Plasma Arc Tweeter Built Inside a Toolbox

Nov 05, 2026 Last reply: 1 day ago 4 Replies
Jump to replies (4)

Plasma speaker builds usually stay in novelty territory, but this one treats the idea as a genuine tweeter. Magneticitist housed the working unit inside a black plastic toolbox, and the maker describes the result as a great tweeter with unusual clarity, though not super loud.



The circuit sits on a flat metal plate that doubles as a heatsink strip, with a red PCB carrying electrolytic capacitors, screw terminals, an IC and a rotary pot. An LM386 amplifier board takes a 3.5 mm input, an XY-PWM board generates the drive signal, and an IRFP4232 MOSFET switches the primary of a TV flyback transformer. A 120 watt Mean Well supply feeds the unit at around 20V and 5A, while clean 20-25V DC runs two buck converters. Inside the box there is an aluminum heatsink with a cooling fan, and the side panel carries a toggle switch, potentiometers and a red voltage readout. A digital panel meter tracks frequency and duty cycle, and the maker also drew a hand-drawn circuit diagram.



Tuning comes down to adjusting gate drive voltage, PWM frequency and the trim pot, and the setup lets the maker change volume and quality on the fly. It runs near 25 kHz at 60% duty cycle, and the maker hears it clearly above about 150 Hz. As a bonus, a radio next to the arc keeps its tuning untouched, and there is no ozone smell at all from the hot whitish arc.



There are trade-offs, of course. Audio pops at times, and dropping the volume stops it but costs loudness; driving closer to 20V or more brings clear volume back without popping or distortion. The electrodes are still being perfected for a longer solid arc that holds steady, and the maker admits not knowing the real consensus on ozone safety when a fairly big arc is needed for good volume. A clear plastic container over the electrode shields the arc from air currents, a practical touch in my opinion.



Project by Magneticitist. Watch it on their YouTube channel:

formatting link


Putting the PWM driver on its own battery feeding its buck driver strikes me as a sharp move. Keeping that supply separate from the main 24V rail should help with clean drive, though it does add one more battery to maintain. Given how fussy these arcs can be, that trade looks worth it to me.

Curious about the monitoring side of this. What exactly does the digital panel meter display while the unit is running? A live readout like that sounds handy for dialing in the arc, so I would like to know what numbers it actually puts up.

The panel meter displays frequency and duty cycle. In the running unit it read 25.0 kHz and 060%, matching the roughly 25 kHz at 60% duty cycle the drive runs at. Separately, the red voltage readout on the side showed roughly 15.4V to 16.4V during use.

The maker reckons a simple 555 timer circuit could probably stand in for the PWM board, and that seems like a worthwhile variant to try. A 555 stage would trim the part count and make replication easier for anyone working from the hand-drawn diagram. Adding the quiet fan and 5A bimetal fuse the maker suggests would round it out nicely.

Join the Discussion

Have something to add? Share your thoughts — no account required.

Didn't find your answer?

Ask the community — no account required