IEEE TRANSACTIONS ON ELECTRON DEVICES, VOL. 68, NO.

4, APRIL 2021 1495

AlGaN/GaN 2-D Electron Gas for Highly

Sensitive and High-Temperature
Current Sensing
Hong Quan Nguyen , Toan Dinh , Hamid Amini Moghadam, Tuan Khoa Nguyen, Thanh Nguyen ,
Jisheng Han, Sima Dimitrijev , Senior Member, IEEE, Yong Zhu , Senior Member, IEEE,
Nam-Trung Nguyen , Senior Member, IEEE, and Dzung Viet Dao

Abstract — This article presents the design and character- drops across a shunt resistor [2], [3]. However, an overcurrent
ization of a direct current (dc) sensor utilizing the Hall effect can permanently damage the shunt resistor, while the high
in AlGaN/GaN 2-D electron gas (2DEG) four-terminal devices power losses make this method unsuitable for measuring
and a flux concentrator. The sensor was fabricated from
an AlGaN/GaN/Si wafer grown by metal-organic chemical high currents. Other current sensing techniques utilize current
vapor deposition. The sensor exhibited excellent linearity transformers and Rogowski coils that are only applicable
and repeatability with a high Hall voltage under the primary for alternating current (ac) sensing [4], [5]. Moreover, these
current ranging from −5 to 5 A. The sensitivity of the sensor techniques yield a low sensitivity at low frequencies and the
was measured to be 0.26 (V/A)/A at 20 ◦ C and independent accuracy depending upon the conductor’s position [6]–[8].
of ambient temperature up to 200 ◦ C. The obtained result is
greater than that of other reported Hall effect-based current Recently, the successful demonstration of 2-D electron gas
sensors. The high sensitivity and thermal stability at varying (2DEG) formed by AlGaN/GaN heterostructures indicates the
temperatures are attributed to the high electron mobility, potential development of high sensitivity ac/direct current (dc)
wide bandgap, and stability of carrier density in 2DEG. Com- current sensors based on the Hall effect due to its high
bining these factors with the excellent mechanical strength, electron mobility, wide bandgap, and superior magnetic field
electrical conductivity, and chemical inertness of GaN, the
proposed sensor is promising for current monitoring in a sensitivity. The operation of the Hall effect current sensor is
wide range of operation temperatures. based on Ampere’s law that a magnetic field is generated
around a current-carrying conductor. This field will be detected
Index Terms — AlGaN/GaN 2-D electron gas (2DEG),
current sensor, energy band analysis, Hall effect. by a Hall sensor. There were several researches on coreless
Hall current sensor [9]–[14]. However, the magnetic field
generated from the conductor is relatively small, consequently
I. I NTRODUCTION creating new challenges for detecting such a small signal.

