Building a ZVS Driver (1500W+ @ 25kV+)

Subscribe & bell for more content like this!
Schematic was lost, but you can easily find them by googling. It's not hard. Change 470 ohm resistors to around 1000 ohms if you want to use higher input voltage. It's not that easy to get as high voltage as you want on ZVS drivers, they're pretty easy to kill. I highly recommend building an IGBT driver (half or full bridge) that way you're actually able to use as high voltage as you want, under about 1000V (on 1200V igbts) and under about 50A or so (on 200A+ igbts). I am currently recording the process of building my IGBT drivers, and once I finish, I will now edit the video and explain as much as I can by voiceovers, how I did it, how it works, how not to kill them, etc. So, to stay tune for that, subscribe and turn the bell on, and wait. It's been about 4 months already, as of December 2022. As of April 2022, almost 8 months.
Discord: / discord
Do not repeat anything I do. I am not responsible for any deaths or injuries.
#highvoltage #arcs #electricity #electrical #electric

[@mathysgobeil1592]

awesome job man. that’s probably the biggest arcs produced by a flyback i’ve ever seen, im impressed

[@trenthighvoltage]

flybacks are from CRTs, this is just a big custom HF transformer

[@kennyfucker1]

I put like 5 of these flybacks in series and gave them all a personal zvs driver. Arcs over 200kv where the result

[@mathysgobeil1592]

@@kennyfucker1 that sounds absolutely crazy man, you should definitely do a video of it

[@kennyfucker1]

@@mathysgobeil1592 i will upload it within an hour

[@kennyfucker1]

I couldnt find the 5 zvs video but i did find a video with 2 and those where already impressive

[@Electro-Dad86]

That's gotta be the best z. V. S driver's circuit. I've ever seen you by far. Have impressed me the most. Be careful, that looks insanely crazy. High voltage, hi, current electro science lives forever

[@shawncalderon4950]

“Okay, it's complete, it's nice and heavy” dude, you had me laughing in tears! I love this channel:)

[@trenthighvoltage]

😁

[@gg4ksa]

hats off for you my brother, respect for taking it to the next level, and respect for coming out alive after all !!

[@trenthighvoltage]

ohh pfft this is nothing, I pulled 150kV 8ft arcs 2 years later hehe

[@Livewire460]

Biggest zvs arcs I've ever seen!!

[@TheTemporalAnomaly]

Feeding the mosfets control circuit (gate drive), is a great idea. Much lower losses in the resistors. I have built circuits this way so as to be more efficient. I have also built circuits using sg3524 switched mode power supply controller, then into mosfet drivers (4427), then into the gates of the push pull stage. The resonant frequency has to be matched to the frequency of the sg3524 clock, but very high efficiency can be reached and as an added bonus the output voltage can be controlled by varying the mosfet on-time (pwm). I have yet to experiment to see if it is possible to have a free running circuit, like the one you show, but drive the mosfets with a gate driver IC so as to get faster switching time and more efficiency.

[@trenthighvoltage]

This is why I like igbt drivers better, it uses a simple TL494 driver (anything else works, even 555 timers would be better and works the same as tl494) and use igbts since theyre more robust than mosfets.

[@johnruscigno5738]

That was badass bruh. I think you had it just right at 60VAC. I love the adrenaline rush you get when pulling arcs that long.

[@trenthighvoltage]

i wanted to push more so bad, i ended up buying 600v mosfets and such but they didn't really work well. lol

[@samueldavies646]

​@@trenthighvoltagehow often do you kill the 200v MOSFETs?

[@trenthighvoltage]

@@samueldavies646 its been like 2 years now, i havent touched zvs drivers since, i now fk with igbt drivers

[@trenthighvoltage]

correction: 2:25 it was 50uH each 2 chokes, not actually 200. Not 40A either, it barely can do that. About 20A. Now i'm using 4 chokes in series, 1 3 5 taps (drain, vcc, drain) seems to be better. much less heat. maybe new better zvs driver vid soon. trying to improve my zvs drivers make it more powerful etc. ALSO: the 12v supply part, actually for some reason i'm able to use lower input voltage with a separate 12v supply, but more input voltage with a not separated 12v supply (that's being powered by the input) and i actually didn't have like MUCH knowledge in this video. But I kinda have more knowledge now. I may make a new video on building a zvs driver, much better and improved, with better explanations etc, but in the future and I actually haven't really researched much about zvs drivers, because I've been focused on my igbt driver(s) since like March i think, and started actually learning more about igbt drivers. currently working on mine, improving it etc, recording footages for like 3 months now, on and off. (100gb+ total, gonna take a while to edit with voiceovers to explain pretty much every aspect possible. Edit: as of June 2023, it's been nearly 10 months now. 200+ gb and 450+ files.)

