Schematic Reference

We’ve all had occasion to be annoyed at inconsiderate dog owners whose animals relieve themselves on the private property of others. This problem can hardly be solved in a lasting manner using verbal (or even physical) means, so recourse to an electronic remedy is better and friendlier. The starting point for this circuit is a ready-made passive infrared sensor (PIR), such as can be found in inexpensive movement detectors. The relay contact of the PIR energizes the power supply of the circuit shown here. The power supply generates a voltage of around 15 V after rectification by D1-D4 and filtering by R1/C3 and R2/C2. This voltage powers a square-wave oscillator comprising IC1a, R3/C1 and IC1b (acting as a buffer). The two unnecessary gates are simply connected in series with the buffer, so that they work with defined levels.

The R-C network is dimensioned such that frequency of oscillation is greater than 20 kHz. The amplitude can be set using P1. An IC power amplifier follows the oscillator to amplify this tone to a level that will be deafening for dogs (and other small creatures). We use the ST Microelectronics TDA2030 (http://us.st.com/stonline/books/pdf/docs/1458.pdf),
The peripheral circuitry corresponds to the specifications in the data sheet. With a supply voltage of 15 V, the TDA2030 can generate around 5W into a 4Ω speaker. According to the data sheet, the supply voltage of the TDA2030 can be increased to as much as 30 V, at which level it generates a hefty 16W into 4Ω (or 11W into 8Ω).

However, the 4093 still must be operated at 15V, which is the maximum allowable supply voltage for a CMOS IC. In principle, any inexpensive piezoelectric tweeter whose frequency response extends past 20kHz can be used for the speaker; it should have the highest possible sound pressure level (>100dB). A suitable type is listed on page 626 of the Conrad Electronics catalog. The impedance of such speakers rises to around 40 to 50 Ω at 20 kHz, so it is naturally not possible to obtain the power listed in the data sheet using this circuit. Nevertheless, it should be more than enough to scare off dog and master.

Using this circuit, you can change the speed of the fan from your couch or bed. Infrared receiver module TSOP1738 is used to receive the infrared signal transmitted by remote control. The circuit is powered by regulated 9V. The AC mains is stepped down by transformer X1 to deliver a secondary output of 12V-0-12V. The transformer output is rectified by full-wave rectifier comprising diodes D1 and D2, filtered by capacitor C9 and regulated by 7809 regulator to provide 9V regulated output. Any button on the remote can be used for controlling the speed of the fan. Pulses from the IR receiver module are applied as a trigger signal to timer NE555 (IC1) via LED1 and resistor R4. IC1 is wired as a monostable multivibrator to delay the clock given to decade counter-cum-driver IC CD4017 (IC2).

Out of the ten outputs of decade counter IC2 (Q0 through Q9), only five (Q0 through Q4) are used to control the fan. Q5 output is not used, while Q6 output is used to reset the counter. Another NE555 timer (IC3) is also wired as a monostable multivibrator. Combination of one of the resistors R5 through R9 and capacitor C5 controls the pulse width. The output from IC CD4017 (IC2) is applied to resistors R5 through R9. If Q0 is high capacitor C5 is charged through resistor R5, if Q1 is high capacitor C5 is charged through resistor R6, and so on. Optocoupler MCT2E (IC5) is wired as a zero-crossing detector that supplies trigger pulses to monostable multivibrator IC3 during zero crossing. Opto-isolator MOC3021 (IC4) drives triac BT136.

Resistor R13 (47-ohm) and capacitor C7 (0.01µF) combination is used as snubber network for triac1 (BT136). As the width of the pulse decreases, firing angle of the triac increases and speed of the fan also increases. Thus the speed of the fan increases when we press any button on the remote control. Assemble the circuit on a general-purpose PCB and house it in a small case such that the infrared sensor can easily receive the signal from the remote transmitter.

Several intercom circuits have appeared in www.circuit-lab.com using integrated circuits. The circuit described here uses three easily available transistors only. Even a beginner can easily assemble it on a piece of veroboard. The circuit comprises a 3-stage resistor-capacitor coupled amplifier. When ring button S2 is pressed, the amplifier circuit formed around transistors T1 and T2 gets converted into an asymmetrical astable multivib-rator generating ring signals. These ring signals are amplified by transistor T3 to drive the speaker of earpiece. Current consumption of this intercom is 10 to 15mA only. Thus a 9-volt PP3 battery would have a long life, when used in this circuit. For making a two-way intercom, two identical units, as shown in figure, are required to be used.

Output of one amplifier unit goes to speaker of the other unit, and vice versa. For single-battery operation, join corresponding supply and ground terminals of both the units together. The complete circuit, along with microphone and earpiece etc, can be housed inside the plastic body of a toy cellphone, which is easily available in the market. Suggested cellphone cabinet, with the position of switches, speakers and mike etc is shown.

This automatic door opener can be made using readily available components. The electromagnetic relay at the output of this gadget can be used to control the DC/AC door-opener motor/solenoid of an electromechanical door opener assembly, with slight intervention in its electrical wiring. A laser diode (LED1) is used here as the light transmitter. Alternatively, you can use any available laser pointer. The combination of resistor R1 and diode D1 protects the laser diode from over-current flow. By varying multi-turn trimpot VR1, you can adjust the sensitivity. (Note that ambient light reflections may slightly degrade the performance of this unit.) Initially, when the laser beam is falling on photo-transistor T1, it conducts to reverse-bias transistor T3 and the input to the first gate (N1) of IC1 (CD4001) is low.

