sensors

Article
A Novel Transient Fault Current Sensor Based on the
PCB Rogowski Coil for Overhead Transmission Lines
Yadong Liu *, Xiaolei Xie, Yue Hu, Yong Qian, Gehao Sheng and Xiuchen Jiang
School of Electronic Information and Electrical Engineering, Shanghai Jiao Tong University, Shanghai 200240,
China; xiexiaolei0116@126.com (X.X.); yuehu@sjtu.edu.cn (Y.H.); qian_yong@sjtu.edu.cn (Y.Q.);
shenghe@sjtu.edu.cn (G.S.); xcjiang@sjtu.edu.cn (X.J.)
* Correspondence: lyd@sjtu.edu.cn; Tel.: +86-139-1733-2013

Academic Editor: Vittorio M. N. Passaro

Received: 15 March 2016; Accepted: 11 May 2016; Published: 21 May 2016

Abstract: The accurate detection of high-frequency transient fault currents in overhead transmission
lines is the basis of malfunction detection and diagnosis. This paper proposes a novel differential
winding printed circuit board (PCB) Rogowski coil for the detection of transient fault currents in
overhead transmission lines. The interference mechanism of the sensor surrounding the overhead
transmission line is analyzed and the guideline for the interference elimination is obtained, and then
a differential winding printed circuit board (PCB) Rogowski coil is proposed, where the branch and
return line of the PCB coil were designed to be strictly symmetrical by using a joining structure of
two semi-rings and collinear twisted pair differential windings in each semi-ring. A serial test is
conducted, including the frequency response, linearity, and anti-interference performance as well as
a comparison with commercial sensors. Results show that a PCB Rogowski coil has good linearity
and resistance to various external magnetic field interferences, thus enabling it to be widely applied
in fault-current-collecting devices.

Keywords: overhead transmission line; fault current; PCB Rogowski coil; differential winding; open
structure; high-frequency transient fault current

1. Introduction
The overhead transmission line is an important part of the power supply network. However,
long-distance overhead transmission lines are exposed to complicated environments and are easily
damaged by lightning strikes and foreign matter. This issue can cause a certain amount of transient
current, which contains considerable fault information and is the basis of fault location and diagnosis
(e.g., transient protection, fault analysis, and travelling wave fault location).
Analyses based on high-frequency components such as high-frequency protection, fault diagnosis,
travelling wave fault location, etc., are more popular than ever [1,2]. Therefore, the real-time transient
current monitoring of overhead transmission lines is the basis for fault detection and diagnosis [3].
Existing current-measuring devices include electromagnetic current sensors [4], Hall current sensors [5],
optical current sensors [6], and Rogowski coils.
Traditional electromagnetic current sensors cannot accurately measure high Ampere currents
because of the saturation problem of the magnetic core. A high-frequency transient fault current is
also difficult to measure because of the working frequency limit of the magnetic core. The Hall current
sensor is significantly influenced by the environmental temperature and has poor stability, so it is
difficult to capture the fault transient current. The newest optical current sensor based on the Faraday
rotation effect can measure both transient and direct currents. Although this new sensor is anticipated
to be the direction of current measurements in the future, it is difficult to use in power systems in the
short run because of its poor anti-interference properties and high cost.

Sensors 2016, 16, 742; doi:10.3390/s16050742 www.mdpi.com/journal/sensors

Sensors 2016, 16, 742 2 of 16

Sensors 2016, 16, 742 2 of 17

to be the direction of current measurements in the future, it is difficult to use in power systems in the
short run because of its poor anti-interference properties and high cost.
A
A Rogowski
Rogowski coil coil is
is the
the optimum
optimum measuring
measuring method method for for high-frequency
high-frequency transient
transient currents
currents at
at the
the
moment. However,parameter
moment. However, parameterconsistency
consistency is hard
is hard to guarantee
to guarantee because
because it is impossible
it is impossible for two for
randomtwo
random
Rogowski Rogowski coilsuniform
coils to have to have windings
uniform windings
and the same and cross-section
the same cross-section
between each between
turn, each
so as turn, so
a result,
as a result, a Rogowski coil can only be used after calibrations by an integrator
a Rogowski coil can only be used after calibrations by an integrator [7]. Whats more, the commercial [7]. Whats more, the
commercial
Rogowski coils Rogowski coils represented
represented by Pearson by (PaloPearson
Alto, CA,(PaloUSA),
Alto, PEM
CA, USA), PEM (Nottingham,
(Nottingham, UK), and Rocoil UK),
and Rocoil (Harrogate, UK) are mainly used as the sensor of precise instruments
(Harrogate, UK) are mainly used as the sensor of precise instruments are costly, thereby limiting their are costly, thereby
limiting their wide applications.
wide applications.
The
The printed
printed circuit
circuit board
board (PCB)
(PCB) Rogowski
Rogowski coil coil has
has attracted
attracted increasing
increasing attention
attention recently
recently because
because
of
of its characteristics of high accuracy, small volume, low cost, and easy mass production [813].
its characteristics of high accuracy, small volume, low cost, and easy mass production [813].
Considering
Considering that that overhead
overhead transmission
transmission lines lines cannot
cannot be be out
out of
of service
service once
once theythey are
are activated
activated anan open
open
structure sensor shall be designed to collect the transient fault currents of
structure sensor shall be designed to collect the transient fault currents of overhead transmission overhead transmission lines.
However, the poor
lines. However, electromagnetic
the poor electromagnetic environment
environment surrounding
surrounding overhead
overheadtransmission
transmissionlineslines also
also
requires
requires aaPCB PCBcoil
coiltotohave
havehighhigh anti-interference
anti-interference capability.
capability.In this paper,
In this a differential
paper, winding
a differential PCB
winding
Rogowski
PCB Rogowski coil was
coildesigned
was designed for open for structure
open structure[14]. A[14].
joint Astructure of two semi-rings,
joint structure as well as
of two semi-rings, as
collinear twisted pair differential windings in each semi-ring, were employed
well as collinear twisted pair differential windings in each semi-ring, were employed to eliminate to eliminate vertical
magnetic field interference,
vertical magnetic moreover,
field interference, moreover,the symmetric
the symmetric wiring of the
wiring of thePCB PCBwaswasusedusedtoto ensure
ensure
symmetrical
symmetrical windings and improve its parallel anti-interference characteristics. This open structure
windings and improve its parallel anti-interference characteristics. This open structure
PCB
PCB sensor
sensor cancan realize
realize hotline
hotline installation,
installation, with with high
high practicability.
practicability. As As aa result,
result, the
the designed
designed PCBPCB
sensor could facilitate the low-cost and large-scale production of transient
sensor could facilitate the low-cost and large-scale production of transient fault current sensors. fault current sensors.

2.
2. The
ThePrinciple
Principleofofthe
theRogowski
RogowskiCoil
Coil
The
The Rogowski
Rogowski coil consists
consists of
of two
two parts
parts (Figure
(Figure 1a).
1a). One
One isis the
thebobbin
bobbin made
made of
ofnon-permeability
non-permeability
magnetic
magnetic materials
materials (the
(the material
material permeability
permeability can
can be
be viewed
viewed approximately
approximately equal
equal to
to the
thepermeability
permeability
of
of vacuum
vacuum 00).). The
Thebobbin
bobbinhas
has aa uniform
uniform texture,
texture, and
and its
its cross
cross section
section is
is generally
generally rectangular
rectangular or
or
round. The other part is wound up on the frame
round. The other part is wound up on the frame tightly andtightly and evenly. The
The Rogowski coil induces aa
Rogowski coil induces
voltage signal that represents the testing current when flux changes.changes.

i(t) 2R
u(t)
h

d
(a) (b)
Figure
Figure 1.
1. Principle
PrincipleofofaaRogowski
RogowskiCoil.
Coil.(a)
(a)Structure
StructureDiagram
Diagramof
ofaaRogowski
Rogowskicoil;
coil;(b)
(b)Equivalent
Equivalent circuit
circuit
of a Rogowski coil.
of a Rogowski coil.

