sensors

Article
Effect of Torsion Stress on the Offset and Sensitivity
of Diagonal and Off-Diagonal GMI in
Amorphous Wires †
Julie Nabias, Aktham Asfour * and Jean-Paul Yonnet
Univ. Grenoble Alpes, CNRS, Grenoble INP, G2ELab, 38000 Grenoble, France; chickichik@gmail.com (J.N.);
Jean-Paul.Yonnet@g2elab.grenoble-inp.fr (J.-P.Y.)
* Correspondence: aktham.asfour@g2elab.grenoble-inp.fr; Tel.: +33-476-826-395
† This paper is an extension version of the conference paper: Nabias, J., Asfour, A., Yonnet, J.-P. “Torsional stress
effect on Giant Magneto-Impedance (GMI) in amorphous microwires”, In Abstract Book of the 12th European
Magnetic Sensors and Actuators Conference (EMSA-2018), Athens, Greece, 1–4 July 2018.




Received: 26 September 2018; Accepted: 19 November 2018; Published: 24 November 2018


Abstract: In this paper, the torsional stress effect on Giant Magneto-Impedance (GMI) was studied
in Co-rich amorphous wires. The study, which was conducted in the context of the development
of a current clamp based on GMI, considered torsion as a parameter of the influence of this sensor.
Both diagonal, Z11 , and off-diagonal, Z21 , components of the impedance tensor were investigated.
The samples were Co-rich wires with a 100 µ diameter. The wires were twisted positive and negative
angles with respect to a reference position. For each component of the impedance, the intrinsic
sensitivity and offset were measured as a function of the rotation angle. The results showed that
the sensitivity of the diagonal component at a given working point slightly increased for angles
between −90◦ to +90◦ , whereas the sensitivity was almost constant for the off-diagonal component at
zero-field. The intrinsic offset in the diagonal configuration was almost unchanged for the rotation
angles considered, whereas this offset increased in the off-diagonal configuration. Furthermore,
the GMI ratio of Z11 was also measured as a function of the rotation angle for comparison purposes
with known data. The maximum of this ratio was obtained for a rotation angle of about 50◦ .

Keywords: GMI sensor; diagonal; off-diagonal; sensitivity; offset; torsion stress; parameter of influence

1. Introduction
Giant Magneto-Impedance (GMI) is a significant change of the impedance of some soft magnetic
materials when they are subjected to an external magnetic field. This change of the impedance is
directly related to the change of the skin depth of the high-frequency current in the magnetic conductor
through the change of the magnetic permeability of the material with the applied magnetic field [1].
The GMI effect is investigated for the realization of magnetic sensors. These sensors are based
on the impedance measurement of the sensitive element. They combine excellent features such
as high sensitivity and a large bandwidth (from DC to several megahertz). Despite its potential,
this technology is in fact still not very mature, especially for industrial applications. To our knowledge,
only a few “concrete” realizations or commercial versions of these sensors have been achieved [2–7].
The systematic use of GMI sensors requires a clear identification of the areas of application for which
these sensors can have clear advantages when compared to other mature magnetic sensor technologies.
Contactless electrical current measurement could be one of these areas. For such applications, the ability
of the GMI sensor to measure both DC and AC magnetic fields with the same sensitive element is clearly
an important feature. Moreover, the mechanical flexibility of some GMI elements (like amorphous

Sensors 2018, 18, 4121; doi:10.3390/s18124121 www.mdpi.com/journal/sensors

Sensors 2018, 18, 4121 2 of 11

wires) is another key and very useful advantage since it enables the sensitive wire to be deformed in
order to be aligned with the magnetic field produced by the measured current.
The work presented in this paper, which deals with the torsion stress effect, was actually conducted
in the context of a specific application. This concerns the use of GMI to realize a toroidal current clamp
(probe), which is mechanically flexible (not rigid) and which allows for both DC and AC measurements
with the same sensitive element. By far, satisfying these features could not be easily achieved using the
magnetic sensor technologies available. In this context, GMI clearly has a decisive advantage.
The basic principle of the current clamp using GMI is quite simple. The GMI wire circles the
conductor that carries the measured current Im . This current produces a circumferential field, Hm ,
which is measured by the GMI element. In practice, the clamp must be opened to circle the conductor
and then closed to perform the measurement. Such a use of the clamp involves repetitive mechanical
stresses. The effect of these stresses, as parameters of influence, on the GMI response needs to be
carefully investigated. Repetitive bending is one such type of mechanical stress. Torsion stress
could also be involved. It actually depends on how the user manipulates the sensor to perform
the measurement. In practice, both bending and torsion stresses could be combined, resulting in a
global change of the GMI response. However, for obvious reasons of simplicity, the two effects were
studied separately.
Unlike the tensile effect, which has been intensively investigated (References [8–16] provide a
non-exhaustive list of these studies), the effect of bending and torsion stresses on GMI have been
investigated far less in amorphous GMI wires. The bending stress effect on the GMI response in
amorphous wires has been addressed in our recent work [17,18] and more recently in Reference [19].
To our knowledge, the torsion stress effect on diagonal GMI was studied in some
publications [16,20–26]. These publications evidently have great merit. However, on the one hand,
most of them deal mainly with the diagonal component of the impedance tensor. Moreover, only the
change of the GMI ratio of this component was considered under torsion stress. While this ratio is
frequently used as a factor of merit to quantify the GMI effect, it is not the most relevant quantity to
take into account for the sensor implementation. In fact, some GMI sensitive elements could have a
large GMI ratio and at the same time exhibit a low sensitivity around the working point [27]. In this
paper, the change of the diagonal component under torsion stress was addressed with particular
attention paid to the change of the most relevant quantities, namely the intrinsic offset and sensitivity
at a given working point. Nevertheless, the GMI ratio was also briefly considered in this study for
comparison purposes with other previous studies dealing with the torsion stress effect. On the other
hand, to our knowledge, the impact of torsion on the off-diagonal component has only been studied in
a few publications [28–30], despite the promising characteristics of this component for GMI sensors.
The change in sensitivity and offset near the zero-field point was not investigated. This is why an
investigation of the torsion stress effect on this off-diagonal component was conducted with a primary
focus on the change of the intrinsic sensitivity and offset near the zero-field. In this study, the torsion
stress was considered as a parameter of influence that affects the response of the GMI current sensor.
Section 2 of the paper presents, firstly, a brief overview of the quantities considered and the
general approach of the study. Secondly, a description of the setup and the experimental conditions is
given. The first results obtained are illustrated and discussed in Section 3.