The high output at pin 3 of gate N1 forward biases the LED-driver transistor (T4) and the green standby LED (LED2) lights up continuously. The rest of the circuit remains in standby state. When someone interrupts the laser beam, photo-transistor T1 stops conducting and transistor T3 becomes forward-biased. This makes the output of gate N1 go low. Thus LED-driver transistor T4 becomes reverse-biased and LED2 stops glowing. At the same time, the low output of gate N1 makes the output of N2 high. Instantly, this high level at pin 4 of gate N2 triggers the monostable multivibrator built around the remaining two gates of IC1 (N3 and N4). Values of resistor R8 and capacitor C1 determine the time period of the monostable.

The second monostable built around IC2 (CD4538) is enabled by the high-going pulse at its input pin 12 through the output of gate N4 of the first monostable when the laser beam is interrupted. As a result, relay RL1 energizes and the door-opener motor starts operating. LED3 glows to indicate that the door-opener motor is getting the supply. At the same time, piezo-buzzer PZ1 sounds an alert. Transistor T5, whose base is connected to Q output (pin 10) of IC2, is used for driving the relay. Transistor T6, whose base is connected to Q output of IC2, is used for driving the intermittent piezo-buzzer. ‘On’ time of relay RL1 can be adjusted by varying trimpot VR2. Resistor R9, variable resistor VR2 and capacitor C3 decide the time period of the second monostable and through it on time of RL1.

The circuit works off 12V DC power supply. Assemble it on a general-purpose PCB. After construction, mount the laser diode and the photo-transistor on opposite sides of the door-frame and align them such that the light beam from the laser diode falls on the photo-transistor directly. The motor connected to the pole of relay contacts is the one used in electromechanical door-opener assembly. If you want to use a DC motor, replace mains AC connection with a DC power supply.

An ordinary automatic room power control circuit has only one light sensor. So when a person enters the room it gets one pulse and the lights come ‘on.’ When the person goes out it gets another pulse and the lights go ‘off.’ But what happens when two persons enter the room, one after the other? It gets two pulses and the lights remain in ‘off’ state. The circuit described here overcomes the above-mentioned problem. It has a small memory which enables it to automatically switch ‘on’ and switch ‘off’ the lights in a desired fashion. The circuit uses two LDRs which are placed one after another (separated by a distance of say half a meter) so that they may separately sense a person going into the room or coming out of the room.

Outputs of the two LDR sensors, after processing, are used in conjunction with a bicolour LED in such a fashion that when a person gets into the room it emits green light and when a person goes out of the room it emits red light, and vice versa. These outputs are simultaneously applied to two counters. One of the counters will count as +1, +2, +3 etc when persons are coming into the room and the other will count as -1, -2, -3 etc when persons are going out of the room. These counters make use of Johnson decade counter CD4017 ICs. The next stage comprises two logic ICs which can combine the outputs of the two counters and determine if there is any person still left in the room or not. Since in the circuit LDRs have been used, care should be taken to protect them from ambient light.

If desired, one may use readily available IR sensor modules to replace the LDRs. The sensors are installed in such a way that when a person enters or leaves the room, he intercepts the light falling on them sequentially—one after the other. When a person enters the room, first he would obstruct the light falling on LDR1, followed by that falling on LDR2. When a person leaves the room it will be the other way round. In the normal case light keeps falling on both the LDRs, and as such their resistance is low (about 5 kilo-ohms). As a result, pin 2 of both timers (IC1 and IC2), which have been configured as monostable flip-flops, are held near the supply voltage (+9V). When the light falling on the LDRs is obstructed, their resistance becomes very high and pin 2 voltages drop to near ground potential, thereby triggering the flip-flops.

Capacitors across pin 2 and ground have been added to avoid false triggering due to electrical noise. When a person enters the room, LDR1 is triggered first and it results in triggering of monostable IC1. The short output pulse immediately charges up capacitor C5, forward biasing transistor pair T1-T2. But at this instant the collectors of transistors T1 and T2 are in high impedance state as IC2 pin 3 is at low potential and diode D4 is not conducting. But when the same person passes LDR2, IC2 monostable flip-flop is triggered. Its pin 3 goes high and this potential is coupled to transistor pair T1-T2 via diode D4. As a result transistor pair T1-T2 conducts because capacitor C5 retains the charge for some time as its discharge time is controlled by resistor R5 (and R7 to an extent).

Thus green LED portion of bi-color LED is lit momentarily. The same output is also coupled to IC3 for which it acts as a clock. With entry of each person IC3 output (high state) keeps advancing. At this stage transistor pair T3-T4 cannot conduct because output pin 3 of IC1 is no longer positive as its output pulse duration is quite short and hence transistor collectors are in high impedance state. When persons leave the room, LDR2 is triggered first, followed by LDR1. Since the bottom half portion of circuit is identical to top half, this time, with the departure of each person, red portion of bi-color LED is lit momentarily and output of IC4 advances in the same fashion as in case of IC3. The outputs of IC3 and those of IC4 (after inversion by inverter gates N1 through N4) are ANDed by AND gates (A1 through A4) and then wire ORed (using diodes D5 through D8).

