Re: Royer induction heater
jnbrex, Sun Aug 14 2011, 02:10AM

That thing is awesome! Some schematics would be great too. That is a very interesting project, but I think it looks a little too complicated for my novice soldering skills...

On another note, does the copper pipe that is the heater get hot? I thought that induction heaters worked through eddy currents produced on the object to be heated and didn't actually heat up themselves.

Re: Royer induction heater
radhoo, Sun Aug 14 2011, 07:43AM

looks and works great!

a few months ago I built something similar, schematics and photos here:

Re: Royer induction heater
Conundrum, Sun Aug 14 2011, 12:46PM

Its a shame you can't buy cheap magnetic solder (yet!)
Something like this would be great for emergency repairs etc and for people who aren't allowed to use a conventional soldering iron for safety reasons.

-A

Re: Royer induction heater
Gabriel35, Wed Aug 17 2011, 02:33AM

Incredible!
How do you got so much power from a simple Royer oscillator? I'm very very interested on the schematic *-*
I got some nice IGBT's and 30vDC Supplys here, and I can't wait to try this scheme out!
Can IGBT's be used instead of mosfets?
I have here a pair of IRFPG40 4.3A @ 1000v
and a Pair of IRFZ48N 64A @ 55v

Can some of them be used?
It's really the same mazilli's driver only with two inductors and strong caps?

Thank you!!

Re: Royer induction heater
Dr. ISOTOP, Wed Aug 17 2011, 03:13AM

Induction heaters will melt normal solder. It's a common misconception that they only heat ferrous objects.

Re: Royer induction heater
Gabriel35, Wed Aug 17 2011, 12:52PM

Can you post some photos of your circuit too?

Thank you!
I can't wait to buy some mosfets and try it out!

Re: Royer induction heater
Marko, Wed Aug 17 2011, 01:12PM


Yes it is, and if you wanted you could even center-tap the coil and use a single inductor if you wished. Another advantage of this would be that you could drive the water supply through both ends of the coil and having it exit through the center tap, removing the temperature difference that would normally be experienced between mosfets cooled with same water flow. But this didn't really matter at this power level and used only a tiny fraction of possible flow rate, so you can do as you wish I guess.


Regarding switches, I would still best recommend IRFP260's and a rewound MOT for the power supply.


Those IRFG40 high voltage mosfets are pretty useless for this circuit, you want low voltage mosfets with ON resistance of 50m ohms or less which is important for keeping the circuit stable. IGBT's, especially 1200V ones have high voltage drop which may again result in the circuit ceasing to oscillate and shorting out the power supply. You can try them but at your own risk, and they will probably not be as efficient as mosfets at such low voltages.

IRFZ44/48 or whatever are another extreme, with them you would be limited to like a 12-15V supply (car battery)? providing tens of amps, and you would have to put a step up transformer between your driver and the tank circuit, or use a 4x larger tank cap than I did which would be impractical.

The radhoo's schematic which uses the transformer is wrong, the tank cap should go after the transformer and not before it by the way.

Re: Royer induction heater
Gabriel35, Wed Aug 17 2011, 02:36PM

Ok!
So the secret of the huge power of the classic Royer Oscillator isn't the two inductors?

I have here a bank of 32x 0.068uF @ 630v - P612 w722 EPCOS Model (P6XX MKP capacitor)
Resulting a capacitance of 2.176uF. I can buy those IRFP260's and try it out!
I've experienced with Royer Oscillator a long time ago using two IRFP250 and a 36v Supply. But it was so weak, It was just capable of making a screwdriver redhot. I think that I still don't know what is the BIG difference between this circuit and the classic ones, like the one I've tryed... It is pretty incredible the power amout that you achieved on that video =O
What about the number of turns and diameter of work coil? are them important variables?
About the water? It's just to maintain the whole thing as cool as possible or it affects the Pratical Power of the circuit?

The capacitors? Can i use the ones mentioned? Do I need more capacitance?
I have a transformer here that is 24v. Rectified and with 8x 470uF capacitors it results in 36vDC, this tranformer is capable of 10~15amps. Can I use it?

About the Zeners, I plan to use 1N5349B 5w.
And the Fast Rectifiers, BYT11-1000
The 470 Ohm 5w Resistors, OK
And the 10K 0.5w too. =D

EDIT: I've found some IRF540N's here...VDSS = 100V RDS(on) = 44mΩ ID = 33A... Can I use them with 36vDC?

Thank you, and nice pics too!

Re: Royer induction heater
Forty, Wed Aug 17 2011, 04:00PM

ah the pictures certainly help. I can close the paused youtube video tab now lol. I wonder how the dual inductor scheme would work with driving a flyback.
I imagine you formed your work coil around a pipe. Did you fill it with sand or anneal it to help with even bending? I've got the tubing sitting here, I just don't want to screw it all up by bending it improperly.
what do you guys think about fairchild's fdp2552
for zvs circuits like this? or should I finally give in and buy the proper irf's?

Re: Royer induction heater
Inducktion, Wed Aug 17 2011, 05:17PM

When I made my induction heater's coil, I filled it up with sand to make sure it didn't bend. Make sure you have it full, with no pockets otherwise you'll end up ruining the pipe. (I have a few kinks in mine because of that, but it still works fine)

and, the dual inductor thing probably wouldn't help with a flyback, because the windings on the core are technically inductors/an inductor as well.

Re: Royer induction heater
Forty, Wed Aug 17 2011, 07:02PM

i wonder if filling it with water and freezing it would work the same as the sand (that's what they do for brass musical instruments, thanks "how it's made")
I didn't mean to improve the driving of a flyback, i just wondered if it would work equally as well as a single inductor. with less turns actually required on the core, I could use thicker wire with some of my smaller transformers.

Re: Royer induction heater
Marko, Wed Aug 17 2011, 11:58PM

Hy everyone,

Gabriel: It's important to use a large tank capacitance, 4uF in my case, if you keep 0.68uF or whatever was used for a mazzili driver the power output will likely be pathetic. Using higher supply voltage as well as a step-up isolation transformer also helps. On the other hand it's completely the same whether you use one DC link inductor or two!