2. Quantities Considered and Experimental Setup

In a GMI sensor, the sensitive wire is supplied by a high-frequency current, iac , of constant
amplitude, as illustrated in Figure 1.
 Sensors
Sensors 2018,
2018, 18,18, x FOR PEER REVIEW
4121 3 of 311of 11
Sensors 2018, 18, x FOR PEER REVIEW 3 of 11

Figure1.1.Principle
Figure Principle of
of the diagonaland
the diagonal andoff-diagonal
off-diagonalGMI
GMI sensor.
sensor.
Figure 1. Principle of the diagonal and off-diagonal GMI sensor.
The impedance isisa atensor
The which has two main
main components:the the diagonal,𝑍 Z,11 , and off-diagonal,
The impedance
impedance is a tensor
tensor which
which has
has two
two main components:
components: the diagonal,
diagonal, 𝑍 ,and
and off-diagonal,
off-diagonal,
𝑍 , components [31–34]. The typical characteristics of these components as functions of the magnetic
, components [31–34]. The typical characteristics of
Z21𝑍, components [31–34]. The typical characteristics ofthese
thesecomponents
componentsas functions
as functionsof the
of the magnetic
magnetic
field, H, are illustrated in Figure 2.
field, H, are illustrated in Figure 2.
field, H, are illustrated in Figure 2.

Figure
Figure 2. 2. Typical
Typical behaviorsofofthe
behaviors thediagonal
diagonal and off-diagonal
off-diagonalcomponents
componentsofof
the impedance
the tensor
impedance in in
tensor
Figure 2. Typical behaviors of the diagonal and off-diagonal components of the impedance tensor in
amorphous wires.
amorphous wires.
amorphous wires.

Since
Since
Since the the behavior
behavior of
the behavior
of the
of the
the modulus
modulus of
modulus of
of 𝑍𝑍11 is
Z is nonlinear,
is nonlinear, aa GMI
nonlinear, a GMIsensor
GMI sensor based based on
sensor based
on this
this component
on this component
component
requires
requires the
requires the
the use
use
use of
of
of aa bias
a
bias field
bias
field (generally
field
(generally static
(generally
static bias),
bias), H
static bias),
Hbias , to
H fix a working point in the almost
, to fix a working point in
bias, to fix a working point in the almost linear
bias thelinear
almost
region.
linear
region. ItIt is
region. is an
an It axial
is anexternal
axial magnetic
axial external
external field,
field, which
magneticmagnetic could
field,
which couldwhichbe
be produced
could beusing
produced aa coil
produced
using coil or aa permanent
using
or a coil or a
permanent
magnet,
magnet, for example. This external magnetic bias gives rise to an offset which is related to theisvalue
permanent for example.
magnet, for This external
example. magnetic
This external bias gives
magnetic rise to
bias an offset
gives which
rise to anis related
offset to
which the related
value
of the
of the
to the impedance,
impedance,
value |Z 11 (H bias)|,
|Z11(Hbias)|, at
of the impedance, at this
|Zthis field.
11 (Hfield. When
When
bias )|,
the
at this measured
thefield.
measured field,
Whenfield, H
the H m, is applied,
m, is applied,
measured the
field, voltage,
theHvoltage, vvacac,,
m , is applied,
across
theacross
voltage, the sensitive
thevsensitive element
ac , acrosselement
the sensitiveis amplitude
is amplitude
element modulated
modulated
is amplitude by this
by this field. After
field. After
modulated amplitude
byamplitude demodulation,
this field. demodulation,
After amplitude
and offset suppression,
and offset suppression,
demodulation, and offset the output voltage of the sensor is obtained. This
the output voltage of the sensor is obtained. This output is proportional
suppression, the output voltage of the sensor output
is is proportional
obtained. This outputto
to
 Z11

 Z11
is proportional  I ac to
∂Z11 
H m ,∂Hwhere Iac is×the
Iac amplitude
× Hm , where of the Iac excitation current. The
is the amplitude intrinsic
of the sensitivity,
excitation current.
H  I ac  H m , where
H = HIbias
ac is the amplitude of the excitation current. The intrinsic sensitivity,

H H  H bias
The intrinsic
H H sensitivity, denoted as S
bias is defined as being derivative of the impedance curve at
 Ω −11,derivative
denoted
denoted asas SS −11, is defined as being
−11, is defined | as | being derivative of
of the
the impedance
impedance curve
curve at
at the
the bias
bias point
point
 ∂ Z11 

the bias point (S
Ω (
Z11−11 biasH ) = ).
(( SS11 ((HZ ∂H
).).
bias )) 

11 H = Hbias
bias  H
It11is Halso H H  H to introduce the voltage sensitivity, denoted as Sv -11 , which takes into account
possiblebias
H H
bias
the amplitude, It
It is
is also
also possible
Iac , of thetoexcitation
possible to introduce
introduce the
the voltage
voltage sensitivity,
current.
sensitivity, denoted
denoted as
This voltage sensitivity
as SSvv--1111,, which
is defined bytakes
which takes into
Equationaccount
(1)
into account
the
the amplitude,
amplitude, IIacac,, of of the
the excitation
excitation current.
current. This
This voltage
voltage sensitivity
sensitivityis
is defined
defined by by Equation
Equation (1)
(1)
∂| Z11 | 
Sv−11 ( Hbias ) = SΩ−11 ( Hbias ) × Iac = Z11 × Iac (1)
Sv 11 ( H bias )  S11 ( H bias )  I ac   Z11∂H  I ac

