In this series George Georgevits has explained the nature of lightning and its effects on communications and power circuits. Now he reviews the characteristics of surge-protection devices.

The energy in a lightning strike is huge, the peak current is enormous and the rise time is very fast. To avoid damage to circuits from a lightning strike, we can’t just stop the stroke energy – it has to go somewhere.

Neither is it feasible to somehow absorb a stroke before any damage is done: the energy is simply too great.

The best way of mitigating damage is to coerce the stroke to attach itself to a predetermined point on the structure of interest, then conduct it safely to earth. The earth is large enough to safely dissipate the energy without causing problems (assuming correct methods are used).

On buildings and towers, appropriately placed air terminations capture the stroke, and down-conductors guide it to earth.

Overhead power lines are usually exposed to the elements, and there is generally nothing to prevent power lines from being struck. However, large high-voltage transmission lines are fitted with overhead earth wires that act as shields.

The overhead electricity network acts like a huge lightning attractor. When an aerial power line is struck, the power surge is propagated many kilometres by the low-loss power conductors.

The electricity network does employ some measures for protection against surges caused by lightning, but they are mainly at substations and not intended to protect consumer equipment.

PROTECTION OF INCOMING MAINS

Surge diverters, typically fitted at the incoming mains switchboard, act as switches that are activated only when the line voltage exceeds some specified value.

One diverter device is usually connected between each incoming phase conductor and earth. When a power surge occurs, the device shunts switches ON and the relevant phase conductor to earth for the duration of the surge.

This sounds fairly straightforward, but there are a few complications.

First, the connection to earth has to be good, and a low resistance path to earth has to be present at the switchboard for the surge diverter to function properly.

Second, all surge diverters rely on some form of non-linear resistor for their operation, usually in the form of a component known as a metal oxide varistor (MOV).

When an increasing voltage is applied across an MOV, it exhibits a high resistance until the predetermined breakdown voltage is reached, whereupon its resistance drops to a low (but still finite) value and the device is said to turn ON.

The electrical power dissipated in any device is equal to the current passing through it squared, multiplied by its resistance.

A diverter will dissipate power during a surge because of its fi nite ON resistance, even though most of the surge energy is conducted through it to ground. Thus the surge-handling capability of a diverter is directly dependent on its size, as larger MOVs can dissipate more power. Surge diverters are available in a range of sizes and operating voltages.

Another characteristic of surge diverters is that they have a finite operational life. Each time one is exposed to a large high-voltage surge, part of it is damaged. Consequently, diverters are rated for their ability to handle a certain number of surges of a specified energy. As the energy increases, the diverter must take fewer hits to remain operational.

For this reason, it is a good idea to have surge diverters checked regularly, and replaced as required. Some types have a small indicator on them to show when they need replacing.

MOVS FOR COMMUNICATIONS

MOV devices are also used on telecommunications circuits where surge protection is required.

They are much smaller than the versions used for mains protection. Again, they come in a range of breakdown voltages and sizes, and the comment on service life versus number of hits also applies.

It is worth noting that devices fitted to telecomms circuits are intended to protect against induced surges only.

Because such circuits are generally underground, the incidence of direct strike is much lower than for aerial power lines. In addition, the cable conductors are much fi ner, so their ability to conduct a large amount of surge energy over a long distance is limited.

On the down side, lightning induced surges are much more common than surges caused by direct strikes. Most lightning strikes are from cloud to cloud, and a bolt has only to pass near a telecomms cable for the accompanying magnetic fi eld to induce a surge in to it.

Electronic devices connected to telecomms lines, particularly if they have a mains earth connection as well (fax machines, answering machines, modems. etc), are susceptible to damage by induced surges.

On the rare occasion when a telecomms cable does get a direct strike, the cable is permanently damaged due to insulation breakdown. The MOV line protection fi tted at the ends of the circuits may not cope with the surge energy (depending on distance from the strike) and will be destroyed. The terminating electronics may also be damaged or destroyed.

GAS ARRESTERS

The gas arrester, commonly used for line protection in telecomms circuits, is much more robust than the MOV and has much greater energy-handling capability, for a given size.

The device consists of a glass tube filled with low-pressure gas. Electrodes inside are so spaced that when a predetermined voltage occurs across them, the gas ionises and conducts. Once the gas ionises, the voltage across the arrester drops to a relatively low value, thereby limiting the power dissipated by it.

The firing voltage of an arrester is set reasonably accurately (typically +/-20%) by careful design and manufacturing techniques.

Choosing the correct fi ring voltage is important, especially if there is a DC voltage on the line under normal operating conditions (e.g. telephone services). Common fi ring voltage values for gas arresters intended for use in telecomms circuits are 90V, 230V, 350V and 500V.

Arresters are generally available in three size ratings: 5KA/5A, 10KA/10A and 20KA/20A. The rating refers to peak surge current and continuous rms sinusoidal current-handling capabilities under specifi c test conditions.

Arresters for communications applications are commonly available as two-electrode or three-electrode types. For balanced-pair applications (e.g. telephony), it is important to use the three-electrode type.

A three-electrode arrester essentially consists of two two-electrode devices in a single envelope. It has two circuit terminals and a centre ground terminal, all fi tted to a common gas chamber.

The idea is that if either side fi res, the discharge spreads rapidly to the other side, thereby maintaining circuit balance and quenching the surge on both legs of the circuit as quickly as possible. (Keep in mind that on balanced-pair circuits, most surges appear simultaneously on both legs of the pair, because induced surges are longitudinal in nature).

Needless to say, a three-electrode arrester will work only if the third leg is connected to a low-impedance earth.

Like MOVs, gas arresters degrade with the number and severity of surges handled. It is a good idea to check them regularly for correct functionality or periodically replace them.

In communications applications employing underground cable, 5KA/5A and 10KA/10A types are generally used.

They are commonly mounted in a magazine holding 10 arresters that plugs directly into a 10-pair punchdown terminal block, as used in telephone exchanges and larger customer premises where key systems or PABXs are installed. The 20KA/20A types are generally used for protection of long aerial circuits.

TWO-STAGE PROTECTION

The main drawback of the gas arrester is that it takes a signifi cant time (in electronics terms) to operate.

Very fast transients sometimes occur on communications circuits, and the damage is done before the gas arrester fi res. For these situations a two-stage protection circuit is necessary.

Primary protection is achieved with a gas arrester, followed by some form of commonmode current-limiting device to restrict or at least delay the surge power (eg: a bifi lar wound common-mode choke for a balancedpair circuit).

Secondary protection comes from a fastacting solid-state device (e.g. MOV, TransZorb or even back-to-back Zener diodes).

CONCLUSION

In the scheme of things, implementation and maintenance of surge protection is easy and relatively low cost. Yet it is surprising how many times equipment is damaged because the appropriate devices were overlooked, incorrectly applied or never maintained.