C URRENT sensing has been typically employed to moni-

tor electric current, energy consumption, and the energy
generated from power plants [1]. One of the conventional
In addition, the magnetic field significantly depends on the
distance between the sensors and the conductor.
In this research, we demonstrated a highly sensitive cur-
current measurement methods relies on measuring the voltage rent sensor based on AlGaN/GaN 2DEG. We employ a flux
concentrator (magnetic core) to concentrate the magnetic field
Manuscript received December 25, 2020; revised January 15, 2021;
accepted January 20, 2021. Date of publication February 22, 2021; onto the sensing area. The flux concentrator was proven to
date of current version March 24, 2021. This work was supported eliminate the positional error, which resulted from the position
in part by the Griffith University’s New Researcher Grants, in part shifting between the conductor and the sensor. It also greatly
by the Australian Research Council under Grant LP150100153 and
Grant LP160101553, and in part by the Queensland node of the Aus- reduces the interference of ambient magnetic fields.
tralian National Fabrication Facility, a company established under the
National Collaborative Research Infrastructure Strategy to provide nano
and micro-fabrication facilities for Australian researchers. The review
II. E XPERIMENTS
of this article was arranged by Editor K. Alam. (Corresponding author: The Hall device was fabricated by growing a GaN layer
Hong Quan Nguyen.)
Hong Quan Nguyen, Toan Dinh, Hamid Amini Moghadam, Tuan Khoa with a thickness of 4 μm on a silicon substrate (thickness
Nguyen, Thanh Nguyen, Jisheng Han, Sima Dimitrijev, and Nam-Trung of 300 μm). The 2DEG layer was then formed by growing
Nguyen are with the Queensland Micro-Nanotechnology Centre, Griffith an AlGaN layer with a thickness of 10 nm on top of the
University, Brisbane, QLD 4222, Australia (e-mail: quan.nguyen@
griffithuni.edu.au; [email protected]; h.aminimoghadam@ as-deposited GaN film, as shown in Fig. 1(a). The electrodes
griffith.edu.au; [email protected]; thanh.nguyen11@ were formed by sputtering Ti/Al/Ni (100 nm/300 nm/100 nm)
griffithuni.edu.au; [email protected]; [email protected]; on top of the AlGaN layer, followed by rapid thermal
[email protected]).
Yong Zhu and Dzung Viet Dao are with the School of Engineering annealing at 800 ◦ C for 30 s in nitrogen-rich ambient. The
and Built Environment, Griffith University, Brisbane, QLD 4222, Australia wafer was then cut into 10 mm × 10 mm square-shaped
(e-mail: [email protected]; [email protected]). chips, as shown in Fig. 1(b).
Color versions of one or more figures in this article are available at
https://doi.org/10.1109/TED.2021.3054360. The current–voltage characteristic of the sample was mea-
Digital Object Identifier 10.1109/TED.2021.3054360 sured using a Keithley 2450 SourceMeter. Fig. 1(c) shows
0018-9383 © 2021 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission.
See https://www.ieee.org/publications/rights/index.html for more information.

Authorized licensed use limited to: Univ of Calif Santa Barbara. Downloaded on May 20,2021 at 17:06:00 UTC from IEEE Xplore. Restrictions apply.
1496 IEEE TRANSACTIONS ON ELECTRON DEVICES, VOL. 68, NO. 4, APRIL 2021

TABLE I
T YPICAL S ENSITIVITY OF C OMMERCIALLY AVAILABLE H ALL S ENSORS

Fig. 1. (a) Schematic sketch of the device structure. (b) Layout of the
device. (c) Current–voltage characteristic of the device. Inset: Raman
spectrum of the AlGaN/GaN device.

the linearity of the current–voltage characteristics between

terminals A and B, as well as between terminals C and D
in the voltage range from −0.5 to +0.5 V, indicating that a
good Ohmic contact was formed between the electrodes and
AlGaN.
The sensitivity of a Hall effect device with respect to supply
current (S I ) is defined as follows
VH
SI = (1)
Is B
where VH is the Hall voltage, B is the magnetic field applied
perpendicular to the plane of the device, and Is is the supply
Hall current. The Hall voltage VH can be obtained by Fig. 2. (a) Geometry and dimension of the magnetic core. (b) 3-D view.
(c) Comparison of experimental and simulation result of the magnetic
Vm,1 − Vm,2 + Vm,4 − Vm,3 flux density versus the primary current. Inset: measurement result of
VH = (2)
4 magnetic flux density generated by 3-(A) primary current at different
distance between the conductor and the sensor.
where Vm,1 is VCD measured at positive field B and positive
current IAB , Vm,2 is VCD measured at positive field B and
negative current IAB , Vm,3 is VCD measured at negative field B a toroidal core to create a 0.5-mm gap for placing the Hall
and positive current IAB , and Vm,4 is VCD measured at negative plate.
field B and negative current IAB . By using this measurement
method, the offset voltages from misalignment and thermal III. R ESULTS AND D ISCUSSION
electric can be eliminated.
The output Hall voltage of a Hall effect device is dependent
The sensitivity of the AlGaN/GaN 2DEG device was mea-
on the electron density in 2DEG and the intensity of the
sured to be 77 V A−1 T −1 , which is greater than that of other
magnetic field [16]
semiconductor-based devices reported in the literature, Table I.
rH
The sensitivity of our device was improved due to several VH = G H Is B (3)
factors, including higher electron mobility and point-like con- q N2DEG
tacts. According to U. Ausserlechner, small contacts make where VH is the Hall voltage, G H is the geometrical correction
the largest Hall signal but also the largest noise (high factor of Hall voltage, r H is the Hall factor, Is is the Hall
impedance) [15]. supply current, B is the magnetic field applied perpendicular
To make the electric current sensor, the chip was bonded to the plane of the device, q is the electron charge, and N2DEG
on a glass film and then attached firmly to the core for the is the sheet density.
characterization of dc current sensing, as shown in Fig. 2(a) From (3), a low sheet density would result in a high
and (b). In this work, the core is formed from two halves of Hall voltage. At room temperature (RT), electron mobil-