Sv 11 ( H bias )  S11 ( H bias )  I ac  H =
H I acHbias (1)
(1)
H H  H bias
H  H bias
One or the other of these quantities will be used, indifferently, in this paper.
One
One or
or the other
other of
thesensor of these quantities
theseon
quantities will
will be
be used,
used, indifferently,
indifferently, inin this
this paper.
paper.no external axial magnetic
A GMI based the off-diagonal component requires, a priori,
A
A GMI
GMI sensor
sensor based
based on
on the
the off-diagonal
off-diagonal component
component requires,
requires, aa priori,
priori, no
no external
external axial
axial magnetic
magnetic
bias since
bias since
both
both
the
the
real,
real,
{ ZZ21 }}, ,and
ReRe{ the imaginary, Im {ZZ21} }, ,parts
partsare
areintrinsically
intrinsically almost linear
bias since both the real, Re{ Z21 } , and the imaginary, Im{Z21 } , parts are intrinsically almost linear
and the imaginary, Im{ almost linear
and asymmetric (odd symmetry) 21 with respect to the zero-field 21 point, as seen in Figure 2. However,
and asymmetric (odd symmetry)
symmetry) with respect
respect to
to the
the zero-field point,
point, as
as seen
seen in Figure
Figure 2.
2. However,
theand asymmetric
appearance (odd
of the off-diagonal with
component in zero-field
wires with circumferential in
anisotropy However,
requires the
the
the appearance
appearance of
of the
the off-diagonal
off-diagonal component
component in
in wires
wires with
with circumferential
circumferential anisotropy
anisotropy requires
requires the
the in
use of a DC current flowing in the wire and producing a circumferential field, as is mentioned
Section 3.2.
 Sensors 2018, 18, x FOR PEER REVIEW 4 of 11

use2018,
Sensors of a18,
DC current flowing in the wire and producing a circumferential field, as is mentioned 4inof 11
4121
Section 3.2.
Without the loss of generality, and for simplicity reasons, only the real part, Re{Z 21} , is
Without the loss of generality, and for simplicity reasons, only the real part, Re{ Z21 }, is considered
considered (the imaginary part exhibits similar behavior). The working point is around the zero-
(the imaginary part exhibits similar behavior). The working point is around the zero- field. In this case,
field. In this case, the offset is defined as the value of the real part at the zero-field, Re{Z }(0) , which
the offset is defined as the value of the real part at the zero-field, Re{ Z21 }(0), which21is almost zero.
Theisintrinsic
almost zero. The intrinsic
sensitivity, denoted
sensitivity,
as SΩ−21denoted as S
, is defined by −21, is defined by the derivative of the curve at
the derivative of the curve at this same point
∂Re{ Z21 }   Re  Z 
(SΩthis
−21 (same
0) = point ( S (0) ).
∂H 21
21
).
H =0 H H 0
In a similar way as for the diagonal component, the voltage sensitivity, denoted as Sv−21 , is defined
In a similar way as for the diagonal component, the voltage sensitivity, denoted as Sv-21, is defined
by Equation (2)
by Equation (2) 
∂Re{ Z21 } 
Sv−21 (0) = SΩ−21 (0) . Iac = Iac (2)
 ReZ∂H
21
Sv  21 (0)  S 21 (0) . I ac 

I acH =0 (2)
H H 0
For the intended application, which is the development of a GMI current clamp, the GMI sensor
should beForimplemented
the intended application, whichThe
in closed-loop. is the development
intrinsic voltageof sensitivity,
a GMI current clamp, the
combined GMI
with sensor
the gain of
should be implemented
amplification in closed-loop.
of the conditioning The determines
electronics, intrinsic voltage sensitivity,gain.
the open-loop combined with the gain
This open-loop gainofhas
amplification
to be “high enough”of theto
conditioning
guarantee aelectronics, determines
well-regulated the open-loop
closed-loop. In thisgain.
way,This
the open-loop gain has
sensor output is less
to be “high
dependent on enough” to guarantee
the imperfections of athe
well-regulated
open-loop. Someclosed-loop. In thisof
advantages way,
the the sensor output isinclude
closed-operation less
dependent
improved on thetemperature
linearity, imperfections of the open-loop.
dependence, Some advantages
and hysteresis of theresponse
of the sensor closed-operation
as well asinclude
a higher
improved linearity, temperature dependence, and hysteresis of the sensor response
dynamic range [35]. The change of the intrinsic sensitivity under quantities of influence like torsion as well as a
higher dynamic range [35]. The change of the intrinsic sensitivity under quantities
stress must be known in order to compensate for it, if necessary. The general goal is to maintain of influence like a
torsion stress must be known in order to compensate for it, if necessary. The general goal is to
“high enough” open-loop gain.
maintain a “high enough” open-loop gain.
That is why the investigation of the change of the intrinsic sensitivity is investigated under torsion
That is why the investigation of the change of the intrinsic sensitivity is investigated under
for both diagonal and off-diagonal components. In addition, the change of the offset, which is another
torsion for both diagonal and off-diagonal components. In addition, the change of the offset, which
practical issue, is also considered. Finally, for comparison purposes with other published works only,
is another practical issue, is also considered. Finally, for comparison purposes with other published
theworks
evolution
only,of the
the GMI ratio
evolution forGMI
of the the diagonal component,
ratio for the Z11 , under torsion
diagonal component, 𝑍 , underis recalled.
torsion is recalled.
A schematic of the experimental setup is shown in Figure
A schematic of the experimental setup is shown in Figure 3. 3.

Figure
Figure 3. Schematic
3. Schematic ofof theexperimental
the experimentalsetup
setup for
for the
the investigation
investigationofofthe
thetorsion stress
torsion effect
stress onon
effect GMIGMI
in amorphous wires.
in amorphous wires.

TheThe samples
samples studiedwere
studied wereCo-rich
Co-richamorphous
amorphous wires
wires (Co-Fe-Si-B)
(Co-Fe-Si-B)with
witha a100
100µmµmdiameter
diameterfrom
from
Unitika.
Unitika. These
These wiresexhibit
wires exhibitnearly
nearlyzero
zeromagnetostriction.
magnetostriction. AApick-up
pick-upcoil
coilwas
waswound
wound around
aroundeacheach
wire.
wire. TheThe
GMI GMI element
element wassoldered
was solderedtotoaarotation
rotation device.
device. The
Thetwisting
twistingofofthe
thewire
wirewas performed
was performed by by
rotating
rotating thisthis device,
device, whichensures
which ensuresthe
theapplication
application of
of aa torsion
torsionwith
withaadefined
defined rotation angle.
rotation Indeed,
angle. Indeed,
this rotation device was graduated to measure the rotation angle starting from a reference position
corresponding to the zero angle.
Sensors 2018, 18, 4121 5 of 11
Sensors 2018, 18, x FOR PEER REVIEW 5 of 11

