DSpace at VNU: Spatial angular positioning device with three-dimensional magnetoelectric sensors

Output signals from the device components are provided in form of sine and/or cosine functions of both the rotation az-imuth and the pitch angles, from which the total intensity as well

Spatial angular positioning device with three-dimensional magnetoelectric sensors

D T Huong Giang, P A Duc, N T Ngoc, N T Hien, and N H Duc

Citation: Review of Scientific Instruments 83, 095006 (2012); doi: 10.1063/1.4752763

View online: http://dx.doi.org/10.1063/1.4752763

View Table of Contents: http://scitation.aip.org/content/aip/journal/rsi/83/9?ver=pdfcov

Published by the AIP Publishing

Articles you may be interested in

Piezoelectric-metal-magnet dc magnetoelectric sensor with high dynamic response

J Appl Phys 114, 027016 (2013); 10.1063/1.4812225

A pencil-like magnetoelectric sensor exhibiting ultrahigh coupling properties

J Appl Phys 113, 134101 (2013); 10.1063/1.4798509

Overlap of the intrinsic and extrinsic magnetoelectric effects in BiFeO3-PbTiO3 compounds: Potentialities for

magnetic-sensing applications

J Appl Phys 113, 034102 (2013); 10.1063/1.4775796

Comparison of noise floor and sensitivity for different magnetoelectric laminates

J Appl Phys 108, 084509 (2010); 10.1063/1.3486483

Geomagnetic sensor based on giant magnetoelectric effect

Appl Phys Lett 91, 123513 (2007); 10.1063/1.2789391

(Received 21 June 2012; accepted 31 August 2012; published online 20 September 2012) This paper reports on the development of a novel simple three-dimensional geomagnetic device for sensing the spatial azimuth and pitch positions by using three one-dimensional magnetoelectric sen-sors assembled along three orthogonal axes This sensing device combines piezoelectric transducer plates and elongated high-performance Ni-based Metglas ribbons It allows the simultaneous detec-tion of all three orthogonal components of the terrestrial magnetic field Output signals from the device components are provided in form of sine and/or cosine functions of both the rotation az-imuth and the pitch angles, from which the total intensity as well as the inclination angle of the Earth’s magnetic field is determined in an overall field resolution of better than 10−4 Oe and an angle precision of±0.1◦, respectively This simple and low-cost geomagnetic-field device is

promis-ing for the automatic determination and control of the mobile transceiver antenna’s orientation with

respect to the position of the related geostationary satellite © 2012 American Institute of Physics.

[http://dx.doi.org/10.1063/1.4752763]

I INTRODUCTION

The intensity as well as the inclination of the terrestrial

magnetic field felt by a suspending object is a well known

function of the object’s geographic position on the Earth or

in the space Geomagnetic-field sensors are, thus, effectively

used for the orientation and local positioning of moving

ob-jects and in motions generally In this sense,

geomagnetic-field sensors are indispensable for applications in the space

engineering and technology sector, especially in

measure-ments of the magnetic field or of the spatial magnetic-field

gradient for different purposes The magnetic field in-orbit

can be sensed for geomagnetic-field measurements, or also

inversely, for the determination of the relative orientation of,

for instance, a spacecraft in the geomagnetic field, i.e., its

rel-ative position and orientation with respect to the Earth This

is indeed the purpose and function of the magnetic sensors in

the attitude control systems (ACS).1The ACS of a spaceship

is devoted to determine and to control the orientation of it, i.e.,

to sense and to adjust its relative orientation within an inertial

reference frame For missions like those of some

communica-tions satellites, for example, the relative orientation between

the geostationary satellite and the mobile transceivers must be

omni-directional controlled For such purposes, angular

posi-tioning devices require a very high sensitivity to accurately

determine both the spatial azimuth (ϕ) and pitch (θ ) angles

with respect to the orientation of the Earth’s magnetic field

Various types of magnetic sensors on the basis of

flux-gate, Hall effect, superconducting quantum interference, and

giant magnetoresistance spin valves have been deployed until

now for these applications.1 , 2In addition, many types of 2-D

and 3-D magnetic sensors have also been proposed for such

applications The proposed devices were manufactured in

dif-a) Electronic mail: giangdth@vnu.edu.vn.

ferent technologies based on various physical phenomena as detection principles.3 5 The recently explored magnetoelec-tric (ME) effect has offered great possibilities to develop a new generation of simple, low-cost but highly sensitive mag-netic sensors, which can be used in those devices.2 , 6 9Indeed,

we have newly reported an optimal design of a ME-based terrestrial magnetic-field sensing device combining elongated high-performance Ni-based Metglas ribbons and piezoelectric transducer (PZT) plates.6 This 1-D device exhibited a field sensitivity of better than 0.850 V/Oe and a field resolution in the order of 10−4Oe

