Comparison and analysis of LED thyristor dimming scheme based on MCU and ASIC

Author: Secom Telecom Bell Rock

As a new and most promising light source, LED lighting has attracted more and more attention because of its energy saving and environmental protection advantages. Coupled with the policy encouragement of the state and local governments, China's LED lighting industry has entered an accelerated development stage, using the market to grow rapidly. In terms of indoor lighting , replacing traditional dimmable incandescent lamps or halogen lamps with LED lamps will also be the trend.

Since the traditional incandescent lamp dimmer uses a thyristor dimmer, when replacing the incandescent lamp with an LED lamp, it is required to be able to change the original line and to adapt to the existing thyristor dimmer. In response to this target market, many large semiconductor manufacturers (including internationally renowned semiconductor manufacturers) have already launched their own LED dimming ASICs. However, due to the inherent light-emitting principle of LEDs, the current LEDASIC dimming cases on the market are still not very Mature, all have their own problems, this article will do a detailed analysis of the current dimming scheme, and introduce our MCU-based dimming scheme.

LED light-emitting characteristics

The current new technology enables LEDs to achieve high power levels, LED power can reach 1W, and even some reach 5W, luminous efficiency reaches 60-85LM/W. This LED device is called high-brightness LED (HB-LED). At present, the LEDs used in lighting are all HB-LEDs. Generally, 1W LEDs are selected to form high-power LED lamps through series-parallel connection, especially in series. This LED is 3.4V ± 2%, and the forward current is about 350mA. Its - curve is shown in Figure 1.1.

Figure 1.1 LED - Curve

As can be seen from Figure 1, before the voltage applied across the LED does not reach 3.4V, it increases with increasing. When the voltage applied across the LED reaches 3.4V, the change is small. Increasing the voltage across the LED will only increase the current flowing through the LED, thereby changing the brightness of the LED until it reaches the maximum (350MA) of the LED, and the LED reaches its maximum. brightness. It has been clamped at around (3.4V). And it will change with the temperature and LED working time, the curve is shown in Figure 1.2.

Figure 1.2 Curve of relative change versus temperature

As can be seen from Figure 1.2, as the temperature rises, it will gradually become smaller. On the contrary, when the temperature decreases, it will increase. When the temperature of the LED rises to 85 degrees, it has become 3.25V (3.4V-0.15V), and when the LED temperature drops to -40 degrees, it becomes 3.6V (3.4V+0.2V).

Therefore, in the LED dimming, in order to adjust the LED to a fixed brightness, it must be guaranteed to be a fixed, that is, to use constant current control. There is another way: constant power control. This is also the common method used in dimming ASICs on the market today. However, constant power control has its inherent drawbacks.

Constant power control LED dimming mode 1 theory According to the constant power control mode, a single-stage FLYBACK topology is generally adopted, and the basic block diagram is shown in Figure 2.1.

Figure 2.1 Block diagram of constant power control

It can be seen from the figure that compared with the FLYBACK of the general switching power supply, the constant power scheme has a dimming signal control loop for detecting the conduction angle and input voltage of the input thyristor, thereby giving the corresponding dimming signal. . At the same time, there is no feedback signal in the scheme, and the open loop control is completely controlled, and the size of the duty ratio is controlled by the primary side to control the magnitude of the output power. The theoretical basis for its control is:

In the formula:

P is the power output to the LED;
η is the efficiency of the converter, mainly determined by the transformer;
IP is the average current of the primary transformer;
LP is the primary inductance of the transformer;
f is the switching frequency of FLYBACK.

The constant power control method first predicts the output power of the LED. In theory, once the transformer is designed, it has been determined. As long as the primary current is changed, the output power can be changed. Because of:

In the formula:

U is the voltage applied to the primary side of the transformer, that is, the input voltage; D is the duty ratio; it is the switching frequency; it is the primary inductance of the transformer. According to Equation 2.2, it is proportional to D. As long as the duty ratio D is changed, the magnitude of the primary current can be changed, and the output power can be changed.

2 Advantages and Disadvantages In theory, this topology is simple and low cost, but careful analysis will reveal that there are many drawbacks in this way. The first is difficult to control, often the deviation is very large, coupled with the use of open-loop control, the accuracy is difficult to guarantee, in the batch, using the same duty cycle will lead to a large deviation of the output power, directly reflected in the LED is the LED The brightness varies greatly and it is difficult to guarantee consistency.

Secondly, the constant power control presupposes that the output power is constant. For example, if 9 LEDs of 1W (=3.4V, =350MA) are connected in series to be used as a 9W PAR lamp, then the design will use =9W to calculate. . But in fact, each LED will be biased, the deviation value is shown in Table 2.1:

As can be seen from Table 2.1, although the deviation of each LED is not too large, when the LEDs are connected in series, the total deviation cannot be ignored. The actual power of the PAR lamp may be more than 9W or less than 9W. Directly reacting on the lamp means that the maximum brightness is not up to the rated value or brighter than the rated value. When the actual power of the PAR lamp is less than 9W, and the output power of the driver is still 9W, the current flowing through the LED has exceeded the rated value. The long-term operation will have certain influence on the life and color rendering of the LED.

In addition, even if there is no deviation when the LED is shipped from the factory, as shown in Figure 1.2, it will change with the change of temperature and working time. When the rated power of the LED is less than 9W due to the change, long-term work will affect the service life of the LED. And color rendering.

