STD 1100-2005 - Part1
STD 1100-2005 - Part1
Std 1100-2005
CHAPTER 4
Considerable current can oscillate between the inductive and capacitive storage elements of the circuit when
nonlinear loads, with a characteristic harmonic near the parallel-resonant frequency, exist in the circuit.
Voltage distortion results from these high oscillating current levels. Under certain conditions, the oscillating
currents can also emit electromagnetic energy, which can interfere with adjacent signal circuits.
In summary, and by comparison, series-resonant circuit currents oscillate through the ac supply system paths
and their source, while parallel-resonant circuits confine such current to the parallel circuits own loop.
Therefore, series-resonant circuits involve the supply, load, and intervening wiring (and all power transport
components in the wiring path), with current at the resonant frequency; and parallel-resonant circuits
impress voltages (at the resonant frequency) on their source, load, and on the wiring system. These two
conditions represent the underlying mechanism for the production of the most common forms of harmonic
voltage waveform distortion on the ac wiring system.
4.7.5.3 Resonance on feeders and branch circuits
The conductors used to form feeders and branch circuits possess both distributed self-inductance and
distributed capacitance. These are called stray or parasitic reactances. The self-inductance portion is series
distributed in longitudinal fashion along the length of the feeder, branch circuit, or both, while the
capacitance portion is shunt distributed between all conductors as well as to equipment ground along the
same path. Equipment ground is defined as any enclosing metallic raceway, a green wire (e.g., an EGC), or
nearby grounded metal if the raceway is nonmetallic. This arrangement forms into a transmission line with
reactive circuit elements connected across the ends in the form of the ac power source and load. Both seriesand parallel-resonant conditions are thus capable of occurring under proper conditions of excitement.
Exciting current is generally provided by switching in the load-source current path, although an exciting
current can also be introduced from the equipment ground path via the shunt capacitance.
Loads, such as ASDs for motors, are known to be capable of producing high-frequency currents sufficient to
excite the resonant circuit in the feeder, branch circuit, or both. Reflected waves from the motor and power
source that oscillate through the intervening wiring path are also associated with this action. This is
particularly true for the modern IGBT (or bipolar) drive used with pulse-width modulation (PWM)
techniques since very fast transition times can be involved during the transistors switching between on and
off states (on the order of between 50 ns and 200 ns), with switching frequencies of between 2 kHz and
20 kHz. The natural resonant frequency of the typical long branch or feeder circuit between 15 m and
1000 m (49.2 ft and 3280 ft) and the involved ac source-ASD combination is typically on the order of from
1 MHz to 10 MHz (see von Jouanne et al. [B67]).
The only real damping or limit on the Q of the resonant circuit in the feeder or branch circuit is generally
provided by the paths resistance, which is in series with the oscillating current, and this is typically not
sufficient by itself to provide rapid damping. As a result, voltage impulses (e.g., transients) on the order of
from 1.3 kV to 1.55 kV can appear across the tank circuit and hence the ASD load terminals (see von
Jouanne et al. [B67]). This is highly detrimental to the motor insulation life span and can have similar
detrimental or disruptive effects to any electronic loads connected into the same circuit.
The foregoing condition can generally be ameliorated by the application of a three-phase, E-core, seriesconnected choke of commercial design that provides some additional series impedance and acts as a lowpass filter in the path of interest. Alternately, a physically smaller filter can be applied that is tuned to once
and twice the carrier frequency of PWM-type drives. The low-pass characteristic of the arrangement ensures
that only the unwanted high-frequency components of the current are attenuated and not the lower frequency
ones involved with the efficient transmission of electrical power between the source and load.
Lightning, and in some cases ground faults, can similarly excite the long feeder or branch circuit resonances
via the shunt capacitance path from ground. Near-field coupling in the H-field from nearby sources of highfrequency noise sources can also induce excitation current into the self-inductance of the resonant circuit of
the feeder or branch circuit.
