THE 240v FLUORO
Basically the 240v circuit relies on heating the cathodes at the ends of the tube so that a relatively low voltage will strike the tube. After this, the filaments can be turned off and the gas will be remain conducting.
In the 12v arrangement, the output of the transformer must be sufficient to strike a cold tube and keep it illuminated without heating the cathodes.
We have all seen fluoro's in the kitchen of almost every home, and thousands in department stores. But little do we realise the complex engineering that has gone into their design (a lot of trial and error) and how the three items of a fluoro circuit work.
These three items are: a fluorescent tube, a starter and ballast.
Figure 2 shows the circuit for this. When a fluorescent is turned on, it flickers a few times then comes on.
What happens is this: When the power is applied, the tube has a very high resistance between its ends and the only path seen by the 240v mains is through the ballast, the filament at one end of the tube, the starter and finally through the filament at the other end of the tube.
The ballast is a single winding on an iron core and it has a number of functions. Firstly it acts as a current limiting device so that when the circuit is first turned on, the voltage across the starter is about 130v. The starter consists of a glass tube with two electrodes. The 130v is the characteristic voltage of the starter and is governed by the mixture of helium and hydrogen in the tube.
When a voltage is applied to its terminals, the gas inside begins to glow and heats up a bi-metallic strip (one of the electrodes) and it touches the other electrode.
This shorts out the starter and turns off the glow. The ballast now provides a voltage of about 8 - 12v for each filament and they begin to glow.
They are only designed to glow to a dull red and this begins to heat up the gas in the fluorescent tube ready for the next part of the cycle.
After a second or two, the bimetallic strip in the starter cools and breaks contact. This causes the current in the ballast to reduce suddenly and a voltage of 800v to 1,000v is produced across its winding.
This voltage is passed to the ends of the tube. Since the gas at the ends of the tube has already been heated by the filaments, the 1,000v is sufficient to strike the rest of the gas along the length of the tube. Once this occurs, current starts to flow and a voltage of about 110v appears across the tube.
This voltage is again governed by the characteristics of the ballast and is not enough to restart the glow in the starter. If the tube does not strike on the first occasion, the starter glows again and this is where the flickering comes from.
The final function of the ballast is to limit the current so that the tube consumes its rated wattage. If a tube that has been "struck" is placed directly across the mains, a very high current will flow and it would blow up as the resistance between the ends is very low.
It is classified as having "negative resistance" as the flow of current causes the impedance of the tube to decrease.

20 WATT FLUORO
The wattage of a fluorescent tube is a characteristic of its size and vice versa; the wattage determines the size. That's why a 20 watt tube is 2 feet long and a 40 watt tube is 4 feet long.
The design of a tube is a complex mathematical equation (a lot of experimentation has gone into its design).
The tubes are filled with a gaseous mixture that produces ultra-violet light when operating and this light hits against the walls of the tube to excite the coating on the inside to produce visible light.
The gas and the coating are all poisonous and the emission of the ultraviolet waveform from an uncoated tube is quite dangerous to the eyes, so experimenting with tubes other than operating them as per this project is to be avoided.
You can buy uncoated tubes (called ultraviolet tubes) for lighting effects, EPROM erasing, special heating, and germicidal applications. Do not use any of these tubes in this project as the voltage we are producing will create different effects and damage your eyes.

12v INVERTER
Now for the 12v version.
The circuit doesn't require many components but its operation is quite complex. The clever component is the transformer. It performs 3 functions.
Firstly it is acting as a feedback component for the transistor to create an oscillator circuit. Secondly it is providing a high voltage (over 1 000 volts) to strike the tube and keep it struck and thirdly it is supplying spikes of energy to illuminate the tube.
The circuit is shown in figure 1 and we will take a detailed look at how the transformer carries out the three functions.
The transformer in this project is not a lethal device as the output wattage is slightly below the value that produces electrocution. However the output is in excess of 1,000v and the ends of the secondary winding should not be touched when the transformer is operating.
To get a shock you must touch both
ends at the same time - it is not sufficient to touch one end and any other part of the circuit as the secondary is an isolated winding.
Even a simple transformer such as the one we are winding in this project will demonstrate a number of interesting features. One of them is the ability to step-up a voltage. This is the main purpose in this project as we require a voltage of approximately 1,000v to strike the tube.
Another interesting feature is the availability to get positive or negative voltage (phase) from a separate winding on the transformer, simply by connecting the winding around one way or the other. In this project we connect the winding to get positive feedback so that a single transistor will drive the circuit.

