18 Sound 128th AES London AIC Model

Audio Engineering Society
Convention Paper
Presented at the 128th Convention
2010 May 22–25 London, UK
The papers at this Convention have been selected on the basis of a submitted abstract and extended precis that have
been peer reviewed by at least two qualified anonymous reviewers. This convention paper has been reproduced from
the author’s advance manuscript, without editing, corrections, or consideration by the Review Board. The AES takes
no responsibility for the contents. Additional papers may be obtained by sending request and remittance to Audio
Engineering Society, 60 East 42nd Street, New York, New York 10165-2520, USA; also see www.aes.org. All rights
reserved. Reproduction of this paper, or any portion thereof, is not permitted without direct permission from the
Journal of the Audio Engineering Society.
Electrical Circuit Model For a Loudspeaker

With An Additional Fixed Coil In The Gap
Daniele Ponteggia1 , Marco Carlisi2 , and Andrea Manzini3
1
Studio Ing. Ponteggia, Terni, TR, 05100, Italy
2
Ing. Carlisi Marco, Como, CO, 22100, Italy
3
18 Sound - Division of A.E.B. S.r.l., Cavriago, RE, 42025, Italy
Correspondence should be addressed to Daniele Ponteggia (daniele@studioponteggia.it)
ABSTRACT
A previous paper by some of the authors investigated a solution to minimize the loudspeaker inductance
based on an additional fixed coil positioned in the gap with 2 additional terminals. This device is referred
as A.I.C. (Active Impedance Control). In this paper a suitable electrical circuit model for such loudspeaker
with 4 terminals and 2 coils is proposed.
1. INTRODUCTION moving coil and made accessible via two additional

In a previous paper[1] by some of the authors an terminals. This coil is almost as long as the gap
alternative approach to inductance and inductance height and being wound around the pole piece turns
modulation reduction is introduced. While there out to be very close to the voice coil.
are several approaches to inductance and inductance A current flowing into this coil, if properly oriented,
modulation reduction, the solution proposed in the generates a magnetic field that is in opposition to
cited paper is based on a device referred as Active the field generated by the moving coil. This tends
Impedance Control (A.I.C.). to cancel out most of the voice coil inductance and
The A.I.C. is an additional fixed, multi turn coil, reduce the flux modulation and the inductance mod-
positioned in the loudspeaker magnetic circuit gap. ulation.
This coil could be made of multiple turns of copper The A.I.C. device can be seen as an “active” shorted
wire, rigidly wound around the pole piece facing the ring in the gap. The two A.I.C. terminals allow to
Ponteggia et al. Electrical Model For a Loudspeaker With An Additional Fixed Coil
I(jω) RE ZL (jω)
+
E(jω) − LCES CM ES RES
Fig. 1: Loudspeaker Model
√
jωKE
I(jω) RE LEB
LE
+
Fig. 2: Thorborg Model
decide how to drive the additional coil. There are and thanks to well known circuit analysis tools is
several choices: driving it with an additional ampli- easy to manage and understand. The model param-
fier, driving it in parallel with the voice coil, driving eters can be used as an example to properly design
it through a filter that shapes the driving level with loudspeaker boxes.
the frequency or simply shorting it. If we look at the loudspeaker model using the electri-
A loudspeaker equipped with an A.I.C. device gives cal only equivalent circuit it is possible to divide the
the loudspeaker system designer the freedom to de- impedance seen from the voice coil terminals ZV C
cide how to use the additional coil. A detailed in two parts: blocked impedance ZB and motional
description of the effects of these different driving impedance ZEM . The motional impedance contains
strategies is treated in depth in the above mentioned the mechanical and acoustical circuit parts. The
paper. blocked impedance models the voice coil immersed
into the magnetic circuit. There are several pro-
This work focuses on another aspect of a loudspeaker posed models available, in particular the coil does
equipped with an A.I.C. device. The classical loud- not behave as a perfect inductor due to eddy cur-
speaker electrical circuit linear models are not able rent losses in the iron. We will not deal here with
to describe correctly such kind of loudspeaker with the details, as there is a vast literature on the sub-
4 terminals. ject, however we need to cite at least[2] [3] [4] [5].
In this paper we propose an electrical circuit linear The voice coil impedance can be written as:
model that can deal with the complexity due to the
ZV C (jω) = ZB (jω) + ZEM (jω)
presence of the two coils: voice coil and A.I.C. coil.
where the motional impedance is:
2. LOUDSPEAKER ELECTRICAL EQUIVA- jωLCES
LENT CIRCUIT ZEM (jω) =
(jω)2 LCES CM ES + jωLCES /RES + 1
A loudspeaker can be modeled with an electrical,
and the blocked impedance is:
mechanical and acoustical lumped parameters cir-
cuit. The model can be used for several purposes ZB (jω) = RE + ZL (jω)
AES 128th Convention, London, UK, 2010 May 22–25

