sensors

Article

Influences of an Aluminum Covering Layer on the

Performance of Cross-Like Hall Devices
Fei Lyu 1 , Xinfu Liu 2 , Yinjie Ding 2 , Eng-Huat Toh 2 , Zhenyan Zhang 1 , Yifan Pan 3 , Zhen Wang 1 ,
Chengjie Li 1 , Li Li 1 , Jin Sha 1 and Hongbing Pan 1, *
Received: 1 December 2015; Accepted: 12 January 2016; Published: 15 January 2016
Academic Editor: Vittorio M. N. Passaro
1

2
3

School of Electronic Science & Engineering, Nanjing University, Nanjing 210093, China;
lvfeieric@gmail.com (F.L.); zhangzhenyan518@163.com (Z.Z.); wzwcwz@foxmail.com (Z.W.);
lichengjie.520@163.com (C.L.); lili@nju.edu.cn (L.L.); shajin@nju.edu.cn (J.S.)
Globalfoundries Singapore, Singapore 738406, Singapore; XinFu.LIU@globalfoundries.com (X.L.);
YINJIE.DING@globalfoundries.com (Y.D.); EngHuat.TOH@globalfoundries.com (E.-H.T.)
Nanjing Foreign Language School, Nanjing 210008, China; panyifanhappy@163.com
Correspondence: phb@nju.edu.cn; Tel.: +86-189-5167-9279; Fax: +86-25-8968-4762

Abstract: This work studies the effects of an aluminum covering on the performance of cross-like
Hall devices. Four different Hall sensor structures of various sizes were designed and fabricated.
The sensitivity and offset of the Hall sensors, two key points impacting their performance, were
characterized using a self-built measurement system. The work analyzes the influences of the
aluminum covering on those two aspects of the performance. The aluminum layer covering mainly
leads to an eddy-current effect in an unstable magnetic field and an additional depletion region
above the active region. Those two points have influences on the sensitivity and the offset voltage,
respectively. The analysis guides the designer whether to choose covering with an aluminum layer
the active region of the Hall sensor as a method to reduce the flicker noise and to improve the stability
of the Hall sensor. Because Hall devices, as a reference element, always suffer from a large dispersion,
improving their stability is a crucial issue.
Keywords: cross-like Hall sensor; sensitivity; offset voltage; aluminum covering

1. Introduction
In recent years, Hall effect sensors inducting magnetic fields have become increasingly
indispensable in various fields, such as the manufacturing industry, consumer electronics, aerospace
industry, and so on [13]. As a typical Hall device implementation, cross-shaped Hall sensors are
considered more reliable, agile, miniaturizable and economical.
The sensitivity and offset are two important performance indicators of a Hall device. Hall sensors
based on silicon processes always offer low sensitivities and suffer from large offsets. Among the
impacting factors, the geometry has a significant effect on the sensitivity and offset [49]. In the past
decade, Hall devices with various symmetrical geometries such as circle, square, octagon, cross-like,
etc. have been designed and analyzed [10,11]. Hall sensors in a cross-like structure, with their high
sensitivity and low offset, are considered to have the most bright prospects. Recently, various new
Hall sensors structures were presented [1217]. Most of the new ones are variants of the cross-like Hall
sensors. For a cross-like Hall sensor, the most frequently used method in the design of the sensor to
reduce the flicker noise [18] and to improve the stability of the Hall sensor [19] is by covering a P-type
region or a metal layer on the active region of the Hall sensor. Our previous work has analyzed the
sensitivity and offset of Hall sensors using the P-type region [20]. The intention of the present paper is
to study the effects of the covering metal layer on the sensitivity and offset of the cross-like Hall sensor.

Sensors 2016, 16, 106; doi:10.3390/s16010106

www.mdpi.com/journal/sensors

Sensors 2016, 16, 106

cross-like Hall sensor. There is a silicon dioxide layer between the covering metal and the active
region and the covering metal must be connected to the ground while the Hall sensor is in operation.
Due Sensors
to the2016,
voltage
between the metal and the active region of Hall sensor, a depletion region
is
16, 106
2 of 9
formed under the silicon dioxide layer. This depletion region can effectively reduce flicker noise and
improve the stability of the Hall sensor. Moreover, the metal layer has an influence on the sensitivity
There is a silicon dioxide layer between the covering metal and the active region and the covering
and offset.
In this work, we use aluminum as the metal layer and the sensitivity and the offset of the
metal must be connected to the ground while the Hall sensor is in operation. Due to the voltage
cross-like sensor caused by the aluminum covering are studied in detail. We designed twenty Hall
between the metal and the active region of Hall sensor, a depletion region is formed under the silicon
cells dioxide
in four layer.
Hall chips
with different structures. Section 2 introduces those designs, the test equipment
This depletion region can effectively reduce flicker noise and improve the stability of
and the
measurement
data.
Section
3 analyzes
from two
the Hall sensor. Moreover,
the metal
layer hasthe
anmeasurement
influence on thedata
sensitivity
andaspects:
offset. InThe
thissensitivity
work,
and offset.
Section
4
summarizes
all
the
work
and
gives
the
conclusions.
we use aluminum as the metal layer and the sensitivity and the offset of the cross-like sensor caused
by the aluminum covering are studied in detail. We designed twenty Hall cells in four Hall chips

