1983_GE_Transient_Voltage_Supression_Data_Library_4ed 1983 GE Transient Voltage Supression Data Library 4ed
User Manual: 1983_GE_Transient_Voltage_Supression_Data_Library_4ed
Open the PDF directly: View PDF 
.
Page Count: 159
| Download | |
| Open PDF In Browser | View PDF |
GENERAL
ELECTRIC
The
GENERAL ELECTRIC
Manual
Publication
Number
Pric
400.3
400.4
400.5
400.6
$5.0
5.01
5.01
5.01
Transient Voltage Suppression
Optoelectronics
Thyristors - Rectifiers
Transistors - Diodes:l=
Manuals are available from your local GE Electronic COrT
ponents Sales office, authorized GE Electronic Distributor!
or by writing to:
General Electric Company
Semiconductor Products Department
West Genesee Street, Mail Drop 44
Auburn, New York 13021
Europe
Keep knowledge of semiconductor technology at your fingertips with GE's new comprehensive
series of technical manuals, the General Electric
International General Electric Company
of New York
The Demesne
Dundalk, County Louth
Ireland
Electronic Data Library.
Far East
The Library is a virtual storehouse of application, technical, and device selection information on
transient voltage suppressors, optoelectronics, signal and power transistors, triacs, unijunctions,
diodes, rectifiers, SCR's and power modules.
Start selecting products from one of the most
complete collections of semiconductor manuals in
the electronics industry today, the GE Electronic Data
Library. It's like having a GE engineer on your
library shelf.
Electronic Components of General
Electric (USA) Pte. Ltd.
201-8 Boon Keng Road
Singapore 1233
Refer to the inside back cover for a complete listing of Gen
eral Electric's worldwide Electronic Components Sales of
fices.
TRANSIENT VOLTAGE
SUPPRESSION
Semiconductor Products Department
General Electric Company
Auburn, NY 13021
GENERAL
e
ELECTRIC
Transient Voltage Suppression manual, Fourth Edition:
Editors
Contributing Authors
Marvin W. Smith
Sal J. Cardella
Sebald R. Korn
Francois D. Martzloff
John May
Bernie I. Wolff
The Semiconductor Products Department of General Electric Company acknowledges the efforts of all the contributing authors and
editors of the previous editions of the General Electric Transient
Voltage Suppression manual.
The circuit diagrams included in this manual are intended mer\lly for illustration of
typical semiconductor applications and are not intended as constructional information.
Although reasonable care has been taken in their preparation to assure their technical
correctness, in the absence of an express written agreement to the contrary, no responsibility is assumed by the General Electric Company for any consequences of .their use.
The semiconductor products, circuits, and arrangements disclosed herein may be
covered by patents of General Electric Company or others. Neither the disclosure of any
information herein nor the sale of semiconductor products by General Electric Company
conveys any license under patent claims covering combinations of semiconductor products with other products or elements. In the absence of an express written agreement to
the contrary, General Electric Company assumes no liability for patent infringement arising out of any use of semiconductor products with other products or elements by any
purchaser of semiconductor products, or by others.
General Electric Company technical information series (TIS) reports are referenced
throughout this manual. Limited copies of these reports are available from:
General Electric Company
Technical Information Exchange
Corporate Research & Development
P.O. Box 43, Building 5
Schenectady, New York 12301
4th Edition
Copyright ©1983
by the
General Electric Company, U.S.A.
Semiconductor Products Department
Auburn, New York 13021
4
TABLE' OF CONTENTS
1. VOLTAGE TRANSIENTS - AN OVERVIEW
1. 1
1.2
1.3
1.4
1.5
1.6
1 .7
1.8
Repeatable Transients . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Random Transients ....... '. .•. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..
Transients on AC Power Lines ...... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..
Telecommunication Line Transients. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . ..
Automobile Transients. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..
Effects of Voltage Transients. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..
Transient Detection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..
Transient Testing and Standards. . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . ..
9
12
13
15
15
16
18
19
2. TRANSIENT SUPPRESSION - DEVICES AND PRINCIPLES
2.1
2.2
2.3
2.4
Transient Suppression Devices. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 31
Transient Suppressors Compared ................................................ '36
Proof Tests ................... '" .................. ,.. . ... . . . . ... .. .. .. ..... 38
Update on New Devices: Low Voltage and High Energy ............................... 39
3. GE·MOV®II VARISTORS· BASIC PROPERTIES, TERMINOLOGY AND THEORY
3.1
3.2
3.3
3.4
3.5
What is a Varistor ................................... '. . . . . . . . . . . . . . . . . . . . . . ..
Physical Properties. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..
Varistor Construction. . . . . . . .. . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..
Electrical Characterization. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..
Varistor Terminology .................................................. ,. ;...
43
44
48
50
55
4. DESIGNING WITH GE·MOV®II VARISTORS
4. 1
4.2
4.3
4.4
Selecting the Varistor. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..
Failure Modes and Varistor Protection ...........................................
Series and Parallel Operation of Varistors ......... ~ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Applications. . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..
59
66
69
71
5. SUPPRESSION - TELECOMMUNICATION SYSTEMS
5.1
5.2
5.3
5.4
5.5
5.6
5.7
5.8
Introduction..................... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..
System Transients ............................. '.' ... '. . . . . . . . . . . . . . . . . . . . . . . ..
Lightning-Induced Transients ............................................... "
Calculations of Cable Transients ........... , . , . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..
Power System-Induced Transients ........ ~ ................................... "
Protectors-Voltage Transient Suppressors ...................................... "
Power Line Transients ...................................... , . . . . . . . . . . . . . . . ..
Relay Contact Protection. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . • . ..
5
87
87
87
89
91
91
93
93
TABLE OF CONTENTS (Continued)
6. SUPPRESSION - AUTOMOTIVE TRANSIENTS
6.1 Transient Environment ............................ ; . '................... .' . . . ..
6.2 Varistor Applications .................... " . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..
95
96
7. VARISTOR TESTING
7.1
7.2
7.3
7.4
7.5
7,6
7.7
Introduction ........ , .................... ~ .............. ',' . . . . . . . . . . . . . . . . .. 101
TestObjectives ...........................' ............ , ..................... 101
Measurement of Varistor Characteristics ........................ , . . . . . . . . . . . .. . . .. 103
Varistor Rating Assurance Tests .............................................. "
Mechanical and Environmental Testing of Varistors .................................
Equipment for Varistor Electrical Testing ....................................... "
Test Waves and Standards ..... '....................... ; .... , .... , . . • . . . . . . . . . ..
109,
112
113
115
8. VARISTOR RELIABILITY
8. r Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .,' .' . . . . . . . . . . . . . . . . . . ..
8.2 AC Bias Reliability ..........................................................
8.3 DC Bias Reliability .................... : .....................................
8.4 Pulse Energy Capability ...................................... , .. : .............
8.5 Mechanical Reliability and Integrity ........................................... "
8.6 Environmental and Storage Reliability ..... : .....................................
8.7 Safety ......... ',' ... ',' ............................. '... '.....................
119
119
122
124
127
128
129
9. GE MOV®U VARISTOR SPECIFICATIONS
Introduction ................................................................ 131
How to Se1eGt a Varistor. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 135
Ratings and Characteristics .... :. '................................... '. . . . . . . . . .. 137
Transient V-I Characteristics ............ " .....................................
143
,
Pulse Ratings ................................................................. 147
High Reliability .. ; ......... " ..... : .................. ~ ........ ~ ................ 150
Outlines and Dimensions .................... '.................................... 151
Mounting ofP Series ....................... ',' .........................' ....... 154
INDEX ...... ' ................................ : . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 159
6
INTRODUCTION
General Electric, a major U. S. manufacturer of electrical equipment, has for many years
researched the causes, effects and frequency oftransitory voltage variations resulting from atmospheric
and other disturbances. Several members of General Electric's Research and Development Center are
recognized worldwide as experts in this field. The need for expertise in this area has become even more
critical as solid-state controls have achieved widespread incorporation into industrial, as well as
consumer products.
Electromechanical and vacuum tube circuits were less susceptible to damage by high voltage
transients, so there was little requirement for superior surge-voltage protection devices. With increased
usage of solid-state controls, surge suppression requirements have become much more stringent.
Traditional methods of suppression were in many instances no longer adequate-a new way had to be
found. One solution was a new, ceramic, nonlinear semiconductor, generically known as a ZnO metal
oxide varistor. General Electric's Semiconductor Products Department introduced the GE-MOV®
metal oxide varistor in 1972, and its acceptance was instantaneous. Here was a new low cost effective
method to suppress transient voltages. It was fast (nanosecond response time), compact (typically 720mm. diameter, 1-3mm. thick), had high energy capability (up to 200J, 6000 A surge) and was more
economical than equivalent devices that had been used to suppress transients.
Since 1972, General Electric has refined suppression technology via improved varistor
performance and has continued extensive research into causes and effects of transient voltages. Product
improvements have evolved through better ceramic material and the introduction of new GE-MOV®
Varistor package configurations to accommodate industry requirements. The late~t of product
developments from the Semiconductor Products Department is the introduction of the GE-MOV® II
Varistor. The GE-MOV® II Varistor is an improved metal oxide varistor with energy capacity rated
significantly above the original product. Now, for the same price and in the same she package, you are
guaranteed more protection than ever before.
To complement the introduction of GE-MOV®II Varistors, General Electric bas produced this
completely revised Fourth Edition Transient Voltage Suppression manual. This is a demonstration that
General Electric will continue to be a world leader in understanding the cause of equipment malfunction
and failure and in devising preventative methods. .
We trust you will find this manual useful in determining the suppression requirements of your
products. General Electric offers a wide range of protection devices, from the small Molded Axial
product rated below 2J and 100A, up to the High Energy product capable of 10,0001 and 70,000A.
With this broad product offering, a solution to most transient voltage problems can be realized. GEMOV®II Varistors are available to you from authorized electronic products distributors throughout the
U.S. and the world. You may also'obtain devices or further information from anyone of our Electronic Components Sales offices listed on pages 162 and 163.
7
~~lJ
VOLTAGE TRANSIENTS - AN 'OVERVIEW
To treat any problem, the scope of the problem must first be established, This chapter is an
overview of the sources and nature of transient overvoltages, the problems they can cause and the
equipment for testing and monitoring them.
Transients in electrical circuits result from the sudden release of previously stored energy. This
energy can be stored within the circuit and released by a voluntary or controlled switching action or it
can be stored outside the circuit and be injected or coupled into the circuit of interest by some action
beyond the control ofthe circuit designer.
Transients may occur either in repeatable fashion or as random impulses. Repeatable transients, such as commutation voltage spikes, inductive load switching, etc., are more easily observed,
defined and suppressed. Random transients are more elusive. They occur at unpredictable times, at
remote locations, and require installation of monitoring instruments to detect their occurrence. In
fact, a direct corollary of Murphy's law states that the best transient suppressor is a transient monitor! However, enough experience has been accumulated to provide reasonable guidelines for the
transient environments in low voltage ac power circuits,1,2 telecommunications equipment,3 and
automotive electrical systems. 4
Effective transient overvoltage protection requires that the impulse energy be dissipated in the
added suppressor at a voltage low enough to ensure the survival of circuit components. The following sections will discuss in detail the two categories of transients, how they occur, their effects and
their detection.
1.1
REPEATABLE TRANSIENTS
A sudden change in the electrical conditions of any circuit will cause a transient voltage to be
generated from the energy stored in circuit inductance and capacitance. The rate of change in current
(di/dt) in an inductor (L) will generate a voltage equal to - L di/dt, and it will be of a polarity that causes
current to continue flowing in the same direction.
It is this effect that accounts for most switching-induced transient overvoltages. It occurs as
commutating spikes in power conversion circuits, when switching loads and under fault conditions.
The effect is brief, since the source is limited to the energy stored in the inductance (1/2 LF ), and
it is generally dissipated at a high instantaneous power (Energy = power x time). But the simple
effect of one switching operation can be repeated several times during a switching sequence (consider arcing in the contact gap of a switch), so that cumulative effects can be significant.
1.1.1 Energizing the Transformer Primary
When a transformer is energized at the peak of the supply voltage, the coupling of this voltage
step function to the stray capacitance and inductance of the secondary winding can generate an
oscillatory transient voltage with a peak amplitUde up to twice the normal peak secondary voltage.
9
Subsequent oscillations depend on the L and C parameters of the circuit. Another important
pOint to remember is that the secondary side will be part of a capacitive divider network in series
with the· transformer interwinaing'capacitance (Cs )' This capacitively coupled voltage spike has no
, direct relationship to the turns ratio of the transformer, so that it is conceivable that the secondary
circuit can see the peak applied primary voltage ..
LINE
VOL~:~
Cs
,
/ \
~
~CLOSED
SECONDARY
VOLTAGE
Vs
PK
Cs
LOAD
FIGURE 1.1: VOLTAGE TRANSIENT CAUSED BY ENERGIZING TRANSFORMER PRIMARY
, 1.1.2 De-Energizing the Transformer Primary
The opening of the primary circuit of a transformer generates extreme voltage transients,
especially if the transformer drives a high impedance load. Transients in excess of ten times normal
voltage have been observed across power semiconductors when this type of switching occurs.
Interrupting the transformer magnetizing current, and the resulting collapse of the magnetic
flux in the core, couples a high voltage transient into the transformer secondary winding, as shown
in Figure 1.2.
OPENING
SWITCH
~-
vp~
~
LlJ
)
VOLTAGE
TRANSIENT
LOAD
.
FIGURE 1.2: VOLTAGE TRANSIENT CAUSED BY INTERRUPTION OF TRANSFORMER MAGNETIZING CURRENT
Unless a low-impedance discharge path is provided, this burst of .transient energy appears
across the load. If this load is a semiconductor device or capacitor with limited voltage capabilities,
that component may fail. The transients produced by interrupting the magnetizing current are
usually quite severe. For example, the stored energy in the magnetizing field of a 150kVA transformer can be 9 J.
10
1.1.3 Fault with Inductive Power Source
If a short develops on any power system, devices parallel to the load may be destroyed as the
fuse clears.
1
OTHER
LOAD
SHORT_
VSUPPLY
(LOAD)
1
FIGURE 1.3: VOLTAGE TRANSIENT CAUSED BY FUSE BLOWING DURING POWER FAULT
When the fuse or circuit breaker of Figure 1.3 opens, it interrupts the fault current, causing
the slightly inductive power source to generate a high voltage ( - L di/dt) , high energy 0/2 Li 2 ) transient across any parallel devices. Suddenly interrupting a high current load will have a similar effect.
1.1.4 Switch Arcing
r------
1
VSUPPLY
Vc:
\
p
SOLID-STATE
EQUIPMENT
rI
--I
..........
I
I
L.._
FIGURE 1.4: VOLTAGE TRANSIENTS CAUSED BY SWITCH ARCING
When current in an inductive circuit, such as a relay coil or a filter reactor" is interrupted by a
contactor, the inductance tries to maintain its current by charging the stray capacitance. Similar
action can take place during a closing sequence if the contacts bounce open after the initial closing.
The high initial charging current will oscillate in the inductance and capacitance at a high frequency.
When the voltage at the contact rises, breakdown of the gap is possible since the distance is still
very small during the opening motion of the contact. The contact arc will clear at the current zero
of the oscillation but it will restrike as the contact voltage rises again. As the contacts are moving
farther apart, each restrike must occur at a higher and higher voltage until the contact succeeds in
interrupting the current.
This restrike and escalation effect is particularly apparent in Figure 1.5, where a switch opens a
relay coil of 1 H, having about O.OOlJ.LF of distributed (stray) capacitance in the winding. Starting
with an initial dc current of lOOmA, the circuit produces hundreds of restrikes (hence, the "white"
11
band on the oscillogram) at high repetition rate, until the circuit clears, but not before having
reached a peak of 3 kV in contrast to the initial 125 V in the circuit.
+
0-
HORIZONTAl.-t, 500,#s,/div.
VERTICAL -v, 1.0 kV/div.
FIGURE 1.5: VOLTAGE ESCALATION DURING RESTRIKES
Electromechanical contacts generate transients which they generally can survive. However, in
the example just discussed, the 2.5 ms long sequence of restrikes and attendant high current may
be damaging to the contacts. Also, the transients injected into the power system during the restrike
can be damaging to other loads.
In an attempt to eliminate electromechanical switches and their arcing problem, solid-state
switches are recommended with good reason! However, if these switches are applied without discrimination in inductive circuits, the very effectiveness of the interruption· can ·lead the solid-state
switch to "commit suicide" by generating high transients.
In the example of Figure 1.6, the transistor used for switching 400mA in a 70mH solenoid is
exposed to 420V spikes, although the circuit voltage is only 150V.
150 VDe
O--+-~>-----,
50 ms/DIVISION
FIGURE 1.6: TRANSISTOR SWITCHING TRANSieNT
Whenever possible, a system should be examined for potential sources of transient overvoltage
so they can be eliminated at the source, for one source can affect many components. If the sources
are many (or unidentifiable) and the susceptible components few, it may be more practical then to
apply suppression at the components.
1.2
RANDOM TRANSIENTS
Frequently, transient problems arise from. the power source feeding the circuit. These transients create the most consternation because it is difficult to define their amplitude, duration and
energy content. The transients are generally caused by switching parallel loads on the same branch
of a distribution system, although they also can be caused by lightning. Communication lines, such
as alarm and telephone systems, are also affected by lightning and power system faults.
12
To deal with random transients, a statistical approach has been taken to identify the nature of line
overvoltages. While recordings of transients have been made, one cannot state that on a specific system
there is an "X" probability of encountering a transient voltage of "Y" amplitude. Therefore, one is
limited to quoting an "average" situation, while being well aware that large deviations from this
average can occur, depending on the characteristics ofthe specific system.
In the following sections, the recorded experiences of three types of systems will be described.
These are: 1) ac power lines (less than 600V); 2) telecommunication systems; and 3) automotive
systems. The more exotic environmental conditions such as NEMP (nuclear electromagnetic pulse)
are intentionally omitted from this discussion because they are the subject of specialized study by
a limited community.
1.3
TRANSIENTS ON AC POWER LINES
Data collected from various sources has provided the basis for this guide to transient overvoltages. 1.5,6,7,8
1.3.1 Amplitude and Frequency of Occurrence
The amplitude of transient recordings covers the range from harmless values just above normal
voltage to several kilovolts. For 120V ac lines, flashover of the typical wiring spacing produces an
upper limit between 6 and 8kV. Ironically, the better the wiring practices, the higher the flashover,
allowing higher transients to exist in the wiring system. Studies of the frequency of occurrence and
amplitude agree on an upper limit and a relative frequency of occurrence. Figure 1.7 shows frequency as
a function of amplitude. Experience indicates that devices with less than 2kV withstand capability will
have poor service life in the unprotected residential environment. Prudent design will aim for 3kV
capability, although, where safety is of the utmost concern, designing for 6kV can cope with these rare
but possible occurrences.
10,000
\
\
\
~
0::
-
....
\
tzlOO
w
\
Cii
IL
0
>-
\
u
z
w 1.0
::J
0
W
0::
~\
. \
\
IL
\
0.1
0.01
\
\
\
\
Z
~.
\
\"
... \
1000
'"
\
.1
./
.2
.4.6
I
\ .. \\
'.
\\
i'\.
\
\'
\.
4
- - - Farmhouse supplied
mission lines.
\
by overhead trans(220V) Service
Landis & Gyr. Plant, Zug, outlet in lab.
\
,\ \
.\
2
_ _ _ Service entrance, 16 family house, under·
ground system (head station).
\,
~\
~..
\
- - - - - - Service entrance of bank building in Basel,
Switzerland.
_ _ _ 16 family house, upstairs living room outlet.
1\
6
10
_
... -
Landis & Gyr., Zug, outlet in furnace room.
(Courtesy of L. Regez, Landis & Gyr., Zug,
Switzerland.)
- - - U.S. composite curve.
kV
PEAK VALUE OF VOLTAGE TRANSIENT
FIGURE 1.7: FREOUENCY OF OCCURRENCE OF TRANSIENT OVERVOL TAGES IN 220V AND 120V SYSTEMS
13
(120V) Service
For systems of higher voltages (220/240V, 480V), limited data is available for u.s. systems.
However, the curves of Figure 1.8 indiCate the difference between the two classes, 120V and 220V
systems, is smaller than the differences within each class. 8 One can conclude that the amplitude of
the transient dependsmbre upon the amount of externally coupled energy and the system impedance
'
than upon the system voltage.
For internal switching transients in the power system, Figure 1.8 shows the relationship (computed and measured) between system voltage and.transient peaks. 8 Oearly, there is no direct liriear
increase of the tran~ient amplitude as the system voltage is increased.
6r--------.--------,-~----~r_--~~~--------_,
5r--------+~~~--~------~r_~~~~--------~
II)
~
o
>
34r-----,,........-7'C--------~------~r_------_+-----,----~
:.:
I
III
C)
o~3~----1.---+--------+_------~k==__-~:::::::==:::::~
>
It:
III
o>
I-
~2r------7q------.,.--~------~r--------+--------~
in
z
ct
It:
I-
O~------~--------~------~~------~--------~
100
200
300
400
500
RATED VOLTAGE OF DISTRIBUTION SYSTEM (VOLTS)
FIGURE 1.8: SWITCHING VOLTAGE TRANSIENTS AS A FUNCTION OF THE SYSTEM VOLTAGE FOR THREE
VALUES OF THE TRANSIENT TAIL (TIME TO HALF·VALUE)
(Data Courtesy of L. Regez, Landis & Gyr., Zug, Switzerland)
Some indication of the uncertainty concerning the expected transient level can be found in the
industrial practice of choosing semiconductor ratings. Most industrial users of power semiconductors choose semiconductor voltage ratings from 2.0 to 2.5 times the applied peak steady-state voltage, in conjunction with rudimentary transient suppression, in order to ensure long-term reliability.
Whether or not this ratio is realistically related to actual transient levels has not been established;
the safety factor is simply chosen by experience. While it is dangerous to argue against successful
experience, there are enough cases where this rule of thumb is insufficient and thus a more exact
approach is justified. Another objection to the indiscriminate rule of thumb is economic. Specifying
2.5 times the peak system voltage results in a high price penalty for these components; It is normally unrealistic and uneconomical to specify semiconductors that should withstand transients without protection. The. optimum situation is a combination of low cost transient protection combined
with lower cost semiconductors having lower voltage ratings.
14
1.3.2 Duration, Waveform and Source Impedance
There is a lack of definitive data on the duration, waveform and source impedance of transient
overvoltages in ac power circuits. These three parameters are important for estimating the energy
that a transient can deliver to a suppressor. It is desirable to have a means of simulating the environment through a model of the transient overvoltage pulse. Suggestions have been made to use standard impulses initially developed for other applications. For instance, the classical 1.2 x 50J.Ls unidirectional voltage impulse specified in high voltage systems has been proposed. 9 Also, the repetitive burst of 1.5 MHz oscillations ("SWC") specified for low-voltage and control systems exposed
to transients induced by high-voltage disconnect switches in utility switch yards is another
suggestion. 10
Working Groups of the IEEE and the Intern(;ltional Electrotechnical Commission have
developed standard test waves and source impedance definitions. These efforts are aiming at moving away from a concept whereby one should duplicate environmental conditions and towards a
concept of one standard wave or a few standard waves arbitrarily specified. The justifications are
that equipments built to meet such standards have had satisfactory field experience and provide a
relative standard against which different levels of protection can be compared. A condition for
acceptance of these standard waves is that they be easy to produce in the laboratory. 11 This is the
central idea of the TCL (Transient Control Level) concept which is currently being proposed to
users and manufacturers in the electronics industry. Acceptance of this concept will increase the
ability to test and evaluate the reliability of devices and systems at acceptable cost.
1.4
TELECOMMUNICATION LINE TRANSIENTS
Transient overvoltages occurring in telephone lines can usually be traced to two main sources:
lightning and 50/60 Hz power lines. Lightning overvoltage is caused by a strike to the conductor
of an open wire system or to the shield of a telephone cable. Most modern telephone lines are contained in shielded cables. When lightning or other currents flow on the shield of a cable, voltages are
induced between the internal conductors and the shield. 12 The magnitudes of the induced voltages
depend on the resistance of the shield material, openings in its construction, and on the dielectric
characteristics and surge impedance of the cable.
Limited data has been collected on lightning voltages in telephone cables because of the difficulty of obtaining such data without interfering with the system operation. 13 Also, the carbonblock or spark gap protectors presently used have provided adequate protection for most installations using electromechanical components.
The close proximity of telephone cables and power distribution systems, often sharing right-ofway-poles and even ground wires, is a source of transient overvoltages for the telephone system.
Overvoltages can arise from physical contact of falling wires, electromagnetic induction, and ground
potential rise. Chapter 5 of this manual presents a detailed discussion of lightning-induced and
power system-induced transients.
1.5
AUTOMOBILE TRANSIENTS
Four principal types of voltage transients are encountered in an automobile. These are "load
dump," alternator field decay, inductive switching and mutual coupling.4 In addition, cold morning "jump starts" with 24 V batteries occur in some areas.
The load dump transient is the most severe and occurs when the alternator current loading is
abruptly reduced. The most demanding case is often initiated by the disconnection of a partially
discharged battery due to defective terminal connections. Transient voltages have been reported
over lOOV lasting up to 500ms with energy levels in the range oftens to hundreds of Joules.
15
Switching of inductive loads, such as motors and solenoids, will create negative polarity transient voltages with a smaller positive excursion. The voltage waveform has been observed to rise to
a level of -210V and +80V and last as long as 320l1s. The impedance to the transient is unknown,
leading some designers to test with very low imped'ance, resulting in the use of more expensive components than necessary.
The alternator field decay transient is essentially an inductive load switching transient. When
the ignition switch is turned off, the decay of the alternator field produces a negative voltage spike,
whose amplitude is dependent on the voltage regulator cycle and load. It varies between ~40V to
-IOOV and can last 200ms.
Other unexplained transients have been recorded with peaks of 600V upon engine shutdown.
Furthermore, removal of regulation devices, particularly the battery, will raise normally innocuous
effects to dangerous levels. For example, ignition pulses up to 75V and 90l1s in duration have bee~
observed with the battery disconnected.
Chapter 6 provides a comprehensive review of automotive transients and practical suppression
techniques to protect automotive electronics.
1.6
EFFECTS OF VOLTAGE TRANSIENTS
1.6.1 Effects on Semiconductors
Most semiconductor devices are intolerant of voltage transients in excess of their voltage ratings.
Even such a short-lived transient as a few microseconds can cause the semiconductor to fail catastrophically or may degrade it so as to shorten its useful life.
Frequently, damage occurs when a high reverse voltage is applied to a non-conducting PN junction. The junction may avalanche at a small point due to the non-uniformity of the electric field.
Also, excess leakage current can occur across the passivated junction between the terminations on
the pellet surface. The current can create a low resistance channel that degrades the junction blocking voltage capability below the applied steady-state voltage. In the avalanche case, thermal runaway
can occur because of localized heating building up to cause a melt-through which destroys the
junction.
If the base-emitter junction of a transistor is repetitively' 'avalanched" or' 'zenered" by a reverse
pulse, the forward current gain may be degraded. The triggering sensitivity of a thyristor will be reduced
in the same manner by "zenering" the gate-cathode junction. Thyristors can also be damaged if turned
on by a high voltage spike (forward breakover) under bias conditions that allow a rate of current increase
(di/dt) beyond device capability. This will occur in virtually all practical circuits because the discharge
of the RC dv I dt protection circuits will exceed device capability for dil dt and destroy the thyristor.
1.6.2 Effects on Electromechanical Contacts
The high voltage generated by breaking current to an inductor with a mechanical switch will
ultimately cause pitting, welding, material transfer, or erosion of the contacts. The nature of ultimate failure of the contacts depends upon such factors as the type of metal used, rate of opening,
contact bounce, atmosphere, temperature, steady-state and in-rush currents, and ac or dc operation.
Perhaps most important is the amount of energy dissipated in each operation of the contacts.
The actual breaking of current by a set of contacts is a complex operation. The ultimate break
occurs at a microscopic bridge of metal which, due to the inductive load, is forced to carry nearly
all the original steady-state current. Ohmic heating of this bridge causes it to form a plasma, which
will conduct current between the contacts when supplied with a current and voltage above a certain
threshold. The inductor, of course, is morethan happy to supply adequate voltage (EL = - L di/dt).
16
As the contacts separate and the current decreases, a threshold is reached, and the current stops abruptly ("chopping"). Inductor current then charges stray capacitances up to the breakdown voltage of
the atmosphere between the contacts. (For air, this occurs at 30kV lin.) The capacitance discharges and
recharges repeatedly until all the energy is dissipated. Thi~ arc causes sufficient contact heating to melt,
oxidize, or "bum" the metal, and when the contacts close again, the contacts may form a poorer
connection. If they' 'bounce, " or are closed soon after arcing, the contacts may be sufficiently molten
to weld closed. Welding can also occur as a result of high in-rush currents passing through the inti ally
formed bridges upon closing.
Good suppression techniques can significantly reduce the amount of energy dissipated at the
contacts, with a proportional increase in operating life. Suppression can also reduce the noise generated by this arcing. Voltage-limiting devices are particularly suited to preventing the noisy high-voltage "showering" arc described above and illustrated in Section 1.IA.
1.6.3 Effects on Insulation
Transient overvoltages can cause breakdown of insulation, resulti~g in either a temporary disturbance of device operation or instantaneous failure. The insulating level in the former case will be
weakened leading to premature failure.
The severity of the breakdown varies with the type of insulation - air, liquid, or solid. The
first two tend to be self-healing, while breakdown of solid insulation (generally organic materials)
is generally a permanent condition.
Air clearances between metal parts in electrical devices and power wiring constitute air gaps,
which behave according to the usual physics of gap breakdown (pressure, humidity, shape of electrodes, spacing). The International Electrotechnical Commission Working Group on Low Voltage
Insulation Coordination has developed a table listing the minimum clearances in air for optimum
and worst case electric field conditions existing between electrodes. 14 Breakdown of the clearance
between metal parts can be viewed as a form of protection, limiting the overvoltage on the rest of
the circuit. However, this protection is dependent upon the likelihood of ac line current that may
follow during the arc breakdown. Normally, follow-on current should cause the system fuse or
breaker to function. If the follow-on current heat is limited by circuit impedance, then the system fusing
may not operate. In that case, sufficient heat could be generated to cause a fire. Experience with
power wiring has shown that metal clearances flash-over regularly and harmlessly under transient
voltage conditions, and power follow-on problems are rare, but can occur.
In liquid dielectrics, an impulse breakdown not followed by a high current is normally harmless.
However, this type of breakdown is of limited interest in low-voltage systems, where liquid insulation
systems are seldom used, except in combination with some degree of solid insulation.
Breakdown of solid insulation generally results in local carbonization of an organic material.
Inorganic insulation materials are generally mechanically and permanently damaged. When no
power follow-on current takes place, the system can recover and continue operating. However, the
degraded' insulating characteristic of the material leads to breakdown at pmgressively lower levels
until a mild overvoltage, even within ac line overvoltage tolerances, brings about the ultimate permanent short circuit. Since the final failure can occur when no transients are present, the real cause
of the problem may be concealed.
Breakdown along surfaces of insulation is the concern of "creepage" specifications. The working
group of IEC cited above is also generating recommendations on creepage distances. The behavior
of the system where creepage is concerned is less predictable than is breakdown of insulation in the
bulk because the environment (dust, humidity) will determine the withstand capability of the creepage surface.
17
When considering the withstand capabilities of arty insulation system, two fundamental facts
must be remembered. The first is that breakdown of insulation is not instantaneous but is governed
by the statistics of avalanche ionization. Hence there is. a "volt-time" characteristic, which challenges the designer to coordinate protection systems asa function of the impinging waveshape. The
second is that the distribution of voltage across insulation is rarely linear. For example, a steep wave
front prod~ces a piling up of voltage in the first few turns of a motor winding, often with reflections inside the winding. Also, the breakdown in the gap between the electrodes, initiating at the
surface, is considerably dependent upon the overall field geometry, as well as on macroscopic sur-,
face conditions.
1.6.4 Effects on Power Consumption
As a result of the increasing emphasis on energy conservation, a number of transient voltage
suppression devices have been offered for sale as energy savers. The premise seems to be that transient
overvoltages woul<,l cause degradation of electrical equipment leading to increased loss~s and thus to a
waste of energy. No convincing proof has been offered to support this claim, and injunctions against
making such claims have been obtained in several states. 15
1.6.5 Noise Generation
I
With sensitive logic gates gaining popularity, noise problems are frequent, especially in environments with electromechanical devices. Noise can upset automatic manufacturing equipment, medical equipment, computers, alarms and thyristor-controlled machinery. Such disruption can cause
loss of product, time, money, and even human life.
'
Noise enters a system either directly on wires or grounds connected to the source or through
coupling to adjacent wires. Noise problems are dealt with by suppression at the source, at the receiver, or by isolation. Noise is induced ,when stray capacitange or mutual inductance links the
susceptible system to the noise-generating system. The amplitude of the induced noise is then
related to the rate-of-change of either the current or the voltage of the noise source. The low-frequency components of the induced noise (which are hardest to filter out) are a result of the
amplitude of the original transient impulses.
Frequently, the source of noise is the arcing of contacts breaking current through an inductor,
such as a relay coil. A low-current, high-voltage arc creates a series of brief discharges of a damped
oscillatory nature, occurring at kHz to MHz frequencies with amplitudes of from .300 to several
thousand volts. These pulses ,and their reflections from loads and line discontinuities tr~vel along
the power wires, easily inducing noise in adjacent wiring. This interference is best eliminated by
preventing it at the source (the inductance) with voltage-limiting devices such as varistors.
1.7
TRANSIENT DETECTION
Voltage transients are brief and unpredictable. These twp characteristics make it difficult to detect
and measure them. Even transients described earlier in this chapter as "repeatable" are subject to
variations resulting from the timing of the switching operation, the erratic bouncing of c'ontacts, and
other random combinations.
The transient detector par excellence is the high-frequency storage oscilloscope, but its cost
limits its availability. Custom systems have been built to monitor transients on location,16.17 but
cost ha~ been a limiting factor in this method of detection as we\I. A conventional oscilloscope with
high-frequency response can be used as a monitor if it is proyided with single-sweep controls for
monitoring transients occurring at random times but at relatively low frequent rates. The operator sets
the trigger controls at some threshold level in single-sweep mode and watches the' 'ready" light on the
oscilloscope panel while a camera with open shutter records the screen display. The film is pulled after
the operator notices that sweep occurred and, thus, a record is obtained. While not very efficient for
18
efficient for extensive monitoring, this method is very effective for short-term panics - the most frequent situation when transients are suspected. Digital storage oscilloscopes with automatic data transfer
to a magnetic disc are now available for unattended monitoring.
In recent years, leading oscilloscope manufacturers have developed improved versions of storage oscilloscopes and high-frequency oscilloscopes, and most laboratories now are equipped with
one or another. The experienced engineer can put them to work and obtain satisfactory recordings
by the technique described above, using normal safeguards against erroneous recordings (check on
noise background, stray ground currents, radiation of noise into preamplifier circuits, etc.).
A wide variety of suitable analog or digital test instruments is commercially available, as indicated in Chapter 7. These allow .economical monitoring of a remote location by providing various
degrees of storage (single-event recording, counting above a threshold, digital memory for playback, etc.).
Trade magazines and engineering papers have also described a number of homemade detectors.
While these are undoubtedly performing to the satisfaction of their creators, one can question the economic wisdom of investing time and engineering resources to duplicate, debug, and calibrate a
homemade device when so many commercial units offering guaranteed and instant performance
are available.
1.8
TRANSIENT TESTING AND STANDARDS
It is desirable to have test criteria and definitions that p~ovide a common engineering language
beneficial to both the user and manufacturer of surge protective devices. Regretfully, different
terms have come into use through industry practice over the years. Testing standards have tended
to proliferate as the measurement objective defines either the characteristics of the protective device
or the environment of the application.
The characteristics of each device will vary according to its basic construction. Protective devices
are diverse, being based on ionized gas breakdown, semiconductor junction breakdown, and "charge
hopping" conduction. For this reason, it seems sensible to group devices by physical category and set
up pertinent standards that are suitable for characterizing their behavior. The standards would use
appropriate stress levels and measure those parameters that are critical to ensuring proper performance.
The application environment has demanded different conditions of transient levels. Standards
vary depending on system usage, whether protection is intended for power lines, telecommunications, automotive, or aircraft, to name a few. Each environment also has been defined with less than
full precision leading to. additional diversity on choice of waveshape, amplitude and duration.
Several organizations such as ANSI/IEEE, IEC, UL, NEMA are currently developing guidelines
and standards to describe what the environment is likely to be, on the basis of accumulated recording and
field experience. From this, test specifications are being preparedI7.18.19.20 that will allow objectives and
realistic evaluation of suppressor applications. A brief review ofthese will be found in Chapter 7.
19
REFERENCES
1. Martzloff, F. D., "The Development of a Guide on Surge Voltages in Low-Voltage AC Power
Circuits," Report 81CRD047, General Electric, Schenectady, New York 1981.
2. Greenwood, Allan, Electrical Transients in Power Systems, Wiley-Interscience, New York,
1971.
3. Bodle, D.W., A.J. Ghaze, M. Syed and R.L. Woodside, Characterization ofthe Electrical Environment, Toronto: University of Toronto Press, 1970.
4. Recommended Environmental Practices for Electronic Equipment Design, Publication SAE
11211, Society of Automotive Engineers, Warrendale, Pennsylvania.
5. Hahn, G.J. and F .D. Martzloff, "Surge Voltages in Residential and Industrial Power Circuits,"
IEEE Trans. PAS-89, No .. 6, July-Aug. 1970, pp~ 1049~1056.
6. Bull, J. H., "Voltage Spikes in L. V. Distribution Systems and Their Effects on the Use of Electronic Control Equipment," Report No. 5254, Electrical Research Assn., Cleeve Rd.,
Leatherhead, Surrey, Great,Britain, 1968.
7. IEEE Surge Protective Devices Committee, "Bibliography on Surge Voltages in AC Power
Circuits Rated 600 Volts or Less," IEEE Trans. PAS-89, No.6, July-Aug. 1970, pp. 10561061.
8. Data contributed by L. Regez (Landis & Gyr. Co., Zug, Switzerland), Swiss representative to
IEC Working Group 28A on Low Voltage Insulation,Coordination.
9. Surge Arresters for Alternating Current Power Circuits, ANSI Standard C62.1, IEEE Standard
28, 1974.
10. Guide for Surge Withstand Capability (SWC) Test,ANSI Standard C37.90a, 1974,IEEE Stand~
ard 472, 1974.
11. Martzloff, F .D. and F .A. Fisher, "Transient Control Level Philosophy and ImplementationThe Reasoning Behind the Philosophy," 77CHI224-SEMC, Proceedings of the 2nd Symposium
on EMC, Montreux, June 1977.
12. Fisher, F . A., "A Way to Evaluate the Effects of Changing Magnetic Fields on Shielded Conductors," Report 77CRDI58, General Electric, Schenectady, New York, July 1977.
13. Bennison, E., A. Ghazi and P. Ferland, "Lightning Surges in Open Wire, Coaxial and Paired
Cables," IEEE Trans., COM-21, Oct. 1973, pp. 1136-1143.
14. "Insulation Co-ordination Within Low-Voltage Systems, including Clearances and Creepage
Distances for Equipment" IEC Report 664, 1980.
15. "Evaluation of Transient Voltage Suppressors for Saving Electric Energy," EPRI EM -1722,
February 1981.
16. Allen, George W., "Design of Power-Line Monitoring Equipment," IEEE Trans. PAS-90, No.
6, Nov.-Dec. 1971, pp.2604-2609.