Note that in this circuit the device has to be rated at least pi*supply voltage, hence 30V is pretty much the macimum for a 100V mosfet. Your supply voltage will probably sag under 30V even without workpiece, but I would use a variac anyway to bring the voltage up if it sags too much when you put the workpiece in. You would need a huge transformer with good regulation if you want to go without variac and drive the circuit to full power.

Also if you have only a 2uF tank cap, your power throughput might be somewhat disappointing if you don't get 200V mosfets. But your current ones should work for the first try.

Forty,Inducktion: Yes, FDB2552 looks great if you have some, although a 200V mosfet would be even better as I think.

I filled my coil with table salt before winding it, and blown it out (with some difficulty) with a compressor later. I didn't think of freezing the coil with water in it, it looks like even better idea actually.

Marko

Re: Royer induction heater
Adam Munich, Thu Aug 18 2011, 12:27AM


Funny, I've built that same circuit in the past as an experiment. It worked, but my mosfets kept exploding so I didn't bother to post it.

Maybe I just used crappy mosfets, but good job for getting it working marko!

Re: Royer induction heater
Dr. ISOTOP, Thu Aug 18 2011, 12:56AM

The giant Eurofarad Snubbers on Ebay are excellent tank caps; I've run mine to 4KVA with only mild heating (cooled through the two large metal terminals).
This has got to be the most beautiful Royer project I've seen

Re: Royer induction heater
Marko, Thu Aug 18 2011, 02:00AM

Hey bwang, those caps you discovered do look superb for induction heaters, so depressing they don't ship them out of US. I see some of them have large flat terminals, are those designed to be conduction cooled? And by 4kVA I assume you mean apparent input power, not the reactive power handled by the cap

WIMA caps do surprisingly well though, I was very surprised to calculate that current in my work coil with no load is over 100A, all of which is handled by just 16 small paralleled caps! It's surprising they are surviving it at all without bulging up or showing decreased capacitance.


Grenadier, your circuit might have been unstable if you used high RDS-on mosfets like IRFP450's. That's the main reason why I enforced low voltages here.

marko

Re: Royer induction heater
radhoo, Thu Aug 18 2011, 03:48PM

I think you got it a little wrong. You can't move the capacitor as the oscillator will not work.
The purpose of the ferrite core is as insulation transformer. It reduces stress on transistors and capacitor.

This approach can be used to power the original winding of flybacks as well, since we know that a concentric primary-secondary topology will put out more power.

Your current design is identical to my second variant, without the insulation transformer, just that you increased the power ratings with the bigger tank capacitor and work coil. Others willing to try this might want to look for mosfets with lower Rds instead, since keeping mosfets cool in this circuit is essential.

Re: Royer induction heater
Marko, Thu Aug 18 2011, 10:40PM

I think you got it a little wrong. You can't move the capacitor as the oscillator will not work.
The purpose of the ferrite core is as insulation transformer. It reduces stress on transistors and capacitor.

Hi radhoo,

the simple answer is, why would you want to have your transformer suffer all the reactive power in the tank circuit when you can have it handle only the real power instead? If I was to place a transformer in my circuit this way, it would need to handle the several tens of kVa present there, requiring a huge ferrite pole pig.
On the other hand having the transformer before the tank circuit it would only need to handle 600 watts of real power that are really going in, which could be done on a flyback core or similar.


I've used these transformers all the time, but on low power levels though (<100W). A 1:1 transformer produces absolutely no observable difference in operation, other than the output being isolated from the input. You can center tap the primary and use a single inductor instead of two and everything stays the same again!

If you use a step-up transformer, it will make the impedance of your tank circuit appear lower to the driver section, divided by square of sec/pri ratio: in other words, the oscillator will feel like you're using more capacitance and less inductance, with everything else staying the same.


I would still take care to minimize the leakage inductance on the transformer as much as possible, which may destabilize the oscillator but simply overloading it could do the same. If I slapped a 1:2 transformer to my circuit I'm pretty certain it would explode, but 1:1 should make no difference at all.

I actually wanted to try it out on this IH, but am unsure how to mount it to the existing configuration of the circuit, would have to isolate the mosfets from the work coil which is some trouble to do now :(

Still if you're going to use a 50Hz transformer like rewound MOT for a power supply it might make no sense to use an additional ferrite transformer - unless you have only very low voltage supply and low voltage mosfets like IRFZ44's and you want to push a lot of power.

Marko

Re: Royer induction heater
Gabriel35, Fri Aug 19 2011, 01:55PM

Hey Marko!
I'm experiencing with the circuit again.
I got some cheap as hell IRF540N's just to try out and but how you said, 36v seems to be so much for them to keep.
I'm waiting for my new IRFP260N's to arrive.
I have 8x 1uF 250 X2 Category Capacitors here.
can I use a capacitance of 8uF or something like that? How does it behave on the circuit?

Can you post a video of your oscillator melting a screw? and some photos of your Power Supply stage?

Can I use 13v Zeners instead of 12v ones?
Thank you!! =D

Re: Royer induction heater
Forty, Fri Aug 19 2011, 03:46PM

the 13v zeners should be just fine. I think his power supply is just a MOT with a rewound secondary and then a big rectifier and some smoothing lytics (unless that's what you meant that you wanted to see)

Re: Royer induction heater
Marko, Fri Aug 19 2011, 04:52PM

Hi guys,

In the video I'm actually using two UPS transformers wired in reverse, in series - rectified and filtered with a bunch of caps. I have a MOT too but I don't want to hammer out it's secondary and I'm actually building a transformer now from some huge cores I had around from old dead OBIT's.

I suggested a MOT as an easy solution for a power supply for soemone who has acces to lots of MOT's.

Gabriel: I don't know what type your caps are so I can't tell whether they'll be good or not, but also their capacitance seems a bit too high individually, unless they can handle 20-30amps per cap I wouldn't use them. If they are polyester, which they most likely are, the circuit might not work at all or they may just burn out.