[@christopherthumm4348]

Finally someone else who wants more power 👍👍👍

[@trenthighvoltage]

I try my best, so sick of seeing people building weak zvs drivers and pulling tiny arcs and call it "powerful" I couldnt get more powerful than this at the time since i dont have much knowledge and couldnt really find good powerful mosfets with much higher voltage

[@B.Cote39]

Crazy man !!! How did you learn all this jesus ??! Thx so much for sharing !!!

[@soofadivision]

Awesome Arcs…!

[@Thrustmaster64]

This might be the best thing I've ever seen.

[@trenthighvoltage]

glad you enjoy it :D

[@Raulley27]

Where do you get big flyback transformer ferrite cores?

[@Jonodrew1286]

My god amazing power and genius!!! Just one concern - NOx NaCl gases hope you have good ventilation going on 🙏🙏🙏

[@jungyeonho]

art of work

[@high1voltage1rules]

a pointer mate next time for your primary'' use a nail and a drill to wind your magnet wires together! you get a nice tight cable at the end

[@lj516]

Oh boy! That pop at the end was great but you got the failure mode wrong! If you watch the last frame before it dies, you will notice the arc path to soda changes from your pole to the legs of the mosfets. This probably vaporized the solder, then a few milliseconds pause, and magic smoke/sparks start after the initial flash. I love seeing interesting failure modes.

[@trenthighvoltage]

i didn't notice that lol

[@unknown-ql1fk]

Damn the low popping noises are just to die for....im in love. Do you have a video detailing the mods to the 1kv zvs board? I think i understand but im new to mods like that

[@trenthighvoltage]

yes that's the arc disconnecting then latching on again over and over again, i love that sound.

[@groundkopi5975]

You could've just used the "1000w" ZVS board and change all the components, and maybe reinforce the solder tracks with some thick wires. That's what I did and able to go up to 4kw running comfortably for a minute

[@trenthighvoltage]

maybe but i like building my own zvs drivers from scrap, it's just more fun that way. i bought some PCBs though so i can put smaller components on etc. also, just got a 1200v 200a igbt, going to build half bridge with it soon with some help from steve arcs. hopefully i'm gonna get some massive outputs hehe

[@groundkopi5975]

@@trenthighvoltage Oh mann you should make a video about it, maybe share the schematics too, I've been meaning to make halfbridge driver but always felt intimidated because i don't have a oscilloscope

[@trenthighvoltage]

@@groundkopi5975 im definitely going to don't worry.

[@Then3xtpart]

So what’s the input power to this?? Dope stuff though!!

[@kahe7436]

That's crazy ass dangerous

[@trenthighvoltage]

yes 👍

[@dalenassar9152]

At 6:18 how did you remove the insulation from the tiny wires to make the LITZ ends?? THANKS FOR THE GREAT VIDEOS!!!

[@trenthighvoltage]

just solder it and it'll melt the varnish, make sure it's hot enough

[@dexterconnolly3696]

Damm that's impressive

[@trenthighvoltage]

ty!

[@unknown-ql1fk]

Cool stuffs. I need a 3d printer for my bobbins. My HF secondary attempts have been limited to single layers insulated with PE sheets and multi layers but have huge capacitance and therefore less current

[@trenthighvoltage]

3d printing makes it a lot easier. it may take hours or even days depending on how big the bobbin (the 25mm core bobbin took like 15 hours i think or a day, i didn't speed up prints in cura that time) but it's always worth it because it's literally just perfect, you can even thicken the walls to withstand even higher voltage, design some supports around the core like i did, so much better.

[@dalenassar9152]

If your mosfets are 200v, then the failure at the end was probably because 70v x PI = 220v (greater than 200v!). I think I saw 68v on the display in the arc illumination, which would still generate 214v. From calculation, the absolute max. input should be 63v. I'm sure other factors come into play, but this was a FANTASTIC show!!!...best I've seen by far! =============================================== BTW: What gauge wire did you wind the secondary with?