17. Herzog, R., "How to Catch a Transient," Machine Design Magazine, March 1973, pp. 170175.
18. Test Specification for Gas Tube Surge Protective Devices, IEEE Std. C62.31, 1981.
19. Test Specifications for Low Voltage Air Gap Surge Protective Devices, Excluding Valve and
Expulsion Type Devices, IEEE Std. C62.32, 1981.
20. Test Specifications for Vati~tor Surge Protective Devices, IEEE Std. C62.33, 1982.
20
The Development of a Guide *
on Su"rge Voltages in Low-Voltage
AC Power Circuits
F.D. Martzloff, Fellow, IEEE
General Electric Company
Schenectady, NY
Surge voltages in ac power circuits become more significant with the increased application of
miniaturized electronics in consumer and industrial products. A Working Group of IEEE has prepared a
Guide describing the nature of these surges in ac power circuits up to _600V , and is currently developing
an Application Guide for using the environmental data defined in the Guide to improve protection.
The Guide proposes two waveforms, one oscillatory, the other unidirectional, depending on the
location within the power system. It also includes recommendations for source impedance or
short-circuit current. This paper describes the data base and approach used by the Working Group and
the recommendations proposed to represent typical surges, in order to obtain feedback for writing the
Application Guide.
INTRODUCTION
Surge voltages occurring in ac power circuits can be the cause of misoperation or product failure
for residential as well as industrial systems. The problem has received increased attention in recent
years because miniaturized solid state devices are more sensitive to voltage surges (spikes and
transients) than were their predecessors.
Although surge voltage amplitudes and their frequency of occurrence on unprotected circuits are
well known, their waveshapes and energy content are less well known. On the basis of measurements,
statistics, and theoretical considerations, a practical guide for outlining the environment for use in
predicting extreme waveshapes and energy content can nevertheless be established. A Working Group
of the Surge Protective Devices Committee has completed such a descriptive Guide; I and is now
developing an Application Guide. This paper presents a brief overview of the Guide and of the approach
taken in the Application Guide. Discussion and comments on the objectives and approach of the
Application Guide are welcome.
SCOPE
Both guides primarily address ac power circuits with rated voltages up to 277V-line to ground,
although some of the conclusions offered could apply to higher voltages and also to some dc power
systems. Other standards have been established,2 such as IEEE Standard 472, Guide for Surge
Withstand Capability (SWC) Tests, intended for the special case of high-voltage substation
environments, and IEEE Standard 28, Standard for Surge Arrestors for ac Power Circuits, 3 covering
primarily the utilitie!i environment. The Guides are intended to complement, not conflict with, existing
standards.
The surge voltages considered in the Guides are those exceeding two per unit (or twice the peak
operating voltage) and having durations ranging from a fraction of a microsecond to a millisecond.
Overvoltages of less than two per unit are not covered, nor are transients of longer duration resulting
from power equipment operation and failure modes. Because these low-amplitude and long-duration
·Reprinted with permission ofIEEE from the 14th Electrical/Electronics Insulation Conference, IEEE, Boston, October 8-11, 1979.
21
surges are generally not amenable to 'Supression by conventional surge protective devices, they require
dlfferent protection techniques.
While the major purpose of the existing Guide is to describe the environment, a secondary purpose
is to lead toward standard tests that will appear eventually in the Application Guide.
THE ORIGINS OF SURGE VOLTAGES
Surge voltages occurring in low-voltage ac power circuits originate from two major sources:
system switching transients and direct or indirect lightning effects on the power system. System
switching transients can be divitled into transients associated with (1) major power system switching
disturbances, such as capacitor bank switching; (2) minor switching near the point of interest, such as an
appliance turnoff in a household or the turnoff of other loads in an individual system; (3) resonating
circuits associated with switching devices, such as thyristors; and (4) various system faults, such as
short circuits and arcing faults.
«
V!
V!
U
.gj
...«
102
0
...>
lV!
w
cr:
...
U
10'
0
V!
V!
w
U
xw
~
cr:
;:j
>
cr:
~
V!
w
(!)
cr:
:J
V!
SURGE CREST - kV
FIGURE 1. RATE OF SURGE OCCURRENCE VERSUS VOLTAGE LEVEL AT UNPROTECTED LOCATIONS
*IN SOME LOCATIONS, SPARKOVER OF CLEARANCES MAY LIMIT THE OVERVOL TAGES.
Measurements and calculations of lightning effects have been made to yield data on what levels
can be produced, even if the exact mechanism of any particular surge is unknown. While the data have
been recorded primarily on 120,220/380, or 277/480V systems, the general conclusions should be
valid for600V systems. To the extent that surge voltages are produced by a discrete amount of energy
being dumped into a power system, low-impedance, heavy industrial systems can be expected to
experience lower peaks from surge voltages than 120V residential systems, but comparable, or greater,
amounts of energy potentially available for deposition in a surge suppressor.
RATE OF OCCURRENCE AND VOLTAGE LEVElS IN UNPROTECTED CIRCUITS
The rate of occurrence of surges varies over wide limits, depending on the particular power
system. Prediction of the rate for a particular system is always difficult and frequently impossible. Rate
is related to the level of the surges; low-level surges are more prevalent than high-level surges.
22
It is essential to recognize that a surge voltage observed in a power system can be either the driving
voltage or the voltage limited by the sparkover of some clearance in the system. Hence, the term
unprotected circuit must be understood to be a circuit in which no low-voltage protective device has
been installed but in which clearance sparkover will eventually limit the maximum voltage. The
distribution of surge levels, therefore, is influenced by the surge-producing mechanisms as well as by
the sparkover level of clearances in the system. This distinction between actual driving voltgage and
voltage limited by sparkover is particularly important at the interface between outdoor equipment and
indoor equipment. Outdoor equipment has generally higher clearances, hence higher sparkover levels:
10kV may be typical, but 20kV is possible. In contrast, most indoor wiring devices used in 120-240V
systems have sparkover levels of about 6kV; this 6kV level, therefore, can be selected as a typical cutoff
for the occurrence of surges in indoor power systems.
Data collected from many sources have led to the plot shown in Figure 1. This prediction shows
with certainty only a relative frequency of occurrence, while the absolute number of occurrences can be
described only in terms of "low exposure," "medium exposure," or "high exposure." These
exposure levels can be defined in general terms as follows:
Low Exposure Systems in geographical areas known for low lightning activity, with little load
switching activity.
Medium Exposure Systems in geographical areas known for high lightning activity, with
frequent and severe switching transients.
High Exposure Rare but real systems supplied by long overhead lines and subject to reflections at
line ends, where the characteristics is the installation produce high sparkover levels of the clearances.
The two lower lines of Figure 1 have been drawn at the same slope, since the data base shows
reasonable agreement among several sources on that slope. All lines may be truncated by sparkover of
the clearances at levels depending on the withstand voltage of these clearances. The' 'high-exposure"
line needs to be recognized,but it should not be applied indiscriminately to all systems. Such
application would penalize the majority of installations, where the exposure is lower.
From the relative values of Figure 1, two typical levels can be cited for practical applications.
First, the expectation of 3kV transient occurrence on a 120V circuit ranges from 0.01 to 10 per year at a
given location-a number sufficiently high to justify the recommendation of a minimum 3kV withstand
capability. Second, the sparkover of wiring devices indicates that a 6kV withstand capability may be
sufficient to ensure device survival indoors, but a withstand capability of lOkV, or greater, may be
required outdoors.
The voltage and current amplitudes presented in the Guide attempt to provide for the vast majority
of lightning strikes but should not be considered as "worst case," since this concept cannot be
determined realistically. One should think in terms of the statistical distribution of strikes, accepting a
reasonable upper limit for most cases. Where the consequences of a failure are not catastrophic but
merely represent an annoying economic loss, it is appropriate to make a tradeoff of the cost of protection
against the likelihood of failure caused by a high but rare surge. For instance, a manufacturer may be
concerned with nation-wide failure rates, those at the upper limits .of the distribution curve, while the
user of a specific system may be concerned with a single failure occurring at a specific location under
"worst-case conditions." Rates can be estimated for average systems, however, and even if imprecise,
they provide manufacturers and users with guidance. Of equal importance is the observation that surges
in the range of 1 to 2kV are fairly \common in residential circuits.
Surges occur at random times with respect to the power frequency, and the failure mode of
equipment may be affected by the power frequency follow current. Furthermore, the timing of the surge
with respect to the power frequency may affect the level at which failure occurs. Consequently, when
the failure mode is likely to be affected, surge testing should be done with the line voltage applied to
the test piece.
23
Lightning surges are completely random in their timing with respect to the power frequency.
Switching surges are likely to occur near or after current zero, but variable load power f~ctors will
produce a quasi-random distribution. Some semiconductors exhibit failure levels that depend on the
timing of the surge with respect to the conduction of power frequency current. Gaps or other devices
involving a power-follow current may withstand this power follow with success, depending upon the
fraction of the half-cycle remaining after the surge before current zero. Therefore, it is important to
consider the timing of the surge with respect to the power frequency. In performing tests, either
complete randomization of the timing or controlled timing should be specified,; with a sufficient number
of timing conditions to reveal the most critical timing.
WAVESHAPEOFREPRESENTATIVE SURGE VOLTAGES
Waveshapes in Actual Occurrences
Indoor Measurements in the field, measurements in the laboratory, and theoretical calculations
indicate that most surge voltages in indoor low-voltage systems have oScill.atory waveshapes, unlike the
well-known and generally accepted unidirectional waves specified in high-voltage insulation standards.
A surge impinging on the system excites the natural resonant frequencies of the conductor system. As a
result, not only are the surges typically oscillatory, but surges may have different amplitudes and
waveshapes at different places in the system. These oscillatory frequencies of surges range from 5kHz
to more than 500kHz. A 30 to 100kHz frequency is a realistic measure of a "typical" surge for most
residential and light industrial ac line networks.
Outdoor and Service Entrance Surges encountered in outdoor locations have also been
recorded, some oscillatory, other unidirectional. The" classical lightning surge" has been established
as 1.2 I 50jls for a voltage wave and 8 I 20jls for a current wave, but these waveshapes should not be
construed as typical waves for low-voltage circuits. Lightning discharges induce oscillations,
reflections, and disturbances that ultimately appear as decaying oscillations in low-voltage systems.
Lenz 4 reports 50 lightning surges recorded in two locations, the highest at 5.6kV, with frequencies
ranging from 100 to 500kHz. MartzloffS reports oscillatory lightning surges in a house during a
multiple-stroke flash.
Because the prime concern here is the energy associated with these surges, thewaveshape to be
selected must invol ve greater energy than that associated with the indoor environment. Secondary surge
arresters have a long history of successful performance, meeting the ANSI C62.1 specification, as
detailed below; consequently, these specifications can be adopted as a realistic representation of
outdoor waveshapes.
Selection of Representative Waveshapes
The definition of a waveshape to be used as representative of the environment is important for the
design of candidate protective devices, since unrealistic requirements, such as excessive duration of the
voltage or very low source impedance, place a high energy requirement on the suppressor, with a
resulting cost penalty to the end user. The two requirements defined below reflect this tradeoff.
Indoor Based on measurements conducted by several independent organizations in 120 and 240V
systems, the waveshape shown in Figure 2 is reasonably representative of surge voltages in these power
circuits. Under the proposed description of a "0.5jls - 100kHz ring wave," this waveshape rises in
O.5jls, then decays while oscillating at 100kHz, each peak being about 60% of the preceding peak.
The fast rise can produce the effects associated with nonlinear voltage distribution in windings and
the dv/dt effects on semiconductors. Shorter rise times are found -in many transients, but, as those
transients propagate into the wiring or are reflected from discontinuities in the wiring, the rise time
becomes longer.
24
The oscillating and decaying tail produces the effects of voltage polarity reversals in surge
suppressors or other devices that may be sensitive to polarity changes. Some semiconductors are
particularly sensitive to damage when being forced into or out of a conducting state, or when the
transient is applied during a particular portion of the 60Hz supply cycle.
Outdoor In the outdoor and service entrance environment, as well as in locations close to the
service entrance, substantial energy, or current, is still available, in contrast to the indoor environment,
where attenuation has taken place. For these locations, the unidirectional impulses long established for
secondary arresters are more appropriate than the oscillatory wave.
Accordingly, the recommended waveshape is .1.2 / 50J.ts for the open-circuit voltage or voltage
applied to a high-impedance device, and 8 / 20J.ts for the discharge current or current in a low-impedance
device. The numbers used to describe the impulse, 1.2 / 50 and 8 / 20, are those defined in IEEE
Standard 28 - ANSI Standard C62.1; Figure 3 presents the waveshape and a graphic description of
the numbers.
0.9
v,.
--v,.
..
0.1
T
= 10,.
II
= 100 kHz) .,
v,.
0.5115-
FIGURE 2. THE PROPOSED 0.5 p.s -100 kHz RING WAVE (OPEN-CIRCUIT VOLTAGE)
ENERGY AND SOURCE IMPEDANCE
General
The energy involved· in the interaction of a power system with a surge source and a surge
suppressor will divide between the source and the suppressor in accordance with the characteristics of
the two impedances. In a gap-type suppressor, the low impedance of the arc after sparkover forces most
of the energy to be dissipated elsewhere: for instance, in a resistor added in series with the gap for
limiting the power-follow current. In an energy-absorber suppressor, by its very nature, a substantial
share of the surge energy is dissipated in the suppressor, but its clamping action does not involve the
power-follow energy resulting from the short-circuit action of a gap. It is therefore essential to the
effective use of suppression devices that a realistic assumption be made about the source impedance of
the surge whose effects are to be duplicated.
The voltage wave shown in Figure 2 is intended to represent the waveshape a surge source would
produce across an open circuit. The waveshape will be different when the source is connected to a load
having a lower impedance, and the degree to which it is lower is a function of the impedance
of the source.
•
To prevent misunderstanding, a distinc\ion between source impedance and surge impedance needs
to be made. Surge impedance, also called characteristic impedance, is a concept relating the parameters
25
of a line to the propogation of traveling waves. For the wiring practices of the ac power circuits
discussed here, this characteristic impedance would be in the range of 150 to 300n, but because the
durations of the waves being discussed (50 to 20l-ts) are much longer than the travel times in the wiring
systems being considered, traveling wave analyses are not useful here.
Souce impedance, defined as "the impedance presented by a soilrce of energy to the input
terminals of a device, or network" (IEEE Standard 100), is a more useful concept here. In the
conventional Thevenin' s description, the open-circuit voltage (at the terminals of the network or test
generator) and the source impedance (of the surge source or test generator) are sufficient to calculate the
short-circuit current, as well as any current for a specified suppressor impedance.
The measurements from which Figure 1 was derived were of voltage only. Little was known about
the impedance of the circuits upon which the measurements were made. Since then, measurements have
been reported on the impedance of power systems,. Bull6 reports that the impedance of a power system,
seen from the outlets, exhibits the characteristics of a 50n resistor with 50l-tH in parallel. Attempts were
. made to combine the observed 6kV open-circuit voltage with the assumption of a 50n/50l-tH
v
0.9 V""
f----,
0.9 Ipk
t---I
0.11""
201"
T, x 1.67 = 1.2 ••
FIGURE 3. UNIDIRECTIONAL (ANSI STANDARD C62.1) WAVESHAPES (A) OPEN-CIRCUIT VOLTAGE WAVEFORM
(B) DISCHARGE CURRENT WAVEFORM
impedance. 7 This combination resulted in low energy deposition capability, which was contradicted by
field experience of suppressor performance. The problem led to the proposed definition of oscillatory
waves as well as high-energy unidirectional waves, in order to produce both the effects of an oscillatory
wave and the high-energy deposition capability.
The degree to which source impedance is important depends largely on the type of surge
suppressors that are used. The surge suppressors must be able to withstand the current passed through
them by the surge source. A test generator of too high an impedance may not subject the device under
test to sufficient stresses, while a generator of too Iowan impedance may subject protective devices to
unrealistically severe stresses. A test voltage wave specified without reference to source impedance
could imply zero source impedance-one capable of producing that voltage across any impedance, even
a short circuit. That would imply an infinite surge current, clearly an unrealistic situation.
Because of the wide range of possible source impedances and the difficulty of selecting a specific
value, three broad categories of building locations are proposed to represent the vas! majority of
locations, from those nea: the service entrance to those remote from it. The source impedance of the
surge increases from the outside to locations well within the building. Open-circuit voltages, on the
26
other hand, show little variation within a building because the wiring provides little attenuation. Figure
4 illustrates the application of the three categories to the wiring of a building.
For the two most common location categories, Table 1 shows the recommended surge voltages and
currents, with the waveforms and amplitudes ofthe surges, and high- or low-impedance specimen. For
the discharge current shown, the last two columns show the energy that would be deposited in a
A
C
B
, '\
I
L ___ -,
I
FIGURE 4. LOCATION CATEGORIES
A. Outlets and Long Branch Circuits
All outlets at more than 10m (30
ft) from Category 8 with wires
#14-10
All outlets at more than 20m (60
ftl from Category C with wires
#14-10
B. Major Feeders and Short Branch
Circuits
Distribution panel devices
Bus and feeder system~ in
industrial plants
Heavy appliance outlets with
" short" connections to the
service entrance
Lighting systems in commercial
buildings
C. Outside and Service Entrance
Service drop from pole to
building entrance
Run between meter and
distribution panel
Overhead line to detached
buildings
Underground lines to well pumps
suppressor clamping at 500V and lOOOV, typical of 120V or 240V applications, respectively. For
higher system voltages (assuming the same current values), the energy would increase in proportion to
the clamping voltage of a suppressor suitable for that system voltage.
The values shown in Table 1 represent the maximum range and correspond to the "medium
exposure" situation of Figure 1. For less exposed systems, orwhen the prospect of a failure is not highly
objectionable, one could specify lower values of open-circuit voltages with corresponding reductions in
the discharge currents.
The 6kV open-circuit voltage derives from two facts: the limiting action of wiring device
sparkover and the un attenuated propagation of voltages in unloaded systems. The 3kA discharge current
in Category B derives from experimental results: field experience in suppressor performance and
27
TABLE 1. SU~Gl~ VOLTAGES AND CURRENTS DEEMED TO REPRESENT THE .INDOOR
ENVIRONMENT AND RECOMMENDED FOR USE IN DESIGNING PROTECTIVE SYSTEMS .
LOCATION
CATEGORY
IMPULSE
COMPARABLE
TO IEC 664
CATEGORY
A. Long branch
circuits and
outlets
B. Major feeders
short branch
circuits, and
load center
II
WAVEFORM
6kV
200A
High impedance(l'
Low impedance<2l
1.2/50/-ls
8/20/-ls
6kV
3kA
High impedance")
Low impedance12l
0.5/-1s - 100kHz
6kV
500A
High impedance(l)
Low impedance
0.5/-1s - 100kHz
,
III
MEDIUM EXPOSURE
AMPLITUDE
TYPE
OF SPECIMEN
OR LOA.D
CIRCUIT
-
ENERGY(JOULES)
DEPOSITED IN A SUPPRESSOR"'
WITH CLAMPING VOLTAGE OF
5DDV
lDDDV
(l20V System) (240V System)
0.8
1.6
-
-
40
80
-
-
2
4
Notes: (1) For high-impedance test specimens or load circuits, the voltage shown represents the surge voltage. In
making simulation tests, use that value for the open-circuit voltage of the test generator.
(2) For low-impedance test specimens or load circuits, the current shown represents the discharge current of the
surge (not the short-circuit current of the power system). In making simulation tests, use that current for the
short-circuit current of the test generator.
(3) Other suppressors which have different clamping voltages would receive different energy levels.
simulated lightning tests. The two levels of discharge currents from the O.5/Ls - 100kHz wave derive
from the increasing impedance expected in moving from Category B to Category A.
Location Category C is likely to be exposed to substantially higher voltages than location Category
B because the limiting effect of sparkover is not available. The' 'medium exposure" rates of Figure 1
could apply, with voltage in excess of 10kV and discharge currents of 1OkA, or more. Installing
unprotected load equipment in location Category C is not recommended; the installation of secondary
arresters, however, can provide the necessary protection. Secondary arresters having 10kA ratings have
been applied successfully for many years in location Category C (ANSI Standards C62.1 and C62.2) .
.Subcommittee 28A of the International Electrotechnical Commission has also prepared a reportS in
which installation categories are defined. These installation categories divide the power systems
according to the location in the building, in a manner similar to the location categories defined in the
Guide. The three categories presented here are comparable to three of the four categories defined by
Subcommittee 28A, with the added specification of a source impedance or discharge current.
Equipment connected to the outlets of Category A corresponds to equipment located in IEC Category I;
Category C corresponds to IEC Category IV.
There are, however, some significant differences between the two concepts. First, the IEC
categories are defined for a "Controlled Voltage Situation, " a phrase that implies the presence of some
surge suppression device or surge attenuation mechanism to reduce the voltage levels from one category
to the next. Second, the IEC report is more concerned with insulation coordination than with the
application of surge protective devices; therefore it does not address the question of the coordination of
the protectors, but rather the coordinatio~ of insulation levels-that is, voltages. Source impedances, in
contrast to the Guide, have not been defined. Further discussion and work toward the Application Guide
and the IEC Report should eventually produce a consistent set of recommendations.
APPLICATION GUIDE
The broad range of surge voltages occurring in low-voltage ac power circuits can be simulated by a
limited set of test waves , for the purpose of evaluating their effects on equipment. Field measurements,
laboratory experiments, and calculations indicate that two basic waves, at various open-circuit voltages
and short-circuit current values, can· represent the majority of surges occurring in residential,
28
commercial, and light industrial power systems rated up to 600V RMS.
Exceptions will be found to a single, broad guide; however, such exceptions should not detract
from the benefits that can·be expected from a reasonably valid uniformity in defining the environment.
Test waves of different shapes may be appropriate for other purposes, and the Guide should not be.
imposed where it is not applicable. The forthcoming Application Guide will detail waves and shapes for
specific applications.
Good Pf.ctices
lOt "otacttOn
(WIring. Sh••ldlng, Bondingl
FIGURE 5. COORDINATION OF PROTECTION SCHEMES
The Application Guide approach will be to compare the known characteristics of the environment,
as defined by the Guide on Surge Voltages, and the known (orto be determined) withstand capability of
the unprotected load circuit to be installed in that environment. Through a systematic process of analysis
and/or test, using the Transient Control Level concept9 for instance, the user will be able to determine
whether or not protection is required, what coordination is necessary, and what economic/performance
tradeoffs can be made. Figure 5 shows a flow chart proposed to guide this evaluation process.
The Working Group, in coordination with other interested organizations, will proceed toward the
preparation of this Application Guide; the Working Group invites comments and suggestions.
29
REFERENCES
1. Guide on Surge Voltages in AC Power Circuits Rates Up to 600V, Final Draft, May 1979.
Document prepared by Working Group 3.4.4 of the Surge Protective Device Committee of the
Power Engineering Society, Institute of Electrical and Electronics Engineers (Now ANSI/IEEE
..
Standard C62.41-1980).
2. Guide for Surge Withstand Capability (SWC) Tests, ANSI Standard C37. 90a, 1974; IEEE Standard
472-1974.
3. Surge Arrestors for Alternating-Current Power Circuits, IEEE Standard 28-1974; ANSI Standard
C62.1-1975; IEC Standard 99-2.
4. Lenz, J .E., "Basic Impulse Insulation Levels of Mercury Lamp Ballast for Outdoor Applications,"
Illuminating Engrg., February 1964,pp. 133-140.
5. Martzloff, F.D. and G.J. Hahn, "Surge Voltage in Residential and Ind~strial Power Circuits,"
IEEE PAS-89, 6, July/August 1970, 1049-1056.
6. Bull, J .H., "Impedance on the Supply Mains at Radio Frequencies," Proceedings of 1st
Symposium on EMC, 75CH1012-4 Mont., Montreux, May 1975.
7. Martzloff, F.D. and F.A. Fisher, "Transient Control Level Philosophy and Implementation: The
Reasoning Behind the Philosophy," Proceedings 2nd Symposium on EMC, 77CHI224-5EMC,
Montreux, June 1977.
8. Insulation Coordination Within Low-Voltage Systems Including Clearances and Creepage
Distances for Equipment, International Electrotechnical Commission Report 664.
9. Fisher, F.A. and F.D. Martzloff, "Transient Control Levels, a Proposal for Insulation
Coordination in Low-Voltage Systems," IEEE PAS-95, 1, January/February 1976, pp. 120-129.
ACKNOWLEDGMENTS
The concepts presented in this paper have greatly benefited from the informed questions and
discussions by members oftheWorking Group on Surge Voltage in AC Power Circuits Rated 600V or
Less, and from interested reviewers; particular appreciation for effective critiques from Catharine
Fisher and Peter Richman is acknowledged. The data base used in developing the Guide was broadened
by the contributions of the Bell Telephone Laboratories and Landis & Gyr, Inc. The flow chart approach
to the Application Guide was suggested by Paul Speranza.
30
TRANSIENT SUPPRESSION - DEVICES
AND PRINCIPLES
This chapter presents a brief description of available transient suppressors and their operation, and
discusses how these devices can be applied.
2.1 TRANSIENT SUPPRESSION DEVICES.
There are two major categories of transient suppressors: a) those that attenuate transients, thus
preventing their propagation into the sensitive circuit; and b) those that divert transients away from
sensitive loads and so limit the residual voltages.
Attenuating a transient-that is, keeping it from propagating away from its source or keeping it
from impinging on a sensitive load-is accomplished with filters inserted in series within a circuit. The
filter, generally ofthe low-pass type, attenuates the transient (high frequency) and allows the signal or
power flow (low-frequency) to continue undisturbed.
Diverting a transient can be accomplished with a voltage-clamping type device or with a
"crowbar" type device. The designs of these two types, as well as their operation and application, are
different enough to warrant a brief discussion of each in general terms. A more detailed description will
follow later in this chapter.
A voltage-clamping deviceis a component having a variable impedance depending on the current
flowing through the device or on the voltage across its terminal. These devices exhibit a non-linear
impedance characteristic-that is, Ohm's law is applicable but the equation has a variable R. The
variation of the impedance is monotonic; in other words, it does not contain discontinuities in contrast to
the crowbar device, which exhibits a turn-on action. The volt-ampere characteristic of these clamping
devices is somewhat time-dependent, but they do not involve a time delay as do the sparkover of a gap or
the triggering of a thyristor.
z ' i.a
1000r---------,---------,---------,--
...
(J)
100
...J
0
>
LINEAR
IIJ
IMPEDANCE:
«
..."
V
I' -
R
...J
0
>
10
NON-LINEAR
IMPEDANCE
(POWER LAW):
0.1
10
CURRENT -
100
1000
AMPERES
FIGURE 2.1: VOL TAGE/CURRENT CHARACTERISTIC FOR A LINEAR 1 OHM RESISTOR
AND NON·LINEAR VARISTOR
31
I' Kv a
With a voltage-clamping device, the circuit is unaffected by the presence of the device before and
after the transient for any steady-state voltage below the clamping level. The voltage clamping action
results from the increased current drawn through the device as the voltage tends to rise. If this current
increase is faster than the voltage rise, the impedance of the device is nonlinear (Figure 2.1). The
apparent "clamping" of the voltage results from the increased voltage drop (IR) in the source
impedance due to the increased current. It should be clearly understood that the device depends on the
source impedance to produce the clamping. One is seeing a voltage divider action at work, where the
ratio of the divider is not constant but changes. However, if the source impedance is very low, then the
ratio is low. The suppressor cannot be effective with zero source impedance (Figure 2.2) and works best
when the voltage divider action can be implemented.
Zs
Zv
VZV =(zv z+v zs) Voe
FIGURE 2.2: DIVISION OF VOLTAGE WITH VARIABLE IMPEDANCE SUPPRESSOR
Crowbar-type devices involve a switching action, either the breakdown of a gas between
electrodes or the tum-on of a thyristor. After switching on, they offer a very low impedance path which
diverts the transient away from the parallel-connected load.
These crowbar devices have two limitations. The first is their delay time, typically microseconds,
which leaves the load unprotected during the initial rise. The second limitation is that a power current
from the steady-state voltage source will follow the surge discharge (called "follow-current" or
"power-follow"). In ac circuits, this power-follow current mayor may not be cleared at a natural
current zero; in dc circuits the clearing is even more uncertain. Therefore, if the crowbar device is not
designed to provide self-clearing action within specified limits of surge energy and system voltage and
power-follow current, additional means must be provided to open the power circuit.
2.1.1 Filters
The frequency components of a transient are several orders of magnitude above the power
frequency of an ac circuit and, of course, a dc circuit. Therefore, an obvious solution is to install a
low-pass filter between the source of transients and the sensitive load.
The simplest form of filter is a capacitor placed across the line. The impedance of the capacitor
forms a voltage divider with the source impedance, resulting in attenuation of the transient at high
frequencies. This simple approach may have undesirable side effects, such as a) unwarited resonances
with inductive components located elsewhere in the circuit leading to high peak voltages; b) high
in-rush currents during switching, or, c) excessive reactive load on the power system voltage. These
undesirable effects can be reduced by adding a series resistor-hence, the very popular use of RC
snubbers and suppression networks. However, the price of the added resistance is less effective
clamping.
Beyond the simple RC network, conventional filters comprising inductances and capacitors are
widely used for interference protection, As a bonus, they also offer an effective transient protection,
provided that the filter's front-end components can withstand the high voltage associated with the
transient.
There is a fundamental limitation in the use of capacitors and filters for transient protection when
the source oftransients in unknown. The capacitor response is indeed nonlinear with frequency, but it is
still a linear function of current.
In Chapter 1, it was explained that to design a protection scheme against random transients, it is
often necessary to make an assumption about the characteristics of the impinging transient. If an error in
the source impedance or in the open-circuit voltage is made in that assumption, the consequences for
a linear suppressor and a nonlinear suppressor are dramatically different as demonstrated by the following comparison.
A SIMPLIFIED COMPARISON BETWEEN PROTECTION WITH LINEAR and
NONLINEAR SUPPRESSOR DEVICES
Assume an open-circuit voltage of 3000V (see Figure 2.2):
1. If the source impedance is Zs = 500
with a suppressor impedance of Zv = 80
the expected current is:
3000
I = 50 + 8 = 51.7A and VR = 8x51.7 = 414V
The maximum voltage appearing across the terminals of a typical nonlinear V130LA20A
varistor at 51.7 A is 330V.
Note that:
Zs x I = 50 x 51.7 = 2586V
VR x I = 8 x 51.7 = 414V:
= 3000V
2. If the source impedance is only 50 (a 10: 1 error in the assumption), the voltage across the same
linear 80 suppressor is:
VR
8
= 3000 5 + 8 = 1850V
However, the nonlinear varistor has a much lower impedance; again, by iteration from the
characteristic curve, try 400V at 500A, which is correct for the V130LA20A; to prove the
correctness of our "educated guess" we calculate I.
I
Zs x I = 5 x 520
Vc =
= 3000 -400V = 520A.
5
= 2600V
400V
= 3000V
which justifies the' 'educated guess" of 500A in the circuit. *
Summary
3000V "'OPEN-CIRCUIT"' TRANSIENT VOLTAGE
Protective Level
Achieved
Linear 80
Nonlinear Varistor
Assumed Source Impedance
500
50
414V
1850V
330V
400V
Similar calculations can be made, with similar conclusions, for an assumed error in
open-circuit voltage at a fixed source impedance. In that case, the linear device is even more
sensitive to an error in the assumption. The calculations are left for the interested reader to
work out.
*An educated guess, orthe result of an iteration - see further in the manual, "Designing with GE-MOViI>II Varistor," Chapter 4.
33
The example calculated in the box shows that a source impedance change from an assumed 500 to
50 can produce a change of about 414V to 1850V for the protective voltage of a typical linear
suppressor. With a typical nonlinear suppressor, the corresponding change is only 330V to 400V. In
other words, a variation of only 21 % in the protective level achieved with a nonlinear suppressor occurs
for a 10 to 1 error in the assumption made on the transient parameters, in contrast to a 447% variation in
the protective level with a linear suppressor for the same error in assumption. Nonlinear
voltage-clamping devices give the lowest clamping voltage, resulting in the best protection against
transients.
2. 1.2 Crowbar Devices·
This category of suppressors, primarily gas tubes (also called "spark gaps") or carbon-block
protectors, is widely used in the communication field where power-follow currentis less of a problem
than in power circuits. Another form ofthese suppressors is the hybrid circuit which uses solid-state or
ionic devices where a control circuit causes tum-on of an active component.
In effect, a crowbar device short-circuits a high voltage to ground. This short will continue until
the current is brought to a low level. A voltage clamping device will never reduce the line voltage below
its steady-state value but the crowbar device often will. Because the voltage (arc or forward-drop)
during the discharge is held very low, substantial currents can be carried by the suppressor without
dissipating a considerable amount of energy within the suppressor. This capability is the major
advantage of these suppressors. However, two limitations must be considered.
Volt-Time Response. When the voltage rises across a spark-gap, no significant conduction can
take place until transition to the arc mode has occurred by avalanche breakdown of the gas between the
electrodes. The delay time, typically microseconds, leaves the load unprotected during the initial rise.
Since the process is statistical in nature, there is aconsiderable variation in the sparkover voltage
obtained in successive operations. For some devices, this sparkover voltage also can be substantially
higher after a long period of rest than after a succession of discharges. From the physical nature of the
process, it is difficult to produce consistent sparkover voltage for low voltage ratings. The difficulty is
compounded by the effect of manufacturing tolerances on very small gap distances. One way to alleviate
the difficulty is to fill the tube with a gas having a lower breakdown voltgage than that of air. However,
this substitution creates a reliability problem if the enclosure seal is lost and the gas is replaced by air.
Some applications require providing a second gap in parallel with the first, with slightly higher
sparkover voltage for backup against failure of the gas tube.
Power-Follow. The second limitation is that a power current from the steady-state voltage source
will follow the surge discharge (called "follow-current" or "power-follow"). In ac circuits, this
power-follow current mayor may not be cleared at a natural current zero; in dc circuits the clearing is
even more uncertain. Therefore, if the crowbar device is not designed to provide self-clearing action
within specified limits of surge energy and system voltage and power-follow current, additional means
must be provided to open the power circuit.
2.1.3 Voltage-Clamping Devices
To perform the voltage limiting function, voltage-clamping devices discussed at the beginning of
the chapter depend on their nonlinear impedance in conjunction with the transient source impedance.
Three types of devices have been used: reverse selenium rectifiers, avalanche (zener) diodes and
varistors made of different materials, I.e., silicon carbide, zinc oxide, etc. i
Selenium Cells. Selenium transient suppressors apply the technology of selenium rectifiers in
conjunction with a special process allowing reverse breakdown current at high-energy levels without
damage to the polycrystalline structure. These cells are built by developing the rectifier elements on the
surface of a metal plate substrate which gives them good thermal mass and energy dissipation
34
performance. Some of these have self-healing charcteristics which allows the device to survive energy
discharges in excess of the rated values for a limited number of operations - characteristics that are
useful, if not "legal" in the unsure world of voltage transients.
The selenium cells, however, do not have the clamping ability of the more modem metal-oxide
varistors or avalanche diodes. Consequently, their field of application has been considerably
diminished.
Zener Diodes. Silicon rectifier technology has improved the performance of regulator-type zener
diodes in the direction of the design of surge-suppression type avalanche diodes, The major advantage
of these diodes is their very effective clamping, which comes closest to an ideal constant voltage clamp.
They are also available in low-voltage ratings.
Since the diode maintains the avalanche voltage across a thin junction area during surge discharge,
substantial heat is generated in a small volume. The major limitation of this type of device is its energy
dissipation capability.
Varistors. A varistor functions as a nonlinear variable impedance. The relationship between the
current in the device, I, and the voltage across the terminals, V, is typically described by a power law: I
= kVO<. While more accurate and more complete equations can be derived to reflect the physics of the
device,2,3 this definition will suffice here. A more detailed discussion will be found in Chapter 3.
The term ex (alpha) in the equation represents the degree of nonlinearity of the conduction. A linear
resistance has an ex = 1. The higher the value of ex, the better the clamp, which explains why ex is
sometimes used as a figure of merit. Quite naturally, varistor manufacturers are constantly striving for
higher alphas.
'
Silicon Carbide. Until recently, the most common type of varistor was made from specially
processed silicon carbide. This material has been and is still very successfully applied in high-power,
high-voltage surge arresters. However, the relatively low ex values of this material produce one of two
results. Either the protective level is too high for a device capable of withstanding line voltage or, for a
device producing an acceptable protective level, excessive stand-by current would be drawn at normal
voltage if directly connected across the line. Therefore, a series gap is required to block the normal
voltage.
A detailed discussion of series gap/silicon carbide block combinations is beyond the scope of this
manual, but many references and standards on the design, testing and application of surge arresters are
available. 4 ,5
In lower voltage electronic circuits, silicon carbide varistors have not been widely used because of
the need for using a series gap, which increases the total cost and reproduces some of the undesirable
characteristics of gaps described earlier. However, this varistor has been used as a current-limiting
resistor to assist some gaps in clearing power-follow current.
Metal-Oxide Varistors. A number of recent developments 6 ,7 have produced a new family of
varistors made of sintered metal oxides, primarily zinc oxide with suitable additives. These new
varistors have ex values considerably greater than those of silicon carbide varistors, typically in the range
of an effective value of 15 to 30 measured over several decades of surge current. One type of varistor,
the General Electric GE-MOV®II Varistor, will be described in greater detail in Chapter 3. For the
moment, the description will be limited to what is necessary for understanding the discussion of
suppression and application in this chapter.
The high exponent values (ex) of the metal-oxide varistors have opened completely new fields
of applications by providing a sufficiently low protective level and a low standby current. The
opportunities for applications extend from low-power electronics to the largest utility-type surge
arresters. However, for this manual, the emphasis is primarily on circuit voltages below 1000V rms and
surge energies of less than 1OOOJ. Higher voltages can be readily obtained by connecting several devices
35
in series. However, increased current values by parallel connection cannot be obtainedwithoutcareful
matching of device characteristics. On the low end of the voltage scale, the metal-oxide varistors, as
they exist today, are limited to 10 Vrms.
The structural charaCteristics of metal-oxide varistors unavoidably result In an appreciable
capacitance between the device terminals, depending on area, thickness and material processing. For
the majority of power applications, this capacitance is not significant. In high-frequency applications,
however, the effect must be taken into consideration in the overall system design.
2.2 TRANSIENT SUPPRESSORS COMPARED
Because of diversity of characteristics and nonstandardized manufacturer specifications, transient
suppressors are not easy to compare. A graph (Figure 2.3) shows the relative volt-ampere
characteristics of the four common devices that are used in 120V ac circuits. A curve for a simple ohmic
resistor is included for comparison. It can be seen that as the alpha factor increases, the curve's
voltage-current slope becomes less steep and approaches an almost constant voltage. High alphas are
desirable for clamping applications that require operation o~er a wide range of currents.
It also is necessary to know the device energy-absorption and peak-current capabilities when
comparisons are made. The table below includes other important parameters of commonly used
suppressors.
JIll
I
(/)
~
~
I
t\
SILlC~~~~BIDE VARISTOR
1000
~ 800
.........
/
25)
SILICON POWER
TRANSIENT SUPPRESSOR
(/)
(Zrn (itlr
z
2
3
'I I
1
4 5
8 10
20
30 40 50
INSTANTANEOUS CURRENT - AMPS
80 100
FIGURE 2.3: V-I CHARACTERISTICS OF FOUR TRANSIENT SUPPRESSOR DEVICES
TABLE 2.1: SUPPRESSOR CHARACteRISTICS
STANDBY
CURRENT
rnA
PEAK
CURRENT
@ 1 ms
A
PEAK
POWER
@ 1 ms
kW
PEAK
ENERGY
JOULES
VOLTAGE
CLAMPING
RATIO
@10A
VOLTAGE
RANGE
DC
1
120
40
70
1.7
14-1200
.005
5.5
1.5
2
1.65
5-200
Selenium
(1'" Sq.)
