high-spatial-resolution magnetic-field measurement …s2is.org/issues/v1/n1/papers/paper10.pdf ·...
TRANSCRIPT
High-Spatial-Resolution Magnetic-Field Measurement by Giant Magnetoresistance Sensor – Applications to
Nondestructive Evaluation and Biomedical Engineering Sotoshi YAMADA
Institute of Nature and Environmental Engineering, Kanazawa University Kakuma-machi, Kanazawa 920-1192, Japan
Abstract
Giant magnetoresistance (GMR) sensor has been developed and widely applied to use as
magnetic read head in data storage industry. This paper describes new applications of
magnetic-field measurement with high spatial-resolution and high sensitivity to
nondestructive evaluation and biomedical engineering. For nondestructive evaluation, the
GMR sensor, used as magnetic sensor based on eddy-current testing technique, was applied to
the detection of micro-crack on micro-conductor for the purpose of printed circuit board
inspection and the detection of micro-solder-ball grid array. For biomedical engineering, the
weight density of magnetic fluid for cancer treatment was measured by the GMR sensor. In
addition, the GMR sensor was applied to measure micro-current and these can lead to the
direct detection of nervous action.
Keywords: giant magnetoresistance, nondestructive evaluation, eddy-current testing, printed
circuit board, conductive microbead, low-invasion, hyperthermia, magnetic fluid, weight
density, nerve, action current
1 Introduction
Magnetic sensing techniques are unique when compared to other sensors because they do not
directly measure the physical property of interest. Magnetic sensors detect changes or
disturbances in magnetic fields that are either created or modified and then derive information
on properties of interest such as direction, presence, rotation, angle, or electrical currents [1].
The giant magnetoresistance (GMR) element as one of magnetic sensors has the sandwich
structure of magnetic and conductive materials that changes its electrical resistance depending
on the alignment of the magnetic layers. External magnetic fields make the magnetic layers of
the GMR device align “up” or “down”, broadly speaking. However, since the layers are made
up of two magnetic materials, one layer (generally referred to as the “free layer”) changes
polarity, while the other layer (generally referred to as the “pin layer”) does not change its
160
INTERNATIONAL JOURNAL ON SMART SENSING AND INTELLIGENT SYSTEMS, VOL. 1, NO. 1, MARCH 2008
polarity. Spin-valve type giant magnetoresistance(SV-GMR) sensors have a maximum
resistance change approximately up to 20 % for external magnetic fields and saturation fields
are 1.0 – 8.0 mT [1]. SV-GMR sensor has a high sensitivity and a high spatial resolution [2].
Nondestructive evaluation (NDE) is a very broad and interdisciplinary area that can be used to
detect defects and measure physical or mechanical characteristics of a material or component.
The popularity and acceptance of NDE as a measurement technique is due to the fact that the
system or material under test is not harmed or damaged during testing, and thus the integrity
of the object under test is retained [3]. There are a range of magnetic sensors developed for
NDE such as superconducting quantum interference device (SQUID), inductive coil, fluxgate,
hall element, and GMR.
In this paper we report novel applications of GMR sensor in the fields of eddy-current testing
(ECT) and medical treatment. For ECT application, the probe based GMR sensor was
fabricated. The proposed probe was used to inspect micro defects on printed circuit board
(PCB) and to detect micro conducting bead. In the application of medical engineering, a novel
GMR probe with needle structure was fabricated to estimate low concentration magnetic fluid
and to detect the magnetic field distribution from nervous action.
2 Comparison of Magnetic Sensors
In order to recognize the advantages of GMR, the comparisons between magnetic sensors are
shown in Table 1. Compared to the aforementioned sensors that are being used in NDE, the
GMR sensor has the advantages of small and very thin structure and high sensitivity under
low magnetic fields up to nT. The temperature dependence is 1/20 of hall element [4-6]. The
GMR sensor is a resistive device with two terminals which means that it can be utilized by the
way of a microstructure and be expanded to an array structure resulting in simple
measurements The SV-GMR sensor also has linear characteristics with a little hysteresis for
small signal. These favourable qualities make the SV-GMR device suitable for industrial
inspection and detection.
161
SOTOSHI YAMADA, HIGH-SPATIAL-RESOLUTION MAGNETIC-FIELD MEASUREMENT BY GIANT MAGNETORESISTANCE SENSOR – APPLICATIONS TO NONDESTRUCTIVE EVALUATION AND BIOMEDICAL ENGINEERING
Table 1: Comparisons between magnetic sensors for NDE applications
3 ECT Technique Based Application
3.1 SV-GMR Sensor and Probe [7, 8]
The applications discussed in this paper require the SV-GMR sensor to operate at low AC
magnetic fields. Hence, the small signal characteristics of the GMR sensor are discussed. The
characteristics of the SV-GMR multi-sensor are shown in Fig. 1(a). Each SV-GMR sensor
consists of four strips where each strip is 100 μm in length and 18 μm in width. The normal
resistance of the sensor is 400 Ω and the maximum MR ratio 12 %.