Page 2 of 9
Sinusoidal – 6'' Loudspeaker Sinusoidal – 6'' Loudspeaker

100.0 180.0 100.0 180.0
Voice Coil Impedance – AIC Separately Driven Voice Coil Impedance – AIC Separately Driven
Ohm Voice Coil Impedance – AIC in Parallel deg Ohm AIC Coil Impedance – Voice Coil Separately Driven deg
80.0 108.0 80.0 108.0
60.0 36.0 60.0 36.0
40.0 -36.0 40.0 -36.0
20.0 -108.0 20.0 -108.0
0.0 -180.0 0.0 -180.0

10 20 50 100 200 500 1k Hz 2k 5k 10k 20k 10 20 50 100 200 500 1k Hz 2k 5k 10k 20k
CH A Ohm 1/48 Octave Unsmoothed 48kHz Delay [ms] 0.000 Distortion Rise [dB] 30.00 SweepTime [ms] 0.00 CH A Ohm 1/48 Octave Unsmoothed 48kHz Delay [ms] 0.000 Distortion Rise [dB] 30.00 SweepTime [ms] 0.00

100.0 180.0 100.0 180.0
Voice Coil Impedance – AIC Separately Driven AIC Coil Impedance – Voice Coil Open
Voice Coil Impedance – AIC Shorted
80.0 Voice Coil Impedance – AIC Open 108.0 80.0 108.0
60.0 36.0 60.0 36.0
40.0 -36.0 40.0 -36.0
20.0 -108.0 20.0 -108.0
0.0 -180.0 0.0 -180.0

10 20 50 100 200 500 1k Hz 2k 5k 10k 20k 10 20 50 100 200 500 1k Hz 2k 5k 10k 20k
Fig. 3: Impedance of voice coil and A.I.C. coil in different configurations - 6” Loudspeaker
where the term ZL (jω) models the “lossy” of the known models we get meaningless results.
impedance.
3. IMPEDANCE MEASUREMENTS
The several proposed models for the loudspeaker
In order to investigate the principle of operation of
differ in the lossy impedance term. The lossy
the loudspeaker with the A.I.C. device we carried
impedance of the model recently proposed by Thor-
out a series of impedance measurements on two loud-
borg is composed of a series of an inductor in series
speakers, a 6” woofer and a 10” mid-woofer.
with a parallel of a semi-inductor and an inductor.
This model is very interesting because each compo- We collected the electrical impedance seen at the
nent has a physical meaning, later in this document voice coil and at the A.I.C. in different driving con-
we will use a modified version of this model for the ditions:
A.I.C. equipped loudspeaker model.
• Voice coil driven and A.I.C. coil open
The above circuits are unable to model a loudspeaker
equipped with the A.I.C. device. A loudspeaker with • Voice coil driven and A.I.C. coil shorted
A.I.C. has two coils, the voice coil and the A.I.C.
fixed coil. If we try to fit the measured impedance • Voice coil and A.I.C. in parallel driven with a
of the voice coil of a loudspeaker with A.I.C. to one single amplifier channel