this
Therotation
wire device was graduated
was supplied by thetohigh-frequency
measure the rotation angle
current, iacstarting
, using from a reference
a signal position
generator and a
corresponding to the zero angle.
voltage-to-current convertor made by the resistor Rg. In the same way, a DC current could be supplied
The wire was supplied by the high-frequency current, iac, using a signal generator and a voltage-
to the wire using a DC source and the resistor Rb . The voltages, vac , across the wire, and vcoil , across the
to-current convertor made by the resistor Rg. In the same way, a DC current could be supplied to the
pick-up coil, were demodulated using a lock-in amplifier. The output voltages of this lock-in amplifier
wire using a DC source and the resistor Rb. The voltages, vac, across the wire, and vcoil, across the pick-
are proportional to the diagonal and off-diagonal components. A low-frequency (0.2 Hz) sweeping
up coil, were demodulated using a lock-in amplifier. The output voltages of this lock-in amplifier are
of the magnetictofield
proportional was applied
the diagonal to the sensitive
and off-diagonal element A
components. using the same pick-up
low-frequency coil. An electrical
(0.2 Hz) sweeping of the
“separation” between this low-frequency section and the high-frequency one was
magnetic field was applied to the sensitive element using the same pick-up coil. An made using a choke
electrical
inductor Lchoke . between this low-frequency section and the high-frequency one was made using a choke
“separation”
inductor Lchoke.
3. Results and Discussion
3. Results and Discussion
3.1. The GMI Ratio of the Diagonal Component
3.1.
ForThe GMI
this Ratio of the the
experiment, Diagonal
GMI Component
wire used was 6.5 cm long with a pick-up coil of about 1000 turns.
For this experiment,
The high-frequency current,the
iacGMI
, hadwire used wasof6.5
a frequency cm long
1 MHz andwith
an aamplitude
pick-up coil
of of aboutNo
3 mA. 1000
DCturns.
current
wasThe high-frequency
supplied current,
to the wire. iac, had
Figure a frequency
4 shows of 1 MHzofand
the modulus thean amplitude
diagonal of 3 mA. No
component forDC currentand
positive
was supplied
negative rotationtoangles
the wire. Figure
of the 4 shows
sensitive the modulus
element. of the diagonal
The zero-degree anglecomponent
correspondedfor positive and
to the position
negative
at which therotation
wire wasangles of thesoldered.
initially sensitive element. The zero-degree angle corresponded to the position
at It
which theseen
can be wire was
that initially
there issoldered.
obviously a net change of the impedance curve with the torsion.
It cancan
This change be seen that there using
be quantified is obviously
the GMI a net change
ratio of the
defined byimpedance
Equation curve
(3) with the torsion. This
change can be quantified using the GMI ratio defined by Equation (3)
|| Z |( H ) − | Z |( Hmax )|
∆| Z11 |/| Z11 |(%) = |𝑍 11|(𝐻) |𝑍 11 × 100 (3)
∆|𝑍 |⁄|𝑍 | (%) = | Z −|( H |(𝐻) )
11 max × 100
|𝑍 |(𝐻 ) (3)

where
where Hmax is the
𝐻 is the maximum
maximum availablefield.
available field.

Figure
Figure 4. 4. Modulusofofthe
Modulus thediagonal
diagonalcomponent
component for
for different
differentrotation
rotationangles
anglesofofthe sensitive
the element.
sensitive element.

ForFor
eacheach rotation
rotation angle,this
angle, thisratio
ratiohas
hasaamaximum
maximum noted ∆| Z11|⁄|/|𝑍| Z11
noted ((∆|𝑍 . This
|)|)max . Thismaximum
maximum is is
plotted
plotted in Figure
in Figure 5 as 5a function
as a function
of theof the rotation
rotation angle.
angle. It It seen
can be can be
thatseen that the behavior
the behavior is not
is not symmetrical
for symmetrical
positive andfor positiverotation
negative and negative
angles.rotation angles. The
The maximum ismaximum
obtained foris obtained
a rotationforangle
a rotation
near angle
50◦ and
notnear
for a50°
zeroanddegree.
not for a zero degree.
Sensors 2018, 18, 4121 6 of 11
Sensors 2018, 18, x FOR PEER REVIEW 6 of 11

Figure
Figure 5. 5. Maximumofofthe
Maximum theGMI
GMI ratio
ratio as
as aa function
functionof
ofthe
therotation
rotationangle.
angle.