Figure 1(b) shows the small-signal characteristics of the SV-GMR sensor at a frequency of
100 kHz when sinusoidal magnetic field of 200 μT is applied. The sensitivity in the sensing
axis (z-axis) is obtained as approximately 150 μV/μT for a bias current of 5 mA. The
sensitivity in the x- and y-axes were lower than 15 μV/μT.
Sensor Materials Range (T)
Sensitivity (%/mT)
Resolution Frequency (MHz)
Operating temperature & variation (μT)
Hall element
InSb, GaAs 10 - 1 0.01 -50 to +150 °C 2 %/°C
Anisotropic magneto-resistance
-50 to +150 °C NiFe, FeCo 0.1 < 2 1 1 0.2 %/°C
Giant magneto-resistance
Magnetic/non-magnetic multi-layer film
-40 to +150 °C 1 < 2 1 > 100 0.09 %/°C
Magnetic/non-magnetic multi-layer film
Spin-valve GMR
-40 to +150 °C 0.01 < 200 0.01 > 100 0.09 %/°C Magnetic impedance element
-40 to +125 °C Amorphous material 0.001 < 1000 0.001 1
162
INTERNATIONAL JOURNAL ON SMART SENSING AND INTELLIGENT SYSTEMS, VOL. 1, NO. 1, MARCH 2008
65
70
75
80
0 10 20 30 40 50 60 70
Am
plitu
de (μ
V)
50 μm
70 μm100 μm
500 μm
150 μm 300 μm200 μm
Position (mm)
70 μm PCB conductor width
Disconnection length
Amplitude
Noise
0.1
1
10
100
1000
0 100 200 300 400 500
PCB width
Noise level
Am
plitu
de v
aria
tion
(μV
)
Conductor disconnection length (μm)
500 μm300 μm200 μm100 μm70 μm
0.1
1
10
100
0 25 50 75 100 125
Noise level
Partial defect size (%)
200 μm100 μm
PCB width
PCB width100 μm
Partial defect size (%)
Am
plitu
de v
aria
tion
(μV
)
Conductordisconnection
(a) ECT signal obtained from (b) ECT signal variations against (c) ECT signal variations against scanning over 70 μm PCB conductor disconnection length partial defects on PCB conductor conductor
Figure 3: Variation of ECT signal on different kinds of defect.
The high frequency ECT probe as shown in Fig. 2 was fabricated for purpose of inspecting
micro-crack on a flat surface. The probe includes meander coil as an excitation coil which is
fed with high frequency excitation current to provide magnetic fields, and magnetic sensor
which will pick up modified magnetic fields BBz due to any defect or boundary on the PCB
conductor, that is perpendicular to the scanning direction. When the SV-GMR sensor was
mounted between the long meander coils, its sensing axis was set to detect only the magnetic
fields Bz. While the long meander coil is primarily used to induce eddy currents on PCB, it
also provides the advantage of fabricating a multi-sensor system.
3.2 PCB Conductor Inspection for Detection of Micro Defect [9]
The proposed ECT probe was used for PCB inspection. A PCB model made from copper with
a thickness of 9 μm was used for experiment. The PCB conductor had disconnections ranging
from 50 to 500 μm.
Fig. 3 (a) shows how the ECT signal varies at defect points and disconnections ranging from
50 - 500 μm can be inspected. Fig. 3(b) shows how the ECT signal variation not only depends
on defect size but also on conductor width. Furthermore, it is also possible to inspect partial
r
r
r
r
r
r
r
r
Scannin
g dire
ction
r
r
r
r
r
r
r
r
Exciting magnetic field
Exciting currentMeander coil
50 μm
200 μm
y
xz
BzBz
Bz
Bz
SV-GMR sensor
Sensing axisPCB conductor
DC biascurrent
Eddy-current flow
Detecting magnetic field
Defect point
Bz Bz
Bz
BzBz
Bz Top view ofPCB conductor
-30
-20
-10
0
10
20
30
Vol
tage
var
iatio
n ac
ross
SV
-GM
R (m
V)
Bz
By
Bx
-220 -110 0 110 220Exciting magnetic fields (μT)
Multi sensor type
Sensingdirection
Bx
Bz
By
200 μm
100 μm
Small-signal haracteristics
Figure1: Proposed SV-GMR sensor and characteristics. Figure 2: SV-GMR based ECT probe.
