April 1967 Electronics World
of Contents] People old and young enjoy waxing nostalgic about
and learning some of the history of early electronics. Electronics World
was published from May 1959 through December 1971. All copyrights are hereby acknowledged.
Electronics World articles.
Even though this article was written more than 40 years ago,
the fundamentals of protecting relays against interference from
either internally or externally generated noise haven't changed.
Sometimes a datasheet will recommend protection and noise suppression
techniques, and when that is that case, the manufacturer's advice
should be followed (unless you have a really good reason to
deviate, possibly voiding a warranty). When you find yourself
on your own with the design, use this article and the very comprehensive
table of application examples.
Here are links to the other relay articles:
Operate and Release Times of Relays
Finding Relay Operate and Release Times
Arc, Surge, and Noise Suppression
Arc, Surge, and Noise Suppression
By R.M. Rovnyak* / Staff Engineer, Product Design Section, Automatic
Electric Co.When relays are used. in switching
circuits, conducted and radiated r.f. interference as well as
contact erosion may occur. Here are some of the techniques that
are employed in order to minimize these harmful effects.
This article will discuss the suppression of interference
associated with a relay. The entire system contributes to the
noise problem and all factors such as grounding, shielding,
bonding, wiring and component layouts. and choice of inter-system
connection points must be adequately engineered to minimize
Switching loads may:
1. Develop voltage
and/or current transients of sufficient magnitude to damage
or destroy components within the system, rendering it inoperable.
2. Produce unwanted disturbances which can cause circuit
malfunction either within or external to the system.
3. Decrease the useful life of the system and its reliability
by, excessive wear and tear on its components.
transients are associated with breaking inductive loads such
as the coils of relays. The problem is most severe when the
inductor is rapidly switched to the "off" state. Under these
conditions the voltage can be very large and is of opposite
polarity to the supply voltage. This presents a hazard to polarity-sensitive
devices or it may initiate a high-energy discharge across a
set of contacts or the insulation between windings on the coil
Load switching may also generate coincident
parasitic disturbances or RFI (radio-frequency interference).
This broad classification spans a frequency spectrum of about
ten decades and can be classified into three types: induced,
conducted, and radiated. The bandwidth associated with these
are approximately: induced - 10 Hz to 106
Hz to 108
Hz; and radiated - 104
Hz to 1011
Disturbances by conduction
are derived from such things as dynamic regulation from the
supply or down the line within the system, loop imbalances,
recirculating currents from inductive loads, and poor connections.
They are not particularly associated with anyone type of load
but are more dependent on the magnitude of the current being
switched. Such disturbances are minimized by applying good techniques
in inter-circuit connections, component layout, wiring practices,
and the proper choice of hardware.
between circuit loops or between an inductor and a loop can
result in more than sufficient energy to cause circuit malfunction.
Careful analysis of normal circuit current paths will pinpoint
the need to either inhibit the source or minimize the pickup.
Physical isolation between source and susceptible pickup points
and the use of twisted-pair leads to reduce the area included
within the loops are the principal cures although magnetic shielding
is frequently required.
Arcing which occurs upon both
make and break of a load by a set of contacts is a source of
radiated electromagnetic interference. All load types (R, C,
L) with open-circuit voltages above about 12 volts can initiate
an arc at the contacts. This occurs in relay circuits as opposed
to semiconductor switching. Suppression of some sort is required,
the type depending on the load. The objective is to prevent
or minimize the energy in the arc.
The final area of
concern is the erosion of the contacts. We first choose a contact
material which is optimum for the load and limit the arc energy
by the application of suppression elements.
If we put a series RC network across the contacts and make R
small, the impedance under the transient condition may keep
the switch voltage (which includes applied and self-induced
voltages) small immediately after the contacts break. This arrangement
is preferred for most inductive loads, the contact voltage at
the instant of opening being limited to IL
is the load current just prior to the break. In
the case of the load being substantially removed from the switch,
however, it may be best to suppress at the load unless the sole
objective is to prevent contact erosion.