Authorized licensed use limited to: Univ of Calif Santa Barbara. Downloaded on May 20,2021 at 17:06:00 UTC from IEEE Xplore. Restrictions apply.
NGUYEN et al.: AlGaN/GaN 2DEG FOR HIGHLY SENSITIVE AND HIGH-TEMPERATURE CURRENT SENSING 1497

ity μ and sheet density N2DEG achieved in the AlGaN/GaN

heterostructure were measured to be 2500 cm2 V−1 s−1 and
2.5 × 1013 cm−2 , respectively. To obtain this result with high
accuracy, Hall effect method was used.
An electric current generates a magnetic field surrounding
the conductor. The magnetic flux density is determined as [16]
μo I p
B= (4)
2πr
where μ0 is the vacuum permeability (4π × 10−7 Tm/A),
r is the distance from the centerline of the conductor to
the sensor (m), and I p is the primary current flowing in Fig. 3. (a) Hall voltage versus primary current at RT and Hall supply
current IS = 0.5 mA. (b) Variation of the Hall voltage with three different
the conductor. However, the magnetic field that is produced primary currents (i.e., 1, 2, and 3 A). The Hall supply current was 0.5 mA.
by the electrical-carrying conductor is normally very small.
As amplifying the magnetic signal would increase the Hall
voltage, we utilized a magnetic flux concentrator in the form magnitude higher than other work using Si as Hall plate [10]–
of a toroidal core, as shown in Fig. 2(a) and (b). The nanocrys- [14], [26].
talline Vitroperm alloy was selected as core material due to To characterize the reliability of the device, repeatability
its high permeability 80 000 and low remanence 0.05 mT. The tests were carried out with a constant Hall supply current
alloy contains 82.8% Fe, 8.8% Si, 1.5% B, 5.6% Nb, and 1.3% Is of 0.5 mA. Under the application of a certain primary
Cu. current I p in the range from 1 to 3 A, the Hall voltage
The magnetic core converges the magnetic flux surrounding was stable without any noticeable signal drift [see Fig. 3(b)].
the conductor into a magnetic flux perpendicular to the chip The Hall voltage also returned to its initial value once the
surface. The magnetic flux density focused by the core to the primary current I p was completely turned off. When the
sensor can be calculated as [23], [24] primary current was OFF, the offset voltage is around 40 nV.
μo μ i I p This excellent repeatability of the Hall voltage under different
B= (5) primary currents is of interest for highly accurate current
l m μo + l g μi
measurements with great sensitivity.
where B is the magnetic flux density, μi is the initial perme- There are two typical noise sources existing in the device,
ability of the core, lm is the mean length of the core, and the including the thermal noise (Johnson noise) and 1/ f noise
length of the air gap is l g = 0.5 mm. (Hooge’s noise). The thermal noise component is calculated as
The flux density of the magnetic field is magnified by a
factor of 61 compared to the case without the core. The VJn(rms) = (4kBTR  f )1/2 (7)
measured results agree well with the simulation result [see
where k B is Boltzmann’s constant, R is the resistance,  f is
Fig. 2(c)].
the measurement bandwidth, and T is the absolute tempera-
To investigate the effect of conductor position on the
ture. The impedance of our device is 3125 . The thermal
magnetic flux density, we conduct an experiment by using
noise of the device is calculated to be VJn = 22.5 nV at a
the Magnetometer (Koshava 5) to measure the magnetic flux
measurement bandwidth of 10 Hz.
density at the gap of the magnetic core. The sensing area at the
The 1/ f noise can be described as [27]
probe’s tip of the magnetometer was 4 × 4 mm. The probe’s
tip was inserted inside the air gap and aligned carefully so αV02 f max
that it was covered completely by the core’s cross section. V1/2 f (rms) = ln( ) (8)
N f min
As shown in the inset Fig. 2(c), the magnetic flux generated
in the magnetic core changed less than 4% due to the shifting where α is the Hooge parameter. For AlGaN/GaN, α was
of the conductor inside the magnetic core. reported to be 10−4 [28]. V0 is the supply voltage to
Fig. 3(a) shows the relationship between the Hall voltage the sensor, and f max and fmin are upper and lower limits
VH and the primary current I P , indicating that the Hall voltage of measurement frequency, respectively. The total number
increases/decreases linearly with increasing/decreasing of the of carriers N is proportional to the area of the 2DEG
primary current. (N = sheet density × width × length = 2.3 × 1013 electrons).
The sensitivity of the Hall current sensor with respect to With V0 = 1.56 V, f min = 2 Hz to f max = 12 Hz, and the 1/ f
primary current can be defined as noise voltage can be calculated to be V1/ f = 5.97 nV. The
total noise voltage
 caused by thermal noise and 1/ f noise is
VH 1 Vnoise_cal = VJn + V1/2 f = 23.2 nV.
2
S= × . (6)
Is Ip ( f ) was also measured at
The voltage spectral density V
The current sensitivity of the fabricated 2DEG Hall effect different supply currents Is in the frequency range from 1 Hz
sensor was found to be as high as 0.26 (V/A)/A. This sensi- to 1 kHz by using the spectrum analyzer (USB-SA44B, Signal
tivity is 13 times higher compared to another report on Hall Hound), as shown in Fig. 4 . V ( f ) in the low-frequency
current sensor using the same material [25] and one order of region is dominated by the 1/ f noise, which decreases with