163
SOTOSHI YAMADA, HIGH-SPATIAL-RESOLUTION MAGNETIC-FIELD MEASUREMENT BY GIANT MAGNETORESISTANCE SENSOR – APPLICATIONS TO NONDESTRUCTIVE EVALUATION AND BIOMEDICAL ENGINEERING
defects on the PCB track width as shown in Fig. 3(c). Amplitude variation then depends on
PCB conductor width, defect length and defect type. The signal-to-noise ratio seems to
increase gradually when the PCB conductor becomes wider.
Simple image processing technique was applied to obtain the ECT images of PCB with
defects as shown in Fig. 4. Disconnection as low as 20 μm width and also partial defects were
located on the PCB model. The scanning results show that the proposed ECT probe can
successfully inspect the defects on the PCB conductor. Even though the photo images are not
very clear, the defects on the PCB conductor can be identified very easily.
PCB, 100 μm width
Gap, 200 μm widthGap, 100 μm width
PCB, 100 μm width
Figure 4: PCB model (bottom:photo image) and scanning result (top: ECT image).
3.3 Conductive Microbead Detection
164
INTERNATIONAL JOURNAL ON SMART SENSING AND INTELLIGENT SYSTEMS, VOL. 1, NO. 1, MARCH 2008
Fig. 5 shows the results obtained for detection of conductive microbead. The ECT signal
waveforms obtained are for conductive microbead of 125 μm radius at frequency 5 MHz. The
results are in good agreement with analytical results. The peak of the ECT signal and signal
gradient is used to determine the microbead diameter and its position. Fig. 6 shows the
maximum signal variation of the ECT signal for a conductive microbead radius range from
125 to 300 μm. Fig. 7(a) shows the results for determining conductive microbead position. To
determine the position of the conductive microbead, the numerical gradient technique is
applied to the ECT signal. The pitches of the conductive microbead are also measured by
considering the peak of the signal gradient. Figs. 7(b) and (c) show the detection results for
microbead array model consisting of microbead of 125 μm radius. The results show that the
conductive microbeads can be clearly recognized and the conductive microbead pitch can also
be accurately estimated with an error of less than 50 μm. This technique enables the detection
of smaller microbead when the SV-GMR sensor is placed as close as possible to the test
-90
-60
-30
0
30
0 0.5 1 1.5 2-5
0
5
10
Distance (mm)
ECT
sign
al (μ
V)
Gra
dien
t (μV
/pitc
h)
ECT signal
Gradient
Bead size
Bead center
Figure 5: ECT signal obtained from the detection of a conductive microbead (PbSn) with 125 μm radius and its signal gradient.
0
100
200
300
0 100 200 300Conductive microbead radius (μm)
Sign
al v
aria
tion
(μV
)
Figure 6: Signal variation vs. microbead(PbSn) radius.
165
SOTOSHI YAMADA, HIGH-SPATIAL-RESOLUTION MAGNETIC-FIELD MEASUREMENT BY GIANT MAGNETORESISTANCE SENSOR – APPLICATIONS TO NONDESTRUCTIVE EVALUATION AND BIOMEDICAL ENGINEERING
specimen.
Distance (mm)
-6-303
Gra
dien
t am
plitu
de(μ
V/p
itch)
Dis
tanc
e(m
m)
1.2
0
425 415459
250
432435459
Unit: μm
Scanning direction0 1 2 3 4
4
3
2
1
0
Distance (mm)
Dis
tanc
e (m
m)
01
23
4
01
23
4
0
1
2
-1
x-axis (mm)y-axis (mm)
Gra
dien
t am
plitu
de (μ
V/p
itch) Unit: μm
380 440 400
420
420380
(a) Detection of bead position (b) Conductive microbead (c) Detection result array model
Figure 7: Identification of the conductive microbead (PbSn) position, conductive microbead array model and its detection results.
4 Applications in Biomedical Engineering
4.1 Proposed SV-GMR sensor for biomedical application
The novel needle-type SV-GMR sensor is shown in Fig. 8(a). The light, low-invasive, and
simple SV-GMR sensor is fabricated especially for biological measurement. The needle is
made so that it can easily be injected into the body for in-vivo experimentation. The SV-GMR
element with sensing area 75 μm × 40 μm is fabricated on the tip of the needle. The size of
the needle of 20 mm is especially designed to be minimally invasive in potential medical
applications. It can be inserted inside the body to an area of interest in a very low invasive
condition.