2. Design
and Experiment
with different
structures. Section 2 introduces those designs, the test equipment and the measurement
data. Section 3 analyzes the measurement data from two aspects: The sensitivity and offset. Section 4
Hall devices of four different structures have been designed and integrated in a 0.18 m BCD
summarizes all the work and gives the conclusions.
TM

lite technology provided by Globalfoundries. Existing implants are employed to form the Hall
devices
without
additional
processes. Five Hall cells with the same structure and in different sizes
2. Design
and
Experiment
are implemented
in one Hall chip. Figure 1 illustrates the structures of the Hall devices we designed.
Hall devices of four different structures have been designed and integrated in a 0.18 m BCD
These
four
structures
all consistbyofGlobalfoundries.
an N-type region
in a P-substrate,
four
contactstoand
a silicon
liteTM technology provided
Existing
implants are
employed
form
the Halloxide
layer.devices
The N-type
is used
as the active
zone
ofwith
the Hall
sensor.
The four
contacts
aresizes
formed
withoutregion
additional
processes.
Five Hall
cells
the same
structure
and in
different
by N+
at four the
branches
of the
Hall
device.
The
biases
areimplantation
implemented symmetrically
in one Hall chip.distributed
Figure 1 illustrates
structures
of the
Hall
devices
wecurrent
designed.
Hall These
sensor
bystructures
one pair all
of consist
opposite
contacts
and the
pair serve
the measurement
contacts.
four
of an
N-type region
in other
a P-substrate,
fouras
contacts
and a silicon oxide
layer. The
N-typeof
region
is used
as the active
of the
Hall sensor.
Theare
fourNWELL,
contacts are
formed
by
The N-type
regions
the first
structure
(S1) zone
and the
second
one (S2)
which
is formed
N+
implantation
symmetrically
distributed
at
four
branches
of
the
Hall
device.
The
current
biases
in the drain extension step of a medium voltage n-channel device. The third (S3) and the fourth
Hall sensor
by one NWELL
pair of opposite
contactsthe
andprocess
the other
as the
measurement
contacts.
structure
(S4) replace
by MVNVT,
ofpair
lowserve
voltage
p-channel
device
substrate.
The N-type regions of the first structure (S1) and the second one (S2) are NWELL, which is formed in
MVNVT is an N-type region with different doping profile of the NWELL and its doping profile is
the drain extension step of a medium voltage n-channel device. The third (S3) and the fourth structure
shown in Figure 2. Globalfoundries provided the doping profiles, which are obtained from the
(S4) replace NWELL by MVNVT, the process of low voltage p-channel device substrate. MVNVT
simulations using SILVACO Technology Computer Aided Design (TCAD). In addition, there is no
is an N-type region with different doping profile of the NWELL and its doping profile is shown in
special
process
above the silica
layer the
in S1
and S3.
However,
S4arefrom
covered
with an aluminum
Figure
2. Globalfoundries
provided
doping
profiles,
which S2
are and
obtained
the simulations
using
layerSILVACO
placed above
the silica
layer. The
silica
layer
is usedIntoaddition,
isolate the
N-type
region
Technology
Computer
Aided
Design
(TCAD).
theremetal
is no layer
specialand
process
above
in order
to eliminate
the
leakage
current.
aluminum
to layer
the ground
can generate
the silica
layer in S1
and
S3. However,
S2 The
and S4are
coveredlayer
withconnected
an aluminum
placed above
the
a depletion
region.
silica layer.
The silica layer is used to isolate the metal layer and N-type region in order to eliminate
the
current. equipment
The aluminum
layer
to the ground
can generate
a depletion
Theleakage
measurement
used
hasconnected
been introduced
in another
recent paper
of theregion.
authors [20]
The
measurement
equipment
used
has
been
introduced
in
another
recent
paper
of
the
authors
[20] Inc.,
and is shown in Figure 3. An Eastchanging 50110 instrument (East Changing Technologies,
and China)
is shownisinresponsible
Figure 3. An Eastchanging
instrument
(East
Changing Technologies,
Inc., Beijing,
Beijing,
for applying50110
a current
to the
electromagnet
and a magnetic
field we
China)
is
responsible
for
applying
a
current
to
the
electromagnet
and
a
magnetic
field
we
need
appears
need appears at the center of electromagnet. The Hall device is put at the center of electromagnet
at the center of electromagnet. The Hall device is put at the center of electromagnet perpendicular
perpendicular to the magnetic field. A Keithley 6220 device (Keithley Instruments, Inc., Cleveland,
to the magnetic field. A Keithley 6220 device (Keithley Instruments, Inc., Cleveland, OH, USA) is
OH, USA) is used to bias the Hall device and a Keithley 2182A (Keithley Instruments, Inc., Cleveland,
used to bias the Hall device and a Keithley 2182A (Keithley Instruments, Inc., Cleveland, OH, USA)
OH, feeds
USA)back
feeds
back
the voltage
two measurement
contacts.
Attemperature,
the room temperature,
the
voltage
betweenbetween
two measurement
contacts. At
the room
the biasing the
biasing
current
is
0.1
mA
and
it
remains
unchanged.
The
value
of
magnetic
field
sampling
50 points
current is 0.1 mA and it remains unchanged. The value of magnetic field sampling 50 points varies
varies
from
1T
to 1T.
from
1T
to 1T.

Figure 1. Schematics of the three-dimensional view (a) and sectional view (b) of the designed
Figure 1. Schematics of the three-dimensional view (a) and sectional view (b) of the designed structures
structures
Hall devices.
are and
the width
lengthofand
S2 andbyS4
covered by
of Hallof
devices.
L and W L
areand
the W
length
Hallwidth
device.ofS2Hall
and device.
S4 are covered
anare
aluminum
an aluminum
layer,
but
S1
and
S3
are
not.
The
N-type
region
of
both
S1
and
S2
is
NWELL
and
S3 and
layer, but S1 and S3 are not. The N-type region of both S1 and S2 is NWELL and S3 and S4 use MVNVT
S4 use
the N-type region.
as MVNVT
the N-typeasregion.

Sensors 2016,
2016, 16, 106
106
Sensors
Sensors 2016, 16,
16, 106

3 of 9

Figure
2.The
Thedoping
dopingprofiles
profilesof
Figure 2.
2.The
doping
profiles
of NWELL
NWELL and
and MVNVT.
MVNVT.

Figure 3.
3. Self-built measurement
measurement system.
Figure
Figure 3. Self-built
Self-built measurement system.
system.

