International Journal of Engineering & Technology IJET-IJENS Vol: 11 No: 02
37
The Development of Hybrid Surge Protection Circuit
with Effect of Adding a Filter
Amizah Md Ariffen , Mohd Zainal Abd Kadir & Wan Fatinhamamah Wan Ahmad
Abstract – Surge protective device is a part of internal lightning
protection and it crucially to divert the certain amount of surges
to the ground before the protected load able to handle.
Therefore, this paper studying the enhancement of SPD through
the effect of adding a filter in hybrid surge protection circuit
using simulation modeling. Ordinarily, filter is used as in
eliminating noises in certain application such as electronic
apparatus. Furthermore, an experiment testing is pursues in high
voltage laboratory for comparison and to validate the simulation.
The analysis involved the surge protector Class 3 according to
IEC or Category A as referring to IEEE standards with a filter
design involving the combination of inductor and capacitor.
From this research is found that a significant improvement of
about 38% obtain when filter is added in surge protective device.
Index Term –
surge protector, transient voltage suppressor
diode, metal oxide varistor, hybrid protection, surges, EMI filter
I.
INTRODUCTION
Surge protection device (SPD) is widely used these days to
withstand surge in power line and telecommunication. Surge
suppressors, surge diverters, surge arresters, transient voltage
suppressors (TVSS) all are the SPD other names with the
similar purpose [1]. SPD can be found from mostly utility type
surge arrester to small suppressors install at equipment. The
protection provided by SPD is attenuating the transient and
their propagation to the ground before entering the protected
load.
However by installing the SPD cannot resolve all
kinds of power system disturbances as notches, sags,
harmonic, swells and etc. [2] as in Table I [3].
SPD functioning either divert to ground or clamp into
safe level the surge before it impinges on the equipment thru
the power system supply and only conduct under the surge
condition by lowering its impedance. Hence, the current will
past, rather than through the protected equipment [4].
Amizah Md Ariffen is with the School of Electrical System Engineering,
Universiti Malaysia Perlis,
P.O Box 77, Pejabat Pos Besar, 01000 Kangar, Perlis, Malaysia (e-mail:
amizah@unimap.edu.my).
Mohd Zainal Abd Kadir was with Faculty of Engineering, Universiti Putra
Malaysia 43400 UPM Serdang, Selangor, Malaysia
Wan Fatinhamamah Wan Ahmad was with Faculty of Engineering, Universiti
Putra Malaysia 43400 UPM Serdang, Selangor, Malaysia
Despite other power disturbances, SPD is still needed
as a protection against surges origin either lightning or
switching phenomenon. Bear in mind, an excellent protection
from surges can only be accomplished if tremendous
coordination all stages of SPD location category starting
upstream until downstream are covered. In addition an
equipotential bonding also must be accomplished according to
the relevant standard available [2]. For ensuring an
appropriate function of SPD, the evaluation must be done if
use in the particular system or with any electrical equipment
which applied [5].
i. Filter typology
Filter’s function is to eliminate the excessive noise before
entering the load which originated from power supply as its
travel along electrical conductors, wires, and printed circuit
board or electronics component for example transformers and
semiconductors. The electrical noise is also in form of radiated
electromagnetic interference (EMI) or radio frequency
interference (RFI) which propagates through air and free
space. Conducted EMI noise can be divided into two types as
common-mode noise (CMN) and differential-mode noise
(DMN) known as normal- mode noise (NMN). CMN is the
noise signals on each of the current-carrying conductors are in
phase and equal in magnitude while the DMN is the noise
emerge between the current carrying conductors [6]. Both
mode of noise as illustrated in Figure 1 [6].
A complete elimination of NM or DM noise can be
accomplished by considering two parts which intrinsic
differential mode (IDM) noise and mix-mode (MM) noise
using typical EMI typology as in Figure 2 below [7].
