Article

Study and Experiment on Non-Contact Voltage

Sensor Suitable for Three-Phase Transmission Line
Qiang Zhou 1, *, Wei He 1 , Dongping Xiao 1 , Songnong Li 2 and Kongjun Zhou 2
Received: 21 November 2015; Accepted: 25 December 2015; Published: 30 December 2015
Academic Editor: Vittorio M. N. Passaro
1 State Key Laboratory of Power Transmission Equipment & System Security and New Technology,
Chongqing University, Chongqing 400044, China; weihe2016@126.com (W.H.);
dongpingxiao2016@sina.com (D.X.)
2 State Grid Chongqing Electric Power CO. Electric Power Research Institute, Chongqing 400015, China;
songnongli2016@126.com (S.L.); kongjunzhou2016@sina.com (K.Z.)
* Correspondence: 20131101031@cqu.edu.cn; Tel.: +86-132-7192-0913

Abstract: A voltage transformer, as voltage signal detection equipment, plays an important role in
a power system. Presently, more and more electric power systems are adopting potential transformer
and capacitance voltage transformers. Transformers are often large in volume and heavyweight,
their insulation design is difficult, and an iron core or multi-grade capacitance voltage division
structure is generally adopted. As a result, the detection accuracy of transformer is reduced,
a huge phase difference exists between detection signal and voltage signal to be measured, and the
detection signal cannot accurately and timely reflect the change of conductor voltage signal to be
measured. By aiming at the current problems of electric transformation, based on electrostatic
induction principle, this paper designed a non-contact voltage sensor and gained detection signal
of the sensor through electrostatic coupling for the electric field generated by electric charges of the
conductor to be measured. The insulation structure design of the sensor is simple and its volume is
small; phase difference of sensor measurement is effectively reduced through optimization design of
the electrode; and voltage division ratio and measurement accuracy are increased. The voltage sensor
was tested on the experimental platform of simulating three-phase transmission line. According to the
result, the designed non-contact voltage sensor can realize accurate and real-time measurement for the
conductor voltage. It can be applied to online monitoring for the voltage of three-phase transmission
line or three-phase distribution network line, which is in accordance with the development direction
of the smart grid.

Keywords: voltage sensor; non-contact measurement; ansoft maxwell; electric field intensity;
three phase transmission lines

1. Introduction
The voltage transformer is an important part of an electric power system and the detection
performance of the transformer can directly influence the reliability of an electric power system
voltage measurement and relay protection device motion as well as the accuracy of electric power
measurement. Presently, voltage transformers extensively applied to electric power system mainly
include potential transformer and capacitance voltage transformer [1–5].
Structure diagram of Capacitance Voltage Transformer (CVT) is shown in Figure 1. The high
voltage terminal is directly connected to electrified body and the other terminal is connected to the
ground; C1 and C2 are adopted for capacitance voltage division, and voltage signal after voltage
division will be transformed into output signal of transformer through potential transformer [6–9].

Sensors 2016, 16, 40; doi:10.3390/s16010040 www.mdpi.com/journal/sensors

Sensors 2016, 16, 40 2 of 21
Sensors 2016, 16, 40
Sensors 2016, 16, 40

Figure 1. Structure diagram of CVT.

Figure 1. Structure diagram of CVT.
Figure 1. Structure diagram of CVT.
Potential transformer is shown in Figure 2, and the high voltage terminal is connected to
Potential
electrified transformer
body.
Potential is is
transformer shown
shownininFigure
Figure 2,
2, and the high
and the highvoltage
voltageterminal
terminal is connected
is connected to to
electrified body.body.
electrified
U 1
2
U 1 3
2
3

1：primary winding
4 2：balanced winding
5 1：primary winding
4 3： iron core winding
2：balanced
N 5 4-5：signal output
3： iron core
N 4-5：signal output
Figure 2. Potential transformer.
Figure 2. Potential transformer.
According to analysis the Figures Figure 12. andPotential
2, the transformer.
traditional voltage transformers shown in
Figures 1 and 2, the
According high voltage
to analysis the terminal
Figures 1of and currently
2, theused voltage transformer
traditional voltage transformers has direct shown electrical in
According
connection
Figures 1 and to2,analysis
with conductor the
the high voltage to beFigures
measured,
terminal 1 ofand
and 2thethe
currently othertraditional
used terminal
voltage of voltage transformers
the transformer
transformer has directis connectedshown in
electrical
to
Figures the
1 and ground.
connection 2 the Therefore,
withhigh voltage
conductor insulation
to terminal
be measured,design of the
of currently
and the other
transformer
usedterminal
voltage isoftransformer
difficult, and,has
the transformer meanwhile,
direct
is it
electrical
connected
involves
to the a large
ground. volume
Therefore, and heavyweight;
insulation designdue to
of the
the iron core
transformer
connection with conductor to be measured, and the other terminal of the transformer is connected designis and multi-grade
difficult, and, capacitance
meanwhile, it to
voltage
involves division
a large of
volumethe transformer,
and heavyweight; serious due timeto delays
the iron occur
core
the ground. Therefore, insulation design of the transformer is difficult, and, meanwhile, it involves during
design and the detection
multi-grade signal and
capacitance
voltage
voltage signal
divisionto beofmeasured in phase serious
the transformer, position.time As adelays
result, detection
occur duringaccuracy of the transformer
the detection signal and is
a large volume and heavyweight; due to the iron core design and multi-grade capacitance voltage
reduced and its application range is restricted. Moreover, three-phase
voltage signal to be measured in phase position. As a result, detection accuracy of the transformer is voltage transformer is
division of the transformer, serious time delays occur during the detection signal and voltage signal to
seldom
reducedadopted
and itsforapplication
the voltagerange grade is above 35 kV in
restricted. electric power
Moreover, systems.voltage
three-phase Therefore, traditionalis
transformer
be measured
voltage in phase position.
transformers can no As a satisfy
longer result,the detection
development accuracy of the ofelectric
the transformer
power system is reduced
[10–12]. and its
seldom adopted for the voltage grade above 35 kV in electric power systems. Therefore, traditional
application
voltage range
In considerationis restricted.
transformers of
canthe Moreover,
noabove-mentioned
longer satisfy three-phase
the problems,
development voltage transformer
thisofpaper
the designed
electric powerisa seldom
systemadopted
non-contact voltage
[10–12]. for the
voltage grade
sensor In consideration of the above-mentioned problems, this paper designed a non-contact voltage can
on above
the 35
basis kV
of in electric
electric field power
coupling systems.
principle, Therefore,
as shown traditional
in Figure voltage
3. In transformers
Figures 1–3 are
no longer
sensorsatisfy
non-contacton the the
voltage development
basis sensors
of electric of the
installed
field electric
on
coupling power system
the three-phase
principle, [10–12].
transmission
as shown line. Positive
in Figure and negative
3. In Figures 1–3 are
induction
In
non-contact electrodes
consideration voltage ofarethedesigned on the
above-mentioned
sensors installed onsensor, and the electric
the problems,
three-phase this field designed
paper
transmission produced by
line. Positive electric
a non-contact charges
and negative voltage
on
sensor the the
transmission
induction
on basis of line
electrodes arewill generate
designed
electric field on induced
the sensor,
coupling charges
and the
principle, on astheshown
electricelectrode ofFigure
fieldinproducedvoltage3. sensor
by electric
In by charges
Figures way 1–3
of are
electrostatic
on the induction.
transmission lineThe
will electric
generatecharges
inducedwill distribute
charges on on
non-contact voltage sensors installed on the three-phase transmission line. Positive and negativethethe electrode
electrode of to form
voltage electric
sensor potential,
by way of
and the potential
electrostatic induction.signal Theofelectric
positive and negative
charges will induction
distribute on the electrodes
electrode is form
to output signal
electric of the
potential,
induction electrodes are designed on the sensor, and the electric field produced by electric charges
non-contact
and the potentialvoltage signal
sensor.of positive and negative induction electrodes is output signal of the
on the transmission line will generate induced charges on the electrode of voltage sensor by way of
Compared
non-contact withsensor.
voltage traditional voltage transformer, non-contact voltage sensor has no electrical
electrostatic
connection induction.
Compared
The electric
with electrified
with traditional
charges will
body. Therefore, thedistribute
voltage transformer, insulation on the is
design
non-contact
electrode
simple and
voltage
to form
sensor thehas electric
cost no low;potential,
is electrical
the
and the potential
induction
connectionelectrodesignal of positive
of non-contact
with electrified and negative
voltage the
body. Therefore, induction
sensor electrodes
is optimized
insulation design through is output
is simple simulationsignal of the
and the costcalculation. non-contact
is low; the
voltage sensor.design
Miniature
induction electrodeof the of sensor is realized,
non-contact voltagethe sensor
voltageisdivision optimized ratio through
is increased, and phase
simulation error of
calculation.
Compared
sensor detection
Miniature withsignal
design traditional
of theissensor
reduced. voltage transformer,
An experimental
is realized, the voltageplatform non-contact
division that isvoltage
ratioconsisted
increased, sensor
of aand
50 kV hassimulated
phase no electrical
error of
three-phase
sensor with
connection detectiontransmission
signal isbody.
electrified line was
reduced. established
An experimental
Therefore, thein insulation
theplatform
laboratory thattoconsisted
design test the of
is simple non-contact
a and
50 kV the voltage
simulated
cost is low;
sensor
the induction designed
three-phase electrode in this
transmission paper.
line was
of non-contact According
established
voltage to the inresult,
sensor the designed
theislaboratory
optimized voltage
to through
test sensor can calculation.
the non-contact
simulation realize
voltage
real-time
sensor detection
designed in for
thisthe voltage
paper. of
According three-phase
to the transmission
result,
Miniature design of the sensor is realized, the voltage division ratio is increased, and phase the line
designed and accurately
voltage sensor measure
can the
realize
error of
voltage
real-timeand phase position.
detection for the voltage of three-phase transmission line and accurately measure the
sensor detection signal is reduced. An experimental platform that consisted of a 50 kV simulated
voltage and phase position.
three-phase transmission line was established in the laboratory to test the non-contact voltage sensor
designed in this paper. According to the result, 2the designed voltage sensor can realize real-time
2
detection for the voltage of three-phase transmission line and accurately measure the voltage and
phase position.
Sensors 2016, 16, 40 3 of 21
Sensors 2016, 16, 40

Sensors 2016, 16, 40

Figure 3. Application of voltage sensor in transmission line.