The needle-type probe has a bridge structure of GMRs, which allows the sensor to measure
the magnetic flux density inside and outside a given cavity simultaneously. These
characteristics make the needle-type sensor a potentially ideal tool to be used during a
surgical procedure for example.
Probe connector
Cover
Sensing axis
Needle withsensor
Length15 mm
tipbonding pad
At the padof theneedle
GMRat the tip of
needle
- 2
- 1
0
1
2
- 120 - 100 - 80 - 60 - 40 - 20 0 20 40 60 80 100 120
GMR
sens
or o
utpu
t (m
V)
Magnetic flux density (µ T)
Bz
By
Bx
Magnetic flux density (µT)
GM
R s
enso
r out
put (
mV
)
Bx
By
Bz
x
z
y
(a) Proposed needle-type SV-GMR probe (b) Small-signal characteristics at 1 kHz Figure 8: SV-GMR needle probe and ac small-signal characteristics.
166
INTERNATIONAL JOURNAL ON SMART SENSING AND INTELLIGENT SYSTEMS, VOL. 1, NO. 1, MARCH 2008
Space factor hs = 0.523
0.91
1.11.21.31.41.51.61.71.8
0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4Weight density Dw
Experimental results
Calculated results
0.91
1.11.21.31.41.51.61.71.8
0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4Weight density Dw
Experimental results
Calculated results
Rel
ativ
e pe
rmea
bilit
y µ*
Weight density Dw
(a) Electron microscopy image and spherical cluster (b) Comparison of theoretical and experimental model results
Figure 9: Microscopic model of magnetic nanoparticles and relative permeability vs. weight density results.
Sensing direction is parallel to the needle. Constant current of 5 mA is supplied to the GMR
sensor. The ac small-signal characteristics at 1 kHz are shown in Fig. 8(b). The sensitivity of
the SV-GMR sensor is approximately 12.5 μV/μT.
4.2 Estimation of Magnetic Fluid Weight Density
4.2.1 Magnetic fluid and its applications
The past decade has seen an unprecedented increase in interest for the utilization of magnetic
nanoparticles in biomedical applications. Magnetic fluid or dextran magnetite (DM) is a
complex of dextran and iron oxide particles. It is not simply a suspension of the two but rather
chemically binds to the circumference of iron oxide particles, and is stable as a colloid
without aggregation or deposition in various solvents or serum [10].
Since the movement of magnetic nanoparticles can be controlled by external magnetic fields,
it is used in targeted drug delivery, where it is possible to carry drugs or medicine to a specific
site. Also magnetic nanoparticles react strongly to ac magnetic fields and thus, enabling
dynamic methods of cancer therapy such as hyperthermia treatment, or they can also be used
as contrasting agents in MRI [11].
Hyperthermia treatment involves magnetic nanoparticles being injected to the tumor area [12,
13]. AC magnetic fields heat up the magnetic nanoparticles and the tissues are heated directly
due to hysteresis loss. Induced heat capacity depends on amplitude of magnetic field, exciting
frequency and weight density Dw of magnetic fluid [14, 15]. Since the injected magnetic fluid
spreads inside the tissue it is critical to estimate the weight density of magnetic fluid inside
body before as well as after treatment. The proposed SV-GMR sensor has the potential to be
indispensable in this sense.
167
SOTOSHI YAMADA, HIGH-SPATIAL-RESOLUTION MAGNETIC-FIELD MEASUREMENT BY GIANT MAGNETORESISTANCE SENSOR – APPLICATIONS TO NONDESTRUCTIVE EVALUATION AND BIOMEDICAL ENGINEERING
4.2.2 Relationship between magnetic fluid weight density and relative permeability
It can be seen from Fig. 9(a) that magnetic particles have a cluster structure with space
between particles. It is assumed that the cluster of magnetite distributes uniformly and that the
nanoparticles have a relative permeability of infinity compared to that of one for liquid.
The permeance of an equivalent magnetic path is estimated through magnetic bead and
through air. Then, regarding the magnetic fluid as a bulk, the magnetic permeance of a unit
volume is calculated [14].
The relative permeability, μ*, can then be expressed by the following equation:
)1(/41* ⟨⟨+= wfsw DhD γμ (1)
where γ = 4.58 (W-35 sample – Taiho Co.) and is the specific gravity of magnetic bead. The
space factor h = 0.523 of spherical magnetite is considered, due to space between the
particles. Fig. 9(b) shows the comparison of theoretical results to experimental data, where the
latter is obtained by filling solenoid vessels with magnetic fluid of varying densities and
measuring relative permeability with the aid of a B-H
f
s
curve tracer.