120
FUNDAMENTALS
IEEE
Std 1100-2005
b)
c)
Note that for well-established purposes of fire and safety from shock, these functional grounding subsystems
are all eventually solidly interconnected to a common earth electrode system at the sites service entrance
section (SES) in accordance with the requirements of the NEC, NFPA 780, or both. Per the NEC, other earth
grounding electrodes may also be involved with the various grounding subsystems (see IEEE Std 142), but
these cannot in any way be isolated (e.g., by electrical isolation or earth resistivity) from one another, or
from the sites main earth grounding electrode at the SES.
Copyright 2006 IEEE. All rights reserved.
121
IEEE
Std 1100-2005
CHAPTER 4
All of the previously described grounding systems must be made electrically common to one another by use
of a grounding conductor(s) so as to form a single, interconnected earth grounding electrode system at the
site. Failure to provide this necessary bonding will place whatever equipment that, or personnel who, may be
deliberately or accidentally connected between the two IG systems at whatever potential that can be
developed between them. Such a potential is typically due to ac or dc system ground faults, lightning, or
other currents that can produce a significant IR drop when flowing in the commonly shared grounding
mediumearth.
The possibility of a voltage difference still exists even when the earth grounding electrode(s) provided have
excellent low-resistance connectivity to the earth, such as in the 1 1 range. For example, a lightning strike
near one electrode with a conservative 20 kA to 40 kA would produce a ground-rise (e.g., offset) potential
of between 20 kV and 40 kV between two earth grounding electrodes with only 1 1 of resistance between
them (e.g., E = IR). This is too high a potential for any reasonably constructed signal-level insulation system
to withstand, let alone any connected solid-state electronic equipment components. Alternately, such high
potentials
entials are known to produce arcing and dielectric breakdown at points along the currents path and
within equipment, and which then may cause fires.6
4.8.2 Earth electrode subsystem
The earth electrode subsystem establishes the facility earth ground reference for lightning, electrical fire,
and shock hazard purposes (i.e., safety purposes). Signal transport processes and the internal signal
processes of equipment are not benefited by this system nor connections made to it except from a safety
standpoint. However, improper connection of these portions of an electronic system into an earth electrode
subsystem is widely known to produce performance, safety, and equipment damage problems under
conditions of power system switching and ground-fault conditions, or lightning. Specific minimum
requirements for the earth electrode subsystem are provided in the NEC.
Ground rods and other types of made electrodes connect the grounding system to the earth. There can be a
great deal of variation in the resistance (and impedance) the ground rod has with respect to the earth. Soil
conditions, particularly with respect to moisture, can cause significant variation in the resistance of a
particular ground rod through the course of the year. In some cases, the soil itself is such a poor conductor
that special grounding methods, such as chemical ground wells, may be necessary to provide an adequate
connection. A typical example of this is in some parts of Arizona and New Mexico, where there is sandy soil
that is very dry most of the year. Both states have periods during the year in which they have a significant
amount of lightning. The resistance and impedance that the grounding system has to the earth is an
important issue in dissipating the energy of a lightning strike and minimizing the voltage gradients.
Testing of various types of ground rods and other types of electrodes has been conducted to determine which
type performs the best. The concrete-encased electrode has been one of the best at maintaining a consistent
resistance to the earth. Concrete is porous, which traps moisture, and therefore, never totally cures. From
this it can be seen that it is a good practice to embed the bare grounding conductors in concrete where the
opportunity presents itself, such as with new construction.
The earth connection generally exhibits an increasing impedance with frequency (see Figure 4-52). This
absolutely limits the effectiveness of the earth grounding electrode in relation to high-frequency noise
control efforts. This clearly means that the earth grounding electrode system alone is not an effective means
of controlling the unwanted effects associated with the higher frequency components of a lightning surge
event.
6The problem of dielectric breakdown, arcing, and possible fire reaches a high level of probability within electrical or electronic equipment where any insulated conductor referenced to a separate, isolated earth grounding electrode system is brought into an equipment
enclosure and where that equipment is itself equipment grounded by connection to the NEC-described safety grounding system consisting of the metal conduit/raceway, EGC system (e.g., green wires), and equipment enclosure.