If the terminal resistance is Rs, the equivalent circuit of a Rogowski coil can be shown as Figure 1b,
If the terminal resistance is Rs , the equivalent circuit of a Rogowski coil can be shown as Figure 1b,
and the output of Rogowski coil can be [15,16]:
and the output of Rogowski coil can be [15,16]:
di t di t
e t Mdi 1ptq Lc di 2 ptq Rc i2 (t ) u t (1)
1 dt
Lc dt ` Rc i2 ptq ` u ptq
2
e ptq M (1)
dt dt
du t u (t )
i2 t C c du ptq uptq (2)
i2 ptq Cc dt ` R s (2)
dt Rs
Combining Equations (1) and (2) yields:
Sensors 2016, 16, 742 3 of 17

Combining Equations (1) and (2) yields:

d2 u ptq

di ptq Lc du ptq Rc
M 1 Lc Cc ` ` Rc Cc ` p1 ` qu ptq (3)
dt dt2 Rs dt Rs

where Cc is the distributed capacitance, Rc is the resistance inside the coil, Lc is the self-inductance of
coil, and M is the mutual inductance of coil. In most cases, Cc is very little and can be omitted, so the
output of Rogowski coil can be:

di2 ptq
e ptq Lc ` pRc ` Rs qi2 ptq (4)
dt
di2 ptq
So when Lc dt " pRc ` Rs qi2 ptq, which means Lc " Rc ` Rs , Equation (1) will be:

di1 ptq di2 ptq u ptq

M Lc Mi1 ptq Lc (5)
dt dt Rs

As the result the relationship between output voltage and input current is:

Lc
i1 ptq u ptq (6)
MRs

Output is proportional to the input, and the Rogowski coil is operating in self-integration model,
di ptq
and when Lc 2dt ! pRc ` Rs qi2 ptq, Equation (1) will be:

di1 ptq pRc ` Rs q u ptq
M pRc ` Rs qi2 ptq i1 ptq dt (7)
dt M Rs

At this time, the input current is equal to the integration of the output voltage, and the Rogowski
coil is operating on an external-integration model.

3. Analysis on the Interference Sources Surrounding Overhead Transmission Lines

The electromagnetic environment surrounding the overhead transmission line is complicated.
Thunder, lightning, discharges, and fault transit currents, except for the power frequency magnetic
fields (PFMFs) around overhead transmission lines, will all generate electromagnetic interferences
which can influence the measurement of a PCB Rogowski coil.
In order to estimate the magnetic field around the overhead transmission lines, [17] calculated the
magnetic field with five different conductor configurations. Assuming all the overhead transmission
lines were 22 m from the ground, the selected point was 1 m above the ground. The calculated
magnetic field distribution when the selected point from the overhead transmission lines is moved to a
farther point is shown in Figure 2, which shows that the PFMFs grow large when close to the overhead
transmission lines. For the transient fault current sensors installed on overhead transmission lines,
the monitoring object is the fault current which is several times larger than the nominal current and
will cause much more interference to the transient fault current sensor, hence, the interference from an
adjacent conductor is the main interference source for the transient fault current sensor, meanwhile,
the interference caused by discharge will affect the sensor at all directions.
 Sensors 2016, 16, 742 4 of 16
Sensors 2016, 16, 742 4 of 17
Sensors 2016, 16, 742 40 4 of 16
35
40
30
35
25

PFMF/T
30
1
20
25

PFMF/T
5
15 1
20 2 3
10 5
15
2 43
5
10
0 4
5
-100 -80 -60 -40 -20 0 20 40 60 80 100
0
-100 -80 -60
Distance
-40 -20
to the0line center/m
20 40 60 80 100

Distance to the line center/m

Figure2.2. Five
Figure Five PFMF
PFMF distributions
distributionsininthe
theUHV
UHVoverhead
overhead transmission
transmissionlineline
route at 1atm1height
route m heightfrom
the ground.
from the 1Parallel arrangement of phase conductor; 2Triangle arrangement of phase
Figure 2. ground.
Five PFMF 1Parallel arrangement
distributions in the UHV of phase conductor;
overhead 2Triangle
transmission line routearrangement of phase
at 1 m height from
conductor;3Common-tower
3Common-tower double-circuit overhead transmission line; 4Compact type
the ground. 1Parallel arrangement of phase conductor; 2Triangle arrangement of splits);
conductor; double-circuit overhead transmission line; 4Compact type (10 phase
(10 splits); 5Compact
5Compact type (12 splits).
type (12 splits).
conductor; 3Common-tower double-circuit overhead transmission line; 4Compact type
(10 splits); 5Compact type (12 splits).
3.1.Analysis
3.1. Analysison
onthe
theInterference
InterferenceofofAdjacent
AdjacentWire
Wire
Whenthe
3.1. Analysis
When theRogowski
on Rogowski
the coilis
Interference
coil ofisinstalled
installedWire
Adjacent ononthe
theoverhead
overheadtransmission
transmissionlineline(Phase
(PhaseA),
A),the
themagnetic
magnetic
fields generated
fieldsWhen
generated by the adjacent
by the adjacent line (Phases
line (Phases B and C)
B overhead will
and C) will also couple
also coupleline in
in thethe coil and
coil A),
andthe produce
produce an
an
the Rogowski coil is installed on the transmission (Phase magnetic
induced
induced electromotive
electromotive force as shown in Figure 3a. The output of the sensor is the sum of all the
fields generated by the force as shown
adjacent in Figure
line (Phases 3a. C)
B and The output
will of the sensor
also couple is the
in the coil sum
and of all the
produce an
induced voltagesproduced
produced bythe thecorresponding
correspondingconductors.
conductors. Inorder
ordertotoestimate
estimate theinterference
interference
induced voltages
electromotive force by as shown in Figure 3a. The outputInof the sensor is thethe sum of all the
fromthe
from theadjacent
adjacent line,phase
phaseBBisistaken
takenasasananexample
exampletotoanalyze
analyze inFigure
Figure3b.3b.
induced voltages line,
produced by the corresponding conductors. Ininorder to estimate the interference
from the adjacent line, phase B is taken as an example to analyze in Figure 3b.

