Fullscreen HTML5Med Res (??MB)HTML4

Discrete microamp LED flasher This LED flasher has an average current of less than one microamp and flashes the LED roughly once every two seconds. The low current means it can be used as an ultra-low-power indicator in places where an indicator is not normally possible, such as a power-on indicator in devices powered by lithium watch batteries. A CR1620 cell has a usable capacity of about 70mAh, which will power this flasher for eight years. The component values in the circuit are for use with a 3V or 3.3V supply. Higher supply voltages can be used, but the average current will increase, and the flash rate will also increase. To use a higher supply voltage while keeping the supply current low, increase the value of all four 2.2MW resistors; for example, 3.9MW will work well for a 4.5V supply. The circuit is basically a relaxation oscillator. It charges a capacitor, then discharges the capacitor through the LED. There are three main sections in the circuit: • A trigger circuit, which detects when the capacitor reaches full charge. • A monostable, which generates a fixed-length pulse. • A power stage to drive the LED. The trigger circuit starts the monostable and the monostable’s output drives the power stage. The 1μF capacitor is charged over time, then powers the LED for each flash. It is charged mostly through 2.2MW resistor R1 and partially through Q2’s emitter and base resis- tors (and its emitter-base junction). PNP transistors Q1 and Q2, and the resistors around them, are the trigger circuit, which monitors the voltage across R1. As the 1μF capacitor charges, the voltage across R1 drops. When it’s below approximately 0.6V, the two transistors act as a Schmitt trigger, turning the gradually changing voltage across R1 into an abrupt signal at the collector of Q1 to start the monostable. NPN transistor Q4 and PNP transistor Q5, along with the resistors around them and the 100nF capacitor, form the monostable circuit. The 100nF capacitor charges through R6 at the same time as the other capacitor. Q4 and Q5 are wired as a discrete thyristor (SCR); once it gets the trigger signal, Q4 starts to switch on, which makes Q5 start to switch on, and they hold each other on. The monostable is powered by the 100nF capacitor, so while it is on, this capacitor discharges, ultimately Above: a 3D render of the Flasher PCB. Left: a plot of the current through LED1 (green) and R1 (blue) for the Flasher. 96 Silicon Chip Australia's electronics magazine dropping to a voltage too low to keep the transistors on, at which time the monostable turns off. NPN transistor Q3 is the power stage. While the monostable is on, the voltage across Q4’s emitter resistor is high enough to drive hundreds of microamps into the base of Q3, which goes into saturation and permits current to flow through the LED. The current through the LED discharges the 1μF capacitor, creating a flash that lasts a fraction of a millisecond. The LED choice is critical in this circuit. As the amount of power available to drive the LED is tiny, an ordinary LED will produce a barely-­visible flash. You need a wide-angle super bright LED; if the angle is too narrow, the flash won’t be visible when your eye is not on-axis with the LED. One suitable LED is Jaycar’s ZD0040 2mm red LED, rated at 600mcd with a 60° viewing angle. Although the circuit shows BC846 (NPN) and BC856 (PNP) surface-mount SOT-23 transistors, it works equally well with BC547 (NPN) and BC557 (PNP) transistors with leads. After building this circuit, clean off all solder flux residue, as flux can absorb moisture from the air and become conductive. Even a fraction of a microamp leaking through flux can prevent the circuit from working! I have designed a small SMD PCB for this circuit, shown in the 3D rendering (siliconchip. au/Shop/6/284). The accompanying LTspice simulation shows the LED current in green and the supply current in blue. We will also be selling a PCB at siliconchip.au/Shop/8/6868 Russell Gurrin, Highgate Hill, Qld. ($100) siliconchip.com.au