Disclaimer:
The author assumes no responsibility for your health, your loved ones, or your pets. If you lack experience working with high voltage or using proper protective equipment, do not assemble this device!
Warning: High voltage! Ozone is generated by a corona discharge between an electrode and an air tube. The voltage can reach 15kV or higher!
Warning: Poison! Ozone is highly toxic, carcinogenic, and mutagenic. It forms oxygen free radicals and can cause severe health issues or death. Always follow strict safety precautions when working with ozone.

This is a personal, non-commercial project created solely for educational and experimental purposes. It is not intended for practical disinfection, and both high-voltage electronics and ozone gas can be dangerous if mishandled. Use of such systems should only be attempted by those with appropriate knowledge and safety precautions.
* This is a personal hobby project created entirely outside of my professional activities. It reflects my private interests in electronics and physics, and is not related to any organization I work for or have worked for.
* The main goal was to explore air ionization and its underlying physical principles — not to build a consumer-ready device.

Background:
During the COVID-19 pandemic, disinfectants became very important. I searched for ways to sanitize groceries and household items. UV lamps were too expensive, so I explored alternatives. In my garage, I found a 250W mercury vapor lamp (similar to the Russian DRL-250) with a ballast. Initially, I considered using it for UV, but due to mercury hazards, I abandoned that idea.

Instead, I discovered a simple ozone generator design online, based on a flyback transformer. It uses a frequency generator driving a flyback, producing a high-voltage corona discharge between the HV output and a grounded metal tube. Thanks to the ionic wind effect, no fan is needed to move the ozone.

Construction:
I found a flyback transformer with an internal diode from an old TV board. The original design used a low ~1kHz frequency, which is inefficient for flybacks (they prefer 15kHz+). I consulted the original author, who said that lower frequencies favored corona discharge rather than arcing — acceptable for this project.

I built the generator on a protoboard. After applying power, the flyback buzzed, and the transistor heated up — a good sign! When I carefully brought the HV electrode closer to the tube, I observed the corona discharge.

Since the flyback’s HV cable was very short, I soldered an extension with silicone-insulated wire and covered it with multiple heatshrink layers. For the grounded electrode, I cut a piece of an old curtain rod (23mm diameter, 120mm length) — luckily, it even had a PVC coating acting as insulation.

The enclosure is a simple plastic junction box. I added an external 12V 3A power supply, a power switch, and a jack connector. To step down the 12V output from the power supply, I used a ready-made LM317 voltage regulator module, adjusted to output 5V for powering the microcontroller.

Troubleshooting:
During initial testing, the device didn’t work — I had reversed the power polarity and fried the NE555 timer IC (rookie mistake!). Without a spare NE555, I adapted an ATtiny13 microcontroller.
I added a 78L05 voltage regulator and replaced the bipolar transistor with an IRFZ34 MOSFET. The microcontroller toggles the output pin at ~20kHz, improving efficiency and eliminating transformer buzzing.

Results:
The ozone generator worked perfectly! I used it to disinfect groceries inside a plastic bucket and to eliminate mould and odors in my basement. The device is reliable and effective — but remember to always work in a ventilated area with a proper respirator and gloves!

PS:
Unfortunately, the original article I based this on is no longer available online.