Figure 3. Application of voltage sensor in transmission line.
Non-contact voltage sensor used at the laboratory test stage adopts closed annular electrode,
and it is mainly
Non-contact appliedFigure
voltage sensor 3. Application
to detection at theof voltage
used performance sensor
and
laboratory in transmission
insulation
test stage line. test of the sensor. In
performance
adopts closed annular electrode,
the follow-up study, open-type annular electrode presented
and it is mainly applied to detection performance and insulation performance in Figure 4 will be designed.
testThe
of sensor
the sensor.
Non-contact voltage sensor used at the laboratory test stage adopts closed annular electrode,
can
In theand be
follow-upconveniently
study, installed
open-type on the
annular line, while
electrode the signal
presented detection
in Figure and
4 processing
will be circuit
designed. Theandsensor
it is mainly applied to detection performance and insulation performance test of the sensor. In
wireless transmission circuit will be designed. In this way, the detection signal can be transferred to
can be conveniently installed on the line, while the signal detection and processing
the follow-up study, open-type annular electrode presented in Figure 4 will be designed. The sensor circuit and wireless
the control terminal in time. In Figure 4, 1 refers to non-contact voltage sensor; 2 is the insulation
transmission circuit will be
can be conveniently designed.
installed on theIn this
line,way,
whilethethedetection signal can
signal detection andbeprocessing
transferred to theand
circuit control
fixture—rubber with good insulation performance and toughness is used in the center fixture so
terminal in
wireless time. In Figure
transmission 4, 1
circuit refers
will betodesigned.
non-contact voltage
In this way, thesensor; 2
detection is the
signal
that when the fixture is compressed, the centers of the line and sensor electrode can be maintained insulation fixture—rubber
can be transferred to
the
with in
good control terminal
insulation in time. In Figure
performance and 4, 1 refersistoused
toughness non-contact
in the voltage
center sensor;
fixture so 2that
is the insulation
when the fixture
the coaxial position, owing to the existence of non-conducting rubber; 3 indicates the signal
fixture—rubber
is compressed, with
the centers good insulation
of the performance
line and sensor and toughness is used in the center fixture so
processing circuit, wireless transmission moduleelectrode
and battery; can4 be maintained
is power in the line;
transmission coaxial
5 isposition,
the
owing that
to when the fixtureofisnon-conducting
the existence compressed, the centersrubber;of3the line andthe
indicates sensor electrode
signal can becircuit,
processing maintained
wireless
insulator; and 6 denotes the connection fitting.
in the coaxial position, owing to the existence of non-conducting rubber; 3 indicates the signal
transmission module and battery; 4 is power transmission line; 5 is the insulator; and 6 denotes the
processing circuit, wireless transmission module and battery; 4 is power transmission line; 5 is the
connection fitting.
insulator; and 6 denotes the connection fitting.

Figure 4. Non-contact open-type voltage sensor.

2. Detection Principle
Figure 4. Non-contact open-type voltage sensor.
The surface of any metal Figure 4. Non-contact
electrified open-type
conductor voltage
has electric sensor.that will generate electric
charges
field. In electric
2. Detection engineering, the power frequency electric field of 50 Hz can be treated as quasi
Principle
2. Detection Principle
static electric field. At this time, when metal electrode is introduced into this electric field, induced
The surface of any metal electrified conductor has electric charges that will generate electric
charges will be produced on the metal electrode, and the quantity of electric charges will change
The
field.surface of any
In electric metal electrified
engineering, the power conductor
frequency has electric
electric charges
field thatcan
of 50 Hz willbegenerate
treated aselectric
quasi field.
with the variation of electric field intensity. Distribution of electric charges on the metal surface will
In electric
static engineering,
electric field. At thethispower
time,frequency
when metal electric fieldisof
electrode 50 Hz caninto
introduced
form electric potential of the metal conductor. When there is potential to ground in this
be this
treated as quasi
electric field, static
induced electric
charges
At this will
field.environment,time,bewhen
produced metal onelectrode
the metaliselectrode,
introduced andinto
the quantity
this of electric
electric field, charges will
induced change
charges will be
potential difference between these two is the induced voltage of metal electrode.
with the
produced on variation
the metal ofelectrode,
electric field
and intensity.
the Distribution
quantity of of electric
electric charges charges
will on the metal
change with surface
the will of
variation
A non-contact voltage sensor was designed according to the above electrostatic induction
form
electric electric
field potential of the of
intensity. metal conductor. When thethere issurface
potential to ground in potential
this
principle. DifferentDistribution
from the traditionalelectric charges
sensor thaton metal
directly measures thewill form
voltage electric
of electrified
environment, potential difference between these two is the induced voltage of metal electrode.
of theconductor,
metal conductor.
this sensor When there
couples is potential
electric to ground
field produced in this environment,
by electric potential difference
charges of the conductor to be
A non-contact voltage sensor was designed according to the above electrostatic induction
measured into voltage sensor, so as to
between these two is the induced voltage of metal electrode.form detection signal. Figure 5 is the schematic diagram for
principle. Different from the traditional sensor that directly measures the voltage of electrified
the
A principle
non-contact of sensor measurement.
conductor, this voltage sensor was
sensor couples designed
electric accordingbytoelectric
field produced the above electrostatic
charges inductiontoprinciple.
of the conductor be
In Figure 5, S indicates the external surface boundary of electrified conductor in any shape; Ω
Different from the traditional sensor that directly measures the voltage
measured into voltage sensor, so as to form detection signal. Figure 5 is the schematic diagram of electrified conductor,
for
means the computational domain of the whole electric field; ε refers to the electric medium constant
this sensor couples
the principle electric
of sensor field produced by electric charges of the conductor to be measured into
measurement.
in the area; r signifies the position vector of the calculation site; r’ denotes the position vector of
voltage sensor,
In Figureso as
5, Stoindicates
form detection
the externalsignal.
surfaceFigure 5 is the
boundary schematicconductor
of electrified
original point; E(r) is the electric field intensity of any point in the computational domain; F(r’)
diagram in forany
theshape;
principle
Ω of
sensormeans the computational domain of the whole electric field; ε refers to the electric medium constant
measurement.
refers to the potential distribution on the external surface of electrified conductor; φ0(t) means the
in the
In area;5,r S
Figure signifies the the position vectorsurface
of the calculation site; r’ denotes the position in
vector of
electric potentialindicates external
of electrified conductor; boundary
A indicates of electrified
the external conductor
surface of metal electrode any
of theshape;
Ω means the computational domain of the whole electric field; ε refers to the electric mediumF(r’)
original point; E(r) is the electric field intensity of any point in the computational domain; constant
refers to the potential distribution on the external surface of electrified conductor; φ0(t) means the
in the area; r signifies the position vector of the 3calculation site; r’ denotes the position vector of
electric potential of electrified conductor; A indicates the external surface of metal electrode of the
original point; E(r) is the electric field intensity of any point in the computational domain; F(r’) refers
to the potential distribution on the external surface of electrified conductor; ϕ0 (t) means the electric
3
Sensors 2016, 16, 40 4 of 21

Sensorsof
potential 2016, 16, 40
electrified conductor; A indicates the external surface of metal electrode of the voltage
sensor; B is the sensor support (connected to the ground); and Rm denotes the sampling resistance of
voltage sensor; B is the sensor support (connected to the ground); and Rm denotes the sampling
voltage sensor.
resistance of voltage sensor.

Figure 5. Principle of sensor measurement.

Figure 5. Principle of sensor measurement.
No free charges exist in the computational domain, i.e., ρ(r’) = 0. The computational domain
No free
meets charges exist
progressive in the
boundary computational
conditions, and anydomain, i.e.,computational
point in the ρ(r’) = 0. The computational
domain domain
meets Poisson
meetsequation of Equation
progressive (1): conditions, and any point in the computational domain meets Poisson
boundary
equation of Equation (1): 
 2( r )   ρ( rp)rq
2
(1)
∇ ϕ pr q “ ´  (1)
ε
Green function is introduced and the following result is gained by solving Equation (1)
Green function is introduced and the following result is gained by solving Equation (1) through
through the first boundary condition φ|s’ = f(r’):
the first boundary condition ϕ|s’ = f (r’):
1
0  E (r ) (2)
1
ϕ0 “  F (r ) Eprq (2)
εFprq
In Equation (2), F(r) is decided by the distance of field point from the conductor to be
measured.
In Equation According to decided
(2), F(r) is this equation,
by thein distance
the computational domain
of field point Ω, the
from theelectric field to
conductor intensity of
be measured.
any point is in direct proportion to electric potential of the conductor to be measured,
According to this equation, in the computational domain Ω, the electric field intensity of any point is in and they have
a linear relation.
direct proportion to electric potential of the conductor to be measured, and they have a linear relation.
Suppose that the electric potential change of electrified conductor with time is φ0(t) and the
Suppose that the electric potential change of electrified conductor with time is ϕ0 (t) and the
frequency is 50 Hz; at this time, the electric field produced by electrified body can be treated as
frequency
quasi is 50 Hz;
static field.atUnder
this time, the electric
the effect field produced
of this electric by electrified
field, induced body
charge q will can be
appear ontreated as quasi
the surface
staticoffield.
metalUnder
electrode theA effect
due toof this electric
electrostatic field, induced
induction principle. charge q will appear
Closed Gaussian surfaceon the surface
is created on of
metaltheelectrode A due to electrostatic induction principle. Closed Gaussian surface
electrode surface with the distance of r from the electrified conductor, and the differential is created on the
electrode
elementsurface
dA is with
takenthe distance
[13–15]. of r from
The electric fieldthe electrified
intensity conductor,
component anddirection
at normal the differential
is En and element
the
dA isangular separation
taken [13–15]. The with E(r,t)
electric fieldisintensity
θ. The component
following equation
at normal candirection
be gained
is Enaccording to
and the angular
Gauss with
separation theorem:
E(r,t) is θ. The following equation can be gained according to Gauss theorem:
dq ¿ dE (r , t )
dq   A dEpr, tq dA (3)
dt “ ε dt dA (3)
dt dt
According to Equation (3), the electric Afield intensity produced by electric charges on the
energized conductor changes with the time t, and the electric field intensity will cause
According to Equation
corresponding changes to(3),the
thequantity
electric of
field intensity
electric produced
charges by electric
on induction charges
electrode on the
A. The energized
induced
conductor changes with the time t, and the electric field intensity will cause corresponding
charges will distribute on electrode A to form electric potential φs(t). Sampling resistance Rm is used changes
to thetoquantity
connect induction electrode A and support B, and the voltage Vs(t) at both ends of the sampling on
of electric charges on induction electrode A. The induced charges will distribute
electrode A toisform
resistance potential ϕs (t). Sampling resistance Rm is used to connect induction
electric voltage:
the detecting
electrode A and support B, and the voltage dE (Vr ,st(t)
) at both
0 Ar
 ends
d of the sampling resistance is the
detecting voltage: V s (t )  Rm 

¿ A

dt
dA 
r 2
Rm
dt
0 (t ) (4)
dEpr, tq ε 0 Ar1 d
By organizing EquationVs ptq(4),“theRmfollowing
ε dA “ be R ϕ0 ptq (4)
dt equation can gained:
m
r2 dt
A

4
Sensors 2016, 16, 40 5 of 21

By organizing Equation (4), the following equation can be gained:

Sensors 2016, 16, 40

r2
ż
ϕ0 ptq “ Vs ptqdt (5)
εARm r1
2
r
 ARm r   s
0 (t )  transmission
When the electrified conductor is cylindrical V (t )dt line conductor, r’ is the section(5)radius
of the cylinder. At this time:
When the electrified conductor is cylindricalr2transmission line conductor, r’ is the section
radius of the cylinder. At this time: K“ (6)
εARm r1
In Equation (6), K is a constant, so: r2
K (6)
żARm r 
In Equation (6), K is a constant, so: ϕ0 ptq “ K Vs ptqdt (7)

 0 (t )  K  Vs (t ) dt (7)
According to Equation (7), the voltage value of electrified conductor can be calculated by
conducting integraltooperation
According Equation for
(7), the voltagevalue
the voltage signalof of sampling
electrified resistance
conductor canand multiplying
be calculated by the
by K; theseintegral
resultconducting two have a linearfor
operation relation. According
the voltage signal toof the above resistance
sampling discussion, a non-contact
and multiplying method
the
can beresult by to
found K; realize
these two have a linear
non-contact relation. According
measurement to the above
for the voltage discussion,
of power a non-contact
frequency high voltage
method
electrified can be found to realize non-contact measurement for the voltage of power frequency high
body.
voltage electrified body.
3. Design of Non-Contact Voltage Sensor
3. Design of Non-Contact Voltage Sensor
Figure 6 presents the lumped parameter equivalent circuit model of electric field coupling
Figure 6 presents the lumped parameter equivalent circuit model of electric field coupling
sensor based on electrostatic induction. ϕ(t) refers to the equivalent voltage of the conductor to
sensor based on electrostatic induction. φ(t) refers to the equivalent voltage of the conductor to be
be measured; C means the equivalent mutual capacitance between the sensor electrode and conductor
measured; aCa means the equivalent mutual capacitance between the sensor electrode and conductor
to be measured;
to be measured; Can indicates the ground
Can indicates equivalent
the ground stray stray
equivalent capacitance of sensor
capacitance electrode;
of sensor Rm signifies
electrode; Rm
the detecting
signifies theresistance;
detectingand R1 andand
resistance; C1 R
constitute integralintegral
1 and C1 constitute circuit circuit
and the andoutput signal
the output of integral
signal of
is V 0 (t),
circuitintegral respectively.
circuit is V0(t), respectively.

