Why ZVS always use regulated PSUs?

[AndreiRS]

I have seen a lot zvs circuits around. But it looks like everybody uses switched power supplies for it. Not sure why. Some months ago I tried to run one from a non regulated power supply but it just destroyed itself. I think the load voltage drop may had caused it to fail...? Currently I'm using a computer PSU. But before I blow 2 more mosfets. Do you think it would work with a rectified transformer and a biiiigggg capacitor? I think I have 140 or 200k uF around.

I have seen people driving them with car batteries and even cell phone batteries for small zvs. I think "slow" voltage drop is not the problem. But sudden drop and drop below gate voltage.

[Teravolt]

generally I think people use what they can find , I prefer just a simple raw dc like a transformer bridge rectifier and filter capacitor of about 30v @ 8 amps
for my ZVS because they are more robust

[RocketScienceSmurf]

I take the old school non-SMPS over SMPS any day of the week. I have destroyed a few ZVS induction heaters (not sure if that is what you menat by ZVS) but I can assure that it was NOT because of a weak power supply. Just as Teravolt wrote I think the reason people use them is that they are so common.

[AndreiRS]

Hmmmm that is really good to know, thank you both.

36 or 48V smps are hard to find here. I was thinking about using my rectified weld transformer to do some tests hahahah. 50V 100A easy. Now I just discovered I don't have those big capacitors anymore. But I can find more on junkyard.

Yes I meant the induction heaters with self ressonance. Simple circuit, but very good.

[Teravolt]

ZVS zero voltage switching or ZCS zero current switching. when you have a resonant circuit and you are switching it with a sqare wave ie a bridge for soft switcing the mosfets switch states toggel right when the resonant circuit also trasitions changes this is true for current also. I believe that with a parralell rsonant circuit like a mazilli flyback circuit is ZVS and also an iduction heater. and in a DRSSTC a series resonant primary circuit gets ZCS. when hard switching occurs the mosfets turn off or on at the wrong time not in sync with the resonant circuit cousing voltage to get real high if opened late or high currents if to late or early. this can couse the body diode that is a intrinsic part of the mosfet to fail. the mur diods are there to take any energy from the body diode and disapate it durring hard swithcing in the event that your resonant frequency changes and your bridge dose not. this is why closed frequency fead back systems ar easier on silicon. a good example is a Mazzillii or a DRSSTC or a well designed induction heater. A bad example would be a SSTC because the teslas resonant frequency can change with arcing or a induction heater with no feedback because the resonant frequency can change with load cousing hard switching

[RocketScienceSmurf]

Hmmmm that is really good to know, thank you both.

36 or 48V smps are hard to find here. I was thinking about using my rectified weld transformer to do some tests hahahah. 50V 100A easy. Now I just discovered I don't have those big capacitors anymore. But I can find more on junkyard.

Yes I meant the induction heaters with self ressonance. Simple circuit, but very good.

Welders work GREAT as power supplies: https://highvoltageforum.net/index.php?topic=32.0

[badpeter]

Funny that you asked...
I was actually been able to run a 1800 w unit off a ordinary transformer with a bridge and a cap.
It happened that I was not been able to find a regulated power supply, so I bought a EU/US "voltage converter" unit because it was cheap. I turned it into a regular non-auto transformer and calculated where I need to tap for the correct sec voltage (then i actually did it 3 times to relieve some of the current the sec be passing).

This setup worked for about 5 mins heating up a large graphite crucible until wires melted and the bridge burned out. (transformer stayed relatively cool). I did use inadequate cooling and wires so that was quite deserved. Since then, my electrician friend gave me some proper wires and when i gather all other parts, I will redo all this and try again. Hopefully I will actually be able to melt something!

The problem with transformer is that voltage seems to drop a lot when drawing current. I do not know why it happens to such a significant extent. The concept is called "voltage regulation" and my findings contradict what i read on the internet (namely, it shouldnt be more than 5% or so, and only when transformer is saturated - not the case here even a bit.)
Maybe there is an engineer who can shed light on what controls significant voltage sagging on a transformer when drawing current.
So yeah... hope to post some updates this week!

[AndreiRS]

Hmmmmm I will find some mur diodes to try to prevent those fails. Thanks Teravolt.

Oooof hahah what a massive power supply you have there kamelryttarn. I will read that link now, looks really interesting.

Hi badpeter. Interesting test. What kind of transformer are you using? How it was originally working?

[Mads Barnkob]

I have used it all.

Iron transformers of all kinds, rewinded microwave oven transformers, scavenged 24VDC AC transformers from industrial power supplies, halogen lamp power supplies and old welders, they are almost impossible to kill but they have a slight disadvantage when it comes to overload. A overloaded / saturated iron core will drop in voltage up until then point of saturation, which often can lead to MOSFETs dying from poor switching if the voltage drops below 10VDC.

SMPS are however becoming very easy to find surplus from old servers and when they are overloaded they make a over-current shutdown instead, which in my experience does not result in near as many MOSFET deaths as a underrated iron core can if a certain high load is suddenly presented to it.

Iron core only fill up its capacitors at 100 Hz vs. SMPS that maybe run at 100 kHz and has a extremely fast load condition monitoring to regulate the output voltage, so SMPS is doing better in my book.

[Teravolt]

whats your buget like and do you have any surplus places or junk yards? if you want something nice a big 24v-36v 60amp switcher supply would be nice but rude and crude like a filtered raw dc supply like a battery charger would work fine in my opinion. can you find some surplus stuff?

[badpeter]

AndreiRS - the name of the product is Simran (or sometimes Goldsource). Costed 90 cad for 3000w unit, no taxes and free shipping=)
Inside, there is a torus with 184 windings, so it is about 1 turn per volt. originally it was connected in autotransformer fashion 1:2, but you can separate prim and sec easily.  Also, it is easy to tap anywhere on it and get any voltage (basically re inventing a variac  of sorts here). the wire is aluminum though (what a piece of crap), you d have to take special measures to solder to that.

[AndreiRS]

Hmmm got it. Aroud here we can't really find anything usefull on junkyards, everything is recycled very fast.

My budget now is basically, what I have at home. But I have some good transformers without a secondary. I can wind some wire there.

The Simram voltage estabilizer, I only found on internet from other countries. But then you said it is like a variac. I can use one from a company here some 50 meters away. Then if it works I will buy one for me.

Thanks.

[Teravolt]

Do you have ant companies or universities or scrap yards that may let you take or get for cheap parts. alot of power electronics like UPS's have lots of usfull parts.