Resultsand
andDiscussion
Discussion
3.
3. Results
Results
and
Discussion
3.1. Hall
Hall Device
3.1.
3.1. Hall Device
Device
When aa current
current I is supplied
supplied through two
two biasing contacts
contacts and the
the Hall device
device is put
put in an
an
When
When a current II is
is supplied through
through two biasing
biasing contacts and
and the Hall
Hall device is
is put in
in an
orthogonal
magnetic
field
B,
the
voltage
V
appears
between
the
other
contacts.
This
voltage
contains
orthogonal
orthogonal magnetic
magnetic field
field B,
B, the
the voltage
voltage V
V appears
appears between
between the
the other
other contacts.
contacts. This
This voltage
voltage contains
contains
two parts:
parts: Hall
Hall voltage
voltage V
VHHand
andoffset
offsetvoltage
voltageVVoffo.f Itf . can
It can
be
expressed
by:
two
be
expressed
by:
two parts: Hall voltage VH and offset voltage Voff. It can be expressed by:

IB VV
Voffoff
VVVSSSI IIB
of f
I IB`

(1)
(1)
(1)

where
SSII is
where S
is the
the current-related
current-related sensitivity.
sensitivity. The
The measured
measured voltage
voltage is
is this
this kind
kind voltage
voltage expressed
expressed in
in
where
I is the current-related sensitivity. The measured voltage is this kind voltage expressed in
above
equation
and
illustrated
in
Figure
4.
The
current-related
sensitivity
S
I
of
Hall
device
has
the
above equation
equation and
and illustrated
illustrated in
in Figure
Figure 4.
4. The
The current-related
current-related sensitivity
sensitivity SSII of
Hall device
device has
has the
the
above
of Hall
following
analytical
expression
[21]:
following
analytical
expression
[21]:
following analytical expression [21]:
Gr H
GrH
SSI Gr
(2)
SII nqteHf f
(2)
(2)
nqt

nqteffeff

Here, G is the correction factor depending on the geometry, r H is the Hall scattering factor, te f f
Here, G
the
factor
depending
on
geometry,
rrHH is
Hall
scattering factor,
tteffeff is
G is
isdepth
the correction
correction
factorregion,
depending
on the
the
geometry,
is the
theof
Hall
factor,
is
is theHere,
effective
of the N-type
n is the
doping
concentration
thescattering
active region
of Hall
the
effective
depth
of
the
N-type
region,
nn is
the
doping
concentration
of
the
active
region
of
Hall
the
effective
depth
of
the
N-type
region,
is
the
doping
concentration
of
the
active
region
of
Hall
device and q is the elementary charge of an electron. L/W ratio that designers have most control of
device
and qq is
elementary
charge
of
L/W ratio
that
designers
have
most
device
is the
theinfluence
elementary
charge
of an
an electron.
electron.
thatin
designers
have[10,11].
most control
control of
of
acts
as aand
common
factor
of geometry
and hasL/W
beenratio
studied
a few papers
acts
acts as
as aa common
common influence
influence factor
factor of
of geometry
geometry and
and has
has been
been studied
studied in
in aa few
few papers
papers [10,11].
[10,11].

33

Sensors 2016, 16, 106

Sensors 2016, 16, 106

4 of 9

(a)

(b)

(c)

(d)

Figure 4. The measured Hall voltage (coupled with offset) as a function of applied magnetic field for
Figure 4. The measured Hall voltage (coupled with offset) as a function of applied magnetic field for
S1 (a), S2 (b), S3 (c), S4 (d) with different L/W ratios.
S1 (a), S2 (b), S3 (c), S4 (d) with different L/W ratios.

As a reference component, it is taken for granted that the Hall device has excellent stability and
As a reference
component,
it isthe
taken
for granted
that the
Hall CMOS
device has
stabilitysuffer
and
low dispersion.
However,
because
uncertainty
of some
factors,
Hallexcellent
devices always
low
uncertainty
ofthe
some
factors, CMOS
Hallsurface
devices
always
suffer
fromdispersion.
dispersionHowever,
caused bybecause
variousthe
reasons,
such as
piezo-Hall
effect and
effects.
There
is
from
dispersion
caused
by
various
reasons,
such
as
the
piezo-Hall
effect
and
surface
effects.
There
is
always an oxide layer covering the Hall device. The oxide layer contains various amounts of charge
always
oxideinduce
layer charges
coveringofthe
Hall device.
oxide
layer
contains
charge5.
carriers,an
which
opposite
polarityThe
in the
Hall
device's
activevarious
zone asamounts
shown inofFigure
carriers,
which of
induce
charges
of opposite
in the
activecharacteristics
zone as shownisincalled
Figurethe
5.
The influence
the charge
carriers
in thepolarity
oxide layer
onHall
the device's
Hall sensors
The
influence
of
the
charge
carriers
in
the
oxide
layer
on
the
Hall
sensors
characteristics
is
called
the
surface effect. The surface effect impact on the current-related sensitivity and offset by the following
surface
effect.
equations
[2]: The surface effect impact on the current-related sensitivity and offset by the following
equations [2]:
SI I
S
QSS
SSI IQ
(3)
(3)
SSI I
S
Bo f f SII QS
Boff H QS

(4)
(4)

Here S I is an instable part of the current-related sensitivity depending on the surface effect, QS
SI isin
anthe
instable
partcharge
of the current-related
depending
surface
effect, Q
is theHere
variation
interface
density, Bo f f issensitivity
an additional
part of on
thethe
offset
depending
onS
is
the
variation
in
the
interface
charge
density,
B
off
is
an
additional
part
of
the
offset
depending
on
the surface effect, and H is the Hall mobility. In order to improve the stability of a CMOS Hall device,
surface
and the
H isdispersion
the Hall mobility.
In surface
order toeffect
improve
a CMOSactive
Hall device,
athe
method
toeffect,
eliminate
caused by
is tothe
getstability
the Hallofdevices
region
a method
eliminate
theintroducing
dispersion caused
by surface
is to get
Hall devices
activea region
away
fromto
the
surface by
a depletion
region.effect
Improving
thethe
stability
by covering
metal
away
from
the
surface
by
introducing
a
depletion
region.
Improving
the
stability
by
covering
a
metal
layer or a P-type region above the Hall device is very effective for this purpose.
layer or a P-type region above the Hall device is very effective for this purpose.