Marko

Re: Royer induction heater
Goodchild, Fri Aug 19 2011, 07:27PM

Beautiful construction, all your compact projects are great!

May I ask what the overall voltage and RMS current rating is of the MMC bank you are running? It must be rather high by the amount of power you are putting into it!

Re: Royer induction heater
Nah, Fri Aug 19 2011, 08:39PM

I knew it was a good idea to get all those WIMA caps!

Also, would it be ok to use a back of caps with different values?

Thank you

Paul

Re: Royer induction heater
Forty, Fri Aug 19 2011, 09:22PM

oh whoops. i thought by "I don't have a proper power source though, I feel sorry killing my only MOT to rewind it." you meant that you already had killed your mot and currently feel bad. sorry for responding incorrectly on your behalf.

i wonder too about using a bank of mismatched capacitors (i've got lots of salvaged mkp x2 box caps.) I suppose it would be fun to try it out and possibly explode a few.
anyone ever used an ac oil filled cap (like a MOC) in a high current tank circuit? i've got 3 10.5uf 520vac oil filled caps from a large ballast and no idea what to do with them.

Re: Royer induction heater
Gabriel35, Mon Aug 22 2011, 01:42AM


I Have the same doubt. about different cap values...

Thank you!

Re: Royer induction heater
Dr. ISOTOP, Mon Aug 22 2011, 02:18AM

Different-valued caps will share current unequally, leading to possible failure of some of the capacitors.
Don't use oil/paper/mylar caps in a tank circuit. You'll be real sad when it fails and spits boiling oil at you.

Re: Royer induction heater
Marko, Mon Aug 22 2011, 03:17AM

hi guys,

If you have the caps from same manufacturer, of same technology and voltage rating but different capacitance I think they should do reasonably well when paralleled. Caps will form a current divider and all of them will share portions of current proportional to their capacitance.

The downside is that not all of the caps may be able to take the same power dissipation punishment due to surface area of their package. Due to square/cube law a lot of small, low-value caps are of benefit here,

So if you have caps that are like 0.27uF and 0.15uF mixing them should work fine, but you wouldn't be able to run 0.15's to their maximum without cooking the other ones. So I say mix them and have fun - just watch them not to overheat. The general rule for polypropylene caps seems to be, if they're too hot to touch, they're too hot, as I found out the hard way.

Goodchild:

I'm not sure what exactly is the current rating of the caps, but dividing the tank voltage with it's characteristic impedance yielded a current value of as much as 150A - and without water cooling the coil gets sizzling hot in seconds. I'm very surprised how well are the caps holding without blowing up so far.

They are 160V DC caps and I'm running them up to 70V AC,which seems to be about the limit considering they get quite hot.

Cheers,

Marko

Re: Royer induction heater
Gabriel35, Mon Aug 22 2011, 02:58PM

Can you post a video of your oscillator melting a screw?
Or Something like that Marko?
It's so nice to see things melting and burning with it, and quite impressive too.

Hey guys! I'm back, and got my new IRFP's260N and everything is working nicely, but now i have one problem... Like Marko said, when the workpiece is introduced on the work coil, the voltage drops a lot, in my case, I got a transformer very simmilar to the one that appears on Marko's video. It's rated 24V AC, Rectified it becomes 36vDC, when i turn the circuit on, it drops to 31, and with work, it drops to 24vDC...
I've made some calculations here, I got 11.8A as maximum current draw, and when the workpiece meets curie point it starts droping and keeps at 9A, so, my power source drops down to 24vDC, it gives me a real power of ~276W and aproximately 230W when curie point is reached...
With only 276W i alredy got the screw to become RedHot and I'm pretty excited to get more power with it, I think something around 1200W

As for the supplys, whats the advantage on using a separate 12v supply for the mosfets gates and a higher voltage for the power stage?

Someone knows a way to avoid that and make the thing work at full power?

As Marko noted, I can go up to 60v, because I'm using IRFP260N as switches.

Thank you!

Re: Royer induction heater
Arcstarter, Tue Aug 23 2011, 12:49AM

To get more overall wattage, you can lower the inductance/resistance of the work coil for lower impedance which raises the current, or use more voltage which would also raise the current.

The 12v supply for the mosfet gates is a good idea. I doubt i will be making any more ZVSs without that kind of drive. With a 12v supply, the gate resistors have to drop less voltage, and thusly heat less. This means more efficiency and smaller resistors, i used a single 1/4 watt resistor for each gate and they remained well within their resistance tolerance range. However, i would probably use less than 500 ohms, because as the voltage drops, so does the speed the gates charge.

A big transformer is needed for the wattage you want, and even a MOT wound with too thin wire could be 50v output with no load and drop to 25v with 10 amp load. You can use ohm's law to calculate the amount of voltage that will be dropped across the resistance of the secondary, if you know the resistance of the secondary.

Re: Royer induction heater
Gabriel35, Tue Aug 23 2011, 02:05AM

And about the two inductors noted 45~200uH , if I lower their inductance, will the current raise?

Re: Royer induction heater
Inducktion, Tue Aug 23 2011, 02:38AM

Your frequency may change...

Re: Royer induction heater
Gabriel35, Tue Aug 23 2011, 04:02PM

Hey guys I'm back with some news...
I got a transformer that gives me 30vAC, properly rectified and filtered it results in 43vDC.
I've tested the circuit with it and the power of it raised a little. I've measured the current and it was at 12,8A at peak. But the voltage drop was giant, When the current was 12.8 the voltage was 28,00v
so it gave me aproximately 360w of power...

So the question is: what do I need to change on the circuit to get more current flowing?

The parts are:
IRFP260N
13v Zeners
BYV26E Ultra fast rectifiers Maximum trr of 75ns

The tank capacitor consists of 27x 0.068uF MKP + 3x 0.47uF + 1x 0.22uF + 2x 0.1uF , resulting in:
3.666uF (all MKP)

Thanks!