On the basis of the above mentioned 1-D geomagnetic device, a new ME-based 3-D one has been developed for de-tecting spatial azimuth and pitch distance in this paper We will show that this 3-D device can simultaneously sense all three orthogonal components of the Earth’s magnetic fields From these data, complete and detailed quantitative informa-tion can be obtained on the resulting magnetic field intensity

as well as the azimuth and pitch angles at a given device’s position and orientation with respect to the Earth’s magnetic field An overall intensity and angular resolution of better than

10−4Oe and 10−1◦, respectively, have been achieved with this device This suggests that the so far developed, simple, low-cost and highly sensitive geomagnetic-field device is promis-ing for potential applications in controllpromis-ing the relative orien-tation between mobile transceivers and a geosorien-tationary com-munications satellite

II THE CONFIGURATION OF THE 3-D ME SENSOR

As shown in Fig 1, the 1-D geomagnetic-field sen-sors were designed and fabricated in a sandwich configu-ration of a Metglas/PZT/Metglas ME laminate composite, reported recently in Ref 6 In this configuration, the ME 0034-6748/2012/83(9)/095006/6/$30.00 83, 095006-1 © 2012 American Institute of Physics This article is copyrighted as indicated in the article Reuse of AIP content is subject to the terms at: http://scitationnew.aip.org/termsconditions Downloaded to IP:

095006-2 Giang et al. Rev Sci Instrum 83, 095006 (2012)

FIG 1 One-dimensional ME sensor construction: (a) SEM images at low

and high (see the image in the inset) magnification of the sandwiched

Met-glas/PZT/Metglas 15 × 1 mm 2ME laminate composite with the vectors hac

and P indicating the applied ac magnetic field and the electrical polarization

direction, respectively The Metglas and the adhesive layer with the

respec-tive thickness of 18 μm and 7 μm are recognized (b) The 1-D ME sensor

prototype where the coil generating an ac field is directly wrapped around the

ME laminates.

laminates consist of one out-of-plane poled PZT plate

sand-wiched between two magnetostrictive Metglas laminates

(Fig.1(a)) The 500 μm-thick piezoelectric plates used were

of the type APCC-855 supplied by The American

Piezoce-ramics Inc., PA The 18 μm-thick magnetostrictive laminates

were cut from Fe76.8Ni1.2B13.2Si8.8 melt-spun (also called

Ni-based Metglas) ribbons with the size of 15 × 1 mm2

The PZT plate was mechanically firmly sandwiched between

the two Metglas layers by using an epoxy layer of around

7 μm-thickness The total laminate volume is of about 15

× 1 × 0.55 mm3 To form the 1-D geomagnetic-field sensor, a

solenoid coil with turn density of 10.5 turns/mm was wrapped

around the entire sandwiched ME laminates (Fig.1(b))

The 3-D geomagnetic-field sensor was then created by

assembling three 1-D sensors S1, S2, and S3 aligned along

the three orthogonal axes, as shown in Fig.2(a) Figure2(b)

shows the fixed geocentric reference frame (XE, YE,ZE) for

determining the directions and angles, in which the XE-axis is

pointing toward the magnetic North pole, the YEaxis pointing

toward the East pole, and ZE-axis is vertical, positive pointing

towards the Earth’s center Here, the azimuth angle is defined

as a horizontal angle measured in clockwise rotating from XE

-axis to the S1-sensor The pitch angle is then determined by

the angle between the S3-sensor and ZE-axis by clockwise

ro-tation of the 3-D sensor around the XE-axis, i.e., in the

verti-FIG 3 Photograph of the simple 3-D rotation system setup.

cal planes (see Fig.2(c)) This 3-D sensor is then mounted on

a simple 3-D rotating system by combining rotations in just horizontal and vertical planes (Fig.3)

III ME LAMINATE CHARACTERIZATION

A Resonating frequency and quality factor

The ME laminate composites were characterized under a

weak ac magnetic field hac (hac = hosin(2π fot)) in the

pres-ence of a bias dc magnetic field H In the our

experimen-tal setup, the output voltage induced across the PZT plate by

the ac field was measured on a commercial lock-in amplifier

FIG 2 The image of 3-D ME sensor prototype (a) and illustrations of the azimuth (b) and pitch (c) angles referred to the fixed axes of the geocentric reference

frame (XE,YE,ZE ) corresponding to North-East-Down (NED frame) with the Earth.