Sensors 2016, 16, 106

Sensors 2016, 16, 106
Sensors 2016, 16, 106

5 of 9

Figure 5.
5.The
effect on
a CMOS Hall
device.
Figure
The surface
surface
Figure 5.The
surface effect
effect on
on aa CMOS
CMOS Hall
Hall device.
device.

3.2.
Current-Related
Sensitivity
3.2. Current-Related
Current-RelatedSensitivity
Sensitivity
3.2.
According
to
Figure
6,
S2
and
S4
with
an
aluminum
covering
layer
have
worse
current-related
According to
toFigure
Figure6,
6,S2
S2and
andS4
S4with
withan
analuminum
aluminumcovering
coveringlayer
layerhave
haveworse
worsecurrent-related
current-related
According
sensitivities
compared
to
S1
and
S3.
Thanks
to
the
paramagnetism
of
aluminum
and
the
depletion
sensitivities compared
compared to
to S1
S1 and
andS3.
S3. Thanks
Thanks to
to the
theparamagnetism
paramagnetism of
ofaluminum
aluminum and
and the
thedepletion
depletion
sensitivities
region
between
the
silicon
dioxide
layer
and
the
active
area
of
the
Hall
sensor,
the
sensitivity
of
region between
between the silicon
thethe
Hall
sensor,
thethe
sensitivity
of the
the
region
silicon dioxide
dioxide layer
layerand
andthe
theactive
activearea
areaofof
Hall
sensor,
sensitivity
of
Hall
sensor
with
an
aluminum
layer
increases
slightly.
However
the
sensitivity
moves
in
the
opposite
HallHall
sensor
withwith
an aluminum
layer layer
increases
slightly.
However
the sensitivity
moves in
the opposite
the
sensor
an aluminum
increases
slightly.
However
the sensitivity
moves
in the
direction
because
of
current
effect
in
unstable
magnetic
field. According
to the
directiondirection
because because
of the
the eddy
eddy
current
effecteffect
in an
an
unstable
magnetic
the
opposite
of
the eddy
current
in an
unstable
magneticfield.
field.According
According to the
measurement
data,
the
negative
aspect
is
the
most
significant
of
the
impact
factors
caused
by
the
measurement
data,
the
negative
aspect
is
the
most
significant
of
the
impact
factors
caused
by
the
measurement data, the negative aspect is the most significant of the impact factors caused by the
covering
sensitivity of the Hall sensor
is reduced
by the
aluminum layer
in our study.
coveringaluminum,
aluminum,so
sothe
covering
aluminum,
so
the sensitivity of the Hall
Hall sensor
sensor is
isreduced
reducedby
bythe
thealuminum
aluminumlayer
layerininour
ourstudy.
study.

Figure 6. Relationship between L/W ratio and current-related sensitivity of S1, S2, S3 and S4.
Figure6.6.Relationship
Relationshipbetween
betweenL/W
L/W ratio
ratio and
and current-related
current-related sensitivity
sensitivity of
of S1,
S1, S2,
S2, S3
S3 and
and S4.
S4.
Figure

As
As aa paramagnetic
paramagnetic material,
material, when
when aluminum
aluminum is
is put
put in
in aa magnetic
magnetic field,
field, magnetization
magnetization appears
appears
As
a
paramagnetic
material,
when
aluminum
is
put
in
a
magnetic
field,
magnetization
appears
inside
inside and
and the
the intensity
intensity of
of magnetic
magnetic field
field increases
increases slightly.
slightly. Even
Even though
though there
there is
is no
no magnetic
magnetic field,
field,
inside
andmolecules
the intensity of magnetic
field increases slightly.
Even thoughmagnetic
there is no magnetic
atoms
atoms or
or molecules inside
inside the
the paramagnetic
paramagnetic materials
materials have
have permanent
permanent magnetic dipoles
dipoles [22].
[22].
field,
atoms
or thermal
molecules
inside the
paramagnetic
materialsoriented
have permanent
magnetic dipoles
[22].
Because
of
the
agitation,
the
dipoles
are
randomly
without
interaction
with
others
Because of the thermal agitation, the dipoles are randomly oriented without interaction with others
Because
of the thermal
agitation, thein
dipoles paramagnetic
are randomly oriented without
interaction
with others
in
in the
the absence
absence of
of an
an external
external field
field in pure
pure paramagnetic materials,
materials, so
so there
there is
is no
no magnetic
magnetic field
field
in
the absence
of anWhen
external
field in pure
paramagnetic
materials,
so there
is will
no magnetic
field
inside
the
material.
the
material
is
placed
in
a
magnetic
field,
the
dipoles
align
inside the material. When the material is placed in a magnetic field, the dipoles will align with
with the
the
inside
the
material.
When
the moment
material appears
is placedininthe
a magnetic
field,
the
dipolesfield.
will align
with the
applied
field
and
a
magnetic
direction
of
the
applied
Generally,
applied field and a magnetic moment appears in the direction of the applied field. Generally, the
applied
field and by
a magnetic moment
appearsquite
in the direction
of the applied
field. Generally,
the
influence
influence caused
caused by paramagnetic
paramagnetic property
property is
is quite small
small and
and for
for most
most paramagnets
paramagnets the
the magnetic
magnetic
3 to
5. Magnetic
influence
caused
by
paramagnetic
property
is
quite
small
and
for
most
paramagnets
the
magnetic
susceptibility
is
of
the
order
of
10
10
susceptibility
is
a
dimensionless
proportionality
3 to 105. Magnetic susceptibility is a dimensionless proportionality
susceptibility is of the order of 10
susceptibility
is indicates
of the order
of
103 to
105 . Magneticof
susceptibilityinisresponse
a dimensionless
proportionality
parameter
that
the
degree
parameter that indicates the degreeof
ofmagnetization
magnetization ofaamaterial
material in responseto
toan
anapplied
appliedmagnetic
magnetic
5. an
parameter
that indicatesthe
the degree ofsusceptibility
magnetization of aaluminum
material in response
to
applied
magnetic
field
current-related
5. The
field [23].
[23]. Otherwise,
Otherwise, the magnetic
magnetic susceptibility for
for aluminum is
is 2.2
2.2 10
10
The
current-related
field
[23]. Otherwise,slightly
the magnetic
susceptibility
for aluminum
is 2.2 105 . The
current-related
sensitivity
sensitivity increases
increases slightly due
due to
to the
the effect
effect caused
caused by
by the
the paramagnetism
paramagnetism of
of aluminum
aluminum to
to the
the
sensitivity
increases
slightlythe
due
to the effect
caused
by the
paramagnetism
of
aluminum
tolayer
the
magnetic
field.
Meanwhile,
depletion
region
[24]
caused
by
voltage
between
the
aluminum
magnetic field. Meanwhile, the depletion region [24] caused by voltage between the aluminum layer
magnetic
field. Meanwhile,
the depletion
region
[24] caused by voltage
between
the aluminum
layer
and
and N-well
N-well gives
gives rise
rise to
to the
the reduction
reduction of
of the
the effective
effective thickness
thickness of
of the
the active
active region
region of
of Hall
Hall sensor.
sensor.
and
N-well
gives
rise
to
the
reduction
of
the
effective
thickness
of
the
active
region
of
Hall
sensor.
According
According to
to Equation
Equation (2),
(2), the
the diminution
diminution of
of the
the thickness
thickness of
of the
the active
active region
region is
is beneficial
beneficial to
to the
the
According
to Equation
(2), the diminution of the thickness of the active region is beneficial to the
current-related
sensitivity.
current-related sensitivity.
current-related
sensitivity.
Furthermore,
Furthermore, aluminum
aluminum can
can weaken
weaken the
the magnetic
magnetic field
field because
because of
of the
the eddy-current
eddy-current effect
effect in
in an
an
unstable
unstable magnetic
magnetic field.
field. In
In our
our measurement
measurement system,
system, the
the magnetic
magnetic field
field cannot
cannot remain
remain absolutely
absolutely
stable.
stable. In
In other
other words,
words, the
the magnetic
magnetic field
field fluctuates
fluctuates in
in aa narrow
narrow range
range around
around the
the set
set point,
point, so
so the
the
power
of
the
magnetic
field
is
dissipated
as
a
result
of
eddy-current
loss
[25].
The
power
dissipation
power of the magnetic field is dissipated as a result of eddy-current loss [25]. The power dissipation
55