[@trenthighvoltage]

yea 70vac, which would be 96 to 98vdc rectified. so that would make sense! i wish i could push more, i definitely would be getting bigger arcs. i'd need 500V mosfets, but it's whatever i have a half bridge now. (i should probably build a full bridge using smaller igbts with higher ratings, or just bricks... and i was using 0.3mm wire i think

[@dalenassar9152]

@@trenthighvoltage The FDL100N50F mosfet is rated 500V @ 100A. ...been looking for the answer to the following question for a long time.... ============================================ Is there any problem with using IGBT's with the standard mosfet ZVS circuit? ============================================= Shouldn't it be virtually a drop-in replacement (with some voltage ratings raised)? THANKS MUCH!!!

[@trenthighvoltage]

@@dalenassar9152 sorry did not see this comment, and if you still want to know, you can use igbts for zvs drivers, i believe they are more robust, since mosfets are more sensitive to voltage spikes and such. you are way better off building an igbt driver though, not zvs driver as they can easily latch up, you have little control over them. with igbt drivers, you have a LOT more control, adjustable frequency, deadtime/duty cycle (can be used to current limit things, very useful), and as high as at least 400-800V, depending on the voltage spikes, just use snubbers and tvs diodes to help suppress down voltage spikes a lot.

[@benjamineduardo8729]

Hello, impressive work! Could you tell me what liquid you use to submerge the high voltage winding? Or your business name? Thank you very much.

[@trenthighvoltage]

it's just mineral oil, you can use any oil that isn't conductive

[@Electronic4081]

Awesome arcs! Where did you find a huge transformer core like that?

[@trenthighvoltage]

from mads lundbaek, or kaizer electronics the youtuber. on facebook. you can find some on mouser i think, and maybe on ebay.

[@Doom2pro]

I have a 500W ZVS that is hotter and longer acrs than your first ZVS... somehow doubting that first ZVS is 1000W

[@trenthighvoltage]

definitely not 1000W. around 500W.

[@RAVI171175]

Man You've got to drive a Tesla coil with that & let us know.

[@trenthighvoltage]

I plan to build a DRSSTC some day.

[@dalenassar9152]

AMAZING VIDEO!!! I am in desperate need of a ZVS driver that is very reliable and can output about 600 to 800VAC peak at about 10A. I would GREATLY appreciate your recommendation. THANKS MUCH!

[@trenthighvoltage]

that would be at least 500+vdc input. you can't use that much on zvs drivers, as mosfets are usually rated for 200V, there are mosfets that's rated for 500V but not enough. i'd suggest using IGBTs with at least 1500V+ rating, or build an IGBT full bridge with a custom transformer that outputs that many voltage.

[@dalenassar9152]

@@trenthighvoltage I may be very wrong here, but if Vpeak= Vin x PI, then wouldn't that make the input be 800v/PI which is about 250v+vdc (instead of 500v+vdc). I guess I am thinking of each mosfet taking only one Vpeak polarity since the source of each mosfet is grounded???

[@trenthighvoltage]

I guess so, for me i seemed to have about 150-200V ac on the output from about 100VDC input. maybe just weird resonant idk. depending on what power you wanna achieve, i strongly recommend building your own igbt driver. it doesn't have to be over $200.

[@dalenassar9152]

At 0:07 is that a burned 470 ohm 2 watt resistor?? I always 5 Watt in this circuit since the pair of gate resistors must also limit the ZVS and UF diode currents.

[@trenthighvoltage]

yep 👍

[@jovannyislassantos6941]

Hey man try corona discharge With ionic trusther

[@trenthighvoltage]

Maybe one day lol

[@jovannyislassantos6941]

@@trenthighvoltage puedes recibir una idea mía es algo super heavy de un proyecto de electricidad y magnetismo Sobre un generador de electricidad como puedo mandarte más electricidad

[@mikhailduplessis1652]

hi. I'm in the process of building a hv supply using the zvs circuit as well. just wanted to know what your inductance value was for the coils1 mate

[@trenthighvoltage]

100 is optimal, more the better (up to 150-200uH) if you mean the chokes.