12
30
9
9
2.3
35-700
Spark Gap
-
>500
-
-
2.4-8.8*
90-1400
-
-
50
4.6
15-300
SUPPRESSOR
TYPE
GE-MOV®II Varistor
(20mm diameter)
Zener Diode (1 W)
Silicon Carbide Varistor
(0.75" diameter)
,
5
*Range of impulse spark-over voltage at 1 kV!Jl.s, then drops to <1.
36
Standby power -the power consumed by the suppressor unit at nonnal line voltage-is an
important selection criterion. Peak standby current is one factor that detennines the standby power of a
suppressor. The standby power dissipation depends also on the alpha characteristic of the device.
As an example, the selenium suppressor in Table 2.1 has a 12mA peak standby current and an
alpha of8 (Figure 2.3). Therefore, it has a standby power dissipation of about 0.5W on a 120V nns line
(l70V peak). A zener-diode suppressor has standby power dissipation ofless than a milliwatt. And a
silicon-carbide varistor, in a 0.75" diameter disc, has standby power in the 200mW range. High standby
power in the lower alpha devices is necessary to achieve a reasonable clamping voltage at higher
currents.
50
ZENER
I-
Z 20
10
z
5
/ v -~
~
Q
~
iii
III
2
is
II::
w
I
~
l( 0.5
>!Xl
o
~ 0.2
a= 35.........
"-.""®v.. "'''~ V
::>
:!j
~IODE OR ICLUSTE+-\
.
I
I
--:.~
~
~
~~
- --
I"SELENIUM}
SILICON
~
a=1La=4
.'\
CARBIDE~
/
IIII
0.1
96
98
100
102
104
106
PERCENT OF RATED VOLTAGE
108
110
FIGURE 2.4: CHANGES IN STANDBY POWER ARE CONSIDERABLY GREATER
WHEN THE SUPPRESSOR'S ALPHA IS HIGH
The amount of standby power that a circuit can tolerate may be the deciding factor in the choice of a
suppressor. Though high-alpha devices have low standby power at the nominal design voltage, a small
line-voltage rise would cause a dramatic increase in the standby power. Figure 2.4 shows that for a
zener-diode suppressor, a 10% increase above rated voltage increases the standby power dissipation
above its rating by a factor of 30. But for a low-alpha device, such as silicon carbide, the standby power
increases by only 1.5 times.
Typical volt-time curves of a gas discharge device are shown in Figure 2.5 indicating an initial
high clamping voltage. The gas-discharge suppressor does not tum on unless the transient pulse exceeds
the impulse sparkover voltage. Two representative surge rates-lkVIILS and 20kVIlLs-are shown in
Figure 2.5. When a surge voltage is applied, the device turns on at some point within the indicated
limits. At 20kVIILS, the discharge unit will sparkover between 600 and 2500V. At lkVIILS, it will
sparkover between 390 and 1500V.
In use for the protection of ac line surges, the gas discharge device may experience follow current.
As the voltage passes through zero at the end of every half cycle the arc will extinguish, but if the
electrodes are hot and the gas is ionized, it may re-ignite onthe·next cycle. Depending on the power
source, this current may be sufficient to cause damage to the electrodes. The follow current can be
reduced by placing a limiting resistor in series with the device, reducing its current, but at a penalty of
increased clamping voltage.
The gas discharge device is useful for high current surges but it is not effective in protecting low
voltage, low impedance circuits. It is often advantageous to provide another suppression device in a
combination that allows the added suppressor to protect against the high in!tial impulse. Several hybrid
combinations with a varistor or avalanche diode are possible. Care in design is required to direct the
initial portion of the impulse to the solid state device and to divert the high current of the later portion of
the pulse to the gas discharge element. Precautions must also be taken against voltages induced in
adjacent wiring by the sharp current rise associated with the gap sparkover.
37
'6000
r---.-----.-~-_,__-__r__r_--_,_--__,
5000r---~--r-~-+--T-+---~--~
4000~-~~--~~-+--.~+---~--~
3000~--~--~--+--r-+---~---~
2000~--~--~~~+-~-+---~--~
>"
I
~ 1000~~-~-~~~~~~~A-+---~--~
~
800~-~b-~~~~~+--~,~,---~--~
g
600r---~~~~~~+---~"~;-~--~
c
500~--~T-~~~~+---+-~!4~~--~
:J
400 r-----j
IIJ
0..
~
300 f----,,:t.f----t~
200
10- 8
10-7
10- 6
10-5
SHORT-TIME SURGE RESPONSE - S
FIGURE 2.5: IMPULSE BREAKOVER OF A GAS·DISCHARGE DEVICE DEPENDS UPON THE RATE OF VOLTAGE RISE
AS WELL AS THE ABSOLUTE VOLTAGE LEVEL
2.3 PROOF TESTS
To consider protective devices while a system is being designed and to follow with proof tests
should be axiomatic, but historical evidence makes it apparent that this is not so. Thus, retrofit of
transient suppressors is common practice. Actually, one can view this retrofit as part of the tradeoff
process, with iterative corrections in a calculated risk approach. It may be justifiable to attempt applying
some device with minimal protection in the harsh outside world, and later, when found necessary, to
take corrective action. Hence, retrofit shouid then be the result of informed choice, not an unforeseen
need for correctIon. Even here some form ofproo(testing will be required to ascertain that the retrofit
will do the job.
The nature of the transient environment comes under examination when a retrofit must be applied.
There is some factual knowledge on the subject, discussed in Chapter 1, but there are also many
tentative' 'generalizations" that require confirmation. Test standards and specifications, then, become
useful guides to the extent that they a,re not applied or enforced blindly. 8
Some test specifications emphasize voltage tests. This is natural because historically, electrical
equipment had dielectric failures as the major consequence of overvoltages. One can, therefore, specify
some voltage test wave that the equipment must withstand without breakdown. However, with the
inclusion of a protective device within the boundaries of an eletronic black box, a simple voltage test is
no longer meaningful. What is needed is the two-step approach discussed below, where the voltage
allowed by the protector is determined first, then the effects on the downstream components
determined. Coordination is the key concept.
One point to remember when specifying or performing a test is the difference between a voltage
and a current test.
In testing a device for voltage withstand capability, proper recognition must be giv:en to the
impedance of the device being tested and to whether or not it already contains a transient suppressor. It
would seem obvious that one does not specify a voltage test and then crank up the generator until that
specified steady-state voltage is achieved at the terminals of the black box containing a, suppressor. Yet
this has been attempted!
Conversely, applying a voltage test to a black box containing a suppressor will be meaningless if
the source impedance of the test generator is too high or too low. It is more appropriate, at least in the
design stage, to separate the test from the design steps. First, one specifies the test circuit which will
38
exercise the suppressor: open circuit voltage, (amplitude and duration) and source impedance. This
determines the clamp voltage that will be developed across the shunt-connected suppressor (Figure
2.6). Second, one designs the protected circuit for this clamp voltage allowing adequate margins. After
the design, this two-step approach can also be applied for demonstrating that withstand capability has
been achieved.
STEP I
OPEN CIRCUIT VOLTAGE
SUPPRESSOR
V ACROSS
PROTECTED
LOAD
FIGURE 2.6: TWO STEPS FOR EVALUATING PROTECTION REQUIREMENTS
Chapter 7 provides detailed information on varistor testing, both for evaluating varistor
characteristics and for conducting realistic proof tests.
2.4 UPDATE ON NEW DEVICES: LOW VOLTAGE AND HIGH ENERGY SERIES·
2.4.1 The New Low Voltage Varistor
New developments and breakthroughs in power techn()logy have made it possible to produce
varistors with voltage ratings as low as 5V dc.
-v
o
+v
FIGURE 2.7 V-I CHARACTERISTICS OF THE V8ZA2 VARIST()R.
Figure 2.7 is an illustration of the dynamic characteristics of the V8ZA2. It is observed that the
device has excellent symmetry of the clamping voltage, and features high alpha and high peak current
ratings.
39
60
Illill
III
55
1'~8Z~ I 1111111
2
50
1
iil 45
40
I I
35
L
f-
..J
0
2:
w
C!1
30
V
C!1
z
Q:
25
2
~ 400
====
u
'-
:i!
~ 200
;(
«
:i!
w
t!l
~
~
0
~-- --
25V
20V
~~
u
,~
I
~
~
~ 15V
~
=""
...
~
~
~ 10V
w
"-
..
::;:
~
X
I,
«
::;:
0.1ms/div
FIGURE 2.10(a) ESO OPEN CIRCUIT VOLTAGE
5V
0
5/ts/div
FIGURE 2.10(b} ESO CLAMPING VOLTAGE
If additional information on the subject of electrostatic discharge, failure threshold levels of different Ie families and the use of low voltage varistors is required, Application Note 200.91 should be
used for reference. It is available from the Semiconductor Products Department upon request. Ratings
and characters for the new low voltage types can be found on Page 138.
2.4.2 High Energy - Large Diameter Devices
The new Series Band D GE-MOV®II devices provide higher energy and current ratings to the
existing product line. They negate the need for paralleling lower current units and matching individual
electrical characteristics to promote current sharing.
Devices are available with voltage ratings from 130V RMS to 750V RMS for the 40mm D Series and
420V RMS to 2800V RMS for the 60mm B Series. Individual devices have energy ratings as high as
10,000 joules. Typical applications for the devices would be to protect large motors and generators, earth
moving equipment, diesel electro locomotives, oil well drilling and pumping rigs, power distribution,
and high energy switching. The devices are available in a new package which enables parallel or series
operation when required. See page 141 for specific details.
41
REFERENCES
1. Sakshaug, E.C., J.S. Kresge and S.A. Miske, "A New Concept in Station Arrester Design," IEEE
Trans. PAS-96,No. 2, March-April 1977, pp. 647'-656.
2. Philipp, H.R. and L.M. Levinson, "Low Temperature Electrical Studies in Metal Oxide
Varistors-A Clue to Conduction Mechanisms," Journal of Applied Physics, Vol. 48, April 1977 ,
pp. 1621-1627.
3. Philipp, H.R. and L.M. Levinson, "Zinc Oxide for Transient Suppression," IEEE Trans. PHP,
December 1977.
4. Surge Arresters for Alternating Current Power Circuits, ANSI Standard C62.1, IEEE Standard 28.
5. Lightning Arresters. Part I: Nonlinear Resistor Type Arresters for AC Systems, IEC
Recommendation 99-1, 1970.
6. Matsuoka, M., T. Masuyama and Y. !ida, Supplementary Journal of Japanese Society of Applied
Physics, Vol. 39, 1970, pp. 94-101.
7. Hamden, J.D., F.D. Martzloff, W.G. Morris and F.B. Golden, "Metal-Oxide Varistor: A New
Way to Suppress Transients," Electronics, October 2, 1972.
8. Martzloff, F.D., "The Development of a Guide on Surge Voltages in Low-Voltage AC Power
Circuits," Report 81CRD047, General Electric, Schenectady, New York, 1981.
42
@k®~~
GE-MOV®II VARISTORS
BASIC P'ROPERTIES,
TERMINOLOGY AND THEORY
3.1
WHAT IS A GE·MOV®II VARISTOR?
GE-MOV® II Varistors are voltage dependent, nonlinear devices which have an electrical behavior
similar to back-to-back zener diodes. The symmetrical, sharp breakdown characteristics shown in
Figure 3.1 enable the varistor to provide excellent transient suppression performance. ~hen exposed to
high voltage transients the varistor impedance changes many orders of magnitude from a near open
circuit to a highly conductive level, thus clamping the transient voltage to a safe level. The potentially
destructive energy of the incoming transient pulse is absorbed by the varistor, thereby protecting
vulnerable circuit components.
The varistor is composed primarily of zinc oxide with small additions of bismuth, cobalt,
manganese and other metal oxides. The structur€? of the body consists of a matrix of conductive zinc
oxide grains separated by grain boundaries providing P-N junction semiconductor characteristics.
These boundaries are responsible for blocking conduction at low voltages and are the source of the
nonlinear electrical conduction at higher voltages.
Since electrical conduction occurs, in effect, between zinc oxide grains distributed throughout the
bulk of the device, the GE-MOV®II Varistor is inherently more rugged than its single P-N junction
counterparts, such as zener diodes. In the varistor, energy is absorbed uniformly throughout the body of
the device with the resultant heating spread evenly through its volume. Electrical properties are
controlled mainly by the physical dimensions of the varistor body which generally is sintered in the
shape of a disc. The energy rating is determined by volume, voltage rating by thickness or current flow
path length, and current capability by area measured normal to the direction of current flow.
GE-MOV®II Varistors are available with ac operating voltages from 4V to 2800V. Higher voltages are limited only by packaging ability. Peak current handling exceeds 50,000A and energy
capability extends beyond 6500J for the larger units. Package styles include the axial device series for
automatic insertion and progress in size up to the rugged high energy device line.
FIGURE 3.1: TYPICAL VARISTOR I·V CHARACTERISTIC.
43
3.2 PHYSICAL PROPERTIES
3.2.1 Introduction
An attractive property of the metal oxide varistor, fabricated from zinc oxide (ZnO), is that the
electrical characteristics are related to the bulk of the device. Each ZnO grain of the ceramic acts as if it
has a semiconductor junction at the grain boundry. A cross-section of the material is shown in Figure
3.2, which illustrates the ceramic microstructure. The ZnO grain boundaries can be clearly observed.
Since the nonlinear electrical behavior occurs at the boundary of each semiconducting ZnO grain, the
varistor can be considered a "multi-junction" device composed of many series and parallel connections
of grain boundaries. Device behavior may be analyzed with respect to the details of the ceramic
microstructue. Mean grain size and grain size distribution playa major role in electrical behavior.
FIGURE 3.2: OPTICAL PHOTOMICROGRAPH OF A POLISHED AND ETCHED SECTION OF A GE·MOV®VARISTOR
3.2.2 Varistor Microstructure
GE-MOV®JI Varistors are formed by pressing and sintering zinc oxide-based powders into
ceramic discs. These discs are then electroded with thick film silver to provide solderable contact areas.
The bulk of the varistor between contacts is comprised of ZnO grains of an average size" d" as shown in
the schematic model of Figure 3.3. Resistivity of the ZnO is
~IOO
'"
I
I
V-'~I/
•
----fJ)
I
~/--3:
~I= kVQ
jl
0)
~J--
°/--i
°
S2
..!.!
"I
tel
tel
I
i
f
(TYPICAL V130LA20A)
1
10-2
10°
10 2
104
CURRENT - AMPERES
FIGURE 3.10: TYPICAL VARISTOR V-I CURVE PLOTTED ON LOG-LOG SCALE
3.4.2 Equivalent Circuit Model
An electrical model for the varistor can be represented by the simplified equivalent circuit of
Figure 3.11.
(TYPICAL V130LA20A)
c --.L
(0.002fL F)
ROFF
T
(JOOO M,Q)
FIGURE 3.11: VARISTOR EQUIVALENT CIRCUIT MODEL
50
3.4.3 Leakage Region oj Operation
At low current levels, the V-I curve approaches a linear (ohmic) relationship and shows a
significant temperature dependence. The varistor is in a high resistance mode (approaching 109 ohms)
and appears as an open circuit. The nonlinear resistance component, Rx , can be ignored because ROFF in
parallel will predominate. Also, RON will be insignificant compared to RaFF'
c
FIGURE 3.12: EQUIVALENT CIRCUIT AT LOW CURRENTS
For a given varistor device, capacitance remains approximately constant over a wide range of
voltage and frequency in the leakage region. The value of capacitance drops only slightly as voltage is
applied to the varistor. As the voltage approaches the nominal varistor voltage, the capacitance abruptly
decreases. Capacitance remains nearly constant with frequency change up to 100 kHz. Similarly, the
change with temperature is small, the 25°C value of capacitance being well within ± 10% from -40°C
to + 125°C.
The temperature effect of the V-I characteristic curve in the leakage region is shown in Figure
3.13. A distinct temperature dependence is noted.
.
100
~P'"
~..",..:....
80
~ ;,-
. /[...-
,r/V "//
/'f'
/
V
r/ V/
VI II
" I, VVI
I 'II II I /
If
IJ V / II
'/ ~fO{r IiI
125°C
10
-9
SPECIM iN'rV130LArA
10-7
10- 6
10- 5
10- 4
VARISTOR CURRENT (AMPERES, DC)
FIGURE 3.13: TEMPERATURE DEPENDENCE OF THE CHARACTERISTIC CURVE IN THE LEAKAGE REGION
The relation between the leakage current, I, and temperature, T, is:
I
= 10
where:
-VB/kT
€
10 = constant
k = Boltzmann's Constant
VB = 0.9 eV
The temperature variation, in effect, corresponds to a change in RoFF • However, ROFF remains at a
high resistance value even at elevated temperatures. For example, it is still in the range of 10 to 100
mega ohms at 125°C.
51
Although ROFF is a high resistance it varies with frequency. The relationship is approximately linear
with inverse frequency.
1
RaFF - f
However, the parallel combination of ROFF and C is predominantly capacitive at any frequency
interest. This is because the capacitive reactance also varies approximately linearly with 1/f.
~f
At higher currents, at and above the milliamp range, temperature variation becomes minimal. The
plot of the temperature coefficient (dv/dt) is given in Figure 3 .14. It should be noted that the temperature
coefficient is negative and decreases as current rises. In the normal operational range of the varistor (I >
1mA), the temperature dependency is less than -O.05%/oC.
~.--- LEAKAGE REGION
.14
NORMAL
I OPERATION
---~
~
.12
'"
........
i'..
~
-0.05%/oC
1 - - - - - - I-----~ f o - . - - - - - - - - .02
.01
.1
I
CURRENT - MILLIAMPERES
10
FIGURE 3.14: RELATION OF TEMPERATURE COEFFICIENT. DV/DT TO VARISTOR CURRENT
3.4.4 Normal Varistor Region of Operation
The varistor characteristic follows the equation I = kV"', where k is a constant and the exponent ex
defines the degree of nonlinearity. Alpha is a figure of merit and can be determined from the slope of the
V -I curve or calculated from the formula:
log (1 2 /1 1 )
a
= log (V !V )
2
=
l
1
In this region the varistor is conducting and Rx will predominate Qver C, RON and R OFF '
many orders of magnitude less than R OFF ' but remains larger than RON'
Rx
becomes
f"
FIGURE 3.15: EaUIVALENT CIRCUIT AT VARISTOR CONDUCTION
During conduction the varistor voltage remains relatively constant for a change in current of
several orders of magnitude. In effect, the device resistance, R x, is changing in response to current. This
can be observed by examining the static or dynamic resistance as a function of current. The static
resistance is defined by;
V
~ =I
and the dynamic resistance by:
dv
Z = - = V/aI = Rx/a
x
di
.
Plots oftypical:resistance values vs. current, I, are given in Figure 3.16.
52
500
~
'"
5
"" '"
" '"
i'...
~
'"~
~
"i'-.
0.01
0·01
0.1
I
10
PEAK CURRENT - AMPERES
I
'""
"
~
........
"0.001
0.01
100
0.1
1.0
10
PEAK CURRENT - AMPERES
100
FIGURE 3.168: Zx DYNAMIC VARISTOR RESISTANCE
FIGURE 3.16A: Rx STATIC VARISTOR RESISTANCE
3.4.5 Upturn Region o/Operation
At high currents, approaching the maximum rating, the varistor approximates a short-circuit. The
curve departs from the nonlinear relation and approaches the value of the material bulk resistance, about
1-10 ohms. This upturn, or saturated region, takes place as Rx approaches the value of RON' Resistor RaN
represents the bulk resistance of the zinc oxide grains. This resistance is linear (which appears as a
steeper slope on the log plot) and occurs at currents of 50 to 50,000A, depending on the varistor size.
FIGURE 3.17: EQUIVALENT CIRCUIT AT VARISTOR UPTURN
3.4. 6 Speed 0/ Response and Rate Effects .
The varistor action depends on a conduction mechanism similar to that of other semiconductor
devices. For this reason, conduction occurs very rapidly, with no apparent time lag-even into the
nanosecond range. Figure 3.18 shows a composite photograph oftwo voltage traces with and without a
varistor inserted in a very low inductance impulse generator. The second trace (which is n,ot
synchronized with the first, but merely superimposed on the oscilloscope screen) shows that the voltage
clamping effect of the varistor occurs in less than one nanosecond.
In the conventional lead-mounted devices, the inductance ofthe leads would completely mask the
fast action of the varistor; therefore, the test circuit for Figure 3.15 required insertion of a small piece of
varistor material in a coaxial line to demonstrate the intrinsic varistor response.
Tests made on lead-mounted devices, even with careful attention to minimizing lead length, show
that the voltages induced in the loop formed by the leads contribute a substantial part of the voltage
appearing across the terminals of a varistor at high current and fast current rise: Fortunately, the currents
which can be delivered by a transient source are invariably slower in rise time than the observed voltage
transients. The applications most frequently encountered for GE-MOV®II Varistors involve current
rise times longer than 0.5 JLS.
53
TRACE I
LOAD VOLTAGE
WITHOUT VARISTOR'
:>
o
....
Ul
~
TRACE 2
LOAD VOLTAGE
CLAMPED
BY VARISTOR
g
o
Q
500 PICOSECONDS/DIV.
FIGURE 3.18: RESPONSE OF A ZnO VARISTOR TO A FAST RISE TIME (500 PICOSECOND) PULSE
Voltage rate-of-rise is not the best term to use when discussing the response of a varistorto a fast
impulse (unlike spark gaps where a finite time is involved in switching from non-conducting to
conducting state). The response time of the varistor to the transient current that a circuit can deliver is
the appropriate characteristic to consider.
1000
~DELJJJJJ
800 ~(LEAD
II
WAVES~PE-==
AREA
1
lLJ
'"
~
3~.:\
8/20 s
~
l-- I-~
400
...J
g
'"n:z
200
a) V-I CharacteristiCS, For Various
Current Rise Times
::Ii
«
...J
u
10
20
40
60
80 100
200
400
PEAK CURRENT -AMPERES
140
.."-
co
I
lodoAlm2
\
!:i
'"
TAT2
120
0
>
'"n:
z
::Ii
I,
IOA/cm 2
110
...J
0
~
l'i' '"
.....
«
u
"-
100
90
0.2
II
1\
lLJ
~
...J
2000
DEVICE: VI30 LA20A
(LEAD AREAclcm 2 l*
\
130
600 800 1000
"
""
~ ~ :-..t-.
0:4 0.50.6 0.8 I
PULSE RISE
b) Overshoot Defined With Reference To The
Basic 8 /20lolsCurrent Pul~
I~~l
~~i
2,
4
5 6
8
10
TIME-~s
*Refer to Section 7.3.2.
FIGURE 3. 19a and b: RESPONSE OF LEAD-MOUNTED VARISTORS TO CURRENT WAVEFORM
54
The V-I characteristic of Figure 3.19a shows how the response of the varistor is affected by the
current waveform. From such data, an "overshoot" effect can be defined as being the relative increase
in the maximum voltage appearing across the varistor during a fast current rise, using the conyentional8
/ 20l-'s current wave as the reference. Figure 3.19b shows typical clamping voltage variation with rise
. time for various current levels.
3.5 VARISTOR TERMINOLOGY
The following tabulation defines the terminology used in varistor specifications. Existirig
standards have been followed wherever possible.
3.5.1 Definitions (IEEE Standard C62.33, 1982)
A characteristic is an inherent and measurable property of a device. Such a property may be
electrical, mechanical, or thermal, and can be expressed as a value for stated conditions.
A rating is a value which establishes either a limiting capability or a limiting condition (either
maximum or minimum) for operation of a device. It is determined for specified values of environment
and operation. The ratings indicate a level of stress which may be applied to the device without causing
degradation or failure. Varistor symbol~ are defined on the linear V-I graph illustrated in Figure 3.20.
Ip
_____________ _
~
L;I)
enw
a:
w
a..
Ix
~
VARISTOR
SYMBOL
-------------
~
I-
Z
en
1000
o
VN(de)
~
W
a:
a:
a
?!.
IN(de)
W
"
~
o
Ipm
ID~~=-~________~___L~~
>
10
10-6
IN (de)
100
CURRENT (AMPERES)
VOLTAGE (VOLTS)
FIGURE 3.20: I·V GRAPH ILLUSTRATING SYMBOLS AND DEFINITIONS
55
3.5.2 Varistor Characteristics (IEEE Standard C62.33-J982 Subsection 2.3 and 2.4)
Term and Description
2.3.1 Clamping Voltage. Peak voltage across the varistor measured under
conditions of a specified peak pulse current and specified waveform. Note: Peak
voltage and peak currents are n9t necessarily coincidental in time.
2.3.2 Rated Peak Single Pulse Transient Currents (Varistor). Maximum peak
current which may be applied for a single 8/20l/-s impulse, with rated line voltage also
applied, without causing device failure.
2.3.3 Lifetime Rated Pulse Currents (Varistor). Derated values of I,m for impulse
durations exceeding that of an 8/20l/-s waveshape, and for multiple pulses which may be
applied over device rated lifetime.
2.3.4 Rated RMS Voltage (Varistor). Maximum continuous sinusoidal rms voltage
which may be applied.
2.3.5 Rated DC Voltage (Varistor). Maximum continuous dc voltage which may be
applied.
2.3.6 DC Standby Current (Varistor). Varistor current measured at rated voltage,
Vm(dC)'
Symbol
Vm(ac)
Vm(dC)
2.4 For certain applications, some of the following terms may be useful. .
2.4.1 Nominal Varistor Voltage. Voltage across the varistor measured at a specified
pulsed dc current, IN(dc) , of specific duration. IN(dc) is specified by the varistor
manufacturer.
2;4.2 Peak Nominal Varistor Voltage. Voltage across the varistor measured at a
specified peak ac current, IN(ac), of specific duration. IN(ac) is specified by the varistor
manufacturer.
2.4.3 Rated Recurrent Peale Voltage (Varistor). Maximum recurrent peak voltage
which may be applied for a specified duty cycle and waveform.
2.4.4 Rated Single Pulse Transient Energy (Varistor). Energy which may be
dissipated for a single impulse of maximum rated current at a specified waveshape, with
rated rms voltage or rated dc voltage also applied, without causing device failure.
2.4.5 Rated Transient Average Power Dissipation (Varistor). Maximum average
power which may be dissipated due to a group of pulses occurring within a specified
isolated time period, without causing device failure.
2.4.6 Varistor Voltage. Voltage across the varistor measured at a given current, Ix.
2.4.7 Voltage Clamping Ratio (Varistor). A figure of merit measureofthe varistor
clamping effectiveness as defined by the symbols VJV m(ac) , VJV m(dc)'
2.4.8 Nonlinear Exponent. A measure of varistor nonlinearity between two given
operating currents, II and 12 , as described by I = kVa where k is a device constant, II :S I
:S 12 , and
a
12 -
V N(ac)
Vprn
P'(AV)m
log 12/11
log V/V 2
2.4.9 Dynamic Impedance (Varistor). A measure of small ',signal impedance at a
given operating point as defined by:
Z = dV,
,
dl,
2.4.10 Resistance (Varistor). Static resistance of the varistor at a given operating
point as defined by:
R = dV x
,
VN(dc)
dl,
56
z,
Term and Description
Symbol
2.4.11 Capacitance (Varistor). Capacitance between the two terminals of the
varistor measured at specified frequency and bias.
2.4.12 AC Standby Power (Varistor). Varistor ac power dissipation measured at
rated rms voltage V m(.e)'
2.4.13 Voltage Overshoot (Varistor). The excess voltage above the clamping
voltage of the device for a given current that occurs when current waves ofless than 8JLs
virtual front duration are applied. This value may be expressed as a % of the clamping
voltage (Ve) for an 8120 current wave.
2.4.14 Response Time (Varistor). The time between the point at which the wave
exceeds the clamping voltage level (VJ and the peak of the voltage overshoot. For the
purpose of this definition, clamping voltage is defined with an 8120JLs current
waveform of the same peak current amplitude as the waveform used for this
response time.
2.4.15 Overshoot Duration (Varistor). The time between the point at which the
wave exceeds the clamping voltage level (VJ and the point at which the voltage
overshoot has decayed to 50 % of its peak. For the purpose of this definition, clamping
voltage is defined with an 8/20JLs current waveform of the same peak current amplitude
as the waveform used for this overshoot duration.
C
3.5.3 Test Waveform
At high current and energy levels, varistor characteristics are measured, of necessity, with an
impulse waveform. Shown in Figure 3.21 is the ANSI Standard C62.1 waveshape, an exponentially
decaying waveform representative of lightning surges and the discharge of stored energy in reactive
circuits.
The 8 / 20ILS current wave (81Ls rise and 20ILs to 50% decay of peak value) is used as a standard,
based on industry practices, for the characteristics and ratings described. One exception is the energy
rating (W,m), where a longer waveform of 10 / lOOOILS is used. This condition is more representative of
the high energy surges usually experienced from inductive discharge of motors and transformers.
GE-MOV@II Varistors are rated for a maximum pulse energy surge that results in a varistor voltage (V N)
shift of less than ± 10% from initial value.
100
;;e
90
--------
W
::J
-'
~
«
'"
w
(L
"-
o
f--
z
50
w
u
a:
w
I
I
I
----~------------I
I
I
I
(L
I
~
I
f--
I
Z
W
a:
a:
I
I
u
VIRTUAL START OF WAVE
::J
10
I
I
I
IMPULSE DURATION
.1
I--VIRTUAL FRONT DURATION
= 1.25 X RISETIME FROM 10% TO 90%
FIGURE 3.21: DEFINITION OF PULSE CURRENT WAVEFORM
57
REFERENCES
1. Harnden, J.D., F.D. Martzloff, W.G. Morris and F.B. Golden, "The GE-MOV® Varistor-The
Super Alpha Varistor," Electronics, Vol. 45, No. 21, 1972, p. 91.
2. Morris, W. G ., "Electrical Properties of ZnO-Bi20 3 Ceramics, " Journal of the Am. Ceram. Soc.,
Vol. 56, 1973.
3. Matsuoka, M., "Non-Ohmic Properties of Zinc Oxide Ceramics," Japanese Jnl. Appl. Phys.,
Vol. to, 1971, p. 736.
4. Mahan, G., L. Levinson. and H. Philipp, "Single Grain Junction Studies. at ZnO VaristorsTheory & Experiment," Report #78CRF160, General Electric, Schenectady, N.Y., 1978.
Forthcoming in Applied Physics Lttt~rs.
5. Mahan, G., L. Levinson and H. Philipp, "Theory of Conduction in ZnO Varistors," Journal of
Applied Physics (in press).
6. Levine, J. D. , "Theory of Varistor Electronic Properties, " Critical Review of Solid State Science,
5, 1975, pp. 597-608.
7. May, J. E., "Carrier Concentration and Depletion Layer Model of Zinc Oxide Varistors, " Bulletin
ofthe American Ceramic Society, Vol. 57, No.3, 1978, p.335.
8. Sze, S.M., "Physics of Semiconductor Devices," John Wiley & Sons, New York, N.Y., 1969.
9. Einzinger, R., "Microcontact Measurement of ZnO Varistors," Ber. Dt. Keram, Vol. 52, 1975,
pp.244-245.
,
to. Lou, L.F., "Current~Vo1tage Characteristics orZnO-Bi 20 3 Heterojunction, " Journal of Applied
Physics, Vol. 50, 1979, p. 555.
11. Lou, L.F., "Semiconducting Properties of ZnO-Grain Boundary-ZnO Junctions in Ceramic.
Varistors," Appl. Phys. Letters, Vol. 36, 1980, pp. 570-572.
12. Lou, L.F. and J.E.,May, Unpublished Research, General Electric, Syracuse, N.Y., 1981.
13. MQrris, W., "Physical Properti~s of the Electrical Barriers in Varistors," J. Vac. Sci. Technol.,
13, 1976, pp. 926-931.
14. Bernasconi, J., S. Strassler, B. Knecht, H. Klein and A. Menth, Solid State Communication, Vol.
21,1977, pp. 867-869.
15. Pike, G. and C. Seager, "The DC Voltage Dependence of Semiconductor Grain-Boundary
Resistance," Journal of Appl. Phys., Vol. 50, 1979, pp. 3414-3422.
58
DESIGNING WITH GE-MOV®II VARISTORS
4.1
SELECTING THE VARISTOR
The varistor must operate under steady-state and transient conditions. Device ratings allow a
selection of the proper size device to insure reliable operation. The selection process requires a
knowledge of the electrical environment. When the environment is not fully defined, some
approximations can be made.
For most applications, selection is a five-step process:
I)
Determine the necessary steady-state voltage rating (working voltage)
Establish the transient energy absorbed by the varistor
Calculate the peak transient current through the varistor
Determine power dissipation requirements
Select a model to provide the required voltage-clamping characteristic
2)
3)
4)
5)
Refer also to page 135 "How to select a GE-MOV®JI Varistor."
4.1.1 Steady-State Voltage Rating
Consider the maximum steady-state voltage applied to the varistor including any high line conditions (i.e., 110% or more of nominal voltage). Ratings are given for sinusoidal ac and constant dc.
If a nonsinusoidal waveform is applied, the recurrent peak voltage should be limited to"';2 x V m(ac).
Specifications for the L Series varistor are shown in Figure 4.1 for 130 V ac rated devices to
illustrate the use of the ratings and characteristics table.
MAXIMUM RATINGS (25°C)
MOOEL
NUMBER
V130LA2
V130LA10A
V130LA20A
V130LA20B
CHARACTERISTICS
TRANSIENT
RMS
VOLTAGE
OC
VOLTAGE
Vm(ac)
Vm(dc)
Wtm
Itm
MIN.
VN(dc)
MAX.
MAX. CLAMPING
VOLTAGE
TYPICAL
V, @ TEST
CAPACI·
MOOEL
CURRENT
TANCE
SIZE
(8/2D/Ls)
(mm)
Ip
Vc
f = 0.1·1 MHz
VOLTS
VOLTS
JOULES
AMPERES
VOLTS
VOLTS
VOLTS
VOLTS
130
175
11
1200
4500
6500
6500
184
200
228
228
228
220
340
340
340
325
CONTINUOUS
ENERGY
(10 11 DDD/Ls)
38
70
70
VARISTOR
VOLTAGE
@ lmA OC
TEST
CURRENT
PEAK
CURRENT
(8 I 2D/Ls)
FIGURE 4.1: RATINGS AND CHARACTERISTICS TABLE
59
AMPS
10
50
100
100
PICOFARAOS
180
1000
1900
1900
7
14
20
20
Indicates an ac voltage rating,of l30V ac RMS.
These models can be operated continuously with up to l30V ac RMS
at 50-60 Hz applied. They would be suitable for l17V ac nominal line
operation and would allow for a 110% high line condition.
Operation is allowable with up to 175V dc constant voltage applied
con tinuously'.
Indicates the minimum varistor terminal voltage that will be measured
with 1 rnA dc applied. This is a characteristic of the device and is a
useful parameter for design or for incoming inspection.
Model Number
V m1ac)
Vm1dC)
V N1dC) Min.
VN'dC)Max. @ 1mA de -
Indicates the maximum limit of varistor terminal voltage measured at
ImAdc.
The format for the model number designation is shown below. Note, the model series are
grouped in two forms.· One group is based on the ac voltage rating for applications primarily across
the power line. The second group is based on VN(dc) characteristic voltage where dc applications are of
more interest.
V
G.ENERAL
ELECTRIC
GE·MOV®
VARISTOR
130
RMS
APPLIED
VOLTAGE
B,D,HE
PA
LA
LA
PRODUCT
SERIES
A
RELATIVE
ENERGY
INDICATOR
SELECTION CLAMPING
VOLTAGE (A OR B)
HIGH ENERGY
POWER VARISTOR
RADIAL LEAD
al Models intended primarily for ac power line applications.
V
GENERAL
ELECTRIC
GE-MOV®
VARISTOR
V.,,,,,,
NOMINAL
VARISTOR
VOLTAGE
MA
ZA
MA
PRODUCT
SERIES
4
B
RELATIVE
ENERGY
INDICATOR
SELECTION CLAMPING
VOLTAGE (A OR B)
MOLDED AXIAL
LOW VOLTAGE
bl Models intended primaiily for dc and circuit applications.
FIGURE 4.2: MODEL NUMBER NOMENCLATURE
60
4.1.2 Energy
Transient energy ratings are given in the W tm column of the specifications injoules (watt-second).
The rating is the maximum allowable energy for a single impulse of 10 / 1000J-ts current waveform with
continuous voltage applied. Energy ratings are based on a shift ofV N of less than ± 10% of intial value.
When the transient is generated from the discharge of an inductance (Le., motor, transformer')
or a capacitor, the energy content can be calculated readily. In many cases the transient is from a
source external to the equipment and is of unknown magnitude. For this situation an approximation technique can be used to estimate the energy of the transient absorbed by the varistor. The
method requires a measurement of the transient current and voltage at the varistor. To determine
the energy absorbed the following equation applies:
E
= [
Vc(t)l(t)Llt
=
KVclr
where I is the peak current applied, V c is the clamp voltage which results, r is the impulse duration and K
is a constant. K values are given in Figure 4.3 for a variety of wave shapes frequently encountered. The
K value and pulse width should correspond to either a current waveform or a voltage waveform. For
complex waveforms, this approach also can be used by dividing the shape into segments that can be
treated separately.
WAVESHAPE
EQUATION
r [] v:
VPK sin
K*
WAVESHAPE
*
1-·
K
(.!: t)
5VPK
0.15
I..T-I
7
lQ:
!--T-!
I
kj:
T---!
kd=
T-----oI
IpK sin
(!!: t)
EQUATION
~
7
VpK e-t / 1.47
0_056
IpK sin (1Tt)e- tlr
0.86
'
. •5I pK
l"
0.637
K*
t
-
- I pK
t
VPK (-)
'
IpK (~)
'
~.
0.038
7
T-I
Q=,
0.5
T---!
7
*Based upon alpha of 25 to 40.
FIGURE 4_3: ENERGY FORM FACTOR CONSTANTS
61
IpK e-t/ 1.447
1.4
IpK
1.0
'
Consider the condition where the exponential waveform shown below is applied to a V130LA1
GE-MOV®II Varistor.
The waveform is divided into two parts that are treated separately using the factors of Figure
4.3: current waveform section (1) 0 to 5/-ts to infinity. The maximum voltage across the V130LA1 at
100A is found to be 500V from the V-I characteristics of the specification sheet.
Section (1)
Section (2)
E
E
= KVc
= KVc
IT
IT
= (0.5)
= (1.4)
(500) (100) (5)(10- 6)
(500) (100) (50 - 5) (10. 6 )
=
=
0.13J
3.15J
3.28J Total
The specifications of Figure 4.1 indicate a rating of 4J for this device which is just adequate for
the application. For more safety factor, a V130LA2 rated at 11J would be a better choice.
4_1.3 Peak Current
The peak current rating can be checked against the transient current measured in the circuit. If the
transient is generated by an inductor, the peak current will not be more than the inductor current at the
time of switching. Another method for finding the transient current is to use a graphical analysis. When
the transient voltage and source impedance is known, a Thevenin equivalent circuit can be modeled.