A
B Phase

A
B Phase
A Phase

B A Phase

BC

(a) (b)
(a) of adjacent conductors. (a) Interferences of adjacent
Figure 3. Interference (b)phases; (b) Effect of the
Figure 3. Interference of adjacent conductors. (a) Interferences of adjacent phases; (b) Effect of the
Phase-B magnetic field.
Phase-B3.magnetic
Figure field.of adjacent conductors. (a) Interferences of adjacent phases; (b) Effect of the
Interference
Phase-B magnetic field.
The distance between Phase-A and Phase-B conductors is dAB as shown in Figure 3b. If one point
Q existsdistance
The between
on the ith turn of Phase-A
Rogowski and coil,
Phase-B which conductors
is r awayis from dAB astheshown
circleincenter,
Figureand
3b. Ifthe
one point
distance
The on
Qbetween
exists distance
the ithbetween
turn of Phase-A
Rogowski and Phase-B
coil, which conductors
is r away is
fromdAB as
the shown
circle in Figure
center, 3b.the
and If one point
distance
Phase-B conductor and point Q is dAB, the angle i is the included angle between r and dAB.
Q exists on
between the ith
Phase-B turn of Rogowski
conductor and pointlaw, coil,
Q isthe dwhich, theisangle
r away i isfrom the circleangle
the included center, and ther and
between distance
dAB .
According to Amperes circuit ABmagnetic induction intensity generated by the Phase-B
between Phase-B
According toconductor
Amperes and point
circuit law, Q is
the d AB, the angle i is the included angle between r and dAB.
magnetic induction intensity generated by the Phase-B
conductor on point Q is on the vertical line to line BQ, and only the magnetic induction intensity
According
conductor to Amperes
on point Q coil
is on circuit
the law, line
vertical the magnetic
to line BQinduction
, and onlyintensity
the generated by theintensity
Phase-B
perpendicular to the cross section, i.e., the tangential direction ofmagnetic
point Q, hasinduction
a contribution to
conductor
perpendicularon point
to the Q is
coil on
cross the vertical
section, i.e., line
the to line
tangential BQ, and only the magnetic induction intensity
direction of point Q, has a contribution to the
the induced electromotive force. The magnetic induction intensity component in the tangential
perpendicular
induced to the coil cross
electromotive section, i.e., induction
the tangential direction of pointinQ,thehas a contribution to
direction of point Q isforce. The
expressed magnetic
as follows: intensity component tangential direction
the induced
of point electromotive
Q is expressed force. The magnetic induction intensity component in the tangential
as follows:
direction of point Q is expressed asB ifollows: i 0 iB (t )
Bv ( t ) B B ( t ) cos i cos i (8)
2 li
i i 0 iB (t )
B Bv ( t ) B B ( t ) cos i cos i (8)
2 li
 Sensors 2016, 16, 742 5 of 17

i 0 i B ptq
BBv ptq BBi ptqcos i cos i (8)
Sensors 2016, 16,16,
742742 2li 5 of5 16
Sensors 2016, of 16
The total turns of a Rogowski coil is N. If a Rogowski coil is wound evenly and tightly, the induced
The total
electromotive
The turns
force
total of of
can
turns abeRogowski
a expressed
Rogowski coil
as
coil is is
the N. If If
a Rogowski
following
N. continuous
a Rogowski coil is is
wound
integral
coil evenly
form:
wound evenlyand tightly,
and thethe
tightly,
induced
inducedelectromotive
electromotive force
forcecan bebe
can expressed
expressed asasthethe following
following continuous
continuous integral form:
integral form:
2 2 2
2 22
di ptq di (t )h h NN ( R (Rd) d) dABdAB
2 pR ` dq `2 d 2(R2(Rd)ddAB
2 2pR ` )dcos
dqd cos

eptq e(te)B(t ) Bdi0B (t )0 0h lnN AB cos
AB
ln ln AB d d d (9)(9)
dt dt4 R 22 R`2ddAB
22 (9)
dt 442 0 202 R 2 2
dABAB
Rd
2 AB cos
2Rd
Rd ABAB
cos
cos
0
OnOn thethe
basis of of
basis numerical calculations
numerical calculationswith MATLAB,
with MATLAB, when
whendABdAB
> (R+d), e(t)=0.
> (R+d), This
e(t)=0. result
This shows
result shows
that On
asas the
long basis
asas of
thethenumerical
adjacent calculations
conductor with
is is MATLAB,
beyond thethe
PCBwhen d AB
Rogowski > (R+d),
coil e(t)=0.
and This
thethe result
winding shows
of thethe
that long adjacent conductor beyond PCB Rogowski coil and winding of
that Rogowski
PCB as long as coil
the adjacent
is conductor
uniform, the is beyond
interference the adjacent
from PCB Rogowski coil and
conductors can theomitted.
be winding of the PCB
PCB Rogowski coil is uniform, the interference from adjacent conductors can be omitted.
Rogowski coil is uniform, the interference from adjacent conductors can be omitted.
3.2.
3.2.Analysis of of
Analysis Vertical
VerticalExternal
External Magnetic
Magnetic Field Interference
Field Interference
3.2. Analysis of Vertical External Magnetic Field Interference
There areare
There various
various interfering
interfering magnetic
magnetic fields around
fields around thethe
sensor,
sensor,which
whichcan bebe
can decomposed
decomposed into
into
There are various interfering magnetic fields around the sensor, which can be decomposed into
BXB, XB,YBparallel to to
Y parallel thethe
plane
planexOyxOywhere
where thethe
bobbin
bobbin is is
located, and
located, andBZBperpendicular
Z perpendicularto to
thethe
plane xOy,
plane xOy,
BX , BY parallel to the plane xOy where the bobbin is located, and BZ perpendicular to the plane xOy,
asasshown
shownin in
Figure
Figure 4 The impact of of
BX and BYBon thethe
Rogowski coil is is
same asas
that of of
anan
adjacent
as shown in Figure 4 4The
The impact
impact of BXBXandand on
BY Yon Rogowski
the Rogowski coil
coil same
is same that
as that of an adjacent
adjacent
conductor,
conductor, so only
so onlythetheimpact
impact of B
ofZBneeds to be considered.
Z needs to be considered.
conductor, so only the impact of BZ needs to be considered.

Rogowski coil
Rogowski coil

Figure
Figure4.4.The
Figure 4.Theimpact
Theimpactofofof
impact orthogonal
orthogonalmagnetic
magnetic
orthogonal field.
field.
magnetic field.

BBz Bis parallel

paralleltototothe
thecross-section
cross-sectionwhere wherethe theRogowski
Rogowskicoil coilwaswaslocated,
located,thus,
thus,no noinduced
z zisisparallel the cross-section where the Rogowski coil was located, thus, no induced
induced
electromotive
electromotive force
force will be
will generated
be generated in the
in winding.
the winding. However,
However, the
electromotive force will be generated in the winding. However, the winding attaches on thethewinding
winding attaches
attaches on onthe coil
the coil
coil
bobbin
bobbin spirally,
bobbinspirally, thereby
spirally, thereby
thereby forming
forming
forming a structure
a structure
a structure similar
similarsimilarto an annular
to an annular
to an annular solenoid.
solenoid.solenoid. After
After the Afterthe coil
the coil
coil winds winds
winds
around
around
around the bobbin
the bobbinfor forone circle,
one an
circle, equivalent
an equivalent large turn
large turnis formed
is formed
the bobbin for one circle, an equivalent large turn is formed as shown in Figure 5a. as shown
as shown in Figure
in Figure 5a.
5a.

return line equivalent big turn

Equivalent turn
Equivalent turn return line equivalent big turn

return line
return line

(a)(a) (b)(b)
Figure
Figure
Figure 5.5.The
5.Thereturn
The return line
return lineofofof
line Rogowski
Rogowskicoil.
Rogowski
coil.(a)
coil. Equivalent
(a)
(a) large
Equivalent
Equivalent turn
large
large without
turn
turn return
without
without line;
return
return (b)
line;
line; Rogowski
(b)
(b) Rogowski
Rogowski
coil with
coil withthe return
the line.
return
coil with the return line. line.