[@ChickenPermissionOG]

Use chisel tips they sink the heat better

[@trenthighvoltage]

wdym

[@ChickenPermissionOG]

@@trenthighvoltage They make better solder joints you get less cold joints especially if you have a large ground plane.

[@janradlinski3787]

I had the same aliexpress zvs driver like in the beginning. I hooked it up to the ordinary flyback to see how it is running. It was switched on for a minute and i made 2 - 3 sparks and suddenly it broke (short detection on PSU). Any idea why that happened?

[@trenthighvoltage]

Idk. Fake maybe. Make your own with genuine fets. I fw igbt drivers now.

[@simonm30]

Did your supply have a lot of voltage ripple?

[@hamitinan6422]

soğutma yağı tutuşursa ve odanda 5kg yangın söndürücü yoksa seni kimse kurtaramaz :)

[@trenthighvoltage]

it's mineral oil... it's not flammable. lol. there was flashovers in the oil, nothing happened and you still say that XD oh and i have a fire extinguisher.

[@hamitinan6422]

@@trenthighvoltage ısındığında yanacaktır. Dikkat et. Çünkü bu yüzden benim evim yandı:)

[@trenthighvoltage]

@@hamitinan6422 it's just food grade mineral oil lol

[@Freddyfanmulti3214]

how many primary and secondary turns, how do i build

[@trenthighvoltage]

it all depends on how you want to design it. it's easy.

[@DickeyHorace]

Hernandez Lisa Hernandez Shirley Gonzalez Sandra

[@newmail3549]

what should i do for super fast zvs, i mean resonance frequency about 500kHz

[@trenthighvoltage]

that may not be possible, your mosfets could easily lock up. consider using an advanced gate driver, using TL494 oscillator and some half/full bridge using igbts or mosfets. you can also adjust frequency and duty cycle.

[@Greenorange0]

Bro can you just tell me a how to make a best powerfull transformer for zvs what type of ferrite core do you and how to buy them

[@trenthighvoltage]

i'm not the guy to answer those questions. i am still learning. zvs drivers really aren't the best choice for high power, like over 2000W or so. If you want a much more powerful driver and transformer, build your own igbt driver (half bridge/fullbridge, video on that on how i build it etc soon!), and use a good ferrite core with lower inductance per one turn as much as possible, also bigger which can be pricey. my ferrite cores are only good for a few kW. I also plan to make a video on making more powerful transformers, in the future.

[@KmgdsHfafjp]

Hernandez Scott Perez Sarah Miller Kenneth

[@buras.]

❤👍👍👍👍

[@mohammedjawadforelectronic9190]

when you connected two of inducters in parallel the results will be 50uH and not 200uH

[@trenthighvoltage]

yes i realized that

[@therealprofessor976]

Amazing! What are the specs of the transformer?

[@trenthighvoltage]

idk

[@TanyaHanson-t8q]

Garcia David Williams Angela Hernandez James

[@yankozlatanov]

The smell of ozone must be unbearable after those arcs lol.

[@trenthighvoltage]

not really

[@jenz-qw9td]

give us a scimatic

[@dalenassar9152]

Is it OK to run the ZVS driver with no load?????

[@trenthighvoltage]

it's recommended not to, it uses a tank oscillating circuit, primary is inductance which also changes frequency. without it it may not work properly or switch too fast and lock up

[@hpsmith2010]

What kind of core is that (the big blue and black one), and does it have an air gap? If so how long is the gap?

[@trenthighvoltage]

big blue and black one? wdym? i'm using 3c80 material 100x25x25 ferrite cores, for the bigger transformer. apparently can handle in a few kW range, so maybe 2-3kW? Idk. the smaller U cores is just from an old CRT flyback. And I don't remember if I had the air gap, if I did then must've been like somewhere under 1 mm? 0.2mm or so not sure. I don't think it matters too much, only for saturation maybe.

[@hpsmith2010]

@@trenthighvoltage The transformer with the blue and black core first appears in the background sitting on the table at about 2:33 in the video. It looks like a ferrite core, part of which has been painted blue. Then you start using the blue and black core with a single layer secondary at about 3:05 in the video, and thereafter with the other secondary coils. So you glued a bunch of 100x25x25 "I" pieces together? Anyway, an air gap in the magnetic path of the transformer would apparently do two things: (1) help prevent core saturation (if you were driving the transformer hard); and (2) increase the "Q" of the circuit (as I understand it, the circuit may stop oscillating if the loaded Q drops below 1). I would guess that if you don't remember deliberately putting some spacers between the core pieces you probably didn't gap the core.

