SPD SPD

SPD

SPD SPD

SPD

SPD SPD Applications Manual

SPD
and Reference Guide
for Surge Protection and
SPD SPD Line Conditioning Products
SPD

SPD SPD

SPD

SPD SPD

SPD

SPD SPD

SPD

SPD SPD

SPD

SPD SPD

SPD

SPD SPD

SPD

® ®
INTRODUCTION
This reference guide provides a comprehensive background in Power Quality issues. A major focus of this guide is
on the causes and effects of transient voltage surges. It covers the strategies available to defend against disruptive
and damaging effects of surges using Surge Protective Devices. It describes critical selection criteria for Surge
Protective Devices including pertinent standards, performance characteristics, and proper installation techniques.
The intention is for the reader to have the information needed to properly implement efficient, cost-effective SPD
installations in a broad range of application environments.

TABLE OF CONTENTS
Chapter 1: Power Quality Pages 2–11
A Perspective
AC Power Basics
FIPS PUB 94
IEEE Standard—1100 Emerald Book
Power Conditioning
Transient Voltage Surges
Regulation
Uninterruptible Power Supplies (UPS)/Standby Power Supplies (SPS)
Noise Isolation
Grounding
Harmonics

Chapter 2: Transient Voltage Surges Pages 12–22

Introduction
Sources of Transients
Effects of Transients
Basic Operation of Surge Protective Devices (SPD’s)
Metal Oxide Varistors (MOV’s) and Other SPD Components
Filters
Characteristics of SPD’s

Chapter 3: SPD Performance Pages 23–33

Overview
IEEE C62 Standards
IEEE C62.41
IEEE C62.45
UL 1449
IEC Standards
The Importance of Product Warranties

Chapter 4: Specifying SPD’s Pages 34–37

Overview
NEMA LS-1 Specification Format Standard

Chapter 5: Protection of Telecommunications and Data Lines Pages 38–46

Chapter 6: Surge Protection for Factory Automation Pages 47–48

Chapter 7: SPD Application Strategies and Installation Practices Pages 49–61

Cascading/Networking Schemes

Chapter 8: Reference Pages 62–64

Industry Standards
Glossary of Terms

1
 Chapter 1: Power Quality

A Perspective
The power quality problems we’re concerned with today are not new to any facility’s AC powerline environment.
Older electro-mechanical equipment, by and large, did not contain microprocessor-based (meaning “computerized”)
circuitry. This older equipment easily endured all but the most severe power quality problems without any resultant
damage or downtime. However, since the late 1960’s, every aspect of business and manufacturing increasingly
depends on some type of microprocessor-based equipment.

Today, the power quality infrastructure has a significant effect on the bottom line in any business. The March, 1993
issue of Electric Light and Power estimated that on average, equipment downtime alone cost each Fortune 1000
company $3.48 million annually. EL&P further estimates that electrical disturbances cost U.S. companies a total of
$26 billion.

The bar graph in Figure 1, published in MIS Week,

estimated that given the current state of power quality
in commercial buildings, data processing downtime
attributed to power quality will increase from 27% in
1990 to 47% by the year 2000. Note that downtime
due to the electrical environment was only 5% in
1980. The reason for this dramatic increase in down-
time is not that the power quality infrastructure is get-
ting markedly worse. Computer chips were much
slower in 1980 and operated at relatively high logic
levels of 5 volts or greater. In the 1990’s, high speed
data transmission is accomplished by computer chips
that operate at lower logic levels. These chips are
even more sensitive to slight changes in voltage lev-
els or transients that appear on the AC powerline. In
other words, accelerated computer technology means
an increased vulnerability to the power quality infra- FIGURE 1
structure.

Power quality problems were first addressed in the 1970’s, when the main goal was to protect sensitive electronic
equipment. In the 1980’s, the emphasis went beyond equipment loss to productivity losses. Today in the 1990’s
when every aspect of manufacturing, sales, customer
service, shipping and finance depends so much on
computer-based technology, the loss of revenue is Estimated LAN Outages
staggering when vital computerized equipment is
(In millions of dollars)

yyyy
,,,, yyyy
,,,, ,,,,
yyyy
down. 4

A study by Infonetics Research on the effects of out-

ages on Local Area Networks or LAN’s illustrates this 3
,,,,
yyyy
,,,,
yyyy ,,,,
yyyy
,,,,
yyyy ,,,,
yyyy
,,,,
yyyy
point. Figure 2 shows that revenue losses in 1989
were a fraction of the productivity losses. Only four
,,,,
yyyy
,,,,
yyyy ,,,,
yyyy
,,,,
yyyy ,,,,
yyyy
,,,,
yyyy
,,,,
yyyy ,,,,
yyyy ,,,,
yyyy
2
years later in 1993, revenue losses begin to overtake
productivity losses.
,,,,yyy
yyyy
,,,, ,,,,,,,,
yyyy ,,,,
yyyy
yyyy,,, ,,,,
yyyy yyyy
1

With the 21st century just a few years away, technol-

ogy may be accelerating at such a pace that there O ,,,,
yyyyyyy
,,,
Productivity Revenue
,,,,
yyyy,,,,
yyyy
Productivity Revenue
may be a need for a National Power Quality Code. In Losses Losses Losses Losses
the meantime, facility managers and design specifiers
must be conscious of the serious impact power quality 1989 1993
has on keeping vital technology up and running. FIGURE 2

2
AC Power Basics
Any discussion of power quality should begin with a brief refresher on AC power basics. Utilities supply electrical
power as ALTERNATING CURRENT, which is usually referred to as AC power. An oscilloscope is an instrument used
to create a graphic image of the generated line voltage. Figure 3 represents the image an oscillo- scope would show
of a single 120 Volt AC cycle. Sixty of these cycles
occur every second. However, rather than saying this
AC frequency is 60 cycles per second, it is referred to The AC Sine Wave
as 60 Hertz, which is usually written in abbreviated
form as 60 Hz. +170V peak
(+120V RMS)
The shape shown in Figure 3 is called a sine wave.
Notice that in the first half-cycle above the line, the volt-
age level rises to 120 Volts. In the second half of the
cycle below the line, the voltage is -120 Volts. So the
voltage and current alternate polarity each half cycle, 90° 180° 270° 360°
hence the term ALTERNATING CURRENT, or AC.

The 120 Volt AC sine waves shown in Figures 3 and

4 represent an ideal condition where perfect, consis- -170V peak
tent, stable AC power is shown. In the real world, this (-120V RMS)
1 CYCLE
is seldom the case. AC line power is subjected to all 1/60 second or 16.7 milliseconds

sorts of unwanted disturbances that produce a variety FIGURE 3

of effects.
200
Commercial Power Envelope
(170 Vpeak +/- 15 percent)
Modes of Measurement
VOLTAGE (Peak)

100

Changes or disturbances of the AC power sine wave

are generally specified in terms of voltage. Voltage is 0

a potential measured between two specific points. In

AC analysis, rather that repeatedly redefining these -100
measurement points, the mode of measurement is
identified. For the three-wire, single-phase branch cir- -200
cuit distribution system commonly used in the U.S. 0° 90° 180° 270° 360°
and Canada (Black wire = Line or Hot, White wire = PHASE ANGLE (Degrees)
Neutral or Common, Green wire = Safety Ground), VOLTAGE: 120 Vrms or 170 Vpeak
(Vpeak = Vms x √2)
the power transfer potential, L - N, is called the
Normal Mode. Sometimes this mode is also called FREQUENCY: 60 Cycles per second
(1/ƒ = 16.67 ms per cycle)
the Transverse Mode.
Likewise, any voltage measurement taken with FIGURE 4
respect to Ground, N - G and L - G, represent
Common Mode potentials. Common mode potentials
are sometimes abbreviated L + N - G, or may be Modes of Measurement
more carefully specified by L- G and N - G. The
modes of measurement for a standard NEMA L5-15 NORMAL
L-N
receptacle are shown in Figure 5.

N L
N-G COMMON COMMON L-G

FIGURE 5

3
FIPS PUB 94
The first official study of the effects of AC powerline disturbances on electronic equipment was published in 1983 by
the U.S. Department of Commerce/National Bureau of Standards. Federal Information Processing Standards
Publication 94, commonly known as FIPS PUB 94, emphasized the importance of the electrical environment for ADP
(Automated Data Processing) installations. FIPS PUB 94 included a list of power quality attributes that still serves as
practical guide to acceptable limits for computerized equipment.

FIPS PUB 94
8.2 Some Representative Power Quality Attributes

Environmental attribute Typical environment Typical acceptable limits for computers and power sources

Normal Critical Units affected and comments

Line frequency ±0.1%-±3% ±1% ±0.3% Disk packs, tape, regulators

Rate of frequency change 0.5-20 x 1.5 Hz/s 0.3 Hz/s Disk packs

Over and undervoltage ±5%- +6,-13.3% +5%,-10% ±3% Unregulated power supplies

Phase imbalance 2%-10% 5% max 3% max Polyphase rectifiers, motors

Power Source: Less than

Tolerance to low 0.85-0.6 lagging 0.8 lagging 0.6 lagging Indirectly limits power source
power factor or or requires greater capacity
0.9 leading unit with reduced overall
efficiency

Tolerance to high steady- 1.3-1.6 peak/rms 1.0-2.5 Greater than 1.414 normal; departures
departures state peak/rms 2.5 peak/rms cause wave shape distortion.
peak current

Harmonics (Voltage) 0-20% total rms 0-10% total 5% max total Voltage regulators, signal
5-10% largest 5-10% largest circuits

dc load current Negligible to 5% or Less than 0.1% As low as 0.5% Half wave rectifier load can
capability of power source more w/exceptions saturate some power sources,
trip circuits.

Voltage deviation 5-50% 5-10% 3-5% Affects regulators, signal

from sine wave circuits

Voltage modulation Negligible to 10% 3% max 1% max Voltage regulators, servo

motors

Transients surges/sags +10%,-15% +20%.-30% +5%,-5% Regulated power, motor torque

Transient impulses 2 to 3 times nominal Varies; Varies; Memory, disks, tapes having
peak value 1,000-1,500 V 200-500V high data transfer rates, low
(0-130% V-s) typical typical level data signals

RFI/EMI and “tone bursts” 10 V up to 20 kHz; Varies widely Varies widely Same as above
normal and common modes less at higher freq. 3 V typical 0.3 V typical

Ground currents 0-10 A rms 0.001-0.5 A 0.0035 A May trip GFI devices, violate|
+ impulse noise current or more or less code, introduce noise in signal|
circuits

FIGURE 6

4
IEEE STD-1100 “Emerald Book”
In 1992, The Institute of Electrical and Electronic Engineers (IEEE), issued Standard 1100, known as
the “Emerald Book”. Standard 1100 is a “Recommended Practice for Powering and Grounding
Sensitive Electronic Equipment”. The Emerald Book is widely respected throughout the electrical
industry and addresses power quality issues that every facility needs to consider. Figure 7 is a
Selection Guide to Power Conditioning Technologies as suggested in the Emerald Book.

Selection Guide to Power Line

Surge Noise Filter Standby Standby
Conditioning Technologies Protection (EMI/RFI)
Conditioner Voltage Power UPS Engine
(Isolation Regulator System
Device Transformer) Generator

Surge Common Mode ✓

✓ ✓
Normal Mode ✓
✓
Noise Common Mode ✓ ✓
✓
Normal Mode ✓
Sag ✓ ✓
Swell ✓
Undervoltage ✓
Overvoltage ✓
Momentary Interruption
✓ ✓
Long-Term Interruption ✓ FIGURE 7
✔ Leviton Products ✓ The level of resolution to power problem will vary from manufacturer to manufacturer

Power Conditioning
Power Conditioning is a broad and complex
industry dealing with the elimination and pre- Facility Load
Generation Transmission
vention of power disturbances to the AC power Distribution Service
sine wave environment. Though the problems
domain may extend from point of generation to
Power Conditioning
load service, the principle focus of the power Area of Focus
conditioning industry begins at the facility ser-
vice entrance. FIGURE 8

Disturbances on the AC power line take many forms, each with its
own unique symptoms and solutions. However, despite their variety Power Quality Cube
and complexity, power quality problems can be grouped in several Regulation
distinct categories as represented by the illustration in Figure 9. Noise Harmonics

The Power Quality Cube is an analogy emphasizing the uniqueness

of each power conditioning area. Each face is distinct and separate,
yet each adjoins the others at an edge or corner. Likewise, the areas
of power conditioning are distinct and require individual, specialized
solution technology and methodologies, yet they are related both in
terms of environment and result.
The power conditioning areas of REGULATION, ISOLATION and Suppression Isolation

SUPPRESSION are the most often mentioned in the major power

quality studies published during the last 20 years. However, these
Grounding
are not the summation of all power quality problem areas. The other
faces of the cube represent the areas of GROUNDING, HARMON- FIGURE 9
ICS and NOISE.
The electrical professional must have some understanding of the types of disturbances that are likely to occur
in the AC powerline environment, as well as the power-conditioning technologies developed to deal with each.
5
Transient Voltage Surges
Transient voltage surges are sometimes called
“spikes”, and by general definition, they are short-
Voltage Trainsient
term deviations or changes from a desired voltage
level (or signal in the case of computers and electron-
ic devices). The deviation can cause an electronic
device to malfunction, or damage it outright.
+120V RMS Amplitude
is typically
Figure 10 shows a transient voltage surge on the 60 double the
Hz AC sine wave. Notice that the transient is very 0 line voltage
brief in duration, and the surge voltage is much high- or greater
-120V RMS
er than the AC line voltage. In fact, the greater the and less than
one cycle
amplitude of the surge voltage, the greater the risk of in duration
it damaging electronic equipment. Furthermore, tran-
sient voltage surges originate from a variety of
sources, so there is no such thing as a transient-free
AC system. Transients are also present on telephone FIGURE 10
lines, coax conductors and a variety of other electrical
environments. Following this chapter, the balance of this reference guide will concentrate on a thorough examination
of transients, and the application of surge protective devices to defend against their disruptive and damaging effects.

Regulation
Voltage Regulation problems are familiar to most people because they occur over a relatively long time period elec-
trically speaking, greater than 1 millisecond, and the effects are discernible to the human eye. Voltage surges which
qualify in the regulation area are generally defined as having a duration of from one quarter cycle to many cycles.

SWELL is defined by the Institute of Electrical and

Electronics Engineers (IEEE) in a publication often Line Voltage Swells and Sags
referred to as the “Emerald Book”. The IEEE Std.
1100 in this publication describes swells as an
increase in the AC voltage, at the power frequency, SWELL

for durations ranging from a half cycle to a few sec-

NORMAL
onds. Figure 11 shows normal AC line voltage in the +120V RMS +120V RMS
SAG
left-hand portion of the sine wave. The center portion
represents a swell, as can be seen by the increased SAG

amplitude of the cycles. -120V RMS -120V RMS

NORMAL
SYMPTOMS: CRT and lights brighten up temporarily,
power supply components blown out, burn marks on SWELL

boards and components.

SAG is defined in IEEE Std. 1100 as a reduction in

FIGURE 11
the AC line voltage, at the power frequency for dura-
tions lasting from a half cycle to a few seconds. The right-hand portion of Figure 11 shows a sag, as can be seen by
the reduced amplitude of the sine waves.
SYMPTOMS: CPU reboot required, motors and fans lug down or stop, lights and CRT dim or sag in and/or blink off.

Blackouts are power outages, meaning a complete absence of line voltage.

SYMPTOMS: No power, system down.