As shown in Fig. 10, a uniform magnetic flux density, BB0, is applied to a magnetic fluid filled
cavity by a Helmholtz tri-coil. The magnetic flux density at the center of the cavity, B1, can be
expressed as,
( )*
01
*
BB
1 N 1
μ
μ=
+ −(2)
where N is the demagnetizing factor of the cavity [16]. Substituting Eq. (1) into Eq. (2), we
obtain the ratio of change between B0
as,
and B1B B
Caviity withmagnetic fluid
Tray withembeddedcavity
SV-GMRneedle-type
sensor
B0
B1
Helmholtz tri-coil
Magnetic fluxlines
( )( )
w1 0
0 s f
4 1 N DB BB h γ
−−= (3)
Hence, it can be seen from Eq. (3) that the
magnetic fluid weight density can be
168
Figure 10: Magnetic flux distribution in an
embedded cavity.
INTERNATIONAL JOURNAL ON SMART SENSING AND INTELLIGENT SYSTEMS, VOL. 1, NO. 1, MARCH 2008
calculated from the difference between the applied magnetic flux density, B0
, inside the fluid filled cavity. While the change in flux density is
proportional to the weight density, the shape of the cavity has an influence on the difference
of magnetic flux density.
B , and the
magnetic flux density, B1
0
1
2
3
4
5
6
0 0.5 1 1.5 2 2.5 3 3.50 0.5 1 1.5 2 2.5 3 3.50
1
2
3
4
5
6
Magnetic fluid weight density, Dw [%]Cha
nge
in M
agne
tic fl
ux d
ensi
ty,
[%
]
Shape ratio of cavity:m =b/a N = 0 (m = )∞
N = 1/3 (m = 1)
N = 0.413 (m = 0.5)
a
b
f = 100 Hz
Tray with cylindricalembedded cavity
Needle typeprobe
Helmholtz tri-coil
[ ]1 0
0
0
B B100 %
BB 100 T
δ
μ
⎛ ⎞−= ×⎜ ⎟⎝ ⎠=
(a) Experimental setup (b) Experimental results Figure 11: Experimental setup and results for estimation of low concentration magnetic fluid weight density.
B
4.2.3 Dependency of shape ratio and density
Experiments were performed to estimate the low concentration magnetic fluid weight density.
The size, shape and composition of magnetic nanoparticles make it an attractive tool in
0
0.01
0.02
0.03
0.04
0.05
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21Position (cm)
Cha
nge
in S
igna
l (m
V)
Dw = 0.24 % Dw = 0.36 %
Dw = 0.47 %
Dw = 0.81 %
m = 0.625
Center ofmagnetic fluid
Potato starch medium
2 cm 5 cm5 cm
5 cm
5 cm
10cm
16 mm
10 m
m
m = 0.625
00.005
0.010.015
0.020.025
0.030.035
0.040.045
0.050.055
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21
Top View of Model
Potato starch medium
2 cm 5 cm5 cm 5 cm 5 cm
10cm
3.2 cm2.4 cm 1.6 cm
4 cm
2 cm
1.5
cm
1 cm 2.
5 cm
m = 0.625
Dw = 0.81 %
Position (cm)
Cha
nge
in S
igna
l (m
V)
(a) Detection of magnetic fluid of various density (b) Detection of 0.81 % weight density at the shape
ratio m (height/diameter)=0.625.
Figure 12: Experimental results detecting magnetic fluid weight density.
169
SOTOSHI YAMADA, HIGH-SPATIAL-RESOLUTION MAGNETIC-FIELD MEASUREMENT BY GIANT MAGNETORESISTANCE SENSOR – APPLICATIONS TO NONDESTRUCTIVE EVALUATION AND BIOMEDICAL ENGINEERING
0
0.01
0.02
0.03
0.04
0.05
0.06
0.07
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21
Dw = 0.81 %
Cha
nge
in S
igna
l (m
V)
Position (cm)
Dendrites
Stimulus
Nucleus
Cell bodyAxon
Nerve impulse
Nodes ofRanvier
Myelin sheathcells
Axonterminalbundle
Chemicaltransmission
Figure 14: Nervous cell
Median nerve
Conductingwire simulatingnerve model
Magnetic field distributionfrom nerve model
SV-GMRelement
NeedleFigure 13: Detection of 0.81 % weight density at the shape ratio m =1.
Figure 15: Measurement of action current in nervous cell with the aid of needle-type SV-GMR sensor.
biomedicine and low concentration magnetic fluid (less than 2.8 % weight density of original
DM) is used in a variety of applications. Fig. 11(a) shows the experimental setup for low
concentration magnetic fluid weight density. Fig. 11(b) shows the results obtained at 100 Hz.
