12vdc Fluorescent

Please support WebConX by visiting this sponsor.

Fluorescent light operational description.

The circuit is built around a hex schmitt trigger inverter package in this case a 40106.IC1 is supplied with approximately 10V by way of R9,C2 and Z1 in order to help ensure steady performance and to protect it from excessive battery voltage swings.IC1B is set up as an oscillator by way of C1 and the resistance established by P2 and R2.IC1C,1D and 1E are set up as inverting buffers to improve the overall quality of the outgoing waveform and thus ensure minimal heating of Q3. In the absence of any other influences the output from the oscillator would be a squarewave ie 50/50 mark space ratio with a working frequency of about 16KHz. R3 pulls the oscillators' feedback pin upwards which causes the oscillator to spend more time in the low state in compensation for this offset. Since the oscillators' output is inverted this translates to a longer conduction time for Q3,more current drawn from the 12V supply and also a very high voltage (about 1200Vpp) delivered to the tube. When the tube 'fires' it draws a much greater current which increases the voltage developed across R12. As this voltage rises it forward biases the B-E junction of Q4 which causes its' collector circuit to draw current down through R13 and thus counteract the original effect of R3. It is by this means that the circuit regulates its' output voltage on the basis of current flow in the fluorescent tube and thus emulates a conventional fluorescent fitting.

Increasing the value of R12 offers the option of reducing the intensity of the emitted light - this ought to be useful if the lights are to be left on at night for the purpose of navigation to the kitchen for that secret midnight snack. Although the filaments of the tube are not used they will be seen to glow to some extent as the current flowing through the tubes' gas envelope heats them. This feature permits the use of tubes that have one or both filaments blown. Since the tube does not have to be heated at its' ends it will light almost instantly when switched on. I have tested these at room temperatures of -5 degrees Celsius and +45 degrees Celsius and have found them to operate without any sign of failure.

Evolution of the circuit.

This circuit is born of experimentation aimed at making a circuit that would be easy to build and maintenance free. The first of my designs worked ok but was not capable of regulating itself in response to changing lamp conditions which often caused the fluorescent tube to give a blinding white light and become too hot to touch. The unit described here has served me well for the last 3 years without any of the problems that conventional mains powered fluorescent lights suffer from.

About the fet.

The fet chosen for the switching function is an IRF540 it will handle up to 100V and 28 Amps. Normally a fet does not require any real current to drive its' gate circuit but as this circuit zips along at about 16KHz into an inductive load things change. Due to capacitance within the fets' drain/gate region there will be an influence on the gate voltage each time the fet changes its' conduction state. As the fet switches on - there is a rapid downward movement of the drain voltage which actually tries to pull the gate potential down with it by virtue of the capacitance mentioned before. The opposite effect will be seen occurring when the fet switches off.

In any circuit that switches a fairly high current this gate swing effect can be very destructive for the transistor.

A current buffer circuit made from Q1 and Q2 deals with this problem by making a high current push/pull signal available to the gate pin of the fet and thus eliminating the undesirable effects described above.

Heatsinking is still a necessary evil here - a piece of aluminum about 2mm thick by about 100mm square bolted to Q3 will perform admirably as a method of thermal relief.

Transformer.

This is probably the most fiddly part of the whole thing. I used transformers that I took from the power supplies of old Philips tv sets. To pull the transformers apart I boiled them in water for about 10 mins to soften the glue that binds them then eased them apart - this needs to be done carefully in order to avoid breakage. Once apart I unwound the transformers and stored the wire for later reapplication to the bobbins.

Rewinding of the transformer starts with the primary which consists of 4 lengths of 0.63mm enameled copper wire wound 10 times around the bobbin. These turns need to be wound neatly so as the conductors are laid side by side without any overlapping. The primary is then completed by covering it with two layers of masking tape. The secondary winding is a lot easier by comparison - simply wind 50 turns of 0.63mm enameled wire onto the tape covered primary winding while taking care to ensure a fairly neat finish.

The need for a tidy finish is not to do with appearance but is due to an annoying little thing called 'leakage inductance'. This becomes worse as the magnetic coupling between the primary and secondary windings is degraded by poor assembly practices. When the leakage inductance reaches a certain point it can cause the transistor that drives the primary to overheat due to the avalanche effect - this can also destroy the transistor. Reducing leakage inductance will also enhance efficiency.

Testing:

After inspecting your work for errors attach a fluorescent tube to the transformer output terminals and apply 12VDC through a 40 watt light globe to the power inputs shown on the circuit diagram. The light globe is simply a means of current limiting in case of catastrophic circuit failure. If all goes well the fluoro tube will glow and maybe flicker due to the effects of the light globe.

Now it is time to remove the globe from the circuit and attach an ammeter in order to confirm that the current draw is about 3amps. Some adjustment of P2 may be required to maximize the effective light output from the tube. This adjustment may also be necessary in order to get the circuit working in the first instance. Once the circuit is operational it will be able to take care of itself by virtue of its' current sensing design which ensures a fairly steady power level being offered to the tube during battery voltage swings.

Have fun lighting the way.

Steven Tetlow.

additional info:

Don Klipstein's Lighting Info Site! - http://www.misty.com/~don/light.html

PMB Electronics (Net-Tech Developments) - http://www.pmb.co.nz/fluor_1.htm

Energy - Efficient Lighting for the Home - http://www.webconx.com/lighting.htm

Home

Please support WebConX by visiting this sponsor.