6
Uninterruptible Power Supply (UPS)/Standby Power Supply (SPS)
The solution to swells, sags and blackouts is voltage regulation. The fundamental requirement for voltage regulation
is that energy is added or subtracted to maintain the integrity of the AC sine wave. This is accomplished by either a
Voltage Regulating Transformer, an Uninterruptible Power Supply (UPS), or a Standby Power Supply (SPS). The
most popular choices today for solving Regulation problems are microprocessor-controlled UPS and SPS units.

Basically, the UPS or SPS is equipped with a switching mode power supply operating off a battery. The battery is
charged while utility power is present. In the event that utility power causes a Regulation problem, an automatic
transfer switch allows the battery to deliver power to the load through a DC-to-AC inverter. Note that the UPS or
SPS deliver “synthetic” AC power (not a perfect sine wave) to their connected loads.

Figure 12 is a block diagram of the ON-Line or true UPS. Figure 13 similarly shows the SPS, which is sometimes
called the BATTERY BACK-UP or OFF-LINE UPS. There are many options available to both the UPS or SPS design-
er, and as a result there is a wide range of both performance specifications and price. A careful examination of perfor-
mance parameters is advised, because a UPS or SPS may only function as a sag-blackout power supply, i.e. it may
not be capable of any degree of overvoltage regulation, suppression, isolation, or even minor sag regulation.

Off-Line or Battery Back-up UPS

AC POWER NC

AC AC
POWER POWER

LOGIC

INPUT OUTPUT

NC
BATTERY
BATTERY CHARGER TRANSFER
SWITCH INVERTER
AC to DC DC to AC
FIGURE 12

On-Line UPS

AC POWER

AC AC
POWER POWER

LOGIC

INPUT OUTPUT

NC NC
BATTERY
BATTERY CHARGER INVERTER TRANSFER
AC to DC DC to AC SWITCH

FIGURE 13

7
Standby Emergency Power
In a facility where Standby Emergency Power is utilized, it is
important to consider the automatic transfer switch that allows
the motor-generator or engine to deliver stored power to the
load. The response of this switch may actually generate tran-
sients on the AC powerline connection to sensitive microproces-
sor-based equipment. It is therefore important to evaluate the
parameters for the automatic transfer switches incorporated in
these power supplies.

Noise Isolation
Sources of noise energy can include motor noise, high-
frequency energy placed on the powerline by thermal heating
devices, copying machines, ballasts and other equipment.
Facilities near broadcast radio and TV towers, or near airport
radar sites, can suffer from noise due to high levels of electro-
magnetic energy radiating from these sources.

Power Conditioning Area: Noise Isolation

Problems: Solutions:
Noise attenuation is usually accom-
NORMAL MODE NOISE plished via filter technology. Hence,
SYMPTOMS: CRT data line filters and specially shielded
bounce, static chatter, "Isolation Transformers" are com-
light and LED flicker, monly used. Since noise is low in
data disruption energy conyent, noise attenuation
technology is not adaptable to high
energy applications.

COMMON MODE NOISE

Description: SYMPTOMS: Data disruption,
"Noise on the line" logic problems
implies the presence
Amplitude: Low
of a low energy,
Frequency: High
random signal of
Energy: Low
higher freqency.

FIGURE 14

8
Grounding
Grounds are a necessary element of any power, signal, or data network. All voltages and signal levels are referenced
to ground. For example, when voltage measurements are made, they’re most often referenced to ground potential.

Service Entrance Power Service Entrance Conductors

Service
Mast Utility
Power
Pole
3Ø
Distribution Meter
Grounding
Metal Transformer
Case Service Head
Neutral 117 V 117 V
120 V Insuulated Grounded Conductor

120 V Service Neutral

Mast Breaker
Load

Bus

Bus

Bus
Equipment
Neutral Grounding Conductor Two-wire
Single Phase
Messenger disconnect Grounding Conductor
Branch Circuit
Breaker
link

Bus
Load
Service Drop Equipment
Grounding Conductor Three-wire
Single Phase
Grounding Strap
Grounding Meter Uninsulated
Feeder Circuit
Conductor Equipment
Grounding
Bus

Protection Neutral
Main Bonding
Grounding Conductor Jumper
Grounding Strap
Service Service Entrance Enclosure
enclosure

Clamp

Grounding
Conductor

Grounding
Grounding electrode / Grounding Systems
Electrode
Grounding
Electrode/Grounding
System

Service Entrance Grounding Bus Typical Facility Power Distribution

To Load

To Service Drop

Neutral
Bus Up to Six Meter Feeder
Disconnecting
Means. Subpanelboard or
Three are Lighting Panalboard
Panelboard
Shown

Feeder
Branch
Service Circuits
Entry
Neutral

Meter
Service
Entrance Subfeeder
Grounding Grounding Bus or Main
Conductor Switchboard

Grounding
Electrode / Grounding System Branch Circuits Branch Circuit

The use of a grounding bus in a service entrance

9
Ground Potential Difference
When changes to ground potential take place, they can cause serious damage or disrupt the operation of electronic
devices. That’s because current flows whenever a difference in potential exists. And this effect can occur in different
ground systems within a single facility, or between the ground systems in separate buildings.

It happens that not all ground systems are at zero volts. In fact, the difference in potential between grounds causes
current to flow in a so-called ground loop as shown by Figure 15.

These ground loops can adversely affect computer

data networks. If one computer is connected to a
power line in one building, and joined by a data net-
work to a second computer in a different building, it is
very likely that the ground systems of the two build- Service Service
Entrance #1 Ground Loop Entrance #2
ings are not at the same equipotential point. Panel Current Panel
Consequently, utility current will flow in the data line
at a level that is inversely proportional to the resis-
tance of the conductors that join the two computers. Computer Computer
System 1 System 2
For example, if the two computers are connected via
coaxial cable with a shield resistance of .05 Ohms,
and the potential difference between the two ground
+3V +1V
Ground Rod System Ground Rod System
systems is 2 Volts, a current of 2/.05 Amps, meaning
40 Amps, will flow on the shield of the coax. Such a Ground Loops Caused by Poor Facility Ground Systems
high current flow can easily cause false signals or
erroneous data pulses. FIGURE 15

Computers rely totally on specific voltage levels to indicate the presence or absence of a data pulse. When ground
loop currents occur, the noise they produce causes false logic commands within the computer’s logic circuits. This
type of anomaly CAN NOT be remedied by the use of surge suppressors. The problem can only be corrected with a
solution that reconciles the voltage differences between two ground systems.

Unfortunately, most facilities really don’t have very good ground systems. First, they often start out with an underper-
forming ground grid. Second, the grounding system may be sized to cope only with low-frequency (meaning 60Hz)
power grounds rather than the high-frequency currents present within lightning strikes. Third, the grounding system
has not been adequately maintained, or it may have degraded or been damaged over the years as a result of
changes made to accommodate other building modifications.

There are several fundamental principles that must be taken into account when planning a grounding system.

SOIL CONDUCTIVITY — is perhaps the most important parameter. Soil conductivity is measured in Ohm-meters,
and the lower it is, the better the ability of the soil to conduct away unwanted lightning surges and leakage currents.

PHYSICAL GEOMETRY OF THE GROUND GRID — meaning length and width, will determine the overall imped-
ance of the entire grid. The lower the impedance, the better the grid will be for conducting away heavy currents. An
overall grid impedance of 5 Ohms or less is desirable for most computer systems.

Other design considerations include inductance of the cable going to the ground system, inductance of the intercon-
nection cabling, and skin effect, which is the tendency of high-frequency currents to flow along the outer surface of a
conductor.

Good design practice in designing ground systems seeks to minimize skin effect and inductance, as these parame-
ters are closely related. There are many reference texts that provide very thorough information on all aspects of
grounding systems and their design. The topic has been briefly noted in this manual because it’s an important con-
sideration in applying a variety of surge-protection strategies.

10
Harmonics
Harmonics are voltages or currents with frequencies
that are integer multiples of the fundamental power
frequency. In the case of the AC powerline environ- Normal AC Sine Wave
Voltage or Current
ment, the fundamental frequency is 60 Hz. The sec-
ond harmonic would be 120 Hz; the third harmonic
180 Hz, and so on. Harmonics occur on the AC pow-
erline whenever the sine wave shape is distorted.
Figure 16 shows the pure AC sine wave, the line volt-
Line Voltage
age with harmonics, and the line current with har- with Harmonics
monics as they would appear on an oscilloscope.
Harmonics are caused by non-linear loads within an
AC power distribution system. Linear loads, such as
a resistive heating element, do not cause harmonic
distortion and the AC current that flows will be a rela- Line Current
with Harmonics
tively pure sine wave. However, if the load is non-lin-
ear, drawing short bursts of current each cycle, the
current wave shape will be non-sinusoidal and har-
monic currents will flow. The total resultant current AC Sine Wave with Harmonics
will be a combination of the fundamental frequency
plus each of the harmonics. FIGURE 16

Switching Power Supplies Generate Harmonics

Power supplies that use semiconductors to switch the line current on and off abruptly during each AC cycle generate
harmonic currents. Switching Power Supplies, sometimes called Electronic or Solid State Power Supplies, are used
in a wide variety of modern electronic equipment found in every health care, commercial and industrial facility. Solid
state ballasts are also used in fluorescent lighting systems. Of particular concern is the fact that all personal comput-
ers, printers and microprocessor-based equipment use switching type power supplies.

Problems Caused By Harmonics

Line voltage harmonics can radiate interference into telephone and communication systems. They can cause over-
loading and malfunctions in circuit breakers, conductors, bus bars, panels, transformers and generators that are
designed to primarily handle 60 Hz. loads. Standby generators may overheat and/or experience internal control cir-
cuit malfunction due to harmonics produced by the very microprocessor loads they’re connected to. Certain motor
drives are particularly vulnerable to higher frequency harmonics and must operate in an environment where the Total
Harmonic Distortion is less than 5%.
Switching supplies use diode-capacitor circuits to convert AC line voltage to lower voltage DC. These large capaci-
tors charge up with narrow pulses of current that are timed with the peak line voltage. This process generates odd
harmonics which are mostly the 3rd and 5th, and to a lesser degree the 7th, 9th, etc. Note that the 3rd and 9th har-
monics will algebraically add in the neutral conductor of a 3-phase distribution system, causing conductor overload-
ing and transformer heating.

Reducing Line Voltage Harmonics

Ironically, the ever-increasing quantity of microprocessor-based equipment used in any given facility may generate
enough harmonics to create a significant power quality problem. Line Filters are valuable devices for reducing har-
monics, and will be discussed in more detail in the next chapter. Another consideration for facility electrical engineers
is lowering the power source impedance in the electrical distribution system. Larger transformers, larger and shorter
conductors, and better connectors will reduce harmonics. Designing for a good power quality Infrastructure requires
that wherever possible, harmonic-generating loads do not share branch circuit wiring with important motor loads.
The IEEE Standard 519, “IEEE Recommended Practices and Requirements for Harmonic Control in Electrical Power
Systems”, is a valuable resource for information on Harmonics.

11
 Chapter 2: Transient Voltage Surges

INTRODUCTION
Power Line Voltage Disturbances
Transient voltage surges are short-term over voltages, Incidence Rate Chart (1974)
usually measured in milliseconds. They are some- Source: Allen-Segal IBM Study 1974
times called “spikes”, with little or no distinction made
between the two terms. However, some reference
Surges
material may define a Spike as a transient that is less 49.0%
than twice the line voltage, and a Transient Voltage
Surge as a transient that is at least twice the line volt- Under/Over
11.0%
age or greater. Regardless of the terminology, they
Outages
are unwanted bundles of electrical energy in AC 0.5%
power lines or communications lines. The energy con-
tent of transients can be enormous, and this unwant- "Spikes"
39.5%
ed energy can damage equipment or cause it to mal-
function. Equipment driven by microprocessors is
especially vulnerable to transient voltage surges. FIGURE 17

Figure 17 shows just how much of a problem voltage

transients present to the Power Quality Infrastructure.
This pie chart is based on a 1974 IBM study. The
study was revisited by National Power Lab in 1994 Current Flow
and yielded very similar results. The chart shows that
11% of the disturbances are due to under/over volt-
- +
ages, which we’ve also called swells and sags.
Spikes account for 39.5% of the disturbances, while
transients account for 49%. With transient voltage
surges accounting for almost 90% of power line volt- Magnetic Field
age disturbances, it is easy to see why we’re empha-
sizing surge protection in this reference guide.
FIGURE 18

Sources Of Transients Transient Magnetic Field Coupling

L
Transients occur whenever line current is interrupted.
N
Therefore, transients can originate inside a facility, or
G
come from the outside utility lines. Sources within a
facility include loads that are switched, for example, a
motor that’s turned on and off. Utility grid switching
can cause transients that originate outside. In addi-
tion, differential ground potentials can be a source of
transients. (See Chapter 1, “GROUNDING” for more
information.) SPD

Another frequent though less obvious source of tran-

sients is inductive coupling. Whenever electric current Lack of Downstream SPD Protection Due to Cable
Resistance and Inductance. Trainsient current flows through
flows, a magnetic field is created. Figure 18 shows electronic system much more readily than through line impedances
the magnetic field around a conductor in which current (inductance) back to the SPD.

is flowing.
FIGURE 19

If this magnetic field extends to a second wire, it will induce a voltage in that wire. This is the basic principle by which
transformers work, where a magnetic field in the primary induces a voltage in the secondary.

In this same way, wires that run adjacent to one another within a building can magnetically couple transients, as
shown in Figure 19.

12
Lightning
But there’s a more dramatic case of magnetic coupling,
and it’s lightning. As Figure 20 shows, a lightning bolt
striking the ground has an enormously powerful mag-
netic field. This field will produce a spectacular tran-
sient voltage surge in nearby power lines by means of
magnetic coupling. The lightning bolt doesn’t have to
actually strike the utility lines. All that’s required is for
the electrical lines to be within the magnetic field. Inductive Lightning - Induced Charge:
DOES NOT require direct strike Transient
to power lines Current
Lightning’s effect on Power Quality is a much bigger
problem in some areas of the country than in others.
Figure 21 is an Isokeraunic Map that shows the frequen- Magnetic Field / Inductive Coupling
Magnetic Field/Inductive Coupling
cy of lightning storms throughout the United States.
FIGURE 20

Where Transients Occur In

The Power Quality
Infrastructure
Figure 22 shows a 1993 study
from Florida Power that breaks
down the source of power quality
problems in the business place.
Lightning accounts for 15% of the
problems. Utility sub-stations may
introduce transients due to grid
switching. The important observa-
tion here is that 60% of transients
are generated by your own office
equipment, and 20% of the tran-
sients come from equipment in the
office next door or equipment com-
mon to whole building such as
HVAC systems. The Florida Power
study is just one of several such
studies that show that the main
source of power quality problems is
FIGURE 21 inside the building.

Transient can be present on any metallic conductor

Power Quality Problems including utility power lines, telephone, data, and sig-
in the Business Place nal lines. Examples of data lines are:
Utility 5.0% ■ Local Area Networks, or LAN’s, using
Lightning
15.0% formats such as RS-232, RS 422, Ethernet and
Token Ring
Neighbor
Office Equip. 20.0% ■ Cable and closed-circuit TV
60.0%
■ Fire and surveillance alarm systems
■ CNC/machine-tool interfaces, especially
non-shielded lines

FIGURE 22

13
THE TRANSIENT ENVIRONMENT

The Transient in the AC Environment Types of Transients

Transients may be thought of as “bundles of energy”
riding on the AC power sine wave; a potential seeking
a pathway to ground. Because they are of short dura-
tion with discernible end points, they are usually
described in terms of their energy content (Joules)
rather than power (energy per unit time, Joules/sec or
watts).
Transient overvoltages, which may be positive or neg-
ative in potential and be positioned at any phase
angle on the wave, originate from a variety of
sources, though it is estimated that only 20 percent of
all transient activity is generated outside a facility.