Then, a load line can be drawn on the log - log, V-I characteristic as shown in Figure 4.4. The two
curves intersect at the peak current value.
w
'"~
Voc
o
>
0:
~
;;::
(f)
:; Vc .fl::.!l!!!~Y.9!J~G~ - - VARISTOR V-I
CHARACTERISTIC
'"o-'
IV
Voc 1Zs
LOG VARISTOR CURRENT - AMPERES
2) Graphical Analysis to Determine Peak 1
1) Equivalent Circuit
FIGURE 4.4: DETERMINING VARISTOR PEAK CURRENT FROM A VOLTAGE SOURCE TRANSIENT
The rated single pulse current, I,m' is the maximum allowable for a single pulse of 8 { 20/-ts
exponential waveform (illustrated in Figure 3.21). For longer duration pulses, I,m should be derated to
the curves in the varistor specifications. Figure 4.5 shows the derating curves for 7mm size, L series
devices: This curve also provides a guide for derating current as required with repetitive pulsing. The
designer must consider the total number of transientpulses expected during the life of the equipment and
select the appropriate curve.
62
1000 .....::-----r--,---,--M-O-O....
EL--:S"'
IZ:::OE"=7m-m----;,----,---,
III
~
SOO
"-
~~~
V2S0LA4
~
100
I-
SO
...a:Z
.,....:---""k::::-~\::::..--+--+- ~~~t~: --+--1
a:
i:l
10
...~
S
:::I
"-
It
C
~
....
Q
O.sl---+--t---+--t---i-----ir---+--i
i
0.11.....----'---'---'---.1...--1.....----''----'---'
20
SOO
1000
2000
50
100
200
SOOO 10,000
IMPULSE DURATION -III
FIGURE 4.5: PULSE RATINGS
Where the current waveshape is different from the exponential waveform of Figure 3.18, the
curves of Figure 4.5 can be used by converting the pulse duration on the basis of equivalent energy.
This is easily done using the constants given in Figure 4.3. For example, suppose the actual current
measured has a triangular waveform with a peak current of lOA, a peak voltage of 340V and an
impulse duration of 500tls.
Then: E = (.5)(10)(340)(500)(10- 6 )
= 850mJ
The equivalent exponential waveform of equal energy is then found from:
= EEXP
= 1.4 Vc I T EXP
ErRIANGULAR
850 mJ
The exponential waveform is taken to have equal Vc and I values. Then,
850mJ
1.4 (340) (10)
=
Or:
T EXP
=
179 tlS
K*
T*
1.4
Where: K* and T* are the values for the triangular waveform and
for the equivalent exponential waveform.
T EXP
is the impulse duration
The pulse rise portion of the waveform can be ignored when the impulse duration is five times
or more longer. The pulse life for the above example would exceed 104 pulses from the pulse life
curves shown in Figure 4.5.
4.1.4 Power Dissipation Requirements
Transients generate heat in a suppressor too quickly to be transferred during the pulse interval.
Power dissipation capability is not necessary for a suppressor, unless transients will be occurring in
rapid succession. Under this condition, the power dissipation required is simply the energy (wattseconds) per pulse times the number of pulses per second. The power so developed must be within
the specifications shown on the ratings tables. It is to be noted that varistors can only dissipate a
relatively small amount of average power and are, therefore, not suitable for repetitive applications
that involve substantial amounts of average power dissipation. Futhermore, the operating values
need to be derated at high temperatures as shown in Figure 4.6.
63
>- 100
CJ
w
Z
w
a:
"
90
f\'" ~
80
ri
w
~
70
Q
60
...
\
\ ~
B, 0 SERI~S i - - -
0
~
50
...0
~
\ \"-
w
~
a:
ZSERIES
__ L SERIES
P SERIES
E $ERIES
40
...Z
\\
~
\
\
\\
30
w
()
a:
20
...w
MA~ERIES
'\
10
o
-55
I'"
['\..
"
1,\
"
50
60
70
80
90
100
110
AMBIENT TEMPERATURE -
120
0
130
C
140
'"
150
FIGURE 4.6: CURRENT, POWER, ENERGY RATING VS. TEMPERATURE
4.1.5 Voltage Clamping Selection
Transient Y":l characteristics are provided in the specifications for all models of GE-MOV®II
Varistors. Shown below in Figure 4.7 are curves for 130V ac rated models of the L series. These curves
indicate the peak terminal voltage measured with an applied 8 / 20lls impulse current. For example, if
the peak impulse current applied to a V 130LA2 is lOA, that model will limit the transient voltage to no
higher than 340Y.
If the transient current is unknown, the graphical method of Figure 4.4 can be utilized. From
a knowledge of the transient voltage and source impedance a load line is plotted on the V-I
characteristic. The intersection of the load line with the varistor model curve gives the varistor
transient current and the value of clamped peak transient voltage.
6000
5000 -
T. - -40 TO 85°C
I I
IL SERIESI
4000
3000 -
M.\XIMUM CLAMPING VOLTAGE,
95 a 130 VAC RATING
MODEL SIZE 7, 14, a 20mm
I
",2000
~
1500
>
V130LA2~
~ 1000
~
800
:::E
::>
:::E
600
:::E
300
~ 400
VI30LAIOA
VI30LA20A . ~
VI30LA20B
-
~.
J.::::--
~
::..-:: V
./
----r--!l
200
10~_2
.--
2
I II i
5 10-1 2
5 100 2
5
101 2
5 102 2
5 103 2
5 104
. PEAK AMPERES
FIGURE 4.7: TRANSIENT V-I CHARACTERISTICS OF TYPICAL L SERIES MODELS
The ability of the varistor to limit the transient voltage is sometimes expressed in terms of a clamp
ratio. For example, consider a varistor applied to protecting the power terminals of electrical
equipment. If high line conditions will allow a rise to 130V ac, then 184Y peak would be applied. The
device selected would require a voltage rating of 130V ac or higher. Assume selection of a Vl30LA2
model varistor. The V 130LA2 will limit transient voltages to 340V at currents of lOA. The clamp ratio
is calculated to be,
64
Clamp Ratio
=
Vc @ lOA
Peak Voltage Applied
340V
= 184V
=
1.85
The clamp ratio can be found for other currents, of course, by reference to the V-I characteristic. In general, clamping ability will be better as the varistor physical size and energy level
increases. This is illustrated in Figure 4.8 which compares the clamping performance of the differellt GE-MOV®II Varistor families. It can be seen that the lowest clamping voltages are obtained from'
the 20mm (L series) and 32mm (HE series) products. In addition, many varistor models are available with two or three clamping selections, designated by an A, B, or C at the end of the model
number. The A selection is the standard model, with B and C selections providing progressively
tighter clamping voltage. For example, the Vl30LA20A voltage clamping limit is 340V at IOOA,.
while the V130LA20B clamps at not more than 325V.
1000
~Do/I
~
'"
600
~
-
-400
300
~
- -----~
I-- ~
...
~" 'V~0/
./
./
V
%$'V ~,'V
~ ~ ....~
NOTE:
CLAMP RATIO EQUALS VARISTOR VOLTAGE DIVIDED
BY VNOM OR 184 VOLTS FOR 130V AC RMS.
Ul
':i0
I I
>
100
.01
.05 0.1
I I
.5 1.0
I I
5
10
50 100
INSTANTANEOUS CURRENT - AMPERES_
I I
500 1000
5000 10.000
FIGURE 4.8 VARISTOR V-I CHARACTERISTICS FOR FOUR PRODUCT FAMILIES RATED AT 130V AC
4.1.6 Summary
The five major considerations for varistor selection have been described. The final choice of
model is a balance of these factors with a device cost trade-off. In some applications a priority requirement such. as clamp voltage or energy capability may be so important as to force the selection
to a particular model. A summary of varistor properties is provided in Figure 4.9 for a quick comparison
of operating ranges.
P.ak
Pul..
Current
Maximum Steady State Applied Voltage
Energy
(Jciul••)
OIIC Size.
Packall"
(Ampa)
40-100
.07-1.7
1004500
.4-35
800-6500
7-360
6500
70-250
20,00040,000
2001050
70.000
150010.000
3mm
P SEAlES
130-660 VRMS
175-850 VOC
HE SERIES. D SERIES
130-750 VRMS
175-970 VDC
r--------
FIGURE 4.9: VARISTOR PRODUCT FAMI-LY SELECTION GUIDE
65
...
32mm
40mm
•
60mm
a
4.2
FAILURE MODES AND VARISTOR PROTECTION
Varistors are inherently rugged and are conservatively rated. Therefore, they exhibit a low
failure rate. Nevertheless, the careful designer may wish to plan for potential failure modes and the
resultant effects on circuitry being protected.
4.2.1 Failure Modes
Varistors initially fail in a short-circuit rn.odewhen subjected to surges beyond their peak
current/energy ratings. They also short-circuit when operated at steady-state voltages well beyond
their voltage ratings. This latter mode of stress may result in the eventual open-circuiting of the
device due to melting of the lead solder joint.
When the device fails in the shorted mode the current through the varistor becomes limit~d
mainly by the source impedance. Consequently, a large amount of energy can be introduced, causing mechanical rupture of the package accompanied by expUlsion of package material in both solid
and gaseous forms. Steps may be taken to minimize this potential hazard by the following techniques: 1) fusing the varistor to limit high fault currents, and, 2) protecting the surrounding circuitry by physical shielding, or by locating the varistor away from other components.
4.2.2 Fusing the Varistor
Varistor fusing should be coordinated to select a fuse that limits current below the level where
varistor package damage could occur. The location of the fuse may be in the distribution line to
the circuit or it may be in series with the varistor as shown in Figure 4.10. Generally, fuse rather
than breaker protection is preferred. Breaker tripping is too slow to prevent excessive fault energy
from being applied.
r-----'
o-----~--~~
I
II
I
I
I
I
I
I
PROTECTED
CIRCUIT
O-------------+---~r_
I
I
iI
I
I
I
I
I
I
I
L _____ J
FIGURE 4.10: FUSE PLACEMENT FOR VARISTOR PROTECTION
In high power industrial circuits the line currents are generally so high as to rule out the use
of a line fuse for varistor protection. The fuse may not clear under a varistor fault condition and
would allow varistor failure. In low power (5-20A) applications it may be feasible to use the line
fuse, FL ' only.
Use of a line fuse, FL, rather than Fy , does not present the problem of having the fuse arc volt-,
age being applied across the circuit. Conversely, with Fv alone, the fuse arc voltage adds to the·
varistor voltage, increasing the Vc , the transient clamp voltage. Since some fuses can have peak arc
voltages in excess of twice peak working voltage, fuse clearing can have a significant effect ,on protection levels.
Another factor in the choice of location is the consequence of system interruption. Fuse location FL will cause a shutdown of the circuit while" location Fv will not. While the circuit can continue to operate when Fy clears, protection no longer is present. For this reason it is desirable to
be able to monitor the condition of Fv .
4.2.3 Fusing Example (Light Industrial Application)
A process control minicomputer is to be protected from transients on a 115V nominal line.
The minicomputer draws 7.5 A from the line, which is guaranteed to be regulated to ± 10% of nomi66
nal line voltage. A Vl30LA20A varistor is chosen on the basis that the worst-case surge current
would be a 10 / 1000l-ts pulse of 100A peak amplitude. The rationale for this surge requirement is
that the incoming plant distribution system is protected with lightning arrestors having a maximum
arrestor voltage of SkY. Assuming a typical son characteristic line impedance, the worst-case transient current through the varistor is 100 A. The I ms impulse duration is taken as a worst-case composite wave estimate. While lightning stroke discharges are typically less than lOOIlS, they can recur
in rapid fire order during a I s duration. From the pulse lifetime rating curves of the t series size 20
models, it is seen that the Vl30LA20 single pulse withstand capability at I ms impulse duration is
slightly in excess of 100 A.
This is adequate for application in areas where lightning activity is medium to light. For heavy
lightning activity areas, an HE series varistor might be desirable to allow a capability of withstanding over 70 transients. In making the choice between the L series and HE series, the designer must
decide on the likelihood of a worst-case lightning stroke and the cost of the fuse replacement should
the varistor fail.
Assuming a low lightning activity area, the Vl30LA20A series is a reasonable choice. To
coordinate the fuse with the vanstor, the single pulse lifetime curve is redrawn as Ft vs. impulse
duration as shown in Figure 4.11. The Ft of the composite 10 / 1000l-ts impulse is found from: 1
I .
·2
= 3' P
Pt
when:
(lOllS) + 0.722 I (T~5)
lOllS)
-
~
200lls (time for impulse current to decay by 0.5)
R:
0.722 I
·2
where:
T(.S)
the first term represents the impulse P t contributed by the lOlls rise portion of the
waveform and the second term is the P t contributed by the exponential decay
portion.
1000
"
~ '\
,<,("Y'
'X). ~
til
0
Z
0
...
...
U
"\
til
0
0::
ct
:J
a
til
100
...
....
til
~~
"\.
~>( ~
~
'" ~~ 1'\
I<'\.X
'V( .A..
"''\
0::
0,.
:Ii
~~~
"'\
ct
II VARISTOR- FUSE COORDINATION CHART
67
Figure 4.11 shows a cross-hatched area which represents the locus of possible failure of the
varistor. This area is equal to an F t value of from two to four times that derived from the data
sheet peak current pulse life curves. The curve extending beyond the cross-hatched area and parallel
to it is where package rupture will take place.
The criteria for fuse selection is given below.
A) Fuse melts; i.e., opens, only if worst-case transient is exceeded and/or varistor fails.
B) If varistor fails, fuse clearing limits 12 t applied to varistor values below that required for
package rupture.
C) Fuse is rated at 130 V RMS.
D) Fuse provides current limiting for solid-state devices.
Based on the above, a Carbone-Ferraz 12 ARMS, 130 V RMS, Class FA fuse is tentatively selected.
The minumum melting 12 t and maximum clearing 12 t curves for the 12A fuse are shown superimposed on
the varistor characteristics.
This fuse is guaranteed to melt at an F t of 40% above the estimated worst-case transient.
Upon melting, clearing F t and clearing time will depend upon available fault current from the
130V RMS line. Figure 4.12 lists clearing times for the selected fuse versus available prospective
circuit current.
PROSPECTIVE CURRENT
AMPS RMS
60
120
240
1200
3600
CLEARING TIME
MILLISECONDS
............................................... 8.0
............................................... 5.6
............................................... 3.5
. : .... ; ........................................ 1.3
............ , .. ; .............................. 0.57
FIGURE 4.12: 12A FUSE - PROSPECTIVE CURRENT VS. CLEARING TIME
As Figure 4.11 shows, a clearing time ofless than 1.5ms is desirable. For fault currents in excess
of 1.2kA, the fuse will clear at less than 24Ns and 1.3ms. This will prevent varistor package rupturing.
However, the distribution line may be "soft," i.e., have a high source impedance at the 60 Hz power
frequency that limits the fault current to values below 1.2kA. Then, it is possible that the fuse would not
protect the varistor package from rupturing, though it would serve to isolate the varistor in any case.
Upon further examination of this example, it is clear that the varistor will be protected from
package rupturing even if the transient pulse current is 50% greater than that of the assumed value,
resulting in an F t of 16 A2 s (Point 2 on Figure 4.11).
Placement of the fuse for this example application could be in the line or in series with the
varistor. If in series with the varistor, the line fuse should be a medium to slow speed, such as a
"slow blow" type 15 A fuse. That would assure a fault in the varistor would be isolated by the
varistor fuse without interrupting the line fuse.
It is desirable to indicate the status of the varistor fuse if one is used in addition to the
line fuse. The circuit shown in Figure 4.13 senses the presence of voltage across the varistor by
use of a photo coupler. When the fuse interrupts the varistor circuit, the LED of the coupler becomes de-energized, and the coupler output signal can be used to annunciate an unprotected
condition. Some fuse manufacturers provide indicating means upon fuse operation that may also
be used to trip an alarm.
68
I
130 V
AC
GE-HIIAA2
AC OPTO COUPLER
~--k2-~"""
I
j
I
~
.
r;t.JI
I
TO STATUS
tl~~~~Clt~~~
TO
PROTECTED
CIRCUIT
I
I1_ _ _ _ _ _ _ ...--1I
FIGURE 4.13: VARISTOR FUSE STATUS SENSING CIRCUIT
In selecting a fuse, the reader is advised to avoid data based on average values or data taken at
operating conditions that are grossly different from the actual application. For example, dc data
does not apply when the fuse will be used on an ac .circuit. Also, test data taken in a resistive circuit
with unity power factor does not hold for low power factor operation.
A list of manufacturers of fast current limiting fuses is given below:
MANUFACTURERS OF CURRENT LIMITING FUSES
4.3
Bussmann Manufacturing Division
McGraw-Edison Company
St. Louis, Missouri 63100
English Electric Corporation
One Park Avenue
New York, New York 10016
Carbone-Ferraz, Inc.
P. O. Box 324
Elm Street
Rockaway, New Jersey 07866
General Electric Company
Power Systems Management Dept.
6901 Elmwood Avenue
Philadelphia, Pennsylvania 19142
Chase-Shawmut Company
347 Merrimac Street
Newburyport,
Massachusetts 01950
General Fuse Corporation (GFC)
7954 Cameron Brown Court
Fullerton Industrial Park
Springfield, Virginia 22153
SERIES AND PARALLEL OPERATION OF VARISTORS
In most cases the designer can select a varistor that meets the desired voltage ratings from
standard catalog models. Occasionally the standard catalog models do not fit the requirements
either due to voltage ratings or energy/current ratings. When this happens, two options are available: varistors can be arranged in series or parallel to make up the desired ratings, or the factory
can be asked to produce a "special" to meet the unique application requirement.
,
4.3.1 Series Operation of Varistors
Varistors are applied in series for one of two reasons: to provide voltage ratings in eXcess of
those available, or to provide a voltage rating between the standard model voltages. As a side
benefit, higher energy ratings can be achieved with series connected varistors over an equivalent
single device. For instance, assume the application calls for a lead mounted varistor with an rms
voltage rating of 375V ac and having a Itm peak current capability of 6000A. The Itm requirement
fixes the varistor size. Examining the L series voltage ratings near 375V ac, only 320V and 420V
69
units are available. The 320V is too low and the 420V unit (V420LA40B) results in too high a
clamp voltage (Ve of 1060V at 100A). For a V130LA20B and a V250LA40B in series, the maximum rated voltage is now the sum of the voltages, or 380V. The clamping voltage, Ve , is now the
sum of the individual varistor clamping voltages, or 925 V at 1000A. The peak current capability
is still 6000A but the energy rating is now the sum of the individual energy ratings, or 140J.
In summary, varistors can be connected in series providing they have identical peak current
ratings (~m)' i.e., same disc diameter. The composite V-I characteristic, energy rating, and maximum clamp voltages are all determined by summing the respective characteristics and/or ratings of
the individual varistors.
4.3.2 Parallel Operation ofYaristors
Application requirements may necessitate higher peak currents and energy dissipation than the
high energy, HE, series of varistors can supply individually. When this OCCj.lrs, the logical alternative
is to examine the possibility of paralleling varistors. Fortunately, all GE-MOV®II Varistors have a
property at high current levels that makes paralleling feasible. This property is the varistor's
series-resistance that is prominent during the' 'up-turn region" of the V-I characteristic. This up-turn is
due to the inherent linear resistance component ofthe varistor characteristic (see Chapter 3). It acts as a
series balancing, or ballasting, impedance to force a degree of current sharing that is not possible at
lower current levels. This is depicted in Figure 4.14. At a clamp voltage of 600V, the difference in
current between a maximum specified sample unit and a hypothetical 2q % lower bound sample would
be more than 20 to I. Thus, there is almost no current sharing and only a single varistor carries. the
current. Of course, at low current levels in the range of lO-lOOA, this may well be acceptable.
1000
(/)
l-
e:
~
600 ,--
tJ
Z
500
J
~ 400
~
<5
300
'"~
200
>
a.
-
--" >-- ~ .--'
...--'
V
L
~ ;..-'
.L-LOWER BOUND (20%)
SAMPLE UNIT
1MB'T
100
0.1
- ---
r-LiMIT SAMPLE
800
.5
IMODIEL ,
r 4 y TO 8 ]
I
I
250
rr
2
500 1000
5 10
50 100
PEAK CURRENT - AMPERES
5000 10,000
FIGURE 4.14: PARALLEL OPERATION OF VARISTORS BY GRAPHICAL TECHNIQUE
At high current levels exceeding 1000 A, the up-tum region is reached and current sharing
improves markedly. For instance, at a clamp voltage of 900V, the respective varistor currents
(Figure 4.14) are 2500A and 6000A, respectively. While far from ideal sharing, this illustration
shows the feasibility of paralleling to achieve higher currents and energy than achievable with a
single model HE varistor.
Practically, varistors must be matched by means of high current pulse tests to make parallel
operation feasible. Pulse testing should be in the range of over IkA, using an 8 / 20ILS, or similar
pulse. Peak voltages must be read and recorded. High current characteristics could then be extrapolated in the range of 100-10,000A. This is done by using the measured data points to plot curves
parallel to the data sheet curves. With this technique current sharing can be considerably improved
from the near worst-case conditions of the hypothetical example given in Figure 4.14.
70
Some guidelines for series and parallel operation of varistors are given in Figure 4.15.
PARALLEL
SERIES
Objective
• Higher Voltage Capability
• Higher Energy Capability
• Non-standard Voltage
Capability
• Higher Current Capability
• Higher Energy Capability
Selection Required
By User
NO
YES
Models Applicable
• All, must have same ~m rating.
L, P, Z, HE, B, D Series
Application Range
• All voltages and currents.
• All voltages - only high currents, i.e., > 100 amperes.
Precautions
•
• Must use identical
rated models.
~m ratings must be equal.
voltage
• Must test and select units for
similar V-I characteristics.
Effect on Ratings
• Clamp voltages additive.
• Voltage ratings additive.
• Current ratings that of single
device.
•
Energy,
~m ,
ratings additive.
• Current ratings function of
current sharing as determined
graphically.
• Energy ratings as above in proportion to current sharing.
• Clamp voltages determined by
composite V-I characteristic of
matched units.
• Voltage ratings that of single
unit.
FIGURE 4.15: CHECKLIST FOR SERIES AND PARALLEL OPERATION OF VARISTORS
4.4
APPLICATIONS
4.4.1 Power Supply Protection Against Line Transient Damage
PROBLEM: It is desired to prevent failure of the power supply shown in Figure 4.16(b) to be
used on residential 117V ac lines. A representative transient generator is to be used for testing, as
50n
100fLH
D
shown in Figure 4.16(a).
,r~ 5kV
~
VT
f\r
=
150fL F
:!: 5kV sin 10 5 71" t X
e- 10- 5t
(b) Typical Power Supply Circuit
{al Transient Generator
FIGURE 4.16: POWER SUPPLY PROTECTION
71
If the transient is applied to the eXisting circuit, the rectifier will receive high negative voltages,
transmitted through the filter capacitor. The LC network is there to prevent RFI from being transmitted into the power line (as in a TY set)~ but also serves to reduce the transient voltage. An
analysis shows that the transient will be reduced approximately by half, resulting in about 2.SkY
instead of SkY at the rectifier.
. This is still too high for any practical rectifier, so some suppression must be added. It is desirable to use the built-in impedance of the coil to drop the remaining voltage, so the suppressor would best
be applied as shown. A selection process for a GE-MOY®II Varistor is as follows:
SOLUTION:
Steady-State Voltage
The 117Y ac, I 10% high line condition is 129Y: The closest voltage rating available is 130Y.
Energy and Current
The 100p,H inductor will appear to be about 300 to the transient. The 300 is derived
from the inductive reactance at the transient generator source frequency of lOS ?r rad. Taking a
first estimate of peak varistor current, 2S00Y/SOO =. 31A. (This first estimate is high, since it
assumes varistor clamping voltage is zero.) With a tentative selection ofa 130V GE-MOY®II Varistor,
we find that a current of 31A yields a voltage of from 32SY to 430V, depending on the model siie, as
shown in Figure 4.17
3000
II)
~ 2000
1000
MAXIMUM CLAMPING VOLTAGE,
,130V AC RATING MODELS-f-
TA = ·55°C TO 85°C
g
~
6
600
500
400
!::4 300
VI30LAIOA~~
VI30LA20A
/
- -
~
z 200
~
II)
~
100
I0-2
,L:
--.
........
><\.:
~
690
,/'
...-
I
l
II
--:±::::
o
./
> 300
II)
"
""
"
""
::>
o
z 200
~
z
ii
II
---
[,-,
L ~
Y'/
500
UJ'
II
II
--
--
~
~/400 ----- - ---r V\
/
103
5 100 \
5 10~J'C.i 102 /
27
31
INSTANTANEOUS CURRENT - AMPERES
510-1
800
UJ
VI30LA20~~
--
~
o
VI30LA2 '
~ 1000
'~ 800
o
'I
I
I
II)
i\
100
5 I,
II
II
~~
...::::
)
I
5
10 ,J\..~/IOO
500100o
"""
27 CURRENT
31
INSTANTANEOU!- AMPERES
FIGURE 4.17: V130LA VARISTOR V·I CHARACTERISTICS
Revising the estimate, I ~ (2S00Y - 32SV)/80n = 27.2A. For model V130LA20B, 27.2A coincides closely with a 320V clamping level.' There is no need to further refine the estimate of peaJc current
If model B remains the final selection.
To arrive at an energy figure, assume a sawtooth current waveform of 27 A peak, dropping to
zero in two time constants, or 20/J.s.
IZ
IV
UJ
II:
II:
::>27-
o
II:
o
lII)
iii:
~
FIGURE 4.18: ENERGY APPROXIMATION
72
Energy is then roughly equal to (27 A x 320V x 20J.l.s)/2, the area under the power waveform. The
result is 0.086J, well within the capability of the varistor (50J). Peak current is also within the
6000 A rating.
Model Selection
The actual varistor selection is a trade off between the clamping voltage desired and the number of
transient current pulses expected in the life of the equipment. A 50J rated varistor will clamp at 315V
and be capable of handling over 105 such pulses. An 8J unit will clamp to approximately 385V and be
capable of handling over 1000 such pulses. Furthermore, the clamping voltage determines the cost of
the rectifier by determining the voltage rating required. A smaller, lower cost varistor may result in a
more expensive higher voltage rectifier diode.
4.4.2 SCR Motor Control
480V
AC 60Hz
FIELD
FIGURE 4.19: SCR MOTOR CONTROL
PROBLEM: The circuit shown in Figure 4.19 experiences failures of the rectifiers and SCR
when the transformer primary is switched off. The manufacturer has tried 600V components with
little improvement.
SOLUTION: Add a varistor to the transformer secondary to clamp the transformer inductive
transient voltage spike. Select the lowest voltage GE-MOV®II Varistor that is equal to or greater than
the maximum high line secondary ac voltage. The V 130LA series fulfills this requirement.
Determine the peak suppressed transient voltage produced by the transient energy source. This is
based on the peak transient current to the suppressor, assuming the worst-case condition of zero load
current. Zero load current is normally a valid assumption. Since the dynamic transient impedance of the
GE-MOV®II Varistor is generally quite low, the parallel higher impedance load path can be neglected.
Determination of Peak Transient Current
Since transient current is the result of stored energy in the core of the transformer, the transformer equivalent circuit shown in Figure 4.20 will be helpful for analysis. The stored inductive
energy is:
=
,...--Zp--,
N
r-zs----,
o-rn~~Nr_.--~
MUTUAL
INDUCTANCE
REPRESENTED
BY IRON CORE
II
IDEAL
TRANSFORMER
FIGURE 4.20: SIMPLIFIED EQUIVALENT CIRCUIT OF A TRANSFORMER
73
The designer needs to know the total energy stored and the peak current transformed in the secondary circuit due to the mutual inductance, Lm. At no load, the magnetizing current, (INL)' is essentially reactive and is equal to im . This assumes that the primary copper resistance, leakage reactance
and equivalent core resistive loss components are small compared to Lm . This is avalid assumption
for all but the smallest control transformers. Since INL is assumed purely reactive, then:
X LM
Vpri
=
INL
and
im =
INL
can be determined from nameplate data. Where nameplate data is not available, Figures
4.21 and 4.22 can guide the designer.
INL
14
I-
Z
~ 12
a:
::>
u
'" 10
\
Z
N
~
8
"'":!<
6
z
"
.......
1'f· 50 . . 60 Hz
IZ
'"~
r-- I--.
4
'"a.
4
6
TRANSFORMER RATING - kVA
10
12
FIGURE 4.21: MAGNETIZING CURRENT OF TRANSFORMERS WITH LOW SILICON STEEL CORE
IZ
Ii:'"
;::
5
~
4
,
""
I'.
I'
.............
N
;::
:i!
3
'"
,
VOLTAGE CLAMP ABOVE
ARC VOLTAGE
w
C>
~ 5(';
g
r----VO-L-TA-G-E-C-L-AM-P~BE~L-O-W--ARC VOLTAGE
u
IX:
u
w
(f)
'oJ
10
'::>
Q.
'"
'"
.q
w
Q.
.q
o.5r------j----t--------+----+---=""""'_.E-~
Q.
2
0·~'~0----;:'50;;:-----::10=0:-----------=50-:-:0:----:-:10~0-:-0------~_:c':_::---:,_J
5000 10,000
IMPULSE DURATION
5
W
I
20
50
100
- I" S
500
1000
IMPULSE DURATION - I" S
(a) V8ZA2 To V100ZA3
(b) V18ZA3 To Vl00ZA15
FIGURE 4.30: ZA SERIES PULSE RATINGS
80
5000 10,000
To use Figure 4.30, the impulse duration (to the 50% point) is estimated from the circuit time
. constants and is found to be l2401Ls. From Figure 4.30(a), the pulse rating is estimated to be slightly
over 108 operations. As this may not be adequate, the designer may wish to go to the next larger size
varistor (V39ZA6). At 0.572A, the approximate impulse duration is now found to be 1280lLs and us•ing Figure 4.30(b), the designer is faced with the problem of extrapolation below IA. This has been
done in Figure 4.31 which is a new plot of the data of Figure 4.30(b) at 12801Ls.
We conclude that the life exceeds 109 operations. The reader may question the extrapolation of
, four orders of magnitude. At low currents the relationship is a straight line extrapolation on log-log
I paper, as seen from Figure 4.30(a), where the pulse rating curves extend to 10 8 pulses.
I
The clipping characteristics of the V39ZA6 model will provide a 61 V maximum peak. The
transistor should have a VeER of 65 V or greater for this application.
15
10
8
~ 6
a::
~ 5
~
""
~
I
Z
ILl
3
a::
a::
VI8ZA3 TO VI00ZAI5
..........
~
~ 4
I-
ZA SERIES
~ v....
.............
............
G2
...........
.......
ILl
(f)
-'
::>
..........
Q.
'"«
ILl
............
I
Q.
-NOTE:
PULSE RATING CURVE FOR 1,280 MICROSECONDS PULSE WIDTH
......
.....
.....
.....
..........
10 6
107
NUMBER OF PULSES
FIGURE 4.31: EXTRAPOLATED PULSE RATING CURVES
4.4.6 Motor Protection
Frequently, the cause of motor failures can be traced to insulation breakdown of the motor
windings. The source of the transients causing the breakdown may be from either internal magnetic
stored energy or from external sources. This section deals with the self-generated motor transients
due to motor starting and circuit breaker operation. Externally-generated transients and their control are
covered in Chapter 2.
In the case of dc moto.rs the equivalent circuit consists of a single branch. The magnetic stored
energy can be easily calculated in the armature or field circuits using the nameplate motor constants. With ac induction motors the equivalent magnetic motor circuit is more complex and the
circuit constants are rarely given on the motor nameplate. To provide a guide for motor protection,
Figures 4.32, 4.33 and 4.34 were drawn from typical induction motor data. While the actual stored
energy will vary according to motor frame size and construction techniques, these curves provide
guidance when specific motor data is lacking. The data is conservative as it assumes maximum motor
torque, a condition that is not the typical running condition. Stored energy decreases considerably
as the motor loading is reduced. Experience with the suppression of magnetic energy stored in transformers indicates that GE-MOV®II Varistors may be used at their maximum energy ratings, even
when multiple operations are required. This is because of the conservatism in the application requirements, as indicated above, and in the varistor ratings. Thus, no attempt is made to derate the
varistor for multiple operations because of the random nature of the transient energy experienced.
81
600
I
i
460 VRMS LINE-LNE
Y CONNECTED
400
"0 ',., '''.J:~
'"
ILl
...J
:::>
0
..,
~'<;~t~
~
I
>- 100
to
0:
80
Z
ILl
0
60
ILl
0:
....0
'"
'"..:
40
ILl
:L
a.
0:
ILl
a.
20
10
8
hI/'
~ '7\/,
n~~~/
200
ILl
.
~V
,,-0
"o_~~,/
/ '"
LI
,,0
-.l."
h Y#
/
/
/
I><
/~ !/4"0.·'
~
/~'0"17
/-.l.~
-.l.">V
~
/
NOTES:
I. Y CONNECTED 60Hz
2. ENERGY AT MAX. TORQUE
SLIP SPEED
3. SEE FIGURE 4.350 FOR
VARISTOR CIRClilT
PLACEMENT
~/
20
40
60 80 100
200
MOTOR HORSEPOWER
400
600 8001000
FIGURE 4.32: STORED ENERGY CURVES FOR TYPICAL WYE-CONNECTED INDUCTION MOTOR
600,-------,--------,----,--.--,--------.-------.----,---,-,
DELTA CONNECTED
400r-------~-------r----+-~--+-------~------~----~--+-~
200
'"
ILl
...J
:::>
~
I
>- 100
to
0:
ILl
80
Z
ILl
0
ILl
60
0:
....0
'"
'":::>
40
ILl
...J
a.
0:
ILl
a.
I. DELTA CONNECTED@ 60Hz
2. ENERGY AT MAX. TORQUE
SLIP SPEED
600 8001000
MOTOR HORSEPOWER
FIGURE 4_33: STORED ENERGY CURVES FOR TYPICAL DELTA-CONNECTED INDUCTION MOTOR
82
600
400
/
MOTOR STORED ENERGY AT START
200
W
...I
::>
..,
0
v"
100
'"a::w
BO
I
z
~~t,;
vO
230 VRMS
lLINE-LINE)\
w
0
w
60
a::
/
0
~
(/)
w
40
(/)
~
X
Q.
a::
w
Q.
20
75-T 9t'
rV
DELTA CONNECTED
APPLIED V.
VARISTOR RATINGS
YCONNECTED
APPLIED V.
VARISTOR RATINGS
230
380
460
550
600
230
380
460
600
550
250/275 420/460 510/550 575/660 660
133
150
220
250/275
266
320
318
420
346
420
FIGURE 4.35: PREFERRED VARISTOR VOLTAGE RATINGS FOR
DELTA· AND WYE·CONNECTED MOTORS
Interruption of motor starting currents presents special problems to the user as shown in
Figure 4.34. Since the stored magnetic energy values are approximately 10 times the running values,
protection is difficult at the higher horsepower levels. Often the motor is started by use of a reduced voltage which will substantially reduce the stored energy. A reduction in starting current of
a factor of two results in a four-fold reduction in stored energy. If a reduced voltage starter is not
used, then a decision must be made between protection for the run condition only, and the condition of locked rotor motor current. For most applications, the starting condition can be ignored
in favor of selecting the varistor for the worst-case run condition.
83
VYARISTOR =
I
./3 VL- L
VYARISTOR
(a) WYE CONNECTED
= VL- L
(b) DelTA CONNECTED
FIGURE 4.36: VARISTOR - 3t:/> INDUCTION MOTOR CIRCUIT PLACEMENT
PROBLEM: To protect a two-pole, 75 hp, 3cp, 460V RMS line-to'-line wye-connected motor
from interruption of running transients.
Specific motor data is not available.
SOLUTION: Consult Figure 4.32 along with Figure 4.35. Standard varistors having the required voltage ratings are the 320V RMS rated models. This allows a 20% high-line voltage condition
. on the nominal 460V line-to-line voltage, or 266V line-neutral voltage. Figure 4.32 shows a twopole 75 hp, wye-connected induction motor, at the running condition, has 52J of stored magnetic
energy per phase. Either a V320PA40 series or a V320HE250 series varistor will meet this requirement. The HE series GE-MOV®II Varistor provides a greater margin of safety, although the PA series
GE-MOV®II Varistor fully meets the application requirements. Three varistors are required, connected directly across the motor terminals as shown in Figure 4.36(a).
4.4. 7 Power Supply Crowbar
Occasionally it is possible for a power supply to generate excessively high voltage. An accidental removal of load can cause damage to the rest of the circuit. A simple safeguard is to crowbar or
short circuit the supply with an SCR. To provide the triggering to the SCR, a high-voltage detector
is needed. High voltage avalanche diodes are effective but expensive. An axial leaded GE-MOV®II
Varistor provides an effective, inexpensive substitute.
PROBLEM: In the circuit below, the voltage, without protection, can exceed twice the normal
240V peaks, damaging components downstream. A simple arrangement to crowbar the supply is .
shown.
The supply shown can provide 2A RMS of short-circuit current and has a lA circuit breaker. A
C 106 SCR having a 4A RMS capability is chosen. Triggering will require at least O.4V gate-to-cathode,
and no more than 0.8V at 200 p,A at 25°C ambient.
I AMP C.B.
",--....,
+
v
~,
CIOSD
NORMAL VOLTAGE < 240V PEAK
ABNORMAL VOLTAGE> 400V PEAK
FULL WAVE
(RECTIFIED)
FIGURE 4.37: CROWBAR CIRCUIT
84
SOLUTION: Check the MA series GE-MOV®II Varistor specifications for a device capable of
supportmg 240V peak. The V270MA4B can handle v'2(l71 V rms) = 242V. According to its specification of 270 V ± 10%, the V2 70MA4 B will conduct 1 mA dc at no less than 243 V. The gatecathode resistor can be chosen to provide 0.4 V (the minimum trigger voltage) at 1 mA, and the SCR
will not trigger below 243 V. Therefore, RcK should be less than 400n. The highest value 5% tolerance resistor falling below 400n is a 360n resistor, which is selected. Thus, RcK is 378n maximum and 342n minimum. Minimum SCR trigger voltage of 0.4 V requires a varistor of 0.4V/378n,
or 1.06 mA for a minimum varistor voltage of ~ 245 V. The maximum voltage to trigger the circuit
is dependent upon the maximum current the varistor is required to pass to trigger the SCR. For the
CI06 at 25°C, this is determined by calculating the maximum current required to provide 0.8V
across a parallel resistor comprised of the 360n RcK selected and the equivalent gate-cathode SCR
resistor of 0.8 V/200 IlA, since the C 106 requires a maximum of 200llA trigger current. The SCR
gate input resistance is 4 Kn and the minimum equivalent gate-cathode resistance is the parallel
combination of 4Kn and RcK(min) , or 360n -5%, 342n. The parallel combination is 325 n. Thus,
lyaristoJor maximum voltage-to-trigger the C106 is 0.8V/315n: or 2.54mA. According to the specification sheet for the V270MA4B, the varistor will not exceed 330V with this current. The circuit
will, therefore, trigger at between 245 and 330V peak, and a 400V rated CI06 can be used. The
reader is cautioned that SCR gate characteristics are sensitive to junction temperatures, and a value
of 25°C for the SCR temperature was merely chosen as a convenient value for demonstrating design
procedures.
Figure 4.3 can be used to determine the maximum energy per pulse with this waveform. It will not
exceed approximately 112 /0.15 / Ipk / Vpk / T (duration of 1/2 wave pulse), or 0.S2mJ for this example.
Since the voltage does not drop to zero in this case, the SCR remains on, and the varistor sees only one
pulse; thus, no steady-state power consideration exists.
4.4.8 General Protection of Solid State Circuitry, Against Transients on 117 V ac Lines
PROBLEM: Modem electronic equipment and home appliances contain solid state circuitry
that is susceptible to malfunction or damage caused by transient voltage spikes. The equipment is
used in residential, commercial, and industrial buildings. Some equipment designs are relatively low
cost consumer items while others are for commercial/industrial use where an added measure of reliability is needed. Since solid state circuits and the associated transient sensitivity problem are relatively new, the knowledge of design in the transient environment is still incomplete. Some test
standards have been adopted by various agencies (see Chapter 7), and further definition of the
environment is underway by the Surge Protective Devices Committee of the IEEE.
