White LED driver analysis and application

White light LED lighting has been widely recognized by everyone for its high efficiency, low power consumption, energy saving and environmental protection. In essence, LED is a light-emitting diode. Its luminous intensity is proportional to the forward current passing through it, and there is a conduction voltage. When the current is 20 mA, the forward voltage drop is generally 3 to 3.5 V. In many cases, the luminous intensity of a single LED cannot meet the needs of practical applications, and multiple LEDs must be used in series or in parallel, which requires a large voltage or current to drive. Different manufacturing processes and even different batches of LEDs have performance mismatches, which also brings difficulties to the reasonable design of the driver. Therefore, although there are many types of original power supplies, none of them can directly power LEDs. This requires driving by boosting or bucking, as well as constant current or constant voltage according to different needs.

1 Working principle of common LED drivers

1.1 Linear Regulator Driver

The earliest complete set of linear voltage regulator drivers appeared in the 1970s, when NPN tubes were used as voltage regulators, as shown in Figure 1. This voltage regulator requires a voltage of 2Vbe between the input voltage and the output voltage. When the input voltage is lower than 2Vbe, the NPN tube enters saturation and the voltage regulator loses its voltage regulation ability. In order to reduce the voltage difference, a combined voltage regulator appeared, as shown in Figure 2, that is, the base of the NPN tube is driven by a PNP tube, but the voltage difference is also close to 1Vbe. In the mid-1980s, low-voltage dropout linear regulators appeared on the market, as shown in Figure 3. Unlike NPN regulators, the voltage difference of PNP regulators is not a function of Vbe, but a function of PNP tube Vce, which is much lower. With the maturity of manufacturing technology, the voltage difference of PNP regulators has been less than 500 mV.

Now the linear voltage regulator driver has developed into the structure shown in Figure 4.

A linear voltage regulator driver refers to a transistor or field effect tube that works in the linear region or saturation region and removes excess voltage from the input voltage to generate an adjustable, stable and accurate DC voltage. It is usually composed of a voltage regulator, an error amplifier, a feedback circuit and a reference voltage. The voltage regulator is usually a MOS tube, which is equivalent to a voltage-controlled resistor, and the size of the resistor is controlled by the gate voltage. The output voltage Vout is Vout=Vin-Vp obtained by dividing the voltage regulator and the load. If the input voltage Vin or the load changes, the control terminal voltage Vc also changes, and the MOS tube resistance is controlled to achieve the purpose of adjusting the size of the MOS tube voltage division Vp, so that Vout is guaranteed to be stable. The linear voltage regulator driver can also be changed into a linear current regulator driver by connecting the sampling resistor in series with the load, the feedback voltage Vo=Iout×R1, and keeping the size of R1 unchanged. The feedback voltage can reflect the change in the output current size, and its specific working principle is basically the same as that of the linear voltage regulator driver.

The efficiency of linear regulators is relatively low. From the principle, we know that the output voltage of the driver is obtained by subtracting the MOS tube voltage divider Vp from the input voltage, and this part of the voltage is completely converted into heat energy and consumed. Therefore, in order to improve the efficiency of the driver, it is generally required that Vp is as low as possible. Linear regulator drivers with low input/output voltage difference are called low voltage dropout linear regulators, or LDO for short.

1.2 Charge Pump Driver

The earliest ideal charge pump model was proposed by Dickson J in 1976, as shown in Figure 5. Its basic idea is to generate high voltage through the accumulation effect of capacitors on charges. Later, Witte-rs J, Toru Tranzawa and others improved Dickson J's charge pump model, proposed a more accurate theoretical model, and verified it through experiments.

Modern charge pumps are mainly composed of switch arrays, oscillation circuits, logic circuits and comparators to achieve DC-DC conversion. The driving mode has also changed from the previous single mode to an adaptive multi-mode. The main forms are single mode (such as 2X mode), dual mode (such as 1X/2X mode) and multi-mode (such as 1X/1.5X/2X mode). The working principle of the charge pump is analyzed in combination with the dual-mode 1X/2X charge pump.