It can be seen that the magnetic fluid weight density and the differential magnetic flux density
has a linear and proportional relationship. The magnetic fluid weight densities of the samples
were adjusted accordingly by thinning original DM. The experimental results are compared
with theoretical estimations for long (N = 0), spherical (N = 1/3), and elliptical shapes (N =
0.413). The experimental results compare favourably with the theoretical estimations.
Further experiments were performed to detect magnetic fluid inside a reference medium. The
magnetic fluid and potato starch model (5 % agar, 5 % magnetic fluid and 90 % distilled
water) was used to simulate a potential in-vivo experimental situation.
Fig. 12(a) shows the detection results for magnetic fluid filled agar pieces of m = 0.625. Each
of the four agar pieces were injected with different magnetic fluid weight density of 0.24,
0.36, 0.47 and 0.81 %. The change in signal is the signal obtained corresponding to the
magnetic flux density inside the magnetic fluid, and in potato starch (reference medium). The
results show that not only can magnetic fluid be detected inside the reference medium but the
change in signal also seem higher for higher weight densities of magnetic fluid. Next, in order
to verify Eq. (3) which shows that N and m are constant, the differential magnetic flux density
should only change with the weight density. Figs. 12(b) and 13 show the results obtained for
170
INTERNATIONAL JOURNAL ON SMART SENSING AND INTELLIGENT SYSTEMS, VOL. 1, NO. 1, MARCH 2008
magnetic fluid filled agar pieces of weight density 0.81 % and different sizes of magnetic
fluid filled agar pieces. The different size magnetic fluid filled agar pieces in Fig. 12(b) has m
= 0.625 whereas it was m = 1 in Fig. 13. For m = 1, the same diameters and positioning were
used as in Fig. 12(b). The results verify Eq. (3) since there does not seem to be a highly
significant change in signal between the samples for m = 0.625 and 1.
4.3 Detection of Magnetic Field Distribution from Nervous Action Model
4.3.1 Background and methodology
The detection of magnetic field distribution from nervous action has the potential to diagnose
as well as treat neurological diseases. Diseases of the central and peripheral nervous system
such as Alzheimer disease, Parkinson disease, and multiple sclerosis are critical for the patient
in most cases. Nervous cells in living organisms transmit electrical signals. Signals travel
along neurons as waves of electrical discharge at a frequency of around 1 kHz. Action
potentials are conducted from dendrites through long axon to other neurons as shown in Fig.
14. The diameter of a typical nerve may vary between 1 and 20 μm [17, 18]. While action
impulse travels along the nerve at speeds between 0.6 -120 m/s [19], action current is not
more than 1 μA.
Fig. 15 shows the idea of the SV-GMR sensor needle inserted into the proximity of the nerve,
and the possible experimental simulation where a magnetic field created around a long thin
conductor can be measured, with the sensor placed in the proximity of the wires’s lateral
surface. Ampere’s law is used to calculate the magnetic field distribution in a long thin
current carrying wire.
IB2
μπρ
= (4)
where: I ; current flowing in the wire (A),
ρ ; distance from the centre of the wire (m),
μ; magnetic permeability (H/m),
B ; magnetic flux density (T).
4.3.2 Experimental results
A wire of 15 μm radius was used to simulate a nerve. To simulate action current a pulse signal
with 10 % duty cycle at frequency 1 kHz was supplied to the wire. The sensor needle was
placed so that it touches the conducting wire. Output signal from the sensor was amplified by
171
SOTOSHI YAMADA, HIGH-SPATIAL-RESOLUTION MAGNETIC-FIELD MEASUREMENT BY GIANT MAGNETORESISTANCE SENSOR – APPLICATIONS TO NONDESTRUCTIVE EVALUATION AND BIOMEDICAL ENGINEERING
an order of 1000 and analyzed by oscilloscope and lock-in amplifier depending if the signals
were above or below 1 μA respectively. The experiments were performed in a shielded area
where the attenuation of noise signals over 120 Hz is 1/100. Fig. 16 shows the logarithmic
plot comparing the signal detected by the SV-GMR sensor for a given action current and
the magnetic flux density calculated theoretically. It can be seen that both plots are linear and
parallel to each other. Next measurements were performed by moving the thin wire in the
vertical and horizontal direction as shown in Fig. 17. With an actual medical procedure in
mind, it is important to know the limit of distance from the wire in horizontal and vertical
direction. In this situation, ρ with respect to Eq. (1) is calculated as follows:
0.01
0.1
1
10
100
1000
1 10 100 10000.01
0.1
1
10
100ExperimentCalculation
Sign
al d
etec
ted
with
SV-G
MR
sens
or m
Vpp
Action current (μApp)
Mag
netic
fiel
d di
strib
utio
nca
lcua
lted
theo
retic
ally
(μT)
0-120 40 80 120 160-40-80-160
Sensorwidth
0
5
10
15
20
25
0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
Out
putv
olta
ge(m
V)
Distance (µm)
Mag
netic
flux
dens
ity(µ
T)
Experimental dataTheoretical calculation
Figure 18: Results obtained by measurement in vertical direction at 100 μA..