The transients which may occur on residential and commercial ac lines are of many waveshapes
and of varying severity in terms of peak voltage, current, or energy. For suppressor application purposes, these may be reduced to three categories.
First, the most frequent transient might be the one represented by a 30 or 100 kHz ring wave.
This test surge is defined by an oscillatory exponentially decaying voltage wave with a peak open
circuit voltage of 6kV. This wave is considered representative of transients observed and reported
by studies in Europe and North America. These transients can be caused by distant lightning strikes
. or distribution line switching. Due to the relatively high impedance and short duration of these
transients, peak current and surge energy are lower than the following categories.
The second category is that of surges produced by nearby lightning strokes. The severity of a
lightning stroke is characterized in terms of its peak current. The probability of a direct stroke
of a given severity can be determined. However, since the lightning current divides in many paths,
the peak current available at an ac outlet within a building is much less than the total current of
85
the stroke. The standard impulse used to represent lightning and to test surge protective devices is an
8 /20 p,s current waveshape as defined by ANSI Standard C68.2, and also described in ANSI/IEEE
Standard C62.41-1980.
A third category. of surges are those produced by the discharge of energy stored in inductive
elements such as motors and transfonners. A test current of 10 / 1000p,s waveshape is an accepted
industry test impulse and can be considered representative of these surges.
Although no hard-and:-fast rules can be drawn as to the category and severity of surges which
will occur, a helpful guideline can be given to suggest varistors suitable in typical applications.
This guideline recognizes considerations such as equipment cost, equipment duty cycle, effect
of equipment downtime, and balances the economics of equipment damage risk against surge protection cost.
GE-MOV® VARISTOR SELECTION GUIDELINE FOR 117 V AC APPLICATION
APPLICATION TYPE
tight Consumer
Consumer
Consumer
tight Industrial
Industrial
Industrial
Industrial
DUTY CYCLE
Very Low
Low
Medium
Medium
Medium
High
High
LOCATION
EXAMPLE
Mixer/Blender
Portable TV
Console TV
Copier
Small Computer
Large Computer
Elevator Control
A
A
A
A
A
B
B
SUGGESTED MODEL
V 130LA2
V 130LA lOA
V130LA20A
i V130LA20A or B
V130PA20A or C
V131DA40
V151DA40
REFERENCE
1. Kaufman, R., "The Magic ofFt," IEEE Trans. IGA-2, No.5, Sept.-Oct. 1966.
86
SUPPRESSIONTELECOMMUNICATIONS SYSTEMS
5.1
INTRODUCTION
Modern telecommunication systems are fast, efficient, and complex. Many improvements have
been made in central office equipment and subscriber. equipment which involves the use of solid
state circuitry. Unfortunately, solid state devices are much more susceptible to malfunction or
failure due to transient voltages and noise than are older devices, such as relays, coils, step-switches,
and vacuum tubes. To complicate matters further, increased usage of telecommunication lines for
data and video transmissions has produced a further intolerance for transient voltages.
Although telecommunications systems have always employed transient protection devices such
as the carbon gap, the gas tube, and the heat coil, these are not 'always adequate to protect solid
state circuitry. The GE-MOV®II Varistor shows promise of providing the extra protection necessary for
even more reliable telecommunications.
5.2
SYSTEM TRANSIENTS
A telecommunication system is made up of subscriber stations linked together through the
cable plant and a central office switching network. Included in the system are repeater amplitiers,
multiplexers, and other electronic circuits. Supplying the electrical energy to run the system is a
main power source.
The cable plant and the power supply provide a path by which damaging transients enter the
system, to be transmitted to vulnerable electronic circuitry. The cable plant consists of conductors
in shielded cables, which are suspended on poles (shared with power lines) or buried in the earth. A
single cable is made up of many conductors, arranged in twisted pairs (tip and ring). Some sections
of open-wire transmission lines are still used, but most of these are remote from central offices, and
transient protectors are usually provided where the open wire enters the shielded cables. All of these
cables (even the ones underground) are capable of picking up transient energy from lightning and
conducting them to the central office or subscriber equipment.
The power used by a telecommunication system is usually obtained from commercial power
lines. These lines, like the telephone cables, are either suspended on poles or buried. Transient
energy is frequently picked up by power lines and transmitted to the central office by direct conduction or by induction into the telephone cable plant. The increased use of off-line power supplies
in telephone equipment makes power line transients even more hazardous to the electronic circuitry.
5.3
LIGHTNING -INDUCED TRANSIENTS
Lightning is the most common source of over voltage, in communication systems. Because of
the exposure to lightning strokes, a knowledge of the effects of lightning is important when designing a transient protective system.
Lightning cutrents may enter the conductive shield of a suspended cable by direct or indirect
stroke,or it may enter a cable buried in the ground by ground currents, as shown in Figure 5.1.
87
SHIELD
=
FIGURE 5.1: LIGHTNING CURRENT IN BURIED CABLE
In the case of a suspended cable, the lightning current that enters the cable is seeking a
ground and will travel in both directions along the cable. Some of the current will leave the shield at
each grounded pole along its path. Studies have shown that all of the lightning current has left the
cable shield after passing 10 poles grounded in high-conductivity soils or 20 poles grounded in high
resistivity soil.
Stroke currents leave a buried cable in a similar way but with a different mechanism. Since the
cable shield has a finite electrical resistance, the current passing through it will produce a potential
gradient along its length. This voltage will produce a potential difference between the cable and the
soil, as shown in Figure 5.2.
SHIELD VOLTAGE
IU
(!)
~
..J
o
>
SOIL VOLTAGE
"TRUE'GROUND
o
DISTANCE
_
FIGURE 5.2: CONDITION FOR PUNCTURE OF CABLE JACKET
At some point (Point A) the shield-to-earth potential will exceed the dielectric strength of the
jacket, causing it to puncture. Some of the lightning current then flows through the puncture into
the soil, thus equalizing the potential at that point. The remaining current continues along the
shield until another puncture occurs, providing another path to ground.
Lightning currents are usually not harmful to the shield itself, but they do induce surge voltages on the conductors of the cable which are often harmful to central office equipment. The surge
voltage that appears at the ends of the cable depends upon the distance to the disturbance, the type
of cable, the shield material, and its thickness and insulation, as well as the amplitude and waveshape of the lightning current in the shield. Since the current-derived potential along the cable
shield is capacitively coupled to the cabled conductors, the wave shape of the surge voltage on the
conductors will closely resemble the waveshape of the lightning current.
Quantitative information on lightning has been accumulated from many sources,2 with research centers in the United States, Western Europe and South Africa. One of the most comprehensive surveys of available data has been compiled by Cianos and Pierce,3 describing the amplitude,
rate-of-rise, duration, etc., in statistical terms.
88
Using these statistics, one can make numerical calculations of induced voltages in various electrical circuits, such as the cable plant of a communication system. The parameters of interest are
the voltages developed as a function of intensity and duration of the lightning impulse. The examples discussed later in this chapter are based on this source of information.
5.4
CALCULATIONS OF CABLE TRANSIENTS
The voltage surge induced into the conductors of a cable will propagate as a traveling wave in
both directions along the cable from the region of induction. The cable acts as a transmission line.
The surge current and voltage are related to each other by Ohm's law where the ratio of voltage to
current is the surge impedance (Zo) of the cable. Zo can also be expressed in terms of the inductance (L) and capacitance (C) per unit length of the cable by the. equation,
Zo
= .jJJC
(n)
The velocity of the surge, as it propagates along the conductors, is also a function of Land C, and
can be expressed as
Velocity = --J lILC (meters/sec.)
The series resistance of the shield and conductors, as well as losses due to corona and arcing, determine the energy lost as the disturbance propagates along the cable.
Tests conducted on telephone cables4 have measured surge impedances of 80n between any
of the conductors and the shield. Shield resistances between 5 nand 6 n per mile were found to be
typical. These values and the applied lightning current waveform of Figure 5.3 were used to compute the worst case transient which would appear at cable terminals in a central office. The
computation assumes the lightning current is introduced into a suspended cable shield at a point
2.75 miles from the central office. An average cable span between poles of 165 feet, with a ground
connection on every fourth pole, was assumed. It was also assumed that the cable will support the
voltage without arcing over.
100
~
(/)
~ 60
UJ
(l.
~
<
, 40
~
IZ
UJ
-...
~ 20
~
::>
u
o
o
25
75
50
100
TIME - MICROSECONDS
FIGURE 5.3: SEVERE LIGHTNING CURRENT WAVEFORM (2/50
355
tI)
::! 284
/
II
III
(l.
~
71
o
o
/
10
J
!
""
16.2
!:i
g
11.0
f
7.3
~
!:i
3.7
~
o
20
30
40
/\
V\
(/) 14.7
"'-......
>
50
TI ME - MICROSECONDS
0
o
/LS)
/
10
/
J
20
~
30
........
50
TIME - MICROSECONDS
FIGURE 5.4: AVAILABLE CURRENT 2.75 MILES
FROM 100kA LIGHTNING STROKE
FIGURE 5.5: OPEN CIRCUIT VOLTAGE - 2.75
MILES FROM 100kA LIGHTNING STROKE
89
The resulting short-circuit current available at the central office is shown in Figure 5.4.
The open-circuit voltage at the cable end is shown in Figure 5.5. This analysis shows that if a
severe, 100kA lightning flash strikes a cable at a point 2.75 miles from a central office, a voltage
transient reaching a peak of nearly 18kV may appear at the cable end, with about 355A of current
available.
Since the cable can be considered to be a linear system, the voltages and currents will bear a
linear relation to the lightning stroke amplitude. A tabulation of the open-circuit voltage and available current which would result from stroke currents of various magnitudes is given in Table 5.
Included in the table is the probability of occurrence, as given by Cianos and Pierce. 3 It should be
realized that voltages in excess of IOkV probably would not be sustained since the cable insulation
will break down:
The transient voltage at the central office in any case would be excessive, so that protectors
would be required. The protector would conduct up to 213A surge current for most (85%)of the
strokes that occur and only up to twice that current on rare occasions.
TABLE 5.1
LIGHTNING TRANSIENTS AT CABLE END 2.75 MILES FROM STROKE POINT
LIGHTNING STROKE,
PEAK CURRENT
PROBABILITY
OF
OCCURRENCE
...
TERMINAL
OPE.N-CIRCUIT
VOLTAGE
TERMINAL
SHORT-CIRCUIT
CURRENT
(kAI
(%1
(PEAK VI
(PEAK AI
175
100
60
20
1
5
15
50
32,200
18,400
11,040
3,680
621
355
213
71
The values shown in Table 5.1 are based on the assumption of a single conductor cable with
the stroke point 2.75 miles from the central station. For closer strokes the peak short-circuit current at the cable end will increase as shown in Table 5.2. These calculations were made assuming a
breakdown at the stroke point, which gives the worst case result.
Since telephone cables actually have many pairs of wires rather than a single conductor,
the peak currents in each wire will be lower. It is assumed that the stroke voltage will be induced
equally in all wires if they are equally loaded. Then, the currents in all wires will be equal if all proTABLE 5.2
PEAK LIGHTNING-INDUCED CURRENTS IN VARIOUS LENGTHS
OF TELEPHONE CABLE (100kA LIGHTNING STROKE)
DISTANCE
TO STROKE
(Miles)
2.75
1.50
1.00
0.5
0.25
PEAK CURRENTS (A)
AT
STROKE
POINT
630
630
734
1110
1480
AT CENTRAL OFFICE
SINGLE
CONDUCTOR
6 PAIR
CABLE
12 PAIR
CABLE
355
637
799
1120
1480
-
-
90
-
-
712
852
453
463
tectors are identical. To predict the individual wire currents, it is assumed that the wire currents are
proportional to sheath current and the ratio of resistances, and are reduced a constant amount by cable
inductance. Worst case calculated values for the shortest distances are shown in Table 5.2
An example of the current which a protective device must handle can now be estimated.
Assume a cable of six pairs (the smallest available) is struck by lightning, inducing a stroke current
of 100kA into the shield, at a distance of 0.25 mile from the protector. The transient current will
be divided up among the twelve suppressors at the cable ends. Each protective device must handle
up to 852A of peak current in order to clamp the voltage to a protected level.
5.5
POWER SYSTEM-INDUCED TRANSIENTS
Since telephone cables very often share a pole and ground wire with the commercial ac utility
power system, the high currents that accompany power system faults can induce over-voltages in
the telephone cables. These induced over-voltages will be at the power system frequency and can
have long duration (compared to the lightning-induced transients) from a few milliseconds to several
cycles of power frequency. Three types of over-voltage can occur in conjunction with power system
faults:
Power Contact
Power Induction
Ground Potential Rise -
(Sometimes called "power cross") The power lines fall and make contact
with the telephone cable.
The electromagnetic coupling between the power system experiencing a heavy fault and the telephone cable produces an over-voltage in
the cable.
The heavy ground currents of power system faults flow in the common
ground connections and cause substantial differences in potential.
There is little definitive data available on the severity of these over-voltages. However, proposals have been made by telephone protection engineers to define the power contact as the most
severe condition. The proposed requirement calls for the suppressor to withstand lOA RMS for a
duration of power Gontact ranging from 10 to 60 cycles of the power system frequency.
5.6
PROTECTORS - VOLTAGE TRANSIENT SUPPRESSORS
5.6.1 Primary Protection
The oldest and most commonly used primary protector for a telephone system is the carbon
block spark gap. The device is made up of two carbon block electrodes separated by a small air
gap of between 0.003 and 0.004 inch. One electrode is connected to the telephone cable conductor and the other to the system ground. When an over-voltage transient appears, the gap breaks
down diverting the transient and dissipating the energy in the arc and the source impedance of the
transient. The carbon gap is a low-cost protector but suffers from a relatively short life and exhibits
sparkover voltage variations. Nominal 3-mil carbon gaps statistically sparkover as low as 300V and
as high as 1000V - this is a serious problem.
Telephone conductors occur in pairs in a cable so that transient voltages induced into the conductors will be common to both tip and ring conductors, as shown in Figure 5.6. This longitudinal
voltage produces no current through the load termination. Normally , there is zero difference in the
potential between conductors. If protector, PR 1 should break down at 400V, while PR2 requires
700V to break down, then only PR 1 would break down on a transient of 600V causing a transient
current flow through the load. Even if PR 2 does break down but responds later in time than PR 1
a transient current will flow.
91
iNDUCED
LONGITUDINAL VOLTAGE
IN CONDUCTOR PAIR
EQUIPMENT
TERMINATION
I
FIGURE 5.6: UNBALANCED LINE PROTECTION
Another common suppressor in telephone systems is the gas tube protector. It consists of two
metallic gaps spaced by distance of 0.010 to 0.015 inch. The electrodes are enclosed in a sealed
glass envelope containing a combination of gases at a low pressure. Such gaps offer higher currentcarrying capability and longer life than do carbon block devices. However, the possibility of seal
leakage. and the consequent loss of protection has limited the use of these devices. Dmit-gap gas
tubes, also called three-electrode gas tubes, have been introduced to alleviate the problem of unbalanced breakdown as described in the preceding paragraph,
a
GE-MOV®II Varistors have properties that make them excellent candidates for telephone
system protectors. These characteristics include .tight tolerance, high reliability, high energy cap- .
ability, and good clamping characteristics. The Vl30LA20A GE-MOV®II Varistor, for instance; is
capable of handling a peak transient current of 6000A (8/20 p.s pulse) and dissipating up to 501 of
energy. The 6000 A current surge would result in the voltage across the varistor being clamped at a
maximum of 600V. A 1000A pulse would be clamped to less than 420V, yet ring voltage peaks of
180V would not be affected by this varistor.
5.6.2 Secondary Protection
Modern solid state communications circuitry can be damaged even if the primary protection is
working normally. It is often advisable to provide a secondary protection system to further reduce
the voltage transient. As shown in Figure 5.7, the secondary protection removes the over-voltage
spike which is passed by the primary protector.
ZLI--
TIME OF
I rRESPONSE
PRIMARY PROTECTION
~ ZL2
....,.r--,...~~::~::------.::----
---.-~~~r--'-----.~~-'-~~----.----'~6~~~l
. GE-MOV@VARISTOR
SECONDARY
PROTECTOR
PRIMARY
PROTECTOR
-=
FIGURE 5.7: SECONDARY PROTECTION
92
In most installations the length of conductor between the primary protector and the telephone
circuit boards is greater than 25 feet. The impedance (ZL2) presented by this length of wire to most
lightning-induced transient voltages will insure that the primary protector will operate first and the
secondary protector will not be exposed to the full surge. In the rare cases where a power cross
occurs, the varistor may fail, but it will still perform its assigned task of protecting the circuit board.
Because its failure mode is a short circuit it will blow the system fuses. Usually the probability of a
power cross is so low that the replacement of a damaged varistor is an acceptable alternative to repairing a damaged circuit board.
5.7
POWER LINE TRANSIENTS
For transients introduced into a telecommunications system through the power lines, the
GE-MOV®II Varistor is a very effective suppressor. Properly selected, the varistor will not effect the
normal operation of the line but will clamp heavy transient surges to an acceptable voltage level.
Refer to Chapters 2 and 4 for information on the selection of a varistor suppressor.
5.8 RELAY CONTACT PROTECTION
Even the most modern telephone equipment requires the use of relays and other electromechanical switching devices. These devices are required to switch currents into inductive loads
causing contact arcing, pitting, and noise generation. The GE-MOV®II Varistor is a useful suppressor
for increasing contact life, improving reliability, and reducing noise. Chapters 1 and 4 contain selection
information for contact protection applications by means of varistors.
REFERENCES
1. Bennison, E., P. Ferland and A.J. Ghazi, "Lightning Surges in Open-Wire, Coaxial and Paired
Cables," IEEE International Conference on Communications, June 1972.
2. Golde, R.H., "Lightning Currents and Related Parameters," Lightning, Vol. 1, Physics of Lightning, Chapter 9, ed. R.H. Golde, Academic Press, 1977.
3. Cianos, N. and E.T. Pierce, "A Ground Lightning Environment for Engineering Usage," Report
No.1, Stanford Research Institute, August 1972.
4. Boyce, C.F., "Protection of Telecommunication Systems," Lightning, Vol. 2, Lightning Protection, Chapter 25, ed. R.H. Golde, Academic Press, 1977.
93
Notes
.~-@
SUPPRESSION\.S2)
AUTOMOTIVE TRANSIENTS
6.1
TRANSIENT ENVIRONMENT
The designer of electronic circuits for automotive applications must ensure reliable circuit
operation in a severe transient environment. The transients on the automobile power supply range
from the severe, high energy, transients generated by the alternator/regulator system to the lowlevel "noise" generated by the ignition system and various accessories (motors, radios, transceivers,
etc.). Transients are also coupled to the input and output terminals of automotive electronics by
magnetic and capacitive coupling in the wiring harness, as well as conductive coupling in common
conductor circuits (especially the chassis "ground"). Steady state over-voltages may be applied by
the circuit power supply due to the voltage· regulator failure or the use of 24 V battery "jump"
starts. The circuits must also be designed against the possibility of the battery being connected in
reverse polarity. Circuits which drive inductive loads must be protect~d against the transients resulting from the energy stored in the field of the inductor. These transients can be defined from the
load inductance and load current. Figure 6.1 summarizes the automotive power supply transients
as documented by the Society of Automotive Engineers (SAE).I
LENGTH OF
TRANSIENT
ENERGY CAPABILITY POSSIBLE FREQUENCY
VOLTAGE AMPLITUDE
OF APPLICATION
CAUSE
00
Steady State
Failed Voltage Regulator
5 Minutes
Booster starts with 24 V battery
4.5ms to 0.1 s
Load Dump - i.e., disconnection
of battery while at high charging
rates.
~
0.32 s
Inductive Load Switching
Transient
~
0.20 s
Alternator Field· Decay
90ms
Ignition Pulse, Battery
Disconnected
1 ms
Mutual Coupling in Harness*
15
~s
Infrequent
+ 18 V
00
Infrequent
±24 V
~ 101
~ 125 V
Infrequent
< 1J
-300Vto +80V
< 1J
-IOOV to -40V
< 0.5 J
~
< 1J
< 0.001
~
Accessory Noise
~
Transceiver Feedback
Often
3V
~ 500 Hz
Continuous
1.5 V
50 Hz to 10 kHz
J
20mV
*These transients may be present on any wire in the vehicle.
FIGURE 6.1: TYPICAL AUTOMOTIVE SUPPLY TRANSIENT SUMMARY
95
Each Tum-Off
~ 500 Hz
Several times in
vehicle life
75V
~200V
Ignition Pulse, Normal
Often
R.F.
Achieving maximum transient protection involves many factors. First, consequences of a failure
should be determined. Current limiting impedances and noise immunity requirements need to be
considered. The state of the circuit during the transient (on, off, unknown) and the availability of
low-cost components capable of withstanding the transient are other factors. Considerable variation has
been noted in the data gathered on automotive transients. Further, the interaction of other parts of the
automotive electrical system with the circuit under transient conditions may require definition. The
empirical evaluation oftransient suppression using SAE-recommended test circuits,2 is invaluable in
many cases. Figure 6.2 illustrates the test waveform for the most common, high energy transients.
6.2
VARISTOR APPLICATIONS
To illustrate the procedures involved in designing transient protection for automotive electronics, two examples are provided. One example illustrates the protection of a solenoid driver circuit consisting of a logic integrated circuit with power transistor buffer; the second is the protection
of an ignition circuit output transistor. These examples also illustrate the difference between protecting against random and repetitive transients. For random transients, energy and clamping vs.
standby power dissipation are dominant constraints. For repetitive transients, transient power dissipation places an additional constraint on the choice of the suppression device. The solenoid driver
protection circuit also illustrates the conflicting constraints placed on automotive transient suppressors by the low maximum voltage ratings of integrated circuits, the 24 V jump start cycle and the
load dump transients.
Vp
SOURCE IMPEDANCE
a) MUTUAL COUPLINj3
~ 0.8.0.
TRANSIENT - TESTED
IN BOTH POLARITIES,
SERIES AND PARALLEL
INJECTED
-O.2Vp
I
10
30
20
v
VOLTS
b) LOAD DUMP
t< 0
14.1
eo
0
T080vl
1500
!
!
OUTPUT
1000
BEFORE
.....
_-LOAD
DUMP
SUPPRESSOR
UNDER
TEST
CI
.03 F
500
O~
-.08
__
~
-.04
__
~
__
0
~
.04
__
~
__
.06
~~-L
.12
__~_
.16
.2
TIME (SECONDS)
FIGURE 6.3: ALTERNATOR POWER OUTPut INTO
FIGURE 6.4: LOAD DUMP SIMULATOR CIRCUit
A CENTRAL SUPPRESSOR
100-
STANDBY CURRENT AT 12V (mA)
THRESHOLO· OF THERMAL RUNAWAY:> 100 (mA)
1
mA
+5
PROPOSED END-POINT QUALIFICATION
-- --- -
- -----------------
0.1
CLAMPING VOLTAGE CHANGE ("to AT 20Al
0.01
N=8
('Yo)
b
~==========::::======:::
10
DUMPS
AT 100J
-1 DUMP
AT 200 J
JI.A1P
START
0.001 L-__________- L____________- L_____________
10 DUMPS
AT 100J
1 DU,MP
AT 200J
24V
-5
FIGURE 6.6: STABILITY OF CLAMPING VOLTAGE
FIGURE 6.6: STABILITY OF STANDBY CURRENT
JUMP
START
24V
6.2.2 Protection Of Electronic Ignition
In the second example, the protection of the output power transistor in an electronic ignition
circuit is analyzed. This power transistor performs the current switching function of mechanical
distributor points in the usual Kettering ignition, thus avoiding the pitting, burning, and erosion
mechanisms associated with the mechanical points. The ignition circuit is illustrated in Figure 6.7.
In normal operation, the coil primary current builds up when the power transistor is on,
storing energy in the coil inductance. The power transistor is then switched off, and the voltage
at the collector rises rapidly as the capacitor, C, charges. Transformer action causes the secondary .
voltage to rise until the spark plug reaches firing voltage, clamping the transistor· collector voltage
at a safe value. If a spark plug is fouled or disconnected, the collector voltage can rise until
either the capacitor contains the stored energy (minus losses), or the transistor breaks down with
resulting damage/failure. Since the capacitor is small, transfer of the stored energy of the coil to the
capacitor would result in a very high voltage requiring transistor protection. A varistor can be used
to tum the transistor on during the period of high voltage, thus dissipating the excess energy safely
as heat. The constraints on varistor selection are: clamp voltage must be low enough to protect the
transistor; clamp voltage must be high enough to not affect normal spark energy; the power dissipation (with two spark plugs disconnected) must be within varistor ratings for an 8-cylinder, 4-cycle
engine at 3300 rpm (misfires at 55 Hz, average). The minimum spark voltage output required is
20,000V, which represents 200V at the transistor collector. The transistor has a breakdown voltage
VB
0-------.;_....,
START
SWITCH
I.en
SYSTEM CONDITIONS
TIMING
SIGNAL
CONDITION
Va
START
SWITCH
RUNNING
12V TO 16V
OPEN
STARTING
5V TO 12V
CLOSED
OPERATING TEMPERATURE RANGE
-40·C TO +IIO·C
JL
POWER
TRANSISTOR
FIGURE 6.7: EXAMPLE 2
TYPICAL ELECTRONIC IGNITION CIRCUIT
rating of 400V with the 47Q base emitter resistor and a current gain over 20. The base emitter onstate voltage, VBE(ON)' is between 1.0 and 1.8 V, and the collector to emitter saturation voltage is
between 0.9 and 1.5V. The varistor clamp voltage range is determined by the 200V needed to
supply minimum spark voltage and the 400V rating of the transistor. At 200V the varistor current
must be less than:
'
IV
VBE(ON)/47Q = 47Q= 0.02A
to prevent unwanted transistor turn-on. The minimum varistor voltage at the 1rnA varistor specification point is found by solving the varistor voltage equation:
I = kya,
assuming a maximum a of 40. The result is 186V. The peak clamping current (at 400VVBE(MAX») is found from the energy balance equation for the coil, using the peak coil current, Ie.
Ie maximum is analyzed under both start and run conditions to determine the worst case:
99
IC(start) .;;;;
12 -,-- 0.9
1.80
= 6.17A
IC(run) .;;;;
16 - 0.9
3.60
=
and,
4.2A
The worst case coil current occurs with the start switch closed and will be less than 6.2A. The maximum peak coil current, Ip ' when clamping is then:
YzLI2
=Y2LI2
+Y2CVP 2
c
P
and with a Vp of 400 V:
Ip 2
= IC 2
-
400 2 C. /L
results in 6.0A starting and 3.6A running. The varistor currents corresponding to this are:
Ip/hFE + vBE/47n;
which gives 0.34A starting and 0.22A running. Peak varistor voltage must be less than:
400V - VBE (i.e., 398V at 0.34A)
The varistor power dissipation at 3300 rpm (55 pps), assuming a triangular current waveform with
constant voltage and no losses, is found from coil energy balance:
solving for t:
t -
-
(7)(10-3) H (3.6A) - 63
400V
p,s
The varistor power dissipation is found to be:
VMAX .
ktf
=
398V (0.~2A)(63)(1O-6)s (55pps)
=
O.15W
Observations indicate that the losses in the coil and reflected secondary load will reduce this by half
to about 75mW. Using the 110°C ambient temperature derating factor of 0.53, it is found that a
varistor of 0.15 W dissipation capability is required. The varistor parameters are now defined as Vx
of at least 186 V at 1 rnA but less than 398 V at 0.34 A and capable of at least 0.15 W dissipation.
The V220MA2A and V270MA4B both fit these requirements.
As these examples have illustrated, the use of the GE-MOV®II Varistor in automotive circuits
for transient protection is both technically and economically sound. Design procedures are identical
to the procedures used in the other environments. Experimental verificatiori of the degree of protection can be made using standard waveforms reported by automotive engineering investigators.
REFERENCES
1. Preliminary Recommended Environmental Practices for Electronic Equipment Design, Society of
Automotive Engineers, 2 Pennsylvania Plaza, N.Y., N.Y. 10001.
2. Electromagnetic Susceptibility Test Procedures for Vehicle Components (except Aircraft),
Society of Automotive Engineers, 2 Pennsylvania Plaza, N. Y., N. Y. 10001.
100
©k!ill~ll
VARISTOR TESTING
7.1
INTRODUCTION
As with any device, metal oxide varistors possess a number of parameters which can be identified and measured in several ways. However, to minimize testing effort, the test parameters should
be reduced to the essential few. Also, tests should be conducted in a standard way to assure correlation of measured values between maker and user. The essential varistor parameters are defined in
Chapter 3. This chapter will detail the tests of these varistor parameters, describe suitable test
methods using simplified test circuits, and list some available test equipment.
It should be noted that all tests are performed at 25°C, unless otherwise specified. Also, the
test circuits and methods given herein are intended as a general guide only, and may not be generally
applicable to the test equipment available to the user. Since the tests frequently entail high voltages
and currents, the user must exercise appropriate safety precautiorts.
7.2
TEST OBJECTIVES
Varistor testing that would be undertaken by a user will depend considerably on prior know ledge of
both the device and the application. Factors are the relative severity of the application (both electrical
and environmental), the number of devices to be used, and the possible adverse effect of device
misapplication or malfunction. Further considerations are resources available to the user and the
economics of alternate uses of those resources versus more extensive varistor testing. Equipment
makers designing transient protection into their products will have different objectives in varistor
testing than a user simply adding a few varistors to existing equipment as a protective step. Finally, the
user may have different test requirements depending on which poin.t in the cycle of system design or
component evaluatio~ and procurement the testing is being done.
7.2.1 Engineering Evaluation
For the original equipment maker, the process of evaluating and procuring a new component
begins with an initial evaluation of the component itself. Typically, the circuit or systems engineer
will obtain a few samples of the candidate component for evaluation in the prototype equipment
design. He may seek recommendations from his component engineer in selecting from devices available in the market. It is important to focus on the key characteristics and ratings to determine if the
component can perform as expected. Typically, varistor voltage, clamping voltage, standby current, insulation resistance, and ~apacitance of the samples should be measured according to the
methods given in section 7.3. Assuming that a varistor type has been selected according to the
design application examples of Chapter 4, the engineer obviously will verify that the component
performs as expected when placed in the breadboard circuit. Also, it should be verified that variation of these parameters within their specification values are consistent with the application requirements. The surge current, or energy, and waveshape available in the circuit together with its
frequency of occurrence should be measured or computed. These characteristics of the expected
transients should then be checked against the pulse lifetime and the power dissipation ratings of the
selected varistor type. Where suitable equipment is available, the rating of the varistor may be
verified by injections of transIents into the varistor alone or into the prototype circuit. See section
7.6 and 7.7 of this chapter for a discussion of transient test equipment and test waves.
101
7.2.2 Product Qualification
In some user organizations, selection and evaluation of the varistor as a component may pass
to a specialized group that evaluates component engineering and reliability. The final output of this evaluation will be a purchase specification detailing the mechanical and electrical requirements and ratings of the component, and possiMe: approvedsources for the part. A product qualification plan often will be used to detail the electrical and environmental tests to which a sample of
the candidate component may be subjected and which it must pass in order to be approved. Frequently the manufacturer will be asked to supply supporting data for his in-house testing to supplement and minimize the qualification testing. The suggested electrical characteristics tests are (with
appropriate conditions and limits): nominal varistor voltage, VN ; maximum clamping voltage, Vc; dc
standby current, ID (optional, especially for ac applications); insulation resistance; and capacitance.
These characteristics will be measured frequently in the component/equipment cycle thereafter, and
care should be excercised that they are neither too many and complex nor too few to be meaningful to the
application. Reliability requirements of operating conditions and expected life will sometimes be
specified and usually tested for early in the qualification phase of the component. These tests may be
performed at. special conditions of environment or temperature to stress the component as proof of its
intended use or design capability. A test to insure surge current withstand capability may be included in
the qualification plan. This test must be carefully performed and specified (by using either 8/20",s or 10
/ 1000",s waveshapes) in line with the recommendations of Chapter 3 and consistent with the pulse
lifetime rating chart of the varistor selected. Other qualification tests may be used to ensure mechanical
integrity, humidity resistance; solderability, and terminalllead strength. These tests should be of a
standard nature wherever possible to assure reproducibility.
7.2.3 Incoming Inspection
Once the component has been qualified, the equipment maker will wish to verify that shipments received consist of correct parts at the expected quality level. Shipments will be sampletested to assure correct markings, appearance, finish, and major or critical electrical parameters. It is .
especially desirable to prevent material with incorrect voltage characteristics from entering assembly operations so as to minimize troubleshooting and rework. For incoming inspection of
GE-MOV®II Varistors, it is recommended that sample testing include nominal varistor voltage, VN,
tested against the minimum and maximum voltages specified on the purchase drawing/specification.
Components below specification limits may lead to premature degradation or circuit failure. If
above specification, they may not deliver the required protection from transients and may possibly
allow other failures. Other electrical sampling tests frequently performed can include insulation
resistance and capacitance. Tests such as maximum clamping voltage, Vc , and dc standby current,
ID , are usually checked only on a periodic audit basis.
7.2.4 Field Maintenance
Field maintenance testing is done to verify that the varistor is still providing the intended
protection function or, in the case of sensitive circuit applications, that the varistor has not degraded.
Since the usual change of GE-MOV®II Varistor characteristics when over-stressed is toward lower
resistance, it is very unlikely that the protection function will deteriorate unless the electrode system is
damaged. The varistor should be physically examined for loose leads, charred or broken areas in the
encapsulant, solder dribbles on the leads, or other evidence of overheating damage. If physically
.
acceptable, the varistor may be tested electrically.
The nominal varistor voltage should be tested against the minimum limits for the model using the
method described in section 7.3.1. If the varistor is open, short, or more than 10% outside either limit, it
should be discarded. The dc standby current also should be measured. If more than twice the
specification, the varistor is significantly degraded and should be discarded. If the varistor is physically
sound and shows no evidence of degradation in these electrical tests, it is fully functional.
102
7.3
MEASUREMENT OF VARISTOR CHARACTERISTICS l
7.3.1 Nominal Varistor Voltage VN
This is measured at a dc test current, IN, of ImA for product models. A simplified circuit for
instrumenting this test, shown in Figure 7.1, is suitable for varistors up through a rating of 300V RMS.
Above the 300V RMS rating, a higher supply voltage will be needed. Resistor Rl has a dual purpose. In
conjunction with the variable voltage supply, E 1, it forms a quasi-current source providing up to 6mA
when switch SI is closed. Also, Rl is used as a current sensor to measure current flowing through the
varistor-under-test. To use the circuit, the operator places switch S2 in position I and S3 into position
_.--+--+-0+
DVM
Rl=100kn. .• l %.1W(VN TEST)
R2=1 ".0.1. 1%.1I2W(IO tEST)
FIGURE 7.1: SIMPLIFIED CIRCUIT FOR VARISTOR
VOLTAGE AND DC STANDBY CURRENT TESTS
V N' A test device is then inserted into the socket and S 1 is closed. El is then adjusted to obtain a reading
of 100 ± 5V on the digital voltmeter. Approximately ImA of current will be flowing in Rl. When
switch S2 is placed in position V, the varistor voltage will be indicated on the voltmeter. The values of
Rl and El supply voltage can be scaled appropriately for other voltage-current test points.
If the varistor voltage test is implemented on automatic test equipment, a "soak" time of 20ms
minimum should be allowed after application of test current before voltage measurement. This is
necessary to allow varistor voltage to settle toward a steady-state value. Figure 7.2 illustrates the
time response of a specimen varistor with a constant 1.0mA current applied. As can be seen, the
varistor voltage initially may rise to a value up to 6% greater than final. With a 20:rp.s or greater soak
time, the measured value will differ by less than 2% from the steady-state value.
For varistor models that are commonly used on 60 Hz power lines, the V Nlimits may be specified
for a 1.0mA peak ac current applied. If an ac test is preferred by the user, a schematic approach similar to
that shown in Figure 7.1 is used, except an ac Variac™ is substituted for the dc power supply, and an
oscilloscope is substituted for the voltmeter. This circuit is equivalent to that of a typical curve tracer
instrument.
v (t)
5V/DIV
250V~--
/o.lms/DIV
Ims/DIV
240V
~--
230V
~--
~ IOms/DIV
~Iooms/DIV
~I,oooms/DIV
FIGURE 7.2: VOLTAGE-TIME VITI CHARACTERISTICS OF A GE-MOV®II VARISTOR (V130LA10A)
OPERATING AT A CONSTANT DC CURRENT OF 1.0mA
103
To avoid unnecessary concern over minor measurement anomalies, three behavioral phenomena
of metal oxide varistors should be noted. First, it is normal for the peak varistor voltage measured
with ac current to be about 2% to 5% higher than the dc value, as illustrated by Figure 7.3. This
"'ac-dc difference" 'is to be expected, since the one-quarter cycle period of a 60 Hz wave is much
less than the 20ms minimum settling time required for dc readout.
... ~ I--'"
en
(60HZ
~!IOO
~
I
W
~o
I
I
>
I
10
-
10
7
130 V RMS RATED
PRODUCT~
/
MEDIUM-VOLTAGE
rTERIT~
,/
10- 6
10- 0
CURRENT - I
10- 4
10-:>
AMPERES
10- 2
10-1
FIGURE 7.3: AC AND-DC CHARACTE.RISTIC CURVES
Second, it is normal for the varistor voltage to increasetslightly when first subjected to electrical current, as shown in Figure 7.4 This might be ,considered a "break-in" stabilization of the
varistor characteristics. During normal measurement the voltage shift typically is less than I %. This
voltage shift is of little consequence for most measurement pUrPoses but might be noticeable when
viewing a DVM as in the test method of Figure 7.1. The visual DVM observation should be made
shortly after power is applied, with measurement to not more than three significant figures.
v(t)
5V/OIV
- - -•• t,50ms/OlV
FI~URE 7.4: (V130LA10) VARISTOR VOLTAGE FOR THE INITIAL CYCLES OF 60 HZ
OPERATION AT A PEAK CURRENT OF 1.0 mA
Third, it is normal for the varistor voltage-current characteristic to become slightlyasymmetrical in polarity under application of dc electrical stress over time. The varistor voltage will increase
in the same direction as the polarity of stress, while it will be constant or will decrease in the
opposite polarity. This effect will be most noticeable for a varistor that has been subjected to unipolar pulse stresses or accelerated de life tests. Therefore, to obtain consistent results during
104
unipolar pulse or operating life tests, it is essential to provide a polarity identification for the test
specimens. However, for initial readout purposes, this effect usually is insignificant.
7.3.2 Maximum Clamping Voltage, Vc
As discussed in Chapter 3, the clamping voltage of a varistor is best defined in terms of the
current impulse impressed on the varistor, rather than in terms of applied voltage. Two typical current impulses that may be used to define the varistor clamping voltage are the 8 / 20ILs and the
10 / 1000ILS pulses. Figure 7.5 shows typical varistor test waveforms for these two impulses.
The clamping voltage of a given model varistor at a defined current is related by a factor of the
varistor voltage. Therefore, a test of the nominal varistor voltage against specifications may be sufficient to provide reasonable assurance that the maximum clamping voltage specification is also satisfied. When it is necessary to perform the Vc test, special surge generators are required. For shorter
impulses than 8 / 20ILS, precautions must be observed to avoid an erroneous "overshoot" in the
measurement of the clamping voltage. Section 7.6 gives general information on surge generators;
a brief description of the' 'overshoot" effect follows.
,
500mV
500mV
±.
I~"""
,j
., ... ... ... .... .. ..... . ...
l'
I'
.. . .
. ...
~
~
~,
....
.