Bz Bisz is
perpendicular
perpendicular to to
thetheplane
plane where
where large
large turns
turns areare
located.
located. When
When Bz Bchanges
z changes withwith time,
time, thethe
Bz is perpendicular to the plane where large turns are located. When Bz changes with time, the
magnetic flux ( ) will produce an induced electromotive force (e
magnetic flux (Z) will produce an induced electromotive force (ez) in the Rogowski coil. Although
Z z) in the Rogowski coil. Although
magnetic flux (Z ) will produce an induced electromotive force (ez ) in the Rogowski coil. Although
thethelarge
large turn
turn has has only
onlyoneone turn,
turn, itsitsarea
areais is
often
oftensignificantly
significantly larger
largerthan
than thethesectional
sectional area
area of of
windings. As a result, it will cause a significant error in the measured
windings. As a result, it will cause a significant error in the measured result. In practice, a return result. In practice, a return
line [18]
line [18]shall bebe
shall considered
considered in in
a practical
a practical Rogowski
Rogowski coil
coilto to
avoid
avoid thethe interference
interference fromfrom Bz Basz as
shown
shown in in
Figure
Figure 5b.5b. TheTheplaneplanewhere
where thethe
return
return line was
line waslocated
locatedwas wasparallel
parallel to to
plane
planexOy.
xOy. If the radius
If the radius of of thethe
circle formed by the return line is R and the induced electromotive
circle formed by the return line is Rre and the induced electromotive force generated by Bz on the
re force generated by B z on the
return
return line is is
line ere,erethe final
, the output
final output induced
induced electromotive
electromotive force
forceezeisz is
equal
equalto to
thethe
sumsum of of
ere eand
re and ez. eIfz. R
IfreRre
Sensors 2016, 16, 742 6 of 17

the large turn has only one turn, its area is often significantly larger than the sectional area of windings.
As a result, it will cause a significant error in the measured result. In practice, a return line [18] shall be
considered in a practical Rogowski coil to avoid the interference from Bz as shown in Figure 5b. The
plane where the return line was located was parallel to plane xOy. If the radius of the circle formed
by the return line is Rre and the induced electromotive force generated by Bz on the return line is ere ,
the final output induced electromotive force ez is equal to the sum of ere and ez . If Rre is determined
equal to the equivalent radius of the large turn (Req ), ere is equal to ez in numerical value but has an
opposite direction. This opposite direction can be attributed to the opposite current in the return line
to that of the equivalent large turn. Therefore, the effect of the vertical magnetic field on Rogowski coil
performance can be effectively eliminated by adding return lines.

4. Design of Differential Winding PCB Rogowski Coil

For the fault current sensors installed on the overhead transmission line, the work temperature
ranges from 40 C to 85 C, so the shape of the bobbin could change with the temperature and cause
measurement errors. Meanwhile, the overhead transmission line requires the sensor to be installed
without power interruption, so an open sensor structure is indispensable, so the main task to develop
a transient fault current sensor based on the PCB Rogowski coil is how to eliminate the errors caused
by temperature and interference.

4.1. Material Selection for the Rogowski Coil Bobbin

The coefficient of thermal expansion was considered when selecting the materials for the Rogowski
coil bobbin. The coefficients of thermal expansion of some common materials are listed in Table 1 [19].

Table 1. Coefficients of (linear) thermal expansion of common bobbin materials.

Material Rubber Polyethylene Epoxy Resin Ceramic

Coefficient of expansion/(106 / C) 200300 126160 6080 1.07.7

Rubber and polyethylene have significantly high thermal expansion coefficients, which will cause
large temperature errors in environments where the working temperature changes greatly. Although
ceramics have a significantly low coefficient of thermal expansion, they break easily and have a low rate
of finished products. Long-term vibrations on overhead transmission lines and sudden electrodynamic
forces on short circuits also occur; these phenomena can destroy the ceramic bobbins easily. As a result,
epoxy resin was selected as the base material of our PCB.

4.2. Differential Winding of the Rogowski Coil

As stated before, if the Rogowski coil only has a signal line, the equivalent large turn is vulnerable
to vertical magnetic field interference. A return line is indispensable to eliminate this interference.
However, the sensor should be an open structure, so how to design the return line on an open structure
sensor is a significant problem for fault current detection. Differential winding, which can achieve
the same effect as return lines (Figures 6 and 7) is chosen to eliminate this interference. Rows of vias
were employed in each semi-ring. The same conductor was folded into two coils that twined on the
bobbin towards the same direction in each semi-ring; this process is known as collinear twisted pair
differential winding. One coil was defined as clockwise winding, and the other was counterclockwise
winding. Given that the currents in these two coils flow in the opposite direction, a 180 difference
exists in their phase position. Two coils were wound tightly in a twisted pair to ensure that the branch
and return line were strictly symmetric. The generated coupling magnetic fluxes had the same absolute
value but opposite directions. They offset each other, thus neutralizing the common-mode signal and
eliminating vertical magnetic field interference.
Sensors 2016, 16, 742 7 of 17

Two rows of vias were set on one side of the external ring of the semicircular printed board,
and three rows of vias were set on one side of the internal ring. They were marked from external to
internal as Rows N, M, C, B and A. The distances between these rows of vias and the circle center were
recorded as RN , RM , RC , RB and RA . One end of the semi-ring comprised the starting via X and end via
Y of the turn. External Rows N and M had 58 vias, which were uniformly distributed at an equivalent
radian interval of /58. Internal Rows C, B, and A had 39 vias, which were uniformly distributed at
an equivalent radian interval of /39. The copper foil connections of turns in this design are shown
Sensors
in 2016,7.16, 742
Figure 7 of 16
Sensors 2016, 16, 742 7 of 16

Figure 6. Structure of rows of vias.

Figure 6. Structure of rows of vias.

Figure 7. Connections of turns.

Figure 7. Connections of turns.
Figure 7. Connections of turns.
In Figure 7, the full lines are front connections and the dotted lines are back connections. The
In Figure 7, the full lines are front connections and the dotted lines are back connections. The
hollow dots are the full
vias on the are clockwise winding,and
and the solid lines
dots areback
the vias on the
hollow Figure
In 7, the
dots are the vias lines
on the front connections
clockwise winding, andthethe
dotted
solid dots areare the connections.
vias on the
counterclockwise
The hollow dots are winding.
the viasThe on clockwise winding
the clockwise and and
winding, counterclockwise
the solid dots winding form
are the vias on one
the
counterclockwise winding. The clockwise winding and counterclockwise winding form one
differential line pair.
counterclockwise The current
winding. The in the counterclockwise
clockwise winding and winding was the
counterclockwise complementary
winding form one signal on
differential
differential line pair. The current in the counterclockwise winding was the complementary signal on
the clockwise
line winding.
pair. The current in The left semi-ring waswinding
the counterclockwise taken as an example. The vias from toponto bottom were
the clockwise winding. The left semi-ring was taken aswas the complementary
an example. signal
The vias from top tothe clockwise
bottom were
marked as Np, Mp, Cq, Bq, and Aq, where p ranged from 1 to 58 and q ranged from 1 to 39. In this
marked as Np, Mp, Cq, Bq, and Aq, where p ranged from 1 to 58 and q ranged from 1 to 39. In this
way, the clockwise winding can be expressed as X-A1-M1-C1-M2-B2-M3-A3-M4-C3, , and the vias
way, the clockwise winding can be expressed as X-A1-M1-C1-M2-B2-M3-A3-M4-C3, , and the vias
of Row M were connected with Rows A, B, and C successively until M58 was connected with C39.
of Row M were connected with Rows A, B, and C successively until M58 was connected with C39.
On the backboard, C30 was connected with N58 and counterclockwise winding began at N58. The
On the backboard, C30 was connected with N58 and counterclockwise winding began at N58. The
counterclockwise winding was from bottom to top, but the turns remained in the same direction as
counterclockwise winding was from bottom to top, but the turns remained in the same direction as
Sensors 2016, 16, 742 8 of 17