HOW THE OSCILLATOR WORKS
The oscillator works on positive feedback. This positive feedback comes from a separate 13 turn winding on the transformer called the feedback winding.
The cycle starts by turning on the transistor a fair bit via the 180R resistor on the base and this causes current to flow in the primary winding. The flow of current causes magnetic flux to be produced by the winding and this passes through the ferrite core.    The feedback winding is also wound around the core and the magnetic lines pass through this winding and produce a voltage.
The winding is connected to the transistor so that the voltage from the winding ASSISTS the voltage from the 180R resistor and causes the transistor to turn on harder.
Thus more current flows through the primary winding and the magnetic flux increases. This causes more voltage to be produced in the feedback winding and the transistor turns on even harder.
This continues until the transistor is fully turned ON and maximum current is flowing in the primary winding.
Now comes the important part.
Even though maximum current is flowing in the primary winding and maximum flux is produced in the core, this flux is a steady flux and not an increasing flux.
The only time a voltage is produced in a secondary (or feedback) winding is when the flux is INCREASING (or decreasing). When the flux is stationary, the voltage in any of these windings ceases to be produced.
Thus we come to a point in the cycle where the current in the primary is a maximum but the voltage in the feedback winding is zero.
The only current flowing into the base of the transistor comes from the turn on resistor but this is note enough to fully turn the transistor ON and so the transistor turns off a small amount.
This has the effect of reducing the flux in the ferrite rod and we now have a situation where the flux is DECREASING. This changes the situation in the feedback winding.   The voltage in the feedback winding is now produced in the OPPOSITE DIRECTION and the transistor begins to turn off even more.
The magnetic flux begins to collapse very quickly and this produces a very high reverse voltage on the base of the transistor (up to about 25v) to turn the transistor off completely.
This is how we get a positive and negative voltage for the transistor.
This is quite an amazing achievement as the voltage through the primary doesn't change direction - it merely increases and decreases in value - but the voltage from any of the other windings changes direction!
The collapsing magnetic flux cuts the turns of the secondary winding and produces a voltage of about 2.5v per turn in the 450 turns, making a total of about 1 ,000 appearing across the ends of the tube. This is sufficient to strike the tube and as we mentioned above, the resistance (or impedance) of the tube reduces as more current flows. In our case the voltage across the tube is about 400v (this voltage depends on how hard the tube is driven and is riot a fixed value).
When the magnetic flux has almost fully been converted to electrical energy, the 180R turn-on resistor on the base of the transistor starts the cycle over again.
The voltages produced by the transformer are very spiky and the gas in a fluorescent tube is very quick to react to these spikes. The gas produces ultraviolet light that strikes the fluorescent material on the inside of the tube and causes it to produce visible light.
The tube forms part of the load for the transformer and has an effect on quenching the spikes to the transistor so it is not advisable to operate the transformer without the tube connected.

WINDING THE TRANSFORMER

The ferrite core of the transformer is an antenna rod from an old transistor radio. You could use a slab antenna but we have chosen a 10mm diameter rod, 8cm long and the first winding to be wound on it is called the primary.
This consists of 58 turns of wire spaced slightly apart so that is occupies the centre 6cm of the rod. The first thing to do is wind two layers of insulation around the rod so that the wire does not touch the rod and create a short.
Leave the first 8 - 10cm of wire and start winding with the thick .61mm wire. But firstly hold the end of the wire in place with a short piece of sticky tape folded over itself and stuck along the length of the rod, where the winding is to be placed. Continue winding and fix the other end in a similar manner, leaving 8 - 10 cm for connecting to the rest of the components.
Place one layer of paper over this winding and secure both the start and end of the paper with sticky tape. Make sure this insulating paper is tight by rolling it like a cigarette before sticky-taping.
The next layer is called the feedback winding and consists of 13 turns of the thin .28mm wire, wound in a spiral fashion so that it takes up the full length of the 6cm.
Terminate both the start and finish of the winding with sticky tape to prevent it unwinding. Cover this with a layer of insulation.
Now, for the final winding, called the secondary.
This consists of 450 turns of .28mm wire, wound in 3 layers of 150 turns.
The winding does not have to be neat and you could quite easily jumble-wind the turns and it would work perfectly ok, however there are two factors to remember.
The voltage between the start and finish of this winding will be about 1,000 volts and the insulation on the wire is only about 100v. So the start and the finish must not be near each other. This also applies to most of the other turns so the best way to prevent the inner-turn sorts is to carefully wind the turns side-by side.
This also produces the best results.
Leave 8 - 10 cm of wire and hold the start in place with a piece of sticky-tape folded over itself and stuck to the insulation. Wind 150 turns neatly across the 6cm of the transformer and hold the last turn in place with sticky tape before placing a layer of insulation over the winding. Continue with the next layer and one more, making a total of 450 turns.
Cover the last layer with insulation, tin the ends of each of the windings with a hot soldering iron and plenty of solder and the transformer is complete.