The net effect is that when persons are entering, the output of at least one of the AND gates is high, causing transistor T5 to conduct and energize relay RL1. The bulb connected to the supply via N/O contact o relay RL1 also lights up. When persons are leaving the room, and till all the persons who entered the room have left, the wired OR output continues to remain high, i.e. the bulb continues to remains ‘on,’ until all persons who entered the room have left. The maximum number of persons that this circuit can handle is limited to four since on receipt of fifth clock pulse the counters are reset. The capacity of the circuit can be easily extended to handle up to nine persons by removing the connection of pin 1 from reset pin (15 and utilizing Q1 to Q9 outputs of CD4017 counters. Additional inverters, AND gates and diodes will, however, be required.

The circuit presented here can be used for connecting two telephones in parallel and also as a 2-line intercom. Usually a single telephone is connected to a telephone line. If another telephone is required at some distance, a parallel line is taken for connecting the other telephone. In this simple parallel line operation, the main problem is loss of privacy besides interference from the other phone. This problem is obviated in the circuit presented here. Under normal condition, two telephones (telephone 1 and 2) can be used as intercom while telephone 3 is connected to the lines from exchange. In changeover mode, exchange line is disconnected from telephone 3 and gets connected to telephone 2. For operation in intercom mode, one has to just lift the handset of phone 1 and then press switch S1.

As a result, buzzer PZ2 sounds. Simultaneously, the side tone is heard in the speaker of handset of phone 1. The person at phone 2 could then lift the handset and start conversation. Similar procedure is to be followed for initiation of the conversation from phone 2 using switch S2. In this mode of operation, a 3-pole, 2-way slide-switch S3 is to be used as shown in the figure. In the changeover mode of operation, switch S3 is used to changeover the telephone line for use by telephone 2. The switch is normally in the intercom mode and telephone 3 is connected to the exchange line. Before changing over the exchange line to telephone 2, the person at telephone 1 may inform the person at telephone 2 (in the intercom mode) that he is going to changeover the line for use by him (the person at telephone 2).

As soon as changeover switch S3 is flipped to the other position, 12V supply is cut off and telephones 1 and 3 do not get any voltage or ring via the ring-tone-sensing unit. Once switch S3 is flipped over for use of exchange line by the person at telephone 2, and the same (switch S3) is not flipped back to normal position after a telephone call is over, the next telephone call via exchange lines will go to telephone 2 only and the ringtone-sensing circuit will still work. This enables the person at phone 3 to know that a call has gone through. If the handset of telephone 3 is lifted, it is found to be dead. To make telephone 3 again active, switch S3 should be changed over to its normal position.

Using modulated rectangular waves of different time periods, the circuit presented here produces ringing tones similar to those produced by a telephone. The circuit requires four astable multivibrators for its working. Therefore two 556 ICs are used here. The IC 556 contains two timers (similar to 555 ICs) in a single package. One can also assemble this circuit using four separate 555 ICs. The first multivibrator produces a rectangular waveform with 1-second ‘low’ duration and 2-second ‘high’ duration. This waveform is used to control the next multivibrator that produces another rectangular waveform. A resistor R7 is used at the collector of transistor T2 to prevent capacitor C3 from fully discharging when transistor T2 is conducting. Preset VR1 must be set at such a value that two ringing tones are heard in the loudspeaker in one second.


The remaining two multivibrators are used to produce ringing tones corresponding to the ringing pulses produced by the preceding multivibrator stages. When switch S1 is closed, transistor T1 cuts off and thus the first multivibrator starts generating pulses. If this switch is placed in the power supply path, one has to wait for a longer time for the ringing to start after the switch is closed. The circuit used also has a provision for applying a drive voltage to the circuit to start the ringing. Note that the circuit is not meant for connection to the telephone lines. Using appropriate drive circuitry at the input (across switch S1) one can use this circuit with intercoms, etc. Since ringing pulses are generated within the circuit, only a constant voltage is to be sent to the called party for ringing.

Note.

• To resemble the actual telephone ringing a 400 Hz tone is switched on in the following sequence: 400ms on, 200ms off, 400ms on and 2000ms off and then repeat.

This water-level indicator uses a 7-segment display, instead of LEDs, to indicate the water level (low, half and full) in the tank. Moreover, a buzzer is used to alert you of water overflowing from the tank. The circuit shows the water level by displaying L, H and F for low, half and full, respectively. The circuit uses five sensors to sense the different water levels in the tank. Sensor A is connected to the negative terminal (GND) of the power supply. The other four sensors (B through E) are connected to the inputs of NOT gate IC 7404. When there is a high voltage at the input pin of the NOT gate, it outputs a low voltage. Similarly, for a low voltage at the input pin of the NOT gate, it outputs a high voltage.