Authorized licensed use limited to: Univ of Calif Santa Barbara. Downloaded on May 20,2021 at 17:06:00 UTC from IEEE Xplore. Restrictions apply.
1498 IEEE TRANSACTIONS ON ELECTRON DEVICES, VOL. 68, NO. 4, APRIL 2021

Fig. 4. Voltage spectral density versus frequency at different Hall supply Fig. 5. Hall voltage versus primary current at temperature ranging from
currents and zero primary current. 20 ◦ C to 200 ◦ C. Inset: the variation in sensitivity due to temperature.

increasing frequency, whereas V ( f ) in the high-frequency

region is mainly thermal noise, and it is independent on the the device size, so it can be considered as point-like contact,
frequency. The measured noise voltage with supply current and therefore, G H = 1 [16]. However, for nonperipheral
Is = 0.5 mA in the frequent range 2–12 Hz can be obtained contacts of the Hall plate, G H may be smaller than 1. The
f max  Hall factor r H was proven to be very weak dependence on
by Vnoise_m = f min [ V ( f )] d f = 23.7 nV. The difference
2

between the calculation and measurement results may arise temperature [30]. Thus, the sheet density is the dominant
from other noise sources (e.g., measurement equipment and temperature-dependent parameter, and it will be analyzed in
electromagnetic noise) rather than just thermal noise and 1/ f more detail in the following.
noise from the sensor. The bandgap of GaN was calculated using the Varshni
The voltage noise can be converted to current noise by using equation [31]
the following formula: T 2 αH
E g−GaN = E g (0) − (11)
Vnoise_m β+T
Inoise = (9)
S where E g (0) is energy bandgap of GaN at T = 0 K, and α H
where the sensitivity of the device S = 0.26 V/A/A. With and β are constants.
the typical measurement frequency range 2–12 Hz and the The bandgap of AlGaN is deduced from the commonly
measurement bandwidth of 10 Hz, the actual noise voltage known compositional dependence [32]
Vnoise = 23.7× 10−9 V. Accordingly, the current noise was E g−AlGaN = (1 − x)E g−GaN (T ) + x E g−GaN (T ) − bx(1 − x)
calculated to be Inoise = 9.11 × 10−8 A·A. This current noise (12)
is also considered as the measurement resolution of this current
sensor, i.e., the sensor cannot measure the current smaller than where b is the bowing parameter and x is the alu-
9.11 × 10−8 A·A. minum concentration. Through solving the Schrodinger
We characterized the operation of the sensor at elevated and Poisson equations by using the Silvaco Atlas soft-
temperatures. The permeability of the magnetic core Vitroperm ware [33], the temperature-dependent energy band diagram
typically changes by less than 5% in the temperature range of AlGaN/GaN heterostructure can be simulated, as shown in
from 20 ◦ C to 200 ◦ C [29]. The Hall voltage versus primary Figs. 6 and 7.
current characteristics in temperature ranges from 20 ◦ C to Fig. 6 shows the calculated temperature dependence of the
200 ◦ C were measured, as shown in Fig. 5. A sensitivity bandgap of GaN and AlGaN and the conduction band offset
of 0.26 V/A/A was monitored at 20 ◦ C then slightly increased of AlGaN/GaN heterostructure. The bandgaps of GaN and
to 0.27 V/A/A at 200 ◦ C, (see the inset in Fig. 5). This increase AlGaN are found to decrease with increasing temperature.
in sensitivity with increasing temperature can be explained by Besides the energy bandgaps, the conduction band offset of
the following equation: AlGaN/GaN heterostructure also drops with the temperature.
rH μo μi According to Fig. 7(a), the conduction and valence band
S = GH . (10) energies of both GaN and AlGaN layers increase with increas-
q N2DEG lm μo + l g μi
ing temperature, especially in the GaN layer. Compared with
The sensitivity of the device S depends on the geometrical the conduction band, the valence band shows a larger shift
correction factor G H , the Hall factor r H , and the sheet density with varying temperatures. Thus, the energy bandgaps of GaN
N2DEG . According to the layout of the Hall plate, as shown in and AlGaN layers decrease with increasing temperature. As a
Fig. 1(b), the contact size is significantly small compared to consequence, the depth of the quantum well at the AlGaN/GaN

Authorized licensed use limited to: Univ of Calif Santa Barbara. Downloaded on May 20,2021 at 17:06:00 UTC from IEEE Xplore. Restrictions apply.
NGUYEN et al.: AlGaN/GaN 2DEG FOR HIGHLY SENSITIVE AND HIGH-TEMPERATURE CURRENT SENSING 1499

Fig. 6. Calculated band profiles of GaN, AlGaN, and the conduction

band offset of the AlGaN/GaN heterostructure. Fig. 8. 2DEG sheet density of AlGaN/GaN sample as a function of
temperature from RT to 200 ◦ C.

