Wiring panel

Most LED driver circuits use a series resistor to control the current through the LED. For applications needing a few LEDs, this is optimal. However, for applications needing many LEDs, this becomes extravagantly inefficient and it is tempting to keep the voltage drop across the resistor as small as possible. That leads to poor control of the current. ICs such as the MM5450 and its relatives and the A6275 and its relatives provide constant current outputs so that the current through the LEDs is well controlled even though the voltage drop across the circuit doing the control is acceptably small. However, the difficulty with these circuits is that because they contain many constant current drivers crowded into a relatively small package, unless the supply voltage is small, they become too hot and can destroy themselves.


This problem is not easy to solve. The solution is to maintain a small voltage across each constant current source. In this circuit, this is accomplished by REG1, the LM317L, which provides a bias of about 1.5V ±5%. Each transistor works as an emitter-follower, presenting the A6275 inputs with about 0.9V. Vled, the LED supply voltage, needs to be high enough to ensure that there will be at least 0.5V across each transistor but it is safe to allow significantly more than this and the supply need not be well regulated. The transistors can be general purpose NPN types such as BC548 and a single LM317L will easily supply a total LED current of at least 1A. A6275s are made by Allegro.

It is a charge pump design. This is where a capacitor (electrolytic) is allowed to charge and is then raised higher and allowed to discharge into a load. The load sees a voltage that can be higher than the supply.

AFX slot car sets are very enjoyable but you can increase the fun with a lap counter. This circuit will count from 00 to 99, with independent counters for each track. The sensing device used is a Hall effect sensor (UGN3503; available from Dick Smith Electronics). One of these sensors is glued under a section of each track (printed side up); between the slot and one of the track rails is the best spot. In this position, it will allow the ground effects magnets on the cars to pass over them. The sensor will provide a voltage of about 3V when a car passes over it and about 2V without a magnetic field. Both counter circuits are identical, with dual op amp IC5 handling the signals from both sensors.


IC5a and IC5b are wired as comparators, with a 2.5V reference derived from zener diode ZD1 via the 10kO and 12kO resistors. Each time the output of IC5a goes high it clocks IC1a, a 4518 BCD counter. NAND gates IC2a & IC2b provide a carry out to the other half of IC1 for a 2-digit display. More counters may be cascaded this way to provide extra digits. The BCD outputs of IC1 drive 7-segment decoders IC3 & IC4 which drive common cathode LED displays. Push-button S1 resets the counters to 00 for both tracks for the start of a new race.

This circuit is very useful while checking/operating counters, stepping relays, etc. It avoids the procedure of setting a switch for the required number of pulses. By pressing appropriate switches S1 to S9, one can get 1 to 9 negative-going clock pulses, respectively. Schmitt trigger NAND gate N1 of IC2, resistor R1, and capacitor C1 are wired to produce clock pulses. These pulses are taken out through NAND gate N3 that is controlled by decade counter CD4017 (IC1). Initially no switch from S1 to S9 is depressed and the LED is glowing. As pins 5 and 6 of NAND gate N2 are pulled up by resistor R3, its output pin 4 goes low. This disables NAND gate N3 to take its output pin 10 to high state, and no pulse is available. IC1 is a decade counter whose Q outputs normally remain low.

When clock pulses are applied, its Q outputs go high successively, i.e. Q0 shifts to Q1, Q1 shifts to Q2, Q3 shifts to Q4, and so on. If any one of switches S1 through S9, say, S5 (for five pulses), is momentarily depressed, pins 5 and 6 of NAND gate N2 go low, making its output pin 4 high, which fully charges capacitor C2 via diode D. At the same time, this high output of N2 enables NAND gate N3 and clock pulses come out through pin 10. These are the required number of pulses used to check our device. The clock pulses are fed to clock-enable pin 13 of IC1, which starts counting. As soon as output pin 1 (Q5) of IC1 turns high, input pins 5 and 6 of NAND gate N2 will also become high via switch S5 because high-frequency clock allowed five pulses during momentary pressing.

This high input of N2 provides low output at pin 4 to disable NAND gate N3 and finally no pulse will be available to advance counter IC1. Before the next usage, counter IC1 must be in the standby state, i.e. Q0 output must be in the high state. To do this, a time-delay pulse generator wired around NAND gate N4, resister R4, diode D, capacitor C2, and differentiator circuit comprising C3 and R5 is used. When output pin 4 of NAND gate N2 is low, it discharges capacitor C2 slowly through resistor R4. When the voltage across capacitor C2 goes below the lower trip point, output pin 11 of NAND gate N4 turns high and a high-going sharp pulse is produced at the junction of capacitor C3 and resistor R5.

