ISMA 2014, Le Mans, France

Experimental Study of a Guitar Pickup

P. Lottona , B. Lihoreaua and E. Brasseurb
a
LAUM UMR 6613, Universite du Maine, Avenue Olivier Messiaen, 72085 Le Mans, France
b
LAUM - UMR CNRS 6613, Avenue Olivier Messiaen, 72085 Le Mans Cedex 9, France
[email protected]

355
ISMA 2014, Le Mans, France

This work deals with experimental study of the string-pickup coupling in an electric guitar. The effect of the
string movement on the pickup output signal is studied by mean of an experimental bench which provides a purely
sinusoidal string movement along only one direction. It appears out that the pickup has a weakly non-linear
behaviour and is primarily sensitive to one of the two polarizations of the string. The effect of the pickup on the
string is examined in a second step. A lab guitar prototype is developed which allows to study this coupling in a
real gaming situation. The results show that the pickup has an impact on the energy exchange between the two
polarizations and reveals that the presence of pickup decreases the guitar sustain if it is set too close to the string.

1 Introduction by a complex signal[15, 18].

In a first part, the effect of the string motion on the pickup
Compared to its acoustic cousin, the electric guitar is a output is studied. For this purpose, an experimental set-up
modern instrument. It was born in the twentieth century, has been developed in order to drive the pickup with a lonely
the need to develop electrified instruments being then polarisation of the string. The effect of each polarisation on
the difficulty of certain instruments to be heard in large the pickup output can then be studied. In a second part, the
ensembles. The first electric guitars were developed in the effect of the pickup on the string vibration is studied. A
late 1920s. They first looked like acoustic guitars equipped prototype of a lab electric guitar has been designed. It allows
with transducers. But, in the early 1940s, the electric to study the decrease of a plucked response as a function of
guitar took a design very similar to those in use today, that to the position of the pickup.
is a solid-body guitar equipped with compact magnetic
pickups[1].
The literature on the acoustic guitar is fairly abundant, z
(see for example [2, 3, 4]). The admittance or the point
mobility at the bridge seems to be a key point to model an
acoustic guitar [5, 6]. This does not seem to be the case
for an electric guitar. Indeed, the body of an electric guitar
consists of a wooden board (solid-body) and the exchange y
energy from the string to the body is not essential. This is the
pickup which senses the vibration of the string and translate
d0
it into an electric signal. There are different types of pickups:
piezoelectric, optic or magnetic. The magnetic pickups
are the most popular. They are comprised of permanent
magnets surrounded by a coil of wire with typically several
thousand turns. The guitar strings consist of wires made of Figure 1: Definition of the axis.
a ferromagnetic material and are parallel to the face of the
magnets. The magnetic field resulting from the interaction
between the pickup and the strings, and thus the magnetic
flux through the coil, depends critically on the position of
2 Effects of the string on the pickup
the strings. Therefore, moving a string changes the magnetic signal
flux through the coil. According to Faradays law the current
induced inside the coil is proportional to the time rate of 2.1 Experimental set-up
change of the magnetic flux through the coil. As the string
To study the effect of each polarisation of the string
moves through the magnetic field, a time-varying current is
movement on the pickup signal, we used the system shown
then produced in the coil. This current is used to produce a
in Fig. 2. A portion of steel string (diameter 1.42 mm) is
potential drop across a resistor, which is then amplified and
fixed on a non-magnetic support, itself fixed to a shaker. The
sent to a speaker.
driven frequency F0 and the amplitude of the displacement
Although the electric guitar is a very popular instrument,
dmax of the shaker can be tuned by the operator. In the
the scientific literature about it is surprisingly quite poor. One
following, the frequency is chosen to be F0 = 85 Hz, which
can cite [1] for an overview, [7, 8, 9, 10] who studied the the
approximatively corresponds to the fundamental frequency
coupling between the string vibrations and those of the body,
of the open low E string. The pickup under test (here a single
[11, 12] who tried to propose a model of the guitar pickup
coil Seymour Duncan SSL-1) is set on a precision movement
and [13, 14] for sound synthesis. The aim of this article is to
device which allows to adjust the distance at rest d0 between
study the string-pickup coupling in an experimental way.
the string and the magnet. Two test configurations are
The pickup is driven by the motion of the strings. This
considered, corresponding to each polarization of the string.
motion is nonplanar and may be seen as the composition of
The motion of the string is measured by an accelerometer
two movements in perpendicular planes. A string vibrates in
fixed on the string support. Since the shaker can exhibit
two polarisations[15, 16], one parallel to the pickup (called
a non linear behaviour, additional content is added at the
y-polarization in the following) and the other perpendicular
harmonics of the excitation frequency, causing a harmonic
to the pickup (called z-polarization)(Fig. 1). Thus, when
distortion of the displacement. To ensure a purely harmonic
considering a single string, the pickup can be seen as a
motion of the string, an active harmonic control technique
system with two inputs and one output. In addition, the
is used [17]. This technique adds higher harmonics to
vibration of a plucked string can be seen as a sum of
the excitation voltage u(t) provided to the shaker, so that
decreasing harmonic components: the pickup is thus driven
u(t) = U1 sin(0 t) + k=2N
Uk sin(k0 t + k ), N being the