~,
- ~~
,..
101-'5
IU
. ..
~...
~
a) 8/20lLs wave I.
'
-
= 50A. V. = 315V
::J'-E
-
I
1mS/dlv.
10fL& /div.
b) 10 /1000lLs wave I. = 50A. Vc = 315V
FIGURE 7.5; TYPICAL CLAMPING VOLTAGE TEST WAVEFORMS
IGE-MOV®II VARISTOR TYPE V130LA10A)
The GE-MOV®II Varistor specification sheets show the V-I characteristic of the devices on the
basis of maximum voltage appearing across the device during a current pulse of 8 / 20ILs. If current
impulses of equal magnitude but faster rise are applied to the varistor, higher voltages will appear across
the device. These higher voltages, described as "overshoot," are partially the result of an intrinsic
increase in the varistor voltage, but mostly of the inductive effect of the unavoidable lead length.
Therefore, as some applications may require current impulses of shorter rise time than the conventional
8ILs, careful attention is required to recognize the contribution of the voltage associated with lead
inductance.'
The varistor voltage, because of its nonlinearity, increases only slightly as the current amplitude of the impulse increases. The voltage from the lead inductance is strictly Ilnear and therefore
becomes large as high current amplitudes with steep fronts are applied. For that reason, it is impractical to specify clamping voltages achieved by lead-mounted devices with current impulses having
rise times shorter than 0.5 p'S, unless circuit geometry is very accurately controlled and described.
105
AREA ~ 22cm 2
AREA
OUTPUT LEAD FROM
TRANSIENT GENERATOR
Rl
0.5 cm 2
,---<~-
COPPER TUBE
SURROUNDING
VOLTAGE PROBE
CURRENT
PATH
--GROUND
CURRENT
PATH
a) Minimal Loop Area
b) Excessive. Loop Area
c) Current Rise of 8lts
d) Current Rise of 0.51'.
FIGURE 7.6: EFFECT OF LEAD LENGTH ON "OVERSHOOT"
To illustrate the effect of lead length on the "overshoot," two measurement arrangements
were used. As shown in Figures 7.6a and 7.6b, respectively, 0.5 cm 2 and 22 cm 2 of area were enclosed by the leads of the varistor and of the voltage probe.
The corresponding voltage measurements are shown· in the oscillograms of Figures 7.6c and
7.6d.With a slow current front of 8J.(s, there is little difference in the voltages occurring with a
small or large loop area, even with a peak current of 2.7 kA. With the steep front of 0.5 J.(S, the peak
voltage recorded with the large loop is nearly twice the voltage of the small loop. (Note on Figure
7.6d that at the current peak, Ldi/dt = 0, and the two voltage readings are equal; before the peak,
L di/dt is positive, and after, it is negative.)
Hence; when making measurements as well as when designing a circuit for a protection scheme',
it is essential to be alert to the effects of lead length (or more accurately of loop area) for connecting the varisto~. This is especially important when the currents are in excess of a few amperes with
rise times of less than I J.(S.
.
With reasonable care in maintaining short leads, as shown in Figure 7.6a, it is possible to describe the "overshoot" effect as an increase in clamping voltage relative to the value observed with
106
1000
~DELJJJJJ
U
2
800 ~'LEAD AREA
1/31':\
8/20
o
I
-I-
~ 400
~
«
~
g
(!)
z 200
ii:
~
«
..J
u
10
20
40
eo
60
100
200
400
600 800 1000
2000
PEAK CURRENT -AMPERES
FIGURE 7.7: TYPICAL "OVERSHOOT" OF LEAD·MOUNTED VARISTOR WITH STEEP CURRENT IMPULSES
a 8 / 20",s impulse. 'Figure 7.7 shows a family of curves indicating the effect between 8 and 0.5",s
rise times, at current peaks ranging from 20 to 2000A. Any increase in the lead length, or area
enclo~ed by the leads, would produce an increase in the voltage appearing across the varistor terminals - that is, the voltage applied to the protected load.
7.3.3 DC Standby Current, ID
This current is measured with a voltage equal to the rated continuous dc voltage, V m(dc) ,
applied across the varistor. The circuit Of Figure 7.1 is applicable where current sensing resistor'R2
has a value of 1000 n. The test method is to set the voltaJ;?;e supply, E I, to the specified value with
switch Sl closed and S2 in the V. position. ThenS2'is placed in position I and S3 in position, ID • Sl
is then opened, the test device is inserted in the test socket, and Sl is closed. The DVM reading
must be converted into current. For example, if a maximum standby current of 200J,LA is specified, ,
the maximum acceptable DVM reading would be 0.200V.
The measurement of dc standby current can be sensitive to the device behavioral phenomena
of "break-in" stabilization and polarization of the V-I characteristics, as described in Section 7.3.1.
If the device under test has prior unipolar electrical history, polarity indicators should be observed
and test values interpreted accordingly.
The value of dc standby current also can be sensitive to ambient temperature. This is unlike
varistor characteristics measured at currents of lmA or greater, which are relatively insensitive to
ambient temperatures. With VM(dc) around 85% of VN , Figure 7.8 shows the typical dc standby
100
80 1--- ~-
--- -- .- -- ,
-- --
I-.
/
~ ~
'fI'/ ~ / ' /
/
/
If
'1'/V /
VI II
" I, I VI
/ 'II '/ I /
If
n,
/J VL 1
10 -9
I {rOVi5
10-7
125°C
SPECIM
10- 6 - 1 0 - 5
r:( r
30LA A
10- 4
10-2
VARISTOR CURRENT (AMPERES, DC)
FIGURE 7.8: TYPICAL TEMPERATURE DEPENDENCE OF DC STANDBY CURRENT
VARISTOR TYPE - V130LA10A
107
current of a model Vl30LAIOA varistor in the order of 10 or 20~A at room temperature. In increases to about 80tLA at 85°C, the maximum operating temperature without derating .
.7.3.4 Capacitance
Since the bulk region ofa GE-MOV®II Varistor acts as a dielectric, the device has a capacitance
that depends directly on its area and varies inversely with its thickness. Therefore, the capacitance of a
GE-MOV®II Varistor is a function of its voltage and energy ratings. The voltage rating is determined by
device thickness, and the energy rating is directly proportional to volume.
.
GE-MOV®II Varistor capacitance can be measured through use of a conventional capacitance
bridge and is found to vary with frequency, as shown in Figure 7.9. Typically, capacitance
measurements are made at 1 MHz. Dissipation factor alSQ is frequency-dependent, as shown in Figure
7.10.
.
1400
a.t-o
1200
-~~
U.
Q.
~
UJ
U
Z
~
--
1000
U
~
800
600
104
10 5
FREQUENCY
(Hz)
FIGURE 7.9: CAPACITANCE VARIATION WITH FREQUENCY
0.12
~,~
o 0.10
ri
~ 0.08
-
u
~
i'\
i1i1
z 0.06
o
~0.04
'~~
J
in
en
00.02
'oQ",
....
~
r"""-o kl.
"
x
~,
.DI
o
10
10 3
10 4
10 5
FREQUENCY
(Hz)
FIGURE 7.10: DISSIPATION FACTOR VARIATION WITH FREQUENCY
. When measured with a dc bias, the capacitance and dissipation factor show little change until the
bias approaches or exceeds the VN value. Furthermore, the capacitance change caused by an applied
voltage. (either dc or ac) may persist when the voltage is removed, with the capacitance gradually
returning to the prebias value. Because of this phenomenon, it is important that the electrical history of a
GE-MOV®II Varistor be known when measuring capacitance.
108
7.3.5 Miscellaneous Characteristics
A number of characteristic measurements can be derived from the basic measurements already
described, including the nonlinear exponent (alpha), static resistance, dynamic impedance, and voltage
clamping ratio. These characteristics are derived characteristics in the sense that they are found by
computation per the defining equations given in Chapter 3. The data, however, may be obtained by
measurement methods similar to those already given for nominal varistor voltage and maximum
clamping voltage. These miscellaneous characteristics may be useful in some cases to enable
comparison of GE-MOV®II Varistors with other types of nonlinear devices, such as those based on
silicon carbide, selenium rectifier, or zener diode technologies.
7.4
VARISTOR RATING ASSURANCE TESTS
7.4.1 Continuous RatedRMS and DC Voltage [Vm(ac) and Vm(dc)]
These are established· on the basis of operating life tests conducted at the maximum rated
voltage for the product model. These tests usually are conducted at the maximum rated ambient
operating temperature, or higher, so as to accelerate device aging. Some test results are given in
Chapter 8. Unless otherwise specified, end-of-lifetime is defined as a degradation failure equivalent
to a VN shift in excess of ± 10% of the initial value. At this point the device is still continuing to
function. However, the varistor will no longer meet the original specifications.
--l
I
VI30LAIOA
I
I
-.J
I
I
I
L ______ ~~_~f! _____
I
FIGURE 7.11: SIMPLIFIED OPERATING LIFE TEST CIRCUIT
A typical operating life test circuit is shown in Figure 7.11. If the varistor is intended principally for a dc voltage application, then the ac power source should be changed to dc. It is desirable
to fuse the varistors individually so testing is not interrupted on other devices if a fuse should
blow. The voltage sources should be regulated to an accuracy of ±2% and the test chamber
temperature should be regulated to within ±3°C. The chamber should contain an air circulation fan
to assure a uniform temperature throughout its interior. The varistors should receive an initial
readout of characteristics at room ambient temperature - i.e., 25 ±3°C. They should then be
removed from the chamber for subsequent readout at 168,500, and 1000 hours. A minimum of 20
minutes should be allowed before readout to ensure that the devices have cooled off to the room'
ambient temperature~
7.4.2 Transient Peak Current, Energy, Pulse Rating, and Power Dissipation Ratings
Special surge generator equipment is required for testing. Data on commercially available
equipment is given in Table 7.3, and. an example test circuit is described in Section 7.6. Since
high energy must be stored at high voltages to perform these tests, especially on larger sizes'of
GE-MOV®II Varistors, the equipment is necessarily expensive and must be operated using adequate
safety precautions.
109
The peak current rating, Itm , of GE-MOV®Il Varistors is based on an 8 / 20JLs test impulse
waveshape. The specifications include a maximum single value in the ratings table. A' pulse rating
graph defines the peak current rating for longer impulse duration as well, such as for a 10 / 1000JLs
wave. A family of curves defines the rated number of impulses with a given impulse duration and peak
current.
Energy rating, W tm , is defined for a 10 /1000JLs current impulse test wave. This waveshape has
been chosen as being the best standard wave for tests where impulse energy, rather than peak current, is
of application concern. A direct determination of energy requires thatthe user integrate over time the
product of instantaneous voltage and current. Such integration is cumbersome to perform, and the
integration feature is not generally available in surge generation equipment.
However, peak voltage and current are readily measured with available equipment. Therefore, the
energy rating can be tested indirectly by applying the rated peak impulse current of a 10 / 1000""s
waveshape to the test specimen. Then, the energy dissipated in the va,ristor can be estimated from the
known pulse waveshape. For a 10 / 1000JLs waveshape the approximate energy is given by the
expression E = 1.4 Vel 7. See Chapter 4 for a discussion of energy dissipation for various waveshapes.
For example, a model V130LAlO varistor has a single pulse rating for a 10 / IOOOJLs impulse
waveshape of about 75A peak, and a maximum clamping voltage at 75A of about 360V. Thus, the
computation of estimated energy dissipation is 38J.
The transient power dissipation rating, Ptam' is defined as the maximum average power of test
impulses occurring at a specified periodic rate. It is computed as the estimated energy dissipation
divided by the test pulse period. Therefore, varistors can be tested against this rating by applying two or
more impulses at rated current with a specified period between pulses. For example, a model
V130LAlOA varistor has a pulse rating of two 10 / 1000JLs test impulses with a peak current of about
65 A. The estimated energy dissipation per pulse computed as per the preceding example is about 30J.
If a period of 50s is allowed after the first test pulse, the estimated average power dissipation can be
computed as about 0.6W, which is the specification rating. It should be noted that GE-MOV®II Varistors are not rated for continuous operation with high-level transients applied. The transient power
dissipation rating is based on a finite number of pulses, and the pulse rating of the varistor must be
observed. See Figure 7.12
10A/div.
100 V/div.
a
1 mS/div.
FIGURE 7.12: SURGE TEST WAVEFORMS
'10/1000~sWAVEFOAM
110
Table 7.1 outlines a suggested program of testing to verify varistor transient and pulse ratings with a
minimum of expensive, time-consuming testing. New specimens should be used for each test level and
failure judged according to the specification criteria.
TABLE 7.1
TESTING OF TRANSIENT CURRENT, ENERGY,
PULSE RATING, AND POWER DISSIPATION RATINGS
NO. PULSES@ RATED CURRENT
(ALTERNATING POLARITY)
TEST PARAMETER
1
(same polarity as readout)
Maximum Peak Current
TEST
WAVESHAPE (}ls)
MINIMUM PULSE
PERIOD (5)
8/20
NA
Pulse/Energy Rating, Power
Dissipation
2
10/ 1000
50
Pulse Rating
10
8/20
25
Pulse Rating
100
8/20
12
7.4.3 Continuous Power Dissipation
Since GE-MOV®II Varistors are used primarily for transient suppression purposes, their power
dissipation rating has been defined and tested under transient impulse conditions. If the devices are to be
applied as threshold sensors or coarse voltage regulators in low power circuits, then a dissipation test
under continuous power is more appropriate. This continuous power test will aid the user in determining
if the device is suitable for his specific application.
A circuit for continuous power dissipation testing is shown in Figure 7.13. The dc power
supply voltage should be set to a value of approximately twice the nominal varistor voltage of the
product model under test. In that case, nearly constant power dissipation is maintained in the varistor.
Since the circuit transfers nearly equal power to the series resistor and varistor-under-test, the series
resistor value is simply chosen to achieve the test design value of power dissipation. In Figure 7.13 a
nearly constant power dissipation of about O.6W is obtained.
400 V DC _
± 2% -
FIGURE 7.13:
CONSTANT POWER LIFE TEST CIRCUIT
111
7.5
MECHANICAL AND ENVIRONMENTAL lESTING OF VARISTORS
7.5.1 Introduction
Many tests have been devised to check the reliability of electronic components when subjected
to mechanical and environmental stresses. Although individual equipment makers may specify their·
own tests on component purchase documents, these tests are often based on an equivalent MIL-STD
specification. Therefore, it is convenient to summarize these tests in MIL-STD terms. Since the
ratings ofGE-MOV®I1 Varistors may vary with product series and model, the test conditions and limits
should be as specified on the applicable detail specification.
GE-MOV®I1 Varistors are available in a high reliability series. This series incorporates most
standard mechanical and environmental tests, including 100 % pre-screening and 100 % process
conditioning. Details are provided in Chapter 9.
7.5.2 UL Recognition Tests
GE-MOV®I1 Varistors have been tested by Underwriters Laboratories, Inc. (UL) and have been
recognized as across-the-line components, varistor type, UL E56529. GE-MOV®I1 Varistors are also
recognized as suppressor components to UL STDl449 per UL File E75961. The tests were designed by
UL and included discharge (withstand of charged capacitor dump), expUlsion (of complete materials),
.
life, extended life, and flammability (UL492) tests.
112
7.6
EQUIPMENT FOR VARISTOR ELECTRICAL TESTING
7.6.1 Introduction
Most tests of GE-MOV®II Varistors can be performed with relatively simple circuits and
inexpensive equipment on the laboratory bench. However, large users with versatile automatic test
systems available may find it more economical to program these systems for the low-current varistor
tests. As noted previously, medium or high-current impulse testing will require specialized test
equipment. Table 7.3 is a partial listing of available test equipment and systems that can be used for
varistor testing. It is intended as a guide only to illustrate the generic type of equipment offered
commercially.
7.6.2 Impulse Generators
A convenient method of generating current or voltage surges consists of slowly storing energy
in a capacitor network and abruptly discharging it into the test varistor. Possible energy storage
elements that can be used for this purpose include lines (lumped or distributed) and simple capacitors, depending on the waveshape desired for the test. Figure 7.14 shows a simplified schematic for
the basic elements of an impulse generator.
"T
R2
L
RI
VARISTOR
UNDER
TEST
OSCILLOSCOPE
vO
COM
I
R3
fIGURE 7.14: SIMPLIFIED CIRCUIT OF SURGE IMPULSE GENERATOR
The circuit is representative of the type used to generate exponentially decaying waves. The
voltage supply, El, is used to charge the energy storage capacitor, C, to the specified open-circuit
voltage when switch Sl is closed. When switch S2 (an ignitron or a triggen~d gap) is closed, the
capacitor, C, discharges through the waveshaping elements of the circuit into the suppressor device
under test. With capacitances in the order of 1tlF to IOtlF and charging voltages of lOkV to 20kV,
the typical 8 / 20ILS or 10 / IOOOILS impulses can be obtained by suitable adjustment to the wave shaping
components L, R I , and Rz, according to conventional surge generator design. 2 ,3,4,5
7.6.3 Measurement Instrumentation
Transient measurements include two aspects of varistor application: (I) detection of transients to determine the need for protection, and (2) laboratory measurements to evaluate varistor
performance. Transient detection can be limited to recording the occurrence of transient overvoltages ina particular system or involve comprehensive measurements of all the parameters which can
be identified. Simple detection can be performed with peak-indicating or peak-recording instruments, either commercial or custom-made. Table 7.4 gives a partial listing of such instruments.
Laboratory instruments and field detection with comprehensive instrumentation can involve
substantial investment, primarily associated with oscilloscopes, cameras, and calibrated sensors. A
detailed discussion of these systems is beyond the scope of this manual; rather, the major oscilloscope manufacturers should be consulted, as well as the available literature.
113
TABLE 7.3 - AVAILABLE EQUIPMENT FOR VARISTOR TESTING* .
TYPE & MANUFACTURER
MODEL
FEATURES
High Voltage Test Systems
E. Haefely and Co., Ltd.
Basel, Switzerland
U.S. pistributor:
Rhode & Schwarz Sales Inc.
14 Gloria Lane
Fairfield, NJ 07006
Storage Curve Tracer
P12/P6R
7CL-6/P6T
P351CP70
CP360
CP1500
Voltage Surges: 1.2/50,0.51700, 101700,
1001700, 0.51100kHz and NEMP up to
35kV
Current Surges: 8/20,20/60, 101l0OOjts up i
t050kA
,
Tektronix, Inc.
P.O. Box 500
Beaverton, OR 97005
315-455 c 6661
5771177
(Also can
use 576)
AC & dc tests up to 1600V peak, with safety
interlock, storage display mode
1687
1MHz test frequency, 3 measurementsl sec,
.01 % accuracy, digital display,
programmable control, IEEE testing.
222F
342
0-2000V, up to lOrnA dc, 100 rnA pulse,
digital readout, front panel programming
O-lOOkVat lOA. Frontpanel programming.
IEEE Standard 4888
T57
or
Z27
0-1200V, up to lOrnA, computer operated,
line printer output, multiple test stations,
data analysis software, tape cartridge
605P
2.2kV, 20kW peak power, pulses 0.3jts lOms, variable PRF or can amplify
external input
Auto Capacitance Bridge
General Radio
300 Baker Ave.
Concord, MA 01742
617-369-4400
Varistor Test System
Mastech, Inc.
478 E.Brighton Ave.
Syracuse, NY 13210
315-478-3133
Semiconductor Test Systems
Teradyne, Inc.
183 Essex St.
Boston, MA 02111
201-334-9470
Pulse Generator
Cober Electronics, Inc.
7 Gleason Ave.
Stamford, CT 06902
703-327-0003
Surge Generator & Monitor
KeyTek Instrument
220 Grove St.
Box 109
Waltham, MA 02154
617-272-5170
System
1000,
including
Model 424
Model 711
6kV, up to 5000A, further expandable,
selectable waveshapes (8 I 20, 10 I 1000,
etc.), measures & displays peak V & I
across test device.
Peale biased differential high voltage probe.
IEEE testing.
360
2.5kV, lOA, pulses, up to 300jts wide,
variable PRF, variable rise-fall available,
plug-ins for higher peak I
Pulse Generator
Velonex '
Varian Div.
560 Robert Ave.
Santa Clara, CA 95050
408-727-7370
*Inclusion of any manufacturer in this listing does not
cOIistitute an endorsement nor does exclusion imply
any judgment upon same.
114
TABLE 7.4 - AVAILABLE TRANSIENT DETECTION EQUIPMENT*
MANUFACTURER
MODEL
FEATURES
Storage Oscilloscopes:
Tektronix
P.O. Box 500
Beaverton, OR 97005
466
7834
100MHz, 3000 divl J.LS speed, portable
Multimode storage, 400MHz, 5500
div/J.Ls
Peak Recording Instruments:
7.7
Micro-Instrument
Memory
Voltmeter
Model 5203
20MHz, records, displays voltage levels up
t02kV
Bermar
Box 1043
Nashua, NH 03061
603-888-1300
Memory
Voltmeter
MVM-I08
Displays peak voltage,
8kV
Dranetz
2285 So. Clinton Ave.
So. Plainfield, NJ 07080
201-287-3680
Model 606
Prints out peak voltages,
pulse
KeyTek
Box 109
Waltham, MA 02154
617-272-5170
Industronic
115 Pleasant St.
Mellas, MA 02054
617-376-8146
424 Surge
Generator &
Monitor
Displays peak voltages,
> 0.5J.Ls duration
Zap Trap
Measures peak voltages,
> 2J.Ls duration
Trott
9020 Wehrle Dr.
Clarence, NY 14031
413-634-8500
TR745A
Detects 0.3J.Ls pulses, up to 3000V
> 0.5J.Ls pulses up to
> 0.5J.Ls duration
TEST WAVES AND STANDARDS
The varistor test procedures described in this chapter have been established to ensure conformity with applicable standards,6 as well as to reflect the electromagnetic environment of actual
circuits 7 which need transient protection. Chapter I presented an overview of the transient environment; some additional background is presented in this section concerning generally accepted
assumptions about this environment.
7. 7.1 Test Waves
A number of test waves have been proposed, to be applied to various electronic "black boxes,"
in order to demonstrate capability of survival or unimpeded performance in the environment. Table
7.5 is a partial listing of these test waves presented to illustrate the variety of proposals rather than
to be an exhaustive listing.
.
115
TABLE 7.5 -
PARTIAL LISTING O,F EXISTING OR PROPOSED TEST WAVES
DESCRIPTION
ORIGIN
ANSI,IEC.
IEEE Std. 472
Guide for Surge Withstand
Capability (SWC)
ANSI/IEEE Std. C62.41-1980
Guide on Surge Voltage in
Low Voltage ac Power Circuits
Ground Fault
Interrupters
WAVESHAPES
•
•
•
•
•
AMPLITUDE
TYPICAL
APPLICATION
Power apparatus
1.2/50p,
8.0 I 20p,s
Specified voltage
Specified Current
1.2SMHz repetitive
at 60Hz
61J.S decay to 50%
lS0n source
impedance
2.5kV Peak
Low-voltage ac circuits and
control lines in substation
equipment·
Dependent on
location
Low-voltage ac circuits and
signal lines .
3kVand6kV
High impedance circuit of
ground fault interrupters.
50 to SOOA
5 to 20kA
Telephone protectors
- 100kHz
• .5p,s
1.2
I
50p,s voltage
• 8 I 20t-is
current
•
rise
• O.SJiS
100kHz
• 2nd peakring60% first
• son source
• impedance
~
IEEE Std. 465.1
Test Specifications for Gas
Tube Surge Protective Devices
FCC Docket 19528
Three requirements:
10 I 1000p,s current
81 20p,s current
Linear voltage ramp
of 100, 500,
5000, 10,000
V/JiS until sparkover
•
•
•
• Metallic
- 10 I 560p,s
- 100A short-circuit current
Longitudinal
- 10 I 160p,s
- 200A short-circuit current
•
FCC Section 68.302
Title 47, Telecommunications
• -2/1Op,s
1000 A short-
800VPeak
Communications
equipment
lS00VPeak
2S00VPeak
Line-powered communication equipment
3a ofProtector
level
Telephone electronics
0.1 to 1000A
Evaluation of components
circuit capability
Rural Electrification Admimstration Spec. PE-60
Nuclear Electromagnetic
Pulse (NEMP)
1000p,s voltage
• 110/
OOV
I p,s rise
•
pulse
• Rectangular
3 nsio lOps
sinewave
• Damped
10 to 10 Hz
sinewave
•, DalJlped
125 kHz
• Unidirectional
- 21 100p,s
1
NASA Space Shuttle
- 3001 600p,s
MIL-STD-704
1.0 to 100A
3
specified,
• Envelope
max. duration
SOps
11tl
Eoc - SOV
Isc - lOA
Eoc - SOV
Isc - lOA
Eoc -O.5V
SA
Isc -
Space Shuttle electronics
600V Peak.
Military aircraft power
A proposal also has been made to promote a transient control level conceptS whereby a few
selected test waves could be chosen by common agreement between users and manufacturers. The
intent being that standard test waves would establish certain performance criteria for electronic
circuits, without resorting to a multiplicity of test waves, each attempting to simulate a particular
environment.
7. 7. 2 Source Impedance
The effective impedance of the circuit which introduces the transient is an extremely important parameter in designing a protective scheme. Impedance determines the energy and currenthandling requirements of the protective device.
Historically, the approach to transient withstand capability was to apply a voltage wave to
a device and to ascertain that no breakdown occurred. Typically, the device offered high impedance to the impulse, so that no significant current would flow (unless breakdown occurred), and the
source impedance was unimportant. But if a transient suppressor is applied, especially a suppressor
of the energy-absorbing type, the transient energy is then shared by the suppressor and the rest of
the circuit, which can be described as the "source."
As in the case of waveshapes, various proposals have been made for standardizing source impedances. The following list summarizes the various proposals intended for ac power lines:
1.
2.
3.
The Surge Withstand Capability (SWC) standard specified alSO n source.
The Ground Fault (UL-GFCl) standard is son source. l l
The Transient Control Level (TCL) proposals of Martzloff et al? include a 50 n resistor
in parallel with a 50 JlH inductor.
4.
The installation category concept of ANSI/IEEE Standard C62.41-1980 implies a range of
impedances from 1 to 500 as the location goes from outside to inside.
5.
The FCC regulation for line-connected telecommunication equipment implies a 2.50 source
impedance. However, the requirement of the FCC is aimed at ensuring a permanent
"burning" of a dielectric puncture and does not necessarily imply that the actual source
impedance in the real circuits is 2.50.
6.
There is no agreement among the above proposals on a specific source impedance.
Examining the numbers closer, one can observe that there is a variance between 2.5 ohms to
about 50 ohms. Going back to ANSI/IEEE Standard C62.41-1980 - by using the OCV (open
circuit voltage) and SCI (short circuit current) for the different location categories, one can
calculate a source impedance.
Any practical power circuit will always have some finite impedance due to the resistance and
inductance of the power line and distribution transformer. Figure 7.15 shows representations of the
impedance between line and ground on typical 120 V and 220 V systems in residential systems.
The impedance of industrial or commercial systems generally. supplied bv underground entrances, or a separate substation of relatively large kVA rating, tends to be low, and the injection
of any lightning transients occurs at a remote point. This results in lower transient peaks than those
that can be expected in residential circuits, but the energy involved may be, in fact, greater. Therefore, transient suppressors intended for industrial use should have greater energy-handling capability
than the suppressors recommended for line-cord-powered appliances.
Clearly, the industry standards have not been able to agree on a single value of the source
impedance, for several reasons. When a transient suppressor is being selected for a particular
application, there is a need for engineering judgment based on a knowledge of the function, and the
capability of the device.
117
6kV / 200A = 300
6kV / 500A = 120
6kV / 3kA = 20
CATEGORY A RING WAVE
CATEGORY B RING WAVE
CATEGORY B IMPULSE
CATEGORY C IMPULSE
lOkV / lOkA = 10
FIGURE 7.15: SOURCE IMPEDANCE AT DIFFERENT LOCATION CATEGORIES
IN LOW VOLTAGE AC SYSTEMS (TO 600VI.
REFERENCES
1. Fisher, F.A., "Overshoot - A Lead Effect in Varistor Characteristics," Report 78CRD, General
Electric, Schenectady, N. Y ., 1978.
2. Heller, B. and A. Veverka, Surge Phenomena in Electrical Machine, ILIFFE Books Ltd. , London,
1968.
3. Greenwood, Allen, Electrical Transients in Power Systems, Wiley Interscience, New York, 1971.
4. Craggs, J.D. and J.M. Meek, High Voltage Laboratory Techniques, Buttersworth Scientific
Publications, London, 1954.
5. Martzloff, F.D., "Transient Control Level Test Generators," Report 77CRD241, General
Electric, Schenectady, N.Y., 1977.
6. Test Specifications for Varistor Surge-Protective Devices, IEEE Std. C62.33, 1982
7. Martzloff, F.D., and F.A. Fisher, "Transient Control Level' Philosophy and Implementation The Reasoning Behind the Philosophy," 77CHI224-5EMC, Proceedings of the 2nd Symposium on EMC, Montreux, June 1977.
8. "Standard for Safety: Ground Fault Circuit Irtterrupters," UL943, Underwriters Laboratories,
May 12, 1976.
9. "Longitudinal Voltage Surge Test #3," Code of Federal Regulations, Section 68.302(e), Title
47, Telecommunications.
118
.@}UID~
®
VARISTOR RELIABILITY
8.1
INTRODUCTION
The GE-MOV®II Varistor is a rugged, reliable voltage transient suppressor designed to improve the
reliability of electronic systems. Proper system design with the GE-MOV®II Varistor, as detailed in other
parts of this manual, will clamp transient voltages to a level compatible with long-life of the electronic
system. To assure GE-MOV®II Varistor reliability, General Electric performs extensive process and quality
control monitoring. This is accomplished via a combination of 100%, periodic, and lot testing. Both
parametric and reliability characteristics are controlled in this manner.
Every year, over 1.7 million device hours of maximum rating and accelerated reliability test data is
accumulated. In addition, a continuous product improvement research program is in effect to provide the
user with an optimal product. As a result of these programs, extensive reliability data and reliability
prediction models have been generated.
Two types of GE-MOV®II Varistors are being manufactured at present; a "line voltage" type (above 100V
RMS) and a "low voltage" type (below 100V RMS). Reliability evaluation has been conducted on both
types under the conditions summarized below:
Test Condition
Stress
Voltage ........................... aC'Bias
dc Power
Temperature. . . . . . . . . . . . . . . . . . . . . 100°C
12SoC
14SoC
Energy . . . '. . . . . . . . . . . . . . . . . . . . . . Pulse
Storage . . . . . . . . . . . . . . . . . . . . . . . . 12SoC
Humidity ................... .40 o C, 95% RH
Mechanical .................... Solderability
Terminal Strength
Shock
Vibration
As improved products, processes, and test procedures evolve, the applicability of past data to
reliability assessment changes. Thus, the data presented in this chapter represents a "snapshot in
time" Of data applicable to the GE-MOV®II Varistors being manufactured now and for the anticipated
future. The test data has been generated at very high stress levels, at or beyond maximum ratings, to confirm
the product's ability to meet these ratings and to obtain the most information in the shortest time period.
Results of ac voltage and dc power bias tests have allowed the generation of models which provide expected
life as a function of stress.
8.2 AC BIAS RELIABILITY
The majority of the applications for the GE-MOV®II Varistor are as transient suppressors on the ac1ine.
The varistor is connected across the ac line voltage and biased with a constant amplitude sinusoidal voltage.
If the varistor current increases with time, the power dissipation will also increase, with the ultimate
possibility of thermal runaway and varistor failure. Because of this possibility, an extensive series 'of
statistically designed tests have been performed to determine the reliability of the GE-MOV®II Varistor
119
, under ac bias combined with temperature stress. This test series contained over one million device hours of
operation at temperatures up to 145°C. The average duration of testing ranges from 7000 hours at low stress
to 495 hours at high stress. The results of this test have shown the GE-MOV®II Varistor to be an excellent fit
to the Arrhenius model, i.e., the expected life is logarithmically related to the inverse of the absolute
temperature (MTBF = e + 10). The definition of failure is a·shift in VN exceeding ± 10%. Although the
GE-MOV®II Varistor is still functioning normally after this magnitude of shift, devices at the lower extreme
ofVN tolerance will begin to dissipate more power. As previously explained, this could ultimately lead to
failure. This choice of failure definition, in combination with the lower stresses found in applications,
should provide life estimates adequate for most design requirements. Figure 8.1 illustrates the Arrhenius
model plot for the line voltage and the low voltage GE-MOV®II Varistor.
C
10 8
0
v
'"w::IE 104
z
'"::IEw
DEVICE TYPE: VI30LA
10 3
~V
....... ~
.....
~~~~:
10 6
w
w
":::;
i"
80% MAX VOLTAGE STRESS
":::; 105
~
10
5
1
..... .'" ~ .......
~~
V
~V
.....
V
AMBIENT TEMPERATURE IN "C
lal Line Voltage GE·MOV
Q
~ 8
....
z
'"'"
o
~
'"
PERCENTILES
6
C>
~
4
l,/
i~
~
PERCENTILES
L~
r--
o
'60
---
150
-:
-
-
-
-
r----
140
~-
90TH
50TH
10TH
PERCENTILES
10TH
120
DC POWER LIFE
Po = 0.6 WATTS AT TA = 55°C
SAMPLE SIZE n = 20
DEVICE TyrE: V24ZA4
500
100
1000
HOURS UNDER STRESS
1500
r-o
200
(a) Low Voltage Varistor
150
DC POWER LIFE
Po =0.5 WATTS AT TA = 100°C
SAMPLE SIZE n =60
DEVICE TYPE: VI30LA
I
500
1000
HOURS UNDER STRESS
1500
(b) Line Voltage Varistor
FIGURE 8.7: ACCELERATED DC POWER LIFE. V. VARIATION
With judicious derating to a modest power level, the varistor may be used at continuous power
dissipation. on a dc line. These applications are limited and highly specialized as the device is intended primarily for intermittent, transient service.
123
2000
I-
-
I-
-
III I
IARrErS(ODEL
DC ,POWER LIFE AT 06 WATTS
(MAXIMUM RATED POWER)
M.L." ,,[-6.34+[6894,5+(T";~73ll]
..,.' ,.'
Ul
~
o
106
:x:
z
;;:; 105
....
::i
z
~ 104
,.
~,
.~
,... V
.vV
V
..... ,~
, ..
rJ)
100~
50 LaJ
V'
~
~
10
LaJ
5
~
I
~
z
::E
:E
DEVICE TYPE: V24ZA4
103
FAILURE CRITERIA: IL ~ O,lmA AT 10V
1 1 1
1
1
'I
I· 1
170 150
130 120 110 100 90 80 70 60 50
180 160
140
AMBIENT TEMPERATURE IN ·C
40
30 25
FIGURE 8.8: RELIABILITY MODEL DC POWER LIFE
8.4
PULSE ENERGY CAPABILITY
The ability of the GE-MOV®II Varistor to absorb large amounts of transient energy is the key
to its utility. No other suppressor device combines equal performance with the same economic ad- I
vantage~ Pulse energy is absorbed throughout the bulk of the device. The effect of pulse stress is to
shift the low current end of the V-I characteristic as illustrated in Figure 8.9. With sufficient stress
(unipolar) the curve. will become asymmetrical as shown in Figure 8.9(a). Other forms of electrical
or temperature stress affect the low current region as well. The general response to most stress is
a shift of the low current V-I segment to the right. That is the main reason for the consistent use of the
failure definition as a change in VN of ± 10% .
LaJ
IIJ
..J
«
..J
«
g~
LaJ
(!)
o
..J
~-----./
4TlME
l&l
........
(!)
;:..J
REVERSE
o
o
>
>
CURRENT (LOG SCALE)
CURRENT (LOGSCALE)
(a) Unipolar Pulsa And de Stress Response
(b) General Response To Stress With Time
FIGURE 8.9: EFFECT OF STRESS WITH TIME ON V-I CHARACTERISTIC
At voltages above VN, little change is observed in response to pulsing or other types of
stress. The varistor will continue to provide adequate clamping protection after stressing, up to the
point of catastrophic failure. At catastrophic failure the device exhibits a short-circuit punchthrough. It takes an extremely high energy pulse to cause this type of failure which is a melting of
the ceramic body. More frequently, it is ac current from the power line that causes the pulse-weakened device to go into thermal runaway.
Voltage stability at several conditions of peak current, impulse duration, and temperature is
Summarized in Figure 8.10 for the V130LA20A model. These results are typical of the excellent
pulse capability observed in all sizes of devices. No significant difference is noted between 25°C
and 75°C testing.
124
220
MAX
MAX
x---f
0
>
~
MAX
MAX
(/)
eJ
200
'z"
0
>
1
fteol
/
X-X
IMED
x~ED
x
fEDi
I
~N
~X
MIN
1
MTN
MIN
180
NO. OF PULSES
0
10
50
0
100
200
0
I
10
0
I
10
PEAK AMPERES
25A
500A
2000A
2000A
TEMPERATURE
25°C
25°C
25°C
75°C
10/ 1000
8/20
8/20
8/20
PULSE DURATION
(fL S )
FIGURE 8.10: TYPICAL PULSE TEST STABILITY, V130LA20A MODEL
7mm
Peak Current
Temperature
Waveshape
No. of Pulses
No. Tested
No. Failed*
5A
25°C
100A
25°C
10 / 1000
8/20
400A
25°C
8/20
400A
75°C
8/20
10 150 1001200 1 110** 1 110**
10 10 10 10 10 10 10 10
0
0
0
0
1000A
25°C
8/20
1
10
0
14mm
Peak Current
Temperature
Waveshape
No. of Pulses
No. Tested
No. Failed*
25A
25°C
10 / 1000
500A
25°C
8/20
2000A
25°C
8/20
2000A
4000A
75°C
8/20
25°C
8/20
1
10
0
10 150 1001200 1 110** 1 110**
10 10 10 10 10 10 10 10
0
0
0
0
20mm
Peak Current
Temperature
Wave shape
No. of Pulses
No. Tested
No. Failed*
I
50A
1000A
4000A
25°C
25°C
25°C
10/1000 8/20
8/20
10 150 1 10 1 110**
10 10 10 I 10 10 10·
0
0
0
0 0
0
4000A
75°C
8/20
1 110**
10 10
0
0
*Catastrophic Failure Definition.
**Designates 5 times the rated value.
FIGURE 8.11: V130LA PULSE CAPABILITY SUMMARY
125
6500A
25°
8/20
1
10
0
Figure 8.11 provides further det,ails on the pulse current capability of the line voltage models
for exponential pulses (reference Figure 3.18). This chart indicates how well units survived peak
currents at rated levels of pulsing and at an over-rated condition of pulsing.
Data defining energy withstand capability is presented in Figure 8.12 for the low voltage
varistor (V68ZA types) and for the line voltage varistor (V130LA types). These curves show a
statistical estimate of the energy to failure distribution. The distributions are shown on normal
probability paper where the estimated percentiles of failure can be obtained. The surge test method
uses a quasi-current source to apply a single surge of 8 / 20",s energy stress after which the rated
continuous voltage is applied, 130V RMS for line voltage units and 40V RMS for low voltage devices.
The failure mode was a catastrophic punch-through of the ceramic body occurring after the surge
stress and during application of rated voltage. Thus, the immediate cause was thermal runaway on
rated voltage induced by overheating from surge energy absorption. A post-test readout of nonfailed devices showed no significant degradation of V-I characteristics.
-30"
-20"'
o
-10"
10"
20"
-30"
30"
300
, 250
:3250
'":> 200
200
Ir
Ir
'"
'"
~ 150
-
'"
o
o.,
o."
5
-10"
~
-20"
~8
~8
10
.01
.1
~ 1--
I-~
-- -
'"
'"'"
Z
~
14mm
tl.