Figure 6. Equivalent circuit model.

Figure 6. Equivalent circuit model.
Through Laplace transformation, transfer Equation (8) of sensor and transfer Equation (9) of
Through Laplace
passive integral transformation,
circuit can be gained: transfer Equation (8) of sensor and transfer Equation (9) of
passive integral circuit can be gained: C R s a m
H M (s)  (8)
1  (C a  C an ) Rm s
Ca R m s
H M psq “ (8)
1 ` pC
H (s)  a
`1 Can qRm s
N (9)
1  C1 R1 s
1
Suppose that ωh and ωl are upper H N psq
limit “ lower limit of sensor measurement bandwidth; the (9)
and 1 ` C1 R1 s
following equations can be gained:
Suppose that ωh and ωl are upper limit and lower limit of sensor measurement bandwidth;
1
the following equations can be gained: H  (10)
(C a  C an ) Rm
1
ωH “   1 (10)
L a ` Can qR m
pC (11)
CR 1 1

The characteristic amplitude-frequency 1 curve of the sensor, integral circuit and the
ω Lresponse
“ (11)
entire measurement system can be obtained through1 R
C Equations
1 (8) and (9), as shown in Figure 7.

5
Sensors 2016, 16, 40 6 of 21

The characteristic amplitude-frequency response curve of the sensor, integral circuit and the entire
measurement system can be obtained through Equations (8) and (9), as shown in Figure 7.
Sensors 2016, 16, 40

Figure 7. Amplitude-frequency response curve.

Figure 7. Amplitude-frequency response curve.
As shown in Equation (12), when (Ca + Can)Rm << 1, the sensor works under differential mode:
As shown in Equation (12), when (Ca + Can )Rm << 1, the sensor works under differential mode:
V ( s)
H M ( s)   Ca Rm s (12)
 ( s)
Vpsq
H M psq “ “ Ca R m s (12)
At this time, in order to meet the measurementϕpsq requirements, integral circuit should be added,
but integral circuit will cause troubles to signal-to-noise ratio design of sensor. Meanwhile, due to
At this time,
problems of in order to meet
component the measurement
parameters in integral requirements, integral
circuit, detecting circuitdistortion
waveform should be added,
and
but integral
inaccuratecircuit will causeproblems
measurement troubles to signal-to-noise
will ratio design
be caused. Therefore, of sensor.
sensor shouldMeanwhile,
work under due to
problems of component
self-integration mode.parameters in integral
In other words, Equationcircuit,
(13) isdetecting waveform
gained when (Ca + Candistortion
)Rm >> 1. Atand
thisinaccurate
time,
the required
measurement detectionwill
problems signal
beof sensor can
caused. be obtained
Therefore, without
sensor integral
should work circuit.
under self-integration mode.
In other words, Equation (13) is gained when (CVa(+ s ) Can )RCma >> 1. At this time, the required detection
H M (s)   (13)
signal of sensor can be obtained without integral circuit.
 ( s ) Ca  Can
In order to make the sensor work under self-integration
Vpsq Ca mode at the power frequency of 50 Hz,
H M
the angular frequency should be ωH < 2πpsq “ “
× 50.CaTherefore, (Ca + Can)Rm must meet the (13)
ϕpsq ` Can
following condition:
In order to make the sensor work under self-integration mode at the power frequency of
(Ca  Can ) Rm  3.18  10 3 (14)
50 Hz, the angular frequency should be ω H < 2π ˆ 50. Therefore, (Ca + Can )Rm must meet the
followingNon-contact
condition: measurement mode is adopted, so the distance between the sensor and conductor
to be measured cannot be short. Thereby,
pCa ` Canthe
qRmutual capacitance
m ą 3.18 ˆ 10
´3 Ca between the conductor to be
(14)
measured and induction electrode cannot be large and the distance between the sensor and ground
Non-contact measurement
is long. Therefore, the ground mode is adopted,Cansoofthe
stray capacitance thedistance
sensor is between the sensorspeaking,
not large (generally and conductor
Ca
to beand
measured
Can are cannot be short.
at the level of pF).Thereby,
In orderthe to mutual condition ofCEquation
meet thecapacitance a between the
(14), conductor
the sampling to be
resistance R should be at the level of GΩ, which will cause problems
measured and induction electrode cannot be large and the distance between the sensor and ground is
m to equivalent internal
long.resistance
Therefore, detection
the groundand impedance matching
stray capacitance Canofofthethe
sensor.
sensorTheis small capacitance
not large (generally values of Ca and
speaking, Ca and
C an also restrict the voltage division ratio of sensor. One terminal of the sampling resistance is
Can are at the level of pF). In order to meet the condition of Equation (14), the sampling resistance Rm
connected to the ground, increasing the difficulty of the insulation design of the sensor. Most
should be at the level of GΩ, which will cause problems to equivalent internal resistance detection and
existing electric field coupling sensors based on electrostatic induction adopt the form of adding
impedance matching of the sensor. The small capacitance values of Ca and Can also restrict the voltage
integral circuit, so as to gain the signal to be detected. Meanwhile, the sensor electrode often adopts
division
plateratio
or of sensor. One
cylindrical terminal
design, whichofmight
the sampling
restrict resistance is connectedvoltage
detection bandwidth, to the ground,
division increasing
ratio,
the difficulty
detectionof the insulation
accuracy, designdesign
and insulation of theofsensor. Most existing electric field coupling sensors based
the sensor.
on electrostatic induction
By directing at theadopt
abovethe form of the
problems, adding integral
equivalent circuit,
circuit so as to gain
of non-contact the signal
voltage sensorto be
detected. Meanwhile,
designed the sensor
in this paper electrode
is given, as shown often adopts8.plate
in Figure The or cylindrical
voltage sensordesign, which
of adopting might restrict
differential
inputbandwidth,
detection structure sets the difference
voltage divisionofratio,
suspended
detection potentials output
accuracy, andby a pair of design
insulation sensor ofelectrodes
the sensor.
working
By under
directing atself-integration mode asthe
the above problems, theequivalent
output of sensor.
circuit In
of Figure 7, φ(t) means
non-contact voltage thesensor
equivalent
designed
voltage source of the conductor to be measured; Ca and Cb refer to the equivalent mutual
in this paper is given, as shown in Figure 8. The voltage sensor of adopting differential input structure
capacitance between induction electrode and conductor to be measured; Can and Cbn signify the
sets the difference of suspended potentials output by a pair of sensor electrodes working under
6
Sensors 2016, 16, 40 7 of 21

self-integration mode as the output of sensor. In Figure 7, ϕ(t) means the equivalent voltage source of
the conductor
Sensors 2016,to 40 measured; Ca and Cb refer to the equivalent mutual capacitance between induction
16,be
electrode and conductor to be measured; Can and Cbn signify the ground equivalent stray capacitance
ground equivalent stray capacitance of induction electrode; Cab indicates the equivalent mutual
of induction electrode; Cab indicates the equivalent mutual capacitance between electrodes; Mark 1
capacitance between electrodes; Mark 1 denotes the suspended potential of voltage source;
denotes the suspended potential of voltage source; Marks 2 and 3 are suspended potentials output by
Marks 2 and 3 are suspended potentials output by two electrodes; and Rm refers to the
two electrodes; and Rm [16].
detecting resistance refers to the detecting resistance [16].

Figure 8. Equivalent circuit of voltage sensor.

Figure 8. Equivalent circuit of voltage sensor.
Transfer function of the system is obtained through Laplace transformation:
Transfer function of the system is obtained through Laplace transformation:
(C C  C C ) R s a an b bn m
H (s)  (15)
(C ab Rm s  1)(C a  C
pCb aCCanan ´
 CCbn C
)  (C a  C an )(Cb  Cbn ) Rm s
b bn qRm s
Hpsq “ (15)
According to FigurepCab Rm8,s `the1qpC a ` Cb ` C
non-contact an ` Cbn
voltage a ` Can qpC
q ` pCdesigned
sensor inb ` Cbn
this qRm s adopts two
paper
induction electrodes and inputs suspended potentials of the two electrodes into the differential
According
amplifier oftosignal
Figure 8, the non-contact
processing voltage sensor
circuit as detection signals.designed in thisno
In this method, paper adopts
terminal twosensor
of the induction
electrodes and inputs
is connected to thesuspended
ground, greatlypotentials of the
reducing the two electrodes
insulation designinto the differential
difficulty amplifier
of the sensor. The of
signalsignal
processing circuit
processing as detection
method signals. In
of differential thiscan
type method, no terminal
effectively eliminateof the sensor is connected
interference of
common-mode signal and increase the measurement accuracy of sensor.
to the ground, greatly reducing the insulation design difficulty of the sensor. The signal processing
method ofBydifferential
comparing type Equation (8) with Equation
can effectively eliminate(15),the after design structure
interference of this paper
of common-mode is and
signal
introduced, transfer function
increase the measurement accuracy of sensor.of the sensor will not be changed. Only mutual capacitance C ab

between electrodes is introduced in the electrode point. Compared with the mutual capacitance Ca
By comparing Equation (8) with Equation (15), after design structure of this paper is introduced,
between the induction electrode and conductor to be measured that cannot be large capacitance, the
transfer function of the sensor will not be changed. Only mutual capacitance Cab between electrodes is
distance between induction electrodes can be designed freely and the inter-electrode capacitance Cab
introduced in the electrode point. Compared with the mutual capacitance Ca between
can be adjusted freely. Therefore, such circuit structure can meet the requirement
the induction
of Equation (14),
electrode and
and the conductor
sensor is able totowork
be measured that cannot
under self-integration be large capacitance, the distance between
mode.
inductionTheelectrodes can be designed freely and the inter-electrode
amplitude-frequency characteristic function of sensor designed capacitance
in this paperCabcan
canbebe adjusted
gained
freely.according
Therefore, such circuit
to Equation (15): structure can meet the requirement of Equation (14), and the sensor is
able to work under self-integration mode. Ca Cbn  Cb Can
The amplitude-frequency characteristic function Rm designed in this paper can be gained
Ca  Cb  Cof sensor
an  Cbn
H ( )  (16)
according to Equation (15): 2 (Ca  Can )(Cb  Cbn ) 2 1
Rm [  Cab ] 
Ca  Cb  Can  Cbn 2
Ca Cbn ` Cb Can
Rm
Its phase-frequency characteristic function
Ca `is:Cb ` Can ` Cbn
|Hpωq| “ d (16)
2
pCa ` Can qpCb `1Cbn q
H ( ) Rarctan
2r
1
` C s `
m
Ca R`C[b(C`a C Can`)(C
 an Cbbn Cbn )  C ] ω 2
ab (17)
m ab
Ca  Can  Cb  Cbn
Its phase-frequency characteristic function is:
According to Equation (16), the inter-electrode mutual capacitance Cab exists in the
denominator, so the transformation ratio of the sensor can1 be increased by magnifying the mutual
=Hpωq
capacitance Cab between “ arctan
electrodes without changing other conditions. In Equation (17), Cab also (17)
pCa ` Can qpCb ` Cbn q
Rm ωr difference between
exists in the denominator only, so the phase ` Cabsignals
output s of the electrified
Ca ` Can ` Cb ` Cbn
body to be measured and voltage sensor can be reduced by increasing the inter-electrode
capacitance toCEquation
According ab, thus measurement accuracy and response speed of the sensor can be
(16), the inter-electrode mutual capacitance Cab exists in the denominator,
further enhanced.
so the transformation ratio of the sensor can be increased by magnifying the mutual capacitance
Cab between electrodes without changing other conditions. In Equation (17), Cab also exists in the
7
Sensors 2016, 16, 40 8 of 21