Sensors 2016, 16, 106

6 of 9

Furthermore, aluminum can weaken the magnetic field because of the eddy-current effect in an
unstable magnetic field. In our measurement system, the magnetic field cannot remain absolutely
stable.
Sensors
2016,In
16,other
106 words, the magnetic field fluctuates in a narrow range around the set point, so the
power of the magnetic field is dissipated as a result of eddy-current loss [25]. The power dissipation
eddy
currents
leads
reduction
sensitivity
of Hall
sensor.
In order
to verify
theory,
of of
eddy
currents
leads
to to
thethe
reduction
of of
thethe
sensitivity
of Hall
sensor.
In order
to verify
thisthis
theory,
Hall
device
biased
with
mA
and
put
center
electromagnet
providing
a 1T
thethe
Hall
device
is is
biased
with
0.10.1
mA
and
put
at at
thethe
center
of of
thethe
electromagnet
setset
providing
a 1T
magnetic
field
and
the
voltage
fed
back
every
0.1
s.
In
order
to
analyze
the
difference
between
two
magnetic field and the voltage fed back every 0.1 s. In order to analyze the difference between two
structures,
translate
measured
voltages
around
zero
and
two
groups
translated
voltages
structures,
wewe
translate
thethe
measured
voltages
around
zero
and
two
groups
of of
thethe
translated
voltages
two
corresponding
structures
illustrated
Figure
of of
two
corresponding
structures
areare
illustrated
inin
Figure
7. 7.

Figure
7. Pulses
generated
byby
thethe
electromagnet.
Figure
7. Pulses
generated
electromagnet.

It is obvious that the electromagnet generates several pulses randomly. In order to ensure that
It is obvious that the electromagnet generates several pulses randomly. In order to ensure that
the measurement data in Figure 4 is detected at the steady part of the wave, the voltmeter feeds back
the measurement data in Figure 4 is detected at the steady part of the wave, the voltmeter feeds back
several results at one point and the results are analyzed to select the most reasonable one. In Figure 7,
several results at one point and the results are analyzed to select the most reasonable one. In Figure 7,
Pulses 1 and 1' have a similar amplitude, as do Pulses 2 and 2. Pulses 1 and 2 coming from the wave
Pulses 1 and 1' have a similar amplitude, as do Pulses 2 and 2. Pulses 1 and 2 coming from the wave
of the structure without aluminum vanish more quickly than Pulses 1 and 2 taken from the other
of the structure without aluminum vanish more quickly than Pulses 1 and 2 taken from the other
structure. The eddy-current effect in aluminum slows the fading of the pulses. In addition to the pulses,
structure. The eddy-current effect in aluminum slows the fading of the pulses. In addition to the pulses,
the magnetic field fluctuates within a very narrow range. In contrast to the pulses, the fluctuation
the magnetic field fluctuates within a very narrow range. In contrast to the pulses, the fluctuation
exists in the whole measurement and is hardly discriminated in the results. This small range causes
exists in the whole measurement and is hardly discriminated in the results. This small range causes
the eddy-current loss, and the power dissipation of eddy currents can be expressed by [25]:
the eddy-current loss, and the power dissipation of eddy currents can be expressed by [25]:

2 BP d 2 f 2
2 B d2 f 2
P 6 DP

(5)

(5)
6D
Here Bp is the magnetic field we set, d is the thickness of the aluminum layer, f is the frequency
Here B p is the
field we
set,aluminum
d is the thickness
ofthe
thedensity
aluminum
layer, f is the
of the fluctuation,
ismagnetic
the resistivity
of the
and D is
of aluminum.
Thefrequency
values
of
the
fluctuation,

is
the
resistivity
of
the
aluminum
and
D
is
the
density
of
aluminum.
The values
of the parameters used in Equation (5) are listed in Table 1. From Equation (5), we can conclude
that
the parameters
used
in Equation
(5) areislisted
in Tableby
1. the
From
Equation
we can conclude
that
theofpower
dissipation
of the
eddy currents
not affected
range
of the(5),
fluctuation.
Although
the
power
dissipation
of
the
eddy
currents
is
not
affected
by
the
range
of
the
fluctuation.
Although
the fluctuation is very small, it can persistently generate power dissipation that we cannot ignore in
fluctuation is very small, it can persistently generate power dissipation that we cannot ignore in
thethe
results.
the results.
Table 1. Values of the parameters used in Equation (5).
Table 1. Values of the parameters used in Equation (5).