of 2DEG falls off with the rising of temperatures, as shown

in Fig. 7(c).
In addition, the 2DEG concentration in AlGaN/GaN het-
erostructures shows a direct proportional relationship to
the conduction band offset, according to the following
equation [34]:
σ ε0 ε
ns = −( )[eφb + E F − E c ] (13)
e dAlGaN e2
where σ is the total bound sheet charge, ε is the relative
dielectric constant of AlGaN, dAlGaN is the thickness of the
AlGaN layer, eφb is the Schottky barriers of the gate contact
on top of the AlGaN layer, E F is the Fermi level with
respect to the GaN conduction band-edge energy, and E c
is the conduction band offset at the AlGaN/GaN interface.
Thus, the sheet density of 2DEG decreases with increasing
temperature due to the reduction of conduction band offset
E c . To characterize 2DEG sheet density of the chip at
different temperatures, we put the chip and permanent magnet
in a furnace. The Hall voltage VH was recorded at a tem-
perature range from 20 ◦ C to 200 ◦ C. Hall supply current
I S was supplied and kept constant at 0.5 mA throughout the
experiment. The magnetic flux density B was measured at
room temperature using Magnetometer Koshava 5. For higher
temperature, B was calculated using the performance loss per
degree rise in temperature factor %/ ◦ C. Based on the recorded
data of VH , I S , and B at specific temperature points, we
obtained the 2DEG sheet density N at temperature range 20 ◦ C
to 200 ◦ C. The simulation result of temperature-dependent
sheet density agreed well with the experimental result shown
in Fig. 8. According to (10), the reduction in sheet density
leading to a slight increment of device sensitivity.
Fig. 7. (a) Calculated band profiles. (b) Conduction band of AlGaN/GaN
Koide et al. [18] and Alpert et al. [35] already reported
heterostructure. (c) The 2DEG distributions in an AlGaN/GaN het- that AlGaN/GaN Hall effect sensors can operate from 0 ◦ C up
erostructure at 25 ◦ C, 100 ◦ C, and 200 ◦ C. Inset: practical and theoretical to 570 ◦ C without dropping its current-scale sensitivity [13].
sheet density at various temperatures.
Therefore, the temperature range of the AlGaN/GaN 2DEG
current sensing system can be expected to be 0 ◦ C–570 ◦ C.
interface becomes shallower and the confinement of electrons However, the system also consists of other components, such
in the well reduces [see Fig. 7(b)]. Therefore, the sheet density as a magnetic core. The characterization of the system above

Authorized licensed use limited to: Univ of Calif Santa Barbara. Downloaded on May 20,2021 at 17:06:00 UTC from IEEE Xplore. Restrictions apply.
1500 IEEE TRANSACTIONS ON ELECTRON DEVICES, VOL. 68, NO. 4, APRIL 2021

