Hysteresis control to achieve constant current LED drive

1 Introduction

As an energy-saving and environmentally friendly green lighting technology, LED (Light Emitting Diode) is widely used in general lighting, backlighting, flash, screen display, signal indication and transportation. Compared with traditional lighting sources, LED has obvious advantages, such as high luminous efficiency, high response speed, low power consumption, small size and long life. At present, there are two driving modes for LED: constant voltage driving and constant current driving. Among them, constant current driving is the most commonly used mode. Constant current driving eliminates the current change caused by the change of forward voltage caused by factors such as temperature and process, and ensures constant LED brightness. Among the LED constant current driving control modes, the hysteresis current control mode has many advantages: simple structure, self-stabilization, and not easy to cause unstable oscillation due to noise [7], etc., and its use is becoming more and more widespread. MAXIM16819 of MAXIM is an LED constant current driving chip.

This paper implements a simple hysteresis control module, which controls the on and off of the power switch tube by building a hysteresis comparison voltage inside the module and combining it with the PWM signal of the DIM control end to achieve constant current control of the LED.

2 Circuit Design and Principle Analysis

2.1 Hysteresis control principle

The application of the hysteresis control module is shown in Figure 1. The change of the LED drive current is reflected in the change of the voltage difference across the detection resistor RSENSE. In this design, the detection resistor is set to 0.5Ω. A smaller detection resistor is conducive to reducing power consumption and maintaining a higher conversion efficiency. The hysteresis current control module has two voltage thresholds built inside. The detection voltage Vcs is compared with the threshold voltage. The comparison result is ANDed with the DIM dimming signal to control the on and off of the power switch tube.

Figure 1 Application diagram of hysteresis control module

Using PWM dimming, full current is supplied to the LED during the reduced current duty cycle. For example, to halve the brightness, full current is supplied during the 50% duty cycle. Usually the frequency of the PWM dimming signal will exceed 100Hz to ensure that the pulse current is not perceived by the human eye.

The internal circuit of the hysteresis control module is shown in Figure 2. When the DIM signal is at a high level and Vcs is greater than the upper voltage threshold, the control circuit outputs a low level and turns off the power switch tube. The loop composed of LED, inductor L, freewheeling diode D and RSENSE allows the inductor to continue to provide current to the LED, and the inductor current gradually decreases, causing the detection voltage Vcs to decrease accordingly; when Vcs is less than the lower threshold voltage, the control circuit outputs a high level and turns on the power switch tube. At this time, D is cut off, forming a loop from the power supply through RSENSE, LED, L and the power switch tube to the ground. The power supply charges the inductor L, the inductor current rises, and the detection voltage Vcs rises accordingly. When Vcs is greater than the upper voltage threshold, the control circuit turns off the switch tube and repeats the action of the previous cycle, thus completing the hysteresis current control of the LED drive current, so that the drive current flowing through the LED, that is, the average value of the inductor current, is constant.

Figure 2 Internal modules of the hysteresis control module

2.2 Hysteresis comparison voltage generation circuit

4. The input voltage of 5V~28V is adjusted and converted into a constant voltage Vcc of 5V to power the subsequent circuits. As shown in Figure 3, the potential at point A is clamped by the operational amplifier and will be equal to the reference voltage 1.2V. Assuming that the output V out is at a high level, M2 is turned on, and the current flowing through M1 is IM1 = V ref / R2, and the voltage at point B is V BL = Vin - IM1 R1; when V out is at a low level, M2 is turned off, and the current flowing through M1 becomes I′M1 = V ref / (R2 + R3), and the voltage at point B increases to V BH = Vin -I′M1 R1, so the voltage change at point B is ΔV B = V BH - V BL = V ref R1 R3/ R2 (R2 + R3), which means that when V out changes from a high level to a low level, a hysteresis voltage is generated at point B. It can be seen that the hysteresis voltage has nothing to do with the input voltage, but is only determined by the reference voltage V ref and the size of the resistor. The size of the hysteresis voltage can be set by selecting the resistance value of each resistor.

