A Light Emitting Diode array for battery-checks.

- and further developments of the very simple design.

The simple project
         The beginner needs easy-to-make, low-cost projects which -- if successful -- yield something useful.
         This simple version of the LED-array is a tool for testing 9 Volt batteries. Depending on the size of the components it takes up very little room and may easily fit into e.g. a matchbox.
The transistor seen in the middle of the picture deals with recharging the battery.

A few hints:
         Some diodes lack the usual markings so the polarity is not very obvious. Most multimeters are able to solve this problem, but such instruments are not always at hand so you may try something else. If the transparent plastic-encapsulation of the LED is not too frosty or turbid, you may distinguish, that one of the terminals ends up in a small plate, which actually is the semiconducting crystal. This one is the cathode (-). The other one -- the anode (+) -- proceeds in a less than hair-thin thread, which bends to make contact somewhere near the middle of the small plate. In case you are not able to peer inside the LED, you may instead connect the diode to some low DC-voltage (3 - 5 Volts) via a 270 Ohms resistor and disclose the right polarity that way.

LEDs are available in several colours. The blue ones are expensive and offers no advantage over the "other" ones, so blue LEDs are not taken into consideration for this project. The other colours work a little differently in terms of forward-voltage and light-output, but will do very well together. The resistor values are calculated for the following order (taken from "below"): red, amber, yellow and green. Any other mix will work too.

Transistors are not critical at all, provided they are able to handle about 300 mW or more. In "Elector" terms any TUP (or TUN for reversed polarity) will fit in the projects seen on this page.

         The Battery Under Test is connected correctly to the tester. Wrong polarity should be avoided, but is not liable to harm the tester. A fully charged 9 Volt BUT will make all of the 4 diodes light up; poorer BUTs will activate fewer diodes -- if any -- and most often the intensity fades during test of a nearly burnt-out battery.
. . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Second version: 1 transistor added!
         The same principles are made to work in a level-meter for use in places (such as cars), where a +12 V. powersupply is already available. The sensitivity is about the same as for the battery-tester, but the input-impedance is considerably higher making the device a suitable match for resistance-based transducers such as photo-resistors or thermo-sensitive resistors -- be it NTC or PTC.

         For a -12 V. powersupply a few changes in the circuit are needed. First the PNP-transistor is replaced by an NPN-equivalent and next all of the unipolar components, such as the electrolytic capacitor and the diodes, are reversed.

. . . . . . . . . . . . . . . . . . . . . . . . . . .

Third version: 2 more transistors added.
         The expansion is meant to make the level meter less dependent on a stabilised power-supply, which in fact it does.
         Reversal of polarity is done in the same way as for the less advanced model.

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

LED-array: experimental model.

         The next version exploits the same principle as seen above. This "inverted action" level-meter offers, however, a more sensitive input and so it may find use in some audio applications.
         I am fully aware that several integrated circuits, doing the same job quite as good or better, are commercially available at fair prices, but in terms of fun and education, they are (to my opinion) less rewarding.
         The keen observer will discover some discrepancies between the scheme on the one hand and the picture of the "prototype" on the other: the 470 Ohms resistor has been replaced by a transistor, a resistor and a yellow LED. The change was made in a bid to improve on the working conditions for the input-transistor. The extra components do work according to the expectation, but the improvement is rather modest.
18 November 1999