356
 ISMA 2014, Le Mans, France

number of harmonics added to the input voltage. The goal 250

d0=0.5
of the active harmonic control is to eliminate the higher
d =1
0
harmonics contained in the displacement signal so that only 200 d =1.5
the fundamental response remains. It searches for the best 0
d =2
combination of amplitudes Uk and phases k , that relate to 0
d =2.5

Vrms (mV)
linearizing the displacement signal d(t). 150 0
d0=3

100

50

0
0 0.2 0.4 0.6 0.8 1 1.2
dmax (mm)

Figure 3: Pickup output signal (RMS) produced by a string

Figure 2: Experimental set-up to study the influence of each motion along z-axis according to the amplitude of the string
string polarisation on the pickup output. On the displacement for several d0 .
photography, the pickup is set to measure z-polarisation. 25
200 0.8

150 0.6

2.2 Results 100 0.4

Displacement (mm)
Pickup (mV)
2.2.1 Displacement along z-axis 50 0.2

0 0
The effect of a displacement of the string along z-axis
is first studied. The output RMS voltage of the pickup 50 0.2
is plotted as a function of the amplitude of the string
100 0.4
displacement for several initial positions d0 in Fig.3. One
can see expected results: the pickup signal RMS amplitude 150 0.6

increases with the amplitude of displacement d(t). Moreover, 200 0.8

0 0.01 0.02 0.03 0.04 0.05
for a fixed amplitude of displacement, the output signal Time (s)
RMS amplitude decreases when the position at rest d0
increases. In addition, the input/output characteristic of the Figure 4: Temporal evolutions of the string displacement
sensor seems to be non-linear when the pickup is close to along z-axis (blue line) and the pickup signal (green line)
the string (weak d0 ). When d0 increases, the input/output for d0 = 1.5 mm and dmax = 0.78 mm.
characteristic of the pickup seems to be linear. Figure 4
represents the temporal evolutions of the accelerometer
signal (blue line) and the pickup signal (green line) for seen more quantitatively on Fig.8 which represents the
d0 =1.5 mm and for a displacement of dmax = 0.78 mm. The relative amplitude of the first three harmonics of the pickup
pickup signal seems to be weakly distorted. This last result output signal for several displacement amplitudes dmax and
can be appreciated quantitatively on Fig.5, which represents for two rest positions d0 = 0.5 mm (top figure) and d0 = 2
the relative amplitude of the first three harmonics of the mm (bottom figure). The response of the pickup according
pickup signal for different displacement amplitudes and for y-polarization is more distorted than the one according
d0 = 0.5 mm (top figure) and d = 2 mm (bottom figure). z-polarization. This is especially the case for d0 = 0.5
Non-linearities remains weak even for large displacements. mm and for strong excitations. Under these conditions,
H2
In any case, H1 < 10 dB and H1
H3
< 20 dB. At last, one can the amplitude of the second harmonic is higher than the
notice that the pickup signal and the accelerometer signal fundamental one. This is an expected result: According
are /2 out of phase : the pickup is a velocity sensor. to the symmetry of the problem, excitation in the y-axis
at F0 results in a pickup response at 2F0 . Nevertheless,
2.2.2 Displacement along y-axis given the symmetry of the problem, this behaviour should be
observed whatever d0 and dmax . It is not the case. This can
In this section, the same measurements as in the previous be explained by the fact that the string may not be perfectly
section are performed but the displacement is now along the positioned at the center of the magnet top face.
y-axis. Figure 6 represents the pickup output RMS level
according to the amplitude of displacement for different 2.2.3 Comparison between both polarisation
initial positions d0 . As expected, the pickup output RMS
voltage increases with the amplitude of displacement. Comparing results presented on Figs. 3 and 6, we can see
Figure 7 represents the temporal evolutions of the that the pickup is much more influenced by the motion along
accelerometer signal (blue line) and the pickup signal (green z-polarisation: the motion along y-polarisation is negligible.
line) for a rest position d0 =1.5 mm and for a displacement This result confirms results obtained by other means in a
amplitude of dmax = 0.8 mm: the pickup signal is clearly previous study [10].
distorted. The non-linear behaviour of the pickup can be