These
These results
results areare very
very closetotothe
close theones
onesreported
reported inin Ref.
Ref. [16],
[16],where
whereaasimilar
similarbehavior
behavior in in
as-cast
as-cast
Co-rich
Co-rich amorphous
amorphous wires
wires hashasbeen
beenobserved.
observed. The
The change
change inin the
theGMI
GMIresponse
responsewas wasinterpreted
interpreted based
based
on aonchange
a change in in
thethe domain
domain structure.Indeed,
structure. Indeed, inin the
the case
case of
of negative
negativemagnetostriction,
magnetostriction, thethe
domain
domain
structure
structure in in
thethe
GMIGMI wirewireconsists
consistsofofaacore,
core, which
which is is axially
axially magnetized,
magnetized,and andananouter
outershell with
shell with
circular magnetization [36–40]. The wires used in our experiments
circular magnetization [36–40]. The wires used in our experiments have a nearly zero magnetostriction have a nearly zero
magnetostriction
constant. In this case,constant. In this assumed
it is usually case, it is usually
that theassumed
magnetic that the magnetic
structure structure
is roughly the issame
roughly
as the
the same as the structure encountered in negative magnetostriction wires [16,24,36–38]. The torsion
structure encountered in negative magnetostriction wires [16,24,36–38]. The torsion modifies this
modifies this spatial distribution of the magnetization close to the surface of the wire. This results in
spatial distribution of the magnetization close to the surface of the wire. This results in a modification
a modification of the circumferential permeability and consequently of the GMI response with
of the circumferential permeability and consequently of the GMI response with torsion. The asymmetry
torsion. The asymmetry in the GMI ratio in Figure 5 may indicate that there were already preexisting
in the GMI ratio in Figure 5 may indicate that there were already preexisting torsional stresses in
torsional stresses in the wire, which may be due to the fabrication process. Some of these preexisting
thetorsional
wire, which mayhave
stresses be due
beento the fabrication
partially compensatedprocess.
for a Some
torsionofcorresponding
these preexisting torsional
to a rotation stresses
angle of
have ◦
50°been partially
[16,23]. In thiscompensated
context, it isforto abetorsion
noted corresponding
that the torsiontoapplied
a rotation angle
could of 50
in fact not[16,23]. In this
completely
compensate
context, it is to for
be these
notedinternal
that thestresses
torsion due to the complexity
applied could in fact of their distributioncompensate
not completely [23]. for these
internal stresses due to the complexity of their distribution [23].
3.2. Offset and Sensitivity of the Diagonal and Off-Diagonal Components
3.2. Offset and Sensitivity of the Diagonal and Off-Diagonal Components
For this experiment, the GMI wire used was 9 cm long with a pick-up coil of about 1400 turns.
AFor
high-frequency current,
this experiment, theiacGMI
, of 1 wire
MHz used
/5 mAwaswas9used. A DC
cm long current,
with Idc = 5 mA,
a pick-up was
coil of also supplied
about 1400 turns.
to the wire. This
A high-frequency currentiacis, required
current, of 1 MHz/5 for the
mAsignificant
was used. appearance
A DC current, of the
Idc off-diagonal
= 5 mA, wascomponent
also supplied in to
the wires
wire. with
This circumferential
current is required anisotropy
for the[31–34]. In addition,
significant appearance this current may make it component
of the off-diagonal possible to obtain
in wires
withancircumferential
anhysteretic field dependence
anisotropy [31–34].ofIn the impedance
addition, whenmay
this current its make
value it exceeds
possible to a obtain
certain an
threshold field
anhysteretic [26,28–30].
dependence of the impedance when its value exceeds a certain threshold [26,28–30].
It is worth
It is worth notingnoting thatthe
that theuse
useofofaarelatively
relatively low
low frequency
frequencyof of11MHzMHzisisjustified
justified bybypractical
practical
considerations related to the parasitic resonance of the pick-up coil. This
considerations related to the parasitic resonance of the pick-up coil. This resonance is due resonance is due to the to
LCthe
circuit formed by the inductance of the coil and the stray capacitance between the turns of the
LC circuit formed by the inductance of the coil and the stray capacitance between the turns of the
winding. Indeed, for the off-diagonal configuration, the voltage sensitivity increases up to a
winding. Indeed, for the off-diagonal configuration, the voltage sensitivity increases up to a maximum,
maximum, which is obtained for an optimum frequency. This sensitivity then decreases when the
which is obtained for an optimum frequency. This sensitivity then decreases when the frequency
frequency increases [29,41]. The optimum frequency is lower than the frequency of parasitic
increases [29,41]. The optimum frequency is lower than the frequency of parasitic resonance of the
resonance of the pick-up coil. For the GMI sensitive element and pick-up coil used, this optimum was
pick-up coil. For the
experimentally GMI
found tosensitive
be about element and pick-upthe
1 MHz. Furthermore, coilamplitude
used, thisofoptimum was experimentally
the high-frequency current
found
was chosen to be relatively low to avoid the nonlinear regime of the GMI and to thereby guaranteetoa be
to be about 1 MHz. Furthermore, the amplitude of the high-frequency current was chosen
relatively low to avoidbetween
linear relationship the nonlinear regime
the voltage andof the
the GMI
current andintothe
thereby guarantee
wire [42]. a linear
Finally, relationship
for a given AC
current, the value (5 mA) of the DC current was used as a trade-off between the sensitivity and(5the
between the voltage and the current in the wire [42]. Finally, for a given AC current, the value mA)
reduction
of the of the was
DC current level used
of low-frequency
as a trade-off intrinsic
betweenmagnetic noise of the
the sensitivity andsensitive element [43].
the reduction of the level of
low-frequency intrinsic magnetic noise of the sensitive element [43].
Sensors 2018, 18, 4121 7 of 11
Sensors 2018, 18, x FOR PEER REVIEW 7 of 11
Sensors 2018, 18, x FOR PEER REVIEW 7 of 11

In summary, in practice, the set of values for the excitation (frequency, amplitude of the AC
In In summary,
summary, ininpractice,
practice,the
theset
setof
of values
values for
for the
the excitation
excitation(frequency,
(frequency,amplitude
amplitude of of
thethe
ACAC
current, and value of the DC current) was chosen to optimize the sensitivity and noise (under zero
current,
current, and and value
value ofofthe
theDC
DCcurrent)
current)was
was chosen
chosen toto optimize
optimizethethesensitivity
sensitivityand noise
and (under
noise zero
(under zero
torsion stress) while avoiding both nonlinear effects of the GMI and the paratactic resonance of the
torsion
torsion stress) while avoiding both nonlinear effects of the GMI and the paratactic
stress) while avoiding both nonlinear effects of the GMI and the paratactic resonance of theresonance of the
pick-up coil. The impact of the torsion stress, as an influence parameter, was studied under these real
pick-up
pick-up coil.
coil. TheThe impactofofthe
impact thetorsion
torsionstress,
stress, as
as an
an influence
influence parameter,
parameter,waswasstudied
studiedunder
underthese realreal
these
conditions of implementation of the GMI current clamp.
conditions
conditions of implementation
of implementation of the GMI
ofproportional current
the GMI current clamp.
Figure 6 shows voltages to theclamp.
modulus of the diagonal component and to the real
Figure 6 shows voltages proportional to the modulus of the diagonal component and to the real
Figure 6 shows voltages proportional to the modulus
part, Re{Z } , of the off-diagonal component for a few rotation of the angles.
diagonal component and to the real
part, Re{Z 21 } , of the off-diagonal component for a few rotation angles.
part, Re{ Z21 }21
, of the off-diagonal component for a few rotation angles.

Figure 6. Voltages proportional to the diagonal andoff-diagonal

off-diagonal components
components as aafunction of the
Figure
Figure 6. 6. Voltagesproportional
Voltages proportionaltotothe
the diagonal
diagonal and off-diagonal components asas
a function of of
function thethe
magnetic field for different rotation angles.
magnetic
magnetic field
field forfor different
different rotationangles.
rotation angles.

A general change of both components withtorsion

torsionisisobserved.
observed. The
The torsion seems to induce an
A general
A general change
change ofofboth
bothcomponents
components withwith torsion is observed. The torsion seems
torsion seemsto to
induce an an
induce
asymmetry of the curve for the diagonal component, whereas the odd symmetry of the off-diagonal
asymmetry
asymmetry of of
thethe curvefor
curve forthe
thediagonal
diagonalcomponent,
component, whereas
whereasthe theodd
oddsymmetry
symmetry ofof
thethe
off-diagonal
off-diagonal
component seems to be well-preserved.
component seems to be well-preserved.
component seems to be well-preserved.
Voltages proportional to the offsets of these components are plotted in Figure 7.
Voltages
Voltages proportional
proportional totothe
theoffsets
offsetsofofthese
thesecomponents
components are are plotted
plottedininFigure
Figure7.7.