200 ◦ C will be conducted in the future work for confirmation [14] TLI4971 High Precision Coreless Current Sensor for Industrial Applica-
of the predicted value. tions in 8×8mm SMD Package, TLI4971 Datasheet, Infineon, Neubiberg,
Germany, Mar. 2020.
[15] U. Ausserlechner, “The signal-to-noise ratio and a hidden symmetry
IV. C ONCLUSION of Hall plates,” Solid-State Electron., vol. 135, pp. 14–23, Sep. 2017,
We successfully demonstrated an AlGaN/GaN heterostruc- doi: 10.1016/j.sse.2017.06.007.
ture 2DEG Hall device for highly sensitive dc current sensing. [16] E. Ramsden, Hall-Effect Sensors: Theory and Applications. Burlington,
MA, USA: Newnes, 2006, pp. 1–10, doi: 10.1016/B978-0-7506-7934-
High linearity of the Hall voltage versus primary current was 3.X5000-5.
observed at different temperatures between 20 ◦ C and 200 ◦ C. [17] H. Lu, P. Sandvik, A. Vertiatchikh, J. Tucker, and A. Elasser, “High
The device exhibited a high sensitivity of 0.26 (V/A)/A with temperature Hall effect sensors based on AlGaN/GaN heterojunctions,”
J. Appl. Phys., vol. 99, no. 11, Jun. 2006, Art. no. 114510, doi: 10.
good repeatability, showing its feasibility for current sensing 1063/1.2201339.
applications. The magnetic core was used to concentrate the [18] S. Koide, H. Takahashi, A. Abderrahmane, I. Shibasaki, and A. Sandhu,
magnetic field generated from the electric current in the con- “High temperature Hall sensors using AlGaN/GaN HEMT structures,”
J. Phys. Conf. Ser., vol. 352, no. 1, pp. 3–7, 2012, doi: 10.1088/1742-
ductor. By using the magnetic field concentrator, the sensitivity 6596/352/1/012009.
of the Hall effect sensor increased significantly compared to [19] K. M. Dowling, “The effect of bias conditions on AlGaN/GaN 2DEG
the case without the magnetic core. The device showed a Hall plates,” in Proc. Solid-State Sens., Actuators Microsyst. Workshop,
2018, pp. 194–197, doi: 10.31438/trf.hh2018.54.
slight increase in sensitivity with increasing temperature due to [20] T. Hara, M. Mihara, N. Toyoda, and M. Zama, “Highly linear GaAs Hall
the conduction band offset lowering. Combining these factors, devices fabricated by ion implantation,” IEEE Trans. Electron Devices,
the AlGaN/GaN 2DEG Hall sensor is promising for current vol. ED-29, no. 1, pp. 78–82, Jan. 1982, doi: 10.1109/T-ED.1982.
20661.
monitoring in a wide range of temperatures. [21] S. R. I. Hout and S. Middelhoek, “A 400◦ C silicon Hall sensor,” Sens.
Actuators, A. Phys., vol. 60, nos. 1–3, pp. 14–22, 1999, doi: 10.1016/
R EFERENCES S0924-4247(96)01412-4.
[1] J. Webster and H. Eren, Measurement, Instrumentation, and Sensors [22] J. Jankowski, S. El-Ahmar, and M. Oszwaldowski, “Hall sensors for