357
ISMA 2014, Le Mans, France

0 35
d =0.5
0
5
30 d0=1
10
d0=1.5
Relative level (dB)

15 25
d0=2

Vrms (mV)
20 d0=2.5
20
25 d0=3
15
30
d =0.1
35 max
10
d =0.2
max
40
d =0.3
max 5
45
1 2 3
Harmonic rank 0
0 0.2 0.4 0.6 0.8 1
0
dmax (mm)
10
Figure 6: Output signal produced by a string motion along
Relative level (dB)

20
y-axis according to the amplitude of the string displacement
30 dmax=0.1
for several d0 .
dmax=0.2
40 40 1
d =0.3
max
dmax=0.4
50
dmax=0.6
20 0.5

Displacement (mm)
60 dmax=0.8
Pickup (mV)
dmax=1
70
1 2 3 0 0
Harmonic rank

Figure 5: Harmonics amplitude for several amplitudes of 20 0.5

string displacement along z-axis for d0 = 0.5 mm (top) and
d0 = 2 mm (bottom). The reference level is the amplitude of
the fondamental H1. Dotted lines are represented for eyes 40 1
0 0.01 0.02 0.03 0.04 0.05
guiding. Time (s)

Figure 7: Temporal evolutions of the string displacement

3 Effects of the pickup on the string along y-axis (blue line) and the pickup signal (green line)
motion for d0 = 1.5 mm and dmax = 0.8 mm.

The previous part of the work allows to study the

behaviour of the pickup for a sinusoidal excitation by string in both polarisations, two Polytec laser vibrometers
decoupling the two polarisations. In this section, we study (OFV 3000 and OFV 3001) and controllers (OFV302 and
the coupling between the string and the pickup in real OFV303) are used.
gaming situation. For that purpose, we have designed a lab The pickup signal and the vibrometers signals are
guitar prototype which allows to precisely adjust the position recorded with a data acquisition card connected to a
of the pickup and provides repeatable pluck responses. computer. The sampling frequency is 10 kHz and the
acquisition time is 30 s.

3.1 Experimental set-up

3.2 Definition of the sustain
The experimental setup is shown in Fig. 9
A guitar string is fixed on an exotical wooden beam The vibration of a plucked string can be seen as the
(sapele) having the following dimensions : 910mm x 140 superposition of several harmonics, each of them having
mm x 140mm. The string has a diameter of 1,42 mm their own decay rate. So a sustain can be defined harmonic
and is tuned as open low E ( f0 = 82 Hz). The string is by harmonic. In this work, the sustain is defined for the
fixed by some mechanical accessories whose references are entire signal. Thus, the definition of the sustain is similar
TonePros LPCM02 C for bridge/tailpiece set, Graph Tech to that used in room acoustics to estimate the reverberation
PQ 6060 00 for the slotted nut and Gibson PMMH-015 time [19]. The main idea is to calculate the remaining energy
for grover tuners. The pickup is set on a mechanical arm on in the plucked response as :
which some precision movement pieces are fixed. Thanks Z
to this system, the pickup position can be adjusted along Er (t) = s(t0 )2 dt0 , (1)
t
the 3 axes. The string is excited by a plectrum attached
to a moving system (ZABER T-LMS025B) controlled by where s(t) represents the pickup signal or the vibrometers
computer via a serial RS232. It allows a good repeatability signals. It is more convenient to express the energy decay in
(see following section). To visualize the motion of the decibel with the total energy of the signal as reference :

