As pn-junction devices, LEDs exhibit V-I characteristics similar to those of conventional diodes, but with higher voltage drops across their junctions. Little current passes through an LED until the forward voltage reaches VF
(which varies from 2.5V for red LEDs to about 4.5V for blue LEDs). When VF
is reached the current increases very rapidly (as in conventional diodes), so the designer must employ current limiting to prevent possible damage. Current limiting can be implemented with three basic methods, each with advantages and disadvantages (see Table 1).
Switching power supply for HB LEDs
The example of an HB-LED power supply shown in Figure 1 is based on a fixed-frequency, highly integrated PWM switching converter (MAX5035) rated for output currents to 1A. A similar device (MAX5033) is rated to 500 mA. These inductor-based buck-regulator ICs accurately control the current through an LED or series-connected string of LEDs with total voltage up to 12 V. The MAX5035 switching frequency is 125 kHz, and its input-voltage range extends to 76 V. (The higher input voltages require higher ratings for the input capacitor and diode.) Thus, the circuit maintains the LED current constant over a wide range of input voltages. Table 2 summarizes the specifications of this circuit.
Control of the LED current is achieved by applying a voltage to the Control terminal of the figure 1 circuit - figure 2 shows the current vs voltage curve, while figure 3 shows the efficiency for this control technique.
The Control voltage is summed with the current-sense voltage (developed across the three parallel-connected current-sense resistors), and applied to the IC’s feedback terminal (FB). A control loop internal to the IC then maintains voltage on the FB pin constant at approximately 1.22V. Thus, a higher Control voltage results in lower current, because the sum of the Control and current-sense voltages (scaled by R1 and R5) must always equal 1.22V. The following equation allows design for output currents and Control voltages other than those in the example:
ILED x R5.Rsense = VREF.(R1 + R5) - Vcontrol.R1
where VREF = 1.22V and Rsense (the parallel equivalent of R2, R3 and R4) equals 5 Ohm.
In many cases it is advantageous to dim an LED by pulsing its current at a low frequency (50Hz to 200Hz) and controlling the width of the pulses (Figure 4). Though the LED illuminates with the same brightness during each pulse, the eye perceives a dimming as the pulse narrows. And, the light spectrum remains constant - unlike the case of dimming via amplitude modulation, in which the light spectrum shifts as the LED current varies.
Figure 4 shows the LED current pulsed at 100Hz in response to a squarewave Control waveform ranging from 0V to approximately 3.9V. For such low-frequency PWM dimming, the typical efficiency is higher than for the linear-current approach shown in Figure 2.
The IC in Figure 1 (MAX5035 or MAX5033) offers an efficient and cost-effective way to generate a constant-current source for driving high-brightness LEDs. Some of the benefits of that circuit are summarized below:
* High-frequency switching (125kHz) allows small reactive components (L1 and C2).
* High efficiency over a wide input-voltage range.
* Output compliance voltage (up to 12V) accommodates as many as three series-connected green HB LEDs.
* No mechanical heatsinks required.
* Voltage range can be extended to 76V to accommodate automotive HB-LED applications.
* Can be used in 24V signage and architectural lighting applications.
* Output current can be extended to 1A by changing the value of current-sense resistors R2, R3, and R4.
* High-integration power IC includes switching power MOSFET on chip.
* Linear amplitude modulation (linear dimming) of LED current through the Control input.
* Low-frequency PWM dimming via the Control input.