[@trenthighvoltage]

@@hpsmith2010 that's masking tape, not blue paint, and at the end i'm pretty sure i glued 3 of them together except for 1 so i can take the coils out etc. what's Q?

[@hpsmith2010]

@@trenthighvoltage As I see it, when you drive a HV transformer with a Mazzilli ZVS driver, you've got a parallel resonant circuit. You have L in parallel with C in parallel with the load R, where the L is the magnetizing inductance of the transformer, C is the value of the capacitance across the transformer and R is the effective load resistance due to the arc or whatever. So at the resonance frequency f (according to one mathematical expression), Q is the ratio of the magnitude of the current in either the L or C branch (i.e. the circulating current at resonance) over the magnitude of the load current: Q = R/(2*Pi*f*L) or Q = 2*Pi*f*C*R. I haven't verified it for myself but someone said that if Q < 1 (perhaps when heavily loaded?), the Mazzilli ZVS driver circuit will stop oscillating. So if this is true and it actually happens in practice then in theory you could reduce L by gapping the core and increase C to maintain the same f while increasing Q. I'm just thinking out loud here as I haven't done any experimenting with this yet. www.electronics-tutorials.ws/accircuits/parallel-resonance.html

[@leonardbrowneda5461]

Anderson Nancy Johnson Michael Harris Jose

[@HVMaster02]

А как так получается гудящая дуга ?

[@Clare-n8y]

Thomas Ruth Allen Timothy Martin Jason

[@patrickedgarjr9186]

Hey any way u can do a more detailed schematic and caponates list

[@trenthighvoltage]

It's a very simple driver, just look up schematics on Google. Replace the 470 ohm resistors with 1000, maybe more than 5 watts. That's it

[@patrickedgarjr9186]

ok i have tried all sorts of different configurations with some working and some not i will try what u said see what happens ,also if you need any caponites let me know i have alot of everything

[@patrickedgarjr9186]

also have you tried that driver with a tesla coil

[@trenthighvoltage]

@@patrickedgarjr9186 no

[@PilotZoomer]

Cool

[@trenthighvoltage]

😁👍

[@hardnachopuppy]

Can you run a magnetron using this?

[@trenthighvoltage]

probably if you use fast diodes. i do not recommend playing with magnetrons.

[@JimAlvis-m8u]

Moore Michelle Gonzalez Mark Garcia Kenneth

[@shawncalderon4950]

Dude, how are you winding your hv transformers? More specifics please!

[@trenthighvoltage]

it's pretty easy. 3d printing or lathe/metal lathe with plastic materials is your best friend

[@shawncalderon4950]

@@trenthighvoltage yes, I can see your nice printer, how many turns for the primary and secondary. Is the litz wire from the primary? What kind of copper wire, what size is your bobbin, what frequency, sorry, I am just so interested in your projects!

[@zulfiqarhaidari8876]

Hey boy where is the circuit diagram give me the diagram man

[@trenthighvoltage]

calm down, don't be greedy. i lost it, i haven't built a zvs driver in a half year now. just look up schematics for those on google and replace these pulldown resistors (470 ohm) to 1000ohms and 200V 50A+ mosfets if you plan to use up to 80vdc 20a+ max. i recommend irfp90n20ds. or more than 200v which is always better.

[@timtrail1845]

That all 🤷🏾‍♀️

[@trenthighvoltage]

?

[@sciencefusion5352]

Can you provide me the schematic

[@trenthighvoltage]

u can just look them up theyre all the same, instead i replace 470 ohm with like 1000ohm for higher input voltage

[@sciencefusion5352]

IRFP 250 N OR IRFP 260 N MOSFET ???

[@sciencefusion5352]

Which one is better

[@trenthighvoltage]

​@@sciencefusion5352 IRFP250N = 200V 30A IRFP260N = 200V 50A What do you think?