Two other factors
associated with inductive load switching are interwinding capacitance
and magnetic flux leakage. The capacitance is effectively in
shunt with the coil and will draw high, short-duration charge
currents. It is occasionally necessary to include a small inductor
or resistor ahead of the load to limit the surge. The leakage
flux will induce a voltage in a coupled loop and, if intolerable,
magnetic shielding must be used.
Capacitor in-rush currents
must be limited to reasonable values or high-energy arcs will
be sustained on contact closure. Intense RFI will be generated
unless the contacts weld first. Capacitive loads per se are
routinely taken care of in the design; however, lead and winding
capacitance, often overlooked, can play havoc with a system
susceptible to such occurrences.
With any type of suppression
scheme it is necessary to first establish the main circuit suppression
requirements - such as to protect a solid-state device, minimize
contact erosion, etc. - before deciding which technique represents
the best compromise. Table 1. Various
suppression techniques used in relay circuits along with comments
on their general characteristics.
The selection of a suppression technique
depends on the objectives to be obtained and the price one is
willing to pay. The objectives can be categorized in three main
areas: protection of components from destruction or abuse due
to the transient; reduction of the erosion rate of the contact
to increase the useful life of the contact; and reduction of
the disturbances produced by switching a load. The relative
ease with which each may be accomplished is in that order and,
as a general rule, one that satisfies the more difficult requirement
will also take care of the less difficult ones, e.g., a suppression
scheme that inhibits RFI will also provide long life and protect
The price paid for achieving
the required degree of suppression may be anyone or more of
the following: cost including component and installation; more
power consumption; increased space or weight requirements; or
longer release times, in the case of relays. The circuit designer
should approach the selection of a suppression technique by:
first, deciding the objective; second, determining the effectiveness
of the various techniques to accomplish the objective; and finally,
resolve the best technique, based on a trade-off in parameters,
including the reliability level of the suppression elements
Table 1 shows various suppression techniques
used in relay circuits, with comments on each regarding general
characteristics. It should be pointed out that, except for meeting
elementary objectives, the final determination of values and
suppression schemes is empirical. Testing Suppression
Evaluation of suppression is a two-phase operation,
observation of the transient magnitudes and degree and nature
of the arc and then testing under operating conditions for the
performance characteristic desired. A fast oscilloscope and
a probe that does not load the circuit are required to observe
the transient and arc. A small resistor, suitably located, may
be needed in order to observe the current in some circuits.
The nature of the load characteristics will determine
the approach and interpretation of the results to arrive at
the most suitable suppression scheme; i.e., the prime factor
in an inductive load is the surge voltage, in a capacitive load
it is the surge current, and in an arc it is its power level
and energy content. If the load is a relay, the effect on timing
may be important.
A common method of measuring this
effect is to trigger the scope at the instant the coil is de-energized
either from the decay slope of the energizing pulse or the induced
transient and measuring the time until contacts change state
by displaying the voltage drop across a resistive load switched
by the contacts. It is good practice to use a dual-trace preamp
and to display the trigger because false triggering can lead
to confusing results.
Switching electromechanical devices
with semiconductors presents no particular hazard if the transient
voltage, current, and time dependency (dV/dt rate in the case
of an SCR) are maintained within rated limits. Three basic things
should be kept in mind: 1. The peak voltage across the switch
is the static off-voltage plus the transient peak when the suppression
is across the load. 2. The peak turn-off current will be the
load current at time of switching. 3. The decaying current must
recirculate decaying to zero, and will seek the part of lowest
impedance which should be, by design, the suppression elements.
Testing the effectiveness of RFI suppression involves
a fairly complex approach. Relative estimates can be made for
conductive disturbances by monitoring with an oscilloscope,
for radiated interference by observations of arc characteristics,
and the coupling interference must be built into the circuit
as previously described. Beyond this, standard apparatus and
measurement schemes must be employed, and the reader is referred
to the following Military Specifications for further guidance
in this area MIL-I-26600, MIL-I-6051C, MIL-S-10379, MIL-I- 11748B,
and MIL-I-6181D.*The author is
a 1961 graduate of Indiana Institute of Technology with a B.S.E.E.
degree. His efforts have been in design and development af electro-
magnetic switching devices. He has published several papers
on related subjects and holds one potent, with another pending.