Re: Royer induction heater
Marko, Tue Aug 23 2011, 04:55PM

Hi gabriel,

well most importantly, what is the power rating of your transformer? And if you don't know it, what is the approximate area of it's core? It might just be too small for what you want, and buying a new transformer of the required power would be very expensive, hence I recommended a rewound MOT. The little transformers I used didn't have much more guts than yours did, I could get full 30V only when I cranked the variac to 11.

If you're sure your circuit is working well you could even consider using a welding supply if you have one, on lowest voltage setting you can get from it.

Marko

Re: Royer induction heater
Gabriel35, Tue Aug 23 2011, 05:02PM

Hey Marko! Thank you for the fast reply!
I don't have a welding supply here =//
The area of the transoformer is aproximately:
11.4cm X 10cm X 6,8cm
It's pretty huge but i Know that it's Power rating is low, because it is a multi tapped transformer...
So, do you thing that whats limiting the current flow of the circuit is the transformer only?
changing the inductors inductance and the work coil's inductance are good ideas? or not?
What about 3,666uF it's good?

A good ideia is getting a transformer capable of 36v and about 22A of current isn't it? or a variac... Or a MOT, but i've never experienced with it, and I don't know how to rewind it...

My rectifier bridge is rated 50A.


Thanks!

Re: Royer induction heater
Forty, Wed Aug 24 2011, 12:55AM

for mot rewinding info.
variacs for that kind of current are pretty expensive. audio amplifiers and subwoofer supplies usually have big transformers too (as long as they're not switch mode supplies)

Re: Royer induction heater
Gabriel35, Wed Aug 24 2011, 02:39AM

Thank you for the information on MOTs.
I have one doubt, help me understanding the following:
For example, I Get 2 transformers, both rated 36v output, one of them is capable of 300VA and the other 700VA...
That means that the 300VA one is capable of 8,33A and the 700VA one is capable of 19.44A
So, the transformer limits the current drain of the whole circuit, or the circuit will take as many amps as It deserve? In other words... Can a transformer be overloaded and start to heat due to high current consumption from the circuit? or it will only work within it's limits, giving only the current that it's projected to give?

Thanks!

Re: Royer induction heater
Dr. ISOTOP, Wed Aug 24 2011, 06:05AM

A transformer will work beyond its current ratings for short duty cycle.
It will _not_ work beyond its voltage ratings (it'll saturate )

Re: Royer induction heater
Marko, Wed Aug 24 2011, 01:09PM

Hi guys,

gabriel, I'm sorry that I didn't clarify that it's cross-sectional area of the core that is required, or whatever it's called (width*length of the centre leg).

Anyway if that's your 700VA transformer it looks like it has more than enough power for this circuit, but if it is the multi-taped transformer you mentioned then it might just have too much resistance in the particular winding you're using.
Oh, and changing your DC link inductors will do pretty much nothing as long as they are not too small, at which the circuit will most likely go into latchup and die.
It looks to me like you just need a bigger power supply.

I think someone mentioned that it's advantageous to use a separate 12V supply for the gates - I used to do it all the time when I tried to conserve every mW of power for wireless power circuits. But now here not only that this is a negligible loss, but I also suspect that big value of gate resistance with high voltage supply actually helps improve stability of the circuit. I'm not sure why is that so, but the circuit seems to like it's gate pullups act more like high impedance current source than a low impedance voltage source. Using discrete gate divers for amplification always turned out a disaster for me.

And yet with this circuit which is the only I've built by bastardly copying the original mazzili schematic, is by far the most stable of them all. It starts up nicely even when I bring the supply voltage up slowly by a variac, a feat unseen before in a royer world.

Cheers,

Marko

Re: Royer induction heater
Gabriel35, Wed Aug 24 2011, 03:01PM

I Agree Marko!
I've experienced with some variations of the Mazzili's driver too , and the most stable one is this, with just one power supply, and using mosfets like IRFP250 & 260N (This secon one, is even better, because of the lower RDS on and higher current capabilty.

So... I think its time to get some MOTs here... But... I'm sacred about how it sags... I hope that the rewinded MOTs don't sag too much... Because if everything works, I pretend to apply something near 60v and see how much power the old and good Royer driver can deliver...

EDIT: Got some photos...







Cheers.

Re: Royer induction heater
Marko, Wed Aug 24 2011, 05:39PM

Hi gabriel... probably a good time to start using water cooling, as your coil looks quite like a roast to me?
I also don't think your mosfets are going to last long once you push 20A through them continuously with their current heatsink.
I used simple pieces of 2mm copper as cooling blocks with pipe soldered underneath which worked well enough, but if you truly wished to be a bastard you could even solder the mosfet backs directly to the pipe... you'd need to have confidence not to have to replace them too often that way, though.

I've also heard some people (including Celem cap manufacturers) complain about the specific arrangement of caps you use, with all of them in a long line from a work coil. I suspect the problem might be in the all extra inductance the cap furthest away from the coil sees, which may be comparable to the work coil inductance - this inductance in turn cancels a portion of the faraway cap's capacitance, making them appear smaller and causing those closest to the work coil to hog current.

I see you placed some of your largest caps on the far end though, which may cancel out this effect to some extent, but I would consider it for future.
If you use double sided PCB for your interconnections, inductance is low enough that you can orient the caps a more relaxed way.

Or you can use another common approach, with one work coil end going around the cap bank and connecting to it from back (some extra copper pipe usage).


Regarding the MOT: I've never rewound a MOT, so I don't know what it's voltage regulation characteristics are, but due do separation of primary and secondary I presume they are poor. Make sure to knock out the magnetic shunts from the mot too along with the secondary, and if it heats up a lot with no load add a 10-20 more primary turns to it somehow.


