Technical issues that need to be considered in the design of dimmable LED lamps

In recent years, the lighting sector has been revolutionized by the advent of LED lamps . In retail stores in Europe and the United States, the old incandescent bulbs are no longer available, or even banned. In the replacement lamp market, low-power compact fluorescent lamps (CFLs) now face new competitors. LED technology is constantly improving, and LED lamps are becoming increasingly powerful in terms of brightness and power . These characteristics, combined with their unique high energy efficiency, make LED lamps the ideal choice for lighting.

Typical incandescent bulbs produce 10 lumens per watt , while LED manufacturers claim that LED lamps can reach up to 100 lumens per watt. However, limitations in form factor and maximum operating temperature make it difficult for LEDs to reach their full potential. The best performing LED replacement lamps currently have efficiencies of 50-60 lumens per watt. Therefore, LED lamps are already on par with CFL lamps in terms of efficiency standards (note: CFL lamps have efficiencies ranging from 60-70 lumens per watt, with limited potential for improvement).

Ecological issues are also gaining increasing public attention. As people become more aware of environmental issues such as global warming and climate change, the concept of saving resources and energy is becoming a key point in many marketing strategies. The reduction in automobile fuel consumption and the ban on incandescent lamps in the past 10 years are good examples of this trend. Many governments around the world have set timetables for the elimination of incandescent lamps. There are also some public policies that specifically provide relaxed tax support or other financial assistance for energy-saving and environmentally friendly "ecological" products.

The inherent advantages of LEDs are very consistent with the public's eager expectations for safe, energy-saving and environmentally friendly lighting solutions. LEDs also have many other advantages, such as higher dimming capabilities, longer service life and smaller form factors, opening the door to convenience in terms of appearance, color, life and cost.

Replacing traditional lighting sources (fluorescent lamps, CFL energy-saving lamps, halogen lamps or other incandescent lamps) with mechanically, optically, electrically and thermally compatible LED lighting solutions is an ongoing industrial revolution. The commercial, office and residential lighting markets call for high-quality LED retrofit lamps. Although cost issues remain the main obstacle to these solutions, the following technical issues also need to be noted:

[1]. Electrical compatibility with existing infrastructure, especially when using standard in-wall socket dimmers.

[2]. The appearance must adopt a screw-on type

[3] LED heat dissipation problem must be solved

1.1 Dimmer compatibility

At present, the control equipment for lighting fixtures in homes, hotels and offices is designed and installed for incandescent lamps. The best way to provide dimming function is to use a "mains phase-cut" dimmer. This dimmer was originally developed to power incandescent lamps. From an electrical point of view, an incandescent lamp can be regarded as a resistive load.

The electrical equivalent load of an electronic lighting source, such as a CFL or LED lamp , is no longer a purely resistive load. This makes a big difference in the way the dimmer works. Improper use of the load can lead to malfunctioning of the system, which can cause very uncomfortable flickering or even damage to the lamp or dimmer. This can cause user dissatisfaction and delay the adoption of LED solutions. Currently, dimmers are still expensive and difficult to install. Therefore, launching a high-quality solution that is compatible with existing dimmers is a must to deliver on the promise of LED lighting promotion.

Phase-cut dimmers come in many different types, but all work on the same basic principle (i.e., cutting off a portion of the power sine wave during each cycle). This is done with a switch . When the switch is on, power is delivered to the load (the light bulb). When the switch is off, no power is delivered. By adjusting the time the switch is on, the total amount of energy delivered to the load can be adjusted.

There are two types of dimmers: leading edge phase cut and trailing edge phase cut.

[1]. A leading edge phase dimmer switches the light at the beginning of the half cycle. After a period of time corresponding to the dimming position, the switch is turned on to supply power to the load until the end of the half cycle. After the zero point, the same operation is repeated.

[2] A trailing edge phase dimmer is one where the switch is turned on at the beginning of a half cycle, turned off after a period of time corresponding to the dimming position, and remains off until the end of the half cycle. After the zero point, the same operation is repeated.