When the tank is empty, the input pins of IC 7404 are pulled high via a 1-mega-ohm resistor. So it outputs a low voltage. As water starts filling the tank, a low voltage is available at the input pins of the gate and it outputs a high voltage. When the water in the tank rises to touch the low level, there is a low voltage at input pin 5 of gate N3 and high output at pin 6. Pin 6 of the gate is connected to pin 10 of gate N9, so pin 10 also goes high. Now as both pins 9 and 10 of gate N9 are high, its output pin 8 also goes high. As a result, positive supply is applied to DIS3 and it shows ‘L’ indicating low level of water in the tank. Similarly, when water in the tank touches the half level, pins 4 and 5 of AND gate N8 become high.

As a result, its output also goes high and DIS2 shows ‘H’ indicating half level of water in the tank. At this time, pin 9 of gate N9 also goes low via gate N4 and DIS3 stops glowing. When the water tank becomes full, the voltage at pin 1 of gate N1 and pin 3 of gate N2 goes low. Output pin 3 of gate N7 goes high and DIS1 shows ‘F’ indicating that the water tank is full. When water starts overflowing the tank, pin 13 of gate N6 goes low to make output pin 12. The buzzer sounds to indicate that water is overflowing the tank and you need to switch off the motor pump. Assemble the circuit on a general-purpose PCB and enclose in a suitable box. Use a non-corrosive material such as steel strip for the five sensors and hang them in the water tank as shown in the circuit diagram. Use regulated 5V to power the circuit.

Not only on ecologically grounds but also economically it makes sense to collect rainwater for use in the garden and increasingly for `grey water` domestic use. People who take rainwater collection seriously use large underground tanks for storage. The problem now arises, how can the water level be determined without lifting the tank hatch and peering in? One solution is to use float switches mounted at different heights in the water tank, and to use a row of LEDs mounted remotely to show the water level in the tank. Transferring this information to a remote display will involve long cable runs so it is of interest to reduce the number of cables to a minimum.

The circuit here shows how information about the water level in a tank can be sent over two wires to a remote LED display. R1 together with the resistor chain made up of R2-R6 form a voltage divider, float switches are wired across the resistors R2-R6 in one arm of the voltage divider. As water flows into the tank and the level rises, switch S5 closes followed by S4, etc. Each time a switch closes, it will short out its parallel resistor in the chain thereby changing the output voltage of the divider. When the tank is full, all five switches will be closed and all the LEDs will be on. The voltage output from this divider chain is applied to the inputs of five op amps that are configured as comparators.

A voltage chain comprising R7-R12 supplies the reference voltage for each of the five comparators. Both divider chains use the same supply so they will be insensitive to supply fluctuations. The maximum supply current for the circuit is less than 25mA. Some of the resistors chosen to make up the voltage dividers are not standard values but can be easily made up from combinations of 10kΩ and 100kΩ resistors. If you need to expand this five level display to give a better resolution of the tank contents, it is a simple job to add more float switches and to expand the voltage divider chain. IC2 also has three spare op amps; these can be pressed into service as further comparators.

Underground tanks inevitably require a pump to move the water to where it will be used. An optional feature of this design is the pump protection circuit. When LED D1 goes off indicating that the tank is almost empty, the solid state relay SSR1 can be used to switch off the mains power to the pump. This will prevent damage to the pump when the tank runs dry. The S202 SE1 solid-state relay (SSR) from Sharp has an isolation voltage between its input and output of 3000V (Class 1). It is important to note here that any mains equipment near the water tank installation must be supplied from an RCD safety socket for the sake of your own health!

With the aid of an (old) moving coil instrument it is very little effort to make a simple voltmeter that, at a glance, indicates the status of a telephone line. Because the input impedance of this circuit is very high, there is no problem in having it permanently connected to the line, since it only draws a tiny amount of current. The schematic shows that the circuit consists of no more than a series resistor, a bridge rectifier and a moving coil meter. The value of the resistor depends on the sensitivity of the moving coil meter. In his prototypes, the author used old VU meters that require 250µA for full-scale deflection. A resistor value of 390 kΩ appeared to be optimal for these meters.


For a 100µA instrument, this resistor value will have to be increased to about 680 kΩ. The starting point, when selecting a resistor value is that when the telephone is not in use, the meter should deflect about 2/3rd of full scale. The amount of meter deflection indicates the three different states of the telephone line: 1. The deflection is very small: the line is in use (voltage 5 to 12 V). 2. The deflection is 2/3rd of full scale: the line is not in use (voltage typically 48 V). 3. Full-scale deflection: ring signal (60 to 90 V AC). Because the idle voltage and certainly the ring voltage are high enough to be dangerous, it is recommended that the circuit is constructed in such a way that it presents no hazard when touched.

Depending on local regulations and the telephone company you happen to be connected to, the voltage on a free telephone line can be anything between 42 and 60 volts. As it happens, that’s sufficient to make a diac conduct and act like a kind of zener diode maintaining a voltage of 38 V or so. The current required for this action causes the green high-efficiency LED in the circuit to light. Line voltages higher than about 50 V may require R1 to be changed from 10 kΩ to a slightly higher value. When the receiver is lifted, the line voltage drops to less than 15 V (typically 12 V) causing the diac to block and the LED to go out.

