Electronics Guide

Lightning and Switching Surges

A direct or nearby lightning strike injects enormous energy into power and communication lines, producing surges of thousands of volts and currents of thousands of amperes that decay over tens of microseconds. Even without a direct strike, the electromagnetic field of a discharge induces transients in nearby conductors. Switching surges arise when inductive loads such as motors, transformers, and solenoids are interrupted, because the collapsing magnetic field drives the inductor to maintain current, generating a voltage spike governed by the rate of change of current. Power-factor-correction capacitor switching and utility operations also create transients that propagate through distribution wiring.

Standard Surge Waveforms

To make protection testable, standards bodies define representative surge waveforms specified by their rise time and duration. A common voltage waveform, designated 1.2/50 microseconds, rises to peak in 1.2 microseconds and decays to half amplitude in 50 microseconds; the companion current waveform, 8/20 microseconds, characterizes the discharge current a protective device must carry. The combination of an open-circuit voltage waveform and a short-circuit current waveform from a generator with defined source impedance, as used in surge-immunity testing, represents the energy a real transient delivers and lets engineers rate devices and verify equipment against a repeatable threat.

Other waveforms serve other threats. The 10/350 microsecond current impulse represents the long-duration current of a direct lightning strike and, at equal peak current, carries far more charge and energy than the 8/20 microsecond waveform used for induced and switching surges; only service-entrance devices are expected to survive it. Outdoor telecommunication lines are tested with a 10/700 microsecond voltage waveform, which reflects the slower transients that couple into long cable runs. Semiconductor suppressors are commonly rated against a 10/1000 microsecond exponential pulse. Because a device rated in one waveform cannot be compared directly with a device rated in another, the waveform must always be quoted alongside the number.

Temporary and Sustained Overvoltages

Not every overvoltage is a microsecond transient. A temporary overvoltage persists for cycles, seconds, or minutes and arises from a broken neutral conductor in a multiphase installation, which can place a much higher voltage across single-phase loads, or from a failed voltage regulator, generator overspeed, sudden load rejection, or accidental contact between a low-voltage line and a higher-voltage circuit. Vehicles face load dump, in which the alternator loses its battery connection while charging and the bus rises for tens to hundreds of milliseconds as the field decays. ISO 16750-2 characterizes this event as pulse 5a for an unsuppressed alternator and pulse 5b for one with internal suppression, with amplitudes reaching roughly 79 to 101 volts on a nominal 12-volt system and durations of about 40 to 400 milliseconds. The energy in such events exceeds what a clamp sized for an 8/20 microsecond surge can absorb by orders of magnitude, so the correct countermeasure is disconnection or regulation rather than suppression.

Clamping Devices

A clamping device exhibits a voltage that rises only modestly as the surge current through it increases, holding the protected line near a defined clamp voltage. As the transient drives current into the device, the voltage across it stays close to the clamping level, and the device absorbs the surge energy as heat. When the transient subsides, a clamp returns smoothly to its high-impedance state without any external action. Transient voltage suppressor diodes and metal-oxide varistors are clamping devices. Their advantage is a controlled, predictable residual voltage; their limitation is that they must dissipate the surge energy internally, which sets a ceiling on the energy they can survive.

Crowbar Devices

A crowbar device, once triggered, switches into a very low-voltage conducting state, effectively short-circuiting the line and dropping the voltage across itself to a small arc or holding voltage. By collapsing the voltage rather than clamping it, a crowbar diverts large currents while dissipating relatively little energy itself, which lets it handle far higher surge currents than a comparably sized clamp. Gas discharge tubes and thyristor surge protectors are crowbar devices. Two consequences follow from the crowbar mechanism. First, a crowbar exhibits a turn-on delay and an overshoot, because the voltage must rise to the breakover point before the device fires, so the protected circuit briefly sees a higher voltage than the holding level. Second, in a direct-current or power circuit a fired crowbar continues to conduct as long as the source can sustain the holding current, a condition called follow current, and the device does not recover until the current falls below its extinguishing value; this behavior requires that follow current be interrupted by an upstream device or limited by the circuit.