Bp (T)
0~1

D (m)
B p (T)
0~1

D (m)

2 106
2

106

(m)
P ( m)

2.83 108

2.83

108

D (kg/m3)
D (kg/m3 )

2.7 103

2.7 103

Consequently, the magnetic field applied on the active area suffers a loss when passing the
covering aluminum layer. The loss of the magnetic field leads to the decrease in current-related
sensitivity. Comparing the data of S1 and S3 or S2 and S4, it is found the measurement data
corresponds to the above analysis, so the Hall sensor with the aluminum covering layer suffers
eddy-current loss in an unstable magnetic field and has a smaller current-related sensitivity.
From the results of a recent paper [20], the P-type covering layer benefits the current-related

Sensors 2016, 16, 106

7 of 9

Consequently, the magnetic field applied on the active area suffers a loss when passing the
covering aluminum layer. The loss of the magnetic field leads to the decrease in current-related
sensitivity. Comparing the data of S1 and S3 or S2 and S4, it is found the measurement data corresponds
to the above analysis, so the Hall sensor with the aluminum covering layer suffers eddy-current loss in
an unstable magnetic field and has a smaller current-related sensitivity.
From the results of a recent paper [20], the P-type covering layer benefits the current-related
Sensors 2016, 16, 106
sensitivity,
because
Sensors 2016,
16, 106 of the decreased thickness of the active region, so in the aspect of the current-related
sensitivity,
thethe
aluminum
covering
toreduce
reducethe
theflicker
flickernoise
noise
and
sensitivity,
aluminum
coveringisisnot
notaarecommendable
recommendable method
method to
and
to to
sensitivity,
the
aluminum
covering
is
not
a
recommendable
method
to
reduce
the
flicker
noise
and
to
improve
the
stability
of
a
Hall
sensor.
improve the stability of a Hall sensor.
improve the stability of a Hall sensor.
3.3.3.3.
Offset
Voltage
Offset
Voltage
3.3. Offset Voltage
Hall
sensors
always
and drifting
driftingoffset,
offset,which
whichlimits
limitstheir
their
use
Hall
sensors
alwayssuffer
sufferfrom
fromaalarge,
large,unpredictable
unpredictable and
use
Hall
sensors
always
suffer
from
a
large,
unpredictable
and
drifting
offset,
which
limits
their
use
in sensing
low
magnetic
fields.
To
obtain
the
offset
of
a
Hall
device,
the
output
voltage
is
the
offset
in sensing low magnetic fields. To obtain the offset of a Hall device, the output voltage is the offset
in
sensing
low
fields.
To
obtain
theFigure
offset of
a Hall device,
thevoltages
output voltage
isL/W
theratio
offset
voltage
when
nomagnetic
magnetic
field
isispresent.
88 shows
the offset
voltages
versusL/W
ratio
voltage
when
no
magnetic
field
present.
Figure
shows
offset
versus
of of
voltage
when
no
magnetic
field
is
present.
Figure
8
shows
the
offset
voltages
versus
L/W
ratio
of
different
structures
and
the
offset
Hallcells
cellsare
arebiased
biasedbybythe
the
current
different
structures
and
the
offsetvoltages
voltagesare
aremeasured
measured when
when the Hall
current
structures
and
the
offset
voltages are
measured
Hall cells
by the
current
of different
0.1
mA.
It is
that
the
relationship
between
the
offset
voltage
andare
thebiased
L/Wratio
ratio
not
very
of 0.1
mA.
It found
is found
that
the
relationship
between
thewhen
offsetthe
voltage
and
the
L/W
is is
not
very
of
0.1 mA.
It is found
that
the relationship
between
the offset
voltage
and
the L/W
ratio
is not
very
clear.
If
covered
with
an
aluminum
layer,
Hall
sensors
with
the
same
N-type
region
have
a
poorer
clear. If covered with an aluminum layer, Hall sensors with the same N-type region have a poorer
clear.
If covered
with
an aluminum
layer, Hall sensors with the same N-type region have a poorer
performance
aspect
theoffset.
offset.
performance
in in
thethe
aspect
ofofthe
performance in the aspect of the offset.

Figure
8. (a)
Measured
offsetvoltage
voltageversus
versusthe
theL/W
L/W ratio
versus
Figure
8. (a)
Measured
offset
ratio of
of S1,
S1,S3;
S3;(b)
(b)Measured
Measuredoffset
offsetvoltage
voltage
versus
Figure
8. ratio
(a) Measured
L/W
andoffset
S4. voltage versus the L/W ratio of S1, S3; (b) Measured offset voltage versus
thethe
L/W
ratio ofofS2S2and
S4.
the L/W ratio of S2 and S4.