Again, power compatibility issues. Since the constant power control mode is preset output power, when the connected lamp power does not match the preset power, the LED lamp cannot work normally. When the power of the connected lamp is less than the preset power, the LED lamp may even be burned.

There is also the problem of no-load loss. Since the output power of the constant power control scheme is only related to the input voltage and the conduction angle, when there is no LED connection, there will still be corresponding power output, which makes the no-load loss become Great.

Finally, the dimmer compatibility issue. Since LEDs are used to replace previous incandescent lamps, LED dimming must also be tuned with thyristors. The incandescent lamp is a purely resistive load, which does not affect the conduction of the thyristor. However, the driving circuit of the LED is composed of a switching power supply, which is not a pure load, and thus has a certain influence on the thyristor. The chopper waveform of the thyristor is easily distorted, and it is particularly difficult to maintain a good conduction state. Therefore, it is generally necessary to add a thyristor stabilization circuit to the driver. Due to its pure hardware implementation, the constant power control method has limited adaptability to the dimmer even if a thyristor stabilization circuit is added.

Since constant power control has the advantages of simple structure and low cost, and its shortcomings may not be reflected in a short time, many semiconductor manufacturers adopt this method. However, if the MCU is used to control the constant current, the above problems can be solved very well.

3. Scheme based on MCU to control constant current

The overall block diagram of the MCU-controlled constant current scheme proposed by Shiqiang Telecom is shown in Figure 3.1. The scheme consists of two stages. The first stage is FLYBACK, which uses closed-loop control to output a stable voltage to the second stage. The second stage uses MCU control to form a BUCK constant current control circuit. The functions to be completed by the MCU include: 1. According to the size of the conduction angle of the thyristor, the corresponding dimming signal is given. 2: Determine whether the dimming signal needs to be changed according to the distortion of the waveform after the thyristor chopping. 3: By detecting the current feedback of the LED, the corresponding duty ratio is given, and the current flowing through the LED is stabilized to achieve constant current control.

3.1 Theoretical basis

The topology of the constant current control circuit is BUCK circuit. When BUCK is used for constant voltage output, the relationship between input voltage and output voltage is:

In the formula:

V0 is the output voltage;

D is the duty ratio;

Vin is the input voltage.

As can be seen from Equation 3.1, the output voltage is proportional to both the duty cycle and the input voltage. In the closed-loop control, the output voltage feedback signal VFB is compared with the preset reference voltage Vref. When the input voltage is constant, if VFB is greater than Vref, the duty ratio D is decreased, and when VFB is less than Vref, D is increased to The effect of voltage regulation is achieved.

Therefore, when the BUCK circuit is used for constant current control, the constant current effect can be achieved by changing the voltage feedback signal to the current feedback signal.

3.2 Advantages and disadvantages

Since the MCU control constant current scheme adopts constant current control, it can solve the problems occurring in the constant power control scheme well;

First, the problem of the influence of the deviation of the VF on the service life and color rendering of the LED. With the constant current control scheme, when the VF is deviated or changes with temperature and working time, as long as D does not change, the current output to the LED does not change too much. Although the forward current IF of the LED changes with temperature and working time, the IF deviation of each LED is not too large, for a high-power LED lamp composed of a plurality of LEDs connected in series, and It will not have any effect on the service life and color rendering of the LED.

Secondly, due to the closed-loop control, a high output current accuracy can be achieved. Can ensure the consistency of the lamp.

Again, power compatibility issues. The MCU constant current control scheme is compatible with LED lamps below the rated power. For series LEDs, changing the output power only changes the number of LEDs n and the voltage VLED output to the LEDs, and does not change the magnitude of the output current. When the output power is lower than the rated power, the clamp voltage n*VF applied to the LED lamp becomes smaller, and the voltage VLED output to the LED is still rated at this time, so the current flowing through the LED becomes larger, so that the current feedback When the signal VIFB is greater than the reference voltage Vref, the duty ratio becomes smaller, the output voltage VLED becomes smaller, and the current flowing through the LED also becomes smaller until the rated current value. Therefore, it can be compatible with LED lamps below the rated power.

For the no-load loss problem, the MCU can be used to detect whether there is load access. When no LED access is detected, the BUCK circuit of the latter stage can be turned off to reduce the no-load loss.

Finally, the thyristor compatibility issue. In the constant power control scheme, the pure hardware mode is used to stabilize the conduction of the thyristor, and the adaptability to the dimmer is still limited. In the MCU control constant current scheme, in addition to the hardware stabilization circuit, there is software to discriminate the waveform after the thyristor chopping. It can be determined by software setting whether the waveform after chopping is distorted or really dimmed, so as to decide whether to change the corresponding dimming signal. Therefore, the MCU control constant current scheme improves the adaptability to the dimmer.

Of course, since the MCU constant current control adopts a two-stage closed-loop control scheme, compared with the constant power single-stage control scheme of the ASIC, there are also some shortcomings. First, the structure is more complicated than the constant power control scheme, and the cost is slightly higher. ~2RMB, in addition, the efficiency is lower than the constant power scheme, but it can fully meet the requirements of "Energy Star".

4. Summary

Through the above analysis and comparison, it is easy to see that MCU control constant current scheme is better than ASIC in terms of consistency, power compatibility, dimmer compatibility and impact on LED service life and color rendering. Controlling the constant power scheme has an absolute advantage.