Page 3 of 9

100.0 180.0 100.0 180.0
Voice Coil Impedance – AIC Separately Driven Voice Coil Impedance – AIC Separately Driven
80.0 108.0 80.0 108.0
60.0 36.0 60.0 36.0
40.0 -36.0 40.0 -36.0
20.0 -108.0 20.0 -108.0
0.0 -180.0 0.0 -180.0

10 20 50 100 200 500 1k Hz 2k 5k 10k 20k 10 20 50 100 200 500 1k Hz 2k 5k 10k 20k

100.0 180.0 100.0 180.0
Voice Coil Impedance – AIC Separately Driven AIC Coil Impedance – Voice Coil Open
Voice Coil Impedance – AIC Shorted
80.0 Voice Coil Impedance – AIC Open 108.0 80.0 108.0
60.0 36.0 60.0 36.0
40.0 -36.0 40.0 -36.0
20.0 -108.0 20.0 -108.0
0.0 -180.0 0.0 -180.0

10 20 50 100 200 500 1k Hz 2k 5k 10k 20k 10 20 50 100 200 500 1k Hz 2k 5k 10k 20k
Fig. 4: Impedance of voice coil and A.I.C. coil in different configurations - 10” Loudspeaker
• Voice coil and A.I.C. separately driven with two The analysis of these measurements shows that the
amplifier channels in parallel A.I.C. coil needs to be modeled together with the
mutual interaction with the voice coil.
• A.I.C. coil driven with voice coil open
4. PROPOSED A.I.C. LOUDSPEAKER MODEL
The impedance measurements were taken using a
The schematic of the proposed electrical equivalent
custom Audiomatica CLIO measurement system
circuit to model a loudspeaker equipped with an
featuring two Audiomatica QCBox model V ampli-
A.I.C. device is reported in figure 5.
fiers with current sensing connected in parallel.
The loudspeaker is modeled with a transformer with
A complete set of impedance measurements is shown
a “mild” coupling of the coils, wound on an iron
in figure 3 for the 6” loudspeaker and in figure 4 for
core, modeled with the lossy inductance proposed
the 10” loudspeaker.
by Thorborg.
If we try to fit the previously defined electrical equiv-
The terms LEB and LAIC are modeling the losses
alent circuit to the impedance curves, in the case of
on the primary and secondary windings (voice coil
voice coil connected in parallel with the A.I.C. and
and A.I.C. coil).
with the A.I.C. separately driven, we obviously get
different parameters. It is possible to draw a simplified schematic of the

Page 4 of 9
IAIC (jω) RAIC LAIC
+
EAIC (jω) −
1:n
√
jωKE
I(jω) RE LEB
LE
+
Fig. 5: AIC Loudspeaker Model
I(jω) ZE ZAIC I
AIC (jω)
E(jω)
+
− ZM − EAIC (jω)
+
ZEM 1:n
Fig. 6: AIC Loudspeaker Model - Compact Form
model, as in figure 6. The parameters are defined as ZT C = ∞

follows:
ZE (jω) = RE + jωLEB In the case of A.I.C. shorted, the impedances are:
√
p jωKE jωLE ZAIC
ZM (jω) = jωKE ||jωLE = √ ZV C = ZE + ||ZM + ZEM
jωKE + jωLE n2
ZT C = 0
We call ZV C the impedance seen from the voice coil
terminals and ZT C the impedance seen from the
In the case of both voice coil and A.I.C. driven with
A.I.C. coil terminals.
voltage source E and EAIC = gE, where g is a
Applying some circuit theory the equations of ZV C generic gain factor :
and ZT C can be calculated in the different connec- 0 0
tion configurations of the two coils. (ZE + ZEM )(ZM + ZAIC ) + ZM ZAIC
ZV C = 0
ZAIC + ZM (1 + g/n)
In the case of A.I.C. coil left open, the impedances
are: (ZAIC ZM ) + (ZE + ZEM )(ZAIC + n2 ZM )
ZV C = ZE + ZM + ZEM ZT C =
[g(ZE + ZEM ) + ZM (n + g)]