Figure
Figure 7. Change
7. Change of of
thethe offset
offset of of
thethediagonal
diagonaland
andoff-diagonal
off-diagonal components
components as
as aa function
functionofofthe
the
Figure 7. Change of the offset of the diagonal and off-diagonal components as a function of the
rotation
rotation angle.
angle.
rotation angle.
Sensors 2018, 18, 4121 8 of 11

Sensors 2018, 18, x FOR PEER REVIEW 8 of 11

For the diagonal component, only a small change of this offset, at a given bias, was observed
for the range
For theof rotation
diagonal angles considered.
component, only a smallInchange
the off-diagonal
of this offset,case, the offset
at a given nearobserved
bias, was the zero-field
for
theexhibited
point range of rotation angleschange.
a significant considered. In the
At this firstoff-diagonal case, the offset
stage of investigation, and near
in athe zero-field
first point
approximation,
the exhibited
change ofathe significant
offset of change. At this first
this component stage
at the of investigation,
zero-field and in ainterpreted
could be roughly first approximation,
by knowing thethe
change of the offset of this component at the zero-field could be roughly interpreted by knowing the
origin of the voltage induced in the pick-up coil. Indeed, in the absence of torsion stress, this induced
originresults
voltage of the voltage
from the induced in the pick-up
appearance of an AC coil. Indeed, in the
longitudinal absence of torsion
magnetization in thestress,
GMI this
wire,induced
which is
voltage results from
circumferentially the appearance
magnetized of an AC longitudinal
by the excitation current, iac .magnetization
To observe this in longitudinal
the GMI wire,component
which is
circumferentially
of the magnetization, magnetized
the staticby the excitation must
magnetization current, iac. Toaobserve
follow helical this
pathlongitudinal
around thecomponent
direction of
of the magnetization, the static magnetization must follow a helical
the excitation current [31–34]. In the case of a GMI wire with an almost circumferential path around the direction of the
anisotropy,
as isexcitation
the case current
with the [31–34].
wire usedIn the incase
thisof a GMI
study, wirestatic
a DC with current,
an almostI circumferential anisotropy, as is
dc , superimposed to the excitation
the case with the wire used in this study, a DC static current, Idc, superimposed to the excitation
current, iac , is required. The static circumferential field produced by Idc , combined with the longitudinal
current, iac, is required. The static circumferential field produced by Idc, combined with the
external measured field, allows the magnetization to have a helical path. Hence, in the absence of
longitudinal external measured field, allows the magnetization to have a helical path. Hence, in the
an external measured field, there is no longitudinal magnetization and the coil voltage is almost
absence of an external measured field, there is no longitudinal magnetization and the coil voltage is
null. In other words, the offset is roughly null at the zero-field. When the wire is twisted, a helical
almost null. In other words, the offset is roughly null at the zero-field. When the wire is twisted, a
magnetic anisotropy could be induced so that an AC longitudinal magnetization component appears.
helical magnetic anisotropy could be induced so that an AC longitudinal magnetization component
Thisappears.
magnetization componentcomponent
This magnetization gives rise togivesa coil voltage,
rise to a coileven when
voltage, thewhen
even externalthe magnetic field is not
external magnetic
applied, that is to say, the offset at the zero-field is not null.
field is not applied, that is to say, the offset at the zero-field is not null.
ForFor
thethe
sensitivity of of
sensitivity these
these components,
components,the thetendency
tendencyofofchange
changewas wasthetheopposite,
opposite, as as illustrated
illustrated in
in Figure
Figure 8. This8. This sensitivity
sensitivity waswas almost
almost unchangedfor
unchanged forthe
theoff-diagonal
off-diagonal component
componentnear near the zero-field
the zero-field
◦ ). In the
point
point (a maximum
(a maximum change
change fromfrom
−19.5–19.5 mV/(A/m) at
mV/(A/m) at zero
zero angle
angle to about − –18
18mV/(A/m)
mV/(A/m) at 90 °). In
at 90 the
casecase of the
of the diagonal
diagonal component,a asignificant
component, significantchange
change in in the
the sensitivity
sensitivityatatthethebias
biasfield was
field wasobserved.
observed.

Figure
Figure 8. Change
8. Change in in
thethe sensitivityofofthe
sensitivity thediagonal
diagonal and
and off-diagonal
off-diagonalcomponents
componentsasas
a function of of
a function thethe
rotation
rotation angle.
angle.

ForFor
thisthis diagonal
diagonal component,aarough
component, roughestimation
estimation ofof the
the intrinsic
intrinsicsensitivity
sensitivitycould
could generally
generally be be
givengiven
by by
thethe ratio
ratio between
between thethefield
fieldofofthe
themaximum
maximum of of the
the impedance
impedance(which(whichisisclosely
closelyrelated
relatedto the
to the
fieldfield of anisotropy)
of anisotropy) andand the
the valueofofthe
value theimpedance
impedance at at this
this field
field [27].
[27].For
Forthe
thedata
dataininFigure
Figure 6, this ratio
6, this ratio
increased, under torsion stress, in some linear behavior, from 0.42 /(A/m)
increased, under torsion stress, in some linear behavior, from 0.42 Ω/(A/m) at −90 to 0.6 Ω/(A/m) at –90° ◦to 0.6 /(A/m) at at
◦ 90°. At least, this general trend is consistent with the almost linear increase in the sensitivity
90 . At least, this general trend is consistent with the almost linear increase in the sensitivity measured measured
in Figure
in Figure 8 in8 the
in the range
range ofof rotationangles
rotation anglesconsidered.
considered.
The experimental results showed a better offset stability for the diagonal component under
The experimental results showed a better offset stability for the diagonal component under torsion.
torsion. The sensitivity change was less important for the off-diagonal component near the
The sensitivity change was less important for the off-diagonal component near the zero-field point.
zero-field point.
The torsion, as a parameter of influence of the GMI current sensor, directly influences the choice
The torsion, as a parameter of influence of the GMI current sensor, directly influences the choice
of one or the other of the two components for sensor implementation. Usually, the offset change is
of one or the other of the two components for sensor implementation. Usually, the offset change is an
an issue, especiallywhen
issue, especially whenthe
thesensor
sensorinin intended
intended to to measure
measure DCDC magnetic
magnetic fields.
fields. In case,
In this this case,
the usetheofuse
of the
the diagonal component may be preferred. However, in this case, attention should be paid to the gain to
diagonal component may be preferred. However, in this case, attention should be paid
theof the of
gain open-loop, which changes
the open-loop, with the torsion
which changes stress,
with the through
torsion stress,thethrough
change in thechange
the intrinsicinsensitivity.
the intrinsic
Sensors 2018, 18, 4121 9 of 11