You could pull more juice if you wound the transformer for a bit high voltage, but just enough that it sags to 60V with royer unloaded. It would sag further when you put a workpiece in, but there's little you can do about that. Note that this way you would have to power your monster up and down from the mains side, to avoid slamming it with too high voltage from overcharged bus caps - and this is in contradiction with royer circuit's inherent desire to latch up and explode when it's supply voltage rises too slowly, as with charging bus caps after turn-on. I don't think it will be a problem in this case though, I couldn't force my circuit to latch up yet no matter what I did :)

Cheers,

Marko

Re: Royer induction heater
Gabriel35, Wed Aug 24 2011, 07:18PM

The work coil used, is from a previous Half bridge IGBT Induction heater, and is a little roasted because of that.
I think this heat sink is pretty large, is larger than my hand, and with 13A it stays very cool.
About the capacitors, I need to get some equally rated caps to make a better arrangement.
This thing is getting complicated... Since I don't have proper caps and proper Supply...
I have that two huge transformers here, and I doubt that a rewound MOT will be much powerful than theese ones.
Does someone knows something about turning two transformers in series to raise thye overall voltage?

I was wondering what happens if i take a big ferrite toroid and wound 5 turns of wire to the driver side and about 15 turns for the tank capacitors side? Its a little hard to find MOTs where I Live... so, using a toroid step up transformer between the driver and the resonant tank could be an option to increase the power? Or I need a flyback ferrite without gaps?

Thanks!

Re: Royer induction heater
Forty, Wed Aug 24 2011, 09:08PM

for a transformer between the driver and tank, you'd need a ferrite transformer anyway, as an iron one would eventually start to glow.
for the voltage sag on the supply side, maybe you could use a high current relay to switch power to the driver once the voltage on the smoothing caps reaches a certain level, thus avoiding the slow climb problem. MOT's are usually rated upwards of 1kW, so they're a bit more powerful than your transformer, but yours should be sufficient.

Re: Royer induction heater
Gabriel35, Thu Aug 25 2011, 01:46AM

So the idea of the ferrite toroid between the driver and tank side is good? Can it be a BIG toroid or i need a flyback ferrite?
Can i achieve more power using it?
I was thinking.... how to amplify the voltage of the transformer's output (the same one that I already have here) Wouldn't be a nice idea to use a voltage doubler circuit?

Re: Royer induction heater
Marko, Thu Aug 25 2011, 02:45AM

Hi gabriel,

Once the voltage on your transformer has sagged to half of it's initial value, it's putting out it's maximum power and there's nothing you can do about it.
700VA transformer should do plenty for this circuit, probably over 1.5kW for short durations. I think it sucks for you because you're using a tiny auxiliary winding which can't support all the current you need.

Your probably only hope is to either find a more useful winding on this transformer, dismantle it and rewind it or rewind a MOT.

If you find a more useful, but lower voltage winding on this transformer you could try the matching transformer approach. Smaller mosfets would then be in order though, like IRFZ44 or IRF1010 for up to 20V supply. I'd wind a 1:2 transformer tightly on your toroid and see what happens. Make sure to give it a sane number of volts/turn for the core crossection area of the toroid. You can also center tap the transformer and get rid of one inductor. Heck, you could even try supplying it with a car battery if you had one around.


Also, don't be fooled by your heatsink feeling cold, as those mosfets seem to be mounted onto a thin fin on it's edge instead of the base, also with sil pads and poor mounting pressure.. I would expect the dies to cook very quickly at 20A +. I can't see your sil-pads by the way, are they trimmed to size of mosfet packages?


Marko

Re: Royer induction heater
Gabriel35, Thu Aug 25 2011, 01:44PM

Yes, I'm using sil Pads... I have many other heatsinks here, i'll try to change them...
I was thinking yesterday and I got a possible solution...
I have here 2 300VA 24v transformers...
Can I plug their primaries together on the wall and their secondaries in series, giving me 48v AC or in parallel giving me same 24v but with more current capability?
I'm a little scared to turn the circuit on with 48v AC, what means something around 70vDC (rectified) It will sag i know... but the range for IRFP260N isn't 60v

What do you think?
What are the risks of everything explode using a 1:2 toroidal step up transformer between tank and driver?

EDIT> I think that I'm still missing something here... I made an association of a 6v tap of the huge transformer, with the other 24v transformer, I've measured the voltage and current when the workpiece was inside the coil and it gave me 32vDC and 14,9A, giving a theorical output of 480W... and I'm pretty away form melting the steel bolt with it, It just became red...

Re: Royer induction heater
Marko, Thu Aug 25 2011, 03:17PM

Hi gabriel,

Your current configuration of the circuit looks pretty much like mine, apart from your work coil being somewhat wider. Note that a jump from 30 to 60V is actually a really big one, if the circuit behaved linearly then my IH would jump in consumption (with same workpiece in) from 600W to 2400W for example.

You'll have to decide whether you want to run at 60V, in which case you'll have to remove a fair bit of caps, I'd go down to 2uF or less.
I think you're safe to plug the circuit straight to rectified and filtered 48V AC, just switch the transformers on and off on primary side (and put a breaker or fuse in series with DC output perhaps).

Or you could try using your transformers in parallel first and see what happens, in which case you can leave the circuit mostly as is. I'd recommend you a bit narrower work coil though, I can't tell exactly from the pics but yours looks like 5-6cm wide while mine is just about 3. Smaller coil will have a better coupling to the workpiece.

Regarding melting a steel bolt, it mostly depends on how big it is, 500W isn't really much power so I'd start with a M4 or M5 bolt. I'll try and see what bolt size can I melt after I finish my new power supply.

I'd recommend you to start some water cooling as soon as possible though, at 500W everything got super hot in seconds in my circuit, and I was afraid that all the emitted heat from the copper tube could cook the caps. You don't really see it at first, you just see decrease in performance as their capacitance drops, after which they bulge up and burst in flames :(


Marko

Re: Royer induction heater
Gabriel35, Thu Aug 25 2011, 04:00PM

My work coil is small as yours, it is 3cm wide too.
Why do i need to take caps out and decrease the capacitance with more power?
I was wondering to use my 8x 1uF bank to see if the power levels become higher but now i don't know...


Thanks!

Re: Royer induction heater
Nah, Thu Aug 25 2011, 04:15PM

Marco,

Would submersing the capacitors in oil help them cool?

