high-spatial-resolution magnetic-field measurement …s2is.org/issues/v1/n1/papers/paper10.pdf ·...

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

[email protected]

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


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


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


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


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


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


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.


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


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


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


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


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


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


[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


high-spatial-resolution magnetic-field measurement …s2is.org/issues/v1/n1/papers/paper10.pdf ·...

Documents