sensitivity. The off-diagonal component still exhibits the advantage of odd symmetry, which was
preserved under torsion stress for the set of excitation parameters used. This preserved odd symmetry
near the zero-field point is important since it still allows for the implementation of the off-diagonal
sensor without making use of an axial bias field to fix the working point. This greatly simplifies the
design of the sensor.
Finally, it should be noted that the results presented were obtained for one set of excitation
parameters that allowed maximum sensitivity under zero torsion stress. In practice, this is a pragmatic
approach for considering the problem. However, both diagonal and off-diagonal components depend
on the high-frequency excitation current and on the DC current.

Author Contributions: Conceptualization, J.N., A.A., and J.-P.Y.; Formal analysis, J.N., A.A., and J.-P.Y.;
Investigation, J.N., A.A., and J.-P.Y.; Writing–original draft, J.N., A.A., and J.-P.Y.
Funding: This research received no external funding.
Conflicts of Interest: The authors declare no conflict of interest.

References
1. Panina, L.V.; Mohri, K. Magneto-impedance effect in amorphous wires. Appl. Phys. Lett. 1994, 65, 1189–1191.
[CrossRef]
2. MI Sensor Electro-Magnetic Product Aichi Steel. Available online: https://www.aichisteel.co.jp/ENGLISH/
products/electromagnetic/mi_censor/index.html (accessed on 15 July 2018).
3. Coillot, C.; Moutoussamy, J.; Chanteur, G.; Robert, P.; Alves, F. On-board hybrid magnetometer of NASA
CHARM-II rocket: Principle, design and performances. J. Sens. Sens. Syst. 2013, 2, 137–145. [CrossRef]
4. Mohri, K.; Uchiyama, T.; Shen, L.P.; Cai, C.M.; Panina, L.V. Amorphous wire and CMOS IC-based sensitive
micro-magnetic sensors (MI sensor and SI sensor) for intelligent measurements and controls. J. Magn.
Magn. Mater. 2002, 249, 351–356. [CrossRef]
5. Honkura, Y. Development of amorphous wire type MI sensors for automobile use. J. Magn. Magn. Mater.
2002, 249, 375–381. [CrossRef]
6. Gudoshnikov, S.; Usov, N.; Nozdrin, A.; Ipatov, M.; Zhukov, A.; Zhukova, V. Highly sensitive magnetometer
based on the off-diagonal GMI effect in Co-rich glass-coated microwire. Phys. Stat. Sol. 2014, 211, 980–985.
[CrossRef]
7. Uchiyama, T.; Mohri, K.; Nakayama, S. Measurement of Spontaneous Oscillatory Magnetic Field of
Guinea-Pig Smooth Muscle Preparation Using Pico-Tesla Resolution Amorphous Wire Magneto-Impedance
Sensor. IEEE Trans. Magn. 2011, 47, 3070–3073. [CrossRef]
8. García, C.; Zhukov, A.; Blanco, J.M.; Zhukova, V.; Ipatov, M.; Gonzalez, J. Effect of tensile stresses on
GMI of Co-rich amorphous microwires. In Proceedings of the INTERMAG Asia 2005, Digests of the IEEE
International Magnetics Conference, Nagoya, Japan, 4–8 April 2005; pp. 1273–1274.
9. Ciureanu, P.; Khalil, I.; Melo, L.G.C.; Rudkowski, P.; Yelon, A. Stress-induced asymmetric
magneto-impedance in melt-extracted Co-rich amorphous wires. J. Magn. Magn. Mater. 2002, 249, 305–309.
[CrossRef]
10. Knobel, M.; Sanchez, M.L.; Velazquez, J.; Vázquez, M. Stress dependence of the giant magneto-impedance
effect in amorphous wires. J. Phys. Condens. Matter 1995, 7, L115. [CrossRef]
11. Knobel, M.; Pirota, K.R. Giant magnetoimpedance: Concepts and recent progress. J. Magn. Magn. Mater.
2002, 242, 33–40. [CrossRef]
12. Atkinson, D.; Squire, P.T. Experimental and phenomenological investigation of the effect of stress on
magneto-impedance in amorphous alloys. IEEE Trans. Magn. 1997, 33, 3364–3366. [CrossRef]
13. Knobel, M.; Vazquez, M.; Hernando, A.; Sánchez, M.L. Effect of tensile stress on the field response of
impedance in low magnetostriction amorphous wires. J. Magn. Magn. Mater. 1997, 169, 89–97.
14. Mandal, K.; Puerta, S.; Vazquez, M.; Hernando, A. The frequency and stress dependence of giant
magnetoimpedance in amorphous microwires. IEEE Trans. Magn. 2000, 36, 3257–3259. [CrossRef]
15. Kim, C.G.; Yoon, S.S.; Vazquez, M. Evaluation of helical magnetoelastic anisotropy in Fe-based amorphous
wire from the decomposed susceptibility spectra. J. Magn. Magn. Mater. 2001, 223, 199–202. [CrossRef]
Sensors 2018, 18, 4121 10 of 11