Oscillatory or Ringwave Transient Spikes can occur randomly on the sine wave
CHARACTERISTICS: Fast rise time with oscillating
exponential decay.
SOURCES: Inductive loads such as elevators,
1.0
copiers, welders, air conditioning equipment, fuse RISE TIME = 0.5µs
V(t)/Vp
clearing, motors and tools of all types. 10µs
0.6
0.4
The Ringwave transient is generally the result of inter-
0.2
nal electrical activity. Its amplitude an energy content
0.0
are determined by its source and environment. The
- 0.2
Ringwave may also be the residual product of an
- 0.4
externally generated impulse and its resultant interac-
- 0.6
tion with the electrical distribution system.
- 0.8
- 1.0
Impulse or Unidirectional Transient - 1.2
0 10 20 TIME,µs 30
CHARACTERISTICS: Fast rise time, slower decay,
high energy content. 100 kHz Ring Wave

SOURCES: Lightning, utility grid switching, industrial

accidents.
1.0
Sometimes called a spike, the Impulse transient if
often defined by both a voltage and a current wave- V(t)/Vp
0.8
form. This takes into account the open circuit of pre-
arc condition (the voltage waveform) and the short cir-
cuit or a condition (the current waveform) of impulse 0.6 RISE TIME = 1.2µs
behavior. For this reason, the impulse “waveform set” DURATION = 50µs
is often called the “Biwave” impulse. The waveform is 0.4

less common than the Ringwave and is thought to be

indicative of externally generated transients. 0.2

0.0
0 20 40 60 80 TIME,µs 100

Combination Wave, Open-Circuit Voltage

FIGURE 23

14
When Can Transients Strike?
Since transients originate with interruptions in current flow, and also by magnetic coupling, they can occur any time
of the day or night. In some cases, it may be necessary to identify the source of transients within a facility. To do this,
some sort of power line monitor is generally used. A number of models are available to record the amplitude, dura-
tion and frequency of transient activity on the line being checked. If the choice is made to use a power line monitor,
sufficient time must be allowed to build an accurate picture of electrical disturbances. Generally, a period of 3 to 6
weeks is needed to establish a useful base of data.

Measuring Transients
As a result of their brief duration, transient pulses can be very high in frequency. That means a typical volt meter
won’t measure them properly because of its limited upper-frequency response capability. What’s needed is an oscil-
loscope or power line monitor with a very fast sampling rate and high upper-frequency response, meaning a mini-
mum of 100-150 MHz bandwidth.

Effects Of Transients
Microprocessor-driven devices can be found in practically
every commercial, industrial and residential setting. A brief
list of electronic equipment includes computers and their Signal amplitude of Operating Systems
peripherals, computer data networks such as LAN’s, med-
ical diagnostic equipment, CNC production machinery,
12V
telecommunications equipment, stereos, televisions,
microwave ovens, bar-code scanners, electronic cash reg-
isters, copy machines, FAXes, security and alarm sensor
equipment, and thousands of others. 5V
All this equipment is especially sensitive to transient volt- 4V
3.3V
age surges because of certain characteristics common to
integrated circuits and IC chips.
1970 1980 1990 2000
SPACING WITHIN THE INTEGRATED CIRCUIT — Most
of the spacing between components of an integrated cir-
cuit is substantially less than the thickness of a human FIGURE 23B
hair. The methods for producing power and signal circuit
paths (called tracks) in an integrated circuit also produce microscopic self-supporting structures. These structures
can become overheated and then sag when hit with surges. Once this happens, tracks which should be isolated can
touch, thereby causing internal shorts that render the IC useless.

APPLIED OPERATING VOLTAGE LIMIT — In striving to extend operating time, computer manufacturers are design-
ing machines with lower operating voltages to allow the use of lower-voltage batteries as shown in Figure 23B. Many
older computers use approximately 5 Volt DC logic levels. Current designs use 3.3 Volt DC, and future units will use
even lower voltages. The result is that any spike above 3.3 Volts that makes its way into the logic IC’s can cause dis-
ruption or permanent damage.

INCREASES IN COMPUTER OPERATING SPEED — The internal heart of a computer is called the clock. Faster
computers have faster clock speeds so, for example, a 33MHz machine is faster than a 16 MHz unit. At these
speeds, electrical noise becomes a threat. When noise enters a computer, it can mimic clock frequencies, and can
be mistaken for a valid logic command. When the computer acts on this false logic command, the keyboard can lock
up, or some other undesired action occurs. Also noise can cause the computer to miss valid operating commands or
clock pulses. If that happens, the computer creates erroneous output, or no output at all.

15
Harmful Effects: The “3 D’s”
The most common failures produced by transient within electronic devices
are disruptive, dissipative, and destructive.

DISRUPTIVE EFFECTS — are usually encountered when a transient enters the equipment by inductive coup-
ling. The energy source for this inductive coupling can act on the data output lines that integrate an electronic
installation. The electronic components then try to process the transient as a valid logic command. The result is
system lock-up, malfunctions, erroneous output, lost or corrupted files, and a variety of other undesirable effects.

DISSIPATIVE EFFECTS — are associated with repeated stresses to IC components. The materials used to fab-
ricate IC’s can withstand a certain number of repeated energy level surges, but not for long. Long-term degrada-
tion begins, and sooner or later, the device fails to operate properly for no apparent reason. Actually, the failure
is due to the cumulative build-up of transient-created stresses which have resulted in arc-overs, shorts, open
circuits, or semiconductor junction failures within the IC.

DESTRUCTIVE EFFECTS — include all conditions where transients with high levels of energy cause equipment
to fail instantaneously. Very often, there is actual physical damage apparent, like burnt PC boards, melting of
electronic components, or other obvious faults. Destructive effects can occur when noise pulses are too fast for
power-supply regulator circuits to respond by limiting transient energy to acceptable levels. Also, transients on
the power line may subject electronic components with overwhelming energy levels. For example, components
like rectifier diodes can fail immediately when their Peak Inverse Voltage rating (PIV) is exceeded. PIV diode rat-
ings in a well-designed computer can be in the 1 kV - 1.5 kV range. Transients on AC lines can easily exceed
1,500 Volts, and often by a wide margin.

Basic Operation of Surge Protective Devices (SPD’s)

As described at the beginning of this chapter, a transient voltage surge is a short-term deviation from a desired volt-
age or signal. Obviously, the higher the transient amplitude, the greater the likelihood of disrupting or damaging elec-
tronic equipment. And transients are generated whenever a current is interrupted. For instance, devices like variable-
speed drives are constantly switching circuits over very short periods of time. They can produce spikes 1 to 7 times
per cycle, or even more often, depending on the drive. That represents over 420 transients per second, at 60 Hz.

Furthermore, transients can occur on any metallic conductor, so they affect not only devices connected to utility
power lines, but also telephones, FAX machines, computer data lines, closed circuit and cable feeds, and others.

A Surge Protective Device, or SPD, attenuates the magnitude of these surges to protect equipment against their
damaging effects. But a SPD doesn’t necessarily reduce the surge to zero amplitude. It just attenuates it to a level
that can safely be passed through to the load. In addition, it’s often advisable to apply a network of SPD’s to provide
a layered defense against transients. Another benefit of Surge Protective Devices is that they can reduce noise ener-
gy as well as transient voltage surges. Since the majority of transient voltage surges are generated inside a building,
understanding and applying SPD’s is a prime consideration for improving the Power Quality in any facility.

16
Remember: SPD’s Do Not Solve Every Power Quality Problem
Surge Protective Devices can’t cure sags and swells in
the AC power provided by electrical utilities. They also
cannot reduce the harmonic conditions produced by Transient Current
non-linear loads like motors and switching-mode power
supplies within computers and some fluorescent lighting
systems. Harmonics reduction requires devices with very
large, specially manufactured capacitors. TVSS-1
Despite some claims that occasionally appear in print,
SPD’s cannot provide utility bill savings. There has
never been any responsible third party testing that
shows SPD’s can cut energy consumption in any way.
SPD’s also can’t remedy power outages. If there’s a Computer
loss of utility line voltage, a device that can temporari-
ly replace utility power is the Uninterruptible Power
SPD Shunt Path For Transients
Supply, or UPS. Conversely, most UPS systems do
not effectively eliminate surges. Although some may
FIGURE 24
have internal components which are protected against
spikes, the UPS itself cannot protect any of its loads
against spikes and, of course, it can’t have any effect
on parts of a facility’s AC lines not connected to the
UPS output. TRANSIENT
PRESSURE
PULSE

What Are Surge Protective Devices?

These are devices that can attenuate, meaning
reduce, transient voltage levels and noise. They’re CHECK
VALVE
also called Transient Voltage Surge Suppressors,
abbreviated as TVSS. The trend in industry is to call
them Surge Protective Devices, or SPD’s. But basi-
cally, the terms TVSS and SPD are interchangeable
since both refer to the same thing.
Surge protective devices are designed to REDUCE
potentially damaging short-duration transients present on
utility power lines, data networks, telephone lines, closed “The Water Wheel”
circuit and cable TV feeds, and any other power or con-
trol lines connected to electronic equipment.
FIGURE 25
One common misconception is that electronic
equipment must have ALL transient voltage surges reduced to zero amplitude on the power or data lines. This is not
the case. With computer systems for example, reducing transients to levels of approximately 150% to 300% of line
voltage (meaning 400 to 500 V peak) will prevent equipment damage. A feature of higher-quality SPD’s is the ability
to eliminate continuous high-frequency noise in addition to attenuating short-lived transients.

How Surge Protective Devices Work

In the simplest terms, SPD’s prevent damaging transient voltage surge levels from reaching the devices they protect. Figure
24 is a graphic representation of how an SPD provides a shunt path for transients before they can enter a computer.
A useful analogy makes this clearer. Consider a water mill protected by a pressure relief valve. The pressure relief
valve does nothing until an over-pressure pulse occurs in the water supply. When that happens, the valve opens and
shunts the extra pressure aside, so that it won’t reach the water wheel. The general arrangement of this system is
shown in Figure 25.
If the relief valve was not present, excessive pressure could damage the water wheel, or perhaps the linkage for the saw.
Even though the relief valve is in place and working properly, some remnant of the pressure pulse will still reach the wheel.
But pressure will have been reduced enough not to damage the water wheel or disrupt its operation.
This describes the action of surge protective devices. They reduce transients to levels that will not damage or dis-
rupt the operation of sensitive electronic equipment.
17
Clamping
“Clamping” is the term used for the process whereby
6000 V
SPD’s reduce or attenuate transients and limit the 6800 V
1700 V
surges reaching the protected load to a specific lower
voltage level. Transformer
1600 V
L 320 V
Figure 26 shows graphically how this happens. An Service
incoming transient voltage surge is shown both before Panel

passing through the SPD (on the left), and after (on

460 V

120 V
Current Current
SPD SPD
the right). After passing through the SPD, the tran- Shunt Shunt

sient has been attenuated to a lower amplitude.

Notice that it has not been attenuated to zero ampli-
N
tude. There is still a reduced level of transient voltage Load
remaining, and this is called the RESIDUAL, or
“let-through” voltage.
G

In fact, reducing transients excessively below needed Unclamped and Clamped Trainsient
levels can do more harm than good. Excessive tran-
FIGURE 26
sient attenuation causes unnecessary strain on the
SPD itself. Furthermore, higher externally-generated
transient currents are drawn into the facility, where
they may couple onto adjacent wires and cause inter- MOV Performance Curve
ference in unprotected branch circuits.

Clamping is one of the very important measures by

Clamping Voltage (Peak)

which an SPD is judged. However, the clamping level BURN OUT

400 Vc

is not the sole parameter used to evaluate an SPD’s

performance. Other factors such as surge current
capability, fusing, and filtering may be equally impor-
tant. These characteristics will be explained in greater
detail later in this reference guide.
200 Vc

NOMINAL VARISTOR VOLTAGE

Metal Oxide Varistors (MOV’s) and Other
SPD Components .001 1 10 100 1000
One type of SPD component is used in so many tran- Transient Amperage
sient voltage surge suppressors that it deserves par-
ticular attention. This component is called a Metal FIGURE 27
Oxide Varistor, (Variable Resistor) and is almost
always referred to as an MOV. An MOV is actually a non-linear resistor with certain semi-conductor properties. The
semi-conductor characteristic making an MOV ideal for use in surge suppressors is that it remains in the “OFF,” or
non-conducting state until a surge appears on the line to which it’s connected.

A typical MOV response curve is shown in Figure 27. It’s clear that the response curve is not linear (the response
characteristics are not graphed in a straight line). As the transient amperage increases (moving from left to right), the
clamping voltage also increases (moving bottom to top). The significance is that as the transient CURRENT increas-
es, so does the clamping level. Therefore, constant rates of increasing current produce disproportionately higher
clamping voltage levels.

Another important consideration is the Maximum Continuous Operating Voltage (abbreviated as MCOV) of surge
suppressors equipped with MOV’s. This is the maximum utility line voltage (or V rms) that may be applied to the sup-
pressor without damaging the MOV.

18
The Essential Advantages and Disadvantages
of the MOV and Other Commonly Used SPD Components.

Metal Oxide Varistor (MOV): Readily available, higher energy capability, excellent
reliability and consistent performance.

Non-linear clamping curve, rapid fatigue at higher

amperage levels, leaky.

Silicon Junction Diode/ Readily available, flatter clamping curve, excellent Avalanche
Diode (SAD): reliability and consistent performance.

Very low energy capability, some capacitive

problems, expensive.

Gas Tubes: Higher energy capability than either diodes or

MOV’s, non-capacitive or leaky in data line
applications.

Unpredictable and unstable repetitive behavior,

tendency to “crow-bar” to ground, higher cost than
MOV’s.

LCR Filters: Excellent noise attenuation, clamping harmonic

elimination and predictable performance at given
frequency.

Expensive, frequency dependent, low energy

capability, leaky and low amperage capable.
(The LCR filter is not a suppressor in and of itself.)

TVSS Hybrid: If properly designed, the Hybrid incorporates all of

the major advantages of many of the available
components while collectively overcoming their
individual faults.

The hybrid is inherently more expensive than the

single component TVSS.

In addition, some less commonly used SPD’ components include: Selenium, Spark Gaps, Zener Diodes, and
“Crowbars” (Zener/SCR combination). These components provide fast response times to transient voltage surges,
but vary in cost, clamping performance and energy-handling capability.

19
Filters
Filters are important for reducing high frequency noise and harmonics on the AC power line. An LCR filter is an
inductive, capacitive and resistive circuit designed to respond to a specific range of frequencies. A filter may be low-
pass, high-pass, band pass or band eliminate (notch). The LCR filters used in high-quality SPD’s are the low-pass
type. They will have no effect on the 60 Hz line voltage, but present a very high impedance to high frequencies that
can potentially disrupt microprocessor circuits. Filters have limited energy capabilities and are not intended to sup-
press transients on their own. They also play an extremely important role in communication line noise suppression,
operating with a fixed load and source impedance of 50 - 300 ohms.

How Noise is Measured

Noise can be measured in terms of either power or voltage. On data systems, voltage amplitudes are usually pre-
ferred. An oscilloscope is used to provide measurements in Voltspeak. Noise reduction/attenuation parameters are
often given in decibels (dB). The formula for noise attenuation is:

dB(db) = 10 Log (Powerout/Powerin)

dB(db) = 20 Log (Voltageout/Voltagein)

This test procedure is performed with a 50Ω load and is referred to as the MIL-STD 220A insertion loss test.