II. STANDARD APPLIED
There are various standards applied for the protection schemes
involving the SPDs as from Institute of Electrical and
Electronics Engineers (IEEE), International Electrotechnical
Commission (IEC), National Lightning Safety Institute
(NLSI), European Association for Electrical, Electronic and
Information Technologies (VDE) and etc. General principles,
design and installation requirements, testing of SPDs and risk
management are overlay almost on these standards,
specification, and guidelines.
The IEC and IEEE standards implied for the
requirement of designing and testing of SPD for low-voltage
system is the following.
113502-0909 IJET-IJENS @ April 2011 IJENS
IJENS
International Journal of Engineering & Technology IJET-IJENS Vol: 11 No: 02
IEC 61643-1: Surge Protective Devices Connected to LowVoltage Power Distribution Systems, Part 1: Performance
Requirements and Testing Methods: This standard describing
the assessment method being made for evaluating the type and
performance of SPD.
IEC 61643-2: Surge Protective Devices Connected to LowVoltage Power Distribution Systems, Part 2: Selection and
Application Principles: It specifies information on location
and category of SPDs including their method of installation,
and failure mode of SPDs.
IEC 61000-4-5: Electromagnetic Compatibility (EMC) Testing and Measurement Techniques-Surge Immunity Test:
It specifies the immunity test, test methods, and array of
recommended test levels for devices to unidirectional surges
due to overvoltages from switching and lightning transient.
IEEE Std.C62.41.2: IEEE Recommended Practice on
Characterization of Surges in Low-Voltage (1000V and less)
AC Power Circuits: It present the designer an additional
selection and specification of surge testing waveforms and
stress levels at which described from surge environment that
should be considered for specific equipment.
IEEE Std.C62.42: IEEE Guide for the Application of
Component Surge-Protective Devices for Use in Low-Voltage
[Equal to or less than 1000V (ac) or 1200V (dc)] Circuits: It
provides assistance in selecting the proper component for SPD
or combination of devices used in surge protection, equipment
or system application.
IEEE Std.C62.45: IEEE Recommended Practice on Surge
Testing for Equipment Connected to Low-Voltage (1000V
and Less) AC Power Circuit: It concentrates on dealing with
surges in low-voltage ac power circuit for testing procedure
using a representative surge waveforms developed by the
previous Std.C62.41.1.
i. Surge Protective Device Category
Each of SPD location is referring is known as category which
each SPD are with different classification involving the rated
current, voltage and specifically different testing waveform
amplitude. As for IEC the classification of SPD is depending
on the lightning protection zone concept applied consist of
Class 1, 2, and 3. The highest exposure is the Class 1 followed
by Class 2, and 3 for the least exposure. The IEEE concept is
conveyed as location categories including A, B and C with A
is the slightest exposure and C is the maximum exposure. The
Table II below indicated the detail category of SPD depending
from both standards and others.
38
crowbar types are gas discharge tube (GDT), spark-gaps,
silicon-controlled rectifiers (triacs) or known as thyristor
SPDs.
The clamping type has a non linear characteristic, as
the voltage across it increases when there is large increase in
current flowing through SPD, then the energy is dissipated in
the device. The benefits of this type are capable to response
rapidly to withstand the high transient surges at same time
maintaining the satisfactory lower clamping voltage [8].
Although it suffer from low energy handling capability but it’s
still an efficient for power protection application.
The crowbar type act contrast to the clamping type, it
only operate when the voltage across their terminal exceed a
certain value, triggered a spark over and therefore provide the
low impedance path to ground during surge propagation. This
is identified a volt-time characteristic as it takes time to
perform. Hence it is also their disadvantage as slower in
operation, other it has a high inrush current in conjunction
with that it’s improper for power application but better for
communication [1], [8]. By the way their energy withstand
capability is excellent which it clamping the transient
overvoltage at a relatively low residual voltage [8].
Connection element of SPD either consists of single
element or multiple stages in parallel which has its own
drawback. A single element connection disadvantage as when
its dysfunctional due to burnout, TOV or else therefore a
protection from surges is unachievable as there no other
backup element to divert the surges. The multiple stages
element connection has this benefit plus it can share all the
surges current so their reliability is higher than single element
connection [8].