Page 5 of 9
Impedances – AIC Separately Driven (6'' Loudspeaker) Voice Coil Impedance – AIC Open (6'' Loudspeaker)
100 100
Voice Coil Measured Voice Coil Measured
90 Voice Coil Modeled 90 Voice Coil Modeled
AIC Coil Measured
80 AIC Coil Modeled 80
70 70
60 60
magnitude (Ω)
magnitude (Ω)
50 50
40 40
30 30
20 20
10 10
0 0
10 100 1k 10k 20k 10 100 1k 10k 20k
frequency (Hz) frequency (Hz)
Voice Coil Impedance – AIC Shorted (6'' Loudspeaker) Voice Coil Impedance – AIC in Parallel (6'' Loudspeaker)
100 20
90 Voice Coil Modeled 18 Voice Coil Modeled
80 16
70 14
60 12
magnitude (Ω)
magnitude (Ω)
50 10
40 8
30 6
20 4
10 2
0 0
10 100 1k 10k 20k 10 100 1k 10k 20k
Fig. 7: Measured vs modeled impedances - 6” Loudspeaker
0
where ZAIC = ZAIC
n2 is the A.I.C. coil impedance bounded non-linear least square error NLSQE
reported to the voice coil. method.
If the voice coil and the A.I.C. coil are physi-
cally connected in parallel the impedance seen from We wrote a Scilab script to load the measured
the amplifier is the parallel between the latest two data[6] and used the NLSQE method implemented
impedances with g = 1: in Scilab with the leastsq[7] function to fit the data.
Z = ZV C ||ZT C
We choose to use an error function based on more
than an impedance curve. In particular we had the
Using the same model, it’s also possible to calculate
best fitting results using a weighted function of the
the A.I.C. impedance when the voice coil is shorted
voice coil impedance with A.I.C. separately driven
or open, of course these condition are not the oper-
and with A.I.C. open, and the A.I.C. impedance
ating condition of the loudspeaker.
with the voice coil separately driven.
5. PARAMETERS ESTIMATION
The model parameters are estimated fitting the The parameters for the 6” and 10” tested loudspeak-
model to the impedance measurements using a ers tested are reported in the following table:

Page 6 of 9
Impedances – AIC Separately Driven (10'' Loudspeaker) Voice Coil Impedance – AIC Open (10'' Loudspeaker)
70 70
Voice Coil Modeled Voice Coil Modeled
60 AIC Coil Measured 60
AIC Coil Modeled
50 50
magnitude (Ω)
magnitude (Ω)
40 40
30 30
20 20
10 10
0 0
10 100 1k 10k 20k 10 100 1k 10k 20k
Voice Coil Impedance – AIC Shorted (10'' Loudspeaker) Voice Coil Impedance – AIC in Parallel (10'' Loudspeaker)
70 16
Voice Coil Modeled Voice Coil Modeled
14
60
12
50
10
magnitude (Ω)
magnitude (Ω)
40
8
30
6
20
4
10 2
0 0
10 100 1k 10k 20k 10 100 10k 20k
frequency (Hz)
Fig. 8: Measured vs modeled impedances - 10” Loudspeaker
Param. 6” Lspk 10” Lspk Units loudspeaker equipped with A.I.C. additional coil.
RE 5.71 5.74 Ω The model is based on the simplified Blasizzo loud-
LEB 0.130 0.0237 mH speaker thermal model [8].
KE 0.05 0.118 semihenrys
LE 1.47 0.983 mH The parameters in same figure are defined as follows:
n 1.29 1.20
LCES 22.4 30.3 mH Pj power dissipated by the voice coil
CM ES 209 142 µF
RES 80.4 58.1 Ω Pe power dissipated by eddy currents
RAIC 18.0 16.0 Ω
Ptc power dissipated by additional fixed A.I.C. coil
LAIC 0.0748 0.253 mH
Cb thermal capacity of voice coil
A comparison of the measured and modeled data is
shown in figures 7 and 8. Cm thermal capacity of magnetic assembly
6. THERMAL MODEL Rcb equivalent conduction thermal resistance from

