Hello all, I'm very new with fluorescent lamps and overall lamps as well. I recently acquired some 18 watt Phillips Master PL-C form factor 2 pin lamp which should have an integral starter to use with an inductive ballast, however where I live it is impossible to find any inductive ballasts because fluorescent lamps are almost entirely phased out. I had several ideas, one being getting an isolation transformer and join one side of the primary with one side of the secondary which would essentially turn it into a big inductor, or I can get a big ferrite toroid (Hesitant because ferrite don't work well 50Hz) and wind 3000 turns of 0.2mm enameled copper wire. Ideally I want to run this long term so I need a transformer with a fairly large core. Thanks in advance.

You need an inductor not hard saturating at the 200V or so and passing 0.2A current. Except the last, transformer primary is that. To get the current,you need to introduce a gap intothe magnetic circuit, so that means disassembling the fore and assembling it so it has the required gap.
To set the gap, you use just some cardboard or so insert, verify the current first when connected directly to mains (to get about 20% above the lamp rating) and then connect with the lamp and fine tune it to get the current where it should be.
You need the winding to handle the current, so it means to start from a 40VA or so transformer with the core made so it is possible to fiddle with it, so an EI or C core (the C would be the best, but they were very rarely used).

Ferrite has no problem with 50Hz per se, but it has limited saturation flux (below 0.3T at temperature; silicon steel is in the 1.5..2T ballpark), plus it is quite expensive material mainly for larger sizes, so rather impractical for such low frequencies, where the steel works better.

Then I'll probably rip a microwave transformer for the EI core and gap between the E and I, could potentially reuse the secondary copper as well. I've heard to test the cumulative resistance (wire resistance + reactance), a resistive load is connected behind the inductor, then measure the voltage across it and hence get the resistance. The lamp is rated at 0.230 amp so the tube voltage should be 18/0.23 at around 78V, should I gap the transformer for this voltage? Thanks for the information!

Lamp voltage is non-sinewave, so usual Ohm/Joule law calculations are not precise. Better google a lamp datasheet and get running parameters there. Also, there is a series of IEC standards for different discharge lamps, you can get recommended ballast impedances from these. IEC 60901 says lamp voltage 100V, working current 0.22A, ballast impedance 800 Ohms @ line voltage 220V

I don't believe you can't get any fluorescent ballasts in Australia. Electrical distributors last stock, classifieds, trash cans in the end


 

The aim of the game is to get a right lamp current, 0.22A here. Use of a good quality RMS meter recommended. Fluorescent lamps don't usually mind if you err on the current in 50% - 150% range briefly.

The ubiquitous 2x18/1x36/40W fluorescent tube choke has an impedance of ~390 Ohms, so for trashcan picks, two of such chokes in series will make almost perfect ballast for 18W CFL lamp.

 

Lamp voltage is non-sinewave, so usual Ohm/Joule law calculations are not precise. Better google a lamp datasheet and get running parameters there. Also, there is a series of IEC standards for different discharge lamps, you can get recommended ballast impedances from these. IEC 60901 says lamp voltage 100V, working current 0.22A, ballast impedance 800 Ohms @ line voltage 220V

I don't believe you can't get any fluorescent ballasts in Australia. Electrical distributors last stock, classifieds, trash cans in the end

The 4 pin version of this lamp is 0.21A @ 80V, so probably slightly higher impedance.

I can't get my hands on any inductive ballasts unfortunately, new installations from the last 5 years are mostly LED, even T8 or T5 lamps are retrofitted with electronic ballast, I believe inductive ballasts are phased out in the 2010s and they were never accessible to regular consumers, sparkies only. Not completely unable to find per se but I'm not paying a hundred bucks for some copper wound around an iron core. Too late to pick anything useful from the dumpsters.