winding. The left semi-ring was taken as an example. The vias from top to bottom were marked as Np,
Mp, Cq, Bq, and Aq, where p ranged from 1 to 58 and q ranged from 1 to 39. In this way, the clockwise
winding can be expressed as X-A1-M1-C1-M2-B2-M3-A3-M4-C3, . . . , and the vias of Row M were
connected with Rows A, B, and C successively until M58 was connected with C39. On the backboard,
C30 was connected with N58 and counterclockwise winding began at N58. The counterclockwise
winding was from bottom to top, but the turns remained in the same direction as the clockwise
winding. The counterclockwise winding was N58-B39-N57-C39-N56-A38, . . . , and the vias of Row N
were connected with those of Rows A, B, and C successively until B1 was connected with Y. When the
front Row
Sensors 2016, N
16,was
742 connected with Rows A, B, and C, it reached across the external side of Row M along 8 of 16
a part of small arc. When Back Row M was connected with Rows A, B, and C, it also reached across the
external
had side of
an equal Row and
radius, N along a part
the total of awas
span small arc. The
equal. The front
entireand
coilback
was arcs had anof
composed equal
two radius, and
completely
the total span
identical was equal.
semi-rings ratherThe
than entire
two coil was composed
mirrored semi-rings. ofThe
twoleft
completely
semi-ringidentical semi-rings
can be the right onerather
after
than twohorizontally.
flipping mirrored semi-rings.
The frontThe left semi-ring
wiring of the leftcan be the right
semi-ring one same
was the after flipping horizontally.
as the back wiring of Thethe
front semi-ring.
right wiring of theAs left semi-ring
a result, was the
two pieces same as
of signal theare
lines back wiring
twisted of the
back andright
forth,semi-ring. As aRresult,
thus making re and

two
R piecesequal.
strictly of signal lines arethe
However, twisted
signalback
phase and forth,
were thus making
completely Rre andThe
opposite. R strictly
inducedequal. However,
electromotive
the signal
forces phase were
produced by thecompletely opposite. The
vertical disturbing induced
magnetic electromotive
field forces produced
in the equivalent large turnbyofthe vertical
clockwise
disturbing
winding and magnetic field in large
the equivalent the equivalent large turn of clockwise
turn of counterclockwise windingwinding
had the andsamethe equivalent
absolute valuelarge
but
turn of counterclockwise winding had the same absolute value but opposite direction.
opposite direction. They offset each other, thus, the disturbance of the vertical disturbing magnetic They offset
each was
field other, thus, the disturbance of the vertical disturbing magnetic field was eliminated.
eliminated.

4.3. Parameters
4.3. Parameters of
of the
the Rogowski
Rogowski Coil
Coil
Figure 8
Figure 8 illustrates
illustrates aa physical
physical Rogowski
Rogowski coil
coil composed
composed of
of two
two semi-rings. The geometric
semi-rings. The geometric
parameters are presented in Table
parameters are presented in Table 2.2.

Figure 8. Physical picture of the designed Rogowski coil.

Figure 8. Physical picture of the designed Rogowski coil.

Table 2. Geometric parameters of the PCB Rogowski coil.

Table 2. Geometric parameters of the PCB Rogowski coil.
Inner Radius of Outer Radius of
Inner Outerthe Bobbin Thickness RA RB RC RM RN
the Bobbin
Radius of Radius of Thickness RA RB RC RM RN
24.0
the Bobbin the Bobbin 39.0 3.0 25.5 26.5 27.5 36.5 38.0
mm
24.0 mm 39.0 mm mm3.0 mm mm
25.5 mm mm
26.5 mm mm
27.5 mm mm
36.5 mm mm
38.0 mm mm

Three rows of internal vias of the PCB coil were successively connected to two rows of external
Three rows of internal vias of the PCB coil were successively connected to two rows of external
vias to form six specifications of small turns (Table 3).
vias to form six specifications of small turns (Table 3).
Table 3. Specifications of small turns.

Turn Type 1 2 3 4 5 6
Outer Row M M M N N N
Outer Row A B C A B C
Turns 38 38 40 38 38 40
Inner radius (mm) 25.5 26.5 27.5 25.5 26.5 27.5
Outer radius (mm) 36.5 36.5 36.5 38.0 38.0 38.0
Sensors 2016, 16, 742 9 of 17

Table 3. Specifications of small turns.

Turn Type 1 2 3 4 5 6
Outer Row M M M N N N
Outer Row A B C A B C
Turns 38 38 40 38 38 40
Inner radius (mm) 25.5 26.5 27.5 25.5 26.5 27.5
Outer radius (mm) 36.5 36.5 36.5 38.0 38.0 38.0

The connection between Y and Y1 was identified as turn type 5, i.e., the turn formed by NB.
The total mutual inductance [20] of the coil was the sum of the mutual inductance of different types
of turns:
i
6 6
h R
M Mi Ni 0 ln Oi (10)
2 R Ii
i 1 i 1
Sensors 2016, 16, 742 9 of 16
where Mi is the mutual inductance of the ith type of small turn; Ni is the number of turns of the ith
where
type Mi is the
of small turn; ROi and
mutual RIi are the
inductance of the
outer ithradius
type ofand small turn;
inner Ni is of
radius thethe ith type
number of turns of the
of small ith
turn,
type of small
respectively. Theturn; ROi andofRmultiple
structure Ii are therows
outerofradius and inner
vias realized radius of the ith
counterclockwise type ofofsmall
winding turn,
coil and
respectively.
made effective The structure
use of of on
the space multiple
the PCB. rows
Theofself-inductance
vias realized counterclockwise
can be expressed as winding
follows:of coil and
made effective use of the space on the PCB. The self-inductance can be expressed as follows:
i
6 6
0 h ROi
LC i 6Ni Mi 6 Ni2 2 hln R (11)
LC N i M i N i 20 lnR Ii Oi (11)
ii11 ii 11 2 R Ii
The
Themain
mainfrequency
frequencycomponent
componentofofthe
thefault
faultcurrent
currentranges
rangesfrom
from5050Hz
Hztotoseveral
severalhundred
hundredkHz,
kHz,
sosothe work frequency band of the sensor should cover most of the frequency band. A
the work frequency band of the sensor should cover most of the frequency band. A T-integrating T-integrating
circuit
circuitwas
wasdesigned
designedfor
forthe
thePCB
PCBRogowski
Rogowskicoil
coilasasshown
shownininFigure
Figure9.9.

C0

R0

Rf Rf
C
R
Vout
+

Figure 9. The schematic diagram of integrator.

Figure 9. The schematic diagram of integrator.

The transfer function of Rogowski coil can be obtained from Figure 1b, so the total transfer
functiontransfer
The function
of the sensor wasof Rogowski coil can be obtained from Figure 1b, so the total transfer
[21,22]:
function of the sensor was [21,22]:
Vout s T s 1 K Ms
H V
s out
psq K0 T 0 s 0` 1 2 2 2 Kcc Ms (12)
H psq I1
s K0 Ts 12 Tc2 s2 2Tc s 1 (12)
I1 psq pTs ` 1q Tc s ` 2Tc s ` 1

2R2 R 11 11 RR L L R`RRCR C

f f s s
K0K , T,0T
0 RRf f `RR00 C
C00 ,, T RRf fCC ` RR0 0CC0 0, ,KKcc , T,cT
c KcK
LccC , c , ac c c s c c s c (13)
a
0 T Lcc C (13)
R R 22 22 R R
Rc c ` sRs 2 2R RRc pR
R s R
c` LcsC
q cLc Cc
 Vout s T0 s 1 Kc Ms
H s K0 (12)
I1 s Ts 1
2 2 2
T s 2Tc s 1
c

2R f 1 1 Rs Lc Rc Rs C c
K0 , T0 R f R0 C 0 , T R f C R0 C 0 , K c , Tc K c Lc C c , (13)
Sensors 2016, R 742 2
16, 2 R c Rs 2 R Rc Rs Lc C c 10 of 17

Figure 10.
Figure 10. The
The frequency
frequency response
response of
of the
the proposed
proposed sensor.
sensor.