358
 ISMA 2014, Le Mans, France
10 22
Pickup
0 Zpol
20 Ypol
Relative level (dB)
10

18
20 dmax=0.1

Sustain (s)
dmax=0.2
30 dmax=0.3 16

dmax=0.4
40
dmax=0.6 14
50 d =0.8
max
dmax=1
12
60
1 2 3
Harmonic rank
10
10
0 2 4 6 8 10
d0 (mm)
0
Figure 10: Sustain according to d0 .
10
Relative level (dB)

20
dmax=0.1
pickup is close to the string, he tends to reduce the sustain.
30 d =0.2
max Beyond d0 3 mm, the sustain no further change : the
dmax=0.3
40 pickup has no influence on the string.
dmax=0.4
50 dmax=0.6
In addition, Fig. 10 shows that the sustains of z-
dmax=0.8
polarisation and pickup signal are almost identical up to a
60
d =1 distance d0 4 mm. This is a confirmation of the previous
max
70 section : the pickup mainly senses the z-polarisation.
1 2 3
Harmonic rank Beyond this distance, the pickup sustain diverges: the
coupling with the string is too weak and the pickup signal is
Figure 8: Harmonics amplitude for several amplitudes of too noisy to estimate a 20 dB decay.
string displacement along y-axis for d0 = 0.5 mm (top) and Regarding the sustain of the y-polarisation, there is an
d0 = 2 mm (bottom). The reference level is the amplitude of unexpected behaviour: the sustain is greater when the pickup
the fondamental H1. Dotted lines are represented for eyes is close to the string. The pickup acts as a vibration sustainer.
guiding. This is particularly the case for d0 = 2 mm where the sustain
of the y-polarisation goes through a maximum.
These results are confirmed by the analysis of figure
11 which represents the energy decrease as calculated
with eq. 2 as a function of time and for different initial
distance d0 . The first observation is that the pickup and
z-polarisation decreases are mingled. This is a conclusion
of the previous section: the pickup mainly senses the string
vibration of the z-polarisation. The divergence between
Figure 9: Lab guitar prototype theses two decreases for d0 = 5 mm is visible. Moreover,
when the pickup is close to the string, the energy decrease
of y-polarisation is slower than the two others. At last,
! one can see an interesting phenomenon when focussing on
Er (t) the decrease of both polarisations. For d0 = 3 and 5 mm,
dec(t) = 10 log . (2)
Er (0) curves intertwine. Everything happens as if there was an
In the following, we define the sustain as the time energy exchange between both polarisations. The decrease
required for a decrease of 20 dB. is locally accelerated or slowed down depending on the
direction of the exchange. Again, one can see the action of
the pickup as these interlacings depend on the position of
3.3 Results the pickup. When the pickup is very close to the string, there
Fig. 10 represents the sustains of the pickup signal is no interlacing. As the y-polarization decrease is slower, it
and vibrometers signals in both polarisations according to is conceivable that the pickup promotes energy transfer from
the distance d0 between the string and the pole piece. The z- to y-polarisation.
pickup is a Seymour Duncan SSL1 positioned at 41.5 mm
from the bridge. For each distance d0 , 10 measurements
are performed. The curves represents the average. Errobars 4 Conclusions
are also given and allow to ensure that the set-up provides a
In this paper, we experimentally study the coupling
sufficient repeatability.
between the pickup and the string in an electric guitar. In
Fig. 10 shows that the sustain of z-polarisation is a
the first part, an experimental system is designed in order to
function globally growing up to d0 3 mm. Beyond this
provide a purely sinusoidal string movement along only one
distance, the sustain no further change. The pickup has
polarization. Results show that the pickup is only sensitive
therefore an influence on the z-polarisation : when the