denominator only, so the phase difference between output signals of the electrified body to be measured
and voltage sensor can be reduced by increasing the inter-electrode capacitance Cab , thus measurement
accuracy and response speed of the sensor can be further enhanced.
Through the above analysis, the sensor is required to work under self-integration mode at the
power frequency of 50 Hz due to the following two reasons:
Sensors 2016, 16, 40
(1) If the sensor works under differential state, integral circuit should be designed. According to
Through
the analysis the above
of circuit analysis,
structure, the sensor bandwidth
the detection is required to work limit
upper underofself-integration mode at
sensor is restricted bythe
corner
power frequency of 50 Hz due to the following two reasons:
frequency of the sensor at this time, and detection bandwidth lower limit is restricted by corner
frequency(1) of Ifthe
thecircuit.
sensor Therefore,
works under bandwidth
differentialof the entire
state, integralsensor
circuitdetector
should beisdesigned.
restricted.According
to the
(2) analysis of and
Capacitance circuitresistance
structure, in
theintegral
detectioncircuit
bandwidth upper limitby
are restricted of sensor
factorsislike
restricted by
temperature
corner As
coefficient. frequency of the
a result, the accuracy
sensor at ofthissimulating
time, and detection bandwidth
integral circuit lower limit
can hardly reachis the
restricted by
requirement.
corner frequency
Meanwhile, due to strayof theparameters
circuit. Therefore, bandwidth of
of components andtheparts
entirein
sensor detector
integral is restricted.
circuit, distortion might
(2) Capacitance and resistance in integral circuit are restricted by factors like temperature
happen to the detection signal.
coefficient. As a result, the accuracy of simulating integral circuit can hardly reach the requirement.
According
Meanwhile, to due theto analysis in this paper,
stray parameters in orderand
of components to meet theintegral
parts in above circuit,
requirements,
distortionthe mutual
might
happen between
capacitance to the detection
sensorssignal.
should be increased to the largest extent. This paper shows that the
method ofAccording
increasing to the mutual in
the analysis capacitance
this paper, between
in order toinduction electrodes
meet the above of the sensor
requirements, is to treat
the mutual
suspended potential
capacitance signal
between of sensor
sensors should electrode as output
be increased to thesignal of extent.
largest the sensor. The purpose
This paper shows that of adopting
the
differential
methodamplifier
of increasingis tothe
restrain
mutual common-mode
capacitance between signalinduction
in suspended potential
electrodes of the signal
sensor of sensor,
is to treat and
suspended
to amplify potential signal
the differential signalofatsensor
the same electrode
time. as output signal of the sensor. The purpose of
adopting differential amplifier is
In the future, solar panel will be adopted asto restrain common-mode
power supply signal in suspended
of signal processingpotential
circuit.signal
The ofcircuit
sensor, and to amplify the differential signal at the same time.
board adopts the mode of DC +5 V power supply, and the differential amplifier uses the mode of +3 V
In the future, solar panel will be adopted as power supply of signal processing circuit. The
and ´3 V power supply. Signal processing circuit and solar panel are designed in the box, and installed
circuit board adopts the mode of DC +5 V power supply, and the differential amplifier uses the
in themode
tower of high-voltage transmission line, so as to realize voltage detection for high-voltage
of +3 V and −3 V power supply. Signal processing circuit and solar panel are designed in the
transmission
box, and line.
installed in the tower of high-voltage transmission line, so as to realize voltage detection
According to thetransmission
for high-voltage above analysis,line. the non-contact voltage sensor designed in this paper is presented
in Figure According
9. Figure 9a shows the physical the
to the above analysis, design of sensor.
non-contact Annular
voltage sensorcopper electrodes
designed in this are arranged
paper is
on thepresented
front and in back
Figuresides of PCB
9. Figure boardthe
9a shows (printed-circuit
physical designboard) as induction
of sensor. Annular copperelectrodes. Theare
electrodes sensor
arranged
diameter is 70 on
mm theand
fronttheand back
pore sides of in
diameter PCB
theboard
middle(printed-circuit
is 18 mm; there board)areas32induction
electrodes electrodes.
on the front
The sensor diameter is 70 mm and the pore diameter in the middle is 18
side and the electrode spacing is 0.3 mm; there are 32 electrodes on the back side and the electrodemm; there are 32 electrodes
on the front side and the electrode spacing is 0.3 mm; there are 32 electrodes on the back side and
spacing is 0.3 mm; and the detection signal is drawn forth by the screened wire. Figure 9b is the
the electrode spacing is 0.3 mm; and the detection signal is drawn forth by the screened wire.
schematic diagram for sensor structure and the electric transmission line passes through the pore in
Figure 9b is the schematic diagram for sensor structure and the electric transmission line passes
the middle of the sensor. In order to enhance the mutual capacitance between electrodes and the sensor
through the pore in the middle of the sensor. In order to enhance the mutual capacitance between
electrode ability
electrodes and of the
holding
sensorelectric
electrodecharges
ability to
of the largest
holding extent,
electric multiple
charges to thePCB boards
largest extent,are connected;
multiple
all electrodes on the front side of each PCB board are connected and all electrodes
PCB boards are connected; all electrodes on the front side of each PCB board are connected and all on the back side of
each PCB boardonare
electrodes theconnected.
back side of each PCB board are connected.

(a) (b)
Figure 9. Design of non-contact voltage sensor: (a) physical design of the sensor and (b) schematic
Figure 9. Design of non-contact voltage sensor: (a) physical design of the sensor; and (b) schematic
diagram for sensor structure.
diagram for sensor structure.
4. Simulation
In this paper, Ansoft software is used to realize optimization for the structure design of
non-contact voltage sensor. Meanwhile, the influence of sensor introduction on measurement and

8
Sensors 2016, 16, 40 9 of 21

4. Simulation
In this paper, Ansoft software is used to realize optimization for the structure design of non-contact
voltage sensor. Meanwhile, the influence of sensor introduction on measurement and the distortion
situations around the electric field are studied when sensor is applied to voltage detection of
three-phase transmission line.
Sensors 2016, 16, 40
The parameters of voltage sensor were shown in Table 1, and those parameters were used
the distortion situations around the electric field are studied when sensor is applied to voltage
in simulation.
detection of three-phase
Sensors 2016, 16, 40 transmission line.
1. Parameters
Table were
The parameters of voltage sensor shown inofTable
sensor.
1, and those parameters were used
the distortion situations around the electric field are studied when sensor is applied to voltage
in simulation.
detection of three-phase transmission line.
Number of Width of Spacing of Diameter of Innermost
Sensor The parameters of voltage sensor were
Table shown inofTable
1. Parameters 1, and those parameters were used
sensor.
in simulation.
Electrode Electrode (mil) Electrode (mm) Electrode (mm)
Board front Number of
32 Width6of Electrode Spacing
0.3 of Diameter of Innermost
Sensor 28
Board opposite Electrode
32 (mil)
Table61. Electrode
Parameters of sensor. 0.3 (mm) Electrode (mm)
Board front 32 6 0.3
Number of Width of Electrode Spacing of 28
Diameter of Innermost
Board opposite
Sensor 32 6 0.3
Electrode (mil) Electrode (mm) Electrode (mm)
The actual sensor
Board front adopts 1232 PCB boards. According
6 to the0.3measurement results of Agilent 4294 A
The actual sensor adopts 12 PCB boards. According to the measurement 28
results
impedance analyzer, electrode mutual capacitance between the output
Board opposite 32 6 0.3 ends of the PCBofsensor
Agilent
is 950 pF,
4294 A impedance analyzer, electrode mutual capacitance between the output ends of the PCB
and the maximum fluctuation deviation value within the frequency range from 45 Hz to 2 MHz is 2 pF.
sensor The
is 950 pF, and
actual theadopts
sensor maximum
12 PCBfluctuation deviation to
boards. According value within the frequency
the measurement results ofrange from
Agilent
45 4294
Hz toA2 impedance
MHz is 2 pF.
analyzer, electrode mutual capacitance between the output ends of the PCB
4.1. Optimization Design of Sensor Structure
sensor is 950 pF, and the maximum fluctuation deviation value within the frequency range from
The4.1.
45Optimization
Hz to 2 MHz
simulation Design
is 2 pF.of Sensor
experiment Structure
model is established according to the electrode structure presented
in Figure 10 where
The 1 represents
simulation experiment the electrode
model
4.1. Optimization Design of Sensor Structure on the front
is established sidetoofthe
according PCB board;
electrode 2 means
structure the electrode
presented
in Figure
on the back side10 of
where
PCB1 represents
board; R,theDelectrode
and d referon the to
front
theside of PCBwidth
radius, board; 2andmeans the electrode
electrode spacing of
on the The
backsimulation
side of PCBexperiment
board; R,model
D andis established
d refer to according
the radius,to width
the electrode structurespacing
and electrode presentedof #1
#1 electrode, respectively;
in Figure 10 where 1 and N indicates
represents theon
the electrode number
the front of electrodes.
side of PCB board; Through post-processing
2 means the electrode for
electrode, respectively; and N indicates the number of electrodes. Through post-processing for the
the simulation
on theresult,
back side equivalent
of PCB board;distributed
R, D and dcapacitances under
refer to the radius, different
width motivations
and electrode
simulation result, equivalent distributed capacitances under different motivations and different
spacing ofand#1 different
electrode electrode, respectively;
structures can becan and N
gained indicates
via the number
calculation, of electrodes.
as shown in in ThroughFigure
Figure post-processing forrelation
the
electrode structures be gained via calculation, as shown Figure11.11. Figure 11a is the
11a is the relation curve
simulation result, equivalent distributed capacitances under different motivations and different
of equivalent
curve of capacitance and electrode
equivalent capacitance and radius;
electrodeFigure
radius;11b presents
Figure the relation
11b presents curvecurve
the relation of equivalent
of
electrode structures can be gained via calculation, as shown in Figure 11. Figure 11a is the relation
equivalent
capacitance and capacitance
electrode and
width; electrode
and width;
Figure and
11c Figure
shows 11c
the shows
relationthe relation
between between
curve of equivalent capacitance and electrode radius; Figure 11b presents the relation curve of equivalent
equivalent capacitance
capacitance
and number and
of electrodes.
equivalent number and
capacitance of electrodes.
electrode width; and Figure 11c shows the relation between equivalent
capacitance and number of electrodes.

Figure
Figure 10.10. Sensor electrode
Sensor electrode structure.
structure.
Figure 10. Sensor electrode structure.

(a)
(a) (b)
(b)

Figure 11. Cont.

