Induction heater by MOSFET IRFN 260

shortbus

Joined Sep 30, 2009
7,857
gold refined from e-waste after passing through chemical process .raw ,aterial will be like fines form
The size of the material has no effect it is the total weight, takes more to melt a pound than an ounce.

One more thing sir the transformer involved in this circuit is 12v 70 ampire as u told the current gradually increases but my case no latched on 6 amp it self
If this is to make money why not do it with the correct thing? Did you read all of the things the man in the link went through before he found one that worked? He blew up a few made like yours until he found what worked.
 

Thread Starter

D Ramesh

Joined Oct 6, 2019
13
Good evening sir , Ur right mainly IAM an electrical and electronics technical .inthe refining feild iam , only a beginer. instead of furncace , tried this easy to reach the goal which negligible fine powder .
Speaking on problem first I thanking u and max sir . Who told me the best links and ideas . Some were I sorted 50 percent .problem is VCC 12v the voltage drop did't measure. Now replaced by an inverter welding rectifier bridge.with tank circuit .Started working this time checked the VCC dc no load 1.76A
After 15mm x20mm ms tube spanner reaching 8amps VCC droped to 7 . 5 hence clear that dc voltage should be more. enough all amp value is onac secondary only sir. Once again thankyou to all also suggest me an rectifier suitably .to proceede and how much VCC I can feed to this circuit safer side .
 

shortbus

Joined Sep 30, 2009
7,857
how much VCC I can feed to this circuit safer side .
I didn't build it, how would I know? It all comes down to the inductors and how much current they can take. Also did you see in the link where he lowered the resonate frequency? Most of these kits or plans use a high 100+KHz( around 150KHz if I remember right) frequency. But when you want to melt or work with thicker metal you need to drop the frequency down to around 70 or 80 KHz.

The higher frequency doesn't penetrate as deeply into the metal like a lower one does, due to the "skin effect".
 

Thread Starter

D Ramesh

Joined Oct 6, 2019
13
I didn't build it, how would I know? It all comes down to the inductors and how much current they can take. Also did you see in the link where he lowered the resonate frequency? Most of these kits or plans use a high 100+KHz( around 150KHz if I remember right) frequency. But when you want to melt or work with thicker metal you need to drop the frequency down to around 70 or 80 KHz.

The higher frequency doesn't penetrate as deeply into the metal like a lower one does, due to the "skin effect".
Thankyou sir
 

panic mode

Joined Oct 10, 2011
1,854
i'm with Dodgydave on this. never heard of Mazzilli oscillator so i tried internet and.. apparently in its basic form it is quite famous - for blowing up transistors. :p
circuit linked in previous post is using RC based astable. that fixes issue with both transistors on at the same time but it is not guaranteed to reach resonance so tuneup and efficiency will likely be an issue. any kind of stability would always require some sort of interlock which DIYers like to skip for simplicity. so why not use classic Royer then?
 

shortbus

Joined Sep 30, 2009
7,857
never heard of Mazzilli oscillator so i tried internet and.. apparently in its basic form it is quite famous - for blowing up transistors
From what I have read it only blows mosfets when powered by switchers. And people don't allow the switcher to come to full voltage before turning on the ZVS. They wire the ZVS directly to the switcher output. If you must use a switcher you also need a relay or other type of switch between it and the ZVS, so the full voltage is there to allow the gates of the mosfets to do their work. If this isn't done, the circuit locks up and blows mosfets.

This guy claims to have blown up more of these ZVS modules trying to learn how to make them live than anyone, and goes on to explain how to make them live. https://www.spaco.org/Blacksmithing/ZVSInductionHeater/1000WattZVSInductionHeaterNotes.htm
 

shortbus

Joined Sep 30, 2009
7,857
any kind of stability would always require some sort of interlock which DIYers like to skip for simplicity.
The biggest thing DIYers do wrong in this is not buying or using the correct toroidal core. They take one from their junk box and wind some wire on it. They don't do any math to make the inductance or amperage of it correct.
 

shortbus

Joined Sep 30, 2009
7,857
that does not prevent both transistors turning on at the same time
All I know about it is what I've read. They don't only use the Mazzilli for induction heaters but also in the solid state Tesla coil power supplies and even some wireless charging applications. And when every thing is done with an eye to design they work.
 

panic mode

Joined Oct 10, 2011
1,854
that may be the clue... who is they? youtubers and banggood/aliexpress sellers or manufacturers of commercial induction heaters? i can't see single patent or scientific paper on it.

unlike Royer, this relies on component differences (large tolerances) rather than circuit design. in the past components had rather large tolerances but nowdays they are getting pretty close. also power levels that one expects from this are increasing. and it only takes one moment of instability to turn it to smoke.
 
Last edited:

panic mode

Joined Oct 10, 2011
1,854
thank you both, i spent enough of time on this trying to find anything that would change my mind. nothing found so for me case is closed.
 

shortbus

Joined Sep 30, 2009
7,857

shortbus

Joined Sep 30, 2009
7,857
thank you both, i spent enough of time on this trying to find anything that would change my mind. nothing found so for me case is closed.
Sorry to hear that. But they if well designed work, and a Royer if not well design won't work any better. The main reason the Mazzilli is used over the Royer is the work coil. The Mazzilli is a simple two ended coil, and the Royer uses a ceter tapped coil. Would this Mazzilli, ZVS be used in an industrial situation? Don't think so, but we're talking DIY, not industrial.
 

cmartinez

Joined Jan 17, 2007
7,064
C, have you ever joined one of those type of places? They want you to just "subscribe" but what happens after you do? A lot of the stuff I've been interested in is only available from those subscription places and have been afraid to try them.
Nope ... I've never joined them, and possibly for the same reasons you haven't either. But if I ever did, I'd do it using a disposable email address. Yahoo and Hotmail hand them out as part of their services. I don't know if they charge you anything for the membership, but if they did, I would never join them.

Why don't you try it first, and then let me know how it went? :D:p

EDIT: I just checked, and they charge $10 dlls/month! ... definitely won't be joining them anytime soon.
 

panic mode

Joined Oct 10, 2011
1,854
The main reason the Mazzilli is used over the Royer is the work coil. The Mazzilli is a simple two ended coil, and the Royer uses a center tapped coil.
no problem with load or output stage. centertap is not the only change, Mazzilli also adds capacitors so waveform is sinusoidal instead of square wave. all of that is fine....

the key issue and center of discussion in this thread is lack of feedback interlock. that means there is nothing to prevent deadlock scenarios that destroy the components. Royer version neatly solves with by simply crossing feedback. For some reason this is missing in most implementations of Mazzilli circuits that float the internet. This can't be right and it is the reason i wanted to see what Mazzilli really did or what circuits of commercial grade units look like. I bet they don't allow both transistors to turn on at the same time. Even sites that try to illustrate differences and operation of Mazzilli version usually start with version that does have an interlock and quietly move over to version that does not, while focusing on benefits of modified output only. For average reader this detail goes unnoticed and they would assume that change only has positive aspects. Then they attribute obviously way too frequent failures to particular implementation rather than looking for flaws in concept.
 
Last edited:
Top