The Proof of the Pudding

I managed to obtain a suitable donor lamp for my inverter project: £3 from Wilkinsons, plus another £1.99 for the bulb (9W, 14mm. screw base). The hollow china base of the lamp will contain the inverter circuit. When all is finished, you won’t be able to tell that it isn’t an ordinary mains lamp — except for the fact that the power lead will have a 6.3mm. jack plug on the end (this being the connector on which I am standardising for my 12 volt distribution boxes).

Nice hollow base:  Plenty of room for circuit.

Nice hollow base: Plenty of room for circuit.



Time for one last test before installing the circuit in its new home. Yes, that’s the light of the sun you’re seeing, after it’s been dissolved in sulphuric acid and liberated by fast-moving charged particles.

Final test before assembly

Final test before assembly

The circuit is built on two pieces of breadboard. The piece I used for the power transistors was an offcut that looked about big enough, but turned out not to be ….. Well, it would have fitted if I’d stood some of my resistors upright. You live and you learn. So the 4047 circuit has a board all to itself. If I was doing this again from the beginning, I’d sketch out the whole circuit first and build it on one board.

You can just about see the readings on the meter — which is measuring the voltage at what is now the secondary winding — and the oscilloscope, which is showing the waveform from the power board.

Oscilloscope showing waveform at transformer input

Oscilloscope showing waveform at transformer input

The two breadboards, the transformer and the voltage at the lamp.

The two breadboards, the transformer and the voltage at the lamp.

I haven’t got a picture showing this but the current from the battery is about 800mA, giving 10 watts power consumption; the bulb is supposedly a 9W device, so that’s 90% efficiency! I suspect it’s a bit less in practice, but will have to borrow a true-RMS meter to make certain.

After an hour’s soak test inside the lamp base, the output transistors were still running cool even without heatsinks; so I’m going to call this one a success.

Comments are closed.