To accomplish this, two main technologies are used: TRIAC (Bidirectional Thyristor ) switches or transistor switches. TRIAC dimmers are mostly leading-edge phase-cut dimmers. Transistor switches can be either leading-edge or trailing-edge phase-cut dimmers. The problem with TRIAC dimmers is that they require special conditions to work properly: the TRIAC can be turned on by triggering the gate, but once triggered, a minimum current is still required to keep it in the on state. This trigger current is called the "latch current" and it must be loaded for a period of time in order to keep the TRIAC on. Once the device is latched, a continuous current must be supplied. This current is called the "holding current". If this current is disconnected or weakened, the TRIAC will turn off. In order to be compatible with dimmers, LED lamps must absorb the holding current required by the dimmer. For example, if a 6W LED lamp (roughly equivalent to a 40W incandescent bulb) is to be used with a dimmer with a minimum load of 10W, some additional circuitry is needed to provide enough holding current. In this case, the bulb efficiency will be reduced, but it is still highly efficient compared to 40W. Without this circuit, the LED lamp will not work properly. The holding current of different dimmers is different. Therefore, the greater the additional loss, the better the compatibility with the dimmer. The design difficulty of dimmable applications lies in finding the best trade-off between lamp efficiency and dimmer compatibility.

NXP 's SSL2101 (and its derivative, the SSL2102) have two integrated bleeder switches controlled by an IC . Two different bleeder currents can be set by external resistors. The IC provides greater flexibility in current selection and optimizes dimmer compatibility. In the SSL2103 (a controller -only version of the SSL2101), the integrated bleeder switches are removed but can be replaced with low-cost external bipolar switches (still driven by the IC ) that can optionally provide higher bleeder currents. This version also removes the internal MOSFETs for converter operation, but can be modified externally to create a solution that meets specific regulation requirements.

1.2 Relationship between size and topology

LED lamps are essentially current driven. Their luminous intensity is roughly proportional to the current passing through them. Therefore, they can be driven from the mains using different topologies. One of the key goals of topology selection is to determine the form factor of the final application. Low-power LED lamps less than 15W are mainly aimed at retrofitting existing lamps. This means that the shape of the bulb must be similar to existing traditional lamps. There are three main topologies currently available on the market:

1.2.1 LED and resistor in series

The LED is connected in series with a resistor, which is directly connected to the mains. This topology is the simplest. As long as the supply voltage is higher than the sum of the forward voltages of the diodes, the LED is turned on. The maximum current of the LED is determined by the resistor value:

Where n is the number of LEDs and Vf is their forward voltage.

This solution is simple but not very efficient. For example, a 12W application may have a maximum current of 500 mA for the LED. The total forward voltage will be 24V. For a 220V AC mains, the required resistor value is 574Ω. This resistor will dissipate 143W at 500mA. Of course, we can consider using a lower current diode, but the Vf required to achieve the same power is higher. The direct consequence of this is that the turn-on time is greatly shortened and flickering will occur.

1.2.2 Buck topology

[2] Buck topology is one of the most efficient topologies. Its basic principle is shown in Figure 5.

Figure 5 Buck topology principle

When the switch is turned on, current flows through the LED coil and emits light. To control the current value, a sensor resistor is connected in series with the ground. By detecting the voltage across this resistor, when it reaches the over-current protection (OCP) value, the switch is turned off. The energy stored in the coil is then released through the freewheeling diode and the LED.

This topology has two major advantages:

[1]. First, the efficiency is very high, especially for low power applications (less than 10W). LED general lighting using this topology is usually claimed to have an efficiency of more than 90%.

[2]. The second advantage is the size. Overall size is very important for the lamp retrofit market, as the final product must be similar in appearance to traditional incandescent or halogen lamp products. The buck topology does not use a transformer or optocoupler, so the coil is relatively small, especially when the switching frequency is relatively high.

This topology also has two main disadvantages

[1]. The main disadvantage of this topology is that it does not provide any electrical isolation. For heat dissipation reasons, LEDs are usually mounted on metal heat sinks. Thus, electrical isolation is mandatory for safety reasons.

[2]. The second disadvantage is that the LED is connected in series with the coil. This requires a trade-off between the total forward voltage of the diode and the maximum losses of the converter. If the voltage difference between input and output is too large, efficiency will be reduced.

1.2.3 Flyback topology

In the flyback topology, the coil is replaced by a transformer and the LED is connected to the secondary side of the transformer, as shown in Figure 6.

Figure 6. Flyback topology schematic

When the switch is turned on, current flows through the transformer, and the diode on the secondary side of the transformer is in a blocking state. When the switch is turned off, the secondary side diode begins to conduct, current flows through the diode and emits light.

