information 100 1000 10000 f [Hz] Accumulated Acceleration Level describes total mechanical vibration is comparable with SPL is never smaller than SPL predicts potential acoustical output neglects acoustical cancellation is identical with SPL for a rigid body mode IEC Standard Project LOUDSPEAKER MEASUREMENTS, 49 Experimental Modal Analysis Expansion into a Series of Orthogonal Modes 90 db 70 60 50 40 Acceleration Level displacement: M m1 Frequency response for each mode Natural Functions describing mode shape x ( jw) ( jw) Ψ m m 30 100 1000 10000 f [Hz] Natural frequencies 70 Hz 840 Hz 3,8 khz 8,1 khz 11,2 khz Natural Function Completely different mode shapes (orthogonal)! Ψ m Ψ T n 0 m n IEC Standard Project LOUDSPEAKER MEASUREMENTS, 50

Modalanalysis of a Microspeaker Total Vibration RRL = -30 db 17 10 1 2 3 Rocking mode Rocking mode IEC Standard Project LOUDSPEAKER MEASUREMENTS, 51 Characteristics for Diagnostics Single-valued parameter derived from AAL (5) How critical is the rocking mode? db 80 70 60 50 40 30 20 10 0 Total AAL RRL Circular Component (AAL) Quadrature Component (AAL) f rock 100 1000 f [Hz] Woofer A with paper cone 1. Search for first maximum in quadrature component in AALon-axis! 2. Determine the relative rocking level RRL defined by RRL(f rock )=AAL quad -AAL in 3. rocking mode is negligible if RRL< -5dB IEC Standard Project LOUDSPEAKER MEASUREMENTS, 52

Desired and Undesired Vibration F 0 µ 1 piston mode generates the sound output rocking modes generate no output but impulsive distortion Voice coil gap voice coil rubbing impulsive distortion one period time IEC Standard Project LOUDSPEAKER MEASUREMENTS, 53 What Causes Rocking Modes? µ k µ m Mass Imbalances Stiffness Imbalances µ Bl Force Factor Imbalance Which root cause excites the rocking? mass, stiffness, force factor Where is the root cause located? angle showing the direction How to assess the magnitude of the excitation? moments IEC Standard Project LOUDSPEAKER MEASUREMENTS, 54

CFR in % Rocking Mode Analysis System Identification µ n x n (r c ) Modal Δ Modal τ n n Excitation Resonator Φ n (r c ) Modal Expansion SCANNER Rootcauses (imbalanc es) Mode Coupling Boosting Mechanism one piston modes, two Rocking Modes Total Vibration Accumulated Acceleration Level (AAL) DIAGNOSTICS IEC Standard Project LOUDSPEAKER MEASUREMENTS, 55 Mass Imbalance Experiment modified loudspeaker with additional mass AAL ~ Energy Excitation lighter total: CFR T = 1.6 % (295 ) components: CFR M = 1.4% (299 ) CFR K = 0.07 % (296 ) CFR Bl = 0 % (- ) CFR in % IEC Standard Project LOUDSPEAKER MEASUREMENTS, 56

Electro-acoustical Efficiency 1. Efficiency in a specified frequency band (e.g. pass band) Ratio between measured acoustical output power P a and measured electrical nominal input power P nom Pa 0 P nom Based on lumped parameter modeling 2 2 Bl Sd ( ) 0 for f >f s and ka<1, 0 2 ReM ms 2c 2. Mean Efficiency in a specified frequency band radiation on one side considered Average of efficiency measured in third-octave bands IEC Standard Project LOUDSPEAKER MEASUREMENTS, 57 Sensitivity Calculated from the frequency response and effective frequency range, as the sound pressure level produced at 1 m on the reference axis by an applied voltage of 2,83 V. Narrow-band sensitivity: the test signal is 1/3-octave filtered noise centered at 1 khz, or at the geometric mean of the limit frequencies of the effective frequency range if different from 1 khz. The frequency shall be stated. Broad-band sensitivity: the test signal is 2-octave filtered noise centered at 1 khz, or at the geometric mean of the limit frequencies of the effective frequency range if different from 1 khz. The frequency shall be stated. Reference: AES recommended practice Methods of measuring and specifying the performance of loudspeakers for professional applications Part 1: Drive units, AES2-1-R IEC Standard Project LOUDSPEAKER MEASUREMENTS, 58