Next hybrid connection is introduced which it can be
both combination of either clamping or crowbar device with
some inductive element in between stages. The examples its
can include the GDT at first stage to clamp the high energy
surges, at second stage is MOV to clamp the medium energy
surges and SAD at last stage to clamp the least energy surges.
The inductive element otherwise can be a series inductor or
filter with inductor and capacitor as can also filtering
harmonic or noise from harming any protected load.
III. SIMULATION CIRCUIT OF SURGE PROTECTIVE DEVICE
The simulation of surge protective device is starting
with the modelling MOV and SAD using the Micro-Cap Spice
simulator which is quite similar and compatible with PSPICE.
The entire varistor SPICE model in Micro-Cap is based on the
manufacturer datasheet of EPCOS and Littelfuse.
ii. Surge Protection Device Technology
The SPD component type used are clamping and crowbar or
switching, with each has different characteristics and
performances. The clamping components are MOV or varistor
and silicon avalanche diode (SAD) or zener diode, whilst the
113502-0909 IJET-IJENS @ April 2011 IJENS
IJENS
International Journal of Engineering & Technology IJET-IJENS Vol: 11 No: 02
39
Table I
Power system disturbance and SPD interaction
Power system
disturbances
Surges
Swells
Temporary overvoltages
(TOVs)
Notches
Sags
Temporary
undervoltages
Harmonics
Noise
Effect on SPD
Some
Possibly adverse
Possibly adverse
SPD effect on
disturbances
Reduce
Possibly reduce
None
None
None
None
None
None
None
Possibly adverse
None
None
Possibly reduce
L
VNM
Sensitive
equipment
N
VCM
VCM
G
Fig. 1. Common-mode (CM) and Normal-mode (NM) noise
Fig. 2. Typical EMI typology
Table II
SPD comparison of location category and exposure
Description
Level of
exposure
IEC 61643
IEEE
C62.41
DIN VDE
Place of
application
SPD Category
Low/Fine
Medium
High/Coarse
Class 3
Class 2
Class 1
Category A
Category B
Category C
Class D
Distributed circuits,
power outlets, long
branch circuit, circuit
remote from point-ofentry
Class C
Class B
Sub circuits or near
to point of entry,
feeder, sub
distribution board
Point of entry inner
city sites, incoming
supply, main
distribution board
113502-0909 IJET-IJENS @ April 2011 IJENS
IJENS
International Journal of Engineering & Technology IJET-IJENS Vol: 11 No: 02
40
Rs
Ls
Cp
V = f(I)
Fig. 3. Varistor equivalent model
ideal
diode
Ld
Rd
Vd
Fig. 4. Equivalent model of unidirectional silicon avalanche diode
Element of filter
MOV
SAD
L
Input
Output
N
Fig. 5. Circuit of hybrid protection
In SPICE model varistor is signify in V/I characteristic curve
with a shunt capacitance and series inductance. Below in
Figure 3 [9] is a basic equivalent model for MOV, as Rs is the
series lead resistant, Ls is the series lead inductance, Cp is the
shunt capacitance and V = f(I) is the V/I characteristic.
Model of V/I characteristic is implemented by a controlled
voltage source V = f(I) with an insertion series resistance, Rs
= 100μΩ to avoid occurring of impermissible state when a
source is connected directly. The approximation mathematical
expression is:
log V  b1  b2 . log( I )  b3 .e  log( I )  b4 .e log( I ) I > 0 (1)
The SAD basic form is a single semiconductor P/N junction
consists of an anode (P) and a cathode (N). In SPICE model
the entire SAD model is based on ST Microelectronics and
Littelfuse manufacturers. In Figure 4 below is an equivalent
model of SAD [10]. Rd is a resistance of SAD, and Ld is
depending on the length of the leads connecting SAD to
external circuit. While Vd, is the dc voltage source or
breakdown voltage which if voltage across SAD is below the
Vd, none current is conducting and vice versa [10].
