Weird Science

ZVS Driver: The Heart of High-Voltage DIY Projects

Polish ver­sion is here

Even during home expe­ri­ments, there is often a need to use high vol­tage, much higher than the 230V found in stan­dard ligh­ting cir­cu­its. At such times, high-vol­tage conver­ters of various kinds are par­ti­cu­larly use­ful. These are elec­tro­nic cir­cu­its that step up the vol­tage. Below, I will pre­sent one of these cir­cu­its, which, in my opi­nion, best com­bi­nes sim­pli­city of con­struc­tion with good per­for­mance. This device was first built by Mr. Vla­di­miro Maz­zilli.

Ori­gi­nal sche­ma­tic:

Ilustracja

As we can see, the sche­ma­tic is very sim­ple and con­ta­ins few com­po­nents. I used IRF450 MOS­FETs, which work very well. The high-vol­tage trans­for­mer comes from an old TV set. The high-vol­tage secon­dary win­ding (usu­ally embed­ded in white pla­stic) sho­uld remain unchan­ged. The pri­mary win­ding sho­uld be wound accor­ding to the instruc­tions on the sche­ma­tic, with a cen­ter tap.

War­ning: I do not recom­mend buil­ding the ZVS cir­cuit for begin­ners without expe­rience in han­dling high vol­tage. The out­put of the device car­ries vol­tage in the range of seve­ral tho­u­sand volts (seve­ral kV). The elec­tric arc is extre­mely hot: it can instan­tly char the skin and flesh of a care­less expe­ri­men­ter. The author assu­mes no respon­si­bi­lity for any inju­ries, dama­ges, or los­ses resul­ting from the con­struc­tion or use of this device; pro­ceed at your own risk!

This conver­ter has very high effi­ciency. This is because the trans­for­mer’s pri­mary win­ding, toge­ther with the capa­ci­tor con­nec­ted in paral­lel, forms a paral­lel reso­nant cir­cuit. Due to the push-pull con­nec­ted tran­s­i­stors, the con­di­tion for oscil­la­tion of the reso­nant cir­cuit at its reso­nant fre­qu­ency is ensu­red. For a paral­lel reso­nant cir­cuit, cur­rent reso­nance occurs. The alter­na­ting cur­rent flo­wing thro­ugh the trans­for­mer win­ding thus rea­ches its maxi­mum pos­si­ble value. Zener dio­des pro­tect the tran­s­i­stor gates from vol­tage sur­ges.

The capa­ci­tor must be of good qua­lity. The best cho­ice would be a pulse capa­ci­tor desi­gned for power sup­plies. I used a bat­tery of four such capa­ci­tors of 0.25µF each, resul­ting in a total of 1µF. These capa­ci­tors must be rated for a vol­tage at least three times higher than the sup­ply vol­tage.

The impe­dance-mat­ching coil is best wound on a toro­i­dal core e.g. from a dama­ged com­pu­ter power sup­ply. Appro­xi­ma­tely 20-30 turns of wire with a dia­me­ter of 0.5 - 1.0mm (AWG 20 - AWG 18) is suf­fi­cient.

All cir­cuit com­po­nents sho­uld be moun­ted secu­rely on an appro­priate board. A prin­ted cir­cuit board (PCB) can be used, but I used a board made of thick press­bo­ard. It was scre­wed onto a sui­ta­ble heat­sink that dis­si­pa­tes the heat gene­ra­ted by the tran­s­i­stors:

The tran­s­i­stors are not visi­ble because they are loca­ted under the board and are scre­wed to the heat­sink thro­ugh sili­cone insu­la­ting pads.

The cir­cuit wor­ked so well that after some time, I made a newer ver­sion:

After assem­bly, the cir­cuit sho­uld work right from the first power-up. Howe­ver, make sure that the power source has suf­fi­cient cur­rent capa­city, as the cur­rent can reach seve­ral ampe­res. The vol­tage at the out­put of the secon­dary win­ding can reach seve­ral or even tens of kilo­volts (kV), with a high cur­rent. This sho­uld be han­dled with extreme cau­tion! An elec­tric arc a few cen­ti­me­ters (about an inch) long, with a tem­pe­ra­ture of seve­ral tho­u­sand degrees Cel­sius (seve­ral tho­u­sand °F), is gene­ra­ted. Do not use cop­per wires to ignite the elec­tric arc, as they may melt. The tran­s­i­stors may gene­rate signi­fi­cant heat during maxi­mum arc stretch.

The conver­ter also works with other types of high-vol­tage trans­for­mers and fly­back trans­for­mers from modern TVs and CRT moni­tors.

Thanks to this device, many inte­re­sting expe­ri­ments can be con­duc­ted:

Using this conver­ter (simi­larly to my Tesla coil), you can gene­rate dischar­ges in the gas fil­ling light bulbs. The inert gas inside the bulb has suf­fi­cien­tly low pres­sure to trig­ger ava­lan­che ioni­za­tion. This mani­fe­sts as visi­ble dischar­ges and glow.

To obse­rve the dischar­ges, it is best to use two secon­dary win­dings con­nec­ted in series and pla­ced on a com­mon core:

One ter­mi­nal of the secon­dary win­ding must be pro­perly gro­un­ded, while the other sho­uld be con­nec­ted to the light bulb soc­ket. Light bulbs of any power rating can be used, from 40 to 100 watts (W), or even more. I used a light bulb with a power of… 800W. In the photo, you can com­pare its size with a 100W bulb.

When the ZVS is tur­ned on, bri­ght blue dischar­ges sho­uld be visi­ble inside the bulb:

When a metal object is bro­u­ght close to the bulb, the dischar­ges converge toward it. This can be seen in the video below.

Have fun and learn a lot! :)

Have fun and learn a lot! :)

Fur­ther rea­dings

Marek Ples

Aa