Operation and Key Voltages

Below its rated standoff voltage, the TVS diode conducts only a small leakage current and is essentially transparent to the circuit. As the voltage rises past the breakdown voltage, the diode enters avalanche conduction and its current increases steeply, holding the voltage near the breakdown level. At the rated peak pulse current the voltage reaches the maximum clamping voltage, the highest voltage the protected circuit will see. The working standoff voltage is chosen above the normal operating voltage so the diode does not conduct or draw excessive leakage in service, while the clamping voltage at full rated current must remain below the damage threshold of the protected components. The ratio of clamping voltage to standoff voltage, sometimes called the clamping factor, indicates how tightly the device controls the transient.

Capability is quoted as a peak pulse power rating tied to a stated waveform, conventionally the 10/1000 microsecond exponential pulse. Widely used surface-mount families are rated at 400, 600, 1500, and 3000 watts, while larger axial and stud-mounted parts reach several kilowatts. Because the rating depends on pulse duration, the same die survives a much higher peak current in a short 8/20 microsecond pulse than in the longer 10/1000 microsecond pulse, and datasheets publish derating curves that translate between durations and between single and repetitive pulses. Junction temperature sets the limit, so ambient temperature and the interval between pulses both reduce the usable rating.

Unidirectional and Bidirectional Devices

A unidirectional TVS diode clamps in the reverse direction and conducts like an ordinary forward diode for opposite-polarity transients, suiting direct-current rails that have a defined polarity. A bidirectional TVS diode, formed by two opposing junctions, clamps symmetrically for transients of either polarity and is required on alternating-current or signal lines that swing both positive and negative. Selecting the correct type prevents the device from forward-conducting and overloading on the polarity it was not intended to clamp.

Response Speed and Capacitance

The avalanche mechanism itself responds in well under a nanosecond, which makes the TVS diode effective against fast threats where slower devices would allow significant overshoot. The very short response times printed on datasheets describe the junction alone, however, and the voltage a circuit actually sees is set by the whole current loop: package inductance, lead length, and board layout add an inductive term that can dominate the clamp during a fast-rising transient. Treating the published figure as an achievable let-through voltage without regard to layout is a common source of disappointing test results. Their principal drawback for high-speed signal lines is junction capacitance, which loads the line and can distort fast data. Low-capacitance TVS structures place a fast steering diode in series with the clamp so that the line sees only the small capacitance of the steering diode, preserving signal integrity on high-data-rate interfaces while still routing transients to a robust clamp.

Characteristics and Ratings

An MOV is specified by its varistor voltage, conventionally the voltage at a defined small measuring current of one milliampere, and by its continuous operating voltage, which must exceed the peak of the line voltage with margin so the device does not conduct during normal operation. Its clamping voltage at a specified surge current, its peak surge current rating for a standard waveform, and its energy absorption rating in joules complete the selection criteria. For a 230-volt mains circuit, a varistor rated for 275 volts rms continuous operation is a typical choice, since the rating must clear the nominal line peak plus the tolerance the supply is permitted to drift. Disc diameter largely sets capability: a 14-millimeter disc commonly withstands a single 8/20 microsecond pulse of several kiloamperes and absorbs on the order of a hundred joules, while a 20-millimeter disc of the same voltage rating raises both figures substantially. MOVs are economical and available in ratings from small signal devices to large blocks that protect service entrances.

Degradation and End-of-Life

Unlike a TVS diode, an MOV degrades with each surge it absorbs. Repeated or high-energy surges gradually lower the varistor voltage and raise the leakage current, so an aged MOV conducts more in normal operation, heats, and can eventually enter thermal runaway. For this reason MOVs used on the power line are commonly paired with a thermal cutoff or fuse that disconnects the device before an overheated MOV becomes a fire hazard, and many surge protectors include an end-of-life indicator. The clamping voltage of an MOV also rises noticeably at high surge currents, so its clamp is looser than that of a TVS diode, which is why the two are often combined.

Operation and Strengths

Because a fired GDT collapses to a low arc voltage, it diverts surge currents of many kiloamperes while dissipating little energy, giving it the highest surge-current capability of the common protectors. Its extremely low capacitance, on the order of a picofarad, makes it nearly transparent to high-frequency signals, so GDTs are favored on antenna feeds and communication lines. Their insulation resistance in the quiescent state is very high, typically in the gigaohm range, so they impose negligible leakage on the protected circuit.

