Interesting circuit ideas which we have checked but not built and tested. Contributions from readers are welcome and will be paid for at standard rates.

12V lead acid battery desulphator

Lead acid batteries often fail prematurely due to over-charging, under-charging, deep discharging and low electrolyte level. All of these can lead to sulphation of the plates which leads to high internal resistance and eventual failure.

Normally, this process is regarded as irreversible but this circuit is claimed to reverse the process by applying high voltage pulses to break down the lead sulphate compounds. The circuit is essentially a high-voltage pulse generator which is powered directly from the battery in question. If the battery is badly sulphated, it will be necessary to connect it to a low power charger as well, say 2A.

We have strong doubts about whether battery sulphation can be effectively reversed but we are publishing this circuit because the subject is of particular interest. This circuit has been submitted to us from a number of sources so we do not know who is the original designer. More information can be found at http://shaka.com/~kalepa/desulf

The 555 timer is connected as an astable oscillator with its output frequency set by R1, R2 and C2. Its output pulses drive the gate of Mosfet Q1 which turns on to charge inductors L1 and L2. At the end of each pulse, Q1 turns off and the inductors develop a high-voltage high-current pulse which is applied across the battery via fast recovery diode D1 and the 100μF capacitor.

The 555 is protected from the high voltage pulses via its isolated supply, by virtue of the 15V zener diode ZD1, the 47μF capacitor and the 330Ωresistor R3.

SILICON CHIP

Low supply rail detection circuit

Here's a simple low supply rail detection circuit that costs peanuts and takes just 20 minutes or so to make. Its power consumption is quite low, so it could easily be built into battery-powered devices.

Instead of using an op amp, the circuit is built around three low-cost transistors (Q1-Q3). Diodes D1-D3 form a 1.8V voltage reference (Vref) for the emitter of Q1. If the voltage across the voltage divider formed by R1 and VR1 is less than this, Q1 turns on and supplies Q2 with base bias current. This turns on Q3 in proportion to this bias current which then drives LED1.

The brightness of the LED gives an indication of the severity of the low voltage condition. The brighter the LED, the lower the supply voltage.

Trimpot VR1 is adjusted so that LED1 just comes on at the desired low-voltage point. The current consumption is typically less than 2mA when LED1 is off.

Finally, the value shown for RLED is suitable for 6-12V operation. For other voltages, RLED can be calculated using the formula RLED = (Vcc - 1.8)/0.01 (this equates to a current of about 10mA).

Trent Jackson
Dural, NSW. ($30)

Fifth channel for code-hopping remote control

This circuit was developed as a refinement to the Code-Hopping 4-Channel Remote Control featured in the July 2002 issue of SILICON CHIP. As mentioned in the article, code-hopping prevents code-cracking of the transmission but you still have a problem if someone steals your remote transmitter.

This circuit adds a fifth hidden channel with its own entry code entered via the four existing buttons. This involves adding three low-cost CMOS ICs, a transistor and several resistors and capacitors (plus an extra relay and relay driver circuit).

The extra ICs include a hex Schmitt trigger IC1, a 4016 quad bilateral switch (IC2) and a 4017 decade counter (IC3). The existing 4013 dual D-type flipflops are retained.

The idea is to input a "sequential code" to the counter (by pressing the correct sequence of buttons on the transmitter), without switching the regular channels. This is achieved by delaying the rising edge of the receiver's output pulses by 0.6s, before they trigger the flipflops which activate the relay drivers. In other words, both the on/off and momentary functions work as before but are delayed by 0.6s.

To activate any of the regular channels, the relevant transmitter button must now be held down for about 1s or longer. Alternatively, to activate the "sequential code" fifth channel, the code must be rapidly entered in the correct sequence, with each button press less than 0.4s.

Here's how it works: initially, IC3 is held reset by the voltage across the 1μF capacitor. When the code is entered, the high on pin 3 of IC3 is applied, via CMOS switch IC2, to the enable pin (pin 13). As a result, Q1 turns on and discharges the 1μF capacitor, thus releasing the reset on IC3.

When this first button is released, pin 13 of IC3 goes low, advancing the counter by 1. Similarly, when the second button is pressed, pin 2 of IC3 goes high and the process is repeated.

After the last button has been pressed and released, the 'O4' output goes high for 0.8s while the 1μF capacitor charges and resets the counter. This high in turn clocks flipflop IC5a.

Note that only three channel buttons have delay networks, to keep the IC count low. This means that one button has to be used twice to get a 4-digit code. And that in turn means that their corresponding counter outputs must be fitted with blocking diodes.

The code for the circuit shown here is 3424 - ie, output O3 from the receiver module goes to the first switch in IC2, output O4 goes to the second and fourth switches, and output O2 goes to the third switch. As a result, blocking diodes are fitted between the O1 and O3 outputs of IC3 and pins 1 & 11 of IC2.