If the appropriate parameters of Rf , Rs , R, C and C0 are selected, the corresponding performance

can be obtained. If the work frequency bandwidth of the sensor was 30 Hz to 1 MHz, the parameters
of the integrator are listed in Table 4, and the corresponding frequency response is shown in Figure 10.

Table 4. Parameters of the PCB Rogowski Coil sensor.

Parameters Values Parameters Values Parameters Values

RC 46.3 M 47.346 nH R0 2 M
LC 1.8283 H Rs 20 M C0 22 F
CC 20.12 pF Rf 1 k C 1 F

5. Performance Test the PCB Rogowski Coil

In order to verify the performance of the proposed sensor, a serial test, including the frequency
response, linearity, anti-interference performance as well as a comparison with commercial sensors
were conducted.

5.1. Linearity Test

An experimental platform was established to test the high-frequency transient current
measurement accuracy of the proposed PCB Rogowski coil (Figure 11a). The impulse current generator
can produce 820 s impulse currents to a peak value 50 kA. The output of the impulse current
generator and the output of the PCB Rogowski coil were connected onto Channels 1 and 2 of a DPO
5204 digital oscilloscope, respectively. When the impulse current is 10,160 A, the output of the PCB
coil is presented in Figure 11b. Figure 11b shows that the output of the PCB coil is almost the same as
the input waveform, so the PCB coil has a sufficient bandwidth to detect the fault current.
generator can produce 820 s impulse currents to a peak value 50 kA. The output of the impulse
current generator and the output of the PCB Rogowski coil were connected onto Channels 1 and 2 of
a DPO 5204 digital oscilloscope, respectively. When the impulse current is 10,160 A, the output of the
PCB coil is presented in Figure 11b. Figure 11b shows that the output of the PCB coil is almost the
Sensors 2016, 16, 742 11 of 17
same as the input waveform, so the PCB coil has a sufficient bandwidth to detect the fault current.

Impulse
current Intergator oscilloscope
generator

Rogowski coil

(a) (b)
Figure
Figure 11.
11.The
Theverification
verificationof
ofthe
thePCB
PCBRogowski
RogowskiCoil.
Coil.(a)
(a)Experimental
Experimentallinearity
linearitycircuit;
circuit;(b)
(b)Waveform
Waveform
comparison between the input and PCB coil output.
comparison between the input and PCB coil output.

Figure 12 shows the output voltage of the PCB Rogowski coil under different input currents.
Figure 12 shows the output voltage of the PCB Rogowski coil under different input currents.
From Figure 12, the linearity of the PCB Rogowski coil was no more than 2% under a 820 s lightning
From Figure 12, the linearity of the PCB Rogowski coil was no more than 2% under a 820 s lightning
impulse wave in a 12,000 A range.
impulse
Sensors wave
2016, in a 12,000 A range.
16, 742 11 of 16

Figure
Figure 12.
12. Linearity
Linearity test
test fitting
fitting curve.
curve.

5.2. Test of Adjacent Wire Interference

5.2. Test of Adjacent Wire Interference
The output line of the impulse current generator was extended to simulate the effect of adjacent
The output line of the impulse current generator was extended to simulate the effect of adjacent
wires on the PCB Rogowski coil. The plane where the PCB Rogowski coil was located was
wires on the PCB Rogowski coil. The plane where the PCB Rogowski coil was located was
perpendicular to the wire (Figure 13a). The amplitude of the impulse current generator was set at
perpendicular to the wire (Figure 13a). The amplitude of the impulse current generator was set
2310 A, and the distance between the wire and coil was controlled at 3 and 6 cm. The inputs and
at 2310 A, and the distance between the wire and coil was controlled at 3 and 6 cm. The inputs and
outputs of the PCB Rogowski coil were recorded (Figure 13b). The different directions of coil and
outputs of the PCB Rogowski coil were recorded (Figure 13b). The different directions of coil and wire
wire were further changed to test the outputs under different conditions (Table 5).
were further changed to test the outputs under different conditions (Table 5).

y
Impulse I
x
current z
generator

Rogowski coil
 The output line of the impulse current generator was extended to simulate the effect of adjacent
wires on the PCB Rogowski coil. The plane where the PCB Rogowski coil was located was
perpendicular to the wire (Figure 13a). The amplitude of the impulse current generator was set at
2310 A, and the distance between the wire and coil was controlled at 3 and 6 cm. The inputs and
outputs of the PCB Rogowski coil were recorded (Figure 13b). The different directions of coil and
Sensors 2016, 16, 742 12 of 17
wire were further changed to test the outputs under different conditions (Table 5).

y
Impulse I
x
current z
generator

Rogowski coil

Intergator oscilloscope

(a) (b)
Figure 13. The test of adjacent wire interference. (a) Testing diagram of adjacent wire interference;
Figure 13. The test of adjacent wire interference. (a) Testing diagram of adjacent wire interference;
(b) The waveform comparison at 2002 A and 3 cm.
(b) The waveform comparison at 2002 A and 3 cm.

Table 5. Interference test of adjacent wire.

Table 5. Interference test of adjacent wire.
Coil Flat Position xz xy yz
742 dAB
Sensors 2016, 16, Coil (cm)
Flat Position 3 xz 6 3 xy 6 3 yz 6 12 of 16
Impulse current
dAB (cm) (A) 19693 19696 19013 19016 20023 2002
6
was generated in Channel
Peak Impulse
value of current1,(A)
channel 1while
(V) the 94.4
Rogowski
1969 coil output
94.4
1969 91.2
1901was only
91.2
1901120 mV. 96 Hence, 2002
2002 the
96 Rogowski
coil with two
Peak
Peak semi-rings
value
value of of could
channel
channel resist the
2 1(mV)
(V) interference
94.4
220 94.4 of adjacent
220 91.2 wires
140 91.2well. 120
140 96 96
160
Peak value of channel 2 (mV)
Error% 220
0.233 220
0.233 140
0.153 140
0.153 120
0.125 160
0.167
5.3. Test of Vertical Error% 0.233
Magnetic Field Interference 0.233 0.153 0.153 0.125 0.167
In Figure 13b, the PCB coil was 3 cm away from the wire, and the coil plane was perpendicular
The impulse current generator was used to produce an 8/20 interference current waveform and
to In
theFigure
wire. When
13b, thea 2002
PCB A impulse
coil current
was 3 cm awaywas setthe
from in wire,
the adjacent
and thewire, a 96.0 was
coil plane V impulse voltage to
perpendicular
to test the effect of vertical wire on the PCB Rogowski coil. A wire that transmitted the interference
the wire. When a 2002 A impulse current was set in the adjacent wire, a 96.0 V impulse voltage was
current was wound on a cylinder to form a solenoid. This solenoid would generate a uniform
generated in Channel 1, while the Rogowski coil output was only 120 mV. Hence, the Rogowski coil
magnetic field perpendicular to the horizontal plane, as shown in Figure 14a. The impulse current
with two semi-rings could resist the interference of adjacent wires well.
was set to 1030, 1565 and 2020 A. The vertical disturbing magnetic field was put perpendicular to the
PCBTest
5.3. Rogowski coil
of Vertical (FigureField
Magnetic 14a).Interference
The outputs of a conditioning circuit were measured (Table 6).