Figure 16: Comparison of detected signal with theoretical calculation for a given action current.
Distance (μm)0 100 200 300 400 500 600 700 800 900 1000
Out
put v
olta
ge (m
V)
Wire radius(15 μm)
5
10
15
20
25
30
0
Figure 19: Results obtained by measurement in horizontal direction at 100 μA .
172
INTERNATIONAL JOURNAL ON SMART SENSING AND INTELLIGENT SYSTEMS, VOL. 1, NO. 1, MARCH 2008
where: z ; vertical height from conductor to
point of interest (m),
(5)2 2cz rρ = +
rc ; radial distance from conductor (m).
It must be noted that during vertical direction measurements the sensor needle touches the
conducting wire as shown in Fig. 17(a) thus making rc negligible (so ρ is assumed equal to z).
However for horizontal measurement as shown in Fig. 17(b), it can be seen that the
conducting wire is moved away from the needle, making rc significant. Fig. 18 compares the
experimental results in the vertical direction for 100 μA to the theoretical data. However, it
must be noted that we assume the sensor as a point in the theoretical calculation, but in reality
it comprises of four thin elements amounting to 40 μm.
A possible reason for the signal in the detecting region not being uniform is that during
experiments the positioning of sensor could have introduced some error. Fig. 19 shows the
results obtained when the nerve was moved in the horizontal direction until no signal could be
deciphered. From the results, it can be seen that the signal can be detected around 750 μm in
the horizontal direction from the needle.
Further measurements were performed to simulate measurements through body tissue.
Measurements were taken through polyimide layers of differing thickness for 1.0 mA of
action current. The results in Fig. 20 show that the detected signal is inversely proportional
with increasing layer thickness of polyimide. In neurological measurements, the signals of
interest are usually less than microamperes. Fig. 21 shows the signal obtained from the SV-
GMR sensor for input currents ranging from 10 μA to 100 nA.
Conductingwire into page
Leng
th o
f GM
Rse
nsin
g el
emen
t
zρ
rcWire
mov
ed in
verti
cal d
irect
ion
40 μ
m
GMR sensingelement
Conductingwire into page
Leng
th o
f GM
Rse
nsin
g el
emen
t
zρ
rc
Wire moved inhorizontal direction
GMR sensingelement
40 μ
m
0
50
100
150
200
250
Thickness of Layer (≅m)
Out
put V
olta
ge (m
V)
15≅m
15≅m - radius of wire
20 40 60 80 100 120 140 160
Out
put v
olta
ge (m
V)
Thickness of layer (μm)15 μm
15 μm - radius of wire
(a) Vertical direction
measurement set up (b) Horizontal direction
measurement set up
Figure 17: Measurement of magnetic field in the horizontal and vertical directions.
Figure 20: Output voltage versus thickness of polyimide layers placed between conducting wire and sensor at 1.0 mA.
173
SOTOSHI YAMADA, HIGH-SPATIAL-RESOLUTION MAGNETIC-FIELD MEASUREMENT BY GIANT MAGNETORESISTANCE SENSOR – APPLICATIONS TO NONDESTRUCTIVE EVALUATION AND BIOMEDICAL ENGINEERING
0.01
0.1
1
10
0.1 1 10Input Current (μΑ)
Out
put V
olta
ge (m
V)
Figure 21: Output signal vs. Input current characteristics.
5 Conclusions
Spin-valve giant magnetoresistance sensors continue to revolutionize the nondestructive
evaluation industry due to its simple micro dimensional structure, wide operating frequency,
and low field operating range among many other advantages. Novel applications in
biomedical engineering employing a SV-GMR sensor included estimating magnetic fluid
weight density and detection of magnetic field distribution from nervous action. The
interesting characteristics of GMR sensor enable us to invent the drastic methodology of
nondestructive evaluation and biomedical measurement by magnetic measurement.
6 References
[1] T.Bratland, M.J.Caruso, R.W.Schneider, C.H. Smith, “A New Perspective on Magnetic
Field Sensing,” Sensors, vol.5, no.12, pp.34-46, 1998.
[2] J. Lenz and A. Edelstein, “Magnetic Sensors and Their Applications,” IEEE Sensors
Journal, vol.6, no.3, pp.631-649, 2006.