9
9
Sensors 2016, 16, 40 10 of 21
Sensors 2016, 16, 40

(c)
Figure 11. Simulation result of electrode structure optimization. (a) Relation curve of equivalent
Simulation
Figure 11.capacitance result of electrode structure optimization. (a) Relation curve of equivalent
and electrode radius; (b) Relation curve of equivalent capacitance and electrode width.
capacitance(c) Relation curve of radius;
and electrode equivalent(b) Relationand
capacitance curve of of
number equivalent
electrodes. capacitance and electrode width;
(c) Relation curve of equivalent capacitance and number of electrodes.
The following conclusions can be gained through the simulation result:
(1) With the increase of the transformer electrode radius R, the mutual capacitance Cab between
The following conclusions can be gained through the simulation result:
electrodes decreases and the ground stray capacitance Cn of electrode rises.
(2) With
(1) With the the increase
increase of theoftransformer
the transformer electrode radius
electrode width D,R,the ground
the straycapacitance
mutual capacitance CC
n of
ab between
electrode rises, while the mutual capacitance Cab between electrodes remains unchanged.
electrodes decreases and the ground stray capacitance Cn of electrode rises.
(3) With the increase of the number of electrodes N, the ground stray capacitance Cn of
(2) With the increase
electrode rises, whileofthe
the transformer
mutual capacitanceelectrode
Cab betweenwidth D, remains
electrodes the ground stray capacitance Cn of
unchanged.
electrode rises,Therefore,
while the mutual capacitance Cab between
when the induction electrode
electrodes remains unchanged.
of sensor is designed, in order to reduce the
(3) With the increase of the number of electrodes N, the ground stray
measurement error disturbance caused by the ground stray capacitance Cn of capacitance Cn of electrode
sensor, the electrode
rises, while the mutual capacitance Cab between electrodes remains unchanged.
radius should be reduced to the largest extent, and meanwhile the width and distance between
electrodes must be decreased. On the premise of comprehensively considering factors like the size
Therefore, when
of sensor themore
space, induction
electrodeselectrode
should beof sensorand
arranged is designed, in order tobetween
the mutual capacitance reduceelectrodes
the measurement
must be increased
error disturbance caused by to the
thelargest
groundextent. In this
stray way, the voltage
capacitance Cn ofdivision
sensor,ratio
theand measurement
electrode radius should
be reducedaccuracy
to the of sensor can be increased, and phase error in measurement will be reduced.
largest extent, and meanwhile the width and distance between electrodes must be
decreased.4.2.
OnInsulation
the premise
Design of comprehensively considering factors like the size of sensor space, more
of Sensor
electrodes should be arranged and the mutual capacitance between electrodes must be increased to
the largest4.2.1. Insulation
extent. Design
In this way,between Electrodes
the voltage division ratio and measurement accuracy of sensor can be
increased, and phase error in measurementcapacitance
In order to increase the mutual Cab between electrodes to the largest extent,
will be reduced.
measures to reduce the distance between electrodes and decrease the electrode width are taken in
electrode
4.2. Insulation Design design. These measures have changed the electric field distribution inside the sensor.
of Sensor
When the sensor is put in regions with high electric field intensity, the insulating ability of sensor
electrode structure should be studied [17–19]. Therefore, simulation calculation should be
4.2.1. Insulation Design between Electrodes
conducted for the electric field intensity between electrodes, and the simulation result is presented
in Figure
In order 12, where
to increase the1mutual
represents the diameter
capacitance Cabof between
10 kV electrified body;to2 the
electrodes refers to the extent,
largest air gap measures
between electrode and sensor; and 3 indicates the spacing between sensor support and innermost
to reduce the distance between electrodes and decrease the electrode width are taken in electrode
electrode. According to the analysis, the electric field intensity between electrodes is far lower than
design. These measures
the electric have changed
field intensity the electric
from electrode surfacefield
to thedistribution
innermost sideinside theInsensor.
of sensor. order toWhen
clearlythe sensor
is put in regions with
reflect the high
electric electric
field fieldbetween
distribution intensity, the insulating
electrodes, the maximum ability ofthe
ruler of sensor electrode
calculation result structure
is 200 V/m, but the actual electric field intensity from electrode surface
should be studied [17–19]. Therefore, simulation calculation should be conducted for the electric to the innermost side of field
sensor is far higher than 200 V/m. Table 2 shows the relative dielectric constant and critical field
intensity between electrodes, and the simulation result is presented in Figure 12, where 1 represents the
intensity of medium involved in the calculation. According to the distribution of electric field
diameter ofintensity,
10 kV electrified
the maximum body; 2 refers
electric field to the airappears
intensity gap between electrode
at the junction and
part of sensor; and 3 indicates
the outermost
the spacing between
electrode andsensor
sensor support
support (50 and innermost
V/m) as well aselectrode.
the interfaceAccording to the
of the sensor analysis,
support and air the electric
field intensity between electrodes is far lower than the electric field intensity from electrode surface to
the innermost side of sensor. In order to clearly reflect
10 the electric field distribution between electrodes,
the maximum ruler of the calculation result is 200 V/m, but the actual electric field intensity from
electrode surface to the innermost side of sensor is far higher than 200 V/m. Table 2 shows the relative
dielectric constant and critical field intensity of medium involved in the calculation. According to
the distribution of electric field intensity, the maximum electric field intensity appears at the junction
part of the outermost electrode and sensor support (50 V/m) as well as the interface of the sensor
support and air (150 V/m). The electric field intensity between electrodes is quite low, almost equal to
0. Therefore, such structure can meet the insulation design between electrodes.
 Sensors 2016, 16, 40
Sensors 2016, 16, 40 11 of 21
(150 V/m).
Sensors The
2016, 16, 40 electric field intensity between electrodes is quite low, almost equal to 0. Therefore,
such structure can meet the insulation design between electrodes.
(150 V/m). The electric field intensity Table 2. Medium
between parameters.
electrodes is quite low, almost equal to 0. Therefore,
such structure can meet the insulation Table 2. Medium
design betweenparameters.
electrodes.
Medium Relative Permittivity Critical Electric Field (kV/cm)
Medium Relative Permittivity Critical Electric Field (kV/cm)
Table 2. Medium parameters.
Epoxy (PCB)
Epoxy(PCB) 3.6
3.6 200–300
200–300
Air
Medium
Air Relative 1
Permittivity
1 Critical 25–30
Electric
25–30Field (kV/cm)
Epoxy(PCB) 3.6 200–300
Air 1 25–30

Figure
Figure 12. 12. Insulation
Insulation intensityof
intensity ofsensor
sensor electrode
electrodeand
andsimulation calculation.
simulation calculation.

4.2.2. InsulationFigure
Design12.of Sensor Support
Insulation intensity of sensor electrode and simulation calculation.
4.2.2. Insulation Design of Sensor Support
Figure 13 shows the electric field intensity distribution of the electrified body under the
4.2.2. Insulation
Figure Design of Sensor Support
voltage 13 shows
grade thekV
of 10 electric field
in electrified intensity
conductor distribution
(conductor), of conductor
the electrified body under
and sensor air gapthe voltage
(air),
gradeinnermost
of 10 kV in electrified
Figure electrode
13 shows and conductor
the electric
support field (conductor),
intensity
of sensor conductor
distribution
(epxoy), and sensor
sensorofelectrode air
the electrified gap (air),
body under
(electrode), innermost
outermostthe
electrode andgrade
voltage
electrode support of 10
and support of kV
sensor (epxoy),
ofinsensor
electrified sensorandelectrode
conductor
(epxoy), (electrode),
(conductor),
sensor and outermost
air conductor electrode
and sensor
(air) successively. and
air gap
According support
to(air),
the
innermost
simulation
of sensor (epxoy),electrode
result,
and and support
thesensor
maximum andvalue airof(air)
ofsensor
electric(epxoy), sensor
field intensity
successively. electrode
appears
According (electrode),
in the
to the sensor outermost
support
simulation part the
result,
electrode
closestvalue
maximum andof
to the support
electricof
electrified bodysensor
field (epoxy (epxoy),
intensity resin) and
and sensor
appears the theand
inelectric air (air)
field
sensor successively.
intensity
support is about
part According
8000 to
closest V/m.
theAsto the
per
electrified
simulation
Table 1, it result,
is far the
lower maximum
than the value
critical of electric
field field
intensity intensity
of epoxy appears
resin. in the
Therefore,
body (epoxy resin) and the electric field intensity is about 8000 V/m. As per Table 1, it is far lower sensor
this support
structure part
can
closest
meet the to insulation
the electrified body
design of (epoxy
sensor resin)
support. andThe
the electric
electric field
field intensity
intensity between
is about 8000 V/m. As
electrodes per
inside
than the critical field intensity of epoxy resin. Therefore, this structure can meet the insulation design
Table 1, it is far
the sensor lower
quite low,than theis critical
but it not equalfield
to intensity of epoxy
0. The electric fieldresin. Therefore,
distribution this
after structure can
magnifying the
of sensor support. The electric field intensity between electrodes inside the sensor is quite low, but it is
meet the insulation
coordinate system isdesign
presentedof sensor support.
in Figure 14. The electric field intensity between electrodes inside
not equal to 0. is
the sensor The electric
quite fieldit distribution
low, but is not equal to after magnifying
0. The thedistribution
electric field coordinateafter system is presented
magnifying the in
Figure 14.
coordinate system is presented in Figure 14.

Figure 13. Insulation design simulation result of electrode.

Figure 13. Insulation design simulation result of electrode.

Figure 13. Insulation design simulation result of electrode.

11

11
Sensors 2016, 16, 40 12 of 21
Sensors 2016, 16, 40

Sensors 2016, 16, 40

Figure 14. Internal electric field intensity distribution of sensor electrode.