As shown in Figure 6, when the charge pump works in 1X mode, the oscillator does not work, S1 and S4 are directly turned on, at this time, Vin = Vout; when the charge pump works in 2X mode, the oscillator outputs a square wave with a duty cycle of 50%, so that S1, S3 and S2, S4 are turned on in turn. When the clock signal is high, S1 and S3 are turned on, S2 and S4 are turned off, Vin is connected to C1, and C1 is charged to make Vc = Vin; when the clock signal is low, S1 and S3 are disconnected, S2 and S4 are turned on, and Vin is connected in series through C1 to supply power to the outside, so in steady state, Vout = Vin + Vc = 2Vin.

The charge pump drive circuit can not only effectively increase or decrease the voltage output, but also easily produce negative voltage output, which is a major advantage of the charge pump driver over the other two drivers.

As shown in Figure 7, its basic principle is the same as that of the Dickson charge pump, but it uses the characteristic that the voltage difference across the capacitor will not jump. When the circuit remains in the charging and discharging state, the voltage difference across the capacitor remains constant. In this case, the original high potential end is grounded, so that a negative voltage output can be obtained.

One of the most important indicators of the charge pump drive circuit is the conversion efficiency. The conversion efficiency of the charge pump is:

In the formula: Pin is the total input power; Lout is the total current flowing through the load LED; VLED is the forward conduction voltage drop of the LED; M is the boost multiple of the charge pump; Iq is the drive current of the charge pump power tube and the static current of other modules. It can be seen from the above formula that the larger the boost multiple M of the charge pump, the lower the conversion efficiency of the charge pump. Therefore, under the condition of satisfying the LED drive voltage, that is, Vout> VLED, the charge pump should be operated in a low boost multiple mode as much as possible.

1.3 Inductive Switching Regulator Driver

The inductive switching voltage regulator driver is referred to as the switching power supply, which is named because the device in the power supply that adjusts the voltage regulation control function always works in a switching mode. The frequency of the early switching power supply was only a few thousand hertz. When the frequency reached about 10 kHz, the magnetic components such as transformers and inductors emitted very harsh noises. It was not until the 1970s that the switching frequency broke through the human hearing limit of 20 kHz, and the noise problem was solved. With the continuous increase in switching frequency, the size of the driver has been reduced and the efficiency has been improved. In the 1980s, zero voltage and zero current switching circuits using quasi-resonance technology, that is, soft switching technology, appeared. This circuit makes the voltage and current zero before the switch is turned on or off, respectively, which solves the switching loss and switching noise problems in the circuit, and allows the switching frequency to be greatly increased, so that the switching power supply can further develop in the direction of small size, light weight, high efficiency and high power density.

The core of the inductive switching voltage regulator driver is the electronic switching circuit. According to the output voltage or current regulation characteristics of the power supply required by the load, the switching circuit is controlled by using the feedback control circuit and the duty cycle control method. When the switch tube is closed, the energy of the power supply is stored in the inductor. When the switch tube is turned off, the energy in the inductor flows into the capacitor, thus realizing energy transmission.

There are usually two control modes for inductive switching voltage regulator drivers: one is to keep the switch duty cycle unchanged and control the switch conduction time pulse width modulation (PWM). In this mode, when the input voltage or load changes, the control circuit performs closed-loop feedback through the difference between the output voltage or current and the reference voltage, and adjusts the conduction pulse width of the main circuit switch device, so that the output voltage or current of the inductive switching voltage regulator driver remains stable; the other is to keep the conduction time unchanged and change the switch duty cycle pulse frequency modulation (PFM). The basic working principle is that when the input voltage or load changes, the control circuit performs closed-loop feedback through the difference between the output voltage and the reference voltage, and controls the length of the switch cycle, that is, controls the switch frequency, to adjust the switch duty cycle, so as to achieve the purpose of stabilizing the output voltage or current. Since the PWM mode circuit is simple and the input/output range is wider than the PFM mode (PFM is usually used in light load, low voltage, and low current conditions), it has been widely used. The following mainly introduces two PWM drive modes.

1.3.1 Voltage Controlled PWM

The structural diagram is shown in Figure 8.

In the PWM controller, the output voltage Vo is detected and added to the inverting input of the op amp, and the fixed reference voltage Vref is added to the non-inverting input of the op amp. After the error is amplified, the DC error voltage Ve is output and added to the non-inverting input of the PWM comparator; the sawtooth wave signal Vosc generated by the ramp signal generator is added to the inverting input of the PWM comparator. After PWM comparison of Vc and Vosc, a square wave signal is output, and the duty cycle of the square wave signal changes with the error voltage Vc. When the output voltage decreases, the Ve value becomes larger. After PWM comparison, the duty cycle of the output square wave decreases, the conduction time of the MOS tube increases, the charging time of Vin to the inductor increases, and Vout increases.