~
z
iii
7mm
99
100
:>
~
10 20304050607080 90
150
99.9 99.99
CUMULATIVE PERCENT FAILURE
.....
50
40
30
20
10
.01
.........
.1
...... 1- 1-1-
..... ...-
.......
~
----
.........
14mm
7mm
10 20304050607080 90
99
99.9 99.99
CUMULATIVE PERCENT FAILURE
(b) Line Voltage Varistor (Type V130LA)
(a) Low Voltage Varistor (Type V68ZA)
FIGURE 8.12: PULSE ENERGY CAPABILITY TO SINGLE PULSE OF 8 x 201'S
The distribution curves reflect the conservatism of the GE-MOV~II Varistor energy ratings. For
example, 7mm and 14mm line voltage devices (V130LA types) are rated at 8J and 30J respectively. Figure 8.12 indicates a statistical estimate at these ener~y levels of not more than I % of the
population failing.
Pulse energy testing also has been performed at 60 Hz for single cycle and ten cycle surges.
This test simulates conditions possible in ac line applications, especially in crowbar circuits and
when used in conjunction with spark gaps to enhance tum-off. In these tests the pulse energy application is immediately followed by maximum rated ac voltage. The results also are presented on
a normal probability graph as distributions of energy vs. percent failure. Figure 8.13 illustrates low
voltage and line voltage varistor performance.
126
-30"
(!)
I
,
50
40
30
20
10
6
o
-10-
7 mm DISC DIA., PULSE WAVEFORM 60Hz ONE CYCLE
10
10
14mm
ONE" _
20mm
10" /
ONE
"
5;
'"
100
200
w
./'"
>'-'
Q
A
8
14mm
V
",/
w
d
./'
V
VV
>-
(!)
«
-20"
250
::J
50
40
30
20
V
....-:
f"'V
'"
/
........- V
..........
...... f--'i-- f-"">- ~ ~
f-""f-""
-
./'"
E-
VO
~C
B_
A_
I0
' . J ._
99.9 99.99
10 20304050607080 90
CUMULATIVE PERCENT FAILURE
.01
(al Low Voltage Varistor (Type V68ZAI
99
99.9 99.99
(bl Line Voltage Varistor (Type V130LAI
FIGURE 8.13: 60 Hz SURGE ENERGY CAPABILITY
8.5
MECHANICAL RElIABILITY AND INTEGRITY
The GE-MOV®II Varistor is constructed by encapsulating a solid piece of ceramic in a rugged,
plastic body. This rugged construction, when subjected to the normal military standard mechanical
tests, again illustrates the conservative design philosophy. Figure 8.14 tabulates the testing performed
and typical results measured. No significant differences are noted between radial and axial devices or
low voltage and line voltage types. It should also be noted that the plastic encapsulant complies with
the flammability requirement of Underwriters Laboratories standard UlA92, superseded by UL141O.
MILITARY TEST
Solderability
RESULTS
(FAI LUR ES/SAMPLE)
METHOD
CONDITION
TYPE OF PACKAGE
RADIAL
AXIAL
MIL-STD-7 5a
Method 2026.2
230°C, 5 Sec. Dip
95% Wetting
0/15
0/25
'Terminal Strength
MIL-STD-750
Method 2036.3
3 Bends, 90° C Arc
8 Oz. Weight
0/15
0/25
Thermal Shock
MIL-STD-202E
Method 107D
-55°C to 85°C
5 Cycle
0/25
0/25
Mechanical Shock
MIL-STD-7 5a
Method 20 16.2
1,500 g's
5 Drops
Xl'Yl'Zj
a/50
0/25
Vibration, Variable
Frequency
MIL-STD-750
Method 2056
20 g's
100 - 2000 Hz
Xj,Yj,Zl
a/50
0/69
FIGURE 8.14: MECHANICAL TEST RESULTS ON GE-MOV®II VARISTOR PACKAGES
127
8.6
ENVIRONMENTAL AND STORAGE RELIABILITY
The construction of the GE-MOV@II Varistor ensures stable characteristics over the wide variety
of environments in which electronic equipment is stored, shipped, and operated. Stress.testing of the
GE-MOV~II Varistor confirms the stability of low voltage and line voltage types subjected to high
temperature storage and accelerated humidity stress. Figure 8.15 presents IOOO-hour stability life data
at 125°e storage conditions.
SAMPLE SIZE n • 30
PERCENTILES
STORAGE LIFE: TA • 125"C
90TH
DEVICE TYPE: V68ZA
75
'"
~
o
>
~ 70
-$ .
--
2?0
210
50TH
PERCENTILES
90TH
10TH
50TH
65
-
180
STORAGE LIFE: TA ·125"C
SAMPLE SIZE n • 45
DEVICE TYPE: VI30Lj
10TH
170
168
500
HOURS UNDER STRESS
o
1000
168
500
HOURS UNDER STRESS
(a) Low Voltage Varistors
1000
(b) Line Voltage Varistors
FIGURE 8.15: ACCELERATED STORAGE LIFE
Figure 8.16 illustrates WOO-hour stability life during accelerated humidity testing. Note that
the low voltage varistor type has been subjected to two tests sequentially. The normal 40o e, 95%
R.H., lOOO-hour test was followed by the very severe 85°C, 95% R.H. test. Excellent stability is
observed through this combined testing sequence.
SAMPLE SIZE n I. 30
DEVICE TYPE: V68ZA
I
220
I
75 t- HUMIDITY LIFE:
HUMIDITY LIFE:
85·C, 95% R.H. FOR 1000 HOURS
40·C, 95% R.H. FOR 1000 HOURS
90TH
~
-
90TH
50TH
210
...'"<5
,1
200
>
z
50TH
168
--
10TH
65 II--
PERCENTILES
500
j
-
-if
1000
HOURS UNDER STRESS
10TH
190
~
PjRCENTILES
180
HUMIDITY LIFE: 40.c,195% R.H
SAMPLE SIZE n· 55
DEVICE TYPE: VI30LA
170
1
1000
o
500
HOURS UNDER STRESS
168
(b) Line Voltage Varistors
(a) Low Voltage Varistors
FIGURE 8.16: ACCELERATED HUMIDITY LIFE
128
1000
8.7
SAFETY
or
The GE-MOV@IIVaristor maybe used in systems where personnel safety equipment hazard
is involved. All components, Including this semiconductor device, have the potential of failing or
degrading in ways whiCh could impair the proper operation of such systems. W~ll-known circuit
techniques are available to protect against the effects of such occurrences. Examples of these techniques include fusing and self-checking. Fault analysis of any systems where safety is in question
is recommended. Potential device reaction to various environmental factors has been discussed
throughout this section. These and any other environmental factors should be analyzed in all circuit designs.
Should the varistor be subjected to surge currents and energy levels in excess of maximum ratings, it may physically f~il by package rupture or expUlsion of material. It is recommended that
protective fusing be used as described in Chapter 4. If not fused, the varistor should be located
away from other components or be l'hysically shielded from them.
GE-MOV@II Varistors have received listing under an Underwriters Laboratories standard for
"Across-The-Line Components," UL E56529, and "Component - Transient Voltage Surge Suppressors," UL E75961.
If the system analysis indicates the need for a maximum degree of reliability, it is recommended
that General Electric be contacted for a customized reliability program.
It is stressed that most GE-MOV@II Varistor parameter and reliability testing requires the use of
voltages of a magnitude that is hazardous. When GE-MOV@II Varistor testing is contemplated,
provisions must be made to insure personnel safety.
129
Notes
GE-MOV®II METAL OXIDE VARISTORS FOR
TRANSIENT VOLTAGE PROTECTION:
SPECIFICATIONS
GE-MOV®II is the latest result in metal oxide varistor technology, offering a significantly higher
energy capability and an improved voltage clamping characteristic.
GE-MOV®II Varistors are voltage dependent, symmetrical, metal oxide semiconductor devices
which perform similary to back-to-back zener diodes in circuit protectiol).. Their characteristics enable
them to protect against high transient voltage spikes (when properly selected) to meet anticipated loads.
When the protected equipment or circuit encounters high voltage spikes, the varistor impedance
changes from a very high standby value to a very low conducting value, thus clamping the transient
voltage to a protective level. The excess energy of the incoming high voltage pulse is absorbed by the
GE-MOV®II Varistor, protecting voltage sensitive components against damage.
The protection afforded by the GE-MOV®II Varistors not only guards expensive and voltage
sensitive equipment from physical damage, but also provides increased reliability in components that
can encounter temporary functional failure from transient voltage of lower amplitude.
FEATURES
• Wide Voltage/Energy Range • Excellent Clamp Ratio • Fast Response Time • Low Standby
Power. No Follow-On Current. UL Recognized
Special Products for Special Applications
MA Series
• Axial Package
• Wide Voltage Range
• Automatic Insertion
ri! ~ f'f! ~
• Radial Package
• Low Voltage Operation
P Series
---
•
•
•
•
Rigid Mountdown
NEMA Creep and Strike Distance
Quick Connect Terminal
UL Recognized
• UL Recognized
HE Series
B, 0 Series
• Isolated Baseplate
• NEMA Creep and Strike Distance
• Rigid Terminals
• Bell Package
• High Energy Capability
• Low Inductance
• Rigid Terminals
• UL Recognized
Hi Reliability Series
Due to our continuing program of
product improvement, specifications
are subject to change without notice.
• Radial Package
• Line Voltage Operation
• 100 % Prescreened
• 100 % Process Conditioning
• Meets Military Specifications
131
• Isolated
• Low Inductance
• Improved Creep and Strike
CONCEPTS OF TRANSIENT VOLTAGE PROTECTION
Varistor characteristics are measured at high current and energy levels of necessity with an inpulse
waveform. Shown below is the ANSI STD C62.1 waveshape, an exponentially decaying waveform
representative of lightning surges, and the discharge of stored energy in reactive circuits.
Based on industry practices, the 8 / 20p,scurrent wave (8p,s rise and 20p,s to 50% decay of peak
value) is used as a-standard for current (ITM) and clamp voltage (V c) ratings shown in the specification
tables and curves. Ratings for other waves of different decay times are shown specifically on the pulse
life derating curves.
For the energy rating (W TM), a longer duration waveform of 10 / 1000p,s is used. This condition is
more representative of the high energy surges usually experienced from inductive discharge of motors
and transformers. GE-MOV®II Varistors are rated for a maximum pulse energy surge that results in a
varistor voltage (V N) shift of less than ± (10 % + 1V) of initial value.
To determine the energy absorbed in a varistor the following equation applies:
where I is the peak current applied, Vc is the clamp voltage which results, T is the pulse width and K
is a contant. K values are 1.0 for a retangular wave, 1.4 for a 10 /lOOOp,s wave, and 1.0 for a 8/ 20p,s
wave.
100
~
90
~3
~~
~~
10Aldiv.
f-w
ffi~
~o
~
10tJS/div.
CJ
-----
SOURCE: ANSI
STD. C62.1-1975
I
I
I
Ww
100Vldiv.
1 ----
~
I
I
-----1------I
I
I
I
I
VIRTUAL START OF WAVE:
I
Peak Current Test Impulse Wave
S/-ts front duration I 20/-ts (impulse duration)
except as noted.
8/20!,-s Test Wave.l p ·315V
V130LA10A
Note that the rated energy (W TM) and the energy absorbed in a varistor may not be identical. A
specimen with lower clamping voltage will absorb less energy. This effect tends to be greatest at rated
peak current (ITM) with an 8 / 20p,s wave.
It is important to note, as demonstrated by the above equation, that poorer varistors must absorb
higher energy levels than the better performance varistors with lower clamp voltages, yet they actually
provi~e less over-voltage protection. For that reason, energy ratings based on an 8/ 20p,s pulse tend to
overstate varistor capability. The 10 / 1000p,s waveform consequently gives a more realistic energy
rating value.
132
SPEED OF RESPONSE
The measured response time of a varistor is influenced by lead configuration and length. In a
typical application, the response time is shorter than the inductive lead effect. In a coaxial configuration, a response time much less than one nanosecond has been shown. See figure 3.18, page 53,54.
DEFINITIONS
DEFINITION
TERM
DC VOLTAGE. V m(dc)
Maximum allowable steady state dc applied voltage. DC standby current, ID = 20l'A typical, 200l'A
maximum at TA = 25°C, except VI8ZA to V36ZA 20mm size: ID = 200l'A (TYP), 3mA max.
RMS VOLTAGE.
Vm(ac)
Maximum allowable steady state sinusoidal voltage (RMS) at 50-60Hz. If a non sinusoidal waveform is
applied, the recurrent peak voltage should be limited to H 1 Vm(ac)'
ENERGY, W tm
Maximum allowable energy for a single impulse of 10 1 lOOOl's current waveform. Energy rating based
on a V N shift of less than ± 10%, ± 1V of initial value.
PEAK CURRENT. I tm
Maximum allowable peak current for a single impulse of 8 1201's waveform with rated continuous
voltage applied. See pulse lifetime rating curves for other conditions.
VARISTOR
VOLTAGE. V N1dc )
Varistor peak terminal voltage measured with a specified current applied. For dc conditions, 1mA is
applied for a duration of 20ms to 5s. For ac conditions ImA peak 60Hz wave is applied.
CLAMPING
VOLTAGE. Vc
Maximum terminal voltage measured with an applied 8 1201's impulse of a given peak current. See V-I
curves and table for product ratings of clamping voltage over the allowable range of peak impulse
current.
CAPACITANCE
Typical values measured at a test frequency of 0.1 to 1.0MHz. Maximum capacitance is two times the
typical value measured at 1MHz.
VARISTOR SAFETY PRECAUTIONS
Should the varistor be subjected to surge currents and energy levels in excess of maximum ratings, it may physically fail by package rupture or expulsion of material. It is recommended that protective fusing be used as described in the Transient Voltage Suppression Manual, Chapter Four. If not
. fused, the varistor should be located away from other components or be physically shielded from
them.
\
133
VARISTOR PRODUCTFAMILYSELECTION GUIDE
Maximum Steady State Applied Voltage
Peak
Pull.
Current
(AmpI)
Energy
(Joule.)
DIIC Sizel
Packagel
3mm
40-100
07-1 7
100-4500
.4-35
800-6500
7-360
6500
70-250
20,00040,000
2001050
70,000
32mm
40mm
... a
•
60mm
1500~
10,000
SERIES
MA
Z
L
P
HE
Operating Ambient
Temperature
-55 to
+75°e
-55 to
+85°e
-55 to
+85°e
-40 to
+85°e*
-40
+85°e
Storage Temperature
-55 to
+ 150°C
-55 to
-55 to
-40 to
-40 to
-' 55 to
+ 125°C
+ 125°C
+ 125°C
+ 125°C
+ 125°C
1000
2500
2500
NA
2500
NA
HiPot Encapsulation, Volts dc
For 1 Minute
Voltage Temperature
Coefficient
B,D
-55 to
+75°e
-0.03%/oe -0,05%/oe -0.05%/oe -0.05%/oe -0.05%/oe -O,lO%/oe
> 1000
> 1000
> 1000
Insulation Resistance (MOl
NA
NA
NA
*8ase Plate Temperature.
Solderability: Per mil std 202E, method 20Se_
,1
>
CI
a:
w
Z
w
ri
w
~
100
""'"
\
90
80
Q.
0
B,DSERIES-
60
SERIES
~ k--P
E SERIES
~
SO
u.
40
IZ
30
«
a:
0
w
0
a:
w
Q.
\
l\
1\\
,.-
o
~
a:
U -,2
w
:)
I-
'"
'\
50
60
70
80
90
100
110
AMBIENT TEMPERATURE -
;2
I'-.
'\
)J
LL
LL
W
MA SERIES
10
-55
2
w
§ -.1
i,\ \ ~
20
o
I-
~
\ \"-
w
l-
T~PE V22ZA3
'~ a
\ M'
70
0
SAMPLE
u
v Z SERIES
V L SERIES
120
130
°c
140
-.3
w
0..
'"
:2E
w
I- -.4
I
150
/
I
/
V-
~
~
r-
-,5
,01
Current, Power, Energy Rating
vs. Temperature
,1
10
100
CURRENT (MILLIAMPERES)
Typical Temperature Coefficient of Voltage
Versus Current Z Series, 14mm Size,-40 to +85°C
134
HOW TO SELECT A GE-MOV®II VARISTOR
To select the correct GE-MOV® II Varistor for a specific application, determine the following
information:
1. What is system RMS or dc voltage?
A. Phase to Ground _ _ _ _ __
B. Phase to Phase _ _ _ _ __
2. How will the GE-MOV®II Varistor be connected?
A. Phase to Ground _ _ _ _ __
B. Phase to Phase _ _ _ _ __
3. Calculate required varistor voltage at 10-25 % above system RMS or dc voltage.
A. V Phase to Ground
B. V Phase to Phase
X
X
1.1 = _______
1. 1
= _______
The maximum continuous RMSor dc varistor voltage should be equal to or greater than either 3A or
3B. This maximum continuous RMS or dc varistor voltage can be found in the rating and
characteristic tables Vm(ac) or Vm(dc)'
4. Selecting the correct varistor voltage is reasonably straightforward,
but selecting the proper energy
rating is more difficult and normally presents a certain degree of
uncertainty. Choosing the highest
energy rating available is expedient, but usually not cost effective.
A
B
c
As economic considerations enter
the selection process, the worst
case size of the transient, the
frequency of occurrence, and the
life expectancy of the equipment
to be protected Cannot be ignored.
ANSI/IEEE C62.41-1980 addresses these considerations and
the reprint on page 21 gives the
background and the environment
description of this standard. From
ANSI/IEEE C62.41-1980 it becomes evident that the equipment
or component to be protected is not
as important as the location in the
electrical system. ANSI/IEEE
C62.41-1980 divides the electrical
distribution system into 3 location
categories. Figure 1 defines these
location categories in detail.
A. Outlets and Long Branch Circuits
All outlets at more than 10m (30
ft) from Category B with wires
#14-10
All outlets at more than 20m (60
ft) from Category C with wires
#14-10
B. Major Feeders and Short Branch
Circuits
Distribution panel devices
Bus and feeder systems in
industrial plants
Heavy appliance outlets with
"short" connections to the
service entrance
C. Outside and Service Entrance
Service drop from pole to
building entrance
Run between meter and
distribution panel
Overhead line to detached
buildings
Underground lines to well pumpfl
lighting systems in commercial
buildings
FIGURE 1. LOCATION CATEGORIES
135
TABLE 1. SURGE VOLTAGES AND CURRENTS DEEMED TO REPRESENT THE INDOOR
ENVIRONMENT AND RECOMMENDED FOR USE IN DESIGNING PROTECTIVE SYSTEMS
LOCATION
CATEGORY
A. Long branch
circuits and
outlets
B. Major feeders
short branch
circuits, and
load center
COMPARABLE
TO IEC 664
CATEGORY
IMPULSE
WAVEFORM
MEDIUM EXPOSURE
AMPLITUDE
TYPE
OF SPECIMEN
OR LOAD
CIRCUIT
ENERGY(JOULES)
OEPOSITEO IN A SUPPRESSOR.'"
WITH CLAMPING VOLTAGE OF
500V
1000V
(120V System) (240V System)
0.8
1.6
II
O,SI-'S - 100kHz
6kV
200A
High impedance(l)
Low impedance(2),
III
1.2/S0l-'s
8/2Ol-'s
6kV
3kA
High impedance")
Low impedance<2l
40
80
O.SI-'S - 100kHz
6kV
SOOA
High impedance(!)
Low impedance
-
4
2
Notes: (1) For high-impedance test specimens or load circuits, the voltage shown represents the surge voltage. In
making simulation tests, use that value for the open-circuit voltage of the test generator.
For low-impedance test specimens or load circuits, the current shown represents the discharge current of the
surge (not the short-circuit current of the power system). In making simulation tests, use that current for the
short-circuit current of the test generator.
(3) Other suppressors which have different clamping voltages would receive different energy levels.
(2)
Table 1 shows the open-circuit voltage and short-circuit current of the transients which can be
expected at location A and B.
The GE-MOV®II Varistor selected must first survive the worst case transient (see "Medium
Exposure Amplitude" in Table 1) and, secondly, clamp the maximum open-circuit voltages to levels
which will not damage equipment or components in the system to be protected.
5. Select proper location category, A or B.
6. Determine worst case transient current and voltage from Table 1.
7. Knowing the maximum continuous RMS or dc varistor voltage (from 3), determine maximum
clamping voltage from V-I curve for the device selected using the worst case transient current found
in 6.
8. Does this clamping voltage provide the required protection level? If not, repeat Step 7 using a higher
energy-rated device. If this process proves to be ineffective, consult your local General Electric sales
office for assistance.
9. In many cases the source of the transient is known. The transient energy can be calculated, and
maximum clamping voltage can be determined from the V-I characteristic since the maximum pulse
current or source impedance is known. Examples of these calculations can be found in the Transient
Voltage Suppression Manual, Chapter 4.
136
MA Series, Axial
Lead Style; 3 mm
GE-MOl/®1I Metal Oxide Varistors
For Transient Voltage Protection
RATINGS AND CHARACTERISTICS TABLE: .
MASERIES
.
MAXIMUM RATINGS (25°C)
CONTINUOUS
MODEL
NuMBER
V18MA1B
V22MA1B
V27MA1B
v33MAiA
V33MA1B
V39MA2A
V39MA2B
V47MA2A
V41MA2B
V51MA2A
."
V51MA2K
Vl8MA3A
VlaMAaB
V82M~3A
V82MA3B
V100M,A4A
V1ooMA4B
V120lVlA1A
V12oMA2B
V150MA1A
V15OMA2B
V1BoMA1A
V180MA3B
V22o"A2~
V22oMA4B
V27dMA2A
V27oMA4B
V33DMA2A
V33OMA5B
V39OMA3A
V311oMA6B
V43OMA3A
V43OMA7B
PEAK
CURRENT
(8/201'11
Vmllel
W,.
I'lm
VOLTS
JOULES
(WATT-SEC)
AMPERES
MIN.
V.(do)
MAX.
VOLTS
PICOFARADS
40
40
40
40
15.0
19.0
24.0
26.0
29.5
31.0
35.0
37.0
42.0
44.0
50.0
54.0
61.0
65.0
73.0
80.0
90.0
102.0
108.0
127.0
135.0
153.0
162.0
187.0
198.0
229.0
243.6
280.0
297.0
331.0
351.0
365.0
387.0
18
22
27
33
21.0
26.0
31.0
40.0
36.5
47.0
43.0
57.0
52.0
68.0
62.0
82.0
75.0
99.0
91.0
120.0
110.0
138.0
132.0
173.0
165.0
207.0
198.0
253.0
242.0
311.0
297.0
380.0
363.0
449.0
429.0
495.0
473.0
44
51
59
73
67
86
79
99
90
117
108
138
127
163
150
200
185
220
205
255
240
310
290
380
360
460
440
570
540
670
640
740
700
550
410
370
300
RMS
VOLTAGE
V..(del
14
18
22
23
26
28
31
34
38
40
45
48
56
60
. 66
72
81
97
101
121
127
144
152
181
191
224
235
257
274
322
334
349
365
MAXIMUM
CLAMPING
VOLTAGE.
Ve@lp=2A
(8/201'11
TEST Ve
VARISTOR
VOLTAGE
@ 1.0mA
DC
CURRENT
ENERGY
(10 /10001'1)
DC
VOLTAGE
VOLTS
CHARACTERISTICS
TRANSIENT
10
14
17
18
20
22
25
27
30
32
35
38
40
45
50
57
60
72
75
88
92
105
110
132.
138
163
171
188
200
234
242
,253
264
0.07
0.10
0.11
0.13
0.15
0.16
0.18
0.19
0.21
0.23
0.25
0.26
0.30
0.33
0.37
0.40
0.45
0.40.
0.50
.0.50
0.60
0.60
0.70
0.80
0.90
0.90
1.0
1.0
40
40
40
40
40
40
100
100
100
100
100
100
1.1
1.2
1.3
1.5
1.7
100
100
NOTE: Power dissipatiorl of transients not to exceed 200 milliwatts.
137
39
47
56
68
82
100
120
150
180
220
.'
270
330
390
430
TYPICAL
CAPACITANCE
f
=0.1·1MHz
250
210
180
150
120
100
40
32
27
21
17
14
12
11
Z Series, Radial
Lead Style; 7, 10, 14,20 mm
T ~,~
GE-MOV®II Metal Oxide Varistors
for Transient Voltage Protection
RATINGS AND CHARACTERISTICS TABLE:
Z SERIES
NOTE: Power dissipation of transients not t.
exceed 0.25, 0.4, 0.6, 1.0 watts for
sizes 7, 10, 14 and 20mm respectively
MAXIMUM RATINGS (25°CI
CONTINUOUS
MODEL
NUMBER
VBZAI
V8ZA2
V12ZAI
V12ZA2
Vl8ZA1
V18ZA2
Vl8ZA3
V18ZA40
V22ZA1
V22ZA2
V22ZA3
V24ZA50
V27ZA1
V27ZA4
V27ZA60
V33ZA1
V33ZA5
V33ZA70
V36ZA80
V39ZAI
V39ZA6
V47ZAI
V47ZA7
V56ZA2
V56ZA8
V68ZA2
V68ZA10
V82ZA2
V82ZA12
Vl00zA3
V100ZA15
V12OZA1
V12OZA6
V150ZAI
V15ozA8
V180ZAI
Vl8OZA10
MODEL
SIZE
DlA.
(mml
DEVICE
MARKING
CHARACTERISTICS
TRANSIENT
VARISTOR
VOLTAGE
@1.OmADC
TEST
CURRENT
MAXIMUM
CLAMPING
VOLTAGE,
Ve@TEST
CURRENT
(8/20 !LsI
TYPICAL
CAPACITANCE
RMS
VOLTAGE
DC
VOLTAGE
ENERGY
(10/1000 !LsI
PEAK
CURRENT
(8/20 !LsI
Vrh1ac)
Vmldcl
Wtm
'tm
MIN.
VN1dC)
MAX.
Ve
Ip
f = 0.1-1 MHz
VOLTS
VOLTS
JOULES
AMPERES
AMPS
PICOFARADS
VOLTS
VOLTS
VOLTS
VOLTS
.8
100
250
6
6
8.2
8.2
II
II
22
20
5
5
4500
12000
.6
1.2
250
250
9
9
16
16
34
30
5
5
3000
7500
14
0.8
1.5
3.5
80.0·
250
500
1000
2000
14.4
18
21.6
42
39
39
37
5
5
10
20
2500
6000
12000
25000
14
18t
0.9
2.0
4.0
250
500
1000
18.7
22
26.0
47
43
43
5
5
10
2000
5000
10000
24Z50
14
18:j:
100.0·
2000
19.2
24t
26.0
43
20
20000
27Z1
27Z4
27Z60
17
22
250
1000
2000
23.0
27
31.1
22
1.0
5.0
120.0·
57
53
50
5
10
20
1700
8500
18000
33Z1
20
26
1.2
29.5
36.5
27
6.0
150.0·
250
1000
2000
68
64
58
5
10
20
1400
7000
15000
7
10
08Z1
08Z2
4
4
5.5
5.5
7
10
12Z1
12Z2
6
6
8
8
7
10
14
20
18Z1
18Z2
18Z3
18Z40
10
7
10
14
22Z1
20
7
14
20
22Z2
22Z3
.4
12
12
18t
27t
7
14
20
33Z70
21
20
36z80
23
31
160.0·
2000
32.0
36t
40.0
63
20
12000
7
14
39Z1
25
31
1.5
7.2
250
1000
35.0
39
43.0
79
76
5
10
1200
6000
7
14
47Z1
47Z7
30
38
1.8
8.8
250
1000
42.0
47
52.0
92
89
5
10
1000
5000
7
14
56Z2
56Z8
68Z2
35
45
2.3
10.0
250
1000
50.0
56
62.0
107
103
5
10
800
4000
33Z5
39Z6
7
14
68ZIO
7
14
33
33t
40
56
3.0
13.0
250
1000
61.0
68
75.0
127
123
5
10
700
3500
82Z2
82Z12
50
66
4.0
15.0
250
1000
74.0
82
91.0
152
147
5
10
600
3000
7
14
100Z
100Z15
60
81
5.0
20.0
250
1000
90.0
100
110.0
180
175
5
10
500
2500
7
14
120Z
120Z6
75
102
6.0
22.0
1200
4500
108.0
120
132.0
205
210
10
50
200
1200
7
14
150Z
150Z8
95
127
8.0
30.0
1200
4500
135.0
150
165.0
250
255
10
50
170
1000
7
14
1'80Z
180ZIO
I 15
153
10.0
35.0
1200
4500
162.0
180
198.0
295
300
10
50
140
800
..
·Energy raling for impulse duralion of. .~O milliseconds minimum (() one half of peak current value.
t IOmA dc lesl current.
:j:Also raled (() wilhsland24V for 5 minules.
138
LA Series, Radial
Lead Style; 7,10,14,20 mm
T ~,~
GE-MOV®II Metal Oxide Varistors
for Transient Voltage Protection
RATINGS AND CHARACTERISTICS TABLE:
L SERIES
CHARACTERISTICS
MAXIMUM RATINGS (2S0 CI
CONTINUOUS
MODEL
NUMBER
MODEL
SIZE
DEVICE
DIA.
(mml ,.,ARKING
V130lA2
V13OlA5
V130lA1OA
V13OlA2OA
V130lA2OB
7
10
14
20
20
V14OlA2
V14OlA5
V14OlA1OA
7
10
14
V15OlA2
V15OlA5
V15OlAl0A
VI5OlA20A
V15OlA2OB
7
10
14
20
20
V175LA2
V175LA10A
7
14
V23OlA4
V23OlA20A
7
14
V25OlA4
V25OlAl0
V25OlA2OA
V25OlA4OA
V25OlA4OB
7
10
14
20
20
V275LA4
V275LAla
V275LA2OA
V275LA4OA
V275LA4OB
7
10
14
20
20
1302
1305
130LlO
130L20
130L20B
1402
1405
140LlO
1502
1505
150LlO
150L20
150L20B
1752
175LlO
Series L Varistors are listed
under UL file #E75961 and
E56529 as a recognized
component.
TRANSIENT
VARISTOR
VOLTAGE
@ 1.0mA DC
TEST
CURRENT
MAXIMUM
CLAMPING
VOLTAGE,
Vc@TEST
CURRENT
18120/lsl
TYPICAL
CAPACITANCE
RMS
VOLTAGE
DC
VOLTAGE
ENERGY
(10/1000 /lsI
PEAK
CURRENT
(8120/ls1
Vmllcl
Vmldcl
Wlm
11m
MIN.
VNldCI
MAX.
Vc
Ip
f = 0.1·1 MHz
VOLTS
VOLTS
JOULES
AMPERES
130
175
VOLTS
VOLTS
VOLTS
VOLTS
AMPS
PICOFARADS
11
20
38
70
70
1200
2500
4500
6500
6500
184
200
228
228
228
228
220
340
340
340
.'>25
10
25
50
100
100
180
500
1000
1900
1900
:HO
140
180
12
22
42
1200
2500
4500'
198
220
242
242
242
360
360
.'>60
10
25
50
160
480
900
150
200
13
25
45
80
80
1200
2500
4500
6500
6500
212
240
268
268
268
268
243
395
395
395
.'>95
.'>60
10
25
50
100
100
150
400
800
1600
1600
175
225
15
55
1200
4500
247
270
303
303
455
455
10
50
130
700
230
.'>00
20
70
1200
4500
.'>24
360
396
396
595
595
10
50
120
600
250
330
21
40
72
130
130
1200
2500
4500
6500
6500
354
390
429
429
429
429
41.'>
650
650
650
650
620
10
25
50
100
100
110
250
500
1000
1000
275
369
23
1200
2500
4500
6500
6500
389
430
45
75
140
140
473
473
473
473
453
710
710
710
710
680
10
25
50
100
100
100
230
450
900
900
V320lA2OA
V320lA4OB
14
20
V42OLA10
V420lA2OA
V42OLA4OB
10
14
20
V48OlA4OA
V48OlABOB
14
20
V510lA4OA
V510lABOB
14
20
2304
230L20
2504
250L
250L20
250L40
250L40B
2754
275L
275L20
275L40
27SL40B
3004
320L20
320L40
420L
420L20
420L40
480L40
480L80
510L40
510L80
V575LA4OA
V575LABOB
14
20
575L40
575L80
575
730
120
220
4500
6500
805
910
1000
960
1500
1410
50
100
370
750
V1000lA1OA
V1000lA 1608
14
20
1000L80
1000
1200
220
4500
6500
1425
1690
~000Ll60
1800
1650
2700
2420
50
100
200
400
V3OOLA4
7
305
405
25
1200
420
470
517
775
10
90
320
420
90
160
4500
6500
462
510
565
540
850
810
50
100
380
750
420
560
45
90
160
2500
4500
6500
612
610
610
680
748
748
720
1120
1120
1060
25
50
100
220
500
1000
480
640
105
180
4500
6500
670
750
825
790
1240
1160
50
100
450
900
510
675
110
190
4500
6500
7)5
820
910
860
1350
1280
50
100
400
800
~O
NOTE: Power dissipation of transients not to exceed 0.25,
0.4,0.6,1.0 watts for sizes 7, 10, 14 and 20mm respectively.
139
P Series; Base
Mount Style; 20 mm
--..
GE-MOV®II Metal Oxide Varistors
For :rransient Voltage Protection
HE Serles,Base
Mount Style; 32 mm
,
RATINGS AND CHARACTERISTICS TABLE:
P/HE SERIES
I'
Series P/HE Varistors are listed under UL file #E75961 as a UL recognized component.
MAXIMUM RATINGS (25°C)
CONTINUOUS
MODEL
NUMBER
V130PA20A
V130PA20C
V130HE150
V150PA20A
V150PA20C
V150HE150
V250PA40A
V250PA40C
V250HE250
V275PA40A
V275PA40C
V275HE250
V320PA40A
V320PA40C
V320HE300
V420PA40A
V420PA40C
V420HE400
V480PA80A
V480PA80C
V480HE450
V510PA80A
V510PA80C
V510HE500
V575PA80A
V575PA80C
V575HE550
V660PA100A
V660PA100C
V660HE600
V750HE700
CHARACTERISTICS
TRANSIENT
VARISTOR
VOLTAGE
@ 1mA DC
TEST
CURRENT
Vlldel
MAX*
MAXIMUM
CLA,MPING
VOLTAGE,.,
Vc @ TEST
CURRENT
(8 J 2op,s)
TYPICAL
CAPACI·
TANCE
RMS
VOLTAGE
DC
VOLTAGE
ENERGY
(10 I 1ooop,s)
PEAK
CURRENT
(8 120p,s)
vm1ac)
Vmldc)
Wtm
Itm
MIN.
VOLTS
VOLTS
JoULES
AMPERES
VOLTS
VOLTS
VOLTS
VOLTS
AMPS
PICOFARADS
130
175
70
6,500
184
200
200
200
80
20,000
6,500
212
240
250
330
220
130
20,000
6,500
354
390
275
369
330
140
20,000
6,500
389
430
320
420
360
160
20,000
6,500
462
510
420
560
390
160
20,000
6,500
610
680
480
640
400
180
25,000
6,500
670
750
675
450
190
25,000
6,500
730
500
220
25,000
6,500
850
550
250
25,000
6,500
970
600
700
25,000
25,000
360
325
365
420
360
425
675
620
690
740
680
760
850
800
860
1160
1050
1200
1280
1160
1320
1410
1280
1450
1560
1410
1600
1820
1650
1850
2100
100
100
300
100
100
300
100
100
300
100
100
300
100
100
300
100
100
300
100
100
300
100
100
300
100
100
300
100
100
300
300
2400
150
243
220
228
284
243
268
453
413
429
494
453
473
565
540
539
790
690
748
860
790
824
963
860
910
1050
960
1005
1210
1100
1160
1320
510
575
660
750
735
805
940
1080
820
910
1050
1200
Vc
p
*With 50-60Hz ac test, the maximum voltage for ImA peak current is 5% higher.
NOTE: Average power dissipation of transients not to exceed 1.0-1.5 watts for PA/HE Series Varistors.
140
f
= o.1·1MHz
4700
2000
4000
1200
2500
1100
2250
1000
1900
1200
3000
1100
-
2700
1000
2400
900
2200
800
1900
1700
-MOV®II Metal Oxide Varistors
Transient Voltage Protection
o Series
Package
40mm
~TINGS
AND CHARACTERISTICS TABLE
B SERIES
B Series
Package
60 mm
ERIES
CHARACTERISTICS
MAXIMUM RATINGS 125°CI
CONTINUOUS
DDEl
MBER
11 DA40
i1DA40
i 1 DA40
'1DA40
!1DA40
!1 DA40
11 DA40
11DA40
"DA40
~1 DA40
51DA40
RMS
VOLTAGE
MAXIMUM
CLAMPING
VOL TAGE.
Vc @200AMPS
18 I 20 !AI
TRANSIENT
DC
VOLTAGE
ENERGY
lID IIOOO!A1
VARISTOR VOLTAGE
@l1.0rnADC
CURRENT
PEAK
CURRENT
IB I 20 !AI
TYPICAL
CAPACI·
TANCE
V m(ae)
Vm(de)
Wtm
Itm
MIN.
VN\dcl
MAX."
Vc
f=O.1·1MH2
VOLTS
VOLTS
JOULES
AMPERES
VOLTS
VOLTS
VOLTS
VOLTS
PICOFARADS
130
150
250
275
320
420
480
510
575
660
750
175
200
330
369
420
560
640
675
730
850
970
270
300
370
30,000
30,000
30,000
30,000
30,000
40,000
40,000
40,000
40,000
40,000
40,000
184
212
354
389
462
610
670
735
805
940
1080
200
240
390
430
510
680
750
820
910
1050
1200
228
268
429
473
539
748
825
910
1000
1160
1320
360
420
660
730
830
1160
1270
1400
1550
1780
2040
10,000
9000
5000
4500
3600
3000
2500
2400
2200
2000
1800
400
460
600
650
700
770
900
1050
Average power dissipation of transients not to exceed 2.0 watts.
;ERIES
CHARACTERISTICS
MAXIMUM RATINGS 125°CI
CONTINUOUS
MODEL
lUMBER
RMS
VOLTAGE
181BA60
511BA60
57tBA60
661 BA60
75tBA60
881 BA60
112BA60
142B860
172BB60
2028860
2428860
2828860
I
V m(de)
ENERGY
11 0 II OOO!AI
VARISTOR VOLTAGE
@l1.0rnADC
CURRENT
PEAK
CURRENT
18 I 20 !AI
MAXIMUM
CLAMPING
VOL TAGE.
Vc @200AMPS
18 I 20 !AI
TYPICAL
CAPACI·
TANCE
Wtm
Itm
MIN.
VN\dcl
MAX."
Vc
f=O.1·1MH2
VOLTS
VOLTS
JOULES
AMPERES
VOLTS
VOLTS
VOLTS
VOLTS
PICOfARADS
420
480
510
575
660
750
880
1100
1400
1700
2000
2400
2800
560
640
675
730
850
970
1150
1400
1750
2150
2500
3000
3500
1500
1600
1800
2100
2300
2600
3200
3800
5000
6000
7500
8600
10,000
70,000
70,000
70,000
70,000
70,000
70,000
70,000
70,000
70,000
70,000
70,000
70,000
70,000
610
670
735
805
940
1080
1290
1620
2020
2500
2970
3510
4230
680
750
820
910
1050
1200
1500
1800
2200
2700
3300
3900
4700
748
825
910
1000
1160
1320
1650
2060
2550
3030
3630
4290
5170
1140
1250
1370
1530
1750
2000
2500
3100
390Q
4500
5500
6600
7800
5100
4700
4300
3900
3300
2700
2200
1800
1500
1200
1000
800
V m(ae)
J21BA60
DC
VOLTAGE
TRANSIENT
50-60 Hz AC test. the maximum voltage for 1 mA peak current is 5% higher.
rE: 1. Average power dissipation of transients not to exceed 2.5 watts.