Figure 14. Internal electric field intensity distribution of sensor electrode.
4.3. Influence of Sensor Introduction on Measurement
4.3. Influence of SensorFigure 14. Internal
Introduction onelectric field intensity distribution of sensor electrode.
Measurement
When non-contact method is adopted to measure the voltage of electrified conductor, the use
of
4.3.sensor
When will
non-contact
Influence inevitably
of Sensor method influence
is adopted
Introduction electric field around
to measure
on Measurement the conductor
the voltage and cause
of electrified electric the
conductor, fielduse of
distortion.
sensor will Therefore,
inevitably the
influence distortion
electric degree
field to of
around electric field
thethe caused
conductor by
and the introduction
cause electric of sensor
fieldthe distortion.
When non-contact method is adopted measure voltage of electrified conductor, use
should be studied [20–22].
Therefore,
of sensor will inevitably influence electric field around the conductor and cause electric field be
the distortion degree of electric field caused by the introduction of sensor should
Figure 15 shows the simulation result of electric field intensity in the cross section direction
studied [20–22].Therefore, the distortion degree of electric field caused by the introduction of sensor
distortion.
under horizontal arrangement of three-phase conducting wire. Sectional area of the wire is 35 mm²,
Figure
should be 15studied
shows[20–22].
the simulation result of electric field intensity in the cross section direction
the wire spacing is 1.38 m, and the vertical height from the ground is 1.8 m. Its size and structure
under Figure 15arrangement
shows the simulation result ofconducting
electric fieldwire.
intensity in thearea
crossofsection direction
are consistent with the testofplatform
horizontal three-phase actually established. Sectional
During the simulation, thethewire is 35 mm²,
voltage
under
the wire horizontal
spacing is arrangement
1.38 m, and theof vertical
three-phaseheightconducting
from thewire.
groundSectional
is 1.8area
m. of
Itsthe wire
size and is structure
35 mm², are
frequency of three-phase conducting wire is 50 Hz and the voltage amplitude is 12 kV. In Figure 15,
the wire
consistent withspacing
the testis 1.38 m, and the vertical
platform height from the ground is 1.8 m. Its the sizevoltage
and structure
A-phase, B-phase and C-phase actually
wires areestablished.
presented from During theside
the left simulation,
successively, and frequency
the initial
are consistent with the test platform actually established. During the simulation, the voltage
phases are conducting
of three-phase 0°, −120° andwire −240°,is respectively.
50 Hz and the It can be seen
voltage that whenisno12sensor
amplitude kV. InisFigure
added, 15, electric
A-phase,
frequency of three-phase conducting wire is 50 Hz and the voltage amplitude is 12 kV. In Figure 15,
field distribution around the three-phase conducting wire is uniform.
B-phase and C-phase wires are presented from the left side successively, and the initial phases are 0˝ , Figure 16 shows electric field
A-phase, B-phase and C-phase wires are presented from the left side successively, and the initial
´120distribution
˝ and ´240around
phases
˝ the three-phase
are 0°,, respectively.
−120° and −240°, It can conducting
be seen that
respectively.
wire
when
It can
afternosensor
be seen sensor is is
that when
installed
added, on the line.
electric
no sensor
According
field
is added, distribution
electric
to the comparison, the introduction of sensor causes distortion to electric field distribution. The
around the three-phase conducting wire is uniform. Figure 16 shows electric
field distribution around the three-phase conducting wire is uniform. Figure 16 shows electric field field distribution around
electric field intensity between phase and phase decreases obviously. As for the reason, the electric
the three-phase conducting
distribution around wire afterconducting
the three-phase sensor is installed
wire after on the is
sensor line. According
installed to the
on the line. comparison,
According
field line sent out by electric charges of the conductor ends in the electrode of voltage sensor.
the introduction
to the comparison, of sensor
the causes distortion
introduction to
of sensor electric field
causes distortion distribution.
to electricThe
Figure 17a,b are diagrams for local electric field distribution of A-phase wire before and after sensor
fieldelectric field
distribution. intensity
The
electric
between phasefieldandintensity
phase between phase
decreases and phase
obviously. As decreases
for the obviously.
reason, the As forfield
electric the reason,
line senttheout electric
by electric
is used, respectively. The application of sensor changes the electric field distribution. However,
field
charges line
of the to sent out
conductor by electric
ends charges
in thefield of
electrode the conductor
of voltage ends
sensor. in the electrode
Figure 17a,b of voltage
are diagrams sensor.
for local
according Figure 17b, electric distribution between the electrified conductor and voltage
Figure
electric field17a,b are diagrams for localwireelectric fieldanddistribution of A-phase wire before and after application
sensor
sensor isdistribution
uniform, and of no
A-phase
obvious electric before after sensor
field mutation is used,
is seen. In this respectively.
section, the The interphase
is used, respectively. The application of sensor changes the electric field distribution. However,
electricchanges
of sensor field intensity is reduced
the electric fieldbydistribution.
using the sensor, but no obvious
However, according electric
to field
Figure mutation is seen. field
17b, electric
according to Figure 17b, electric field distribution between the electrified conductor and voltage
The electric field distribution is still uniform.
distribution between the electrified conductor and voltage sensor is uniform, and no obvious electric
sensor is uniform, and no obvious electric field mutation is seen. In this section, the interphase
field electric
mutation fieldisintensity
seen. Inisthis section,
reduced the interphase
by using the sensor, butelectric field intensity
no obvious is reduced
electric field mutationby using the
is seen.
sensor,
Thebut no obvious
electric electric field
field distribution mutation
is still uniform.is seen. The electric field distribution is still uniform.

Figure 15. Electric field distribution without sensor.

12
Figure 15. Electric field distribution without sensor.
Figure 15. Electric field distribution without sensor.
12
Sensors 2016, 16, 40 13 of 21
Sensors 2016, 16, 40
Sensors 2016, 16, 40

Figure
Figure 16.16. Electricfield
Electric fielddistribution
distribution after
afterintroducing sensor.
introducing sensor.
Figure 16. Electric field distribution after introducing sensor.

(a) (b)
(a) (b)
Figure 17. Influence of sensor on electric field. (a) A-phase electric field distribution without sensor;
Figure 17. Influence of sensor on electric field. (a) A-phase electric field distribution without sensor;
Figure 17.(b)Influence
and of sensor
A-phase electric fieldon electric field.
distribution after (a) A-phasesensor.
introducing electric field distribution without sensor;
and (b) A-phase electric field distribution after introducing sensor.
and (b) A-phase electric field distribution after introducing sensor.
Measurement points a and b are put 5 and 45 mm away from the left side of B-phase wire.
Measurement points a and b are put 5 and 45 mm away from the left side of B-phase wire.
Electric field intensity at the above two points is calculated at the same moment with the interval of
Measurement pointsatathe
Electric field intensity and b are
above twoput 5 and
points 45 mm away
is calculated frommoment
at the same the leftwith
sidethe
1 kV (the phase spacing of three phases is 120°) when the line voltage is within the range of 1–12 kV.
of interval
B-phase of wire.
1 kV
Electric (theintensity
field phase spacing
at ofabove
the three phases
two is 120°)iswhen
points the lineatvoltage
calculated the is within
same the range
moment with of 1–12
the kV. of
interval
Table 3 shows the electric field intensity when sensor is not used, and Table 4 presents the electric
1 kV Table 3 shows
(the phase
field intensity
the electric
spacing fieldphases
of three
after sensor
intensity
is used.
when
is 120 sensor the
˝ ) when is not used,
line and Table
voltage 4 presents
is within the electric
the range of 1–12 kV.
field intensity after sensor is used.
Table 3 shows the electric field intensity when sensor is not used, and Table 4 presents the electric field
Table
intensity after sensor isTable 3. Electric field intensity at A and B when sensor is not used.
used.3. Electric field intensity at A and B when sensor is not used.
Voltage Phase = 35 Electric Field Intensity Electric Field Intensity of
Voltage Phase = 35 field Electric
3. Electric intensity Field Intensity Electric
is Field Intensity of
Table
[deg] (kV) of PointatAA(kV/m)
and B when sensorPointnotBused.
(kV/m)
[deg] (kV) of Point A (kV/m) Point B (kV/m)
1 2.10 0.15
Voltage 12Phase = 35 Electric Field 2.10
Intensity
4.20 Electric Field0.15
Intensity
0.30
2 (kV)
[deg] of Point A4.20
(kV/m) of Point B 0.30
(kV/m)
3 6.30 0.48
3 6.30 0.48
41 2.10
8.39 0.150.60
42 8.39
4.20 0.300.60
5 10.49 0.75
53 10.49
6.30 0.480.75
6 12.59 0.90
64 12.59
8.39 0.600.90
7 14.69 1.04
75 14.69
10.49 0.751.04
8 16.79 1.19
86 12.59
16.79 0.901.19
9 18.89 1.34
97 14.69
18.89 1.041.34
108 20.99
16.79 1.49
1.191.49
10 20.99
119 23.089
18.89 1.64
1.341.64
11 23.089
1210 25.189
20.99 1.491.79
12 25.189 1.79
11 23.089 1.64
12 25.18913 1.79
13
Sensors 2016, 16, 40 14 of 21

Sensors 2016, 16, 40

Table 4. Electric field intensity at A and B when sensor is introduced.

Table 4. Electric field intensity at A and B when sensor is introduced.

Voltage Phase
Voltage Phase= 35
= 35 ElectricField
Electric FieldIntensity
Intensity Electric Field Intensity of
[deg]
[deg] (kV)(kV) of
ofPoint
PointAA(kV/m)
(kV/m) of Point B (kV/m)
1 1 13.35
13.35 0.008
0.008
2 2 25.86
25.86 0.012
0.012
3 3 39.00
39.00 0.016
0.016
4 4 52.00
52.00 0.019
0.019
5 5 65.00
65.00 0.023
0.023
6 6 78.00
78.00 0.028
0.028
7 7 90.99
90.99 0.032
0.032
8 103.99 0.035
8 103.99 0.035
9 116.99 0.040
9 116.99 0.040
10 129.98 0.045
10 11 129.98
132.15 0.045
0.058
11 12 132.15
155.99 0.058
0.062
12 155.99 0.062

TheThe
linearity
linearityfitting
fitting curve graph shown
curve graph shownininFigure
Figure
18 18 is drawn
is drawn according
according to thetoresults
the results
in in
Tables 2 and
Tables 3. 3.
2 and

(a)

(b)
Figure 18. Fitting curve of electric field intensity and voltage grade: (a) phase A; and (b) phase B.
Figure 18. Fitting curve of electric field intensity and voltage grade: (a) phase A; and (b) phase B.
Least-square fitting is conducted for electric field intensity data at A and B acquired before and
Least-square
after fitting is conducted
sensor is introduced, for electric
and Equations (18) andfield
(19) intensity datawhere
are obtained, at A and B acquired
x means before and
electric field
intensity of A and B before using the sensor, and y means electric field intensity of
after sensor is introduced, and Equations (18) and (19) are obtained, where x means electric field A and B after
using of
intensity theAsensor. In the equation
and B before using thethe constant
sensor, and term is generated
y means under
electric field the combined
intensity of A andaction of using
B after
boundary condition setting and fitting method when simulation model is established.
the sensor. In the equation the constant term is generated under the combined action of boundary
y simulation
condition setting and fitting method when 6.0291x  1.06
model is established. (18)
y  0.0317 x  0.7024 (19)
y “ 6.0291x ` 1.06 (18)
14
y “ 0.0317x ` 0.7024 (19)
Sensors 2016, 16, 40 15 of 21

According to the above simulation result, a linear relation is presented between electric field
intensity around
Sensors 2016, 16,the
40 circuit and line voltage when sensor is not used. When voltage sensor designed in
this paper is used, a linear relation is still presented between electric field intensity around the circuit
AccordingTherefore,
and line voltage. to the above simulation
though result,
the use a linear sensor
of voltage relation will
is presented between
influence electricfield
the electric fieldto be
intensity around the circuit and line voltage when sensor is not used. When voltage sensor
measured, electric field distribution is still uniform, and electric field intensity and line voltage present
designed in this paper is used, a linear relation is still presented between electric field intensity
linear changes.
around the circuit and line voltage. Therefore, though the use of voltage sensor will influence the
electric field to be measured, electric field distribution is still uniform, and electric field intensity
5. Experiment
and line voltage present linear changes.
The working performance of non-contact voltage sensor should be verified by experiment.
5. Experiment
The experiment is composed of sensor linearity test and sensor measurement accuracy test.
The experimental platform isofshown
The working performance in Figure
non-contact voltage19.sensor
In Figure
should19a, 1 refersbytoexperiment.
be verified the voltage sensor
The
experiment is composed of sensor linearity test and sensor measurement accuracy
and this sensor installed on all three-phase lines. Through design of the voltage sensor, its voltage test.
The experimental
division ratio is set as 1500:1. platform is shown inthe
2 represents Figure 19. In Figure
high-voltage 19a, of
probe 1 refers to theitsvoltage
P6051A, voltage sensor
division
ratio and this
is set assensor
1000:1. is installed
3 indicates on all
the three-phase lines. Through
wiring terminal design of line.
of three-phase the voltage sensor,
4 denotes theitstransformer,
voltage
division ratio is set as 1500:1. 2 represents the high-voltage probe of P6051A, its voltage division
its rated capacity is 50 kVA, and the output voltage of transformer can be adjusted through the control
ratio is set as 1000:1. 3 indicates the wiring terminal of three-phase line. 4 denotes the transformer,
cabinet. 5 is the standard voltage transformer. 6 signifies the oscilloscope. 7 means the high-precision
its rated capacity is 50 kVA, and the output voltage of transformer can be adjusted through the
digital multimeter.
control cabinet. In 5 Figure 19b, 1 refers
is the standard voltageto the load part6composed
transformer. signifies theof oscilloscope.
resistive load 7 and
means capacitive
the
load, high-precision
and the load part is equipped with a large number of fans for heat dissipation
digital multimeter. In Figure 19b, 1 refers to the load part composed of resistive load and 2 represents
the control cabinetload,
and capacitive which andcanthecontrol
load part the transformation
is equipped rationumber
with a large of transformer.
of fans forStepless adjustment
heat dissipation
can beandrealized for thethe
2 represents voltage
controlwithin
cabinetthe range
which of control
can 0–35 kV. theMeanwhile,
transformationthe ratio
control cabinet can also
of transformer.
switchStepless
the load,adjustment
and thecan linebecurrent
realizedcan forchange
the voltage within
within the the range
range of 0–35AkV.
of 0–200 Meanwhile,
through the
the matching
control
relation cabinet
between can also
voltage and switch
load.the load,voltage
Thus and theandline current
current can change within
of three-phase the range
changes canofcoexist
0–200 Ain the
through
line and the matching
the purpose relation between
of simulating actualvoltage and load.
three-phase Thus voltage
transmission and
line current
will of three-phase
be realized. Moreover,
changes can coexist in the line and the purpose of simulating actual three-phase transmission line
the control cabinet can realize real-time display for voltage, current and power factor of different
will be realized. Moreover, the control cabinet can realize real-time display for voltage, current and
phases. Sectional area of the wire is 35 mm², the distance from the wire to the ground is 1.8 m, and the
power factor of different phases. Sectional area of the wire is 35 mm², the distance from the wire to
spacing
the between
ground is phase andthe
1.8 m, and phase is 1.38
spacing m. phase and phase is 1.38 m.
between