The circuit diagram indicates + and – with the phone lines. However, in a number of countries the line polarity is reversed when a call is established. To make sure the circuit can still function under these circumstances, a bridge rectifier may be added as indicated by the dashed outlines. The bridge will make the circuit independent of any polarity changes on the phone line and may consist of four discrete diodes, say, 1N4002’s or similar. Finally, note that this circuit is not BABT approved for connection to the public switched telephone network (PSTN) in the UK.

The Maxim MAX 6665 (www.maxim-ic.com) provides a complete temperature-dependent fan controller. It can switch fans operating at voltages of up to 24 V and currents of up to 250 mA. The IC is available from the manufacturer in versions with preset threshold temperatures between +40 °C (MAX6665 ASA40) and +70 °C (MAX6665 ASA 70). The device’s hysteresis can be set by the user via the HYST input, which can be connected to +3.3 V, connected to ground, or left open. The following table shows the hysteresis values available:

HYST = Hysteresis
open = 1 °C
ground = 4 °C
+3.3V = 8 °C
The other pins of the SO8 package are the FORCEON input and the status outputs WARN, OT and FANON. The test input FORCEON allows the fan to be run even below the threshold temperature. The open-drain output WARN goes low when the temperature rises more than 15 °C above the threshold temperature, while the open-drain output OT indicates when the temperature is more than 30 °C above the threshold. The push-pull output FANON can be used to indicate to a connected microcontroller that the fan is turned on.

Nowadays, just about every house has an outside lamp with a motion sensor. Such a device eliminates the need to feel your way to the front door, and it apparently also scares away intruders. The only problem is that free-running dogs and cats in the neighborhood have little regard for such lamps and continue to deposit their excrement in the garden, once they have found a habitual location there for this purpose. This gave rise to the idea of connecting a sort of siren in parallel with the outside lamp to clearly advise dogs and cats that they are not welcome.

Naturally, it would be nice to avoid startling the entire neighborhood with this alarm signal. Here we can take advantage of the fact that dogs and cats have a significantly better sense of hearing than people. Not only are their ears more sensitive, they can also perceive significantly higher frequencies. With people, the upper limit is around 18 kHz, but dogs and cats can hear frequencies in excess of 20 kHz. We can take advantage of this by building a siren that emits a frequency just above 20 kHz. This will scare off dogs and cats, but people will simply not hear it.

All we need for this is an oscillator with an amplifier stage and a tweeter that can reproduce such high frequencies, such as a piezoelectric tweeter. The schematic diagram shows how easily this can be implemented. The power supply for the entire circuit is formed by the components up to and including C2. The 230-V leads are connected in parallel with the motion-sensor lamp. C1 and R1 provide capacitive coupling to reduce the 230 V to an acceptable voltage. A DC voltage of approximately 9.1 V is generated from this voltage using a bridge rectifier and D1, filtered and buffered by C2. The oscillator is built around R3, C3 and IC1a.

The frequency of this oscillator is rather dependent on the specific characteristics of IC1, so the values shown here should be regarded as guidelines. If the oscillator frequency is too high, it can be reduced by increasing the value of R3 and/or C3. If the frequency is too low (which means that the siren tone it is audible), the value of R3 and/or C3 should be increased. The square-wave signal from the oscillator is applied to the input of an H bridge composed of several Schmitt triggers in combination with the final output stages (T1–T4). This approach causes the peak-to-peak value of the square wave signal to be twice the supply voltage.

As a result, a respectable 18 V is obtained across the piezoelectric tweeter, which is sufficient to produce a quite loud whistle tone. When building the circuit, you should bear in mind that it is directly powered from 230 V and not electrically isolated from the mains network. It is thus necessary to avoid contact with all of the components when the circuit is in use. In practice, this means that the circuit must be fitted into a well-insulated, waterproof box. If you want to test the circuit, it is a good idea to first discharge C1 using a resistor, since it can hold a dangerous charge. You must also ensure that components F1, C1, R1 and B1 all have a mutual insulation separation of at least 6 mm!

The electronic dog repellent circuit diagram below is a high output ultrasonic transmitter which is primarily intended to act as a dog and cat repeller, which can be used individuals to act as a deterrent against some animals. It should NOT be relied upon as a defence against aggressive dogs but it may help distract them or encourage them to go away and do not consider this as an electronic pest repeller. The ultrasonic dog repellant uses a standard 555 timer IC1 set up as an oscillator using a single RC network to give a 40 kHz square wave with equal mark/space ratio.

This frequency is above the hearing threshold for humans but is known to be irritating frequency for dog and cats. Since the maximum current that a 555 timer can supply is 200mA an amplifier stage was required so a high-power H-bridge network was devised, formed by 4 transistors TR1 to TR4. A second timer IC2 forms a buffer amplifier that feeds one input of the H-bridge driver, with an inverted waveform to that of IC1 output being fed to the opposite input of the H-bridge.

This means that conduction occurs through the complementary pairs of TR1/TR4 and TR2/TR3 on alternate marks and spaces, effectively doubling the voltage across the ultrasonic transducer, LS1. This is optimised to generate a high output at ultrasonic frequencies. This configuration was tested by decreasing the frequency of the oscillator to an audible level and replacing the ultrasonic transducer with a loudspeaker; the results were astounding. If the dog repellent circuit was fed by a bench power supply rather than a battery that restrict the available current, the output reached 110dB with 4A running through the speaker which is plenty loud enough!