Telecommunication and signal-line tubes are commonly specified with direct-current sparkover voltages from roughly 75 volts to several hundred volts, with tolerance bands of about twenty percent, and with 8/20 microsecond surge ratings ranging from a few kiloamperes in small two-electrode parts to tens of kiloamperes in larger devices. Three-electrode tubes share a common gas chamber between two lines and ground, so that when one line fires the other is drawn into conduction almost simultaneously; this keeps the differential voltage between the two conductors small and protects balanced pairs from the longitudinal-to-differential conversion that independent single-line protectors would allow.

Limitations

The crowbar mechanism imposes two limitations. The GDT exhibits a turn-on delay and a sparkover overshoot that depends on how fast the voltage rises, so a fast transient can momentarily exceed the static breakdown voltage before the tube fires; faster transients produce higher impulse sparkover voltages. After firing on a power line, a GDT can sustain follow current from the mains, so it must be coordinated so that follow current is interrupted. These behaviors make the GDT excellent at diverting large, slower surges but less precise than a clamp, which is why protection schemes pair a GDT with downstream clamping devices that handle the residual overshoot.

Operation

Below its rated standoff voltage the device blocks and draws only leakage current. When the applied voltage reaches the breakover voltage, the internal structure switches regeneratively into conduction, and the voltage across the device collapses to an on-state voltage of a few volts, far below the level that triggered it. The device remains latched for as long as the current through it exceeds the holding current, and it returns to the blocking state once the current falls below that value. Bidirectional parts protect lines that swing both polarities, and multiline arrays place several protectors in one package with a common ground reference.

Strengths and Constraints

Because the switching is a silicon process rather than gas ionization, turn-on takes nanoseconds and the overshoot is far smaller and far more repeatable than a gas tube's sparkover, while the low on-state voltage keeps internal dissipation modest. Unlike a metal-oxide varistor, a thyristor surge protector does not degrade progressively with each surge, so it suits equipment that must survive repeated qualification surges without a change in characteristics. This combination makes it standard on telecommunication line cards and other ports exposed to both lightning surges and accidental power contact.

The latching behavior is the principal design constraint. On a circuit able to supply more than the holding current, the protector stays on after the transient has passed, so the design must either keep the available current below the holding value or provide an upstream element, such as a fuse or a positive-temperature-coefficient resistor, that interrupts the follow current. Surge-current capability sits below that of a gas discharge tube, and capacitance is higher, which limits the device on the fastest data lines unless a low-capacitance topology is used.

The Staged Protection Concept

A typical multi-stage network places a high-energy crowbar device, such as a GDT or a large MOV, at the point of entry to divert the bulk of the surge current to ground. A series impedance, a resistor on signal lines or an inductor on power lines, follows; it develops a voltage drop under the high surge current that helps the first stage fire and limits the current reaching the next stage. A fast, low-voltage clamp, such as a TVS diode, then forms the final stage close to the protected circuit, trimming the residual overshoot to a safe level. The first stage absorbs the energy, and the last stage sets the protected voltage.

Ensuring the Stages Cooperate

For staged protection to work, the high-energy stage must turn on before the low-energy clamp is overwhelmed. The decoupling impedance is essential: without it, the fast clamp would conduct the entire surge before the slower crowbar fired, and would be destroyed. The series element drops enough voltage at high current to raise the voltage at the entry point to the crowbar breakover level, transferring the burden upstream. Designers verify coordination by confirming that, across the range of surge rise times and amplitudes, the entry device fires before the let-through energy exceeds the rating of the downstream clamp.

Layout and Parasitic Inductance

The voltage a protected circuit actually sees is the clamping voltage of the device plus the inductive drop along the path the surge current takes to reach it and return. That second term is easy to underestimate. A straight wire contributes roughly a nanohenry per millimeter, and a kiloampere 8/20 microsecond surge rises at something over a hundred amperes per microsecond, so a centimeter of lead can add several volts. That is negligible beside a mains clamp of several hundred volts and ruinous beside a clamp meant to hold a logic rail near five volts.

