This single transistor flyback driver topology was created in response to achieve higher efficiency and higher output voltages from ordinary CRT television flybacks (diode split flybacks), for experiments such as x-rays or ionic lifters, without having to make any external HV multipliers. Since these flybacks are normally sealed in epoxy and can withstand an output voltage of 50 to 75 kilovolts, why not use them in drivers such as this one… 🙂 Moreover, the simplicity of this circuit also adds a possibility of some simple audio modulation of the arc. So, let’s begin!
But to say the least, the first driver I have built with this topology had an AC flyback and an input power supply rated just 60 watts. The result was a compact, short-circuit proof, high voltage “lab” supply with frequency control, some active cooling, audio modulation (singing arc, or a plasma speaker), two 5-15 kV high voltage outputs (AC and DC) to play with things such as plasma globes, small Jacob ladders, multipliers, and so on. The machine, due to its intended purpose, did not provide big fat arcs – however I think it’s still impressive and worth mentioning. Behold!
High voltage enthusiasts, who are familiar with the classic NE555 flyback driver, will notice the foil capacitor across the primary winding, which really makes wonders. (More on this here.) The output voltage depends on the oscillator frequency (set it to variable 15-30kHz for output voltage fine-tuning), number of primary turns and on the resonant capacity. Fewer turns, lower frequency and lower cap values such as 100-330 nF are going to produce over 60 kV out of a DST flyback with ease; more turns, higher frequency and higher capacity (up to 1 uF) will yield smaller output voltage with more current. Tune these factors to get the best output which suits for you. Do not forget to include the gate protection circuit (diode+resistor) on the MOSFET, but if the resonant mode is not used, omit this. Lastly, use the fifth pin for PWM audio modulation – you are going to need a 0.5-1.5W amplifier for best results. If you are not going to use this, ground the 100n cap.
The Monster Flyback Driver !
Tired of measly sparks? This flyback driver is for you then! To allow higher voltage inputs and bigger power outputs, get a separate 12-16V DC supply (a few watt transformer, e.g.) for the NE555 oscillator part. I do not suggest using a stabiliser like 7812, LM3xx’s since they are prone to the strong EMI this machine generates, and you do not want to fry the 555 chip with excessive voltage spikes… Then, substitute the “IRF5x0” transistor for a better type with at least 200V Uds and low Rds(on), i.e. IRFP250, IRFP460 or similar. Change the heatsink and the resonant capacitor if needed (a 330n-680n 250V AC MKT/MKP is enough), then connect the primary coil to at least 20 volts rectified d.c., and you are good to go!
I have given the nickname “Monster” to my second flyback driver based on this topology, because when you switch it on, all hell breaks loose. High-pitched whine, strong hiss and vast amounts of ozone are produced, high voltage wires are moving on their own, nasty static charges build up on everything conductive, strong ionic wind and corona discharge are felt even 0.5m far from the anode wire, some serious EMI is sent back to the mains: speakers buzz and hiss violently, ADSL router loses connection at times….These are just a few signs that the machine is alive and kicking.
When properly tuned, this topology draws between 3 to 8 amps on load, in a supply voltage range of 18 up to 2 volts, excluding the first example of my driver, which drew 3 amps and had been constructed for low power emphasis in mind.
Basically, you want to drive the flyback with a square wave between 15 to 30 kHz with a duty cycle between 35% (for low power) and 60% (higher power), but don’t oversaturate the ferrite core, or the flyback will get hot and short internally (especially old AC flybacks).
The thing you see here has an IRFP250N and the oscillator frequency is freely tunable between 16-30 kHz, giving an output voltage circa 20 to 55 kilovolts DC unloaded (at full blast, the arc ignition distance was 7.5 cm which might be roughly 50 kV). That corona discharge between the high-voltage anode and a concrete wall is also spectacular..
Pay attention to correct coil phasing, if using a DC flyback! You need to connect your new primary coil in a polarity so that the internal diodes in the flyback rectify each pulse in the forward way – otherwise you’ll destroy the diodes, yielding A.C. output on the flyback that will eventually eat through the insulation and fry the whole thing from inside and out.
To detect this, try connecting your primary coil in both polarities and then use a kilovoltmeter – or just observe arc striking distance at the same frequency and power supply you’re using. A correctly rectified flyback will yield more kilovoltage (bigger arc striking distance) when the polarity is correct.
Nevertheless, if you overvolt the diodes (especially with cheap small flybacks), the flyback will eventually fail even if they are forward-biased.
ESD protection (important): If you have a big flyback (like from a projection CRT) that can provide more than 50 kilovolts with this circuit, and your driver chip is close to the anode so that it accumulates static electricity, you can observe the driver chip AND the MOSFET become randomly destroyed with the flyback “hissing” free to air. If this happens, shield your driver properly, put it further away from the high voltage, and especially use e.g. 16 to 18 volt Zeners in anti-parallel with the gate of the MOSFET and ground, and also in anti-parallel with pin 8 and ground of your driver NE555, so that the static build-up gets shunted safely. A gate charge of more than 20 volts (or in some cases, 30) will kill the transistor instantly, and you want to avoid this.
Audio modulated “Singing arc”
Old prototype version