The Dog and Cat repellent was activated using a normal open switch S1 to control the current consumption, but many forms of automatic switching could be used such as pressure sensitive mats, light beams or PIR sensors. Thus it could be utilize as part of a dog or cat deterrent system to help prevent unwanted damage to gardens or flowerbeds, or a battery powered version can be carried for portable use. Consider also using a lead-acid battery if desired, and a single chip version could be built using the 556 dual timer IC to save space and improve battery life.

The consumer unit (or ‘electricity meter’) cupboard in some older houses is a badly lit place. If the bell transformer is also located in this cupboard, it may be used to provide emergency lighting by two high-current LEDs. These diodes are powered via a small circuit that switches over to four NiCd batteries when the mains fails. The output voltage of the bell transformer is rectified by bridge B1 and buffered by capacitor C1. The batteries are charged continuously with a current of about 7.5 mA via diode D1 and resistor R2. The base of transistor T1 is high via R3, so that the transistor is cut off. When the mains voltage fails, C1 is discharged via R1; when the potential across it has dropped to a given value, the battery voltage switches on T1 via R3 and R1, provided switch S1 is closed. When T1 is on, a current of some 20 mA flows through diodes D4 and D5. The light from these LEDs is sufficient to enable the defect fuse or the tripped circuit breaker to be located.

This circuit will illuminate a LED if one of your telephones is in use. It should work in all countries (Including UK) that have a standing line voltage above 48 Volts DC. Please note that it is illegal to make a physical permanent connection to your telephone line in some countries (this includes the UK and Ireland). If building this circuit it is advisable to use a plugin cord so that the unit can be unplugged should a fault occur. If in doubt consult either your telephone or cable operator.

If all extension phones are on-hook and the line voltage is around 48 V, Q1 will conduct thus effectively shorting the gate of Q2 to its source, so it will be off and the LED will be disabled. Lifting the handset of any phone on the line causes the line voltage to drop to 5-15 V. The gate voltage of Q1, equal to some 6% of the line voltage, will then be too low and Q1 will be turned off. So Q2's gate is now biased at approximately 1/2 of the line voltage, Q2 turns on and the LED indicates that the line is in use. The circuit itself is practically invisible to the other telephone devices using the same line. LED1 must be low-current and its current-limiting resistor must be 2k2 or more.

The other components' ideal values may vary slightly, depending on the local telephone line parameters. The circuit is powered off the telephone line. If other types of MOSFETs are used, the 500k trimmer can be adjusted to ensure that Q1 is biased fully on while the line is not in use (LED1 off), and vice versa. If Q2 is not a BS108 but some other 200 V MOSFET with a higher G-S threshold voltage, it might be necessary to increase the value of the lower (or decrease the value of the upper) one of the two resistors connected to the gate of Q2. Plain (bipolar junction) transistors can be used instead and the circuit also works fine.

But the resistor values are then much lower - letting ten times more microamps of current pass through while the line is not in use, and even this MOSFET design still could not meet formal minimum on-hook DC resistance specifications. Both prototypes' PCBs were 4x1 cm. The current-limiting resistor for LED1 is 2k2 in both cases. DO NOT ground any of the leads or conducting surfaces in this circuit. A more reliable design would also include some kind of over-voltage protection etc.

Warning:

In their normal course of operation, telephone lines can deliver life-threatening voltages! Do not attempt to build any of the circuits/projects unless you have the expertise, skill and concentration that will help you avoid an injury. There are also legal aspects and consequences of connecting things to telephone lines, which vary from country to country. Keep away from telephone lines during a lightning storm!

This remotely controlled watering system is both inexpensive and easy to expand. It is designed to operate in conjunction with a conventional watering timer and allows remote switching between nine zones. The prototype is used in a bore system, where a deep-well pump must be started and kept running while zones are being changed. This is necessary to minimise cycling and results in maximum pump life. A standard portable telephone is used as the transmitter and receiver. The system’s range is therefore limited only by the telephone specifications. The prototype uses an Audioline model CDL1A, set to pulse-dial mode via a switch in the side. Selecting zones from the telephone keypad couldn’t be simpler.

For the first nine zones, each key number (1-9) corresponds directly to a zone number. If additional zones were added to the basic circuit, "0" would represents zone 10, while further zones are "dialed-in" by simple addition. For example, to select station 15, you’d press "0" and then "5". Looking now at the circuit, the telephone base station is wired to one input of a hex Schmitt-trigger inverter (IC5a), which functions as a low-pass filter and pulse shaper in conjunction with two 1kO resistors, a 10µF capacitor and a second inverter (IC5c). Glitch-free pulses are fed to the clock inputs of two 74HC164 8-stage shift registers (IC3 & IC4).

The A & B inputs of IC4 are permanently pulled high, so the first pulse results in a logic high at output O0 (pin 3). Each additional pulse causes the next successive output to go high. After eight pulses, output O7 (pin 13) goes high and this is propagated to the second shift register (IC4) via its A & B inputs. The shift register outputs are wired to a collection of 74HC86 exclusive-OR gates (IC6-IC8) in such a way that only one of the 74HC86 outputs can be high at a time. For example, after three clock pulses, outputs O0-O3 of IC4 are high, which results in IC7c’s output going high.