Characteristics of Surge Protective Devices

The two most frequently used types of SPD are divided into two very general categories based on the type of surge
suppression circuitry they feature. One type uses MOV’s as the suppression component for high-energy transients.
The other type also uses MOV’s, but has additional components to make the SPD more effective in suppressing
high-frequency transients.

Figure 28 shows how a single MOV SPD operates.

The top portion shows a potentially damaging tran-
sient being transmitted along the AC line conductor PROTECTED
directly to the load. The single MOV device solves the ELECTRICAL
SYSTEM
problem by shunting the transient current between the 6000V, 200A
100 KHz
line and neutral conductors, preventing the bulk of the
400V
energy from reaching the load. Only the residual tran- CLAMPING LEVEL
L
sient energy travels on to the load. RESIDUAL TRAVELS
TRANSIENT CURRENT
IS SHUNTED L-N BY ON TO THE LOAD
SPD COMPONENT
SINGLE SPD
COMPONENT
TRANSIENT RESIDUAL
SHUNTED TO NEUTRAL ENERGY DISSIPATED
AS HEAT DURING
CLAMP LOAD
N

FIGURE 28

20
MOV Clamping
SPD’s using only MOV’s are referred to as an “enve-
lope clamping” devices. That’s because, as Figure 29
shows, they clamp transient surges by limiting their
amplitude within a broad band above and below the
60 Hz AC sine wave.
Clamping
Level

Typical Clamping Envelope of MOV - based SPD’s

FIGURE 29

Hybrid SPD’s
Figure 30 shows an SPD with relatively simple combi- HYBRID SPD
nation of suppression components in addition to
LINE INDUCTOR
MOV’s. This is an example of a hybrid suppression
circuit. A properly designed hybrid circuit will vastly L
LOAD
outperform any single-component SPD.
MOV PRIMARY

MOV SECONDARY

FIGURE 30

Hybrid SPD Clamping

Hybrid suppression circuits are often referred to as
“sine-wave tracking” devices. As shown in figure 31,
their clamping profile follows the contour of the 60 Hz
AC sine wave, and tightly attenuates both positive
and negative-going spikes.

57120-M3
0 Commercial
120/208V
Power Envelope
(170 Vpeak +/– 15%)

0 90 180 270 360

PHASE ANGLE (Degrees)

FIGURE 31

21
Types of Surge Protective Devices
There are many types of surge protective devices Load Current
available. Two broad classifications include those
devices used to attenuate transients on low-voltage
conductors, meaning data lines, phone lines, coax
feeds, and others. Then there are SPD’s intended to
suppress transients on AC utility power lines. TVSS
LOAD
DEVICE
The application settings for SPD’s fall into three gen-
eral categories: First, SPD’s that are wired into the AC
or low-voltage lines at some distance from the equip-
ment they protect. An example would be branch-panel
mounted suppressors. Second, SPD’s that provide Parallel Operated SPD’s
point-of-use protection at the same location as the
equipment. Examples include plug-strips, though
FIGURE 32
some point-of-use SPD’s are wired in rather than
plugged in. Third, is integral SPD’s that are a built-in
component of the equipment they protect.
Among SPD’s for AC line applications, there are some units designed to be wired in parallel, and others designed to
be wired in series. Figure 32 shows a parallel-wired SPD.

The advantage of parallel-wired suppressors is that they don’t have to handle any of the load current. Theoretically,
these units can be placed on any size main current power bus. Of course, mechanical features of these SPD’s must
conform to requirements of the National Electrical Code (NEC) and, sometimes, NEMA requirements to meet the
demands of various installation specifications.

Series-operated suppressors work in a different way.

These devices have one or more series inductors that
rely on a principle called Lenz’s law which states that
a current cannot instantaneously change through an
inductor. Figure 33 shows a typical series-connected TVSS DEVICE
SPD.
Load Current
LOAD
Series devices have inductors in line with power leads
that ultimately supply downstream loads. In addition
to the inductors, other elements such as capacitors
and high surge-current protection components are
employed to reduce incoming transients to very low
levels.
Series Operated SPD’s
A disadvantage of series SPD’s is that they must be
able to handle not only high surge currents, but the FIGURE 33
load current as well. Therefore, the inductors tend to
be quite large when designed for high-current levels,
particularly for three-phase devices rated over 100
Amps, with at least one coil per phase.

22
 Chapter 3: SPD Performance Standards

The purpose of performance standards for SPD’s is two-fold. First, SPD’s can be described by literally dozens of
technical parameters, and it is usually not simple to make an apples-to-apples comparison. Second, there is a basic
need to define certain upper limits for expected transient voltage and current levels in a number of well-defined facili-
ty locations.

There are two important sets of standards that are applied to surge protective devices. One is the Underwriters
Laboratories UL 1449 collection of standards, and the other is the Institute of Electrical and Electronics Engineers
EEE C62.41 standards.

Using the measures of performance provided in UL 1449 and IEEE C62.41 standards, it is possible to make a mean-
ingful comparison of surge protective devices, and objectively judge which devices are most effective in attenuating
transient voltage surges of the sort most likely to occur in their intended application environments.
The remainder of this chapter will deal with the specifics of the UL and IEEE standards.

Technical Note On Determining Transient Voltage Levels

Before we examine standard measurements of transient voltage waveforms, it’s useful to take a quick look at an
important formula.
High-energy transients occur whenever a current is interrupted. The higher the current, the greater the amplitude of
the transient. The following formula can be used to determine the transient voltage level (represented by V in the
equation):
V = -L di/dt
L is the circuit’s total inductance. di represents the rate of change in the current. dt is the interval of time over which
the current changed.
Note that since dt is the denominator in this fraction, the faster the transient (meaning the smaller the number repre-
sented by dt), the larger the transient amplitude (represented by V) becomes.

The IEEE C62.41 Standards That Define Transient Voltage Surges

Prior to 1970, very little information regarding the transient environment was available. By the end of the decade,
professional committees of IEEE and others had compiled sufficient data to publish a bench-mark waveform guide
thought to describe both transient waveforms and their probable locales of occurrence. The guide became known ini-
tially as IEEE 587. Later on, after adoption by ANSI, it became IEEE C62.41.

23
Two Types of Surges
Two basic types of transient voltage surges are 6000V
defined by the IEEE C62.41 Standards: First, a
100%

PERCENT OF V
“combination-wave” transient, and second, a “ring-
wave” transient. For purposes of comparison, a com- 90%
bination wave is associated with lightning-induced
transients on utility power lines. It has a significantly 50%
higher current than ring waves. 30%
A ring wave is an oscillatory surge with relatively high 0%
voltage levels at relatively high frequency, but with 1.2µs TIME
limited energy content. Ring waves are associated
with fuses opening, power factor/capacitor switching 50µs
action, or load switching of motors, pumps, compres-
sors, and other electrical loads. Combination Wave, Open-Circuit Voltage

Combination Wave 3kA

The combination-wave transients that could be PERCENT OF I 100%
expected from lightning were characterized, and these 90%
are displayed in Figure 34. One waveform shown
comprises the test CURRENT, and is defined by an 8
microsecond (written 8µs) rise time, with a 20µs trail- 50%
off. At that point, the wave has diminished to 50% of
its peak value. 10%
0% TIME
By way of explanation, the rise time of any wave is 8µs 50µs
the time needed for it to transition from 10% of its 20µs
maximum amplitude to 90%.
Combination Wave, Short-Circuit Current
The accompanying VOLTAGE waveform for lightning
has a 1.2µs rise time with a 50µs trail-off. FIGURE 34

The test parameter just described is called a combi-

nation wave because the test source must provide
6kV, 500A, tr =0.5µs
both the current and voltage waveforms simultane-
ously. Vpeak
0.9Vpeak
The second waveform, called a ring wave, is impor- T=10µs (f=100kHz)
tant to testing SPD’s higher-frequency response to
transients created within a facility by interrupted load
currents. 0.1Vpeak

0.5µs
Ring Wave
As shown in Figure 35, the ring wave is characterized 60% of Vpeak
as having a fast rise time of only 0.5µs along with a
10µs period, which yields a natural frequency of 100
100 KHz Ring Wave
kHz.
FIGURE 35

24
The IEEE C62.41 Standards That Define SPD Operating Environment Categories
In order to properly test SPD’s, it was also necessary to define the operating environment WITHIN a facility. The
IEEE C62.41 Standard defines three operating location environments called Category A, Category B, and Category
C, as shown in Figure 36.

Category C environments are located

A B C
on the LINE side of the service disconnect.
• Outside and service entrance
Service • Service drop from pole to building
Entrance
Meter • Run between meter and panel
• Overhead line to detached building
• Underground line to well pump
Category B environments are immedi-
ately adjacent on the LOAD side of the
service disconnect breaker. Category B
environments are characterized as hav-
ing short branch circuits and feeder lines.
• Distribution panel devices
Service • Bus and feeder industrial plants
Entrance
Outbuilding
• Heavy appliance outlets with
“short” connections to service
Meter
entrance
• Lighting systems in large
Underground Service buildings
Category A environments have long
branch circuits and outlets more than 30
feet from a Category B environment, or
more than 60 feet from Category C
locations.

Service
Entrance Outbuilding
Transformer

Underground
Service

FIGURE 36

25
IEEE Test Standards
With combination waves and ring waves clearly defined, the IEEE has specified test standards using both wave
forms as they could occur in Category A, B, and C locations. There are standards for three levels of exposure within
the categories. So, for example, in Category C, the C1 test standard represents the least severe combination wave
exposure. C2 represents a moderate exposure, and C3 is the most severe. The test values for exposures in each
category are shown in the following table:

Standard 0.5 ms-100 kHz Ring Wave Standard 1.2/50 ms-8/20 ms Combination Wave
Voltages and Current Surges Expected Voltages and Current Surges Expected
in Location Categories A and B in Location Categories B and C
Low, Medium and High Exposures Low, Medium and High Exposures
Single-Phase Modes: L-N, L-G, and [L&N]-G Single-Phase Modes: L-N, L-G, and [L&N]-G
Polyphase Modes: L-L, L-G, and [L’s]-G Polyphase Modes: L-L, L-N, L-G, and [L’s]-G

Peak Values Effective Peak Values Effective

Location System Voltage Current Impedance Location System Voltage Current Impedance
Category Exposure (kV) (kA) (Ω) Category Exposure (kV) (kA) (Ω)

A1 Low 2 0.07 30 B1 Low 2 1 2

A2 Medium 4 0.13 30 B2 Medium 4 2 2
A3 High 6 0.2 30 B3 High 6 3 2

B1 Low 2 0.07 12 C1 Low 6 3 2

B2 Medium 4 0.33 12 C2 Medium 10 5 2
B3 High 6 0.5 12 C3 High 20 10 2

Chart 1 Chart 2

Notice that Category C environments are subjected only to combination wave transients, while Category B environ-
ments are tested using both ring waves and combination waves. Category A environments are tested with ring
waves only.
The reason for limiting the test transients to 6 kV ( meaning 6,000 Volts ) in Category B environments is because of
certain characteristics of typical wiring devices used in commercial, industrial and residential Category B application
settings. Specifically, the IEEE determined that terminal screw spacing between line, neutral and ground of recepta-
cles and metering pans would cause arcing when voltages in excess of 6 kV were placed on the lines connected to
device terminals.
Furthermore, the combination waves associated with lightning-induced transients were limited to 3 kA in Category B.
The IEEE concluded that inductance of wiring within a facility would limit the amplitude of conducted lightning cur-
rents to no more than 3,000 Amps. Some experts hold a different opinion and believe that currents in Category B
locations may actually be capable of reaching 35,000 Amps. Yet even this figure is far short of the unreasonable
claim made by some protective device manufacturers who suggest facilities must be capable of absorbing lightning
currents from 250 kA to 1 million Amps.
Such claims are dubious because research shows that the statistical probability of any lightning-induced transient is
less than 5%. But more important, the largest lightning-current test facilities are equipped with generators that can
produce only up to 180 kA reliably with an accurate 8µs x 20µs wave form per phase. Maximum surge current rat-
ings in excess of these values for any surge protective device are meaningless, because there is no way to carry out
tests that can confirm performance.

Note on “Passing or Exceeding” IEEE Standard C62

Be wary of manufacturers that claim their SPD’s “PASS or EXCEED IEEE Standard C62,” because such claims are
meaningless. The C62 Standards cannot be passed or failed, or exceeded. They simply serve as a yardstick for use
when testing and documenting the performance of Surge Protective Devices.

26
IEEE C62.45 Performance Testing
The ANSI/IEEE C62.45, “Guide on Surge Testing for Equipment Connected to Low-Voltage AC Power”, established a
world-wide accepted performance evaluation for Surge Protective Devices. The tests are performed using off-the-
shelf hardware and conducted in strict accordance with ANSI/IEEE C62.45 test procedures. The test results should
be repeatable in any adequately equipped laboratory in the world.

“Let-through”, residual or clamping voltage of various suppression products is easily measured by using IEEE wave-
forms to approximate actual on-line transients at A, substituting SPD samples at B, and measuring the residual
clamp with a digital storage oscilloscope at C.

Every SPD sample will demonstrate a unique “signature” with each specified waveform input. These signatures can
be compared by waveform. Obviously, the lower and cleaner (implying smooth and without significant harmonic con-
tent) the signature, the better the performance.

6000V 200 A
100 KHz

Residual travels
PROTECTED on to the load
Transient current
ELECTRICAL is shunted L-N by Single SPD
SYSTEM TVSS component Component

C
TYPICAL TEST SETUP A B
Velonex 587 Suppression Tektronix
Surge Device Digital Storage
Generator Under Test Oscilloscope

TEST WAVEFORMS
Cat A Ring Wave
6KV, 200 A, 100 KHz IEEE Resultant
Waveforms Output
Cat B Impulse
6KV 1.2 x 50 us
3KV 8 x 20 us Tektronix
P6009 100X
Change voltage constant NEMA 15-5 probes to scope
Scope settings constant
at 6KV, 90 degrees plug-in connection as shown during test.
to generator Readout on screen

ELECTRICAL ENVIRONMENT PROTECTION MICROPROCESSOR ENVIRONMENT

FIGURE 37
27
The New UL 1449
An overview of the second edition requirements for transient voltage surge suppression and the
impact of the revisions to the standard.
Introduction:
Underwriters Laboratories satisfied a much called for need to the industry at large when the first edition of UL 1449,
the Standard for Safety for Transient Voltage Surge Suppresors was published August 25, 1985. The standard adopt-
ed select test waveforms from the IEEE C62.41 . However, soon after the introduction of the standard it became evi-
dent that certain issues needed to be addressed, most notably with regard to the "fail-safe" operation of surge sup-
pression components contained in these devices under abnormal service conditions or "end-of-life" degradation
modes specific to metal-oxide varistors.
After a number of years the proposed second edition of UL1449 was published August 25, 1996. The effective date
assigned to these new and revised requirements is February 16, 1998. It should be noted however that Hardwired
products (panels and receptacles) have been extended to August 17, 1998.
It is the responsibility of manufacturers and specifiers to make available the safest product to the end-user and con-
sumer. The following outlines a comparison of the "old" versus "new" UL1449 and the impact of these Second
Edition requirements.

The Scope of UL1449:

A good place to start a review of UL1449 is the scope of the standard. This is where issues can be clearly defined as
to what does and does not apply to the Listing of a product bearing the UL – TVSS mark. It should be noted that
unlike most other standards bodies including ANSI, IEEE, NEMA and IEC; UL has elected not to use the term Surge
Protective Device (SPD). It is generally accepted that the terms TVSS and SPD are one in the same.
The Scope of UL1449 contains the following:
1. The requirements cover transient voltage surge suppressors intended for permanently connected, cord-connected
and direct plug-in applications on 50 or 60 Hz power circuits not exceeding 600v ac.
2. These requirements are intended for installation on the load side of the main overcurrent protection.
3. These requirements do not cover the interconnection of multiple field installed TVSS.
4. These requirements cover TVSS employing circuit components specifically intended to function as filters for con-
ducted electromagnetic interference (EMI) or noise, in addition to limiting transient voltage surges.
5. These requirements do not cover Secondary Surge Arresters intended for use on the line side of the main over-
current protection.

