By Andrew Fawcett, Senior Product Manager, element14

The light-emitting diode (LED) is a seemingly tireless worker and it’s solid-state construction promises a long useful lifetime. With the growing availability of high-intensity devices, LED is quickly becoming a mainstream component for area lighting. Although LED lighting promises much longer lifetimes than older incandescent and fluorescent technologies, users can be disappointed by the early failure of the luminaires they buy. Online reviews and word-of-mouth quickly convey to the market how well different brands perform in terms of reliability and there is reputation risk to the manufacturer as well as the potential for lower sales relative to competitors who are seen to be providing higher-performance products.

Why LEDs can fail

Despite their reputation for reliability as solid-state devices, LEDs can be fragile and a key reason for early failure is damage to the LEDs themselves caused by transient over voltages, which can come from many sources. As LEDs are often used in low-voltage environments, as indicators, and rarely encounter damaging surges, designers often underestimate the risk to the LEDs when using them in harsher and more demanding applications such as outdoor lighting.

Despite their reputation for reliability, LEDs have known failure mechanisms, both mechanical and thermal in nature. For example, an LED operated at full power for long periods will generate large quantities of heat that, over time, damage the wire bonds connecting the device to its package. As the metal wires oxidise in the heat, they become brittle over time, which increases the likelihood of an LED failure especially in systems prone to vibration. Other causes of LED failures are, like other semiconductors, electrostatic discharge (ESD) events or surges induced by nearby lightning events.

A further challenge when using solid-state lighting design is that an entire string of LEDs in a luminaire can fail due to a problem with just one of the LEDs in the array. LEDs are typically connected in series and driven by a constant-current source to the desired brightness, colour and intensity required by the application. Series wiring is favoured because it provides more consistent behaviour across the LEDs in the string that provides more uniform brightness. An open circuit in a single LED in a string, perhaps caused by the breaking of a single wire bond, can cause the entire string to fail. In streetlight or airport runway illumination applications, the loss of the string may dim or shut down the light, which then creates a safety hazard. LED display advertising signs are less mission critical, but the blank spots caused by string failures will lead to complaints and potential lost revenue as well as incurring the costs of more frequent maintenance visits. In home applications, brands that become known to be prone to early failures will encounter problems in the market.

Designing in protection to avoid common failure

Failure can nevertheless be reduced, if not avoided, with careful circuitry. Many systems will have some level of ESD and surge protection, but this is often focused on the power-supply inlets. Naturally, the AC input is a key area of focus for protection, but in any LED lighting application, there are three areas that require circuit protection. As well as the AC inlet, the DC side of the power supply and the LEDs themselves need protection and designers should address all the different types of protection needed at each point in the circuit.

Protecting the LED itself

The problem of a single LED failure causing an entire string to go dark can be avoided simply: by placing an open-LED protection device in parallel with each LED on the string. One type of component that can act as an open-LED protector is an electronic shunt. This acts as a current bypass for the circuit that has fallen open and allows power to flow to the remaining LEDs in the string. The shunt is a two-terminal device that automatically resets if the open-LED connection later heals itself or is replaced.

A good design for such a shunt protector is a voltage-triggered switch that has leakage on the order of microamps. Once an LED falls open, there is enough voltage in the circuit to trigger the protector to the on-state. An advantage of this type of protector is its built-in surge immunity – bypassing the LED in the event of a voltage surge that may be induced by nearby lightning strikes or ESD events. Examples of such electronic shunts can be found in the PLED family; devices in this series also protect LEDs from accidental reversed voltage.

PLEDs are relatively easy to specify. With parameters such as forward voltage and forward current and the connection scheme used in the LED strings, all that is required is to determine the size of the PLED and its protection ratio. The PLED switching current needs to be less than the value provided by the constant current source and the turn-on voltage needs to be less than the “compliance” voltage, which is the maximum, open-circuit output voltage provided by the LED string’s power supply.

The next step is to determine the number of LEDs protected by a single PLED: a designer may opt to risk the dimming that results from having three LEDs unpowered if just one in a group fails. Typically, a PLED6, which triggers at 6V, protects one LED, a PLED9, which triggers at 9 volts, is suitable for use with two LEDs and the PLED13 can be used with a group of three LEDs.

Other components which could potentially be used to provide open-LED protection often have drawbacks. For example, silicon-controlled rectifiers (SCRs) and Zener diodes both have attributes that may make them seem suitable for the job. A Zener provides effective ESD and lightning protection as well as reverse-polarity protection. However, it would not survive long in the actual application: when engaged the string current will easily overload the diode and shorten its useful life. An SCR will protect against open-LED conditions but will not guard against ESD and lightning strikes, nor will it provide reverse-polarity protection. An SCR is also typically a larger and bulkier device, which will be hard to accommodate in many high-brightness luminaires where LEDs may be tightly packed.

Protection for the AC inlet

When it comes to selecting protection for the AC inlet, a key consideration is that this is the area that is most susceptible to nearby lightning surges. Any protective device must be robust enough to withstand lightning surge requirements. A minimum of 3kA is required but it can be important to guarantee 6kA. The response also needs to be fast to limit any downstream damage. The criteria for selecting the line fuse for the AC input include the voltage and current as well as the I2T rating. The third parameter provides a gauge of the amount of energy the fuse element can withstand before it opens. As a result, time-delay fuses have higher I2T ratings than faster-acting fuses. Additionally, the I2T value increases in proportion to the current rating of the fuse.

Another key component for the AC power supply is a transient voltage suppressor (TVS) or metal-oxide varistor (MOV), using to steer over voltages away from sensitive components.

Protection for the DC inlet

Further downstream, a crucial component in the DC section is the high-voltage DC fuse, which is designed to open during overcurrent events. A secondary TVS in the DC section additionally protects against over voltages, preventing damage to the power-converter’s control electronics as well as limiting the amount of damaging charge that can make it through to the LED strings themselves.

In conclusion, although LEDs can exhibit much higher lifetimes than traditional lighting solutions, they need adequate circuit protection to ensure they can deliver on their promise. By focusing on the three key elements of a luminaire’s electrical designs, LEDs will live longer and, increasingly, will be able to move into increasingly harsh environments.