122
FUNDAMENTALS
IEEE
Std 1100-2005
From a wave-theory standpoint, the earth and the connection to it represents a serious impedance
discontinuity to almost all short-duration, fast transition time impulse currents. Hence, the connection
generates large reflections of both current and voltage when such waveforms are imposed on the earth
grounding electrode from either the direction of the earth or the connecting grounding electrode conductor
acting as an unbalanced transmission line. Such reflections are then propagated back into the grounding
electrode conductor where they are spread to all items of equipment connected into it via the power system
and any related EGC networks.
The earth grounding electrode subsystem alone is not generally capable of controlling the hazards associated
with power system ground faults. This is reflected in the NEC requirements that clearly state that the earth
shall not (ever) be used as the sole path for a current flow. Instead, good grounding/bonding techniques are
required as are discussed elsewhere in this chapter, and in Chapter 8 and Chapter 9.
123
IEEE
Std 1100-2005
CHAPTER 4
Per the foregoing, the common-mode voltage situation is actually quite serious. Take the example of an
insulated and isolated signal ground terminal located within equipment that is metal enclosed and properly
connected into the ac power systems EGC system. If the insulated-isolated signal ground terminal is
connected into an isolated earth grounding electrode (a non-recommended practice in all cases), then any
current flow in the impedance between the power systems earth grounding electrode and the signal
terminals isolated earth grounding electrode will produce a common-mode voltage between the terminal
and the equipments frame/enclosure in which it is installed. Since signal level wiring and components are
not normally provided with other than low-voltage insulation means, and minimum air spacing from
exposed terminals to the frame/enclosure are also the rule, a very real probability of voltage breakdown
exists between the two (e.g., kVG1G2 = kAG1G2 1G1G2 is the case with the two separate grounding
systems). Thus, the signal level circuits and associated logic and signal level semiconductors, etc., are
placed into great risk by this approach, and are not benefited in any manner.
4.8.3 Basic grounding and bonding concepts
There are actually two sections in a proper grounding system: the connection to the earth with an electrode
(grounding) and what the NEC calls bonding (see Figure 4-53):
The permanent joining of metallic parts to form an electrically conductive path that will ensure
electrical continuity and the capacity to conduct safely any current likely to be imposed.
The NEC defines ground as (see Figure 4-54):
A conducting connection, whether intentional or accidental, between an electrical circuit or equipment and the earth, or to some conducting body that serves in place of the earth.
It would eliminate some of the confusion of terminology if we referred to grounding as earthing when we
were talking about the connection into the earth. Notice also that the NEC says or to some conducting
body. In the case of an airplane, it would be the frame and the outer skin. For a ship, it would be the hull.
Figure 4-53Bonding
124
IEEE
Std 1100-2005
FUNDAMENTALS
Figure 4-54Grounded
Bonding is essential to ensure overcurrent protective device operation and prevent shock hazards when
electrical equipment fails. By bonding the outside metal parts (to the grounding system) with a conductor of
sufficiently low impedance and sized to carry the fault current and allow the overcurrent protection to
operate, a person in contact with the faulted equipment is not exposed to hazardous voltages. The NEC
would consider this effectively grounded.
Many of the grounding problems electronic equipment experiences that prevent proper operation are
bonding problems per NEC terminology. Once again the key question is: Is the earth in the circuit?
Lowering the resistance of the grounding system to the earth only impacts problems in which the earth is in
the circuit.
For purposes of grounding, effective bonding consists of a set of grounding conductor interconnections and
terminations that, taken together, form a usefully low-impedance path at all frequencies of interest, for the
flow of current through them. If done properly, this arrangement then can be used to successfully limit the
development of unwanted potentials across the ends of the bonding connection.
The objective is that each termination (bond) be such that the electrical properties of the total path are a
function of all of the connected elements and not just the interconnections. Conversely, poor bonding is
often the principal cause of many hazardous and noise-producing situations, e.g., unacceptable voltage
drops, heat generation, intermittent operation, electrical noise, and high-resistance grounds. A large crosssectional area grounding conductor is of little use if it is terminated via a poorly bonded connection.
It must be stressed that the low- and high-frequency characteristics of most grounding and related bonding
techniques are quite different. Their high-frequency characteristics are of particular importance for most
electronic equipment applications where both low-level and high-level, short-duration, rapid transition time
Copyright 2006 IEEE. All rights reserved.