28
Test Requirements:
UL1449 has basic safety tests no different from most UL standards for safety. These tests include leakage current,
dielectric withstand, insulation resistance, temperature rise and mechanical integrity tests such as impact, drop,
crush and mold stress relief distortion for plastics among others. For the sake of this discussion only tests specific to
surge protectors will be covered here.
Measured Limiting Voltage Test:
Suppressed Voltage Ratings (SVR) are assigned based on subjecting products to the following waveforms. These
waveforms are combination impulses described in IEEE C62.41. The rise and decay wavforms are; 1.2 by 50 us for
open-circuit voltage and 8 by 20 us for short-circuit current. Ratings are measured from zero ground to the peak of
the “clamped” transient level.
The “Old” Table The “New” Table
Product Type Voltage Current Impulse Voltage Current Impulse
Panels & Receptacles 6 kV 500 A 6 kV 500 A
Strips & Plug-ins 6 kV 3000 A 6 kV 500 A

It is also important to note that Hardwired products are measured at the end of Six inches of lead length recommend-
ed in the manufactures installation instructions. The “old” UL1449 would measure directly at the terminals of hard-
wired equipment; devices with integral leads were always measured at six inch lead length.
Duty Cycle:
SVR is assigned based on the before and after levels described in the table below. It should be noted that three
samples are tested and the before and after duty cycle measurements shall not deviate by more than ten percent.
The rating is then assigned based on the average of the total of six before and after measurements.

The “Old UL1449” The “New UL1449”

Product Type 24 Shots 20 Shots
Panels & Receptacles 6 kV 750 A 6 kV 3000 A
Strips & Plug-ins 6 kV 125 A 6 kV 500 A

Maximum Surge Current Test:

This is a new test added to UL1449 with the intent for a device to withstand a relatively high surge current and not
be degraded to the point of causing a risk of fire or shock hazard. Panels and receptacles are subject to two shots of
a 10 kA / 6kV impulse.
Strips and plug-ins are subjected to a 3 kA / 6 kV impulse.
Abnormal Overvoltage Tests:
These are the most notable of new tests introduced in the standard. There are three:
1. Temporary Over Voltage: The device is connected to 125 percent of the normal system voltage for seven hours
and must not result in risk of fire or shock hazard
2. Full-Phase System Overvoltage: This test subjects the device under test to a simulated "Loss of Neutral". For
example, a device intended for 120/240v systems will be connected to 240v. Again, after the test the device shall
not exhibit evidence of a risk of fire or shock hazard.
3. Limited Current Overvoltage: This test subjects the device under test to increasing levels of "standby current"
through each mode where an MOV is connected within the device. The test basically simulates an "end-of-life"
degradation, or thermal runaway scenario. Again, at the conclusion of the test there shall be no evidence of a risk
of fire or shock hazard. Verify also includes a dielectric withstand (hipot) and leakage current tests.
There may be various approaches in design of a product to pass these tests including several methods of properly
fusing MOVs, or just over-rating the maximum continuous operating voltage for the MOVs employed in the device.
Obviously the latter design will result in a safe, but high SVR, or clamping level.

29
Product Markings:
Believe it or not, UL has elected not to require any changes in the "UL – Listed TVSS" marking. In the case of the
consumer purchasing a basic surge strip it will be left to honest labeling on packaging to state any claims to compli-
ance to the new, second edition UL1449.
Fortunately for the specification community, compliance to the new UL 1449 can be verified by the actual UL Listing
Card. UL has changed the format for the TVSS Listing Card from merely listing catalog numbers of Listed products
under the TVSS category to providing a table with suppressed voltage ratings (SVR) for each model by mode.
Suppressed Voltage Ratings are required to be marked on the product by each mode. The measured limiting voltage
is rounded up to the following SVR assigned levels:
330v pk 600v pk 900v pk 1500v pk 2500v pk 5000v pk
400v pk 700v pk 1000v pk 1800v pk 3000v pk 6000v pk
500v pk 800v pk 1200v pk 2000v pk 4000v pk

Summary:
UL1449, the Standard for Safety for Transient Voltage Surge Suppressors—2nd Edition
• A UL Listed TVSS must be installed on the LOAD side of the main service disconnect. Devices connected LINE
side are Secondary Surge Arresters which are Listed by UL under "desk" standard ANSI/IEEE C62.11.
• A UL Listed TVSS may be install OUTDOORS if evaluated to UL50 with appropriate NEMA Type rating assigned to
the device.
• A UL Listed Series-operated/connected TVSS must be tested and assigned an Available Fault Current Withstand
rating (AIC). Series-operated devices are also known as Two-Port Surge Protectors.
• A UL Listed Parallel-operated/connected TVSS does not require an AIC rating.
• The term SVR denotes Suppressed Voltage Rating. This is the assigned "Clamping" voltage with respect to zero
volts (ground) to the peak of the "clamped" voltage.
• Cord-connected and Direct plug-in TVSS are assigned SVR based on 6kV / 500A (Duty Cycle (20 shots)
• Hardwired Panel and Receptacle TVSS are assigned SVR based on 6kV / 500 A at the end of 6 inch leads.

Hospital Grade Surge Protective Devices

For Hospital Grade surge protective devices, there is an additional safety test performed to measure leakage current.
Figure 38 shows how the test apparatus is arranged.
Devices that are Listed for use in health-care settings are limited to a maximum leakage current level of 300 micro-
amps (written 300 µA) per UL 544, Standard for Safety for Medical and Dental Equipment, third edition. For devices
that aren’t Listed Hospital Grade and not intended for use in the health care environment, acceptable leakage current
levels are somewhat higher at 0.5 milliamps (written 0.5 mA). Devices with leakage currents greater than 0.5 mil-
liamps but less than 3.5 milliamps must carry a WARNING label.

L
120 Volt Product Meter
N
Gr Insulating Table

FIGURE 38

30
Methodology for Testing Permanently Connected SPD’s
If the device is supplied with wire leads, these are
cut back to a six-inch length and connected to the
surge generator. Clamping voltage is measured Surge
at the end of the leads. If the device is equipped Attachment Generator
with terminal lugs, the surge generator is still Points
attached with 6 inches of leads from the lugs, and
clamping voltage is measured at the ends of the
leads.

6"
LEADS
6"
LEADS

SPD with
Terminal Block

SPD with
Clamping Voltage Wire Leads
Measurement Test Set Up

FIGURE 39

31
Understanding UL 1449 Ratings

Section 37.1 of the UL 1449 Standard defines the range of clamping levels demonstrated by SPD’s that have been
tested by UL. These levels are assigned by taking the ACTUAL clamping level and ROUNDING UP to the nearest
UL rating level. The UL rating levels are:

330 Volts 1500 Volts

400 Volts 1800 Volts
500 Volts 2000 Volts
600 Volts 2500 Volts
700 Volts 3000 Volts
800 Volts 4000 Volts
900 Volts 5000 Volts
1000 Volts 6000 Volts
1200 Volts

Notice that the lowest possible UL assigned clamping level is 330 Volts. Remember that UL rounds UP to the nearest
level, so for example, an SPD with an actual clamping level of 290 Volts would still be issued the 330 Volt UL clamp-
ing rating. Likewise, a device that actually clamps at, say, 634 Volts would be assigned an 800 Volt UL clamping
level.

UL adopted this policy to assure that the clamping levels provided for UL-tested SPD’s would indicate an absolutely
reliable minimum level of performance that could be expected by the users of these devices.
Furthermore, all UL 1449 Listed surge protective devices MUST be furnished with a label applied directly to the prod-
uct that plainly shows UL clamping ratings verified by UL testing.

The label plainly indicates “UL Listing – Transient Voltage Surge Suppressor.” It shows the clamping levels of the
SPD, and indicates modes of protection. These are the so-called NORMAL MODE, meaning SPD protection is pro-
vided from Line to Neutral, and COMMON MODE, meaning there is protection from Line to Ground and Neutral to
Ground. The protection modes are important to verify because it is possible for a manufacturer to get UL 1449
Listing on an SPD that doesn’t offer both normal and common-mode protection.

Moreover, real caution on the part of the

buyer is needed when choosing plug-in
SPD’s. That’s because there are some
devices on the market described as surge Cat. No. 5300-P Transient Voltage
suppressors by their manufacturers. But Assembled in Mexico Surge Suppressor
they are NOT UL 1449 LISTED AS TRAN-
SIENT VOLTAGE SURGE SUPPRESSORS. U® L Listed
TVSS
55N7
Rating: 125VAC
60Hz, 15A
Instead, they are often listed under UL 1363
Standard for Safety - Temporary Power CAUTION: to reduce the risk of
Taps. This is actually the set of standards electric shock, use only indoors
that applies to devices like cube taps or out-
let adapters. It DEFINITELY DOES NOT Plug-in Diagnostics: U.L. 1449 Suppression
indicate satisfactory clamping performance, Indicator Light L-N....330 V peak
and in fact is proof that the device has defi- ON Off L-G....330 V peak
nitely not been subjected to clamping perfor- Power ON OFF N-G....330 V peak
mance testing by UL. Of course, any prod- Protection ON NONE
uct claimed by its manufacturer to be a Polarity GOOD REVERSE If fault is indicated,
surge protective device, but without the UL & consult qualified
1449 Listing, should never be used as a electrician.
Ground GOOD FAULT
surge suppressor, and should be avoided
entirely.
FIGURE 41

32
IEC 1000-4 Series Standards (Formerly IEC 801)
The International Electrotechnical Commission, known as the IEC, issued the 801 Series Standards for Immunity to
Electromagnetic Interference (EMI). IEC 1000-4 addresses EMI that reaches electronic equipment via both conduc-
tion and radiation at every point of the equipment. In particular, the IEC addresses the Power Quality problem creat-
ed by Electrical Fast Transients (EFT) or burst noise. EFT is caused by the showering arcs that accompany heavy
power switching, such as in air conditioning compressors, oil burner relays, etc. The IEC 1000-4-4 EFT/Noise test
waveform is a repetitive, fast transient wave rather than the single pulse wave selected in earlier ANSI/IEEE
Standards. The repetitive wave simulates the phenomena of EFT pulses building up on microchip inputs, eventually
causing an incorrect bit or count. The latest ANSI/IEEE C62.41 cites IEC as its EFT requirement.

The Importance of Product Warranties

After reviewing the performance characteristics of SPD’s, the second most important feature of these devices is the
manufacturer’s warranty. The warranty should be carefully read, because its fine print often invalidates the very pro-
tection features desired by the user.
Generally speaking, the track record and reputation of the device manufacturer is a major consideration. Many man-
ufacturers promise warranty periods that are longer than the time the firm has been in existence.
Some manufacturers claim to offer a “Lifetime Warranty”, but after a short period, the end-user is required to pur-
chase additional insurance to keep the warranty in effect. Also, some SPD’s are sold at grossly inflated prices as a
hedge against future warranty claims. Finally, some smaller manufacturers could be bankrupted by warranty claims
on major system failures involving their products.
Some manufacturers offer “DOWNSTREAM” warranty provisions. This type of warranty backs not only the SPD
itself, but also the electronic equipment being protected by the SPD. Typically, if the electronic equipment is dam-
aged, repairs are made at an authorized shop. After notification of the SPD manufacturer, and a finding that damage
was the result of a transient surge event, payment is made to the claimant.
There are many forms of warranty applied to surge protective devices. The one constant is that they must be careful-
ly examined, and viewed in light of the manufacturer’s longevity and reputation.

33
 Chapter 4: Specifying SPD’s

Overview
Because of the volume of data published by the manufacturers of Surge Protective Devices, creating a specification
for these devices is vulnerable to omissions or inconsistencies. Recognizing the need for a consistent, reliable speci-
fication structure, the National Electrical Manufacturer’s Association, or NEMA, has developed the NEMA LS-1
Specification Format for Surge Protective Devices.

Navigating Through Irrelevant Performance Specifications

It’s also important to understand certain IRRELEVANT characteristics of SPD’s that often muddy the waters, mislead,
and confuse the issues when making product comparisons and specifying SPD’s. Perhaps the two characteristics of
SPD’s that are most often improperly reported involve Joule energy ratings and response time or turn-on time.
Despite the emphasis that these parameters are given by some manufacturers, they can be safely disregarded when
evaluating SPD’s, because of the many ambiguous ways these values are reported. Indeed, the NEMA LS-1 format
does not include any reference to either Joule energy rating or response time, because NEMA recognizes
that these are not critical parameters.

Joule Energy Ratings

To begin with, the “Joule” is a unit of measure applied to the ability to do work, or absorb energy. In the case of
SPD’s it relates to the ability for absorbing heat energy. One Joule per second is equal to 1 Watt, a unit of electrical
power. Likewise, horsepower is another unit of work. At 100% efficiency, 1 hp is equal to 746 Watts. How is this relat-
ed to SPD energy absorption?

Well, horsepower is related to British Thermal Units, or BTU’s. One horsepower (746 Watts at 100% efficiency) is
equivalent to 2545 BTU’s per hour. A single BTU is the amount of energy needed to raise one pound of water one
degree Fahrenheit. This very roundabout path is the basic link of the “Joule” to electrical energy, and the temperature
rise within an MOV. To put it in simpler terms, when a transient current passes through an MOV, this component will
heat up because of the BTU equivalent energy within that transient current, and because of the resistance value of
the MOV itself.
Although this general link exists, it’s often reported in a way that can be misleading. For instance, some manufactur-
ers describe the Joule energy rating of a device as the sum total of the rating for each phase the SPD protects.
Obviously, for a three-phase device with neutral-to-ground protection, this figure will be four times higher than any
real-world Joule rating available per phase.
Moreover, the Joule ratings are often determined by calculations which depend on waveforms that are very different
from the IEEE C62.41 8x20µs standard.

In conclusion, Joule energy ratings of SPD’s are, at best, unimportant and at worst, seriously misleading.

34
Response Time or “Turn-On” Time

The parameter usually called response time is basically supposed to be the interval of time it takes an SPD to recog-
nize there is a transient present, then react by suppressing the transient. But, at the present time, no standard exists
for such a measurement, and it’s also not certain just what such a measurement would mean. That’s because of the
basic mathematical relationship first stated in chapter 1 by the formula

Frequency = 1/Time

Many manufacturers state their response times in nanoseconds. ( A nanosecond is one thousandth of one millionth
of a second.) Using the formula above, a nanosecond represents a transient with a frequency of 1000 megahertz, or
1 GigaHertz. Typical facility wiring just can’t support transmission of these frequencies.

In addition, simply measuring this phenomenon would require an instrument several orders of magnitude more accu-
rate than the thing it measures. As a rule of thumb, 5 to 10 times better, so the frequency bandwidth of an oscillo-
scope would have to be 5 to 10 GigaHertz. This is not possible with the typical instrumentation available in today’s
laboratories.

Worse yet, some manufacturers claim picosecond response times, meaning 1,000,000 MHz, or 1,000 GHz. This is a
frequency that’s beyond the radio-frequency spectrum, and it could not be propagated in any solid metallic conduc-
tor. Also, the parameter could never be measured because no oscilloscope exists with this performance capability.