(a) (b)
Figure 19. Experimental platform structure: (a) experimental platform structure 1; and
Figure 19. Experimental platform structure: (a) experimental platform structure 1; and (b) experimental
(b) experimental platform structure 2.
platform structure 2.
5.1. Sensor Linearity Test
5.1. Sensor Linearity Test
Sensor linearity test aims to study whether the detection value of sensor and the voltage of
Sensor
conductorlinearity test aimsstill
to be measured to study
presentwhether the detection
linear changes when thevalue
use ofofsensor
sensorcauses
and the voltage
electric
field distortion.
of conductor to be measured still present linear changes when the use of sensor causes electric
The schematic diagram for experimental platform is shown in Figure 20. The high-voltage
field distortion.
probe is connected to A-phase, B-phase and C-phase wires, respectively; meanwhile, lines of the
The schematic diagram for experimental platform is shown in Figure 20. The high-voltage probe
three phases all pass through the non-contact voltage sensor. Output connection to oscilloscope is
is connected to A-phase, B-phase and C-phase wires, respectively; meanwhile, lines of the three phases
all pass through the non-contact voltage sensor. Output connection to oscilloscope is realized between
15
Sensors 2016, 16, 40 16 of 21
Sensors 2016, 16, 40

the high-voltage
realized betweenprobe and
the sensor, theprobe
high-voltage voltage
andof control
sensor, thecabinet
voltageisofadjusted, and waveform
control cabinet and
is adjusted, andphase
Sensors 2016, 16, 40
difference data are recorded every 1 kV within the range of 1–12 kV.
waveform and phase difference data are recorded every 1 kV within the range of 1–12 kV.
realized between the high-voltage probe and sensor, the voltage of control cabinet is adjusted, and
Voltage Sensor
waveform and phase difference data
Phaseare
A recorded every 1 kV within the range of 1–12 kV.
High-Voltage Probe
Phase B
Voltage Sensor
Phase A
Phase C
High-Voltage Probe
Phase B

Phase C

Voltage
Transformer
Regulator Power Load
Oscilloscope Oscilloscope
Voltage
Transformer
Regulator Power Load
Oscilloscope
Figure 20. Schematic diagram for the structure ofOscilloscope
linearity test platform.
Figure 20. Schematic diagram for the structure of linearity test platform.
Figure 21a shows
Figurethe
20. line voltage
Schematic waveform
diagram for thedetected
structureby high-voltage
of linearity probe under the voltage
test platform.
Figure
grade of21a shows
8 kV, and the line21b
Figure voltage
showswaveform
the detectiondetected
signal by high-voltage
waveform probe
of voltage under
sensor the the
under voltage
gradevoltagekV,grade
of 8Figureand21aof 8 kV.21b
Figure
shows According
the shows to Figure
the
line voltage 21, detection
detection
waveform signal signal
waveform
detected of the designed
of voltage
by high-voltage voltage
sensor
probe transformer
under
under the voltage
the voltage
has
of the
gradegrade ofsame
8 kV. kV,waveform
8 According
and Figure towith theshows
Figure
21b actual linedetection
21, detection
the voltage signal
signal detected
of the
signal by of
designed
waveform high-voltage probe,
voltagesensor
voltage and the
transformer
under the
has the
voltage
samevoltage
waveform division
gradewith ratio
of 8 kV. of voltage
the According transformer
actual linetovoltage
Figure 21,is 1500:1.
detection
signal signal
detected byofhigh-voltage
the designed voltage transformer
probe, and the voltage
Besides
has ratio
the same resistive
waveform load,
withcapacitive
the actual and
line inductive loadsdetected
voltage signal also exist
by in the part ofprobe,
high-voltage experimental
and the
division of voltage transformer is 1500:1.
platform
voltage design
division and
ratio platform
of voltage load implementation,
transformer is 1500:1. so there are some errors between phase and
Besides resistive load, capacitive and inductive loads also exist in the part of experimental
phaseBesides
in the three lines.load,
resistive Therefore, the difference
capacitive is notloads
and inductive exactly 120°.
also exist in the part of experimental
platform design and platform load implementation, so there are some errors between phase and phase
platform design and platform load implementation, so there are some errors between phase and
in the three lines. Therefore, the difference is not exactly 120˝ .
phase in the three lines. Therefore, the difference is not exactly 120°.

(a)

(a)

(b)
Figure 21. Waveform figure: (a) Signal of high-voltage probe under 8 kV; and (b) signal of voltage
sensor under 8 kV. (b)
Figure 21. Waveform figure: (a) Signal of high-voltage probe under 8 kV; and (b) signal of voltage
Figure 21. Waveform
Figure 22a shows figure: (a) Signal of high-voltage probe under 8 kV; and (b) signal of voltage
sensor under 8 kV. the output waveform of A-phase line high-voltage probe and non-contact
sensor
voltageunder 8 kV.
sensor under the voltage grade of 4 kV, and Figure 22b shows the output waveform of
A-phase line22a
Figure high-voltage
shows theprobe and
output non-contact
waveform voltage sensor
of A-phase under the voltage
line high-voltage probe grade of 6 kV.
and non-contact
voltage 22a
Figure sensor underthe
shows theoutput
voltagewaveform
grade of 4 of
kV,A-phase
and Figure
line22b shows the output
high-voltage probe waveform of
and non-contact
A-phase
voltage sensorline high-voltage
under probe
the voltage and of
grade non-contact
4 kV, andvoltage
Figuresensor underthe
22b shows theoutput
voltagewaveform
grade of 6 kV.
of A-phase
line high-voltage probe and non-contact voltage sensor under the voltage grade of 6 kV.
16

16
Sensors 2016, 16, 40 17 of 21
Sensors 2016, 16, 40

(a)

(b)
Figure 22. Comparison figure of high-voltage probe and voltage sensor: (a) signals comparison
Figure 22. Comparison figure of high-voltage probe and voltage sensor: (a) signals comparison under
under 4 kV; and (b) signals comparison under 6 kV.
4 kV; and (b) signals comparison under 6 kV.
The phase error of the three-phase line between these two under different voltage grades is
gained,
The as shown
phase error ofinthe
Table 5. In the line
three-phase table,between
φS indicates
thesethe
twooutput
undervoltage phase
different of voltage
voltage grades sensor
is gained,
and φ signifies the output voltage phase of high-voltage probe. According to Table
as shown in Table 5. In the table, ϕS indicates the output voltage phase of voltage sensor and ϕ H
H 5, voltage
sensor
signifies theworks
outputunder self-integration
voltage mode. According
phase of high-voltage probe.toAccording
Table 4, the maximum
to Table phasesensor
5, voltage error ofworks
voltage sensor and high-voltage probe is 33.9′ and the average error is 27.925′. The phase detection
under self-integration mode. According to Table 4, the maximum phase error of voltage sensor and
accuracy of voltage sensor under steady state meets the accuracy class of 0.5-level voltage sensors.
high-voltage probe is 33.91 and the average error is 27.9251 . The phase detection accuracy of voltage
sensor under steady state meets the accuracy Tableclass of 0.5-level
5. Phase error. voltage sensors.

Voltage (kV)
Table φH-φS (')
5. Phase error.
1 33.2
2
Voltage (kV) ϕ H30.7
-ϕS (')
3 25.4
14 33.2
23.9
2 30.7
5 25.7
3 25.4
46 22.3
23.9
57 31.7
25.7
68 32.6
22.3
79 25.4
31.7
810 22.8
32.6
911 25.4
33.9
1012 22.8
27.5
11 33.9
5.2. Sensor Measurement Accuracy Test 12 27.5

Sensor measurement accuracy test aims to verify the measurement accuracy grade of voltage
5.2. Sensor Measurement Accuracy Test
sensor, and it is the important index for the working performance of voltage sensor.
Sensor Themeasurement
experimental platform
accuracyoftest
simulating
aims tothree-phase
verify the transmission
measurement lineaccuracy
shown ingrade
Figureof
19voltage
is
still adopted, but the verification test is conducted by adopting the method
sensor, and it is the important index for the working performance of voltage sensor. of phase measurement.
Figure 23 is the schematic diagram for the structure of sensor measurement accuracy test platform.
The experimental platform of simulating three-phase transmission line shown in Figure 19 is
The standard voltage transformer adopts EVT1-20 electronic voltage transformer. The rated voltage
still adopted, but the verification test is conducted by adopting the method of phase measurement.
Figure 23 is the schematic diagram for the structure of sensor measurement accuracy test platform.
17
The standard voltage transformer adopts EVT1-20 electronic voltage transformer. The rated voltage of
primary side is 20 kV, the working frequency is 50 Hz, the rated phase shift is 0, the rated output of
Sensors 2016, 16, 40 18 of 21
Sensors 2016, 16, 40

of primary side is 20 kV, the working frequency is 50 Hz, the rated phase shift is 0, the rated output
secondary side is 0–4 V(AC), and the accuracy class is 0.2. 34410A high-performance digital multimeter
of secondary side is 0–4 V(AC), and the accuracy class is 0.2. 34410A high-performance digital
of Keysight Truevolt
multimeter series Truevolt
of Keysight is adopted as high-precision
series digital multimeter,
is adopted as high-precision and the measurement
digital multimeter, and the
accuracy of measuring
measurement the alternating
accuracy of measuringvoltage of 50 Hz voltage
the alternating and 0–50of V50isHz
0.006%. Therefore,
and 0–50 the above
V is 0.006%.
experimental
Therefore,equipment can guarantee
the above experimental the accuracy
equipment of the test
can guarantee theconclusion.
accuracy of the test conclusion.