125
IEEE
Std 1100-2005
CHAPTER 4
impulses are the typical currents propagated through the bonding conductor (see MIL-STD-188-124A
[B47]).
The following factors are important when characterizing alternative bonding methods:
a)
Contact resistance. Contact resistance of conductor and shield terminations, and their aging, are of
importance.
b)
Dissimilar materials. Dissimilar materials are problematic in that they often set up galvanic halfcells or rectifying junctions that result in EMI generation at the junction, corrosive failure of the
connection, or both.
c)
Skin effect. High-frequency currents do not penetrate deeply into high-conductivity materials.
Therefore, the high-frequency impedance of bonds must be assessed.
d)
Bond reactance. Bond size, geometry, and the physical relationship between conductors being
bonded can introduce reactive components into the impedance of the bond. The minimization of
self-inductance in the path is of utmost importance at high frequency.
e)
Conductor resistance. The total resistance of the bonding conductors path is of importance when
the IR characteristics (at dc and low frequency) are such that, for a specific current level, too much
potential can be developed across the path for the connection to be effective.
f)
Overheating and fusing point. The selection of the conductor and bonds must not place the
arrangement into safety conflict with its NEC rms ampacity limits or the instantaneous fusing I2t
characteristics of the conductor or its bonding terminations.
(4.18)
IEEE
Std 1100-2005
FUNDAMENTALS
where
Lresonance is the resonant conductor length (m)
n
c
is the speed of light in free space (3 u 108 m/s)
fresonance is the frequency of excitation in the conductor (Hz)
127
IEEE
Std 1100-2005
CHAPTER 4
The actual problem involves the consideration of both the peak current through the paths inductance and
how much voltage can be developed in the distributed or lumped capacitance that is associated with the
grounding/bonding conductor. Hence, the real-world problem is represented in Equation (4.19).
L path
E max = I peak --------------------C stray path
(4.19)
where
Emax
Ipeak
Lpath
Cstray path
The use of Equation (4.19) is derived from Chapter 7 in Ott [B49] and is necessary in view of the fact that
the stray or parasitic capacitance involved with the typical grounding/bonding paths connections and
conductor may not be ignored. In short, the problem does not involve a theoretical inductance in free space
not coupled to anything. Further, the stray capacitance may be quite large, as when the grounding/bonding
connection exists across two closely spaced-apart metal cabinets or racks, each with significant surface area
showing to the other. This forms a neat air-dielectric capacitance that is integral to the understanding of the
circuit and that cannot be ignored.
From the foregoing, it should be apparent that there is no real substitute for directly abutting equipment units
and multiply bolting them together so as to obviate the need for discrete grounding/bonding conductors.
This action results in a two-fold attack on the problem:
a)
b)
As a result, even for a high peak value of current, the developed voltage must be low. Signal circuits routed
between two such cabinets are thereby not subjected to high values of common-mode disturbance.
An additional attack on the problem is provided when the designer provides for a an externally applied
grounding/bonding structure of a type that discourages the concentration of any current flow in any one or a
few of the grounding/bonding conductors making up its network. This is typically accomplished by use of
an SRS such as one comprised of a plane or grid. These are typically and respectively called a signal
reference ground plane (SRP) and signal reference grid (SRG).
When direct, unit-to-unit grounding/bonding and SRS techniques with low self-inductance grounding/
bonding conductors are properly combined, the common-mode transient voltage and current problem is
largely eliminated in practice.
4.8.3.3 Length restrictions on grounding/bonding conductors
For the reliable operation of the grounding/bonding conductor, it is recommended practice (see FIPS Pub
94) to install grounding/bonding conductors whose electrical length is a fraction of a wavelength long and
where it does not exceed (1/20)h at the highest expected frequency of interest. While this is an adequate
restriction for most commercial practice, more critical applications may require limits approaching (1/50)h
or less. The highest frequency of interest in this case is defined as the fastest transition time in the expected
waveform, and not its duration or repetition rate. For example, the important part of a 100 kHz square-wave
clocking or data signal that might appear on the grounding/bonding conductor with a 50% duty cycle is the
time it takes the impulses first transition (e.g., leading edge rise time) to go from 10% to 90% of its peak
amplitude.
128