FIG 4 Frequency dependence of the MEVC for ME laminate composites

corresponding to sensors S1, S2, and S3 The inset shows the signals in details

around the resonating frequencies.

(Model 7265, Signal Recovery), which simultaneously

sup-plied the input current for hac An electromagnet was used

to provide the bias field while the oscillating field with

am-plitudes of hac = 10−2 Oe was generated by a Helmholtz

coil The magnetoelectric voltage coefficient (MEVC-αE)

was then determined from the magnetoelectric voltage

re-sponse (MEVR) VMEversus the applied magnetic field by the

equation

αE= VME

hac· tPZT

(1)

with tPZTas the PZT plate thickness

For illustration, Fig.4shows the frequency dependence

of the MEVC for investigated ME composite laminates at

a fixed bias dc magnetic field of 2 Oe Note that the

well-pronounced peaks of resonance are clearly observed and their

height is enhanced with the increasing bias dc magnetic fields

up to a certain value (see Fig.5below) These peaks are found

to locate at the resonating frequencies fr of 99.95, 100.18,

and 100.13 kHz for the S1, S2, and S3 sensors, respectively,

with a quality factor of around 1.5% The difference between

the observed resonating frequencies, however, is of less than

0.5%, i.e., still within the bandwidth range of the resonating

frequency In the present work, the slight difference in the

res-onating frequencies of the S1, S2, and S3sensors may relate to

mechanical coupling modifications which could be caused in

the different manually manipulated lamination processes

B Magnetic-field intensity dependence of the MEVC

The MEVC at the individual resonating frequencies are

presented in Fig 5 in the dependence on the bias dc

mag-netic field for all investigated ME laminate composites As

can be seen, despite a few slight discrepancy related to

var-ious detailed aspects of the sensor’s technical (geometrical,

fabrication) parameters, the MEVC data for all measured ME

laminate composites show a significantly similar behavior in

the curves as well as in the absolute magnitudes The MEVC

initially increases with the applied magnetic field and reaches

a maximum value as high as 130 V/cm Oe at the applied field

intensity of around 7 Oe and then decreases as the applied

magnetic field further increases

FIG 5 Bias magnetic-field dependence of the MEVC at the resonating

fre-quencies of the ME laminate composites for the corresponding sensors S1 ,

S2, and S3

IV GEOMAGNETIC SENSOR CHARACTERIZATIONS

A Sensor calibration

In the working mode of the ME geomagnetic-field sensor, the solenoid coils in all of the 1D-ME sensors always function

as ac magnetic-field generators at their individual resonating

frequencies and they all are fed by an ac current source On

the other hand, in this measuring arrangement, the terrestrial magnetic field, which should be sensed and determined, plays

the role of the dc magnetic field H For an operation test of

the as-fabricated sensor in the conventional intensity range

of the geomagnetic field, however, a commercial Helmholtz coil supplied by a Keithley 230 current source was used to generate the terrestrial-like magnetic field of the intensity in the range up to 1.5 Oe with the accuracy of 10−5 Oe In this

mode, the observed MEVR from the S1, S2, and S3 sensors

in the presence of the low external magnetic fields are shown

in Fig.6 The figures obviously indicate a linear variation of ME-voltage with the external magnetic field in the field range

of interest From this result, it turns out that the magnetic

field calibration coefficient k of the sensors could be derived

as k1 = 192.6, k2 = 200.8, and k3 = 205.5 mV/Oe corre-sponding to the field resolution of 3× 10−4 Oe for the S

1,

FIG 6 MEVR at low external magnetic fields for the corresponding sensors

S1, S2, and S3 The included fitting curves indicate the field sensitivity (and/or field calibration coefficients).

This article is copyrighted as indicated in the article Reuse of AIP content is subject to the terms at: http://scitationnew.aip.org/termsconditions Downloaded to IP:

095006-4 Giang et al. Rev Sci Instrum 83, 095006 (2012)

FIG 7 MEVR data versus azimuth ϕ-angle excited by two ac magnetic fields with an 180◦phase shift difference (hacand –hac) described in the Cartesian (a)

and polar coordinate (b).