Advanced modeling of an electro-dynamical transducer mechanical admittance (FI) type analogy R E(t) L E(ω,x,i) F=Bl(x,t)i q i R L(ω,x,i) Frel(x,i) F A SD(ω,x) Zload(ω) u Bl(x,t)v Bl(x,t) v M MS C MS(ω,x,t) RMS(ω,x,v) -1 p p out(r) dl(q) electrical domain mechanical domain (fundamental mode) higher order modes acoustical domain using on lumped elements where some parameters are time variant ( due climate, aging, heat) are frequency dependent have a nonlinear dependency on state variables (displacement, current ) IEC Standard Project LOUDSPEAKER MEASUREMENTS, 59 Ranking List of Transducer Nonlinearities 1. Force Factor Bl(x) 2. Compliance C ms (x) tweeter 3. Inductance L e (x) 4. Flux Modulation of L e (i) 5. Mechanical Resistance R ms (v) 6. Nonlinear Sound Propagation c(p) 7. Nonlinear Cone Vibration 8. Doppler Distortion (x) 9. Flux Modulation of Bl(i) 10. Port Nonlinearity R A (v) 11. many others... microspeaker woofers microspeaker horns microspeaker Full band IEC Standard Project LOUDSPEAKER MEASUREMENTS, 60

Full Dynamic Measurement of transducer and system nonlinearities Noise or music current voltage described in IEC Standard PAS 62458:2008 IEC Standard Project LOUDSPEAKER MEASUREMENTS, 61 Characteristics derived from Nonlinear Curve Shape NONLINEAR FORCE FACTOR Force factor limited displacement X Bl generating 10 % distortion Symmetry point X sym (X ac ) depending on AC amplitude Xac Offset x off of the voice coil from a defined reference rest position NONLINEAR STIFFNESS K ms (x) Compliance limited displacement X c generating 10 % distortion Suspension asymmetry A k NONLINEAR INDUCTANCE L e (x) Inductance limited displacement XL generating 10 % IEC Standard Project LOUDSPEAKER MEASUREMENTS, 63

Force Factor Limited Displacement x Bl defined according IEC standard 62458 6,0 Bl N/A 4,0 Bl min=82 % Bl(x=0) Bl(x Bl) Steps: 1. Operate transducer in large signal domain 3,0 2,0 1,0 0,0 x Bl -5,0-2,5 0,0 2,5 5,0 << Coil in X mm coil out >> 2. Read displacement X Bl where force factor Bl(x ac ) decreases to 82 % of the value Bl(x=0) at rest position IEC Standard Project LOUDSPEAKER MEASUREMENTS, 64 Compliance Limited Displacement x C defined according IEC standard 62458 C ms(x) mm/n 1 0.8 0.6 0.4 0.2 0 C MS(x=0) 0.75C MS(x=0) x C -5,0-2,5 0,0 2,5 5,0 coil in x mm coil out Steps: 1. Operate transducer in large signal domain 2. Read displacement X C where compliance value C ms (x ac ) decreases to 75 % of the value C ms (x=0) at rest position IEC Standard Project LOUDSPEAKER MEASUREMENTS, 65

Bl [N/A] 7 6 5 4 3 2 1 0 KLIPPEL -5-4 -3-2 -1 0 1 2 3 4 5 << Coil in X [mm] coil out >> Bl [N/A] 7 6 5 4 3 2 1 0 [Ohm] KLIPPEL -5-4 -3-2 -1 0 1 2 3 4 5 << Coil in X [mm] coil out >> Bl [N/A] 7 6 5 4 3 2 1 0-5 -4-3 -2-1 0 1 2 3 4 5 << Coil in X [mm] coil out >> Peak Displacement limited by Nonlinearities Compliance Bl(X) N/A KLIPPEL 3,5 Force Factor Cms(X) Cms (-X) mm/n KLIPPEL 0,9 25 Inductance Magnitude of electric impedance Z(f) x= 0 mm x = - 4 mm x = + 4 mm KLIPPEL Doppler 3,0 0,8 2,5 2,0 1,5 1,0 0,5 0,0 C lim = 75 % -4-3 -2-1 -0 1 2 3 4 << Coil in X [mm] coil out >> 0,7 0,6 0,5 0,4 0,3 0,2 0,1 0,0 Bl lim = 82 % -4-3 -2-1 -0 1 2 3 4 << Coil in X [mm] coil out >> 20 15 10 5 0 Z lim = 10 % 101 102 103 104 Frequency [Hz] X c X limited by Cms(x) 10 % THD X Bl X limited by Bl(x) 10 % IMD minimum X L X limited by Le(x) 10 % IMD X D X limited by Doppler 10 % IMD X max,10% Generating not more than 10 % THD or 10 % IMD IEC Standard Project LOUDSPEAKER MEASUREMENTS, 66 How to check the Voice Coil Rest Position Force factor Bl (X) magnet pole plate magnet pole plate Induction B voice coil voice coil rest position Induction B voice coil pole piece x=0 displacement pole piece x=x b displacement Offset from Symmetry Point Offset from Reference Curve Force factor Bl (X) Force factor Bl(X) Symmetry Reference Symmetry point Offset Can not cope with B field asymmetry No reference curve required Important for product development Transducer diagnostics Reference curve is required (Golden Reference DUT) Coil height and gap depth are constant Important for QC end-of-line testing can cope with asymmetrical curve shape IEC Standard Project LOUDSPEAKER MEASUREMENTS, 67