141
6000
I
I
-
GE-MOV®II Metal Oxide Varistors
for Transient Voltage Protection
STANDBY POWER DISSIPATION - B SERIES
STkNDBY POWER DISJIPATION VS.
APPLIED RMS VOLTAGE AT
VARIED TEMPERATURES
TYPICAL TEMPERATURE COEFFICIENT
OF POWER DISSIPATION = 2.2%/oC
1.0
0.9
0.8
0.7
a: W
w~
a. ce
........
111..1
, ."
1"'oC
.05
... 0
.04
.... a:
Ze
0.3
@"'>~
./
V
,/
,/
~
~z
~/
2w
......
~
j..\~\J~
0.2
V
iii II.
1110
E...
a:..1
wO
~~i
0..1
~
~p..~
I
~~
::
/
I'
.06
...
>
ce III
./
0.i2 '
0.1
.09
.08
@"'>
/'
,/
./
",oC/'
~z~
~@>
",oo(~\~
,,/"
.07
.06
~
.05
~
.04
90
80
100
110
PERCENTAGE OF MAXIMUM RATED RMS VOLTAGE - %
1"~ ..
.6
JJ
a: w .5
w~
o.~
(i)..1
... 0·
... > .4
TYPICAL STABILITY OF STANDBY POWER
DISSIPATION AT RATED RMS VOLTAGE
FOR 1000 HOURS, T A ;" 75 0 C
celli
~~
.... a:
Ze
2w .3
......
cece
r---- _____
a. a:
011.
1110 .2
E ...
a:..1
wO
.~~ .1
0..1
:w::
0.-
o
o
,..,..
JJ
100
10
TIME AT RATED RMS VOLTAGE (HOURS)
142
1000
GE-MOY®II Metal Oxide Varistors
for Transient Voltage Protection
TRANSIENT V-I CHARACTERISTICS:
MASERIES
400
v
- t:::: v t:::-
200
III
150
~
f--
tog 100 F""'"
~ 80
-~
~ t--
I--- ~ l -
V33 A1
200
~~
V
r--t-
V100M MA
V82MA3A
V68MA3A
V56MA2A
MA
t~ 60 F""
2
;:)
2 40
400r--r---r--.--'---'---r--r---r-~--'---.--'
./.
t:::- ~
I=--:: v
~150BBa~
~ t:/'
g
~
2
i
t--
~
~
lOa
80
60
L~--r
40
::E
2
T. = -55 TO 75°C
20
MAXIMUM CLAMPING VOLTAGE.
33 TO 100 VDC VN
A SELECTION
I
I
I
I
I
I
10
2
5
5
0.01
0.1
20
T. = -55 TO 75°C
MAXIMUM CLAMPING VOLTAGE,
18 to 100 VDC VN
B SELECTION
10L-L__-L__L--L__-L__L--L__-L~~~__~~
5
2
2
10
ioo
5
0.01 2
5
5
0.1
PEAK AMPERES
4000
4000
T. = -55 TO 75°C
MAXIMUM CLAMPING VOLTAGE.
120 VDC & HIGHER VN
A SELECTION
I
2000
III
1500
~ 800
2
~
V430MA3A
600
--I-::::
I---
k::::
-- t:::- -
V390MA3A
~~
V
I-- ~ ~
V330MA2A
;:)
2400 I - - I-V270MA2A
V220MA2A
V180MAIA
-V150MA1A
200
v'2! I\AA1A
i
10
5>
2
100
I
~I -
1500
~
g 1000
-'"
rV390MA6B
~ 800
V
~
I-
5
T. = -55 TO 75°C
MAXIMUM CLAMPING VOLTAGE.
120 VDC & HIGHER VN
BSELECTION
2000
III
g~ 1000
-
2
PEAK AMPERES
~ ~V
600
V330MA5B
400 F'" r-V2701\AMB
V220MA4B
V180MA3B
tV150MA2B
200
V120M
t-
i
I--
100
V430MA7B
_
2
;:)
2
-
i.--::: ~
V :::.,......
~~
- lI--- I---t:-:
-- t::: I--t::: ~V
I--"
-
100
0.01
2
5
5
0.1
2
5
10
2
5
100
0.01
5
2
0.1
PEAK AMPERES
2
5
5
2
1
10
2
5
100
PEAKAMPE+S
ZSERIES
200
100
f-+-
tJJ-51lt IIIII
a
~AnM3~~g~~~PING V~LTAGE.
80 I-f- MODEL SIZE 20mm
l,..--
60
j...-- I-"" j;..o-
50
40
l-
j...- ~I:: ~ ~ t-"" 10- i--" l--' ~
....,
c:......
200
V82ZA2
~ V/jJ
~
f.;
~~
v
V
~II--' l- i--"
I- I1-1 ~ I..--' i--"
30
l- I-"
I- I---
20
1-1- \-I
~~
I'-V36ZA80
V33ZA701
\~ ~ V27ZA&0
~~~~--~~~~~~~~~~~V~~1
r\ ~ V24ZA~0
~
10
2
0.00 I
5
2 5
0,01
2
O. I
5
V33ZAI
I I I
Y18f A
1°
V27ZAI
5252525
10
100
1,000
10,000
PEAK AMPERES
5
0.01
·2
0.1
2
5
2
10
PEAK AMPERES
143
5
2
100
5
1,000
GE-MOV®II Metal Oxide Varistors
for Transient Voltage Protection
TRANSIENT V-I CHARACTERISTICS:
Z SERIES (cont'd)
700
TA
600
500
III
~.
0
>
~
-c
w
Go
:I
;:)
::Ii 200
><
-c
::Ii
150
WL-~~L-~-L__~-L~__~-L~__-L~__~~~
0.01 2
5
0.1
2
5
2
5
2
10
5
2
100
-
:--l-I--
---
-
/
yV
,,"" k:.o' ~ ~
/
-,,'
300
/
~~ ~
I
400
!:i
.I
= -5510 85°C
'MAXIMUM CLAMPING VOLTAGE, J180iAl ~
120 VDC & HIGHER VN
I-V150ZA~ ___ !-..
MODEL SIZE 7 & 14mm
V120ZAI '")I
:.:-I~
~ ~ ..-; ~ "" ~ v~V
"
~
~v
"120~r-
I- f-P
100
1000
0.01
2
5
0.1
2
5
2
5 10
2
5 1000 2
5 100 2
5 10,000
PEAK AMPERES
PEAK AMPERES
LSERIES
6000
5000
4000
3000
TA = -5510 B5°C
MAXIMUM CLAMPING VOLTAGE
130 VAC RATING MODELS
MODEL SIZE 7, 10, 14, 20 mm
-
r!
....
g 1000
~
800
Go
600
500
400
.:E
;:)
:!i
X
~
1\
V 130LA208
V 13QlA20A
V130LA10A
V130LA5
Vl,30LI\2
2000
-c
w
6000
I
I
SOOO riTA = -55 to 850C,
4000 f-M'AXIMUM CLAMPING VOLTAGE
140 VAC RATED MODELS
3000 f- MODEL SIZE 7, 10, 14 mm
LI
h.;
1\
:
r!....
o
>
"«~
./
v .... :...:: ::;.--
~
X
300
~
200
5
/'
""'"
,300
5
10
5
100
~2000
V150LA10A
vlSOLA20A
1,000
g
VISOLA208
-
,,1000
~ 800
5
5
10,000
600
500
--
-400
300
5
0.1
0,01
10
100
1,000
5
10,000
~
6000
I
5000 'TA .. -55 10 85°C
4000 MAXIMUM CLAMPING VOLTAGE
175 VAC RATED MOLELS
3000 MODEL SIZE 7 & 14 mm
1"\
[;\
2000
r!
...
~~
g
"
:;...-
V175LA2
1000
~ 800
Go
-
Q.
:I
.,..,/'
::l
600
SOO
400
~
300
~
I-'"
..-
--
V175LA1Oi\-
200
200
100
100
0.01
...-:
PEAK AMPERES
6000
5000 'TA = -5510 85°C
4000 MAXIMUM CLAMPING VOLTAGE
V150LA2
ISO VAC RATING MODELS
3000 MODEL SIZE 7, 10, 14, 20mm
-VISOLA5
~c
~
100
5
0.1
PEAK AMPERES
~
::Ii
~\ ~
1000
BOO
200
100
0.D1
--.,
V140LA10A '"'
:!i 600
;:) 500
:!i400
"'
~
V140LA5
2000
~
./
I--V14bLA~
5
0,1
10
100
5
5
5
0.01
1,000
0.1
5
5
10
PEAK AMPERES
PEAK AMPERES
144
100
1,000
10,000
GE-MOV®II Metal Oxide Varistors
for Transient Voltage Protection
TRANSIENT V-I CHARACTERISTICS:
L SERIES (cont'd)
6000
5000
4000
3000
6000
5000
4000
TA = -55 to 85°C
M'AXIMUM CLAMPING VOLTAGE
230 VAC RATED MODELS
MODEL SIZE 7 & 14 mm
i!?
~
w
1000
800
::;:
600
500
400
~
300
0.
~
X
V230LM
0.
V230LA20A
::;:
::l
::;:
X
«
::;:
10
PEAK AMPERES
2000
1,000
100
V275LA40B V275LA20AI----- V275LA10 _
V275LM
V300LA4
-
h
~ BOO
6000
TA - -55 to 85°C
5000
MAXIMUM CLAMPING VOLTAGE,
4000 320 VAC & HIGHER RATING
3000 MODEL SIZE 14mm
A SELECTION
2000
V57SLA40A ~
to-'
0 1500
h
+-.
f-\ {\
b ~\\
~c..-: ~~
::;:
300
.
::l
600
'''''ll
::;:
X
'«::;:
400
II
100
100
2
5
1,000
100
10
0.Q1
10,000
5
2
2
0,1
10
PEAK AMPERES
PEAK AMPERES
I!?
-'
o
>
.
~
«w
::;:
6000
SOO 0 'TA = -55 to 85°C
4000 MAXIMUM CLAMPING, VOLTAGE,
320 VAC & HIGHER RATING
3000
MODEL SIZE 20mm
B SELECTION
2000
Vl000LA160B1500
1000
BOO
i
600
~
400
X
jll'
"'~~w
V
I<-
k:: ~
-
~t:::==
-"
V
VS10LA80B
V480LA80B
V420LA40B
V320LA40B
300
200
lOa
0,01
2
S
0,1
2
S
10
2
5
PEAK AMPERES
145
t:. '-"'"
V480LA40A
V420LA20A
V320LA20A
300
200
0.1
-
k
Vl000LA80A
~ ~~
'II
200
0.Q1
flL-
1000
800
::;:
./
1- I--"
>
~
«
w
600
!g;;
10,000
1,000
100
I I
'"
0.
i
10
0.1
0.01
10,000
PEAK AMPERES
~V275LMOA-
~ 1000
~
300
100
0.1
~k%
600
500
400
100
~
::;:
1000
800
200
6000
I- TA = -55 to 85°C I
5000
I-MAXIMUM CLAMPING VOLTAGE
4000
275,300 VAC RATING MODELS
3000 MODEL~ZE~l~l~Wmm
en
~
«
w
=\
V250LA40B ~
>
/
200
0.01
V250LMO~
0
1/
\
V250LA20A --.\
'" 2000
to-'
..J
>
i-V250~MI.
3000
2000
o
TA - -55 to 85°C
MAXIMUM CLAMPING VOLTAGE
250 VAC RATING MODELS
MODEL SIZE 7,10,14,20 mm
i-V250LAlO
100 2
5
1,000
2
5
10,000
100
5
2
1,000
S
10,000
GE-MOV®II Metal Oxide Varistors
for Transient Vol tag, Protection
TRANSIENT V-I CHARACTERISTICS:
P SERIES
6000
5000
4000
3000
......
I/)
2000
0
1500
...'"
1000
>
C
I
T. = -40 TO io°c
MAXIMUM CLAMPING VOLTAGE,
130 VAC & HIGHER RATING
A SELECTION
::E
;:)
::E
-
W \\
400
::E
200
,I
....-: ~
-
"~,.~~
......
0
I/)
'\
~
~
>
II.
v
~
V480PA80A
V420PA40A
r- ~
2
10
2
100
5
1,000
2
~
V320PA40~
0.01
'10,000
'\ ~
-JIlL
2
5
0.1 2
5
1
2
5
:;..:: v
-- -::::: ~
:- t-
V480PAsOC
: ' - V420PA40C
V275PA4011'l
V250PA40C
V150PA20C
VI30PA20C
100
5
"""l\
1/1
200
II II
0.1
~
400
::E
100
2
~
600
i(
C
:::: ..... ~
-----
~
1000
800
'~
::E
~
_f--~~ ~
1500
...'C"
/.
:- -: ::::: f:::
V275PA40A
V250PA40A
VI50PA20A
V130PA20A
om
~
-: --::: ~
'\
600
i(
C
V660PA100A ~
V575PA80A
V150PA80A ~
-
-::: ~ ~ ~
800
II.
-
6000
5000 T.\ -40 TO 700C
I I I I I
4000 MAXIMUM CLAMPING VOLTAGE,
V660PA looC
3000 130 VAC & HIGHER RATING
~ V575PABOC ~
C SELECTION
V510PA80C ~~)
2000
10 2
I III
5
100
2
5
1,000
~
•'10,000
2
PEAK AMPERES
PEAK AMPERES
HE SERIES
6000
-
5000
T. = -40 TO 85°C
MAXIMUM CLAMPING VOLTAGE,130 TO 320 VAC RATING
MODEL SIZE 32mm
4000
-
3000
......
0
'...c"
1000
800
;:)
::E
::E
600
~
400
>
II.
......
0
I/)
2000
I/)
V320HE300,
V275HE250 ~\\
V2S0HE2S0
'""I,
1500
\4
-
I-~
't'
.... - """
,/
...- r----:::::
::E
~
V150HE1S0 ....
V130HE150
::E
;:)
::E
i(
C
::E
S
0.001
10
0.1
0.01
5
0,0
S
5
100
---:::
~
100
5
r
~
@
200
II II
100
-
1000
800
600
500
400
300
II.
-
~
300
...'"c
T. - -40 TO 8So C
MAXIMUM CLAMPING VOLTAGE,
420 TO 750 VAC RATING
MODEL SIZE 32mm
V480HE450
V510HESOO
V57SHE550
V660HE600
V750HE700
2000
>
~7
"~""l\
6000
SOOO
4000
3000
5
1,000
0.01
0.1
5
5
5
10,000 100 ,000
5
10
100
S
1,000
5
S
10,000 100,00(.
PEAK AMPERES
PEAK AMPERES
B SERIES
D SERIES
6000
5000
4000
3000
~
g
T A =-55 TO 75°C
MAXIMUM CLAMPING VOLTAGE
130T0750Vm (ac) RATING
MODEL SIZE 40mm
V751DA40
V661DA40
V571DA40
V511~
2000
~~
1000
II.
BOO
::E 700
600
500
-- ---
i
X
c(
........ , , /
400
V151 OA40
V131 OA40
200
r-
......
0
I/)
>
:l'"
II.
ii(
C
:.
-55 TO 75'C
MAXIMUM CLAMPING VOL TAGE
420 TO 2800 Vm(acl RATING
MODEL SIZE 60mm
./... V2828860
V242BBeo
15.000
10.000
8000
6000
5000
4000 :
3000
--
2000
---- -
2
2
0.1
5
~
2525252525
10
100
PEAK AMPERES
1000
10,000
0.1
100,000
10
100
~
~
5 ,000
'
PEAK AMPERES
146
/'/J ~~~~~~~~
142BB60
V112BA60
,.., . / ~~8818A60
r-- ...... / ~~~l~~~g
..4'~~m~~g
~- .....
~V4818A60
1000
100
om
~
:.
~
300
I
r-- T.
DA40
V4fl DA40
10MO
~V32 50A40
~V27
V2500A40
.......
'~,"
::E
I
I
:::v
~Y481
::::v
V4218Aeo
~
5 10,000 5 '00 ,000 5
GE-MOV®II Metal Oxide Varistors
for Transient Voltage Protection
*PULSE RATINGS:
MA SERIES
100~---------r-------r------~----------r-------'
50
'"
20
::E
10
A-
C
'"a ZI
Method 2056
Method 1041
Method 2031
MIL-202E, Method 215
MIL-202E, Method n lA
20 G's; 100-2000Hz; XI' Y I, ZI
35°C; 24 Hr.; 10-50 G/M2/Day
260°C; 10 Sec.; 3 Cycles; Test
Marking Permanence; 3 Solvents
15 Sec. Torching; 10 Sec. to Flame Out
Note: High reliability varistors are rated to withstand a low temperature storage of - 65°C.
Please contact your local General Electric Sales Office for any specific high reliability requirement or for
types presently available.
150
GE-MOV@II Metal Oxide Varistors
for Transient Voltage Protection
OUTLINES AND DIMENSIONS:
MA SERIES
M1WMETERS
MIN.
MAX.
.60
.83
SYMBOL
cf>b
cf>D
3.43
8.01
26.0
G
L
3.68
8.50
29.0
INCHES
MIN.
MAX.
0.024
.135
.315
1.03
.033
.145
.335
1.14
Typical Weight· O.35g.
Z AND L SERIES
Dimensions: MM PN.)
VARISTOR MODEL SIZE
14MM
lOMM
20MM
A (MAX)t
11.7
(0.461 )
16.0
(0.630)
18.9
(0.744)
25.5
(1.01)
cf>D(MAX)
8.7
(0.343)
12.5
(0.492)
16.4
(0.646)
22.5
(0.886)
5.0
(.197)
7.5
(0.296)
7.5
(0.2%)
7.5
(0.2%)
0.5
1.2
2
~
SYMBOL
e±IMM
(.039) ,
10. 14,20mm
Models
.81 ± .05
(.032 ± .002)
Typ. Weight
(g)
7MM
VARISTOR MODEL
7mmModeis
.635 ± .05
(.025 ± .002)
V8ZAV56ZA
V68ZA Vl00ZA
Vl20ZA Vl80ZA
Vl30LA V321lA
el
• 2.0
(,079)
E (MAX)
5.0
(.197)
SYMBOL
'il MM
(.039)
:j:±l.5MM
(0.59)
V421lA
V57ftA
Vl000LA
• 2.5
(.098)
:j:4.0
(.157)
:j:7 .3
(,287)
5.6
(.220)
7.3
(.287)
10.8
(.425)
AVAILABLE LEAD STYLE CHANGES
m~
.~
S~t,rmlj
T
25.4
(1.00)
MIN.
t
Crimped
Lead
Dimension A: MM (IN.)
.
L
TRIM
3.55(0.140)
± 1.14 (.045)
SYMBOL
A (MAX)
,
7MM
VARISTOR MooEL SIZE
lOMM
14MM
15,0
<.591 )
19.5
(.768)
22.5
(.890)
20MM
29.0
(1.140)
Crimped and
Trimmed Lead
Tape and ReeltFor Models V420LA-V1000LA, A(MAX) for 10, 14, 20MM
is 17.0(.669), 20.0(.787), 28.0(1.10) respectively.
*for Tape and Reel availability and specifications, Contact Factory.
151
GE-MOV®II Metal Oxide Varistors
for Transient Voltage Protection
INCHES
SYMBOL
MILLIMETERS
MIN. NOM. MAX. MIN. NOM. MAX.
A
b
b2
b3
B
C
0
.57
.26
NOTES:
1. Tab is designed to fit
'I." quick connect terminal.
2. Case temperature is measured at Tc on top surface of
base [plate.
3. H, (I30-150VRMS devices)
H, (250-320V RMS devices)
H, (420-660VRMS devices)
4. Electrical connection: top terminal and base plate.
5. Typical weight - l00g.
14.3
6.6
.16
.13
E
.44
F
h
.30
.03
H,
H2
H3
J
c/>p
.22
1.99
Q
S
T
c/>T
Tc
2.00
.76
12.9
6.5
66.2
33.5
11.2
7.7
.8
.04
1.01
1.12
1.29
.32
.24
2.01
23.2
24.6
26.3
5.8
50.6
50.8
19.2
.04
.11
1
4.1
3.2
.51
.26
2.61
1.32
C/>O,
NOTES
.9
25.5
28.3
32.6
8.1
6.0
51.0
3
3
3
1.0
2.8
.13
3.2
2
HE SERIES
DIMENSION
A
B
.-
r~,
C,
0
E
F
G
H
J
K
L
M
N
NOlA (TYP)
~~~:1 I"
K:I
P--.-I
P
I-
MILLIMETERS
61 MAX.
41. MAX.
44.45 ± .75
25.40 ± .75
16.5 NOM.
3.2 NOM.
23 MAX.
41 MAX.
13 NOM.
1.6 NOM.
6.4 NOM.
'6.4NOM.
sA NOM.
40.5 NOM.
INCHES
2.40 MAX.
1.60 MAX.
1.75 ± .03
1.00 ± .03
.65 NOM.
.13 NOM.
.91 MAX.
1.60 MAX.
.51 NOM.
.06 NOM.
.25 NOM:
.25 NOM.
.21 NOM.
1.6 NOM.
Typical Weight .. ,........... '....................... l00g.
Mil1imum Strike and Creep Distance
Terminal to Terminal ........................... 1.4 in.
(3.5 em.)
Terminal to Baseplate .......................... 0.80 in.
(2.0 em.)
152
GE-MOV@II Meta. Oxide Varistors
for Transient Voltage Protection
MM(IN) .
OUTLINE DIMENSIONS:
BSER'ES
.
/
PACKAGE B SERIES TYPE
ALL MAXIMUM DIMENSIONS
EXCEPT WHERE NOTED;
i40mmll.5811
I
I
I
I
I
i
~
,
!
I
I
95mm
(3.741
-79mm
(3.111
MAX
I
:
" ,
I
Typical Weight: BA-250g.
BB-600g.
o SERIES
4.3mm
(.1701
=::::'==
f
TYPICAL WEIGHT: BOg.
ALL MAXIMUM DIMENSIONS
EXCEPT WHERE' NOTED.
Typical weight: 6ag.
153
GE-MOV®II Metal Oxide Varistors
for Transient Voltage Protection
SUGGESTED MOUNTING OF THE P SERIES VARISTOR
.
j- d
.
THERMAL
~
1"" ,.
GREASE~
LAYER
I
I
.j.7
.
~
'4--#10_32 PAN HEAD
\
SCREW
TERMINAL
\\" Q. UICK----IIICONNECT
~:~
~m'
I
I
[t..~
I
++--#10 FLAT WASHER
I
~VARISTOR
,
::.::=-_-:;=+-MOUNTING SURFACE
,-:.
+ - - LOCK WASHER
.--#6-32 x %" LG.
.
i . - - # 6 FLAT WASHER
. +--PHENOLIC SHOULDER
_ . .
WASHER
.--SPACER
\
II
4---'-VARISTOR
(1) THERMAL~
G.REASE
==:;::.====.=!==.::>+-MICA INSULATOR (2)
.. I
; = '+-MOUNTINGSURFACE
LAYER
,
+--#10-32 NUT
II
SCREW
.
. _-:.::.::=-_-_-
.
4 - - - LOCK WASHER
4 - - - #6-32 NUT
Typical Non-Isolated Mounting
Typical Isolated Mounting
NOTE&:
(1) GE G623, Dow Coming, DC3, 4, 340, or 640 Thennal
Grease is recommended for best heat transfer.
(2) lOOO-volt isolation kit containing the following parts can be
ordered by part #A7811055.
(l)MICAinsulation I" /3.1" /.005" thick.
(2) #6-32 / 3/4 screw.
(2) #6 flat washer.
(2) Phenolic shoulder washer.
(2), #6 intemt\l tooth lock washer.
(2) #6-32 nut.
(1) '14" quick connect tenninal.
(1) Spacer.
154
ACBias
AC-DC Difference
Across-The-Line Components
Accelerated Humidity Life
Accelerated Storage Life
Accelerated Testing
Acceleration Testing
AC Power Lines, Transients
Activation Energy
Air Gaps
120,122,123,128
112
13
120
17
Alpha
Alpha Stability
Alternator Field Decay
ANSI
ANSI Standard
Applications, Automotive
Applications, Crowbar
Application Examples
Applications, Motor Control
Applications, Motor Protection
Applications, Noise Reduction
Applications, Power Supply
Applications, Relays
Arcing
Arcing, Switch
52
121
15,95
116
19,30,86,116,117,139
95
84
71-86,91-93,96-100
73
81-84
78
71,72
75-77
75
11
Arrhenius Models
Assurance Tests
Asymmetrical Polarity
Automotive Ignition
Automotive Test Waveforms
Automotive Transients
.Avalanche Diodes
Branch Circuits
Break-In Stabilization
Building Locations
Buried Cable
Cable Plant
Cable Shield
Cable, Suspended
Cable Transients
Capacitor Clamping
Capacitance, Stray
Capacitance, Varistor
Carbon Blocks
Carbonization
Catastrophic Failure
Ceramic Structure
Channeling
Characteristics, Electrical
Characteristic Measurement
Characteristics Table
Circuit Models, Varistors
Clamp Ratio
Clamp Voltage Measurements
Clamping
Clamping Voltage
Clearing Time
Comparison of Suppressors
Conservation, Power
-Construction, Varistor
Contacts
Contact Arcing
Contact Bounce
Contact Bouncing
Contact Welding
Continuous Power Dissipation
Coupling, Mutual
Creepage Specifications
Crowbar, Power Supply
Crowbar Type Suppressors
119
104
112
128
128
Current, Derating
Current, Peak Rating
Current Ratings
DC Bias
DC Power Life
DC Standby Current
Degradation
Degradation, Semiconductot
Detection of Transients
Detector, R. F. Noise
Delay Time, Gas Tube
Dielectric Breakdown
Disc Diameter
Disc Thickness
Dissipation Factor
Dynamic Resistance
Electromechanical Contacts
Electrical Characteristics
Electromechanical Switching
Electronic Ignition
EMP
Energy Conservation
Energy Calculations
Energy, Inductive
Energy, Pulse
Energy Pulse Testing
Energy Rating
Energy, Stored
Energy Withstand
Energy, 60 HzSurge
Encapsulation
Engine Shutdown
Engineering Evaluation
Environment, Automotive
Environmental Testi'ng
Environmental Reliability
Epoxy
Equipment, Test
Examples, Applications
Failure Definition
120,123,124
109
104
105,106
96
15,95
35
27,28
104
26
87,88
87
87
88
89
76
11
47,51
34,89
17
124
43,44,45
16
50,51,52
113
Failure, Insulation
Failure Modes
Failure, Semiconductor
Fault Conditions
FCC Standard
Field Decay
Field Failures
Field Maintenance
Field Testing
Filters
Fire Retardant
Flame Test
Flashover
Follow Current
Fuse Clearing
Fuse Selection
Fusing, Varistor
Gas Tubes
Gas Tube Protector
GE-MOV Varistor Properties
59
50,51
65
105
32
64
68
36
18
48
Grain Structure
Ground Currents
Ground Potential Rise
Hazardous Voltage
High Energy
16
75-77
11,17
76
16
111
15,96
17
84
31,34
High Reliability
Humidity Life
Idle Power
IEEE Standard
IEC Standard
159
63,64
62,63
109
122
123
107
123
16
18,113
78
34
17
48
44,45
108
52,53
11,12
50,51,52
16
97
13
18
61,62
10,11
124
110
61,109,110,126
10,15
126
127
48
16
101
y
95
127,150
128
48
114
71-86,91-93,96-100
120,124
17
66
16
11
116,117
96
120
102
102
32
112,127
112,127
13,17
34
66
66
66,129
34
90
43,44,45
43,44
87
91
129
39
150
128
37
116
17,19,28
Ignition, Electronic
Impedance, Source
Impulse Generators
Incoming Inspection
Inductance, leads
Induction Motor
Inductive load Switching
Inductive loads, Transistor Switching
Inductive Switching
Induction Motor
Impulse Waveform
In-Rush Current
Instrumentation, Detection
Insulation
Ionization
12t
Jump Starts
Junction
Kettering Ignition
lead Inductance
lead length
leakage Current
leakage Region
life, DC Power
life, Humidity
life, Storage
life Testing
lifetime Current Rating
line Voltage Varistor
lightning
lightning Strikes
lightning Transients
load Dump
load line
location Categories
loop Area
low Voltage Varistor
Magnetizing Current
Matching Characteristics
Measurement, Capacitance
Measurement, Clamp Voltage
Measurement of Characteristics
Measurement of Varistor Voltage
Mechanical Reliability and Integrity
Mechanical Shock
Mechanical and Environmental Testing
Measuring Transients
Metal Oxide Properties
Model, Equivalent Circuit- Varistors
Model Number
Model, Reliability
Motor Control
Motor Protection
Mutual Coupling
Noise Contacts
Noise Detection
Noise Generation
Noise Simulator
Noise Suppression
Nomenclature
Non-linear Resistors
Operation Theory
Oscilloscope
Outline & Dimensions
Overshoot Effect
Package Material Expulsion
Package Rupture
Package Type
Paralleling Varistors
Peak Current Rating
99
32,117,118
113
102
53
82
16
79
75
82
15
32
113,115
17
18
67
15
44,45
99
53,105
106
123
50-51
123
128
128
109,112
63
119
15,67
87
90
95,96,97
64
27
106
39
10
70
108
105
103
103
127
127
127,150
113,115
35,36,44,45
50,51
58
124
73
81,82,83
15,96
18
78
18
78
78
60
35,36,44
46
18
151-154
55,105,106,107
129
67,129
48,49
69,70,71
62,109
PNJunction
Power Circuits lAC)
Power Circuits lAC), Source Impedance
Power Circuits lAC), Waveforms
Power Contact
Power Cross
Power Derating Curve
Power Dissipation
Power, Follow
Power Induction
Power line lAC) Transients
Power Savings
Power Supply Crowbar
Power Supply Protection
Power, Standby
Product Series V /1 Curves
Product Series Pulse Rating Curves
Primary Protection
Product Family Guide
Product Ratings and Characteristics
Product Qualification
Proof Tests
Prospective Current
Pulse Current Testing
Pulse Current Waveform
Pulse Energy
Pulse Generator
Pulse Rating
Pulse Rating Curves
Pulse Test Stability
Punch-Through
QuaHfcation Testing
Quality Control
Random Transients
Rate Effect
Rating, Assurance
Ratings, Current vs Temperature
Rating, Energy
Rating, Peak Current
Rating, Power Dissipation
Rating Table
RCNetwork
RC Network, Snubber
REA Standard
Relay
Relay Switching
Reliability
Reliability, AC Bias
Reliability, DC Bias
Reliability, DC Power
Reliability, Environmental
Reliability, Mechanical
Reliability, Model
Reliability, Storage
Resistance, Dynamic
Resistance, Static
Response Time
Restrike
RFI Filter
SAE Test Circuits
Safety
Saving Power
Secondary Protection
Selection Guide
Selenium Suppressors
Semiconductor Degradation
Semiconductors, Transient Effects
Seriesing Varistors
Service Entrance
Shielding, Physical
160
16
21
1.5
15
91
91
64
63
34
91
13
18
84
71,72
37
143-146
147-149
91
65
137-141
102
38
68
110
57
124,126
114
63,80,81
147,148,149
125
124
102'
119
13
53-54
109
64
61,126
62,63
110,1,11
59
77
32
116
11
76
14,119
119,120
122
122
128
127
124
128
52
53
51,52,133
11
72
96
129,133
18
90
65,86
34
16
16
69,70,71
24
129
Shock Testing
Short Circuit
Silicon Carbide
Soak Time
Solenoid Driver
Solderability
Solderability, Reliability
Source Impedance
Source Impedance, AC Power Circuits
Spark Gaps
Speed of Response
Stability, long-Term
Stability, Pulse Test
Standards, Transients
Standby Power
Starting Currents
Static Resistance, Varistor
Storage Life
Storage Oscilloscope
Storage Reliability
Stored Energy
Stored Energy, Motor
Stress levels
Suppression Devices
Suppressors, Avalanche Diodes
Suppressors Comparison
Suppressor, Selenium
Suppressor, Zener
Surge Energy (60 Hz) Capability
Surge Impedances
Surge Testing
Suspended Cable
SWC
SWC Standard
Switch Arcing
Switching, Contacts
Switching, Inductive loads
Switching, Transformer
Symbols, Varistor
TCl
Telecommunications
Telecommunication Line Transients
Telephone Lines
Temperature Coefficient
Temperature Dependence
Temperature Derating
Temperature Effects
Terminal Strength
Terminology
Test Circuit
Test Equipment
Testing
Testing, Clamp Voltage
Testing, Current
Testing, Current Surge
Testing, Environmental
Testing, In-Field
Testing, Mechanical
Testing, Pulse Energy
Test System
Testing, Transient Power
Testing, Transients
Testing, Varistors
Test Objectives
Test Waves
Test Waveform
Thermal Runaway
Thevein Equivalent Circuit
Thermal Shock
Transmission Lines
150
124
35
103
77
150
127
32,117,118
15
34
53,54,133
121
125
19
37
83
53
128
18
128
10
82
119
31
35
36
34
34,35
127
89
110
88
15
117
11
75
75
10
55-56,57
15,117
87
15
15
52
107
64
51
127
55
78
113,114,115
38,39
104
107
110
t27,150
102
150
110
114
110
19
101
101
115
57
63,66
62
127
87
Transformer
Transformer, Magnetizing Current
Transients, AC Power Lines
Transient!?, Automobile
Transients, Cable
Transient Control level
Transient Detection
Transient Effects
Transient Environment
Transient Measurements
Transient Power
Transient Standards
Transient, Switching
Transient Testing
Transients, Random
Transients,Repeatable
Transients, Telecommunication
TV Receiver
Ul Recognition
Ul Standards
Upturn Region
Varistor
Varistor Capacitance
Varistor Terms
Varistor Circuit Model
Varistor Construction
Varistor Degradation
Varistor Failure
Varistor Fusing
Varistor Matching
Varistors, Metal Oxide
Varistor Rating
Varistor Reliability
Varistor Selection
Varistor Series (MA)
Varistor Series (Z)
Varistor Series (l)
Varistor Series (PA,HE)
Varistor Series (D,8)
Varistor Shielding
Varistor Stability
Varistor Symbols.
Varistor Terminology
Varistor Test System
Varistor Testing
Varistor Voltage Measurement
Varistor Voltage Ratings
Varistor Voltage, VN(dcl
Vibration
Vibration Fatigue
Vibration Variable
Volt-Ampere Characteristics
Volt-Time Curves
Volt-Time Response
Voltage Clamping
Voltage Dependent Device
Voltage, Hazardous
Voltage Rating
Waveform, Test
Zener Suppressors
Zinc Oxide
161
9,73
10
13
15
89
15,116
18
16
21-30
113,115
110
19
9
19
12
9
15
72
112,127,129
112,127
50,53,70
35
47,51
56
50
48
102
129
66,129
70
35,43
109
119
59,65,73,77;86
137
138
139
140
141
129
121
56,57
55
113
101
103
83
45
127
150
150
31,36
37,38
34
64
43
129
59,109
57
35
43
.b. !t~~_~ IE~:~~=~,tL~i~f~~~~~:~:S
UNITED STATES OF AMERICA
ALABAMA
INDIANA
NORTH CAROLINA
Huntsville 35801
Holiday Office Center
3322 S. Memorial Parkway
Suite 17
(205) 837-0411
Ft. Wayne 46805
Lakeside 1 Office Bldg.
Suite 225
2200 Lake Ave.
(219) 422-8551
Charlotte 28234
700 Tuckaseegee Rd.
P.O. Box 34396
(919) 379-8474
ARIZONA
Indianapolis 46268
6321 Lapas Trail
P.O. Box 68543
(317) 298-5317
Phoenix 85016
5320 North 16th St.
(602)241-7224
CALIFORNIA
San Diego 92138
. P.O. Box 85014
(714) 565-4114
Santa Monica 90405
342d Ocean Park Blvd.
Suite 1000
(213) 450-0353
Palo Alto 94304
1801 Page Mill Rd.
Suite 223
(415) 493-2600
COLORADO
Deriver 80201
201 University Blvd.
Mailing Address:
P.O. Box 2331,80201
(303) 320-3031
CONNECTICUT
Meriden 06450
1 Prestige Dr.
Mailing Address:
P.O. Box 910,06450
(203) 238-6894
FLORIDA
Palm Beach Gardens 33410
10800 N. Military Trail
Suite 207
(305) 622-8823
GEORGIA
Atlanta 30341
1835 Savoy Dr.
Suite 215
(404) 458-8401
ILLINOIS
Des Plaines 60018
2860 S. River Road
Suite 400
(312) 827-9100
Greensbciro 27408
2105 Enterpri$e Rd.
Mailing Address:
P.O. Box 9476,27408
(919) 379-8474
MARYLAND
OHIO.
Columbia 21046
Appliance Park East - Bldg. 1
(301) 992-7250
Cleveland 44132
26250 Euclid Ave.
(216) 266-2900
MASSACHUSETTS
Dayton 45439
3430 S. Dixie Hwy.
Mailing Address:
P.O. Box 2143
Kettering Branch 45429
(513) 297-3287
West Lynn 01910
40 Federal St., Bldg. 20
(617) 594-7270
MICHIGAN
Southfield 48034
25900 Telegraph Rd.
(313) 356-8000
MINNESOTA
Minneapolis 55435
4600 W. 77th St.
Suite 201
(612) 835-2550
MISSOURI
st. Lou is 63132
1530 Fairview Ave.
(314) 997-8437
NEW JERSEY
Fairfield 07006 '
420 Route 46
(201) 227-6050
NEW YORK
Albany 12205
11 Computer Dr., W.
(518) 454-2576
Jericho 11753
400 Jericho Tnpk.
(516) 681-0900
PENNSYLVANIA
(Philadelphia)
Wayne 19087
999 Old Eagle School Rd.
(215) 964-2991
Pittsburgh 15220
3 Parkway Center
Room 304
(412) 921-4134
TEXAS
Dallas 75234
14673 Midway Rd.
Suite 117
(214) 661-8582
Houston 77210
7100 Regency Square
Suite 128 - Box 4367
(713) 978-4347
VIRGINIA
Charlottesville 22901
1843Seminole Trail
(804) 978-5040
. WASHINGTON
Rochester 14618
130 Aliens Creek Rd.
Room 101
(716) 422-3230
Seattle 98188
112 Andover Park, E.
P.O. Box 88850
(206) 575-2865
Syracuse 13221
Bldg. 1, Room 223
Electronics Pk.
(315) 456~3421
WISCONSIN
162
Milwaukee 53202
615 E. Michigan St.
(414) 226-1625
Semiconductor Products Department
General Electric Company
Auburn , NY 13021
"'
400, 10/ 83 (§.RM} P
S upersedes 400.3 4/82
GENERAL
ELECTRIC
PrlntAld in U.S.A.
Source Exif Data:
File Type : PDF File Type Extension : pdf MIME Type : application/pdf PDF Version : 1.3 Linearized : No XMP Toolkit : Adobe XMP Core 4.2.1-c041 52.342996, 2008/05/07-21:37:19 Create Date : 2017:08:07 08:14:42-08:00 Modify Date : 2017:08:07 10:19:39-07:00 Metadata Date : 2017:08:07 10:19:39-07:00 Producer : Adobe Acrobat 9.0 Paper Capture Plug-in Format : application/pdf Document ID : uuid:3617f278-0caf-5a4f-8597-e503eec8c2fc Instance ID : uuid:cc9e4094-bb18-2b49-a3b4-0759be18e53e Page Layout : SinglePage Page Mode : UseNone Page Count : 159EXIF Metadata provided by EXIF.tools