Re: Royer induction heater
Marko, Thu Aug 25 2011, 04:49PM

Hi guys,

Gabriel, you need to remove the caps in order to get a better impedance match to the transformers. You want your IH to draw much less current (about 4 times) at 30V in order not to overlaod the supply at 60V.

Nah, I'm actually already building a MMC that is cooled by water flow around the caps. Transformer oil is hard to get here and would require heat exchangers and mess.


By the way, I challenge anyone of you guys to attempt an offline IH of this type with 1200V IGBT's and insulation transformer, as I don't ahve any suitable igbt's around myself. It would be quite challenging, as far as I know there have ben few attempts but their stability is questionable.
I would start by using a 30-40V supply for the gates, and use 470 ohm resistors and zeners much like original mazzili circuit.
Then if this is not stable enough, I have another secret little improvement that I'll only tell about if someone attempts it :)

Cheers,

Marko

Re: Royer induction heater
Gabriel35, Thu Aug 25 2011, 07:26PM

I Have some IGBTs here... The model is...
SKM40GD123D
3 phase brick.

So taking off caps is just to decrease the current consumption?
Impedance match between transformers? are you talking about the supply transformers? because I still do'nt have a step-up transformer between the tank and the driver...


I'm getting quite bored with my circuit at all...
No matter what I try, I can't go any further on it's power...
Ive tryed a lot of transformer combinations, i got the voltage stabilized @ 32v even with work on the coil, using 6v +24V in series, the current was something around 14A, and nothing seems to change.
Now I got the both transformers wired in Paralel, giving 25vAC and plenty of current, but It seems like the circuit don't draw any more current...
I'm still impressed like a young child, with the power that you've achieved with this simple IH Marko...

Should I try the 50vAC (72vDC) that the transformers give me wired in series with the two 24v outputs? do I need to decrease the number os capacitors to do that? Something tells me that it is going to explode...

Re: Royer induction heater
Marko, Fri Aug 26 2011, 12:15AM

Well gabriel, your power level looks pretty close to mine, your capacitor is somewhat smaller, and you're likely using a smaller workpiece. Try putting some coins in, they have high flux catching area and should heat up very quickly, and you should be able to melt quite large pieces of aluminum too. If you kept the heating time too long without water cooling I'd even worry whether your caps got damaged by the conducted heat, you could at least use two buckets and soem tube if you don't have a water tap nearby.

I would try stripping your caps down to 2uF and try a 60V run, then add some back if the voltage doesn't sag too much.

Re: Royer induction heater
Gabriel35, Fri Aug 26 2011, 01:49AM

Well Marko!
I got that idea of the coin too... It really heats very nice and almost started melting...
I think that the secret is on the inductive coupling that the coil have with the material and it's shape...
Maybe I'll need some bigger (And separate) heatsinks for my mosfets, some equally rated caps, and a rewound MOT...
I'm a little scared to try a 72 run (or 60v when it sags) I don't know if the driver will "Like" that initial peak of 72v (I'm afraid to kill my mosfets) What do you think?
Coz I think that if I get things to work with 60v, so maybe i don't need the MOT...(pretty diffcult to find here)



Thank you, and lets keep the progress of the topic and the project!

Re: Royer induction heater
Marko, Fri Aug 26 2011, 02:22AM

Hi gabriel,

the idea is to turn your supply on on mains side so that the circuit never even sees the 70V peak of fully charged caps. But I think it would be fine even if you plugged it straight in any way, I've overvolted mine a few times and it seemed quite resillent.

Marko

Re: Royer induction heater
Gabriel35, Fri Aug 26 2011, 02:59AM

So let me see if i understood, you're talking me to maintain everything wired out, and just turn on the both transformers using a switch on the 220v wall side?
If is that... I've always done like that... Because my switch here is pluging and unpluging the transformers from the 220v mains...
Isn't a good idea changinh the 470 Ohm resistors and maybe the 10K too, with higher values like 1K? (Using 60v)?
And what about the ideia of the step up transformer? Is it easy to blow the driver up with it?
I heard some folks saying that Inductance value of the two inductors may vary, (increase for more voltage (output), decrease for more current (output)). Is that true?
Is that's true Ill try to take some windings of the inductors out.

Re: Royer induction heater
Marko, Fri Aug 26 2011, 04:17AM

I'm not sure why would you want astep up transformer with high supply voltage. I think it only makes sense if you're using like 12V and a lot of current.

I woulld leave the 470 ohm resistors, if they and the zeners are suitably rated, as they are.

Again, if the inductors re big enough that ripple current in them is minimal (as it should be for the operation of this circuit) then increasing them will have no effect at all. Too small inductors might lead to atypical operation with blown mosfets as the most likely result :(

Re: Royer induction heater
Gabriel35, Fri Aug 26 2011, 12:33PM

Oh, OK!...I understand... So the two Inductors are not critical componentes of the circuit...
So... Turning the transformers on by pluging and unpluging them directly from the wall, the circuit will never "see" the 72v Peak?

I'll make some changes to Heatsinks of my Mosfets today, and see If I can get a MOT...
I'll post news soon, as well as photos.

Thank you!

Got New Photos...







And a Video



What Do you guys think? Lets keep the progress of the topic!
Post more videos of yours Marko!

Thank you friends!

Re: Royer induction heater
Nah, Thu Sept 15 2011, 07:46PM

Hi!

I just bought 4 0.004 uf, 1.5 kilovolt, transmitting mica caps at 40 amps at 1 kilocylcle. Could I parellel them to get 160 amps at 1kilocylcle? Would they work?

Thank!

Paul

Re: Royer induction heater
Gabriel35, Fri Sept 16 2011, 02:18AM

The capacitance seems a little low for me

Re: Royer induction heater
Nah, Fri Sept 16 2011, 09:17PM

oops...
They are all 0.09 mfd.

