Ozone Generator High Power Ionizer Circuit

Here is a more powerful ionisator and ozone generator than described in my version-2. It is meant for continuous operation in problem spaces such as cellars with some rot & mould. It can also be operated in normal living rooms when you are not at home. But it is too aggressive for your lungs, if used continuously in inhabited spaces. It is not powerful enough to carry out short disinfection actions, for example after a burning frying pan has put your kitchen under smoke.

The transformer is from an old TV set, which still used a single rectifier instead of a tripler or even a diode-split rectifier. The latter ones you cannot use because they output positive polarity against ground, and it is very difficult to “turn around”. Here the rectifier diode was inverted to generate “Minus” at the output.

Schematic of the Ozone Generator Circuit

High power ozone generator

I have employed lacquered paper layers between the ferrite U-cores, thickness 0,04mm. This made it easier to trim the primary side for resonance.

The capacitor of 3,9nF on the primary side is to avoid voltage peaks, which would activate the built-in Zener diode of the IRF Mosfet. A complete snubber circuit is still more effective, but dissipates too much energy. The value was chosen to limit the voltage peak to a rounded half-sine shape of some 45V. (Zener voltage was ca 60V) The primary voltage is rough. The rectangular base impulses of 14 kHz are overlaid by a damped oscillation after each switching of ca 50kHz. Additionally, spurious resonances of the secondary coil can be seen on the primary side.

The capacitor is highly loaded and has to be of best quality, with a generous margin in rated voltage. If it becomes hot, it will not have a long life.

The base frequency is 14 kHz with the given values of R and C around the 4093, because this is no longer audible for most persons. If you can hear a whistling noise, push the frequency up a bit, at 18 kHz there should be no more noise. In this case take 2 windings less on the primary side. The potentiometer is for the impulse/pause-ratio, not for the frequency. Tune it to give an optimum relation between current consumption and output voltage.

You need to test by experiment, if the voltage peak caused by the primary winding in the secondary winding is negative in the moment of switching off, as the concept is a flyback converter. We want to get a soft but high voltage. Reverse the polarity of the primary winding for test to find out which direction is best.

The energy consumption is 150mA * 14,7V = 2,2W on the secondary side, a small transformer of 3,2W rated power is used. On the primary side of the grid transformer I did not measure, but it will be 3,5 – 4W.
This is acceptable for continuous operation, and the device was put into the under-floor space of a house in Inkilä-Luumäki.


After a year of operation the perforated positive-charged aluminum plate (0,3mm thickness) was found to be attacked, partly eaten up. The aluminum had very probably been converted to aluminum-oxide by the ozone. Next time I try to get 1mm stainless steel. The negative-side copper wires were still ok but possibly had become a bit shorter. In the dark, all 33 wire tips show a slight blue corona discharge.

Sniffing on the output can be done shortly for test, but it is not just “fresh air” but aggressive, so don’t take too much.

Ionizer pcb

ionizer pcb

The device generates around 10 kV and is dangerous. It has to be touch-proof for adults, and if there are small children, you need to take it away. It is nearly impossible to protect against knitting needles or nails without obstructing the flow of the ion wind.


Casings for electronic devices cost far more than what is inside. So I bought a plastic container in a household store, which was meant to store berries in the deep-freezer. The bores on the side admit air, but with a finger you cannot reach any high voltage parts. The cut-out window on the upper side passes the ion wind out, but the fingers just reach the green protection plastic. And anyhow, the aluminum plate is close to ground potential.

Sorry, comments are closed!