In order to drive more than one high brightness white LED, the design engineer needs to choose whether to connect the LED in series or in parallel. The parallel connection only needs to apply a lower voltage at both ends of each LED, but the ballast resistance or current source needs to be used to ensure the consistent brightness of each LED. If the bias current flowing through each LED is different, their brightness is also different, resulting in uneven brightness of the whole light source. However, using ballast resistance or current source to ensure consistent brightness of LED will shorten the service life of battery.
Using series connection can ensure the consistency of current, but it is necessary to apply high voltage to the LED string. In order to achieve proper lighting brightness, ordinary white LEDs need a bias voltage of 3.6V and a maximum bias current of 20mA. Figure 1 shows a low-cost inductive boost circuit that can adjust the brightness of seven white LED strings.
This circuit can be divided into two parts: the boost circuit composed of Q1 and Q2 and the control circuit composed of Q3 and jfet1. Assuming Q1 is off, when the battery voltage is slightly higher than the VVB of Q2, the base of Q2 will flow a positive current (IB = (battery voltage VBE) / rjet1). At this time, Q2 is on and inductor L1 is grounded.
As the current on L1 increases at the speed of di / DT, the energy is stored in the L1 magnetic field. As the current gradually increases, it also flows through the resistance RSAT of Q2 (sd1 and LED string are in the cut-off state). The collector voltage of Q2 is high enough to turn Q1 on. The base voltage of Q1 is connected to the collector of Q2 through a feedforward network composed of R1 and C1. R1 is also used to limit the base current of Q1.
After Q1 is turned on, the base of Q2 is driven to ground, so Q2 is cut off, and the energy of L1 is released into the LED string with the weakening of the magnetic field.
The fast zero return action of L1 applies a forward bias voltage higher than 26V on the LED string to make the LED emit white light. Since the human eye cannot feel the high-frequency flicker of the LED, the circuit can provide illumination with constant brightness. When L1 discharge ends, Q1 returns to the cut-off state.
During normal operation, this self oscillation action is repeated until the battery voltage drops to the sum of VBE less than Q2 and jfet1 voltage drop (about 1V), and Q2 is no longer conductive. The RSAT of L1 and Q2 and the switching characteristics of Q1 and Q2 will also affect the oscillation period and duty cycle.
The voltage of the battery pack (four alkaline batteries) is increased to more than 26V to provide a forward bias to the LED string composed of seven white LEDs in series.
The small DC current flowing through R4 (less than 20ua) biases Q3 to adjust the channel resistance of jfet1, so as to adjust the battery leakage current to prolong the battery life. The grid voltage of jfet1 is about 0.9V higher than the battery pack voltage. Here, p-jfet is used as a depletion device. When VGS is equal to zero, p-jfet turns on.
The source of ET is connected to the battery terminal. The design engineer can turn off the channel by increasing the grid voltage (higher than the battery positive voltage). The higher the grid voltage is than the battery voltage, the greater the channel resistance.
Therefore, when the battery pack voltage decreases from 6V to 3V, the oscillation frequency decreases (the VGS of jfet1 will change slightly). At this time, the brightness of the LED decreases slightly. Ideally, the control loop will keep the LED current constant. However, the sensitivity of human eyes to light obeys the quasi logarithmic relationship, so the small linear decrease of brightness is not easy to be detected before the battery voltage drops to about 2V.
Another scheme is to keep the output power (the product of current and voltage) of the battery unchanged. Due to the internal resistance loss of the battery, although this can keep the brightness of the LED unchanged, it will shorten the battery life, and the complexity of the circuit will be greatly improved. In short, the LED brightness of this simple circuit will change little throughout the battery life.
The brightness of the LED string can be slightly adjusted. For example, the design engineer can adjust the manufacturing deviation of the triode and led by slightly changing the resistance value of R2, so that the light output (unit: lumen) can be set to a fixed value.
When the battery pack is about to run out of energy, the LED string with dim light can be short circuited and only one LED can be connected. At this time, as long as the battery pack has 1V residual voltage, this led can emit strong light. This single LED connection can use waste batteries to provide final emergency lighting.
In terms of safety, when using alkaline batteries, all batteries must be matched. When the energy of the battery with the least energy in the battery pack is completely exhausted, and other batteries have enough energy to form a reverse bias to the energy exhausted battery, the energy exhausted battery will overheat and leak emulsion acid, resulting in safety problems.
In order to achieve battery matching, it shall be ensured that all 4 batteries are replaced at the same time with new batteries in the same package. The rated capacity of 4 AA alkaline batteries is 4 × 1000mah, which means that the LED can be illuminated continuously for about 61 hours. The test results of the circuit prototype show that the continuous lighting time is a little more than two days (48 hours).