359
ISMA 2014, Le Mans, France

[2] N. H. Fletcher and T. D. Rossing, Physics of Musical

Instruments, Springer-Verlag, New York, (1998)
[3] O. Christensen and B. B. Vistisen, Simple model for
low frequency guitar function, J. Acoust. Soc. Am.
68(3), 758-766 (1980)
[4] J. E. Popp, Four mass coupled oscillator guitar
model, J. Acoust. Soc. Am. 131(1), 829-836 (2012)
[5] J. Woodhouse, On the synthesis of guitar plucks Acta
Acust. Acust., 90(4), 928-944 (2004)

[6] J. Woodhouse and R. S. Langley, Interpreting the Input

Admittance of Violins and Guitars Acta Acust. Acust.,
98(4), 611-628 (2012)
[7] H. Fleischer and T. Zwicker, Mechanical vibrations of
electric guitars, Acustica, 84(4), 758-765 (1998)
[8] H. Fleischer and T. Zwicker, Investigating dead spots
of electric guitars, Acustica, 85(1), 128-135 (1999)
[9] H. Fleischer, Vibration of an electric bass guitar, Acta
Acust. Acust., 91(5), 247-260 (2005)

[10] A. Pate, J.-L. Le Carrou, B. Fabre, Predicting the

decay time of solid body electric guitar tones, J.
Acoust. Soc. Am., 135 (3045), (2014)
[11] G. Lemarquand and V. Lemarquand, Calculation
method of permanent-magnet pickups for electric
guitars, IEEE Trans. Magn., 43(9), 3573-3578 (2007)
[12] N. G. Horton and T.R. Moore, Modeling the magnetic
pickup of an electric guitar, Am. J. of Phys. 77(2), 144-
150 (2009)

[13] R. C. D. Paiva and J. Pakarinen and V. Valimaki,

Figure 11: Energy Decrease for several d0 . a. d0 =0.5, 3 and Acoustics and Modeling of Pickups, J. Audio Eng.
5 mm soc., 60(10), 768-782 (2012)
[14] N. Lindroos and H. Penttinen and V. Valimaki,
to one of the two polarizations. In addition, the pickup Parametric Electric Guitar Synthesis, Comp. Music
output signal appears to be fairly linear except when the J., 35(3), 18-27 (2011)
excitation is very strong and/or when d0 is very low. [15] C. Valette, Mechanics of Musical Instruments, chapter
The effect of the pickup on the string is examined in a The mechanics of vibrating strings, edited by A.
second step. An lab guitar prototype is developed which Hirschberg, J. Kergomard and G. Weinreich, Springer-
allows to study this coupling in a real gaming situation. Verlag, vienna, (1995)
Results show that the pickup has an effect on the string
motion. This effect is investigated through the energy decay [16] O. OReilly and P. J. Holmes, Non-linear, non-planar
of the pickup output and the energy decay of the vibrometers and non-periodic vibrations of a string, J. Sound Vib.,
signals in both polarisations. The pickup has an effect on the 153(3), 413-435 (1992)
energy exchanges between both polarisations depending on
the distance d0 between the string and the pickup. At last, it [17] A. Novak, P. Lotton, L. Simon, Dynamic
is shown that the pickup has an impact on the sustain: when Measurement of Loudspeaker Suspension Parameters
d0 is weak, the pickup slow down the z-polarisation and the Using an Active Harmonic Control Technique.,
sustain diminishes. Convention paper 9054, 136th Audio Engineering
Society Convention, Berlin (2014)
[18] H. A. Conklin, Generation of partials due to nonlinear
References mixing in a stringed instrument, J. Acoust. Soc. Am.,
[1] C. Gough, The science of string instruments, chapter 105(1), 536-545 (1999)
Electric guitar and violin, edited by T. Rossing, [19] M. R. Schroeder, New method of measuring
Springer-Verlag, New York, (2010) reverberation time, J. Acoust. Soc. Am., 37(3), 409-
412 (1965)

360