The impulse currentTablegenerator

6. Testwas used
results to produce
of vertical an 8/20
magnetic fieldinterference
interference. current waveform and
to test the effect of vertical wire on the PCB Rogowski coil. A wire that transmitted the interference
current was woundImpulse current (A)
on a cylinder to form a solenoid. This1030solenoid would 1565generate a uniform
2020 magnetic
Peak value
field perpendicular of channel
to the 1 (V)plane, as shown
horizontal 48in Figure 14a.64The impulse current94 was set
Peak value of channel 2 (mV) 420 450
to 1030, 1565 and 2020 A. The vertical disturbing magnetic field was put perpendicular to the PCB 480
Rogowski coil (Figure Error%
14a). The outputs of a conditioning0.875 circuit were0.7measured (Table 0.51
6).

Impulse
current
generator

Rogowski coil Integrator oscilloscope

(a) (b)
Figure14.
Figure 14.The
Thetest
testofofvertical
vertical magnetic
magnetic field
field interference.
interference. (a) Testing
(a) Testing diagram
diagram of vertical
of vertical magnetic
magnetic field
field interference; (b) The waveform comparison at vertical magnetic
interference; (b) The waveform comparison at vertical magnetic field interference.field interference.

According to Table 6 and Figure 14b, when the interference current increased from 1030 A to
2000 A, the output of the Rogowski coil was about 450 mV. Thus, the designed PCB Rogowski coil
had good resistance to the vertical magnetic field interference.

5.4. Test of Frequency Response

 Sensors 2016, 16, 742 13 of 17

Table 6. Test results of vertical magnetic field interference.

Impulse current (A) 1030 1565 2020

Peak value of channel 1 (V) 48 64 94
Peak value of channel 2 (mV) 420 450 480
Error% 0.875 0.7 0.51

According to Table 6 and Figure 14b, when the interference current increased from 1030 A to
2000 A, the output of the Rogowski coil was about 450 mV. Thus, the designed PCB Rogowski coil had
good resistance to the vertical magnetic field interference.

5.4. Test of Frequency Response

In order to test the frequency response of the proposed sensor, 50 Hz, 8/20 s and 30/80 s
currents were selected as the current source, so the frequency could cover 50 Hz to 43.75 kHz. The
parameters of the current source were listed in Table 7 and the test results were listed in Table 8, while
the output waveforms of the PCB Rogowski coil with different source were listed in Figures 1517. The
output waveforms are almost the same with the standard current source, meanwhile, the input, the
output ratio is 1185:1, 1197:1 and 1208:1, corresponding to the 50 Hz, 8/20 s and 30/80 s, respectively,
so the work frequency bandwidth covers the bandwidth of the three sources.

Table 7. Models and equipment parameters.

Parameters Source 1 Source 2 Source 3

Current parameters 50 Hz 8/20 (s) 30/80 (s)
Source models PY002 KV2103-G-020 MWG001
Sensors 2016, 16, 742 13 of 16
Standard CT FCT 300/50 150 A/V Pearson 4997 100 A/V Pearson 1080 200 A/V

Figure 15. 50 Hz output waveforms.

Figure 15. 50 Hz output waveforms.
Sensors 2016, 16, 742 14 of 17
Figure 15. 50 Hz output waveforms.
Figure 15. 50 Hz output waveforms.

Figure 16. 8/20 impulse current output waveforms.

Figure 16.8/20
16.
Figure 8/20 impulse currentoutput
impulse current outputwaveforms.
waveforms.

Figure 17. 30/80 impulse current output waveforms.

Figure 17. 30/80 impulse current output waveforms.
Figure 17. 30/80 impulse current output waveforms.

Table 8. Results of the frequency experiments.

Current Waveforms 50 Hz 8/20 (us) 30/80 (us)

Standard CT outputs (V) 2.77 41.42 20.78
Input current (A) 415 4142 4156
PCB R-coil outputs (V) 0.35 3.46 3.44
PCB R-coil ratio 1185:1 1197:1 1208:1

5.5. Performance Comparation Between the Commecial Sensors

In order to compare the performance among the newly designed sensor and commercial ones,
two commercial sensors, a Pearson 4997 and a FCT 200 were selected. Their main specifications are
shown in Table 9.

Table 9. The main specifications of two commercial sensors.

Sensor Pearson Model 4997 Normal Rogowski Coil FCT 200

I/O ratio 100 A/1V 500 A/1V
Bandwidth (3 dB) 0.5 Hz15 MHz 1 Hz1 MHz
Accuracy 1% 1%
 shown in Table 9.

Table 9. The main specifications of two commercial sensors.

Sensor Pearson Model 4997 Normal Rogowski Coil FCT 200

I/O ratio 100 A/1V 500 A/1V
Sensors 2016, 16, 742 15 of 17
Bandwidth (3 dB) 0.5 Hz15 MHz 1 Hz1 MHz
Accuracy 1% 1%
After setting the input current range from 1000 A to 5000 A, and recording the output of different
After setting the input current range from 1000 A to 5000 A, and recording the output of different
sensors, Figure 18 shows that when the input current was 4000 A, the output waveform of the proposed
sensors, Figure 18 shows that when the input current was 4000 A, the output waveform of the
sensor was almost
proposed sensorthewassame as the
almost the output
same as ofthethe Pearson
output 4997
of the and FCT
Pearson 4997 200
and but
FCTfor
200the
butoutput
for the ratio.
If theoutput
Pearson 4997 is considered as the standard, the outputs among the three sensors are
ratio. If the Pearson 4997 is considered as the standard, the outputs among the three sensors listed in
Tableare
10.listed
From in Table 10. From Table 10, the linear correlation coefficient of the PCB Rogowski coil is and
Table 10, the linear correlation coefficient of the PCB Rogowski coil is 0.99959,
that of the FCT
0.99959, and 200
that is
of 0.99969, so there
the FCT 200 is nearly
is 0.99969, noisdifference
so there between the
nearly no difference proposed
between sensor and
the proposed
sensor and
commercial commercial ones.
ones.

(a) (b)
Figure 18. Output waveforms of 820 s impulse current. (a) PCB R-coil and Pearson output
Figure 18. Output waveforms of 820 s impulse current. (a) PCB R-coil and Pearson output
waveforms; (b) PCB and Normal R-coil output waveforms.
waveforms; (b) PCB and Normal R-coil output waveforms.
Table 10. Outputs of different sensors.
Table 10. Outputs of different sensors.
Setting Input Current Pearson 4997 Output PCB R-Coil Output FCT 200 Output
No.
(A) (V) (V) (V)
Setting Input Pearson 4997 PCB R-Coil FCT 200
1 No. 1000 12.0 1.00 2.54
Current (A) Output (V) Output (V) Output (V)
2 2000 23.3 1.98 4.62
3 1 30001000 12.0
31.9 1.00
2.66 2.54
6.52
4 2 40002000 23.3
41.8 1.98
3.46 4.62
8.49
5 3 50003000 31.9
51.7 2.66
4.38 6.52
10.41
4 4000 41.8 3.46 8.49
5 5000 51.7 4.38 10.41

Sensors 2016, 16, 742 15 of 16

6. Field Application
6. Field Application
The newly designed sensor has been implemented in an overhead transmission line online
monitoring The newly designed sensor has been implemented in an overhead transmission line online
device named distribution fault locator as shown in Figure 19. The low frequency
monitoring device named distribution fault locator as shown in Figure 19. The low frequency
components under 1 kHz were eliminated by a high pass filter because only the high frequency
components under 1 kHz were eliminated by a high pass filter because only the high frequency was
was of concern ininthe
of concern theapplication.
application.

(a) (b)
Figure 19. The field application of the proposed sensor. (a) Locators installed in field; (b) The captured
Figure 19. The field application of the proposed sensor. (a) Locators installed in field; (b) The captured
fault current waveforms.
fault current waveforms.
Figure 19 shows the amplitude of the captured fault travelling wave current was 1254 A, the half
wavelength was about 26 s, while the rise time was about 5 s, which was fully adequate for
transient fault current detection, so the designed PCB sensor has excellent performance, as good as
that of the commercial ones.