Manfred Schmidt,
Edgewater, WA.

In-situ battery test probe This item describes a simple probe that allows in-situ AA and AAA battery voltage and current drain measurements without removing the cells from the holder. In use, the probe is simply pushed between the positive end of one cell and either the negative end of the adjacent cell or the battery holder terminal. The probe leads are then plugged into a multimeter set to a voltage range and the device switched on so that the no-load voltage can be read. Switching the multimeter to a current range allows the device to power up so that the current drain can be read. Warning: be sure your meter has a fuse in its current metering circuit, in case the device being tested has an internal short. The prototype probe was made by gluing strips of brass shim (using plastic adhesive) to both sides of a strip of tough, flexible plastic. Robin Stokes, Armidale, NSW. ($25)

Automatically armed engine immobiliser

Most immobilisers, once armed, do not allow the vehicle to be started. By contrast, this design is automatically armed each time the ignition is switched on and allows the engine to be started. The engine will then continue to run for about 9 seconds but will stop unless the immobiliser is disarmed using a small transmitter.

Alternatively, the immobiliser can be disarmed after the ignition has been turned on but before the engine is started.

The logic behind this approach is to fool the thief into thinking that the car is faulty and abandon it.

When the ignition is switched on, the LED in optoisolator OPTO1 is turned on (D1-D3 and R1 limit the voltage across the LED to about 1.8V). The transistor in the optoisolator then conducts and resets IC1b, part of a 4013 dual-D flipflop, via a 10nF capacitor. As a result, pin 13 of IC1b is initially low.

At the same time, the 10nF capacitor on pin 3 of flipflop IC1a charges via a 10kΩ resistor and causes pin 2 of IC1a to go low. This in turn triggers both sections of IC2, a 556 dual timer. As a result, pins 5 & 9 of IC2 both go high. Pin 9 turns on LED1 (a flashing LED fitted in the dashboard) to indicate that the immobiliser is armed and also supplies +5V to the emitter of Q1.

During this time, Q1's base is also at +5V (pin 5 of IC2a is high) and so Q1 is off. By contrast, Q2 is biased on by the high on pin 9 of IC2b (via D5). This activates the relay and allows the vehicle to start.

Both sections of IC2 now start their timing sequence. After about 6s, pin 5 of IC1a goes low (as set by the 5.6MΩ resistor and 1μF timing capacitor) and this turns on Q1 and sounds the warning buzzer. This alerts the driver that the immobiliser is still active.

If the immobiliser isn't disabled, pin 9 of IC2b goes low after about 9 seconds. When that happens, LED1 stops flashing, the buzzer stops and the relay turns off, stopping the engine.

Because IC1b is reset when the ignition is switched on, its Q output is low. As a result, the circuit will remain in this condition, even if the ignition is turned on and off repeatedly. The vehicle is thus effectively immobilised and cannot be restarted without the transmitter.

Pressing the button on the keyring transmitter (Oatley Electronics TX2) before the timing sequence is complete resets the timers (and holds them in the reset condition). It works like this: when the UHF receiver (Oatley Electronics RX4) receives an encoded signal, its output is fed to decoder IC3 (Oatley Electronics A5885M). If the code is valid, IC3's output goes high and clocks IC1b.

As a result, IC1b's Q output (pin 13) goes high to ensure that Q2 and the relay remain on. At the same time, pin 12 of IC1b goes low and resets both timers via a 10nF capacitor, causing their outputs to go low. This in turn causes LED1 to stop flashing and silences the buzzer (if it is sounding), ready for the next timing sequence.

IC3's output also resets IC1a, so that it is ready to change state next time the ignition is turned on.

The contacts of the relay can be connected in various ways to disable an engine. These include interrupting the +12V supply to any of the following: the ignition coil, fuel and gas cutoff solenoids or fuel injectors. In addition, you can open the connection to the starter motor solenoid to prevent the engine being cranked.

Note that the relay should be a heavy-duty automotive type with 30A contacts.

Don Adamson,
Mt. Duneed, Vic. ($50)

Simple headlight reminders These two headlight reminder circuits are easy to install and operate on the KISS (Keep It Simple Stupid) principle. The simple circuit involves adding just a 12V piezo buzzer between the lights circuit and a door switch. The buzzer sounds if the lights are left on and you open a door. The disadvantage of this simple circuit is that it's annoying to have the buzzer sound continuously if you want to leave the door open while the lights are on. The improved circuit overcomes that problem by adding a 1000μF capacitor and a parallel 100kΩ resistor in series with the buzzer. Now, when a door is opened, the buzzer gives a brief burst of sound only, while the 1000μF capacitor charges. The 100kΩ resistor discharges the capacitor when the lights are switched off. Andrew Gibbs, Jindalee, Qld. ($30)