Re: Royer induction heater
Tonskulus, Mon Sept 19 2011, 04:22PM

Looking very good!
But, can we push this simple circuit even further? As we know, using low voltage/high current is making quite a lot switching losses. Can we resolve the problem using maybe 100Volts, even 200Volts but lower current for IGBT/mosfets?
Ok, first of all we need separated bias supply for gates.

But the main problem is that quite low impedance tank circuit. There would be huge circulating tank current if we simply put 200VDC to that circuit!

Somekind of impedance matching network is required. Transformer? Yes, maybe.. but tank circuit has to be after the transformer (tho, it works on primary side but this method will make some big power losses in transformer and requires heavy duty transformer anyway).

Then what happens to the feedback?
Im not so sure if it can be taken from the secondary (center tapped and grounded?) side of matching transformer...

Re: Royer induction heater
Inducktion, Mon Sept 19 2011, 06:50PM

Hey Tonskulus, I recently made a schmitt trigger induction heater much like your design, except I successfully made it into a series resonant design instead! It works decently on 33 volts input. (I had a small tank circuit, .3 uf and a very small coil, it was an experiment)

But, it works, and it's ZCS, so very low losses in the switches! But, I had issues with grounding, as well as keeping the tank oscillating with a larger load.

Re: Royer induction heater
RogerInOhio, Tue Oct 25 2011, 04:42PM

Hi Marko, this is a wonderful project. I have built a scaled up version of the one you have shown here with 4 IRFP260 mosfets using them in pairs connected in parallel. The tank circuit uses 24 .22uf capacitors for a total capacitance of 5.2 uf and the resonant frequency is about 40 KHZ.

So far I have taken it up to 2500 Watts input power using 70 volts and heating the pipe shown in the picture. The load depends greatly on the size of the object being heated. Running at 70 volts the capacitors get just a little bit warm.

Re: Royer induction heater
Marko, Tue Oct 25 2011, 07:38PM

Hi sir,

Well, that looks like an epic-enough scale-up. Have you had any unreliabilities/mosfet deaths upon startup and during runs? Your work coil also looks far higher inductane than mine. How much power can you push into small items like coins and such?

The obvious drawback of this method is the requirement of huge iron step-down transformer for power supply... I've been thinking about a possibility of using an arc welder for this job (the voltage would be even less stable this way though).

Hence I recommended the circuit as a newbie project where a rewound mot could serve as a cheap power supply with loads of fun to be had out of it still :)

If one was to scale this up to a mains powered design, one would have to use 1200V IGBT's along with a ferrite stepdown transformer, and think out ways to prevent the application from blowing up (some sort of overcurrent protection is mandatory). Drive could be classic mazzili style from a separate low voltage supply, or even from a feedback winding on the main transformer. I've tried this method only on low power so far.

Marko


PS. Nah - where did you get your 90nF mica caps? I'm very interested in them.

Marko

Re: Royer induction heater
RogerInOhio, Wed Oct 26 2011, 02:37AM

Marko, I have burnt up quite a few mosfets. Some times too much of a load will cause one to short out and I have experimented with different mosfets and some variations to the circuit and have lost some that way.

With the work coil I am using now small objects heat up slowly and they put very little load on the system. I am planning to make another smaller work coil for smaller objects. A smaller coil will make the frequency go up so it will be interesting to see how that works out.

The transformer I am using is kind of over kill. It is so heavy it is all I can do to lift it. I got it out of an old UPS unit they scraped out where I work. It has lots of different taps so I can get about any voltage I would need.

Re: Royer induction heater
Marko, Wed Oct 26 2011, 07:15PM

Hi,

Have you thought of implementing an overcurrent protection of some sort? I'm thinking of using something as simple as a comparator + NE555 monostable. At first I thought about using a CT with mosfet sources wired through it in reverse (so it effectively sees AC). But after a while I realized this may not be the best solution for relatively slow current rise as it may occur in fault conditions of this circuit.

Hence I think it's by far most bulletproof to use a shunt, made up of a bunch of low-value resistors in parallel. This is especially practical for the high voltage igbt version of the circuit where currents aren't as high, for the IRFP260 circuit you'd have either to use a very low value shunt (pulled from a DMM or UPS's) or opt for a hall sensor (not very in spirit of this project).

Marko

Re: Royer induction heater
Inducktion, Wed Oct 26 2011, 07:28PM


Why not use the same way that uzzors did with his 50 watt ZVS cap charger? Just wire the comparator to pull the gates low once OC is reached.

Re: Royer induction heater
Marko, Wed Oct 26 2011, 07:42PM


[/quote1319657294]

Why not use the same way that uzzors did with his 50 watt ZVS cap charger? Just wire the comparator to pull the gates low once OC is reached.

[/quote1319658060]

Hi,

Well, that basically is the idea, only with some time delay introduced to keep the whole thing of oscillating too fast (and blowing the igbt's again) in case of persistent fault condition.

Marko

Re: Royer induction heater
Marko, Tue Nov 01 2011, 06:10PM

Hey guys,

I just tried the offline version of the circuit, with 1200V igbt's and using two ATX PSU transformers for isolation.. In short, the conclusion is that diode feedback is not going to work for this. UF 4007's blew at about 100V in and when I resorted to higher voltage and power diodes the overall voltage drop got badly over treshold voltage of the igbt's (which is, to make things worse, a badly negative-tempco parameter).

I have tried to get the circuit going from secondary side feedback, but failed miserably at it - there's no theoretical reason why shouldn't it work, provided that transformers are "ideal enough", but apparently this wasn't the case here as more low-frequency response is required.

One could use a dedicated driver circuit, but I think that would sort of defeat the idea of this topology in favor of using bridges (BTW, I think a current fed H bridge is a very attractive topology to explore for an induction heater, and other things, but I'm not sure if I'll be building something that ambitious any time soon.


One idea I'd like to test out is capacitively divided feedback. SInce mosfet gate is a capacitor basically just seriesing a small capacitor and connecting it to drain of opposing mosfet should do the job actually - ofcourse some zeners and balancing resistors would be highly advisable for HV operation.