The general circuit of hybrid protection which consists of
MOV and SAD as the element of SPD with a inductorcapacitor (LC) as the element of filter for this paper as
illustrated in Figure 5. LC as a filter is placed between the
stages of MOV and SAD. The simulation of SPD circuit
without a filter also is applying which the circuit diagram
similar in Figure 5 except without the element of LC in
between.
i.
Simulation of impulse voltage and current
The simulation of SPD involving Class 3 (IEC) or Category A
(IEEE), hence a combination waveform is applied for testing
as according to both standard. The standard equation of
impulse waveform for voltage as in mathematical expression
below:

V  Vc e t  e  t

(2)
with front time 1.2s and duration 50s as the front voltage
waveform is 1.67 (t90 – t30) which t90 and t30 are the times of
113502-0909 IJET-IJENS @ April 2011 IJENS
IJENS
International Journal of Engineering & Technology IJET-IJENS Vol: 11 No: 02
90% and 30% amplitude of the leading edge of waveform.
The equation for current impulse waveform as:

I  A e t  e  t

IV. ACTUAL EXPERIMENT OF SURGE PROTECTIVE
DEVICE CIRCUIT
The surge waveform is generated by modular impulse
generator Haefely which has internal resistor of 2 ohm, beside
other useful equipment are the high voltage probe, Le-Croy
digital storage Oscilloscope and current clamper. Before
proceed to the actual experiment testing, the fabricating of
SPD circuit is needed and the component selections are based
on the chosen type components on simulation earlier.
Since described by IEEE [12] and IEC [13] [14], the
procedure of testing equipment is thought according the
guidance and meet the standard requirement to avoid the
misleading results. Therefore, the standard of procedure
defines the range of test levels, test equipment, test setup and
test procedure.
(3)
which the front time 8s and duration of 20s as the front
current waveform is 1.25 ( t90 – t10) with t90 and t10 are the
times of 90% and 10% of amplitude. The amplitude for
combination waveform tested in this simulation are 6kV/ 3kA
as in normal nature despite the IEC suggesting the maximum
amplitude can be reach are 20kV and 10kA. Figure 6 show
the simulation of combination wave applied for testing in
Class 3.
ii.
Element of filter
In this simulation only a single inductor-capacitor
(LC) filter is applied as using the X capacitor (CX) with a
series of inductor as in Figure 7. The matrix equation
described in equa. (4) is represent the simple L consists with a
series inductor and a shunt capacitor that outline a simple
chain matrix. The mathematical expression of inductive
capacitive reactance can be finding by substituting equa. (5)
into equa. (4) [11].
2

JKRd  V 
Vi  1  K
 o

JK
I  
  I o 
1
 i  R

 d
R
1
C
L d
2F0 Rd
2F0
41
i. Generation of Impulse voltage and current
The modular impulse generator generates the combination
waveform at 6kV (1.2/50µ) and 3kA (8/20µ) as shown in
Figure 8. Each mode of SPD; Live, Neutral and Ground are
injected with surge waveforms and the performance output is
analyzed in term of let-through voltage.
(4)
(5)
Fig. 6. Simulation of combination wave, open circuit voltage 6kV/3kA
L
C
Fig. 7. LC Network Description for the above figure.