Handbook: Spatial, Mechanical, Thermal, and Radiation Measurement. extreme temperatures,” Sensors, vol. 11, no. 1, pp. 876–885, Jan. 2011,
Boca Raton, FL, USA: CRC Press, 2014, doi: 10.1201/b15474. doi: 10.3390/s110100876.
[2] P. S. Filipski, M. Boecker, and M. Garcocz, “20-A to 100-A AC– [23] A. Bermudez, D. Gomez, and P. Salgado, Mathematical Models and
DC coaxial current shunts for 100-kHz frequency range,” IEEE Trans. Numerical Simulation in Electromagnetism, vol. 74. Cham, Switzerland:
Instrum. Meas., vol. 57, no. 8, pp. 1637–1641, Aug. 2008, doi: 10. Springer, 2014, pp. 151–182, doi: 10.1007/978-3-319-02949-8.
1109/TIM.2008.923783. [24] J. R. Barnes, Robust Electronic Design Reference Book. Boston, MA,
[3] P. S. Filipski and M. Boecker, “AC–DC current shunts and system for USA: Springer, 2004, pp. 126–155, doi: 10.1007/1-4020-7830-7.
extended current and frequency ranges,” IEEE Trans. Instrum. Meas., [25] T. P. White, S. Shetty, M. E. Ware, H. A. Mantooth, and G. J. Salamo,
vol. 55, no. 4, pp. 1222–1227, Aug. 2006, doi: 10.1109/TIM.2006. “AlGaN/GaN micro-Hall effect devices for simultaneous current and
876535. temperature measurements from line currents,” IEEE Sensors J., vol. 18,
[4] D. A. Ward and J. L. T. Exon, “Using Rogowski coils for transient no. 7, pp. 2944–2951, Apr. 2018, doi: 10.1109/JSEN.2018.2794264.
current measurements,” Eng. Sci. Edu. J., vol. 2, no. 3, p. 105, 1993, [26] C. Sun, “Self-contained wireless Hall current sensor applied for two-
doi: 10.1049/esej:19930034. wire zip-cords,” IEEE Trans. Magn., vol. 52, no. 7, pp. 22–25, Jul. 2016,
[5] S. Ziegler, R. C. Woodward, H. H.-C. Iu, and L. J. Borle, “Cur- doi: 10.1109/TMAG.2016.2524671.
rent sensing techniques: A review,” IEEE Sensors J., vol. 9, no. 4, [27] F. N. Hooge, “1/f noise is no surface effect,” Phys. Lett. A, vol. 29,
pp. 354–376, Apr. 2009, doi: 10.1109/JSEN.2009.2013914. no. 3, pp. 139–140, Apr. 1969, doi: 10.1016/0375-9601(69)90076-0.
[6] D. E. Shepard and D. W. Yauch, “An overview of Rogowski coil current [28] M. E. Levinshtein, M. S. Shur, S. L. Rumyantsev, M. A. Khan, and
sensing technology,” in Proc. LEM High Current Syst., 2000, p. 13. R. Gaska, “Low frequency and 1/f noise in wide-gap semiconductors:
[7] L. A. Kojovic and R. Beresh, “Practical aspects of Rogowski Silicon carbide and gallium nitride,” IEE Proc.-Circuits, Devices Syst.,
coil applications to relaying,” in Proc. IEEE PSRC Spectr. Rep., vol. 149, no. 1, pp. 32–39, Feb. 2002, doi: 10.1049/ip-cds:20020328.
2010, pp. 1–72. [Online]. Available: https://www.pes-psrc.org/kb/ [29] Nanocrystalline Vitroperm EMC Products, Vacuumschmelze GmbH co.
