Thursday 4 August 2011

Zero voltage switching (ZVS) technique.

Though the 555 driver i ve been using is quite good in running the flyback, however the MOSFETs were easily heated up . I ve broken about 3-4 MOSFET since i ve been using it. Therefore, when i was searching for a more robust flyback driver in order to use it as my tesla coil power supply afterwards, i realized that zero voltage switching (ZVS) flyback driver can give high voltage without heating the MOSFET too much ( sometimes even none!). Hence, i turned to ZVS from that instance.

The most popular ZVS flyback driver nowadays amongst hobbyists is that designed by Vladimiro Mazilli. This driver is called zero voltage switching driver cuz it switches current exactly when the voltage across the MOSFET equals zero, and give a pure sine wave output, hence behaving like an DC-AC converter or 'inverter'. Due to its ability to convert the input DC into sine wave rather than sawtooth signal (from 555 driver circuit), hence the noise or the transformer's hum can be reduced or sometimes, entirely eliminated at all.

I ve been testing the ZVS driver with all electronic components i ve got so far ( since i dont wanna buy too much to reduce cost, i m short of money right now), and yesterday i finally found the most optimal configuration or variant of the circuit, where no more mosfet, diode and zener diode are broken ( honestly, i ve broke several zener diodes, some mosfets and also common rectifier diodes in the process).

The spark generated from the transformer is quite solid and smooth, no hum from the transformer's core, and the mosfet and all of the components on the circuit are not even getting warm at all. The only disadvantage from ZVS driver i recognised so far is that it generates less voltage than its 555 counterpart ( hence smaller spark), but i can increase it with Villard cascade since it is a sine wave alternating current anyway. Furthermore, my pack of 1000 diodes had arrived yesterday from Hong Kong since bought it last week from ebay. So maybe i can use these diodes to create Villard cascade and some juice glass bottles for the high voltage capacitors.

Here is the schematic of the ZVS driver circuit, with the components are my own selection and the original circuit can be seen here.


About the circuit ( with reference to the schematic);
Speaking from experiences, I think the most critical parts of this circuit are the diodes, and the primary inductor. Prior to using 1N4007 diodes, i used 1N4148 and broke about 4 to 5 of them, maybe due to high current ( the high current split them into two!). Therefore, if u wanna choose suitable type of diodes, make sure their ratings can stand the current u wanna use. I also tested salvaged diodes from old TV, 1N5401 ( diodes for 3A current) and it works as well.

For the inductor, i just followed the recommended value by Mazilli which is between 40 to 200 microHenries. I chose 167microHenries, by winding breadboard wire with 35 turns around a ferrite core of 1cm diameter and 6cm length. Too low an inductance ( for example, lower than the recommended range) will cause overheating, while too high ( higher than the recommended range) will cause less voltage on the output.

Other parts are not quite critical, you may test by yourself, broken mosfets and diodes are common and something u always have to bear in mind when testing a power circuit like this.

Here are photos taken:
Overall setup

Flyback transformer. Notice 6 turns on each primary winding, both primary windings are connected to each other with a center tap.

Flyback transformer and the circuit ready for running

Power supply and inductor. Yellow thing is the inductor while the black wire is the laptop power supply that gives 9V, 3.16A power.


2 comments:

  1. pernahkah terlintas di fikiran saudara yg kertas tu blh terbakor? :)

    saya terfikir, pnah cuba redirect spark tu guna magnet? Kalau tk silap saya, cas elektrik ni dipengaruhi oleh cas ion (betulkan saya kalau silap). selalunya elektrik 'zap' ke permukaan yg ada cas tertentu. Tapi adkah tanpa permukaan tu, zap masih blh terbentuk?

    apa yg brlaku jika sparx itu dilalukan di dalam lingkaran gelung yang dilalukan elektrik? Adkah sparx tu akan terbias/bengkok?

    FYI, saya cuma tahu bagi idea/soalan saja. Letrik2 bkn bidang saya :)

    ReplyDelete
  2. Indeed, yes! It always happen. There are several burns on the paper ( which i havent show u any).

    I dont have magnet in my possesion so i havent try it yet :(. But anyway, magnet field indeed has an ability to deflect electron movement.

    About sparks in electrified coil, I bet that it will be deflected as well since electrified coil possesses magnetic field. Thats the principle behind electron gun in TV, where acceleration and deceleration of electron beam is done by controlling voltage to the anode coil. Correct me if i m wrong.

    ReplyDelete