Figure 9 reports a proposed thermal model for a voice coil to ambient

Page 7 of 9
Ptc
Pe
Cb Cm
Tb
Tm
Pj Rcb Rtbm Rtma
Ta
Fig. 9: AIC Loudspeaker Thermal Model
Rtbm thermal resistance between voice coil and at least three measurements are required: Pj , Tb and
magnetic assembly Pe with the A.I.C. coil open; Ptc and Tm with the
voice coil open; Pe with both voice coil and A.I.C.
Rtma thermal resistance between magnetic assem- driven. In the latter case the Pe and Ptc sources are
bly and ambient connected in parallel and can be reduced to a single
T b voice coil temperature source, in fact it is possible to use a single Pe0 current
source that takes into account both the effects of the
T m magnetic assembly temperature power dissipated from the A.I.C. coil and the power
dissipated by eddy currents. We have to point out
T a ambient temperature that the effect of the A.I.C. is the reduction of the
eddy currents.
There are two basic modifications to the Blasizzo
simplified model: This measurement procedure is very time consum-
ing, unfortunately at the time of writing tests are
• A current source Ptc that models the power dis- still undergoing.
sipated by the A.I.C. coil is introduced. The
current is injected into the Tm node, since the 7. CONCLUSIONS
additional coil is wound around the pole piece The proposed electrical equivalent model for the
and is at same magnet temperature. loudspeaker equipped with A.I.C. fits quite well with
measurements. This model can help system design-
• An Rcb thermal resistance is added to model the
ers to better understand A.I.C. usage, in particular
heat conduction from voice coil to ambient, al-
the parameters of the proposed model allow to cor-
though the heat flux is low this element cannot
rectly simulate the low frequency electro-acoustical
be neglected from the model to correctly con-
behavior of the loudspeaker.
sider the raise of voice coil temperature due to
the additional fixed coil heating. Regarding to the practical application of this study,
it must be noted that a simple approach to get
The power dissipated by eddy currents is dependent the correct motional impedance parameters of the
on the A.I.C. coil connection and driving. In order loudspeaker is to simply measure the voice coil
to correctly estimate the thermal model parameters impedance with the A.I.C. shorted or open.

Page 8 of 9
Also at low frequencies the A.I.C. can be seen as a [6] D. Ponteggia, “CLIO 10 Sinusoidal File Struc-
simple resistor, thus the current that flows into the ture With Import Examples In SCILAB - Ap-
voice coil can be easily calculated and used to model plication Note 1”, Audiomatica, 2009.
the force on the voice coil in a mechanical simulation
of the loudspeaker. [7] S.L. Campbell, J. Chancelier, R. Nikoukhah,
“Modeling and Simulation in Scilab/Scicos”,
A thermal model of a loudspeaker equipped with Springer, 2006.
A.I.C. is currently under study and can lead to in-
teresting results in term of energy balance of such [8] F.Blasizzo, “A New Thermal Model for Loud-
loudspeaker. speakers”, J. Audio Eng. Soc., Vol. 52, No. 1/2,
2004 January/February.
An extension to a non linear model is very interest-
ing and can be regarded as the natural extension of
this paper.
8. ACKNOWLEDGEMENTS
This work has been supported by Eighteensound,
we would like to thank everyone at their R&D de-
partment. We would like also to thank Mr. Fabio
Blasizzo for his support in thermal modeling and in
the review of this paper.
9. REFERENCES
[1] M. Carlisi, M. Di Cola, A. Manzini, “An Al-
ternative Approach to Minimize Inductance
and Related Distortions in Loudspeakers”, pre-
sented at the AES 118th convention, Barcelona,
Spain, 2005 March 28–31.
[2] J. Vanderkooy, “A Model of Loudspeaker Driver
Impedance Incorporating Eddy Currents in the
Pole Structure”, J. Audio Eng. Soc., Vol. 37,
No. 3, (1989 March).
[3] W. M. Leach, Jr., “Loudspeaker Voice-Coil
Inductance Losses: Circuit Models, Parame-
ter Estimation, and Effect On Frequency Re-
sponse”, J. Audio Eng. Soc., Vol. 50, No. 6
(2002 June).
[4] M. Dodd, W. Klippel, and J. Oclee-Brown,
“Voice Coil Impedance as a Function of Fre-
quency and Displacement”, presented at the
AES 117th Convention, San Francisco Octo-
ber 2004, Convention Paper 6178.
[5] K. Thorborg, A. D. Unruh, “Electrical Equiv-
alent Circuit Model for Dynamic Moving-Coil
Transducers Incorporating a Semi-Inductor”, J.
Audio Eng. Soc., Vol. 56, No. 9 (2008 Septem-
ber).