Surge Counters

Some SPD manufacturers sell auxiliary devices that

claim to “count” the number of surges on the lines
being monitored. The surges tallied are supposed to Threshold Level
indicate the need for an SPD, or the need to replace
an SPD after it has experienced a certain number of
transients.
Counted
The critical question is what these counters actually
count. Counters are basically measuring voltage Surges
amplitude, often called voltage threshold levels, as
shown by Figure 42.

However, the destructive energy in a transient

depends on the voltage and current of the transient. Threshold Level
Knowing one without the other is meaningless. For
example, there’s a lot of difference in a 600 Volt tran-
FIGURE 42
sient at 5 milliamps compared with a 600 Volt tran-
sient at 500 Amps. But the counters don’t distinguish
between the two. If any transient simply exceeds a pre-set voltage level, it’s added to the tally.
Since a typical threshold for counters is 300-500 Volts, recorded transient events will include noise, and spikes of
unknown severity. Some may have truly damaging energy content, some will be harmless. However, if the manufac-
turer’s suggested replacement schedule is keyed to the counter’s tallies, there’s no way to be sure just how much
remaining life may actually be left in the SPD.

35
Using the NEMA LS-1 Format for Preparing
Surge Protective Device Specifications
A brief review of the elements comprising the NEMA LS-1 format
will be helpful in defining the information to be shown.
1. DEVICE NAME AND MODEL NUMBER: Available from manufacturer’s published literature and specifications.
2. DEVICE CIRCUIT DESCRIPTION: This defines the components within the Surge Protective Device that actually sup-
press transient voltage surges. Examples include single or multiple Metal Oxide Varistors (MOV’s), gas-tube design, hybrid
circuit, and others. This also indicates whether the device is series or parallel operated.
3. NOMINAL LINE VOLTAGE: Whether the device is panel-mounted, a plug-in unit, or a hard-wired unit, the intended
line voltage for its application should be specified.
4. MAXIMUM CONTINUOUS OPERATING CURRENT: This rating must be specified for Surge Protection Devices that
contain in-line series-connected components in their circuit design to confirm that the device will be operated within the
circuit ampacity limits and not be overheated when installed.
5. MAXIMUM CONTINUOUS OPERATING VOLTAGE (MCOV): The MCOV is typically dictated by the assigned rating of
the MOV’s within the surge protection device. The MOV rating is established by the MOV manufacturer rather than by the
maker of the Surge Protective Device. This value provides a ceiling for the device’s vulnerability to being degraded as a
result of line-voltage swells.
6. CONNECTION MEANS: Depending on the type of device, this could be lug terminals, terminal screws, wire leads, plug-
in, or others.
7. PROTECTION MODES: For point-of-use devices, three modes of surge protection should be provided: line to neu-
tral, line to ground, and neutral to ground. Of course, clamping data should be furnished for each mode. In the case of
panel-mounted units, especially those installed on delta systems or at service entrances where ground and neutral are
bonded, the devices may provide adequate protection even though every possible suppression mode is not applicable.
8. MAXIMUM SURGE CURRENT: This is a measure of how robust a Surge Protective Device may be in the the face of
extremely high peak currents associated with, for example, lightning-induced surges. The standard of measure is a sin-
gle-pulse 8x20µsec waveform specified in IEEE Standard C62.45. It is important that this information be provided by
actual testing of the device, and not only from calculations based on manufacturer’s component specifications.
9. CLAMPING VOLTAGE: The clamping rating of a Surge Protective Device is the result of tests conducted using the IEEE
C62 Standards Collection waveforms. It’s important for these values to show clamping levels with respect to zero ground
level, and with line voltage applied. Also, clamping data should be provided for each protection mode the device offers.
10. EMI/RFI NOISE REJECTION: Electromagnetic interference (EMI) and radio frequency interference (RFI) should be
attenuated by Surge Protection Devices in the frequency ranges specified. Test methods are based on 50Ω insertion
loss procedures outlined in MIL-STD 220A.
11. SAFETY AGENCY APPROVALS: Certification organizations like UL, CSA, and NOM, should be specified along with
their appropriate test standards, product categories, and reference file numbers. For example, in the case of a plug strip,
the UL listing would be based on evaluation under standards UL 1449 TVSS (XUHT) and also UL 1363 Temporary Power
Taps (XBYS). In addition, manufacturers are assigned file numbers for their listed devices. XUHT and XBYS are UL prod-
uct identifiers (CCN Directory Codes).
12. SAFETY AGENCY RATINGS: Safety agencies assign suppression (clamping) ratings based on requirements in
applicable test standards. For instance, the UL 1449 rating levels that appear in Chart 3 are used to determine
clamping ratings for UL listed Surge Protective Devices. It is important to note that CSA and UL both prohibit outdoor
installation of these devices unless they’re protected by an additional suitable enclosure like a NEMA-3R box, for
example.
13. PHYSICAL DATA: Available from manufacturer’s specifications, this entry spells out device dimensions, weights,
materials composition and other physical characteristics.

36
 NEMA LS-1
SPECIFICATION FORMAT
FOR SURGE PROTECTION DEVICES

SPD Model:

SPD Circuit Description

Nominal Line VoItage:

Maximum Continuous Line Current:

Maximum Continuous Operating VoItage:

Connection Means:

SPD Protection Modes:

MaximumSurgeCurrent:
(Single Pulse, 8/20 µs, Mode, Data obtained from actual tests)

Clamping VoItage during Maximum Surge Current (VPeak measured):

Clamping Voltage:
(Data taken at 90 degrees phase angle of power frequency voltage, positive polarity only,applicable surge current)

EMI-RFI Noise Rejection:

Safety Approvals (Agency, Standard, File):

Safety Ratings (UL1449, VPeak):

Application Environment:

PHYSICAL DATA:

Dimensions:

Weight:

Materials:

Wire size and length:

Accessories:

37
 Chapter 5: Protection of Telecommunications and Data Lines

Surge Protection — Solving Problems Without Creating Them

The threat posed by electrical disturbances, particularly those of an impulse nature, is well known to most computer
users. These disturbances go by many names: spikes, surges, transient surge voltages. The effects, however, are
universal; disruption, degradation and damage. Of equal importance, especially with the increasing popularity of
computer networking, is the effect of transients on communication lines. Communication lines entering a building
(underground or aerial) can import high-magnitude transients into a facility. Through complex coupling mechanisms,
voltage surges may damage an inner building’s communication interfaces. This complexity makes the “band-aid”
approach to surge protection a short-sighted, inappropriate strategy.
The mechanisms by which transient voltages
couple into data circuits can be divided into
two categories. The first is inductive coupling;
so called because the dataline forms part of
an “inductive” loop. Surge currents flowing
through metallic conductors produce a chang- Data Line
ing magnetic field, which will induce a voltage
in the loop formed by the dataline. The greater
the rate of change of surge current, the high-
er the voltage that will be induced in a given "Inductive" Loop
dataline loop. formed by AC supply and
datalink to remote terminal
The second and most destructive mechanism AC
Power
is ground potential difference. A dataline con- Line
necting two items of computer equipment will
be subjected to any difference in voltage that CPU or Host
may exist between the two equipment
grounds. The destructive effects of this coupling
mode can readily be seen when a dataline
interconnects two separate buildings. FIGURE 43
During a lightning storm, the ground potential
between buildings may differ by many thou-
sands of volts. The resulting physical damage
to interface chips is a familiar sight to service
engineers. Bldg
"B"
Bldg Dataline
"A"

Avoiding or Minimizing Noise on Data and

Communication Systems
Terminal CPU or Host
Noise on data lines can often be avoided by
proper consideration of the data line layouts
200 KV
within a facility. Data lines should not be
draped over power line conduits. Metal con-
duits should not be considered a shield from
Voltage
the noise and transients occurring on power
lines. Other areas to avoid when running data
lines are ballast transformers, lightning down- V=12KV 15KV
conductor areas, and building steel (especially
in the vicinity of lightning downconductors).
100m Distance
Ground connections within facilities should 3KV
500m
only be connected to a single tie point located
at the service entrance panel. This single-point FIGURE 44
connection to earth ground precludes the inad-
vertent development of multiple ground points. Multiple ground points create differences in utility voltage, causing
undesired currents to flow on low voltage data lines. These undesired currents flow in the form of noise which conta-
minates the transmission of data on Local (LAN) and Wide Area Networks (WAN’s).

38
Objectives of Implementing Surge Protective Devices (SPD’s)
In an uncontrolled environment, transient volt-
ages can reach significant amplitudes; many THE THREAT
times greater than the immunity of the elec-
tronic systems. The objective of implementing
Surge Protective Devices (SPD’s) is to
achieve a controlled transient environment,
bridging the gap between the equipment and AC Power Cable
the environment.
The level at which transients are controlled
should always be lower than the immunity of
the equipment.
Ground potential problems may also occur
within a building, particularly if equipment is Potential threat to
supplied from separate transformers. equipment operation
One solution to mitigate the effects of both
coupling mechanisms is to keep high magni-
tude surge currents out of the building.
Transient Equipment
Environment Immunity

FIGURE 45

Protective Shield — the Systematic Approach

The protective shield approach to Transient
Voltage Surge Suppression establishes a THE GOAL
series of defined transient environments or
zones. The two guiding principles are:
1. Keep the highest surge currents out of
inner building wiring.
2. Ensure that at specific interfaces (AC or AC Power Cable
data) transients are controlled to a level SPD
below equipment susceptibility and
vulnerability.

Slightly separated

Transient Equipment
Control Immunity
Level
FIGURE 46

39
PROTECTIVE SHIELD CONCEPT

Structure is Protection Shield #1

Zone 0 Zone 1 (Controlled Environment)
(Uncontrolled
Transient
Environment)

Computer room (shield #2)

Zone 2 Zone 3 CPU

Dataline

Ground Bar Ground Bar

FIGURE 47

ZONE 0 ZONE 1 ZONE 2 ZONE 3

ORDER OF
MAGNITUDE OF X 1000A X 100A X 10A X 1A
SURGE CURRENT

Outside the Building (Zone 0)

Cables, lines, etc., outside a building (underground or aerial) may carry very high magnitude surge currents, particu-
larly during lightning.

Uncontrolled surge currents flowing on inner building wiring will produce both inductive dataline transients and
ground potential differences – a recipe for disaster.

As these cables enter the building, the first protective shield must be established. Ideally, the AC supply and all com-
munication lines should enter the facility at the same general location. Surge Protective Devices are installed at this
point, thereby preventing high-magnitude surge currents from entering. Both AC and communication line surge sup-
pressors should be bonded to the same electrical ground.

On AC systems, it is common practice to install this first line of defense at the service entrance panel. On incoming
telephone lines, the telephone company may provide a course level of protection at the main distribution frame.

40
Within the Building (Inner Zones 1,2,3 etc.)
In large facilities, local control of transients is often mandatory for either specific equipment or for a whole computer
room. The installation of SPD’s on both AC and communication lines, as shown in Figure 48, creates a second zone
with lower transient voltage levels. The surge currents allowed to flow within this area are many orders of magnitude
small than the currents in an uncontrolled environment.
The final protective shield is always the equipment chassis itself, into which some inherent protection has been
designed.

Local Ground Window

In some sites where a systematic approach is not practical or economical, a local protective shield can be estab-
lished for individual equipment. A typical example would be a remote terminal. By using a combined AC and Dataline
SPD, a local ground reference (ground window) is established. The maximum transient voltage present between any
line and local ground within the protective shield is precisely controlled. Therefore, the problem of potential differ-
ences has been eliminated for the equipment within the shield.

“PERCEIVED” Protection
It is important to consider the possible
results of the inappropriate application of
Protective Shield
AC protection products. For example, a
small UPS, containing suppression com-
ponents connected to ground, may cause Combined
ground potential differences. This in turn AC & Dataline
may lead to disruption or damage of Surge Suppressor
dataline ports. There appears to be a sim-
ple solution to this problem: Use products
Terminal
that do not employ suppression compo-
nents connected to ground. Closer inspec- HOT
tion, however, shows that the use of such NEUT
products will generate large neutral to
GND
ground voltages which may overstress
power supplies. As so often happens with Data Line
“band-aid” solutions, the problem is
moved from one area to another.
In reality, the problem of multiple ground
paths already exists in all but the most FIGURE 48
simple facilities. Each item of electronic
equipment contains some filtering and/or
suppression components connected to ground. In a large building where we find a wide variety of computerized
equipment, a multiplicity of current paths to ground exists.
The correct solution to this problem is to prevent high magnitude surge currents flowing on inner building wiring. In
the case of AC, this is achieved by the installation of surge suppression at the service entrance – the systematic
approach.

Effective Protection
Correctly applied, the protective shield approach to surge protection represents one of the most powerful cost effec-
tive tools for preventing equipment disruption, degradation and damage.

41
 Communication Line Protection
for Modems when used with
UPS’s, Power Conditioners & SPD’s

Short Run
Data Cable within
Unprotective Zone UPS Protected Zone

Terminal
Line
Conditioner

Telephone
Wall Jack

5350 MODEM
CPU

2 & 4 Wire Protectors Protective Zone

Available for:
Dial Up, DDS, Leased Line
and T1 Applications

Protection For: Models Available For:

• Modems • Dial-Up (Ringing) Lines
• Point of Sale Terminals • Leased Or Dedicated Lines
• FAX Machines • T1 Applications
• Credit Card Readers etc. • DDS Applications

FIGURE 49

42
 Communication Line Protection
for use with
UPS’s, Power Conditioners & SPD’s
in Network Applications

Short Run
Data Cable within
Unprotective Zone UPS Protected Zone

Terminal
Line
Conditioner
Telephone
Wall Jack

5350 MODEM
5360 CPU
UTP for Token Ring
Data 5360
Wall Jack Hard Wired RS232
5360
IBM BNC 3270

Protective Zone
Add 5360 Modules
As Required...
Up to 4 Modules

FIGURE 50

43
 10 BASE2
An Overview ThinNet
of Ethernet PC PC
LAN
Applications...