113502-0909 IJET-IJENS @ April 2011 IJENS
IJENS
International Journal of Engineering & Technology IJET-IJENS Vol: 11 No: 02
42
Fig. 8. Combination waveform testing
Input of Combination wave
7
6.104
6
6.027
Exp: T 90% = 0.8µs
Sim: T 90% = 0.6µs
5
Voltage (kV)
4
Exp: T 50% = 45µs
Sim: T 50% = 49µs
3
2
Exp: T 30% = 0µs
Sim: T 30% = 0µs
1
0
-10
0
10
20
30
40
50
-1
time (µs)
experiment input
simulation input
Fig. 9. The comparison input of combination waveform 6kV/3kA
Output at L-N
270
Exp: 251.95V
Sim: 240.10V
230
voltage (V)
190
Exp: 162.00V
Sim: 151.92V
150
110
70
30
-10
-10
0
Experiment (filter)
10
20
time (µs)
Simulation (filter)
30
Experiment (no filter)
40
50
Simulation (no filter)
Fig. 10. Output at Live to neutral for SPD with filter compare to SPD without filter
113502-0909 IJET-IJENS @ April 2011 IJENS
IJENS
International Journal of Engineering & Technology IJET-IJENS Vol: 11 No: 02
43
Output at L-E
290
Exp: 262.78V
Sim: 260.48V
240
voltage (V)
190
Exp: 168.13V
Sim: 158.92V
140
90
40
-10
-10
0
Experiment (filter)
10
20
time (µs)
Simulation (filter)
30
Experiment (no filter)
40
50
Simulation (no filter)
Fig. 11. Output at Live to earth for SPD with filter compare to SPD without filter
Output at N-E
290
Exp: 253.19V
Sim: 238.86V
voltage (V)
240
190
Exp: 161.67V
Sim: 151.02V
140
90
40
-10
-10
0
Experiment (filter)
10
20
time (µs)
Simulation (filter)
30
Experiment (no filter)
40
50
Simulation (no filter)
Fig. 12. Output at Neutral to earth for SPD with filter compare to SPD without filter
V.
RESULTS AND DISCUSSION
i. Evaluation of input waveform
The SPD circuit is being tested with by injecting the
combination waveform of 6kV/3kA on each mode of
protection which are Live to neutral (L-N), Live to earth (L-E)
and Neutral to earth (N-E). The output after the SPD
components clamping is measured for voltage difference
between the output terminal which known the let-through
voltage. Figure 9 show the input of both combination
waveforms from simulation and experiment for first injection.
As observe the different between the simulation input and
the experiment input is faintly which is 1.26 %. While
evaluation with the standards; IEC [4], [13] and IEEE [10], the
requirement of open circuit voltage are the front time margin
is T1 = 1.2µs ± 30% and time to half value, T2 = 50µs ± 20%.
Hence, in analysis of experiment input the T1 is 1.34µs and T2
is 45µs, that both are within the margin time. Though the
simulation time, T1 and T2 are 1 µs and 49µs respectively
which within the proposed margin time. For that reason, the
experiment and simulation input are accurate. Moreover, the
value of Vpeak (Note 1) for simulation and experiment also is
within 10% margin endowed that 6.027kV and 6.104kV
correspondingly.
ii. Live to neutral
The graph in Figure 10 indicated the result when impulse
voltage is injected at between Live and Neutral (L-N) line.
The output result signifies after SPD clamping is known as the
let-through voltage. Shown from the Figure 10 the let-through
voltage for hybrid SPD with filter is much lower than without
the filter whether in simulation or experiment. The let through
voltage of SPD with filter added much lower than SPD
without filter about 35.7% to 36.7% for both in experiment
and simulation subsequently. As observe, the simulation result
is predict a bit lower than the experiment but yet still tolerable
which the difference is around 4.7% to 6.2% compare to the
113502-0909 IJET-IJENS @ April 2011 IJENS
IJENS
International Journal of Engineering & Technology IJET-IJENS Vol: 11 No: 02
experiments’ result. Above all, the pattern of all output graphs
is similar whether in actual experiment or forecasted in
simulation.
iii. Live to earth
Let through voltage at the Live to earth is in Figure 11. In
analysis, the results signify the let through voltage of SPD
with filter is 36.02% lower than without filter in experiment
and 38.99% much lower for filtered SPD than non filtered
SPD in simulation. The dissimilarity of output for SPD with
filter between simulation and experiment is 5.48%. While, for
SPD without filter the difference is 0.88% comparing in
experiment and simulation.
iv. Neutral to earth
Figure 12 demonstrates the results of let through voltage
at the Neutral to earth mode. As analyze, the let through
voltage of SPD with effect of filter has the 36.15% peak lower
than non filter in actual experiment result. Moreover, in
simulation the peak voltage lower than 36.78% for filtered
SPD than the non- filtered. Hence, the dissimilarity among
simulation and experiment results is minimal 5.65% and
6.59% for both types of SPDs.
