#video #toread about #manufacturing a simple DIY #ozone generator
on 02024-05-08#manufacturing a high-power #ozone generator #hardware #video using 3000 volts on each of several heatsunk plates to get corona discharge, plus a waterproof industrial fan to blow air through; he says the overall machine, with 3 modules totaling 12 plates, produces 120g/hour of ozone. The frame is welded up from steel structural tubing, with rivet nuts to add threads to the holes drilled in the tubing to which to screw the fan, ozone modules, frosted plexiglass side panels, aluminum top and bottom and fan panels, and air intake grille. The high-voltage wires have all the panels in parallel and are electrical-taped together, potted with household silicone inside a snap-together plastic conduit, zip-tied to the steel frame. He dremels holes in the side panels to epoxy-mount the electronic modules such as the timer, power meter, power entry module, and power switch. The modules’ screw terminals are potted in silicone presumably to prevent ozone damage. Uses about 350W. He doesn’t show the high-voltage power supply.
on 02024-04-06#bigclive high-power shady £16 #ozone generator teardown with schematic #electronics #hardware #video. Arrived with a broken ceramic plate through which it arced. Operates around 4000V and uses the “ridiculously high power” of 155 watts with a power factor of 94%. He says it really needs to be mounted in an enclosure with a fan so it doesn’t overheat. The wires to the high-voltage conductive-ink grids (very hard ink, harder than his box cutter) are silicone-insulated for high temperature, and the solder connections are silasticked over, presumably to prevent lower-voltage corona discharge. He explains that capacitive coupling through a high-dielectric-strength insulator gives you lots of tiny little sparks when you’d get an arc without the insulator. The input bridge rectifier is made out of 1N5399 1.5A 1kV diodes. An interesting oscillator design using a sort of totem pole of two npn transistors, perhaps FJP13009 (700V 12A; they’re labeled J13009-2), driven from (opposite-sense?) feedback windings on a high-current toroid, only used for feedback; its primary is in series with the primary of the resin-potted high-voltage transformer (2Ω primary, 100Ω secondary). Surprisingly to me, there’s a diac in the oscillator, which he says is the “start circuit”. The transistors are mounted on a giant aluminum heatsink through an insulating film (mica?). The primary of the toroid is tied to both rails of the power supply through rather large (0.47μF 400V) film caps, which discharge through the transistors when they come on. There’s no storage cap, so the oscillation dies down when the 50Hz signal is close to its zero-crossings. Originally I thought it was a blocking oscillator, but apparently the feedback through the toroid is positive, not negative; what terminates a pulse is when the film cap, discharging through the two transformers, runs out of charge, and thus stops turning on the transistor it was discharging through. “Normally, with transistors to a positive rail, you’d have a pnp transistor, but in this instance, because it effectively has its own transformer-coupled supply to its base here from the emitter, you can use standard npn transistors for both of them. (...) Much simpler circuitry than I was expecting.” The diac to the base of the negative-rail transistor is activated by a .01μF cap charged up through a 470kΩ resistor pulling it up toward the positive rail, though I’m not sure why you need the cap. The size of the heatsink makes me think it can’t be dissipating more than 5–10W, which means better than 92% efficiency.
on 02024-04-06