LV Fluro-tube driver

Fluorescent tubes are used in offices, libraries, and other places where people read things, because reading under white light feels so much better than yucky yellow light.

You may or may not know that these tubes that you see everyday are being powered by mains power, which in Australia is 240V.

Something else you may not have ever cared less about: how these tubes actually work...

Image courtesy of Sela-Light

Image courtesy of HowStuffWorks

I hear you saying, "who gives a f**k, just apply electricity to the tube right?".

No, that's not how they work. Underneath each fluorescent tube in the tube fitting is something called a starter, and beside that (hidden from view) is a coil of wire known as a ballast.

Basically, to get any light output from a fluro tube, there must be the following conditions in this order:

1. The electrodes must be heated, so that they start to emit atoms.

2. Immediately after, there must be a huge spike in voltage, that ignites the tube.

3. If the ignition of the tube is successful, it will continue the heating of the electrodes. If not, then steps 1 and 2 must be repeated.

4. The current fed to the tube must be automatically limited, as fluorescent tubes are negative resistance devices, they will self destruct if allowed to draw infinite current from a source.

The LV Fluoro-driver MARK I

So what if you want the tube to run off a low voltage, such as from a battery?

Then there must be complicated semiconductor ignition circuitry to achieve steps 1 to 4.

Basically the idea is to convert low voltage DC into higher voltage AC by feeding it into the output transformer.

The output transformer has 7 outputs:

4x 240VAC tube supply, 2x filament supply, 1x High voltage ionizing supply (~2000V)

Basically the circuit works by effectively "chopping" up the DC input into something that roughly resembles AC (but square wave). This is then fed into the output transformer that ramps up the voltage and provides the appropriate filament power. I remember the efficiency of the circuit to be up around 80%.

Here it is lighting up an 11W green fluoro tube. Incidentally, I was given this tube by the lecturers at CDU. Apparently they thought it didn't work, but as we now know, they're pretty silly.

The LV Fluoro-driver MARK II

As you saw before, the light given off the tube was quite dim, because the circuit could not provide enough power. The Mark II version employed modifications to my original circuit. One of the mods included running the transistors at a faster rate, by use of high power resistors.

These resistors had to be fan and heat sink cooled or else they would start to shoot bits of ceramic around the room.

And, just proving that it works. Here is a 20W pure white tube lit up fairly brightly, although nowhere near its maximum potential.

The circuit can power up the 11W tube to maximum brightness, but has to struggle to cope with this 20W tube. So I then went on to design a better circuit as the next version.

And the Mark III lighting up the same 20W tube as before. Notice how saturated this photo is. The whole room lights up like sunlight.

This setup draws 2A at 12V, which is 24W. Because it is powering a 20W tube, the efficiency is around about 83% (ballpark figure). Ahhh...not too bad I guess.

And just for fun, the Mark III lighting up a 10W green fluoro tube. Its not the same as the first one, but I actually also got this green tube from the guys at CDU. They gave me this one after hearing of success with the Mark I version, which was a very nice thing to do.

So there it is, a very harshly assembled, but fully operational low voltage (12VDC) fluorescent tube driver circuit. This makes it suitable for camping and hiking trips, since there won't be any 240VAC power out there.