Page 9 of 9

18 Sound 128th AES London AIC Model

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

18 Sound 128th AES London AIC Model

Uploaded by

Copyright:

Available Formats

Audio Engineering Society

Electrical Circuit Model For a Loudspeaker

Correspondence should be addressed to Daniele Ponteggia (daniele@studioponteggia.it)

1. INTRODUCTION moving coil and made accessible via two additional

Fig. 1: Loudspeaker Model

Fig. 2: Thorborg Model

AES 128th Convention, London, UK, 2010 May 22–25

Sinusoidal – 6'' Loudspeaker Sinusoidal – 6'' Loudspeaker

80.0 108.0 80.0 108.0

60.0 36.0 60.0 36.0

40.0 -36.0 40.0 -36.0

20.0 -108.0 20.0 -108.0

0.0 -180.0 0.0 -180.0

Sinusoidal – 6'' Loudspeaker Sinusoidal – 6'' Loudspeaker

60.0 36.0 60.0 36.0

40.0 -36.0 40.0 -36.0

20.0 -108.0 20.0 -108.0

0.0 -180.0 0.0 -180.0

AES 128th Convention, London, UK, 2010 May 22–25

Sinusoidal – 10'' Loudspeaker Sinusoidal – 10'' Loudspeaker

80.0 108.0 80.0 108.0

60.0 36.0 60.0 36.0

40.0 -36.0 40.0 -36.0

20.0 -108.0 20.0 -108.0

0.0 -180.0 0.0 -180.0

Sinusoidal – 10'' Loudspeaker Sinusoidal – 10'' Loudspeaker

60.0 36.0 60.0 36.0

40.0 -36.0 40.0 -36.0

20.0 -108.0 20.0 -108.0

0.0 -180.0 0.0 -180.0

AES 128th Convention, London, UK, 2010 May 22–25

IAIC (jω) RAIC LAIC

Fig. 5: AIC Loudspeaker Model

Fig. 6: AIC Loudspeaker Model - Compact Form

model, as in figure 6. The parameters are defined as ZT C = ∞

AES 128th Convention, London, UK, 2010 May 22–25

Fig. 7: Measured vs modeled impedances - 6” Loudspeaker

AES 128th Convention, London, UK, 2010 May 22–25

Fig. 8: Measured vs modeled impedances - 10” Loudspeaker

6. THERMAL MODEL Rcb equivalent conduction thermal resistance from

AES 128th Convention, London, UK, 2010 May 22–25

Fig. 9: AIC Loudspeaker Thermal Model

AES 128th Convention, London, UK, 2010 May 22–25

AES 128th Convention, London, UK, 2010 May 22–25

You might also like