1.3.2 Current Control PWM Principle

The structural diagram is shown in Figure 9. The difference between this circuit and the voltage control type is that this circuit has two parts: the outer control loop and the inner control loop. When the output current Iout decreases, the error amplifier output increases, and the PWM output is 0; when the rising edge of the oscillation wave arrives, the MOS tube is turned on, Vin charges the inductor, the current increases, and the feedback voltage increases through the sampling resistor R3. When the feedback voltage exceeds Ve, the PWM output is 1. When the falling edge of the oscillator arrives, the MOS tube is turned off, and the current on the inductor is output to the outside. The current control mode has the same inversely proportional relationship between the duty cycle and the output voltage as the voltage control mode, and also has the following characteristics: the outer control loop controls the minimum current; the inner loop controls the maximum current.

2 Comparison of the advantages and disadvantages of various drivers

As for LED driving methods, each LED driver has its scope of application, as well as their own advantages and disadvantages. Understanding their respective advantages and disadvantages can help us better design a reasonable LED driving circuit based on actual conditions. This can be done through comparative analysis of efficiency, operating voltage, noise interference, output regulation, response speed, installation size and cost.

2.1 Overall efficiency

The overall efficiency of the linear voltage regulator driver is relatively low, mainly because the linear voltage regulator driver relies on the power tube to divide the excess voltage to achieve the voltage regulation effect, and this part of the power consumption is completely useless, resulting in a decrease in the driver efficiency. Therefore, when using a linear voltage regulator driver, the input and output voltage difference should be minimized as much as possible, and its actual conversion efficiency is usually between 50% and 95%; since the basic charge pump can only provide output voltage in multiples, its output voltage cannot be stabilized at a certain value, so an additional LDO is usually connected to the outside of the charge pump circuit to convert it into a voltage-regulated charge pump, which results in the efficiency of the charge pump driver being based on its own power consumption, and the additional power consumption of the LDO driver, and the efficiency is usually between 70% and 85%; the loss of the inductive switch driver and the basic charge pump driver mainly comes from the static current loss of the internal MOS device, the power loss of the external capacitor and the sampling resistor, and its efficiency can reach 80% to 90%.

2.2 Operating Voltage

Due to the voltage division working principle, the linear voltage regulator driver can only step down the output, which means that it can only work when the input voltage is higher than the voltage required for the LED drive; the charge pump driver can step down or step up, but if the output voltage needs to be adjusted by a high multiple or in multiple modes, a large number of switches and capacitors need to be connected, which greatly reduces the efficiency, so it is generally used for over-voltage drive, that is, when the input voltage is not much different from the output LED drive voltage; the inductive switch driver utilizes magnetic field energy storage, so no matter it is stepping up, stepping down, or both at the same time, it only needs to adjust the ratio of the sampling resistor to adjust the output voltage over a wide range, and the driver efficiency will not be changed due to the output adjustment, so it has the widest application range and can be widely used under various input voltages.

2.3 Noise Interference (EMI)

The working principle of the linear voltage regulator driver is to stabilize the output voltage by voltage division. It does not require capacitors or inductors for voltage regulation. The MOS tube always works in a linear state and does not need to be turned off or on, so it does not generate noise voltage, current and electromagnetic interference. The charge pump does not use inductors, so its EMI effect can be basically ignored. In the process of output voltage, the MOS tube needs to perform switching operations, so a certain amount of power supply noise will be generated, but since no inductor is used, the noise is small and can be eliminated by connecting a very small capacitor. The inductive switch driver is the main source of power supply noise and EMI. Due to the frequent switching operations of the MOS tube, PWM will generate large EMI interference within the fixed frequency of the MOS tube switch, and PFM will generate interference within the variable range of its frequency. Therefore, suppliers usually need to take measures to increase the operating frequency of the inductive switch so that its EMI falls outside the system frequency band. In addition, due to the inductance, the MOS tube will generate a large peak current or voltage at the moment of switching on and off, and there is also a phase difference between the output current and voltage.