published/reports/Practical%20Aspects%20of%20Rogowski%20Coil% KG, Hanau, Germany, 2010.
20Applications%20to%20Relaying_Final.pdf [30] R. S. Popovic, Hall Effect Devices. Bristol, U.K.: IOP, 2003,
[8] A. Cataliotti, D. D. Cara, A. E. Emanuel, S. Nuccio, and G. Tine, pp. 159–161, doi: 10.1201/NOE0750308557.
“Characterization and error compensation of a Rogowski coil in the [31] Y. P. Varshni, “Temperature dependence of the energy gap in semicon-
presence of harmonics,” IEEE Trans. Instrum. Meas., vol. 60, no. 4, ductors,” Physica, vol. 34, no. 1, pp. 149–154, 1967, doi: 10.1016/0031-
pp. 1175–1181, Apr. 2011, doi: 10.1109/TIM.2010.2090694. 8914(67)90062-6.
[9] J. Jiang, “CMOS wide-bandwidth magnetic sensors for contactless [32] N. Nepal, J. Li, M. L. Nakarmi, J. Y. Lin, and H. X. Jiang, “Temperature
current measurements,” Ph.D. dissertation, Electron. Instrum. and compositional dependence of the energy band gap of AlGaN alloys,”
Lab., Delft Univ. Technol., Delft, The Netherlands, 2019, Appl. Phys. Lett., vol. 87, no. 24, Dec. 2005, Art. no. 242104, doi: 10.
doi: 10.4233/uuid:38dfa450-82c2-4b63-92d1-26b2eb3adeed. 1063/1.2142333.
[10] A. Girgin, M. Bilmez, H. Y. Amin, and T. C. Karalar, “A silicon Hall [33] A. Abou-Elnour and K. Schuenemann, “A comparison between different
sensor SoC for current sensors,” Microelectron. J., vol. 90, pp. 12–18, numerical methods used to solve Poisson’s and Schroedinger’s equations
Aug. 2019, doi: 10.1016/j.mejo.2019.04.020. in semiconductor heterostructures,” J. Appl. Phys., vol. 74, no. 5,
[11] T. Funk and B. Wicht, “Hall current sensor front-end,” in Integrated pp. 3273–3276, Sep. 1993, doi: 10.1063/1.354548.
Wide-Bandwidth Current Sensing. Cham, Switzerland: Springer, 2020, [34] O. Ambacher et al., “Two-dimensional electron gases induced by
pp. 117–124, doi: 10.1007/978-3-030-53250-5_6. spontaneous and piezoelectric polarization charges in N- and Ga-
[12] MLX91210 Integrated Current Sensor IC, 3901091210 Datasheet, face AlGaN/GaN heterostructures,” J. Appl. Phys., vol. 85, no. 6,
Melexis, Ypres, Belgium, Aug. 2017. pp. 3222–3233, Mar. 1999, doi: 10.1063/1.369664.
[13] Fully Integrated, Hall-Effect-Based Linear Current Sensor IC With 2.1 [35] H. S. Alpert et al., “Sensitivity of 2DEG-based Hall-effect sensors
kVRMS Isolation and a Low-Resistance Current Conductor, ACS712 at high temperatures,” Rev. Sci. Instrum., vol. 91, no. 2, Feb. 2020,
Datasheet, Allegro Microsyst., Manchester, NH, USA, Jan. 2020. Art. no. 025003, doi: 10.1063/1.5139911.

Authorized licensed use limited to: Univ of Calif Santa Barbara. Downloaded on May 20,2021 at 17:06:00 UTC from IEEE Xplore. Restrictions apply.