16. Blanco, J.M.; Zhukov, A.; Gonzalez, J. Effect of tensile and torsion on GMI in amorphous wire. J. Magn.
Magn. Mater. 1999, 196, 377–379. [CrossRef]
17. Nabias, J.; Asfour, A.; Yonnet, J.P. Investigation of Bending Stress Effect on the Diagonal and Off-Diagonal
Impedances for GMI Sensor Implementation. IEEE Trans. Magn. 2017, 53, 1–8. [CrossRef]
18. Nabias, J.; Asfour, A.; Yonnet, J.P. The Impact of the Bending Stress on the Performance of Giant
Magneto-Impedance (GMI) Magnetic Sensors. Sensors 2017, 17, 640. [CrossRef] [PubMed]
19. Beato-López, J.J.; Vargas-Silva, G.; Pérez-Landazábal, J.I.; Gómez-Polo, C. Giant Stress-Impedance (GSI)
sensor for diameter evaluation in cylindrical elements. Sens. Actuators A Phys. 2018, 269, 269–275. [CrossRef]
20. Chizhik, A.; Gonzalez, J.; Zhukov, A.; Stupakiewcz, A. Torsion stress induced magnetic switching in
amorphous microwires. IEEE Magn. Lett. 2017, 8, 1–5. [CrossRef]
21. Blanco, J.M.; Zhukov, A.; Gonzalez, J. Torsional Stress Impedance and Magneto-impedance in
(Co0.95 Fe0.05 )72.5 Si12.5 B15 Amorphous Wire with Helical Induced Anisotropy. J. Phys. D Appl. Phys. 1999, 32,
3140–3145. [CrossRef]
22. Tejedor, M.; Hernando, B.; Sanchez, M.L.; Prida, V.M.; Vazquez, M. The torsional dependence of the
magneto-impedance effect in current-annealed Co-rich amorphous wires. J. Phys. D Appl. Phys. 1998, 31,
3331–3336. [CrossRef]
23. Blanco, J.M.; Zhukov, A.; Gonzalez, J. Asymmetric torsion stress giant magnetoimpedance in nearly-zero
magnetostrictive amorphous wires. J. Appl. Phys. 2000, 87, 4813–4815. [CrossRef]
24. Blanco, J.M.; Zhukov, A.; Chen, A.P.; Cobeño, A.F.; Chizhik, A.; Gonzalez, J. Asymmetric torsion giant
impedance in nearly-zero magnetostrictive amorphous wires with induced helical anisotropy. J. Phys. D
Appl. Phys. 2001, 34, L31–L34. [CrossRef]
25. García, K.L.; García-Beneytez, J.M.; Valenzuela, R.; Zhukov, A.; González, J.; Vázquez, M. Effects of Torsion
on the Magnetoimpedance Response of CoFeBSi Amorphous Wires. J. Magn. Magn. Mater. 2001, 226,
731–733.
26. Antonov, A.S.; Buznikov, N.A.; Granovsky, A.B. Asymmetric Giant Magnetoimpedance of Amorphous
Microwires under the Action of Torsional Stresses. Tech. Phys. Lett. 2014, 40, 267–270. [CrossRef]
27. Ménard, D.; Seddaoui, D.; Melo, L.G.C.; Yelon, A.; Dufay, B.; Saez, S.; Dolabdjian, C. Perspectives in giant
magnetoimpedance magnetometry. Sens. Lett. 2009, 73, 339–342. [CrossRef]
28. Ipatov, M.; Zhukova, V.; Zhukov, A.; Gonzalez, J. Magnetoimpedance sensitive to dc bias current in
amorphous microwires. Appl. Phys. Lett. 2010, 97, 252507. [CrossRef]
29. Ipatov, M.; Chizhik, A.; Zhukova, V.; Gonzalez, J.; Zhukov, A. Correlation of surface domain structure and
magneto-impedance in amorphous microwires. J. Appl. Phys 2011, 109, 113924. [CrossRef]
30. Buznikov, N.A.; Antonov, A.S.; Granovsky, A.B. Asymmetric magnetoimpedance in amorphous microwires
due to bias current: Effect of torsional stress. J. Magn. Magn. Mater. 2014, 355, 289–294. [CrossRef]
31. Sandacci, D.; Makhnovskiy, D.; Panina, L.; Mohri, K.; Honkura, Y. Off-diagonal impedance in amorphous
wires and its application to linear magnetic sensors. IEEE Trans. Magn. 2004, 40, 3505–3511. [CrossRef]
32. Makhnovskiy, D.P.; Panina, L.V.; Mapps, D.J. Measurement of field-dependent surface impedance tensor in
amorphous wires with circumferential anisotropy. J. Appl. Phys. 2000, 87, 4804–4805. [CrossRef]
33. Antonov, A.S.; Iakubov, I.T.; Lagarkov, A.N. Nondiagonal impedance of amorphous wires with circular
magnetic anisotropy. J. Magn. Magn. Mater. 1998, 187, 252–260. [CrossRef]
34. Zhukov, M.; Ipatov, M.; Zhukova, V.; García, C.; Gonzalez, J.; Blanco, J.M. Development of ultra-thin
glass-coated amorphous microwires for HF magnetic sensor applications. Phys. Status Solidi 2008, 205,
1367–1372. [CrossRef]
35. Asfour, A.; Yonnet, J.P.; Zidi, M. A high dynamic range GMI current sensor. J. Sens. Technol. 2012, 2, 165.
[CrossRef]
36. Kabanov, Y.; Zhukov, A.; Zhukova, V.; Gonzalez, J. Magnetic domain structure of wires studied by using the
magneto-optical indicator film method. Appl. Phys. Lett. 2005, 87, 142507. [CrossRef]
37. Mohri, K.; Humphrey, F.R.; Kawashimay, K.; Kimura, K.; Mizutani, M. Large Barkhausen and Matteucci
effects in FeCoSiB, FeCrSiB, and FeNiSiB. IEEE Trans. Magn. 1990, 26, 1789. [CrossRef]
38. Zhukova, V.; Blanco, J.M.; Ipatov, M.; Zhukov, A. Effect of transverse magnetic field on domain wall
propagation in magnetically bistable glass-coated amorphous microwires. J. Appl. Phys 2009, 106, 113914.
[CrossRef]
Sensors 2018, 18, 4121 11 of 11

39. Zhukova, V.; Usov, N.A.; Zhukov, A.; Gonzalez, J. Length effect in Co-rich amorphous wire. Phys. Rev. B
2002, 65, 134407. [CrossRef]
40. Phan, M.; Peng, H. Giant magnetoimpedance materials: Fundamentals and applications. Prog. Mater. Sci.
2008, 53, 323–420. [CrossRef]
41. Dufay, B.; Saez, S.; Dolabdjian, C.P.; Yelon, A.; Menard, D. Characterization of an optimized off-diagonal
GMI-based magnetometer. IEEE Sens. J. 2013, 13, 379–388. [CrossRef]
42. Seddaoui, D.; Ménard, D.; Movaghar, B.; Yelon, A. Nonlinear electromagnetic response of ferromagnetic
metals: Magnetoimpedance in microwires. J. Appl. Phys. 2009, 105, 083916. [CrossRef]
43. Paperno, E. Suppression of magnetic noise in the fundamental-mode orthogonal fluxgate. Sens. Actuators A
2004, 116, 405–409. [CrossRef]

© 2018 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC BY) license (http://creativecommons.org/licenses/by/4.0/).