PC PC

PC PC

PC PC

10 BASES-T UTP
10 BASE5 BACKBONE
Runs from 1st
10 BASE2
through 4th floors. ThinNet
PC PC

PC PC

10 BASE5 ThickNet
UL 497
PRIMARY
Line
File Conditioner
Server
Zone 0

SPD
SPD
UPS
SPD HOST
SPD
FIGURE 51
WATER/GAS PIPES

Zone 1
BUILDING
GROUND
SPD Surge Protection Devices

44
 Leviton Communications and Data-Line Surge Protective Devices
Stand-Alone and Stackable Modules

Leviton Industry Protocol or No. of Wires or Pins Cable *Operating Maximum Series Clamp Maximum Item Operating Line
Part Number Standard Application Protected (Active Pins) Type Frequency Capacitance Resistance Voltage Surge Current Description Voltage
5350-TLL NEC Art. 800 Telco Leased 4 RJ11 UTP N/A 60pF 22 Ohms 105 Volts (1) 3.8kV/1.9kA 4 Wire Lease Line 70V AC
(Stand-Alone) (UL497A) Lines (2,3,4,5) Connector-RJ11

5350-RS2 EIA RS232 8 Wire RJ45 UTP 1MHz N/A 4.3 Ohms 27 Volts (1) 4kV/2kA 8 Wire RS232 18V DC
(Stand-Alone) (1 turu 8) Maximum Connector-RJ45

5350-TBT IEEE 10 Base T** 4 Wire RJ45 UTP 130 MHz 19pF None 5 Volts (1,2) 2kV/1kA 4 Wire Ethernet 4V DC
(Stand-Alone) (1,2,3,6) Maximum Connector-RJ45 L-L

5350-BNC IEEE Video or 2 (Center + Shield) Coaxial 70MHz 45pF 4.3 Ohms 15 Volts (1,3) 5kA 2 Wire Video/3270 8V DC
(Stand-Alone) IBM 3270 Maximum Connector-BNC

5350-TTR IBM Token Ring 4 Wire RJ45 UTP 130MHz 19pF None 5 Volts (1,2) 2kV/1kA Token Ring 4.4V DC
(Stand-Alone) (3,4,5,6) Maximum Connector-RJ45 F/F L-L

5350-TEB EIA RS422 4 Wire Terminal Strip A/R 70MHz 45pF 4.3 Ohms 15 Volts 4kV/2kA 4 Wire RS422 8V DC
(Stand-Alone) (All) Maximum Terminal Strip (422)

5350-PC IEEE CATV 2 (Center + Shield) Coaxial 1GHz N/A 4.3 Ohms 25 Volts (1,3) 5kA 2 Wire Coax 10V DC
Maximum Connector-COAX-F

5350-SAT IEEE DSS 2 (Center + Shield) Coaxial 1.5 GHz Max. N/A 4.3 Ohms 25 Volts (1,3) 5kA DSS F-Connector 10V DC

5350-PT4 IEEE 4-Wire Telco 4 Wire RJ11 (2,3,4,5) UTP N/A 60pF 22 Ohms 220 Volts (1) 3.8kV/1.9kA 4 Wire Lease Line 90V DC
Connector-RJ11

5350-PT8 IEEE 8-Wire Telco 8 Wire RJ45 UTP N/A 60pF 22 Ohms 220 Volts (1) 3.8kV/1.9kA 8 Wire Lease Line 90V DC
Connector-RJ45

5360-DDS NEC Art. 800 DDS 4 Wire RJ45 UTP 5.6 MHz 60pF 22 Ohms 105 Volts (1) 3.8kV/1.9kA 4 Wire DDS Lines 70V DC
(Stackable) (UL497A) (1,2,7,8 ) Maximum Connector-RJ45

5360-RS2 EIA RS232 8 Wire RJ45 UTP 1MHz N/A 4.3 Ohms 27 Volts (1) 4kV/2kA 8 Wire RS232 18V DC
(Stackable) (1 thru 8) Maximum Connector-RJ45

5360-TBT IEEE 10 Base T** 4 Wire RJ45 UTP 130MHz 19pF None 5 Volts (1,2) 2kV/1kA 4 Wire Ethernet 4V DC
(Stackable) (1,2,3,6) Maximum Connector-RJ45 L-L

NOTES: LAN TERMINOLOGY

(1) Maximum Clamping Voltage measured at 2kV (1.2/50 µs) 1kA (8/20 µs). * Operating Frequency equates to the Maximum Standard Name Convention for IEEE 802.3 & Ethernet
(2) Maximum Clamping Voltage measured between twisted pair during the Bit Rate by using the Formula: ** 10 Base T explanation
common mode application of the test waveform. 10 = Frequency in Mbps
(3) Maximum Clamping Voltage Line to Shield. Base = Signal Type (Baseband or Broadband)
No external power is required for any of the above applications. Max Operating Frequency = Maximum Bit Rate
T = Segment Length in hundreds of meters
5

45
Telco & Data Line Surge Protection Guide

SYMPTOMS: What to look for —

1. Large Loop Areas (Inductive Coupling)
2. Facility Entrance (Copper!)
3. Different Grounds/Multiple Grounds
4. “Noisy” Electrical Environment (i.e.: Ballasts)
5. Multiple Service Entrances
6. Multiple Lines in One Area
7. Line to Ground AC Protectors (electrically pollute
ground references)

Characteristics of EFFECTIVE Surge Protection:

1. Control Transient Voltage.
2. Withstand Electrical Environment.
3. Compatibility.

Reasons for Surge Protector Failures:

1. Poor Transient Control Level — Device clamps above equipment
susceptibility level.
2. Length of Protector’s Ground Wire — avoid pigtail leaded devices!
Voltage drop across ground pigtail is added to overall clamp level.
3. Different Grounds — loop potential differences.
4. Unprotected Lines — “backdoors”, “cover all bases”.
5. Protector circuitry exhibits high capacitance – results in signal
interference.

Approvals compliance for Telco Protectors:

NEC ARTICLE 800 — Communications Circuits
UL 497A — Safety Standard for Secondary Protectors

SYSTEM COMPATIBILITY TABLE: Protocols, Connectors and

Operating voltages, frequencies, transmission speeds.

46
 Chapter 6: Surge Protection for Factory Automation

Introduction
Protection of factory automation equipment follows the same guidelines used for any other type of electronic devices,
since the mainstay of such systems is comprised of one or more programmable logic controllers (PLC’s) working in
conjunction with multiple input/output (I/O) devices. I/O devices are typically controlled by PLC’s. I/O devices perform
a variety of tasks including; the application or removal of voltages to control valves, start, stop or control motors or
vary motor speed upon demand. Other I/O functions include the control or monitoring of manufacturing based
processes such as measurement of product (assembly line) weight (packaged goods products), measurement and
control of solution ph levels, or batch fluid levels during manufacturing or packaging operations.
Automation equipment may be located/housed in an instrumentation rack, or independently mounted in small frame
enclosures. Today many automation systems are personal computer (PC) based, using an open (data line) architec-
ture or priority data protocol to communicate with I/O devices.

Analogies
Protection of a PLC system can be compared
to a castle surrounded by a moat (Fig 1). In
order to effectively protect the castle, all input
and output access points must be guarded.
The same is true for PLC systems. All input
and output lines must be protected from tran-
sients. “All lines” means both power and sig-
nal/data lines. Most often people forget to
protect data lines. Transients do not care
where they enter a system or if the lines are
input, output or power; therefore protection
must be applied to each of them. The costs
associated with protecting all lines will more
than pay for itself if only one damaging tran-
sient is eliminated!

Surge Protection and Grounding

Systems
The start of any evaluation program leading
to an effective surge protection installation is FIGURE 1
to verify that all electronic equipment has
been properly installed and grounded in
accordance with manufacturers instructions
and recommendations.
All electronic system components must be
properly connected to a ground system.This
is the point to which all surges are directed in
their path away from the electronics being
protected. The grounding system must also
be connected to a good building grounding
system that is only connected to an earth
Service PLC Racks
ground rod system, at one point (Fig 2). This Entrance
one point is usually at the service entrance Grounded Wire
location. Randomly grounding building frame-
work to an earth ground rod system can be
counterproductive, as it could lead to the Earth Ground Rod System
formation of ground loop currents that can
corrupt data line (LAN, WAN) communication FIGURE 2
within the building.

47
Relatively speaking, protection of input power lines to PLC or PC based systems is the most straight foward. Placing
a strip, plug-in (for PC’s) or wired-in device (for rack mounted PLC’s) in the power line forms a defense against both
internal and externally generated facility transients. It is also preferred to have a panel mounted SPD located at
branch and/or service entrance locations. The SPD placed at the PLC or PC will act as a downstream networked
device that mitigates the effects of any remnant transient that may pass branch or service entrance mounted SPD’s.
Data line or signal line SPD’s must be mounted (and grounded) as close to PLC or I/O cards as possible, and at
each end of the line (when such lines exceed 5 ft in length). Having SPD’s grounded near PLC’s or I/O cards is
especially important when data or signal inter-connection lines are unshielded, and also when they are required to
exit the shielded confines of their grounded rack enclosure.

Data Line SPD Selection Criteria

Critical parameters when selected SPD’s for non standard data protocols include:
• Peak Voltage Levels
• Frequency of Operation
• Protection Modes (Line-Line, Line-Ground)
• Capacitance Loading L-L and or L-G
• Type of Fusing
• Underwriters Laboratory (UL) Listing Per UL 497B
• CE Marking (Required for European Operation)
• Clamping Voltage Levels with associated test current and test waveform

48
 Chapter 7: SPD Application Strategies And Installation Practices

Accurate Assessment of Electrical Systems

Precise identification of the electrical system is critical in the proper selection and application of the most effective
surge protective device. Since SPD performance is directly related to nominal line-voltage parameters, line voltage
measurements should always be taken by a qualified electrician before the SPD’s are specified and installed.
Measurements should be taken even when the electrical system configuration is known.

Figure 52 shows some common electrical systems used throughout the USA and Canada. Specific systems present
in a given facility should be identified by measuring with a voltage meter across each line to neutral, and line to line.
The readings should be written down and referred to when ordering the required SPD’s.

Voltage Measurements Can Be

EXTREMELY DANGEROUS! PLEASE BE VERY CAREFUL!

Single Phase 2-Wire COMMON VOLTAGES (VAC)

L
V(L-N) 120,220,277,346
V(L-N) or V(L-L) V(L-N) 208,240,380,480,600
N or L
Single Phase 3-Wire
L V(L-N) 120 240
V(L-N)
N V(L-L) V(L-L) 240 480
V(L-N)
L

Three Phase 4-Wire WYE

L
V(L-L) V(L-N) 120 277 346
L V(L-L) 208 480 600
V(L-N) V(L-L)
L
N

Three Phase 3-Wire Delta

L V(L-L) 240 480 600
V(L-L)
L
V(L-L)
L

Three Phase 4-Wire Delta (Split Phase)

L
V(L-N)1 120
V(L-N) 1
L V(L-N)2 208
V(L-N) 2 V(L-N)3 120
L
V(L-N) 3
N

FIGURE 52

49
SPD Networks
PI Filters and SPD Networks
Remarkable improvement can be made in SPD performance
through the use of a relatively simple concept: inductive cou-
pling. With this concept, SPD networks can be created within
SPD circuit designs or which span entire electrical systems and TRADITIONAL LOW-PASS
utilize the existing features and components of the system to PI NETWORK FILTER
great advantage.

This concept is certainly not new and has its roots in electrical
engineering circuit design. For example, inductively coupled net-
works are often used to achieve enhanced levels of perfor- SPD SPD
mance beyond that possible through multiplication of compo-
nents in a circuit. In the traditional PI Network, an inductor “cou-
ples” the two identical capacitors of value C, yielding better low- TRANSFORMER - INDUCTANCE
pass filter performance than could otherwise be achieved by a COUPLED TVSS NETWORK
single capacitor of size 2C.

If the capacitors of the PI Network filter are replaced by SPD components, the result is a “coupled” SPD Network
with better overall performance capability than a single SPD module.

Simply adding and additional SPD module to a single-module circuit without the inductor would result in only slight
improvements.

In PI Network Filters or SPD Networks, the inductor serves as the coupling medium and is primarily responsible for
the increase in performance capability, though “matching” of the inductor is an engineering challenge.

Component-Based Hybrid Networks

SPD Hybrid circuits which utilize inductors may be regarded as SPD “Component Networks”

MOV, DIODE SPD HYBRID

LINE INDUCTOR

L
LOAD
MOV
PRIMARY
SILICONE
DIODE
SECONDARY
N

50
Transformer-Based SPD Networks

A convenient way of achieving SPD Network “coupling” is through the use of a transformer. Because of its unique
power conditioning and electrical properties, the transformer is an ideal coupling medium.

In the single-phase, progressive action schematic illustration below, a 6800 Volt transient of specified energy content
(perhaps the residual of lighting arrestor activity at the service entrance) enters the electrical system through the ser-
vice panel and is attenuated by an SPD Network consisting of two appropriately sized SPD modules and a shielded,
step-down transformer.

Had only one side of the transformer been “protect-

6800 V
1700 V
ed”, the resulting stress at the load would have been
500 A
TRANSFORMER much higher. For example, with the 480V module
1600 V
alone, the resulting 1600V transient would have had
L 320 V LOAD access to the load. With the 120V module alone, the
resultant or clamping voltage of the module would
120 VOLTS
480 VOLTS

have been significantly higher than 320 Volts since

Current
Current
Shunt
SPD Shunt SPD the module would have had to contend with the
SERVICE
PANEL entire energy content of the original transient. The
attenuating effect of the transformer on clamping har-
N monics and random noised associated with SPD
activity would not have been realized.

An installation schematic for a Three-Phase,

Transformer-Based SPD Network is shown below.

480/120 VAC TRANSFORMER

L3 L3
L2 L2

N Conduit connected
L1
L1 Safety Ground

G
L1 L1
SERVICE PANEL
BP-4 SPD

SERVICE PANEL

L2 1 1
BP-4 SPD

L2
L3 2 2
L3
N 3 3 N
G
N N G

51
Branch Circuit Based SPD Networks
Transformers are not absolutely necessary in achieving optimum SPD Network performance. Since it is inductance
which is primarily responsible for the “coupling” effect, very effective networks can be developed using the natural
inductance of branch circuit wiring.
In the example below, two networks are established. One is between the service entrance lightning arrestor and the
distribution panel SPD, the other is between each of the two loads and the distribution panel SPD. Of course, the
overall effectiveness of these networks depends upon the relative capabilities of the SPD components involved and
the actual inductance (or length) of the branch circuit wiring.

SERVICE ENTRANCE
LIGHTING ARRESTOR
SPD ZONE
OF EFFECT

PLUGSTRIP SPD
3 PHASE
POWER

Terminal
SERVICE
PANEL

PROCESS
PANEL CONTROLLER
MOUNT JUNCTION
SPD BOX

OEM SWITCH MODE

SPD POWER SUPPLY
SPD ZONE
OF EFFECT

Network Zones of Effect

In the above example, a zone called the SPD “Zone of Effect” or “Zone of Effectiveness”, is depicted about the point
of application of the secondary or load-side SPD device and next to the specific loads requiring protection.
The zones pictured indicate the area or greatest protection in the network. Typically, networks are derived in order to
provide enhanced protection for a specific load. All loads connected to the branch circuit within the network are SPD
protected, but the greatest or optimum protection is in the vicinity of the secondary or load-side SPD within the zone.
This is why SPD networks are generally constructed from a point where specific protection is needed, backwards
through the electrical system, to a panel or a transformer.
How these zones work and their performance (protection) characteristics are illustrated in the example on the next page.
52
Branch Circuit Network Example

SURGE In order to illustrate the properties of a Branch Circuit Network, 50 feet

GENERATOR of #14 AWG 3 conductor wire is laid out from points A to B. A surge
generator with back-filter is employed to inject IEEE C62.41 type transients
onto the sine wave at 90 degrees of phase angle. Simple SPD’s are
employed at both A and B as indicated below.

BACK
FILTER INPUT: IEEE C62.41 CAT. A
AC RINGWAVE, 6KV, 200 A
L = .6µH
50 FEET #14 AWG, 3 CONDUCTOR
POWER A B

1µs 1µs

200V 488.0V 200V 606.0V

OUTPUT
@ B W/O
SPD

OUTPUT @ A WITH SPD 5µs

200V 268.0V
1µs

OUTPUT
2.880V @ B WITH
SPD

5µs

200V 346.0V
AMPERAGE WAVEFORM @ A
WITH 100 A / V OUTPUT
@ B WITH
SPD
BUT W/O
SPD @ A

V = L (di / dt) = 0.6 µH (200 A / 600 ns) = 200 VOLTS

53
Applications for Commercial/Light Industrial Facility

4.

2.

3.

1.

1. Power Distribution 3. Building Controls

■ Load centers ■ Building management systems
■ Switchgear ■ Motor controls
■ Distribution equipment ■ Drives

2. Electrical Loads 4. Computers & Communications

■ Lamps ■ Voice/data equipment
■ Ballasts (PBX system, LAN’s, WAN’s)
■ Lighting controls ■ Security system
■ Fire alarm system
■ Patch Panels
■ UPS systems

54
SPD Network Example: Commercial/Light Industrial Facility
L Indicates suggested placement of
120/208 Leviton SPD product. Other solutions may
200A also apply. Attention should be paid to
available options with panel mounted SPD.