2.4 Output Regulation and Response Speed

The linear voltage regulator driver can control the output voltage by adjusting the ratio of the external sampling resistor and the voltage division of the MOS tube according to the needs of the product, and its circuit is simple and the response speed is fast; the charge pump driver itself cannot adjust the output voltage value at will, and can only adjust the multiple through the sampling voltage feedback. The voltage can also be adjusted by connecting the LDO in series, but this will also reduce the response speed of the driver, and the response speed is slower than that of the LDO. The advantage of the charge pump in regulating the output voltage is that it can output both positive and negative voltages through a simple design. The other two driving methods require additional circuit design to achieve the effect of outputting negative voltage; the inductive switch driver only needs to adjust the output voltage by adjusting the ratio of the sampling resistor and changing the control square wave duty cycle to adjust the output voltage. The adjustment is relatively simple, but due to the complex circuit structure, multiple comparisons and amplifications are required. The presence of inductance also further slows down the reaction speed of the voltage adjustment, so the response speed is the slowest.

2.5 Installation size and cost

The linear voltage regulator driver circuit is simple in structure, generally only 20 to 40 components are needed, and the cost is low. However, since the MOS tube is always in the linear or saturated region, the heat generated is large, so a large heat sink needs to be installed to ensure good heat dissipation and ensure system stability. The complexity of the charge pump circuit is moderate, and the volume of the external capacitor can also be reduced by increasing the switching frequency. In addition, the chip capacitor has also been well applied, which greatly improves the integration of the charge pump and further reduces the required installation size. The inductive switch driver circuit is the most complex and the most expensive, and at least an external inductor, capacitor and Schottky diode are required. Especially when shielding is required, a shielding device needs to be installed, which is more expensive and larger in size. The attached comparison table is shown in Table 1.

3 LED lighting actual circuit application

After understanding the working principles, advantages and disadvantages of various drivers, we can make a simple classification and summary of the application of some common LED lighting driver circuits.

3.1 Lithium battery powered LED driver

The voltage of lithium batteries is usually between 2.5 and 4.5 V. Portable devices, including mobile phones, MP4s, and laptops, are usually powered by lithium batteries. In order to be easy to carry, the devices are small in size, light in weight, and highly integrated. Large electromagnetic interference will affect other circuits. According to the actual situation of portable devices, LED drivers need to meet the following requirements: boost drive; small footprint; low electromagnetic interference; high conversion efficiency. Small device LCD lighting only requires 3 to 9 LEDs in series and parallel, but requires high light consistency. Large device LCD lighting usually adopts the backlight module. The backlight module has handled the light uniformity problem of white light LED light sources through refraction, light guide, and other process technologies. Therefore, the requirements for light consistency are relatively low. The best circuit structure is to use a charge pump drive with LDO, which can boost voltage, occupy a small area, and has little EMI interference.

3.2 Battery powered LED driver

The battery voltage is usually between 12 and 36 V, and the input voltage is always higher than the LED tube voltage drop. In this case, only the voltage reduction operation is required, such as solar street lights, motor vehicle lighting systems, etc. In this case, the consistency of LED light emission is often lower, but the brightness of light emission is higher, so high-power LEDs are often used. At the same time, considering the highest possible efficiency and low cost, the best solution is a buck inductive switch driver.

3.3 General lighting applications under mains power conditions

Mains power supply is the most meaningful power supply method for LED lighting. It is an application that determines whether LED can be truly popularized, so it is an important issue that must be studied and solved for LED lighting. When using mains power supply, the first thing to solve is the problem of voltage reduction and rectification. At the same time, considering that the LED drive power supply will have an impact on the power grid, it is also necessary to solve the power supply noise, electromagnetic interference and high power factor problems. Therefore, circuit isolation is required to reduce pollution to the power grid. For small and medium power LEDs, the best circuit structure is to use an isolated single-ended flyback switching converter. Since the single-ended

The output power capability of the flyback circuit is limited. For applications above several hundred watts, a bridge switching converter with a larger output power should be selected.

4 Conclusion

White light LEDs are widely used, and the design of their drive circuits is very critical. Only by adopting reasonable drive methods for different application environments can the most stable and reasonable drive circuits be designed in practical applications.