S2, and S3sensors, respectively These calibration coefficients

(and/or sensor sensitivities) reveal an amplitude discrepancy

of about 5% They are, however, two orders of magnitude

bet-ter than those previously reported for similar magnetic-field

sensor devices, but comparable with that of available

com-mercial geomagnetic-field sensors.10 , 11 This also reveals the

excellent capability and suitability of our as-developed and

as-described sensor for an accurate determination of the

geo-magnetic field

B Offsets calibration

Presented in Fig.7(a)is the S1-sensor output MEVR

ver-sus azimuth ϕ-angle measured by rotating the sensor system

in the horizontal plane about ZE-axis It is clearly seen that

the recorded sensor signal well corresponds to the harmonic

cosine function of the rotation angle ϕ The signal, however,

is accompanied by a rather high non-zero offset value This

offset shift can be manually evaluated from zero to the

cen-ter value between the maximum and minimum peaks In our

sensor, this problem was automatically solved by averaging

the two output signals excited by the two ac magnetic fields

of a 180◦phase shift difference (denotes as hacand –hac) The

non-zero offset contribution to the MEVR is represented by

the cardioids in the polar coordinate in Fig.7(b) The offset

shall centralize the circle by averaging where the circle radius indicates the amplitude of the offset signal

C Terrestrial magnetic-field intensity and spatial azimuth angle positioning

The amount of azimuth rotation of the sensor S1 with

respect to the North magnetic pole (XE-axis) is sensed by

turning the 3-D sensor system about the ZE-axis Here, the

sensors S1and S2are in the Earth surface’s (horizontal) plane

The offset-compensated signals from the S1, S2, and S3sensor

versus the ϕ-angle in a complete circular cycle are plotted

in Fig 8(a) The two V1 and V2 curves correspond well to

the cosine and sine functions of the rotation angles ϕ These signals always reach a maximum of 77 and 80.3 mV in the S1

and S2sensors, respectively, when the sensors are pointing to the North direction When the sensor was oriented along the East and/or West direction in this plane, its output diminished

to zero As seen in the figure, the V3curve of the respective

S3-sensor is an almost perfectly flat line around 40.9 mV The positive sign means that the magnetic field of the Earth

is pointing down at our location (located in the Northern hemisphere)

By using the magnetic-field calibration coefficients of the

respective sensors as mentioned above, three components (H1,

FIG 8 (a) MEVR from the S1, S2, and S3as a function of the azimuth angle plotted in the Cartesian coordinate, (b) the three derived orthogonal Hi (=Vi/ki ),

horizontal component (Hxy) components and the total intensity (Htot ) of the Earth’s magnetic field described in the polar coordinate and (c) the angles in degree

unit calculated from the arctangent function of ratio (H2/H1) and (H3/Hxy ) corresponding to azimuth and inclination angle of the Earth’s field.

FIG 9 (a) MEVR from the S1, S2, and S3as a function of the pitch angle plotted in the Cartesian coordinate, (b) the three derived orthogonal Hi (=Vi/ki ),

vertical (Hz) components and the total intensity (Htot ) of the Earth’s magnetic field described in the polar coordinate and (c) the pitch angles in degree unit

calculated from the arctangent function of ratio (H2/H3 ).

H2, H3) of the Earth’s magnetic field were derived and they

are plotted in the polar coordinate system (see Fig.8(b)) In

this description, the data are distributed on almost perfect

cir-cles From these orthogonal components, the horizontal (Hxy)

and the total (Htot) terrestrial magnetic-field intensity can be

computed using the expression

H xy=

H2

1 + H2 2



, H tot =

H2

1 + H2

2 + H2 3



. (2)

The plots of Hxy and Htotin the polar coordinate system

are also shown in Fig.8(b) It turns out in this experiment that

the strength of the horizontal terrestrial magnetic field Hxy

in our localized laboratory conditions (Hanoi, Vietnam) is in

the order of 0.3998 Oe and the strength of the total Earth’s

magnetic field Htot equals to 0.4466 Oe Combining the

de-rived terrestrial magnetic-field horizontal (Hxy) and the

ver-tical components (H3), the inclination (or dip) angle of the

Earth’s field to the surface of the Earth can be directly

de-termined by using the arctangent function of (H3/Hxy) The

calculated results presented in Fig.8(c)are almost invariant

at around θi = 26.5 ± 0.1◦ An acceptable discrepancy of

0.1◦ is appropriate with an extremely small change in the

magnetic field strength estimated 10−4 Oe that can be

re-liably resolved by this sensor Although the measurements

were performed under best arranged experimental conditions

for the purpose of studying the geomagnetic field, the results

may still be influenced by indoor objects This finding,

how-ever, is in good consistency with standardized data reported

previously.12–14

For the determination of the azimuth ϕ-angle, only the

two sensors S1and S2of the respective H1and H2components

are involved by using the relationship tanϕ = tan(H2/H1) To

account for the tangent function being valid over 180◦and not

allowing the H1division calculation, the following equations

can be used:

H1= 0 and H2 < 0 : ϕ = 90o,

H1= 0 and H2 > 0 : ϕ = 270o,

H1> 0 and H2 < 0 : ϕ = − arctan (H2/ H1) ,

H1< 0 : ϕ = π − arctan (H2/ H1) ,

H1> 0 and H2 > 0 : ϕ = 2π − arctan (H2/ H1)

(3)

The derived results converted to degrees as presented in Fig.8(c)show a perfect linear variation with the experimental

ϕ-angle of a slope exactly equal to 1

D Spatial pitch angle positioning

The sensor signals depending on the pitch angle (θ ) be-tween the S3-sensor and the reference ZE-axis were measured Figure 9(a) illustrates the S2 and S3 sensor readings when

turning the 3-D sensor around the S1-sensor axis, i.e., around the North direction This plot again shows a sine and cosine

output response to the θ -angles during the rotation for the two respective sensors S2and S3 In this case, the V1curve of the

corresponding S1sensor is again an almost perfectly flat line around 77 mV The maximum values of 40 and 40.9 mV

for the sensors S2and S3, respectively, always were reached when the sensor axes are pointed vertically downward (i.e.,

along the ZEaxis) The plots of the derived components (H1,

H2, H3) of the Earth’s field in the polar coordinate system (shown in Fig.9(b)), again, are well fitted in perfect circles

By using the relationship tanθ = tan(H2/H3) and the same equation (1)applied for the ratio of (–H2/H3), the pitch

an-gle θ was again determined Results presented in Fig. 9(c)

give exactly the experimental pitch angle in complete circular rotation cycle

V CONCLUSIONS

A novel 3-D geomagnetic device for detecting the az-imuth and pitch positions has been developed on the basis

of three simple, low-cost and high-sensitivity 1-D ME sen-sors arranged in a perpendicular configuration The device al-lows simultaneous detections of all the three orthogonal com-ponents of the terrestrial magnetic field Its output signals in form of sine and cosine function of the rotation azimuth and pitch angles provide complete and detailed quantitative infor-mation on the intensity of the terrestrial magnetic field as well

as the value of the azimuth and pitch angles with respect to the direction of the Earth’s magnetic field The overall intensity and angular angle accuracy of the device has been determined

as better than 10−4Oe and 10−1◦, respectively This sensor is

This article is copyrighted as indicated in the article Reuse of AIP content is subject to the terms at: http://scitationnew.aip.org/termsconditions Downloaded to IP:

095006-6 Giang et al. Rev Sci Instrum 83, 095006 (2012)

being integrated with an electronic interface on a mobile

satel-lite signal receiver for the automatic determination and

con-trol of the latter antenna direction with respect to the satellite

position in space

ACKNOWLEDGMENTS

This work was supported by the Vietnam National

Foun-dation for Science and Technology Development

(NAFOS-TED) under the granted Research Project No 103.02.86.09

and by the National Research Program on Space Technology

of Vietnam

1 M Díaz-Michelena, “Small magnetic sensors for space applications,”

Sen-sors9, 2271 (2009).

2 J Zhai, S Dong, Z Xing, J Li, and D Viehland, Appl Phys Lett.91,

123513 (2007).

3 N H Duc and D T Huong Giang, J Alloys Compd.449, 214 (2008).

4 D T Huong Giang and N H Duc, Sens Actuators A149, 229 (2009).

5 L Ding, J Teng, X C Wang, C Feng, Y Jiang, G H Yu, S G Wang, and

R C C Ward, Appl Phys Lett.96, 052515 (2010).

6 D T Huong Giang, P A Duc, N T Ngoc, and N H Duc, Sens Actuators

A179, 78 (2012).

7 F Burger, P A Besse, and R S Popovic, Sens Actuators A67, 72 (1998).

8 S Lozanova and Ch Roumenin, Sens Actuators A162, 167 (2010).

9 S Lozanova, A Ivanov, and Ch Roumenin, “A novel three-axis Hall

mag-netic sensor,” Procedia Engineering 25, 53 (2011).

10M Johnson, Magnetoelectronics (Elsevier, Amsterdam, 2004).

11M J Haji-Sheikh, in Sensors, edited by S C Mukhopadhyay and R Y M.

Huang (Springer-Verlag, Berlin, 2008), p 23.

12T T Ai, Geomagnetism and Magnetic Prospecting (Vietnam National

Uni-versity, 2005).

13 See http://www.ngdc.noaa.gov/seg/geomag/jsp/struts/calcIGRFWMM for information about the inclination angle and the total terrestrial magnetic-field intensity located at Hanoi, Vietnam.

14 See http://magnetic-declination.com/ for information about the inclination angle and the total terrestrial magnetic-field intensity located at Hanoi, Vietnam.