The entire the cases whether the surge is implied at L-N,
L-E or N-E mode resulting the same lowers let-through
voltage which represents the lower let-through voltage the
improved protection as the supply voltage entering the load at
safe level. This is indicated by more than 30% of less let
through voltage of SPD with filter than the SPD without a
filter being added. Thus, the simulation of the design hybrid
SPD is effectively functioned by clamping the certain amount
of surge voltages injected as well in the actual experiment
implied which with a minimal error percentage and a parallel
pattern of result graphs. Attributable to this, in addition filter
also functioning as rejecting the higher frequency of current
which in this case is more than 150 kHz as well as reduce the
noises in power line.
V.
CONCLUSION
It is proven that by adding a filter as element of inductive
and isolating part between stages for hybrid SPD indeed has a
excellent impact to assist a much lower let-through voltage
after the surges impinges. In average, the SPD with added
filter was obtain 38% reduce of let through voltage contrast to
non added filter whether by experiment or simulation.
Therefore the enhancement of SPD Class 3 for a better reliable
protection is accomplished. Beside, the simulation as a tool for
forecasting analysis of performance SPD circuit in further
study is acknowledgeable as it resembles the actual
experiment practice.
44
REFERENCES
[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
S. Ioannou, E. K. Stefanakos, P. H. Wiley, (2007). New MOV
Failure Mode Identification Invention, IEEE Transactions on
Consumer Electronics, vol. 53 n.3, pp1068 – 1075.
D. Kadar, F. Martzloff, (2006). Facts, Fiction, and Fallacies in SPD
Design and Applications, IEEE, pp. 1- 8.
IEEE Guide on Interactions between Power System Disturbances
and Surge-Protective Devices, (2005). IEEE Standard C62.48.
Low-Voltage Surge Protective Devices-Part 12: Surge Protective
Devices Connected To Low-Voltage Power Distribution SystemSelection And Application Principles, (2002). IEC Standard 6164312.
R. W. Hotchkiss, (2008). Application of Low-Voltage SPDs:
Approach of he IEEE, IEEE.
Dev Paul, (2001). Low Voltage Power System Surge Overvoltage
Protection, IEEE Transaction on Industry Application, vol. 37 n. 1,
pp. 223 – 227.
Hung-I Hsieh, Jhong-Shu Li, Dan Chen, (2008). Effects of X
Capacitors on EMI Filter Effectiveness, IEEE Transactions on
Industrial Electronics, vol. 55 n. 2, pp. 949-955.
K. Samaras, C. Sandberg, C. J. Salmas, A. Kaoulaxouzidis, (2005).
Electrical Surge Protection Devices For Industrial Facilities (A
Tutorial Review), IEEE, pp. 165 – 175.
SIOV Metal Oxide Varistors, (2008). EPCOS Data Book.
A. Beutel, J. V. Coller, (2005). The Application of Silicon
Avalanche Diodes on Low Voltage Power Systems, IEEE
Transactions on Industry Applications, vol. 41 n. 4, pp 1107-1112.
R. L. Ozenbaugh,(2001). EMI Filter Design, (2nd ed. revised and
expanded) New York: Marcel Dekker, Inc.(Chapter 18).
IEEE Recommended Practice on Surge Testing for Equipment
Connected to Low-Voltage (1000V and Less) AC Power Circuit,
(2002). IEEE Standard C62.45.
Electromagnetic Compatibility (EMC)-part 4-5: Testing and
Measurement Techniques-Surge Immunity Test (First Revision),
(2001). IEC Standard 61000-4-5.
Low-voltage Surge Protective Devices-Part 1: Surge Protective
Devices Connected to Low-Voltage Power Distribution SystemsRequirements and Tests (First Revision), (2005). IEC Standard
61643-1.
IEEE Recommended Practice on Characterization of Surges in
Low-Voltage (1000 V and less) AC Power Circuits, (2002).
IEEE Standards C62.41.2.
ACKNOWLEDGMENT
This work was supported by Centre of Excellence on
Lightning Protection (CELP).
113502-0909 IJET-IJENS @ April 2011 IJENS
IJENS