As shown in Figure 9, there is a reverse biased pn-junction between the P-SUB and the N-type
AsAs
shown
ininFigure
there
biased pn-junction
pn-junction betweenthe
the P-SUB
and
N-type
shown
Figure9,9,
thereisisaareverse
reverse
and
thethe
N-type
regions.
In other
words,
a depletion
region isbiased
formed because ofbetween
the voltage P-SUB
difference
between
the
regions.
InInother
words,
aadepletion
region
is formed
becauseofofthe
thevoltage
voltagedifference
difference
between
regions.
other
words,
depletion
region
formed
because
between
theisthe
substrate and the active region. This phenomenon redefines the bottom edge of Hall device and
substrate
the
active
region.
This
phenomenon
redefines
the
bottom
edge
ofof
Hall
device
and
is is
substrate
the
active
region.
This
phenomenon
redefines
the
bottomthe
edge
Hall
device
and
known and
asand
the
JFET
effect.
The JFET
effect
always exits
no matter
whether
active
region
of the
Hall
known
the
JFETeffect.
effect.
TheJFET
JFETor
effect
region
of of
thethe
Hall
known
asas
JFET
The
effect
exitsno
nomatter
matterwhether
whetherthe
theactive
active
region
Hall
sensor
isthe
covered
with
aluminum
not.always exits
sensor
coveredwith
withaluminum
aluminumor
ornot.
not.
sensor
is is
covered

Figure 9. The depletion regions in integrated Hall sensors covered by aluminum.t and Xj are the
Figure
9. thickness
The depletion
regions
in integrated
Hall sensors
covered by aluminum.t and Xj are the
effective
and the
total depth
of the N-type
region, respectively.
Figure
9. The
depletion
regions
in integrated
Hall sensors
covered by aluminum.t and X j are the
effective thickness and the total depth of the N-type region, respectively.
effective thickness and the total depth of the N-type region, respectively.

If the Hall sensor is covered with an aluminum layer and this aluminum layer is connected to
If thea Hall
sensor
is covered
with
an aluminum
and this
aluminum
layersensor
is connected
to
ground,
depletion
region
appears
between
SiO2 andlayer
the active
region
of the Hall
due to the
ground,
depletionbetween
region appears
betweenlayer
SiO2 and
of[25].
the Hall
due region
to the
voltage adifference
the aluminum
and the
the active
N-typeregion
region
Thissensor
depletion
voltage
difference
between
the
aluminum
layer
and
the
N-type
region
[25].
This
depletion
region
isolates the N-type region from SiO2 and forms a new top on the Hall devices active region. The two
isolates
thelayers
N-type
region
from
SiOthe
2 and forms a new top on the Hall devices active region. The two
depletion
above
and
under
active region could slightly decrease the effective depth of the
depletion
layers
above
and
under
the
region could
slightly decrease
thethe
effective
depth of
active region and are conducive to theactive
current-related
sensitivity.
However,
offset voltage
ofthe
the

Sensors 2016, 16, 106

8 of 9

If the Hall sensor is covered with an aluminum layer and this aluminum layer is connected
to ground, a depletion region appears between SiO2 and the active region of the Hall sensor due
to the voltage difference between the aluminum layer and the N-type region [25]. This depletion
region isolates the N-type region from SiO2 and forms a new top on the Hall devices active region.
The two depletion layers above and under the active region could slightly decrease the effective depth
of the active region and are conducive to the current-related sensitivity. However, the offset voltage of
the Hall sensors depends on the depth of the depletion region [26]. A greater depth of the depletion
region leads to a larger offset voltage. It is obvious that the aluminum layer leads to a additional
depletion region, which will increase the offset voltage, so S1 and S2 without covering aluminum are
much better than S3 and S4 in the offset voltage aspect. In the method mentioned in [20], a P-type
layer is formed above the N-type region. Therefore, one more process with the mask misalignment is
necessary in the active region. The mask misalignment that is inevitably introduced by the additional
process increases the offset. Moreover, the offset caused by mask misalignment is more serious than
that caused by the increase of the depletion regions thickness, so in the aspect of the offset, covering
with an aluminum layer is a more suitable method to reduce the flicker noise and to improve the
stability of the Hall sensor.
4. Conclusions
In order to analyze the influences of the covering aluminum on the current-related sensitivity and
the offset voltage of Hall sensors, four different Hall sensor structures are designed and fabricated in
0.18 m BCD liteTM technology provided by Globalfoundries. Each structure is implemented in a Hall
chip with multiple Hall cells. We found the impact on the current-related sensitivity caused by the
aluminum covering by comparing S1, S3, S2 and S4. In an unstable magnetic field, the Hall sensor
with the aluminum covering suffers a loss, which leads to a decrease of the current-related sensitivity.
Moreover, the influence of the aluminum layer on the offset voltage of the cross-like Hall sensor has
been studied. Due to the aluminum covering on the active region of the Hall sensor, a depletion region
appears between the N-type region and silicon oxide layer. This depletion layer leads to an increase
of the offset voltage. When designing a Hall device, a covering aluminum is not a perfect method to
reduce the flicker noise and to improve the stability of the Hall sensor. If the current-related sensitivity
is the main concern in the design, the method of covering with an aluminum layer the active layer
of the Hall sensor is not a good way, however, if the offset is the most important parameter of the
assessment criteria for the Hall devices, covering with an aluminum layer is a suitable method to
reduce the flicker noise and to improve the stability of the sensor.
Acknowledgments: This work was supported by National Nature Science Foundation of China under Grant
No. 61370040, 61376075, 61176024; Research Fund for the Doctoral Program of Higher Education of China under
Grant No. 20120091110029; the key Research and Development Program of Jiangsu Province under Grant No:
BE2015153; the project on the Integration of Industry, Education and Research of Jiangsu Province under Grant
BY2015069-05, BY2015069-08; A Project Funded by the Priority Academic Program Development of Jiangsu Higher
Education Institutions (PAPD); the Research Innovation Program for College Graduates of Jiangsu Province under
Grant No. KYZZ15_0029, SJZZ15_0015.
Author Contributions: Fei Lyu wrote and revised the manuscript, and also designed the Hall devices.
Eng-Huat Toh, Xinfu Liu and Yinjie Ding analyzed the measured data in aspect of technology process.
Zhenyan Zhang, Yifan Pan, Zhen Wang and Chengjie Li carried out the experiments. Li Li and Jin Sha gave a
keen insight in this manuscript. Hongbing Pan arranged all the work in this project.
Conflicts of Interest: The authors declare no conflict of interest.