The main advantage of the flyback topology is that the LEDs are connected to the secondary side of the transformer. Since the winding turns ratio can be selected, there is no need to trade off between the number of LEDs and efficiency. This topology allows for electrical isolation, which improves the safety of the device and makes the generation of VCC simpler. It is also possible to add an auxiliary winding to the transformer to provide power to the controller.

The main disadvantage of this topology for the retrofit market is the large physical size of the application. The transformer (and possibly an optocoupler if feedback is required) takes up a lot of space. Some controllers, such as the SSL210x series, are compatible with both buck and flyback topologies. Depending on the requirements, a very compact buck application can be used when safety and good thermal performance are required, or a flyback topology can be used if very high LED power or electrical isolation is required. To achieve a smaller form factor, the SSL2101 and SSL2102 both integrate the power switch.

1.3 Dimming

LEDs are easy to dim. Their luminous intensity is roughly proportional to the supply current. As long as the average current is reduced, the light energy emitted will be reduced. However, designing a dimming topology is more complicated and requires the use of different techniques such as PWM dimming, frequency modulation or I peak modulation.

[1]. When using PWM dimming, the instantaneous current delivered to the LED has only two values: 0 or Imax

– When PWM=0, ILED= 0 A

– When PWM=1, ILED = Imax

By adjusting the duty cycle of the PWM signal, the average LED current can be varied. This technique requires the converter to act like a fast-switching current source. The goal of this technique is to accommodate large currents and maintain precise control. Switching losses can be high, especially for deep dimming.

[2]. When using frequency modulation, the maximum LED current does not change, but the frequency of the converter changes. In other words, the energy delivered to the LED in each cycle of the converter is the same, but the number of cycles per second varies. The problem with frequency modulation is that when the minimum frequency becomes very low, it will generate noise, and when the maximum frequency becomes very high, it will generate huge switching losses. These factors make deep dimming difficult.

[3]. When using I peak modulation, the controller frequency is fixed, but the on-time of the power switch varies. Therefore, the maximum current I peak of the LED varies. In other words, the number of cycles per second remains constant, but the energy delivered to the LED per cycle decreases. When using this technology for deep dimming, the pulses must be very small. Due to the large switching losses, this method is generally not recommended.

Some controllers, such as the SSL2101 and its derivatives (SSL2102/03), can combine frequency modulation and I peak modulation. This minimizes the disadvantages of each method. At full power, the switching frequency can be optimized to limit switching losses. In dimming mode, reducing I peak and frequency simultaneously can avoid too short an on time and too low a frequency. As a result, dimming levels below 1% can be achieved.

1.4 Lifespan

The main advantage of LEDs is their long lifespan, which is claimed to be up to 50,000 hours. This value is much higher than the lifespan of energy-saving lamps and far exceeds the lifespan of traditional incandescent lamps. However, LEDs cannot be used alone (as mentioned above) and need to be driven by other devices. In the semiconductor industry, the standard for lifespan testing is 1000 hours. Depending on the operating temperature of the final application, some factors are also used to estimate the device lifespan. This 1000-hour test is not enough to ensure the estimated lifespan, and accelerated life tests must be performed to ensure that the lifespan of the controller is comparable to the lifespan of the LED.

The SSL210x series has successfully passed the 8,000-hour lifetime test at a junction temperature of 150°C. Depending on the actual junction temperature in the final application, the estimated lifetime is 45,000 hours at 115°C and 75,000 hours at 105°C. These values ​​are sufficient to support the claim that the SSL210x series can last as long as the LED.

1.5 Conclusion

LED lighting fixtures are still a minority in the general lighting market in terms of quantity. However, as the market heats up, LED solutions will inevitably dominate the market in the next few years. This article discusses several aspects of dimmable LED lighting applications. Buck and flyback topologies are two commonly used topologies. Buck topology provides a very compact and efficient solution for low-power applications, while flyback topology is still the best choice for systems that require safety and flexibility due to its internal galvanic isolation. Dimmer compatibility and dimming range are two challenging characteristics. Weighing various options around related issues will be the key to product success. Currently, NXP's SSL210X series covers almost all possibilities of retrofit market dimmable applications, so each design can be optimized to obtain the best solution.