[@sciencefusion5352]

@@trenthighvoltage 260 N 😂😂😂 And can i use these semiconductors in parallel for more current??? Or should i use and igbt module ( if you suggest please mention igbt number )

[@JulieCafdsh]

Martinez Elizabeth Martin Christopher Johnson Nancy

[@PabloTRUE-v8d]

Johnson Brenda Walker Eric Young Mark

[@NxndbhduKnbsuxdh]

Martinez Kevin Hall Larry Gonzalez Susan

[@Luiz_inventor]

Did you try using crt flyback's ?

[@trenthighvoltage]

no they're weak and rated for like 200-300w

[@Doom2pro]

@@trenthighvoltage My 500W CRT flyback is pumping white hot arcs longer and faster than your first "1000W" setup... Dodgy...

[@davestorm6718]

@@Doom2pro What model flyback are you using? I had a weak one - one or more of the internal diodes burned out so will have to manually wind one up (ugh).

[@Doom2pro]

@@davestorm6718 I have several, from TVs, Monitors, and from eBay, only thing they have in common is gap between core and flyback so I can ram 5+5 turns of automotive wire around them. They all produce the same.

[@trenthighvoltage]

@@Doom2pro I never really said it was around 1000W. It's actually about 500W at most. I couldn't really increase the voltage more due to some components limiting it.

[@StevenFellows-n8w]

Anderson Steven Smith Laura Anderson Timothy

[@LillieAlvarez-x4h]

Hall Larry Young Melissa Clark Kevin

[@navarajaraja7889]

Why you not mention resistance number?

[@trenthighvoltage]

uh wym?

[@LillianJackson-z1q]

Lee Edward Hall Dorothy Martinez Angela

[@melmentereaudley6546]

Thompson Donna Thompson Sharon Rodriguez Mark

[@LorenaGraves-j7b]

Davis Larry Allen Paul Young Eric

[@FredCamille-i1b]

Walker John Thompson Kenneth Wilson Donald

[@williamralphs270]

Brown Linda Williams John Garcia Donald

[@NoraBerg-g8p]

Thomas Edward Gonzalez Dorothy Hall Nancy

[@williamralphs270]

Martin Deborah Rodriguez George Jackson Dorothy

[@EatuopLyreqzgj-f5l]

Harris Larry Miller Brian Martin Deborah

[@oussamathedjfreeman]

3d printer sound at background hhh

[@trenthighvoltage]

lol

[@axelbelleris]

It's mute. 🔇

[@trenthighvoltage]

it is not lmao

[@Lesa-j7q]

Taylor Donna Brown Nancy Young Cynthia

[@BrettMeadows-y3z]

Wilson William Miller Jeffrey Williams Dorothy

[@ььььььььььььь-ц2ы]

Bizze maestro. I'm interested. ZVS dreever.

[@TehnoMan77]

Если каждый маэстро так будет экспериментировать, людям скоро будет нечем дышать, весь кислород сжигаешь. Думай о будущем поколении чувак.

[@trenthighvoltage]

😂😂

[@billgeitdeandada3588]

Wilson Robert Harris Jennifer Brown Jennifer

[@trenthighvoltage]

Just noticed I got 120+ subscribers just from this video. That's insane! Considering I only got 13K views and 380 likes.

[@GardenChasel-z1l]

Rodriguez Larry Thomas Sarah Robinson Sharon

[@sheiladonaldson3143]

Johnson Carol Walker Jose Gonzalez Kenneth

[@DaisyHayes-n2k]

Jones Patricia Hernandez Elizabeth Jackson Gary

[@ClaraIda-x1e]

Hernandez Thomas Williams Thomas Wilson Gary

[@VictorChapman-b1s]

Lopez Carol White Jessica Martinez Laura

[@Fadara-t8x]

Rodriguez Daniel White Eric Thompson Richard

[@MrKor4k]

is that safe, what if it's accidentaly shock you? what is the current?

[@trenthighvoltage]

It's never safe but it's pretty high freq, so you won't really feel a shock I don't think. But can damage your nerves if you touch it more often and burn your skin. It's never always safe

[@SeokSbk]

Robinson Anna Harris Shirley Anderson Jose

[@ChireDeville-mh4cu]

Garcia George Martinez David Taylor Eric

[@MoniNaha-b2x]

Harris Eric Garcia Donald Lopez Mary