L 52120-M3

L 5200P L 5280 L 5200PS

120V
13A

PROCESS
PC
COPIER

INPUT
TO MINI

L 5400 L 5400 L 5400

PROCESS PROCESS PROCESS

PC PC PC

INPUT INPUT INPUT

DATA BUSS TO MINI

L SECURITY
51020-WM SYSTEM
WAREHOUSE
INVENTORY CONTROL
5350
1000 FEET
TELEPHONE L L
KSU 5350-PT4

L 5350-PT4
L L L
MODEM

500-LC
1600-LC
MODEM

5350
MINI 5360 PROCESS
120V COMPUTER 5360 PC
20A
DATA BUSS TO
PROCESS PC'S

(DATA CONNECTIONS INPUT

VIA RS 232)

55
Applications for Industrial Facility

3.
2.
1.
4.

1. Power Distribution 3. Industrial Controls

■ Load centers ■ Programmable logic controllers
■ Transformers ■ Motor controls
■ Generators ■ Motor control centers
■ Variable-frequency drives
Electrical Loads ■ Proximity sensors
2.
■ Motors ■ Bar coding equipment
■ Industrial
fans, heaters ■ Machine vision systems
and blowers
■ Ballasts Computers & Communications
4.
■ LAN’s and WAN’s
■ Voice/data (PBX)
■ Fire alarm systems
■ Intercoms
■ Security systems
■ UPS systems
■ Building management systems
56
SPD Network Example: Industrial Facility L Indicates suggested placement of
Leviton SPD product. Other solutions may
also apply. Attention should be paid to
available options with panel mounted SPD.

PROCESS
CONTROL
L 57480-DM3 PANELS
480 VOLT L
3 PHASE
1000 A
51020-WM
STEP- STEP- OR
DOWN DOWN 1200-LC
480V TRANS- TRANS-
3 PHASE FORMER FORMER 120V
STEP- 15A
DOWN
52120-M2 240V
L DELTA L
TRANS-
120V FORMER
1 PHASE 32277-DY3
8A L
WORKBENCH 42412-DS3
CNC
MACHINE WORKSTATION
ARC 240V
TOOL CNC DELTA
CONTROL WELDER
PROCESS SPLIT
CPU PC PHASE

L 120V 5300-IPS
51020-WM 1 PHASE L
20A

INTERGRATED
L PROCESS
MACHINE
5100-IPS
TOOL
WITH
BUILT-IN
CPU
PRINTER
PROCESS
PC

INPUT

57
SPD Network Example: Medical Radiology Environment
L Indicates suggested placement of
480 VOLT Leviton SPD product. Other solutions may
3 PHASE also apply. Attention should be paid to
available options with panel mounted SPD.

480V X-RAY
L 3 PHASE GENERATOR / CAT SCAN / MRI
57480-DM3
(For service up to 100A wired via "feed through" lugs)

L 57480-DM3

GANTRY
480V
3 PHASE

120/208
STEP 3 PHASE
DOWN L DISTRIBUTION
TRANSFORMER PANEL
57120-M3 TABLE
120V
SINGLE PHASE
OUTPUT

L
208V 42120
DISC
DRIVES

L SYSTEM
42120 CPU

PROCESS DATA LINES

L
5350 DATA LINE PROTECTION
SPECIFIED FOR APPROPIATE PROTOCOL
120V
20A

8380 SURGE
PROTECTED L L
HOSP. GRADE 5300P
OUTLET

EQUIPMENT CART
L 5300-HPS HOSPITAL GRADE
PROCESS
SURGE PROTECTED STRIP
PC

58
SPD Network Example: Retail Check-Out L Indicates suggested placement of
Leviton SPD product. Other solutions may
also apply. Attention should be paid to
available options with panel mounted SPD.

5300-PS
PANEL 1 PROCESS
L PC

L
52120-M3 TO PAYROLL CIRCUITS

5280
L
480V TELEPHONE
ISO. KSU
3 PHASE PANEL 2 TRANS
STEP-
DOWN 277/480 LIGHTING
L 3 PHASE L
TRANS.
52277-M3

52480-DM3
L SECURITY
51020-WM SYSTEM

POINT
PANEL 3 CONNECTIONS

L
PANEL 4 52120-M3 120V 5350-PT4 L
STORE
REGISTER MODEM
120V MINI
120V
COMPUTER MAIN
DEDICATED
STORE
LINES
CONNECTION

L 32120-DY3 REGISTER CONNECTIONS RS-232

L 5350 DATA LINE PROTECTORS

HEAVY
LOADS
REGISTER
EXAMPLE: REFRIGERATION SYSTEM
COMPRESSORS
INCLUDES
500-LC L 500-LC L 500-LC L CARD
READER

CHECK-OUT CHECK-OUT CHECK-OUT

REGISTER REGISTER REGISTER

INPUT INPUT INPUT

5350 5350 5350

DATA LINE DATA LINE DATA LINE
PROTECTORS PROTECTORS PROTECTORS
MATCHED
TO APPROPIATE
PROTOCOL

59
SPD Network Example: Home Electrical System
L Indicates suggested placement of
Leviton SPD product. Other solutions may
also apply. Attention should be paid to
available options with panel mounted SPD.

L L L
5280 5280 5280

METER
REF.
L TV
51120-3

MICROWAVE

120/240
1 PHASE
200A
20 AMP
L
5380
20 AMP

STERO
FURNACE HOME
ENTERTAIN- DRYER WASHER
MENT
CENTER

5300-PSC
L TO DISH
5200P L COAX L
CABLE
5350-SAT

SATELLITE
CONTROLLER

5280 SECURITY TO POINT

L SYSTEM CONTACTS

VCR

HOME
OFFICE

5300-PST BIG
MODEM PC SCREEN
L
TV

60
Chapter 8: Reference

ABBREVIATIONS, REFERENCES AND DEFINITIONS

Abbreviations

ANSI: American National Standards Institute

CSA Canadian Standards Association
EMP Electromagnetic Pulse
EMI Electromagnetic Interference
IEC: International Electrotechnical Commission
IEEE Institute of Electrical and Electronics Engineers
NEMA National Electrical Manufacturers Association
RFI Radio Frequency Interference
UL Underwriters Laboratories

References

Institute of Electrical and Electronics Engineers (IEEE) Standard 100-1988

Standard Dictionary of Electrical and Electronic Terms

C62-IEEE Guides and Standards for Surge Protection

C62.41-IEEE Guide for Surge Voltages in Low Voltage AC Power Circuits

C62.45-IEEE Guide on Surge Testing for Equipment Connected to Low Voltage AC Power Circuits

UL 96 Standard for Safety-Installation Requirements for Lightning Protection Systems

UL 452 Standard for Safety-Antenna Discharge Units

UL 497A Standard for Safety-Secondary Protectors for Communication Circuits

UL 498 Standard for Safety-Receptacle and Receptacle Plugs (Including Direct Plug-In Devices)

UL 544 Standard for Safety—Medical and Dental Equipment

UL 1283 Standard for Safety—Electromagnetic Interference Filters

UL 1363 Standard for Safety—Temporary Power Taps (Power Strips)

UL 1449 Standard for Safety—Transient Voltage Surge Suppressors

NEMA LS-1 Low Voltage Surge Protective Devices

61
 Definitions:

(Per IEEE Standard 100 and IEEE 1100)

Capacitance: The property of a system of conductors and dielectrics that permits the storage of electrically separat-
ed changes when potential differences exist between the conductors.

Common-Mode Noise: The noise voltage that appears equally and in phase from each current-carrying conductor
to ground.

Crest Factor: Ratio between the peak value (crest) and rms value of a periodic waveform.

Differential-Mode Noise: See: noise, transverse-mode

Decibel (dB): One-tenth of a bel, the number of decibels denoting the ration of the two amounts of power being ten
times the logarithm to the base 10 of this ratio.

PowerdB=10 Log10 (Powerout/Powerin)

VoltagedB=20 Log10 (Voltageout/Voltagein)

Equipment Grounding Conductor: The conductor used to connect the noncurrent carrying parts of conduits, race-
ways, and equipment enclosures to the grounding electrode at the service equipment (main panel ) or secondary of
a separately derived system (e.g., isolation transformer). (The term is defined more specifically in the NEC (2),
Section 100).

Impulse–(Surge Arrestors): A surge of unidirectional wave of current or voltage of very short duration containing
no appreciable oscillatory components.

Inductance: The property of an electric current by virtue of which a varying current induces a electromotive force in
that circuit or in a neighboring circuit.

Ground/Ground system

Ground: A conducting connection, whether intentional or accidental, by which an electric circuit or equip-
ment is connected to the earth, or to some conducting body of relatively large extent that serves in place of
the earth.

Ground Loop: A potentially detrimental loop formed when two or more points in an electrical system that
are nominally at ground potential are connected by a conducting path such that either or both points are not
at the same ground potential. (4)

Ground System: A system in which a least one conductor or point (usually the middle wire of neutral point
of transformer or generator windings) is intentionally grounded, either solidly or through an impedance.

Grounded Conductor: Another name for the neutral conductor. A conductor which is intentionally ground-
ed, either solidly or through a non-interrupting current limiting device.

Grounded Conductor–(NEC): The conductor that is used to connect the equipment or the wiring system
with a grounding circuit to a grounding electrode or electrodes.

Solidly Grounded: Connected directly through an adequate ground connection in which no impedance has
been intentionally inserted.

62
Harmonic: A sinusoidal component of a periodic wave or quantity having a frequency that is an integral multiple of
the fundamental frequency.
Impulse: (Metal Oxide Varistors) A surge of unidirectional polarity.
Input Power Factor (of a system): Specifies the ratio of input kilowatts to input kilovoltamperes at rated or speci-
fied voltage and load.
Induced Current: (General)—Current in a conductor due to the application of a time-varying electromagnetic field.
Induced Voltage: (General)—A voltage produced around a closed path or circuit by change in magnetic flux linking
that path.
Induced Voltage: (Lightning Strokes)—The voltage induced on a network or electric installation by an indirect
stroke.
Inrush: The amount of current that a load draws when it is first turned on.
Isokeraunic Level: (Lightning)—The average annual number of thunderstorm days.
Isolated Equipment Ground: An insulated equipment grounding conductor run in the same conduit or raceway as
the supply conductors. This conductor is insulated from the metallic raceway and all ground points throughout its
length. It originates at an isolated ground type receptacle or equipment input terminal block and terminates at the
point where neutral and ground are bonded at the power source. (This term is defined more specifically in the NEC
(2), Section 250-74 and 250-75).
Isolation: Separation of one section of a system from undesired influences of other sections.
Isolation Transformers: Provides a local ground reference point. Attenuates common-mode disturbances on the
power supply conductors.
Joule: The work done when the point of application of a force of one newton is displaced a distance of one meter in
the direction of the force. (A newton is that force when applied to body having a mass of one kilogram, gives it an
acceleration of one meter per second squared.)
Leakage Current: (Health Care Facilities) This is any current, including capacitively coupled current, not intended
to be applied to a patient but which may be conveyed from exposed metal parts of an appliance to ground or to
other accessible part of an appliance.
Linear Load: An electrical load device which, in steady state operation presents an essentially constant load
impedance to the power source throughout the cycle of applied voltage.
Maximum Continuous Operating Voltage: The maximum designated root mean square (rms) value of power fre-
quency voltage that may be applied continuously between the terminals of an arrester.
Mutual Inductance: The common property of two electric circuits whereby an electromotive force is induced in one
circuit by a change of current in the other circuit. The coefficient of mutual inductance M between two windings is
given by the following equation:
emf=-M (di/dt)
Noise: Electrical noise is unwanted electrical signals that produce undesirable effects in the circuits of the control
system in which they occur (4). (For this Recommended Practice, “Control systems” is intended to include sensitive
electronic equipment in total or in part.)
Noise, Normal-Mode: See: transverse-mode noise
Nonlinear Load: Electrical load that draws current discontinuously or whose impedance varies during the cycle of
the input as voltage waveform.
Residual Voltage: (Arrester) The voltage that appears between the line and ground terminals of an arrester during
the passage of discharge current.
Safety Ground: See: equipment grounding conductor

63
Self-Inductance: (Inductance) The property of an electric circuit whereby an electromotive force is induced in that
by a change of current in the circuit. The coefficient of self-inductance ”L” of a winding is given by the following
expression:
Voltage (e)=-L (di/dt)
Shield: As normally applied to instrumentation cables, a conductive sheath (usually metallic) applied over the insu-
lation of a conductor or conductors, for the purpose of providing means to reduce coupling between the conductors
so shielded and other conductors that may be susceptible to, or that may be generating unwanted electrostatic or
electromagnetic fields (noise).
Shielding: Shielding is the use of a conducting barrier between a potentially disturbing noise source and sensitive
circuitry. Shields are used to protect cables (data and power) and electronic circuits. They may be in the form of
metal barriers enclosures, or wrappings around source circuits and receiving circuits.
Spike: (pulse terms) A distortion in the form of a pulse waveform of relatively short duration superimposed on an
otherwise regular or desired pulse waveform.
Surge Impedance: The ratio between voltage and current of a wave that travels on a line of infinite length and of
the same characteristics as the relevant line.
Transient Overvoltage: The peak voltage during the transient conditions resulting from the operation of a switching device.
Transverse-Mode Noise: (With reference to load device input ac power) Noise signals measurable between or
among active circuit conductors feeding the subject load but not between the equipment grounding conductor or
associated signal reference structure and the active circuit conductors.
Voltage Regulation: The degree of control or stability of the rms voltage at the load. Often specified in relation to
other parameters such as input-voltage changes, load changes, or temperature changes.
Watt: The unit of power in the International System of units (SI). The watt is the power required to do work at the
rate of 1 joule per second.
Withstand Current: The crest value attained by a surge of a given wave shape and polarity that does not cause
disruptive discharge on the test specimen.
Withstand Voltage: The specified voltage that, under specified conditions, can be applied to insulation without
causing flashover or puncture.
Sag: A rms reduction in the AC voltage, at the power frequency, for durations from half-cycle to a few seconds.
(Note IEC terminology is dip).
Swell: An increase in the AC voltage, at the power frequency, for durations from a half-cycle to a few seconds.
Transient: A sub cycle disturbance in the AC waveform that is evidenced by a sharp brief discontinuity of the wave-
form. May be of either polarity and may be additive to or subtractive from the nominal waveform.

General Information:
Clamp Circuit: Circuit which limits the applied voltage amplitude to a desired level, based on the transients intrin-
sic current.
Failure Mode: The effect by which a failure is observed.
Clamping Voltage: The (peak) voltage occurring on a conductor, measured at the output of a Surge Protection
Device (the point connected to the load/device being protected), to either the ground or neutral conductor.
Residual (voltage): The amplitude (level) that remains after a Surge Protective Device has attenuated the ini-
tial transient.
Static Charge: The electricity generated when two dissimilar substances come into contact. (Conveyor belts are
active producers of static electricity).
Transient Overvoltage: The temporary overvoltage of short duration associated with the operation of a switching
device, fault, lighting, or arcing ground faults.

64
Leviton Manufacturing Co., Inc.
59-25 Little Neck Parkway, Little Neck, NY 11362-2591
1-800-323-8920 • FAX 1-800-832-9538
Visit Leviton’s Web site at http://www.leviton.com

TECH LINE: 1-800-824-3005 • 8:30 am — 6:00 pm Eastern Time

Leviton Mfg. of Canada, Ltd.
165 Hymus Boulevard, Pointe Claire, Quebec H9R1G2
1-514-954-1840 • FAX 1-514-954-1853 ©1998 Leviton Manufacturing Co., Inc. All Rights Reserved G-5337/J8-10M rs