7. Conclusions
Sensors 2016, 16, 742 16 of 17

Figure 19 shows the amplitude of the captured fault travelling wave current was 1254 A, the half
wavelength was about 26 s, while the rise time was about 5 s, which was fully adequate for transient
fault current detection, so the designed PCB sensor has excellent performance, as good as that of the
commercial ones.

7. Conclusions
An open structure PCB Rogowski coil has been presented. The mechanism of interference
around the overhead transmission line and the guidelines of anti-interference were established. The
proposed PCB Rogowski coil according to the guidelines was designed and tested. For transient fault
current detection, the proposed PCB Rogowski coil has almost the same excellent performance in
work frequency bandwidth and linearity as commercial ones. Furthermore, the strictly symmetrical
structure of the PCB Rogowski coil could eliminate almost all the magnetic field interference from the
radial direction, so it is immune from the interference of adjacent conductors, while, the differential
winding of the PCB Rogowski coil could offset the axial magnetic field interference, and as a result, the
proposed sensor has perfect anti-interference performance, which was the key to measuring transient
fault currents. Through the tests, the proposed PCB Rogowski coil was proved to be more machinable,
repeatable, and of commercial value.

Acknowledgments: This research was funded by a grant (No. 51307109) from the National Natural Science
Foundation of China and a grant (No. 2013DFG71630) from the International S & T Cooperation Projects of China
and the State Grid Corporation of China.
Author Contributions: The idea of this work was proposed by Yadong Liu, Qian Yong, Gehao Sheng and
Xiuchen Jiang. Xiaolei Xie and Yue Hu performed the experiments and analyzed the results. Yadong Liu and
Xiaolei Xie wrote the paper.
Conflicts of Interest: The authors declare no conflict of interest.

References
1. Le Blond, S.T.; Deng, Q.; Burgin, M. High frequency protection scheme for multi-terminal HVDC overhead
lines. In Proceedings of the 12th IET International Conference on Developments in Power System Protection
(DPSP), Copenhagen, Denmark, 31 March3 April 2014; pp. 15.
2. Lei, A.Y.; Dong, X.Z.; Shi, S.X.; Wang, B. A novel current travelling wave based single-ended fault location
method for locating single-phase-to-ground fault of transmission line. In Proceedings of the 2015 50th
International Universities on Power Engineering Conference (UPEC), Shropshire(Shrewsbury), UK, 14
September 2015; pp. 16.
3. Ouyang, Y.; He, J.; Hu, J.; Wang, S.X. A current sensor based on the giant magnetoresistance effect: Design
and potential smart grid applications. Sensors 2012, 12, 1552015541. [CrossRef] [PubMed]
4. Varskyi, H.; Tankevych, S.; Hrechko, V. Design features of electromagnetic transducer as a part of electronic
current transformer. In Proceedings of the 2014 IEEE International Conference on Intelligent Energy and
Power Systems (IEPS), Kyiv, Ukraine, 26 June 2014; pp. 6062.
5. Sun, C.; Wen, Y.; Li, P.; Ye, W.; Yang, J.; Qiu, J.; Wen, J. Self-contained wireless Hall current sensor applied for
two-wire zip-cords. IEEE Trans. Magn. 2016, 99, 1127. [CrossRef]
6. Karasawa, D.; Sonehara, M.; Kitazawa, S.; Sato, T. Development of optical probe current sensor with Kerr
effect for power electronics. In Proceedings of the IEEE Sensors 2014, Valencia, Spain, 25 November 2014;
pp. 19281931.
7. Zhang, G.; Liu, Y.; Xie, S. Positional error analysis of PCB Rogowski coil for high accuracy current
measurement. Adv. Mech. Eng. 2013, 5, 375301. [CrossRef]
8. Di, Z.G.; Jia, C.R.; Zhang, J.X. PCB Rogowski coil for electronic current transducer. Sens. Transducers 2014,
171, 206213.
9. Han, R.Y.; Wu, J.W.; Ding, W.D.; Jing, Y.; Zhou, H.B.; Liu, Q.J.; Qiu, A.C. Hybrid PCB Rogowski coil for
measurement of nanosecond-risetime pulsed current. IEEE Trans. Plasma Sci. 2015, 43, 35553561. [CrossRef]
Sensors 2016, 16, 742 17 of 17

10. Fritz, J.N.; Neeb, C.; de Doncker, R.W. A PCB Integrated Differential Rogowski Coil for Non-Intrusive
Current Measurement Featuring High Bandwidth and dv/dt Immunity. Available online: https://eldorado.
tu-dortmund.de/handle/2003/33955 (accessed on 16 May 2016).
11. Ghanbari, T.; Farjah, A. Application of Rogowski search coil for stator fault diagnosis in electrical machines.
IEEE Sens. J. 2014, 14, 311312. [CrossRef]
12. Ardila-Rey, J.A.; Albarracn, R.; lvarez, F.; Barrueto, A. A validation of the spectral power clustering
technique (SPCT) by using a Rogowski coil in partial discharge measurements. Sensors 2015, 15, 2589825918.
[CrossRef] [PubMed]
13. Chen, Q.; Li, H.B.; Huang, B.X.; Dou, Q.Q. Rogowski sensor for plasma current measurement in J-TEXT.
IEEE Sens. J. 2009, 9, 293296.
14. Sheng, Z.Z.; Han, B.; Qing, M.A. The analysis of the factors affecting the performance of Clamp Rogowski
coil. In Proceedings of the 2014 International Conference on Power System Technology (POWERCON),
Chengdu, China, 2022 October 2014; pp. 14461450.
15. Metwally, I.A. Design of different self-integrating and differentiating Rogowski coils for measuring
large-magnitude fast impulse currents. IEEE Trans. Instrum. Meas. 2013, 62, 23032313. [CrossRef]
16. Ahmed, A.; Coulbeck, L.; Castellazzi, A.; Johnson, C.M. Design and test of a PCB Rogowski coil for very
high dI/dt detection. In Proceedings of the 2012 15th Conference on International Power Electronics and
Motion Control (EPE/PEMC), Novi Sad, Serbia, 46 September 2012.
17. Shao, F.Y. Phase conductor configuration and power frequency electromagnetic environment of UHV
transmission lines In China. Power Syst. Technol. 2005, 08, 17.
18. Tao, T.; Zhao, Z.; Ma, W.; Pan, Q.; Hu, A. Design of PCB Rogowski coil and analysis of anti-interference
property. IEEE Trans. Electromagn. Compat. 2016, 58, 344355. [CrossRef]
19. Gong, Y.Z. Study on Current Sensor Based on PCB Rogowski Coil; Jiangsu University: Zhenjiang, China, 2009.
20. Ferkovic, L.; Ilic, D.; Lenicek, I. Influence of axial inclination of the primary conductor on mutual inductance
of a precise Rogowski coil. IEEE Trans. Instrum. Meas. 2015, 64, 30453054. [CrossRef]
21. Farjah, E.; Givi, H.; Ghanbari, T. Application of an efficient Rogowski coil sensor for switch fault diagnosis
and capacitor ESR monitoring in non-isolated single switch DC-DC converters. IEEE Trans. Power Electron.
2016. [CrossRef]
22. Heng, S.K.; Han, Y.C.; Zhang, J.S.; Wei, H.R.; Zheng, X.L.; Wang, C.L.; Zhu, W. Design of PCB Rogowski coil
current transformer for fault indicators in distribution networks. Water Resour. Power 2015, 33, 186189.

2016 by the authors; licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC-BY) license (http://creativecommons.org/licenses/by/4.0/).