Effective layout therefore treats the protector as part of a current loop rather than as a component with a datasheet number. The protector belongs directly at the connector, on the incoming side of the circuit it defends, with short leads and the shortest available return to the chassis or ground plane, so that surge current never shares a path with a signal return whose voltage drop would appear as interference elsewhere. A clamp placed downstream of the circuit it protects does nothing about energy that has already passed through. On multilayer boards, a dedicated protection zone near the connector, separated from the sensitive ground by the decoupling impedance, keeps the surge return current out of the quiet ground entirely.

Crowbar Overvoltage Protection Circuits

In a regulated power supply, the failure that matters most is a shorted series pass element, which applies the unregulated input directly to the load. The classical remedy is a thyristor crowbar: a reference and comparator watch the output, and when it exceeds a set threshold the circuit gates a thyristor across the rail. The thyristor short-circuits the output, collapsing the voltage and forcing an upstream fuse or circuit breaker to open. The action is deliberately destructive to the supply and deliberately protective of the load, and it is chosen where downstream damage would cost far more than a replaced fuse. Design attention goes to the trip threshold, which must sit above normal load transients, and to a short intentional delay, which prevents nuisance firing on ordinary ripple and switching noise.

Overvoltage Controllers and Series Disconnects

Modern systems more often disconnect than short. An overvoltage protection controller monitors the input rail and drives a series field-effect transistor, or a back-to-back pair where reverse blocking is also required, turning it off within microseconds when the input exceeds a programmed threshold. This series-switch approach protects a portable device from an out-of-specification charger and shields automotive electronics from load dump without absorbing the event's energy. Some controllers instead regulate the transistor in its linear region, holding the downstream rail at a safe voltage for the duration of the event so that the system keeps running; that choice moves the energy into the transistor, whose safe operating area must then cover the full duration of the disturbance. Hot-swap and electronic-fuse controllers combine the same series element with current limiting, giving overvoltage, overcurrent, and inrush control in one stage.

IEC 61000-4-5 Surge Immunity

IEC 61000-4-5 is the international standard for surge immunity of equipment against transients from switching and lightning. It specifies a combination wave generator that delivers a 1.2/50 microsecond open-circuit voltage and an 8/20 microsecond short-circuit current, with a defined effective source impedance: two ohms for line-to-line coupling, and twelve ohms for line-to-ground coupling, obtained by adding a series resistor to the generator output. A separate 10/700 microsecond generator is specified for ports connected to outdoor symmetrical communication lines. The standard's preferred test levels step the open-circuit voltage through 0.5, 1, 2, and 4 kilovolts, allowing the severity to be matched to the expected installation environment, with the applicable level for each port selected according to how it is wired and where the equipment will be installed. Tests apply the surge at specified phase angles of the alternating-current waveform, typically the zero crossings and the positive and negative peaks, and through defined coupling and decoupling networks, so that the surge reaches the equipment under test without propagating back into the laboratory supply. Equipment is assessed against performance criteria that distinguish normal operation, temporary degradation with self-recovery, and unacceptable damage or loss of function.

Related Surge and Immunity Standards

The broader IEC 61000-4 family addresses related electromagnetic disturbances, including electrostatic discharge in IEC 61000-4-2 and electrical fast transient bursts in IEC 61000-4-4, which complement surge immunity in defining a product's transient robustness. Surge protective devices are themselves classified and tested under the IEC 61643 series rather than by the immunity standards. IEC 61643-11 covers devices for low-voltage power distribution and sorts them by where they are installed and the waveform they must withstand: Type 1 devices, tested with the high-charge 10/350 microsecond impulse, go at the service entrance where a direct lightning current may arrive; Type 2 devices, tested with the 8/20 microsecond waveform, serve distribution boards; and Type 3 devices, tested with a combination wave, sit close to the equipment. IEC 61643-21 covers devices connected to telecommunication and signaling networks, where the hazards include both lightning coupled into long outdoor runs and accidental contact with power lines, and where transmission requirements such as capacitance, insertion loss, and longitudinal balance constrain the protector as much as its surge rating does. In North America, UL 1449 governs surge protective devices with its own type designations and test regime, so a product sold in both markets is usually qualified to both. Together these documents let designers select protective devices and verify equipment against the full spectrum of overvoltage threats.