This sharp pulse resets counter IC1 and its Q0 output (pin 3) goes high. This is represented by the glowing of LED. Ensure the red LED is glowing before proceeding to get the next pulse. Press any of the switches momentarily and the LED will glow. If the switch is kept pressed, the counter counts continuously and you cannot get the exact number of pulses.

Consumer appliances these days hardly ever have a proper mains switch. Instead, appliances are turned on and off at the touch of a button on the remote control, just like any other function. This circuit shows how a device (as long as it does not draw too high a current) can be switched on and off using a pushbutton. The approach requires that a microcontroller is already available in the circuit, and a spare input port pin and a spare output port pin are required, along with a little software. When power is applied T1 initially remains turned off. When the button is pressed the gate of T1 is taken to ground and the p-channel power MOSFET conducts. The microcontroller circuit is now supplied with power. Within a short period the microcontroller must take output PB1 high. This turns on n-channel MOSFET T1 which in turn keeps T1 turned on after the push-button is released.

Now the microcontroller must poll the state of the push-button on its input port (PB0) at regular intervals. Immediately after switch-on it will detect that the button is pressed (a low level on the input port pin), and it must wait for the button to be released. When the button is next pressed the device must switch itself of f: to do this the firmware running in the microcontroller must set the output port pin to a low level. When the button is subsequently released T1 will now turn off and the supply voltage will be removed from the circuit.

The circuit itself draws no current in the off state, and for (rechargeable) battery-powered appliances it is therefore best to put the switch before the voltage regulator. For mains-powered devices the switch can also be fitted before the voltage regulator (after the rectifier and smoothing capacitor). Since there is no mains switch there will still be a small standby current draw in this case due to the transformer. Be careful not to exceed the maximum gate-source voltage specification for T1: the IRFD9024 device suggested can withstand up to 20 V. At lower voltages R2 can be replaced by a wire link; otherwise suitable values for the voltage divider formed by R1 and R2 must be selected.

Circuit diagram:

 The Gentle Touch Circuit Diagram

The author has set up a small website for this project at http://reweb.fh-weingarten.de/elektor, which gives source code examples (which include dealing with pushbutton contact bounce) for AVR microcontrollers suitable for use with AVR Studio and GNU C. Downloads are also available at http://www.elektor.com.

Rainer Reusch - Elektor Electronics 2008

Manufacturers of cordless drills generally recommend a battery charging time of three hours. Once the charging time is up the battery must be disconnected from the charger: if you forget to do this there is a danger of overcharging the battery. This circuit, which sits between the charger circuit and its battery socket, prevents that possibility: the contact of relay Re1 interrupts the charging current when the three hours are up. Ten LEDs show the remaining charging time in steps of 20 minutes. The timer is reset each time power is applied and it is then ready for a new cycle. When power is applied IC3 is reset via C4 and R5. When the charging time has elapsed, Q9 (pin 11) goes high, which turns the relay on and interrupts the charging current.

Since Q9 is connected to the active-low EN (enable) input, the counter will now remain in this state. The charging time can be adjusted from about 2 hours 15 minutes to 4 hours 30 minutes using P1. The author set P1 to 30 kΩ, giving a charging time of 3 hours 7minutes. The greater the resistance of P1, the shorter the charging time. The timing of the circuit is not particularly precise, but its accuracy is entirely adequate for the job. When adjusting the charging time it is worth noting that the first clock cycle after the circuit is turned on (from Q0 to Q1) is longer than the subsequent ones. This is because initially capacitor C3 has to be charged to around half the supply voltage.

This little circuit is invaluable for quick go/no-go testing of just about any remote control transmitting infra-red (IR) light. The tester is battery-powered, built from just a handful of commonly available and inexpensive parts, and fits in a compact enclosure. Schmitt trigger gate IC1f is used as a quasi-analogue amplifier with, unusually, an infra-red emitting diode (IRED) type LD274 acting as the sensor element. An R-C network, C1-R2, is used at the output of the gate because all IR remote controls transmit pulse bursts, and to prevent the output LED, D2, lighting constantly when day-light or another continuous source of IR light is detected.

Circuit diagram:



Cased project:
This creates a useful ‘quick test’ option: point the tester at direct daylight, and the indicator LED should light briefly. The sensitivity of the tester is such that IR light from remote control is detected at a distance of up to 50 cm. The circuit is designed for very low power consumption, drawing less than 1 mA from the battery when IR light is detected, and practically no current when no light is detected. Hence no on/off switch is required. The construction drawing shows how the tester may be ‘cased’ using a small ABS case from Conrad.

COMPONENTS LIST
Resistors:
R1,R2 = 10MW
Capacitor:
C1 = 10nF
Semiconductors:
D1 = LD274 (Siemens)
D2 = LED, 3mm, low-current
IC1 = 74HC14
Miscellaneous:
Bt1 = 3V Lithium cell with solder tags, e.g.type CR2045 (560 mAh)