The exclusive-OR gates feed a pair of ULN2001A Darlington drivers (IC9 & IC10), which in turn drive relays to switch power to the water solenoids. If a wrong key is pressed at the remote end and 10 pulses arrive at the shift register inputs, output O1 of IC3 will go high, triggering both 555 timers (IC1 & IC2) via inverter IC5e. The 555s are configured as monostables, so their outputs immediately swing high. IC2 resets the shift registers, returning all outputs to their initial (low) state. The reset signal is held for about three seconds, which ensures that any number of additional pulses (within reason) above the maximum of nine will be ignored. In the meantime, IC1 energises one of the water solenoids via diode D2 and the zone #1 driver circuit.

This solenoid is held on for about 20 seconds, giving sufficient time for the number to be redialled after the 3-second redial "hold-off" period. This solenoid "hold-on" period is important as it prevents overheating of the pump motor that might otherwise occur without continuous water flow. The circuit operates from +5V, which is generated by a conventional bridge rectifier (BR1), filter and regulator arrangement. 24VAC for the water solenoids is obtained from the water system timer transformer and is external to this circuit.

Editor’s note:

• For the "sorry, wrong number" feature to be effective, some form of operator feedback would be required if all of the sprinklers are not visible. Perhaps a siren could also be driven by IC1’s output to alert the operator that a valid sector number must be dialled within 20 seconds!

If your letterbox is some distance from your house, you will find a monitoring device useful to indicate when new post has arrived. This can take the form of a US-style visible flag; a more modern alternative uses a 433 MHz radio transceiver. The big advantage of the solution presented here is that is can use an existing two-core bell cable, without requiring any further power source. The arrival of post is signalled by a blinking LED; for added effect, a digital voice recorder can also be connected which will, at regular intervals, announce the fact that the letterbox needs emptying. The device is silenced by a reset button.

The circuit uses one half-cycle of the AC supply to power the bell or buzzer, and the other half-cycle for the post indicator. Suitably-oriented diodes in the device and in the letterbox ensure that the two signals are independent of one another (Figure 1). The bell current flows from K1.A through D3, bell-push S2, D1 and the relay back to K1.B. C1 provides adequate smoothing of the current pulses to ensure that the relay armature does not vibrate. The bell is operated by the normally-open relay contact. If the bell is actually a low-current piezo buzzer, then it can be connected directly and the relay dispensed with.

During the half-cycle for the letterbox monitor current flows from connection K1.B on the bell transformer through current-limiting resistor R1, the LED in the optocoupler, reed contact S1 (a micro-switch can also be used) and D2 and finally back to K1. If the reed contacts are closed, the LED in the optocoupler will light and switch on the phototransistor. A positive voltage will then appear across R3 which will turn the thyristor on via C6. The red LED will indicate that post has arrived. Pressing S3 shorts out the thyristor, reducing the current through it below the holding value. A small extra circuit can be added to provide continuous letterbox monitoring.

This takes the form of a voice recorder whose ‘play’ button is operated by transistor T1. T1 in turn is driven by a 555 timer IC. In the usual 555 timer circuit, where the device is configured as an astable multi-vibrator, the mark-space ratio cannot be set with complete freedom. Here two diodes provide separate charge and discharge paths for capacitor C4. When capacitor C4 is charging, D5 conducts and D4 blocks: the charge rate is determined by R5. When discharging, D4 conducts and R6 and the potentiometer determine the rate. The values shown give a pulse length of approximately 0.5 s with a delay of between 15 s and 32 s.

The short pulse is sufficient to trigger the voice recorder module via transistor T1 connected across its ‘play’ button. The voice recorder module (e.g. Conrad order code 115266) is designed to run from a 6 V supply. The maximum recording time is 20 s and the current consumption is 20 mA when recording and between 40 mA and 60 mA when playing back. Since our supply is at 8 V, the excess voltage must be dropped using between 1 and 3 series-connected 1N4148 diodes (shown as Dx in the circuit diagram). The final voltage should be checked using a multimeter. Alternatively, a 7806 can be used without suffering a significant loss in volume.

If it is desired to use a piezo buzzer to provide an acoustic signal, the pulse length should be increased to at least 2s. In this case, R5 should be increased to 560 kΩ or 680 kΩ: the pulse length, t on, is 0.7.R5.C4, and the interval between pulses, t off, is 0.7. (R6+R7).C4. Suitable buzzers are available with a wide range of rated voltages.

Most electric kettles do not produce a whistle and just switch off when they have boiled. Fitting a box of electronics directly onto an electric kettle (or even inside!) to detect when the kettle has boiled is obviously out of the question. The circuit shown here detects when the kettle switches off, which virtually all kettles do when the water has boiled. In this way, the electronics can be housed in a separate box so that no modification is required to the kettle. The box is preferably a type incorporating a mains plug and socket. In this application, the current flowing in coil L1 provides a magnetic field that actuates reed switch S1. Since the current drawn by the kettle element is relatively large (typically 6 to 8 amps), the coil may consist of a few turns of wire around the reed switch.