Figure 3 Hysteresis comparison voltage generation circuit

2.3 Operational amplifier circuit

From the above analysis, it can be seen that the operational amplifier plays an important role. It must have a high gain to make the voltage at point A accurately follow the reference voltage, so as to accurately set the level at point B and the hysteresis voltage. In addition, since the change frequency of V out is the same as the system switching frequency (the maximum switching frequency of the system is about 2MHz), the current flowing through M1 also switches quickly between IM1 and I′M1 at the same frequency, so the unit gain bandwidth of the operational amplifier must be greater than the maximum switching frequency of the system. The designed operational amplifier structure is shown in Figure 4. The folded input structure can obtain a larger common-mode input voltage range.

From the frequency characteristic simulation diagram of the op amp, we can see that the gain reaches 84.266dB, the phase margin is 108°, and the unit gain bandwidth is about 12MHz, which meets the circuit requirements.

Figure 5 Op amp frequency characteristic simulation

2. 4 Average drive current Setting

The op amp clamps the potential of point A to the bandgap voltage reference. The cascade current mirror composed of M7 - M8, M6 - M9 mirrors the bias current I1 to the branch where M8 - M9 - R5 is located, so the voltage Vn of one input terminal of the Compara2tor module remains constant, and the voltage Vp of the other input terminal will change with the detection voltage Vcs. When the comparator output Vout is high (the switch tube is turned on), the voltage at point B is VBL, that is, the lower threshold detection voltage VCSMIN. When Vcs drops to this threshold, the symmetrical circuit structure composed of M6~M11 makes the current flowing through R5 and R6 equal, and Vn = Vp. If Vcs < VCSMIN, that is, Vp < Vn, the comparator flips and the output Vout is low. When V out changes to a low level, M2 is turned off, and the voltage at point B becomes V BH , which is the upper threshold voltage V CSMAX . The average drive current flowing through the LED is set by the average voltage at point B:

Hysteresis current range:

The above formula determines the ripple size of the driving current.

3 Simulation Results Analysis

The circuit in this paper adopts 0.5μm 5V/18V/40V CDMOS process and is simulated by Hspice Z-2007.03. The simulation results are shown in Figure 6 under the combined effect of DIM signal with pulse width of 200μs and period of 300μs and Vin = 12V (typical value).

Figure 6 Circuit simulation when Vin = 12V

The LED driving current was simulated again under the conditions of Vin = 2.5V and Vin = 28V respectively. The three simulation data results are shown in Table 1 respectively.

Table 1 Driving current under three input voltage conditions

When Vin = 12V, the temperature characteristic of the LED driving current is simulated, and the three simulation waveform results are shown in Table 2. It can be seen that the temperature characteristic of the chip is good.

Table 2 Driving current at three ambient temperatures when Vin = 12V

Since the fixed delay τ of the system has an impact on the current ripple, the actual driving current peak is IMAX + τoff di/ dt, and the current valley is IMIN - τON di/ dt, τoff is the system delay from the driving current being greater than the set value to the power switch being turned off, τon is the system delay from the driving current being less than the set value to the power switch being turned on, and di/ dt is the rate of change of the inductor current. If the inductance takes a larger value, it will have little effect on the average driving current, but it can reduce the current ripple. Otherwise, this is at the expense of increasing the volume of the external inductor.

The circuit can achieve very high efficiency. On the one hand, the power consumption in the detection resistor will cause the power dissipation of the power supply, but in this design RSENSE = 0.5Ω, then PRSENSE is quite small. On the other hand, the system efficiency is defined as the ratio of the power consumed by the LED to the power provided by the power supply, that is, η = PLED/PPOWER. Among them, PPOWER =V in3 Ivin, PLED = V LED* , from the simulation, it can be seen that the average value of Ivin is much less than , so the system efficiency can be very high.

4 Conclusion

This paper designs a hysteresis control circuit suitable for step-down LED constant current driver chip. The high-side current detection scheme is adopted, and the hysteresis current control method is used to perform hysteresis control on the drive current, so as to obtain a constant average drive current. By adjusting the external detection resistor, the constant LED drive current can be adjusted. The chip adopts 015μm 5V/18V/40V CDMOS process, and the power supply voltage range is 4.5V~28V. It can provide about a constant 350mA drive current for the LED, and the temperature characteristics are -40℃~125℃, which can achieve a fairly high efficiency. When Vin changes from 4.5V to 28V, the average drive current changes by 22mA, and the maximum constant current accuracy is 6.2%.