80V / 0.21A /16.5W is a rating of this lamp for HF operation at >20kHz according to IEC 60901

All that pains involved, I personally would consider getting this nice cheap electronic ballast https://www.ebay.com.au/itm/144125025642 and removing lamp's starter from the base by carefully opening it, converting the lamp to 4-wire version. Or just get two/or triple-fold 18W 4-pin tube nearby on Ebay, whichever you like better

That particular model of Tridonic.ATCO HF ballast actually seem to strike 2 pin 18W lamps fine even with the starter in circuit! Certainly did when I tried it, it of course didn't preheat the cathodes and I don't know what would happen long term or at EOL so in no way advisable to do but worked for a quick test.

Then I'll probably rip a microwave transformer for the EI core and gap between the E and I, could potentially reuse the secondary copper as well. I've heard to test the cumulative resistance (wire resistance + reactance), a resistive load is connected behind the inductor, then measure the voltage across it and hence get the resistance. The lamp is rated at 0.230 amp so the tube voltage should be 18/0.23 at around 78V, should I gap the transformer for this voltage? Thanks for the information!

MOTs are too huge for this, but you may try... But be sure to physically remove the secondary from the core first (e.g. by a hacksaw; so you won't get zapped even accidentally, the kV's there are no joke, it is really instantly lethal). Then you may try the primary current when connected directly to the mains, it could already be something ballpark close; unless the MOT itself does not have significant gap by itself - to compensate of the otherwise lead phase of the HV circuit (the series capacitor...). If this gap is formed by the center pillar of the E side being cut shorter, you are done, you can not make the gap smaller that easily...

But the question is, where and how to get something other, when potted transformers became so popular in the last at least 30 years or so...
Maybe gutting some trashed "hifi" set...

They are too big indeed, quite wasteful making a low power fluorescent ballast out of it, would be good for a higher power MV/MH ballast though. I'll try find some more alternatives.

All that pains involved, I personally would consider getting this nice cheap electronic ballast https://www.ebay.com.au/itm/144125025642 and removing lamp's starter from the base by carefully opening it, converting the lamp to 4-wire version. Or just get two/or triple-fold 18W 4-pin tube nearby on Ebay, whichever you like better

The sole reason I bought these lamps is to replace the old halogen (which recently went out) in the garage, they were a dollar a pop so no point buying it and not use it, will probably get an electronic ballast then hack off the base, sounds good.

With an electronic ballast the starter presence is not that a problem. But an absence of the capacitor which decent makers put parallel to the starter but some cheepeese tend to omit, will be a problem. If your multimeter has capacitance measurement,you may try to measure the lamp. If you get few nF reading, the capacitor isthere and it will work with the electronic. But if you measure nothong (reading below 1nF),there is no capacitor inside so it won't work with electronic. Well unless you open the starter compartment and put some there (needs to be 1kV rated, some 3.3 or 4.7 nF).

To my understanding, the capacitor forms a LC circuit with the output choke and determines the frequency right? At least in simpler self oscillating designs. I've opened a cheapo Chinese ballast before, there's only a few components. So by lowering the value of the cap, is it possible to increase the oscillating frequency and therefore increase the impedance of the choke? Cause I've got a 36W T8 ballast collecting dust and if I can modify that it would be fun, provided it's got 13003 transistors which would work up to 2MHz. Probably not good for the lamp if the frequency is too high, but oh well.

The capacitor controls the frequency (and also the maximum voltages and currents) practically only during ignition. But for that the system is not that much sensitive (it is mainly the lamp what is the ultimate voltage limitter), Once the lamp ignites, the arc becomes low enough impedance so the capacitor is practically invisible in the impedance.
The only exception could be some older programmed start ballasts using open loop style frequency control for preheat, where the programmed frequency assumes exact L and C to get the desired preheat current and limited voltage across the lamp (so it won't ignite when electrodes are not warm yet), because the capacitor then determines how detuned the LC is and what are the impedances.
But many more modern ballasts use just timed frequency sweep down, intentionally to make it more robust against component tolerances (the resonant capacitor tends to lose capacitance as it wears out by the starts, if the system is immune to it, it offers either longer lifetime or allows the use of cheaper components).
In your case the preheat will be affected by the starter anyway, so you will be even better off with some really simple selfoscillating power stage ballast.