[3] S.C.Mukhopadhyay, “A Novel Planar Mesh Type Micro-Electromagnetic Sensor: Part I –
Model Formulation,” IEEE Sensors Journal, vol.4, no.3, pp.301-307, 2004.
[4] H.Wakiwaka, “Characteristics of Giant Magneto-resistance Sensor and Their
Application,” Journal of the Magnetics Society of Japan, vol.28, no.7, pp.825-833, 2004.
[5] T.Dogaru, S.T.Smith, “Giant Magneto-resistive-Based Eddy Current Sensor,” IEEE
Transactions on Magnetics, vol.37, no.5, pp.3836 -3838, 2001.
[6] C.H.Smith,R.W.Schneider, “Low-Field Magnetic Sensing with GMR Sensors Part 1:The
Theory of Solid-State Sensing,” Sensors, vol.16, no.9, pp.76-83, 1999.
[7] S.Yamada, K.Chomsuwan,Y.Fukuda, M.Iwahara, H.Wakiwaka, S.Shoji, “Eddy-Current
Testing Probe With Spin-Valve Type GMR Sensor for Printed Circuit Board Inspection,”
IEEE Transactions on Magnetics, vol.40, no.4, pp.2676-2678, 2004.
174
INTERNATIONAL JOURNAL ON SMART SENSING AND INTELLIGENT SYSTEMS, VOL. 1, NO. 1, MARCH 2008
[8] K.Chomsuwan, S.Yamada, M.Iwahara, H. Wakiwaka, S.Shoji, “PCB Conductor
Dimension and Alignment Inspection Using an ECT Probe with an SV-GMR Sensor,”
Transactions on Magnetics Society in Japan, no.5, pp.93-96, 2005.
[9] S.Yamada, K.Chomsuwan, T.Hagino, K. Minamide, M.Iwahara, “Conductive Microbead
Array Detection by High-Frequency Eddy-Current Testing Technique,” IEEE Transaction
on Magnetics, vol.41, no.10, pp.3622-3624, 2005.
[10] I.Nagano, H.Nagae, S.Shiozaki, I.Kawajiri, S. Ygitani, K.Katayama, K. Tazawa,
“Development of a Portable Cancer Treatment System Using Induction Heating: a New
Weapon for Killing the Cancer,” 2nd Kanazawa workshop, pp.11-15, 2006.
[11] G.Reiss, A.Hütten, “Magnetic Nanoparticles: Applications Beyond Data Storage”,
Nature Materials, vol.4, issue 10, pp.725-726, 2005.
[12] J. van der Zee “Heating the Patient: a Promising Approach?,” Annals of Oncology,
13(8), pp. 1173-84, 2002.
[13] R.Hergt, W.Andra, C.G.d’Ambly, I.Hilger, W.A. Kaiser, U.Richter,
H.G.Schmidt,“Physical Limits of Hyperthermia Using Magnetic Fine Particles,” IEEE
Transactions on Magnetics, vol.34, no.5, pp.3745-3754, 1998.
[14] S.Yamada, K.Chomsuwan, S.C.Mukhopadhyay, M.Iwahara, M.Kakikawa, I.Nagano,
“Detection of Magnetic Fluid Volume Density With SV-GMR Sensor,” Journal of the
Magnetics Society of Japan, vol. 31, no. 2, pp. 44-47, 2007.
[15] S.C.Mukhopadhyay, K.Chomsuwan, C.P. Gooneratne, S.Yamada, “A Novel Needle-
type SV-GMR Sensor for Bio-Medical Applications,” IEEE Sensors Journal, vol. 7, issue.
3, pp. 401-408, 2007.
[16] R.M.Bozorth, Ferromagnetism, 3rd ed, vol.2, Oxford: Clarendon, pp. 845-873, 1989.
[17] J.P.Wikswo, J.P.Barach, J.A.Freeman,“Magnetic Field of a Nerve Impulse: First
Measurements”, Science, vol.208, issue 4439, pp. 53-55, 1980.
[18] K.R. Swinney, J.P. Wikswo Jr, “A Calculation of the Magnetic Field of a Nerve
Action Potential,” Biophysical Journal, vol.32, pp.719-731, 1980.
[19] E.Tkacz, P.Borys, Bionika, WNT, pp.41-47, 2006.
175
SOTOSHI YAMADA, HIGH-SPATIAL-RESOLUTION MAGNETIC-FIELD MEASUREMENT BY GIANT MAGNETORESISTANCE SENSOR – APPLICATIONS TO NONDESTRUCTIVE EVALUATION AND BIOMEDICAL ENGINEERING