The reed switch is so fast it will actually follow the AC current flow through L1 and produce a 100-Hz buzz. The switching circuit driven by the reed switch must, therefore, disregard these short periods when the contacts open, and respond only when they remain open for a relatively long period when the kettle has switched off. The circuit is based on a simple voltage controlled oscillator formed around T2 and T3. Its operation is best understood by considering the circuit with junction R4/R5 at 0 V and C4 discharged. T2 will receive base current through R5 and turn on, causing T3 to turn on as well. The falling collector voltage of T3 is transmitted to the base of T2 by C4 causing this transistor to conduct harder.

Since the action is regenerative, both transistors will turn on quickly and conduct heavily. C4 will therefore charge quickly through T2’s base-emitter junction and T3. Once the voltage across C4 exceeds about 8.5 V (leaving less than 0.5 V across T2’s b-e junction), T2 will begin to turn off. This action is also regenerative so that soon both transistors are switched off and the collector voltage of T3 rises rapidly to +9 V. With C4 still charged to 8.5 V, the base of T2 will rise to about 17.5 V holding T2 (and thus T3) off. C4 will now discharge relatively slowly via R5 until T2 again begins to conduct whereupon the cycle will repeat. The voltage at the collector of T3 will therefore be a series of short negative going pulses whose basic frequency will depend on the value of C4 and R5.

The pulses will be reproduced in the piezo sounder as a tone. The oscillation frequency of the regenerative circuit is heavily dependent on the voltage at junction R4/R5. As this voltage increases, the frequency will fall until a point is reached when the oscillation stops altogether. With this in mind, the operation of the circuit around T1 can be considered. In the standby condition, when the kettle is off, S1 will be open so that C1 and C2 will be discharged and T1 will remain off so that the circuit will draw no current. When the kettle is switched on, S1 is closed, causing C1 and C2 to be discharged and T1 will remain off. C3 will remain discharged so that T2 and T3 will be off and only a small current will be drawn by R1.

Although S1 will open periodically (at 100 Hz), the time constant of R1/C1 is such that C1 will have essentially no voltage on it as the S1 contacts continue to close. When the kettle switches off, S1 will be permanently open and C1/C2 will begin to charge via R1, causing T1 to switch on. C3 will then begin to charge via R4 and the falling voltage at junction R4/R5 will cause T2/T3 to start oscillating with a rising frequency. However, once T1 has switched off, C3 will no longer be charged via R4 and will begin to discharge via R3 and R5 causing the voltage at R4/R5 to rise again. The result is a falling frequency until the oscillator switches off, returning the circuit to its original condition.

As well as reducing the current drawn by the circuit to zero, this mimics the action of a conventional whistling kettle, where the frequency rises as more steam is produced and then falls when it is taken off the boil. The circuit is powered directly by the mains using a ‘lossless’ capacitive mains dropper, C6, and zener a diode, D2, to provide a nominal 8 V dc supply for the circuit. A 1-inch reed switch used in the prototype required about 9 turns of wire to operate with a 2-kW kettle element. Larger switches or lower current may require more turns. In general, the more turns you can fit on the reed switch, the better, but do remember that the wire has to be thick enough to carry the current. It is strongly recommended to test the circuit using a 9-volt battery instead of the mains-derived supply voltage shown in the circuit diagram. A magnet may be used to operate S1 and so simulate the switching of the kettle.

Warning:
This circuit is connected directly to the 230-V mains and none of the components must be touched when the circuit is in use. The circuit must be housed in an approved ABS case and carry the earth connection to the load as indicated. Connections and solder joints to components with a voltage greater than 200 volts across them (ac or dc) must have an insulating clearance of least 6 mm. An X2 class capacitor must be used in position C6.

This circuit was designed to produce an audible alarm when the mains power is interrupted. Such an alarm is essential for anyone whose livelihood depends on keeping perishable foodstuffs in cold storage. The circuit is powered by a 12-V mains adapter. LED D5 will light when the mains voltage is present. When the mains voltage disappears, so does the +12 V supply voltage, leaving the voltage regulator IC1 and relay driver T1-T2 without power. The relay driver, by the way, is an energy-saving type, reducing the coil current to about 50% after a few seconds. Its operation and circuit dimensioning are discussed in the article ‘Relay Coil Energy Saver’. The value of the capacitor at the output of voltage regulator IC1 clearly points to a different use than the usual noise suppression.


When the mains power disappears, Re1 is de-energized and the 0.22 F Gold-cap used in position C4 provides supply current to IC2. When the mains voltage is present, C4 is charged up to about 5.5 volts with IC1 acting as a 100-mA current limit and D10 preventing current flowing back into the regulator output when the mains voltage is gone. According to the Goldcap manufacturer, current limiting is not necessary during charging but it is included here for the security’s sake. The CMOS 555 is configured in astable multivibrator mode here to save power, and so enable the audible alarm to sound as long as possible. Resistors R5 and R6 define a short ‘on’ time of just 10 ms. That is, however, sufficient to get a loud warning from the active buzzer. In case the pulses are too short, increase the value of R5 (at the expense of a higher average current drawn from the Goldcap).