References
1.
2.
3.

Ramsden, E. Hall-Effect Sensors: Theory and Application; Elsevier: Philadelphia, PA, USA, 2011.
Popovic, R.S. Hall Effect Devices; CRC Press: Boca Raton, FL, USA, 2003.
Boero, G.; Demierre, M.; Popovic, R. Micro-Hall devices: Performance, technologies and applications.
Sens. Actuators A Phys. 2003, 106, 314320. [CrossRef]

Sensors 2016, 16, 106

4.
5.
6.

7.
8.

9.

10.
11.
12.

13.
14.

15.

16.
17.
18.

19.
20.
21.
22.
23.
24.
25.
26.

9 of 9

Paun, M.-A.; Sallese, J.-M.; Kayal, M. Comparative study on the performance of five different Hall effect
devices. Sensors 2013, 13, 20932112. [CrossRef] [PubMed]
Paun, M.-A.; Sallese, J.-M.; Kayal, M. Offset and drift analysis of the Hall Effect sensors. The geometrical
parameters influence. Dig. J. Nanomater. Biostruct 2012, 7, 883891.
Paun, M.-A.; Sallese, J.-M.; Kayal, M. Geometrical Parameters Influence on the Hall Effect Sensors Offset and
Drift. In Proceedings of the 7th Conference on Ph.D. Research in Microelectronics and Electronics (PRIME),
Trento, Italy, 37 July 2011; pp. 145148.
Paun, M.-A.; Sallese, J.-M.; Kayal, M. Hall effect sensors design, integration and behavior analysis. J. Sens.
Actuator Netw. 2013, 2, 8597. [CrossRef]
Paun, M.-A.; Sallese, J.-M.; Kayal, M. A specific parameters analysis of CMOS Hall effect sensors with
various geometries. In Proceedings of the 19th International Conference Mixed Design of Integrated Circuits
and Systems (MIXDES), Warsaw, Poland, 2426 May 2012; pp. 335339.
Paun, M.-A.; Sallese, J.-M.; Kayal, M. Hall effect sensors performance investigation using three-dimensional
simulations. In Proceedings of the 18th International Conference on Mixed Design of Integrated Circuits and
Systems (MIXDES), Gliwice, Poland, 1618 June 2011; pp. 450455.
Versnel, W. Analysis of symmetrical Hall plates with finite contacts. J. Appl. Phys. 1981, 52, 46594666.
[CrossRef]
Versnel, W. The geometrical correction factor for a rectangular Hall plate. J. Appl. Phys. 1982, 53, 49804986.
[CrossRef]
Heidari, H.; Gatti, U.; Bonizzoni, E.; Maloberti, F. Low-noise low-Offset current-mode Hall sensors.
In Proceedings of the 9th Conference on Ph. D. Research in Microelectronics and Electronics (PRIME),
Villach, Austria, 2427 June 2013; pp. 325328.
Heidari, H.; Bonizzoni, E.; Gatti, U.; Maloberti, F. A CMOS Current-Mode Magnetic Hall Sensor with
Integrated Front-End. IEEE Trans. Circuits Syst. I Regul. Pap. 2015, 62, 12701278. [CrossRef]
Heidari, H.; Bonizzoni, E.; Gatti, U.; Maloberti, F. A 0.18-m CMOS Current-Mode Hall Magnetic Sensor
with Very Low Bias Current and High Sensitive Front-End. In Proceedings of the IEEE SENSORS, Valencia,
Spain, 25 November 2014; pp. 14671470.
Heidari, H.; Bonizzoni, E.; Gatti, U.; Maloberti, F.A. Current-Mode CMOS integrated microsystem for current
spinning magnetic hall sensors. In Proceedings of the IEEE International Symposium on Circuits and
Systems (ISCAS), Melbourne, Australia, 15 June 2014; pp. 678681.
Paun, M.-A. Three-dimensional simulations in optimal performance trial between two types of Hall sensors
fabrication technologies. J. Magn. Magn. Mater. 2015, 391, 122128. [CrossRef]
Huang, H.; Wang, D.; Xu, Y. A Monolithic CMOS Magnetic Hall Sensor with High Sensitivity and Linearity
Characteristics. Sensors 2015, 15, 2735927373. [CrossRef] [PubMed]
Schweda, J.; Riedling, K. A nonlinear simulation model of integrated Hall devices in CMOS silicon technology.
In Proceedings of the IEEE International Workshop on Behavioral Modeling and Simulation, Santa Rosa, CA,
USA, 68 October 2002; pp. 1420.
Popovic, R. Hall-effect devices. Sens. Actuators 1989, 17, 3953. [CrossRef]
Lyu, F.; Zhang, Z.; Toh, E.-H.; Liu, X.; Ding, Y.; Pan, Y.; Li, C.; Li, L.; Sha, J.; Pan, H. Performance Comparison
of Cross-Like Hall Plates with Different Covering Layers. Sensors 2014, 15, 672686. [CrossRef] [PubMed]
Xu, Y.; Pan, H.-B. An improved equivalent simulation model for CMOS integrated Hall plates. Sensors 2011,
11, 62846296. [CrossRef] [PubMed]
Miessler, G. DA Tarr Inorganic Chemistry; Prentice Hall: Upper Saddle River, NJ, USA, 2013.
Chakravarty, A.S. Introduction to the Magnetic Properties of Solids; John Wiley & Sons: Hoboken, NJ, USA, 1981.
Anderson, B.; Anderson, R. Fundamentals of Semiconductor Devices; McGraw-Hill: New York, NY, USA, 2005.
Fiorillo, F. Measurement and Characterization of Magnetic Materials; Elsevier: Amsterdam, Holland, 2004.
Bellekom, S. CMOS versus bipolar Hall plates regarding offset correction. Sens. Actuators A Phys. 1999, 76,
178182. [CrossRef]
2016 by the authors; licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons by Attribution
(CC-BY) license (http://creativecommons.org/licenses/by/4.0/).