Alternative circuit for a white LED torch

This is an alternative approach to the circuit for the White LED Torch in the December 2000 issue. It will give a usable light from a battery with an open circuit terminal voltage of 0.5V (that’s so flat that it will not run anything else at all).

The circuit will drive an 8000mcd high intensity white LED to 20mA at 3.6V from a single 1.5V cell or, by changing a resistor, from a single 1.2V rechargeable cell. The LED used for the prototype was from Dick Smith Electronics (Cat Z-3982).

The circuit is a blocking oscillator type and the components are not critical – it is almost guaranteed to oscillate. L1 and L2 are wound on an "H" shaped ferrite bobbin which measured 3.25mm inside diameter, 4.25mm inside length. These bobbins are in abundance in switchmode power supplies from old computers, monitors, colour TVs, etc. The precise dimensions are not critical although if the size is too different you may have to alter the value of R1 to compensate.

L1 is wound by first stripping the enamel from one end of 0.25mm enamelled copper wire (not critical but physical size needs to be considered) and soldering it to one of the mounting pins of the bobbin. This done, wind 100 turns and strip and solder the other end to the other pin.

L2 is wound straight on top of L1 and consists of 30 turns of the same wire. The ends of this winding are left floating and held in place by hot glue or wax. The reason for such a close turns ratio is to keep the circuit oscillating at very low voltages and very heavy loads.

R1 is 22Ω for a 1.5V battery or 10Ω for a 1.2V battery. For different LEDs or multiple LEDs you may wish to experiment with other values. Instead of risking your expensive white LED you can temporarily use two red LEDs in series. If the LED is out of circuit when the oscillator starts the voltage across C1 (output) quickly rises above 9V. Connecting this to the LED would result in its immediate destruction. Switch off and discharge the capacitor before connecting the LED or make sure the LED is never out of circuit.

L1 measures 300μH. The transistor used is not critical as long as it can handle the input current. The 1.5V circuit draws 130mA for the full 20mA output at 3.6V. If the unit fails to oscillate, as indicated by no or little voltage across C1, reverse either (not both) L1 or L2.

Philip Chugg,
Rocherlea, Tas. ($40)

Central Locking Interface Some cheap car alarms do not have a connection for the central locking system. However, in most it should be possible to find a point in the alarm circuit which is high when the alarm is activated and low when it is off. This signal can then be used to drive this relay circuit to operate the central locking system. The interface circuit converts each toggle of the alarm signal to a brief pulse to operate the two relays which then are then connected in parallel with appropriate contacts on the master solenoid in the central locking system. Frank Keller, via email. ($40)

Temperature-controlled soldering iron

One reason why commercial soldering stations are expensive is that, in general, they require the use of soldering irons with inbuilt temperature sensors, such as thermocouples. This circuit eliminates the need for a special sensor because it senses the temperature of a soldering iron heating element directly from its resistance. Thus this circuit will, in principle, work with any iron with a resistance which varies predictably and in the right direction with temperature (ie, positive temperature coefficient). A soldering iron that’s ideally suited for use with this controller is available from Dick Smith Electronics (Cat T-2100).

This circuit runs from a 12V battery or a mains-operated DC source. It works as follows: a DC-DC converter (IC1, Q1, D1, Q2, T1, D2, L1, etc) steps up the 12V DC input to about 16V. The higher voltage boosts the power to the iron and reduces warm-up time. This output voltage is applied to a resistance bridge in which the heating element of the iron forms one leg.

The other components of the bridge include resistors R7-R9 and pots VR2-VR4. When the iron reaches a preset temperature, as set by VR4, the output of IC2a goes high, sending a signal to switching regulator IC1. This forces the output of the converter to a relatively low voltage. A bi-colour LED indicates that the iron has reached the preset temperature by changing from red to green. The iron now begins to cool until it drops below the preset temperature, at which point the output voltage from the DC-DC converter goes high again and the cycle repeats.

A degree of hysteresis built into the circuit makes the LED flicker between red and green while the iron is maintained at its preset temperature.

Calibrate the circuit as follows: while the iron is still relatively cold, monitor the input voltage and current and adjust VR1 so that the input power (Volts x Amps) is about 50W. When you have done that, set VR4 to maximum and adjust VR2 so that the LED flickers between red and green when the iron has reached the desired maximum temperature.

Finally, set VR4 to mid-position and adjust VR3 so that the LED flickers when the iron reaches the desired mid-range operating temperature. As an example, you might choose to set the maximum temperature to about 400°C and the mid-range operating temperature to about 350°C. The overall temperature range, in that case, should be approximately 280°C to 400°C.

Check that the calibration is correct and repeat the adjustment procedure if necessary. Use a temperature probe, preferably one designed especially for soldering irons, rather than guesswork, when making the adjustment.

Note: VR4 should have a logarithmic taper to compensate for non-linearity in the temperature-resistance characteristic of the soldering iron.

Herman Nacinovich,
Gulgong, NSW.

Using a LED as a light sensor This circuit shows how to use an ordinary LED as a light sensor. It makes use of the photovoltaic voltage developed across the LED when it is exposed to light. LEDs are cheaper than photodiodes and come with a built-in filter, which is useful when the application involves colour discrimination. The photo-voltage of a red LED (its bandgap voltage) is typically about 2V. The source impedance of this voltage is about 800MΩ in daylight, rising to infinity in darkness. A TL071 JFET input op amp is used to amplify and buffer this extremely high impedance signal. Resistor R1 ensures that the op amp "sees" a 0V input when the LED is in total darkness. To avoid undue loading of the signal, R1 would ideally be a 100MΩ or larger resistor but since such high values are rare and expensive I used a smaller value and increased the gain of the op amp to compensate for the voltage loss. To avoid the need for a second variable resistor to set the op amp’s input offset to zero, R1 must be large enough for the reduced voltage across the LED to swamp the op amp’s input offset voltage. With a 30MΩ resistor for R1, the voltage at the op amp input when the LED is exposed to bright light is reduced to about 60mV. This is just over four times the 13mV maximum input offset of the TL071 op amp. R1 can be three 10MΩ resistors in series. Alternatively, I have found that a reverse-biased 1N4148 diode has an impedance of about 30MΩ (connect it in the circuit with the anode to ground). The output of the circuit is about 0V when the LED is in darkness. VR1 sets the gain of the op amp and it should be adjusted to give the required output voltage when the LED is exposed to bright light. Andrew Partridge, Kuranda, Qld. ($30)