Figure 23. Schematic diagram for the structure of accuracy test platform.
Figure 23. Schematic diagram for the structure of accuracy test platform.
The output ends of standard electronic voltage transformer and non-contact voltage sensor are
The outputtoends
connected of standard
digital electronic
multimeter, voltage voltage
the three-phase transformer and non-contact
is adjusted, voltage
and the output sensor
results of are
A-phase,
connected B-phase
to digital and C-phase
multimeter, standard voltage
the three-phase transformer
voltage as and
is adjusted, wellthe
as output
the output results
results of
of A-phase,
non-contact
B-phase voltage
and C-phase sensor voltage
standard designedtransformer
in this paper are recorded
as well every results
as the output 1 kV within the range voltage
of non-contact of
1–12 kV. The ratio error ε% between sensor signal and actual voltage signal is gained according
sensor designed in this paper are recorded every 1 kV within the range of 1–12 kV. The ratio error to ε%
the provisions in IEC60044-7 standard, and its definition is as follows:
between sensor signal and actual voltage signal is gained according to the provisions in IEC60044-7
standard, and its definition is as follows: %= KU S  MU H  100%
(20)
MU H
KUS ´ MUH
where K refers to the voltage divisionε%=ratio of voltage ˆ 100%designed in this paper and the rated (20)
sensor
MUH
transformation ratio after calibration is 1500:1; Us means the detection signal of voltage sensor; M
wheresignifies
K refersthetostandard electronic
the voltage voltage
division transformer
ratio of voltageand the transformation
sensor designed in ratio is 5000:1;
this paper and and UHrated
the
denotes the output voltage of standard electronic voltage transformer. The experimental results
transformation ratio after calibration is 1500:1; Us means the detection signal of voltage sensor; M are
shown
signifies theinstandard
Tables 6–8.
electronic voltage transformer and the transformation ratio is 5000:1; and U H
denotes the output voltage of standardTable
electronic voltage transformer. The experimental results are
6. Results of Phase A.
shown in Tables 6–8.
Voltage (kV) Us (kV) UH (kV) ε%
1 0.6582 0.1967A.
Table 6. Results of Phase 0.37
2 1.3245 0.3961 0.31
Voltage3(kV) Us1.9763
(kV) 0.5912
U H (kV) 0.28
ε%
4 2.6496 0.7929 0.24
1 0.6582 0.1967 0.37
25 3.4023
1.3245 1.0231
0.3961 0.21
0.31
36 4.0518
1.9763 1.2138
0.5912 0.14
0.28
47 4.6164
2.6496 1.3834
0.7929 −0.11
0.24
58 5.3271
3.4023 1.6010
1.0231 −0.17
0.21
69 4.0518
6.0846 1.2138
1.8301 0.14
−0.21
710 4.6164
6.6743 1.3834
2.0079 ´0.11
−0.23
811 5.3271
7.3257 1.6010
2.2027 ´0.17
−0.27
912 6.0846
8.0710 1.8301
2.4293 ´0.21
−0.33
10 6.6743 2.0079 ´0.23
11 7.3257 2.2027 ´0.27
12 8.0710 2.4293 ´0.33

18
Sensors 2016, 16, 40 19 of 21

Table 7. Results of Phase B.

Voltage (kV) Us (kV) U H (kV) ε%

1 0.6732 0.2011 0.41
2 1.3411 0.4008 0.37
3 2.0177 0.6030 0.39
4 2.6714 0.7994 0.25
5 3.3376 0.9987 0.26
6 4.0365 1.2088 0.18
7 4.6585 1.3943 0.23
8 5.3436 1.6014 0.17
9 6.0094 1.7990 0.21
10 6.6832 2.0106 ´0.28
11 7.3431 2.2080 ´0.23
12 7.9763 2.4018 ´0.37

Table 8. Results of Phase C.

Voltage (kV) Us (kV) U H (kV) ε%

1 0.6641 0.1998 ´0.31
2 1.3276 0.3994 ´0.27
3 1.9883 0.5985 ´0.33
4 2.6475 0.7958 ´0.21
5 3.3416 1.0062 ´0.31
6 4.0117 1.2010 0.21
7 4.6774 1.3992 0.27
8 5.3169 1.5901 0.31
9 6.0712 1.8154 0.33
10 6.6573 1.9918 0.27
11 7.3448 2.1955 0.36
12 7.9659 2.3829 0.29

According to the above two tests, some errors exist in detection signals of the non-contact voltage
sensor designed in this paper and standard electronic voltage transformer, and the errors mainly
come from two sources. On the one hand, though the sensor is specially designed, the position of
its electrode arrangement has inevitable errors due to the restriction of craftsmanship. As a result,
errors between transformer phase position and actual voltage phase are caused. However, such errors
can be ignored in actual use. On the other hand, complete coaxial arrangement cannot be realized
between the voltage sensor and line, so the electric field lines passing through sensor electrode are
imbalanced and the measurement accuracy will be affected. However, such errors can be ignored in
actual use. The voltage ratio error of the non-contact voltage sensor designed in this paper within the
voltage range of 0–12 kV is ε% < 0.5%, and the phase difference is ϕ < 351 . Therefore, the measurement
accuracy of such non-contact voltage sensor can meet the actual measurement requirement during
steady state measurement.

6. Conclusions and Prospect

According to this paper, the following conclusions can be gained:
(1) When non-contact voltage sensor is applied to steady state voltage measurement within 12 kV,
the phase error is smaller than 351 , which meets the phase detection accuracy class of 0.5-level voltage
transformer used for measurement.
(2) When non-contact voltage sensor is applied to steady state voltage measurement within 12 kV,
the ratio error ε% is <0.5%, so the accuracy of steady state measurement meets the accuracy class of
0.5-level voltage transformer used for measurement.
Sensors 2016, 16, 40 20 of 21

(3) Through simulation and experiment, structure design of sensor is optimized, voltage division
ratio of voltage sensor is increased, insulation design difficulty of sensor is reduced, and the sensor is
able to work under self-integration mode.
In the future, in-depth studies will be made into the aspects of standardization, anti-interference
and signal processing circuit of non-contact voltage sensor. In order to meet the electrified installation
requirement of sensor, open-type design should be conducted for the sensor electrode; in order to
solve the coaxial arrangement problem of sensor and line, a corresponding fixture should be designed.

Acknowledgments: This work was supported by the Funding Program from State Grid Chongqing Electric
Power Co. Electric Power Research Institute (SGCQDK00JZJS1400018).
Author Contributions: Wei He is the head of the research group that conducted this study. He contributed to
the research through his general guidance and advice. All authors contributed to the writing and revision of
the manuscript.
Conflicts of Interest: The authors declare no conflict of interest.

References
1. John-Paul, H.K.; Cheri, W. An Innovative Approach to Smart Automation Testing at National Grid.
In Proceedings of the 2012 IEEE PES Transmission and Distribution Conference and Exposition (T&D),
Orlando, FL, USA, 7–10 May 2012; pp. 1–8.
2. Liu, G.Y.; Kang, L.; Peng, W.N. On-Line Monitoring System for Transformer Partial Discharge.
Appl. Mech. Mater. 2013, 303, 464–467. [CrossRef]
3. Sarkar, B.; Koley, C.; Roy, N.K.; Kumbhakar, P. Condition monitoring of high voltage transformers using
Fiber Bragg Grating Sensor. Measurement 2015, 74, 255–267. [CrossRef]
4. Kaveri, B.; Saibal, C. Electric stresses on transformer winding insulation under standard and non-standard
impulse voltages. Electr. Power Syst. Res. 2015, 123, 40–47.
5. Zhao, S.P.; Li, H.Y.; Peter, C.; Forooz, G. Test and Analysis of Harmonic Responses of High Voltage Instrument
Voltage Transformers. IET 2014, 3, 1–6.
6. Cecati, C.; Citro, C.; Siano, P. Combined Operations of Renewable Energy Systems and Responsive Demand
in a Smart Grid. IEEE Trans. Sust. Energy 2011, 2, 468–476. [CrossRef]
7. Browning, C.A.; Vinci, S.J.; Zhu, J.; Hull, D.M.; Noras, M.A. An evaluation of electric-field sensors for
projectile detection. In Proceedings of the 2013 IEEE SENSORS, Baltimore, MD, USA, 3–6 November 2013;
pp. 1–4.
8. Farhangi, H. The path of the smart grid. IEEE Power Energy Mag. 2010, 8, 18–28. [CrossRef]
9. Pan, F.; Xiao, X.; Xu, Y.; Ren, Y. An optical AC voltage sensor based on the transverse pockels effect. Sensors
2011, 11, 6593–6602. [CrossRef] [PubMed]
10. Metwally, I.A. Comparative measurement of surge arrester residual voltages by D-dot probes and dividers.
Electr. Power Syst. Res. 2011, 81, 1274–1282. [CrossRef]
11. Metwally, I.A. D-dot probe before fast-front voltage measurement. IEEE Trans. Instrum. Meas. 2010, 59,
2211–2219. [CrossRef]
12. Harun, T.; Seddik, B.; Daniel, C.; Ahmad, H. Low-Voltage Loss-of-Life Assessments for a High Penetration
of Plug-In Hybrid Electric Vehicles. IEEE Trans. Power Deliv. 2012, 27, 1323–1331.
13. Wang, Y.A.; Xiao, D.M. Prototype design for a high-voltage high-frequency rectifier transformer for high
power use. IET Power Electron. 2011, 4, 615–623. [CrossRef]
14. Metwally, I.A. Coaxial D-dot probe: Design and testing. In Proceedings of the 1995 Annual Report Conference
on Electrical Insulation and Dielectric Phenomena, Virginia Beach, VA, USA, 22–25 October 1995; pp. 298–301.
15. Barker, P.P.; Mancao, R.T.; Kvaltine, D.J.; Parrish, D.E. Characteristics of lightning surges measured at metal
oxide distribution arresters. IEEE Trans. Power Deliv. 1993, 8, 301–310. [CrossRef]
16. Zhou, Q.; He, W.; Li, S.N.; Hou, X.Z. Research and Experiment of a Unipolar Capacitive Voltage Sensor.
Sensors 2015, 15, 20678–20697. [CrossRef] [PubMed]
Sensors 2016, 16, 40 21 of 21

17. Kubo, T.; Furukawa, T.; Itoh, H.; Hisao, W.; Wakuya, H. Numerical electric field analysis of power status
sensor observing power distribution system taking into account measurement circuit and apparatus.
In Proceedings of the 2011 IEEE SICE Annual Conference (SICE), Tokyo, Japan, 13–18 September 2011;
pp. 2741–2746.
18. Koziy, K.; Bei, G.; Aslakson, J. A low-cost power quality meter with series arc-fault detection capability for
smart grid. IEEE Trans. Power Deliv. 2013, 18, 1584–1591. [CrossRef]
19. Bayer, C.F.; Baer, E.; Waltrich, U.; Malipaard, D.; Schletz, A. Simulation of the electric field strength in the
vicinity of metallization edges on dielectric substrates. IEEE Trans. Dielectr. Electr. Insul. 2015, 22, 257–265.
[CrossRef]
20. Kubo, T.; Furukawa, T.; Fukumoto, H.; Ohchi, M. Numerical estimation of characteristics of voltage-current
sensor of resin molded type for 22 kV power distribution systems. In Proceedings of the ICCAS-SICE 2009,
Fukuoka, Japan, 18–21 August 2009; pp. 5050–5054.
21. Richard, C.J.; Michael, W.; Shi, G.Y. Efficient DC Fault Simulation of Nonlinear Analog Circuits: One-Step
Relaxation and Adaptive Simulation Continuation. IEEE Trans. Comput. 2006, 25, 1392–1400.
22. Wang, D.; Zhang, Y.R.; Wang, Y.H.; Lee, Y.S.; Lu, P.J.; Wang, Y. Cutting on Triangle Mesh: Local Model-Based
Haptic Display for Dental Preparation Surgery Simulation. IEEE Trans. Vis. Comput. Graph. 2005, 11, 671–683.
[CrossRef] [PubMed]

© 2015 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/).