Stiffness Asymmetry A K defined according IEC standard 62458 K ms(x) N/mm 5 4 3 2 1 0 K MS( -x peak) K MS( x peak) -x peak x peak -5,0-2,5 0,0 2,5 5,0 coil in x mm coil out Steps: 1. Operate transducer in large signal domain 2. Read stiffness values X ms (X peak ) and X ms (- X peak ) at maximal peak displacement 3. Calculate stiffness asymmetry according A ( x K peak 2 K ) K MS MS ( x ( x peak peak ) K ) K MS MS ( x ( x peak peak ) 100%, ) IEC Standard Project LOUDSPEAKER MEASUREMENTS, 69 Thermal Measurements Increase of mean voice coil temperature ΔT v based on monitored DC voice coil resistance R e (t) Increase of magnet temperature ΔT m IEC Standard Project LOUDSPEAKER MEASUREMENTS, 70

[Ohm] R tv Temperature Measurement By Using a Steady-State Pilot Tone Benefit of adding an additional tone: Quasi-dc measurement with ac-power amplifier possible (f < 4 Hz) High speed monitoring of variation of Re(t) Long term averaging using low amplitude No external stimulus required active during cooling phase (OFF-cylce) Impedance measured at one frequency power of pilot tone is negligible Transducer: 1-4 Hz Systems: 0.01... 3 khz Stimulus Pilot Tone power amplif ier loudspeaker system Magnitude of electric impedance Z(f) v oltage sensor current sensor 50 45 40 35 30 25 20 Measured Most accurate measurement for transducer KLIPPEL Impact of woofer, tweeter and crossover Conductivity of Coil Material U(t) Fourier Transform Temperature Calculation - I(t) 15 10 5 2 5 10 20 50 100 200 500 1k 2k 5k 10k Resistance of cold speaker Increase of VC Temperature Frequency [Hz] IEC Standard Project LOUDSPEAKER MEASUREMENTS, 71 Thermal Characteristics Basic Characteristics Parameters of a thermal model P tv P g R tg P mag R tc (v) R tt (v) P coil T v T v C tv P con R ta (x) C ta P eg T g T g C tg T m T m C tm R tm Can be used to predict heating and cooling for any stimulus T a Derived Characteristics Effective total thermal resistance R therm = ΔT v /P real Thermal time constant of the voice coil τ v and magnet τ m Bypass factor to assess convection cooling and heating by eddy currents IEC Standard Project LOUDSPEAKER MEASUREMENTS, 72

Advanced Nonlinear Thermal Modeling dome v coil R tv P tv P g R tg P mag T v P con R tc (v) R tt (v) T g T m P coil P eg T v C tv Air convection cooling R ta (x) C ta Direct heat transfer T g C tg T m C tm R tm T a IEC Standard Project LOUDSPEAKER MEASUREMENTS, 73 Time Variant Parameters Shift of resonance frequency Shift of the voice coil rest position Electrical characteristics (input impedance, power, ) Efficiency, sensitivity, Lumped parameters (TS, other linear, nonlinear) Coil and magnet temperature, thermal parameters Mechanical characteristics and distributed parameters (cone) Long-term testing Time varying parameters (aging, fatigue, ) Climate impact IEC Standard Project LOUDSPEAKER MEASUREMENTS, 74

Variation of Suspension Stiffness K(t) versus Measurement Time t Performing a power test with pink noise of constant amplitude K(t) K(t=1h) break-in Stiffness ratio after 1 h and 100 h power testing R 100h K( t 100h) K( t 1h) K(t=100h) Speaker 2 Speaker 1 fatigue Disadvantages of : measurement results depends on the properties of the stimulus assumes constant excitation during power test can not be transferred to other stimuli neglects the slope of the stiffness variation 0 20 40 60 80 100 120 140 160 180 accumulated load hour Idea: Replacing time t by a quantity describing the dosage of the mechanical load t IEC Standard Project LOUDSPEAKER MEASUREMENTS, 75 Mechanical Load Model K(W) K(W=0) loss of stiffness K(W) K P=const. Measurement Condition: same stimulus of constant amplitude during the power test 0 W50% W90% W Stiffness of loudspeaker suspension versus accumulated work W Kˆ ( W) K( W 0) K( W) K( W ) N i1 i C 1 e W / wi N=2 sufficient for most cases IEC Standard Project LOUDSPEAKER MEASUREMENTS, 76

Your feedback is appreciated There are the following opportunities: Join or contact your national IEC committee Attend the AES standard group SC-03-04 Attend the ALMA symposium 2017 (before CES) Contact the German standard group (ringpfeil@kurtmueller.com) Or just contact me (wklippel@klippel.de) IEC Standard Project LOUDSPEAKER MEASUREMENTS, 77 Thank you! IEC Standard Project LOUDSPEAKER MEASUREMENTS, 78