Marko

Re: Royer induction heater
radhoo, Wed Nov 02 2011, 07:49AM

too bad, I was hoping for some good news on this.

Re: Royer induction heater
Marko, Wed Nov 02 2011, 11:44AM

Hi -

Well, my OCD circuit was actually quite sloppily made, as I hacked a "monostable" out of a comparator instead of NE555 and it turned out to be a messed up approach... a mistake not worth killing igbt's.

I'll be back with this in a month or so, and I wondered if anyone would play with this on their own.







PS. there's an error with the OCD "output pin", I hope you can figure that out... also I didn't draw an isolation transformer in any of schematics but it can be freely inserted between the inverter and the work circuit.

Marko

Re: Royer induction heater
Raka, Mon Nov 14 2011, 06:51PM

I was woundering if one can better the performance of such a IH a bit by concentrating the flux as these guys did.

Im thinking of something like a crt yoke on the outside of the work coil, provided that the resonant frequency is designed to be about 15KHz to suit the yoke.

Re: Royer induction heater
SWA_Flyer, Wed Dec 21 2011, 07:23AM

how is the frequency set on the oscillator?

Re: Royer induction heater
Patrick, Wed Dec 21 2011, 07:52AM

What is that material they used to control the flux?

Re: Royer induction heater
Marko, Wed Dec 21 2011, 08:03AM

Hi guys

I'm afraid normal high permeability ferrite would get induction heated itself if applied like that. I suspect the material they use is something like those low-mu carbonyl iron materials, I might be wrong though.

Marko

Re: Royer induction heater
Gabriel35, Fri Feb 17 2012, 02:21AM

Yesterday, i was wondering, what about using a series ressonant tank circuit with the old and good Royer / Marko Oscillator?

Re: Royer induction heater
m4ge123, Fri Feb 17 2012, 04:01AM

How do you plan to do that?

Re: Royer induction heater
Gabriel35, Fri Feb 17 2012, 06:51PM

And what about Roy Induction heater?
Isn't that what we were searching for?

Re: Royer induction heater
m4ge123, Fri Feb 17 2012, 10:26PM

I mean how do you plan to use series resonance with a ZVS?

Re: Royer induction heater
Tonskulus, Sat Feb 18 2012, 10:20AM

Now I dont understand.. series resonant for what?
Parallel resonant circuit is a main part of royer/mazzilli oscillator.

Series resonant oscillator would be totally different thing..

Re: Royer induction heater
Davebmx, Tue Feb 28 2012, 12:06PM

Hi Beam tetrode may i ask a couple of questions? Is your royer setups final power output into the workpiece controlled by the input voltage? If the voltage was lowered to say 15v or even 5v for example would this just cease the induction heating process or would it just become very slow and less powerful? or would the circuit cease to function once the voltage dropped below a certain level? I say this as i am interested in the ability of your circuit to be able to melt metal but also to be throttled back to only heat to a core temp of around 100 degree centigrade for an experiment.

many thanks

Re: Royer induction heater
Gabriel35, Wed Feb 29 2012, 01:01PM

Lowering the input voltage, you will consequently lower the power output.

I think it won't cease the process, but voltages lower than 10v can make the gate drive signals too weak.

When I was imagining how a Series ressonant circuit would be, I was imagining something like this:



Crazy idea?
Impossible?

Re: Royer induction heater
Davebmx, Mon Mar 05 2012, 04:17PM

Gabriel if the voltage was below 10v would there be the chance that this might cause the mosfets to not switch? could this possibly result in the mosfets getting stupidly hot eventhough they are well within tolerance?

Re: Royer induction heater
m4ge123, Mon Mar 05 2012, 04:39PM

Gabriel, you could do that, but I think you have to leave in the resonant cap on the the primary. Try simulating it.

Re: Royer induction heater
Gabriel35, Mon Mar 05 2012, 06:18PM

m4ge123 nice idea! I'll try to simulate that, any suggestions on a program to do that?

What does the others think about series ressonant Royer IH?

Re: Royer induction heater
m4ge123, Mon Mar 05 2012, 06:37PM

Re: Royer induction heater
Gabriel35, Tue Mar 13 2012, 01:11PM

Marko, is there any difference between soldering the two inductor's legs directly to the mosfet's drains or connecting them before the capacitor bank on work coil side?

Isn't there a difference between this two points in waveforms?

I'm asking that because I'm using the two inductor's legs soldered almost directly to the mosfets's drains, but at schematic it shows them connected as the tank side...

I got a misterious cool mosfet failure here, one of them shorted and I don't know why.

My circuit was working properly, and it happened... all the other components were good, just one mosfet failed. (Took my 50A Bridge with it too)

I was using 32vDC (dont sag under this) and it was draining 33Amps from two transformers that I Have here...

Everything was cold...

Re: Royer induction heater
Marko, Thu Mar 15 2012, 08:39PM

Hi guys

Well, I think the latch-up failures of Mazzili circuits are pretty common, just not so many people tell about them! As the circuit can enter a "stable" state with one mosfet permanently on it's no surprise at all. I was actually impressed that my circuit never failed in normal use (until I miswired it's power supply and blew one mosfet up)

Over last months I experimented a lot with various modifications of this circuit, attempting to make an oscillator that starts up from linear region so the mosfets don't blow up instantly on turn-on if something goes wrong, but just conduct some current and produce little heat instead.

To stay honest, it was all pretty much a complete fail.

A severe hurdle in acheiving this with mosfets and IGBT's is their severe negative tempco of treshold voltage. I may set the gate bias so that mosfets start oscillating fine but they'll still blow up after they get hot. Trying to compensate this with source resistors failed, they ust seemed to kill the gain too much for reliable oscillation.

Alternative would be measuring the temperature of the mosfets by PTC or NTC thermistor and using feedback from this to control the gate bias, in hopes of preventing a catasthrophic runaway.



Regarding the experiment with BJT's, I'm not even sure why is it failing. I ideated using a big ferrite toroid stepdown transformer feeding a tank circuit on secondary side, and I think the parasitics of the transformer are hampering the circuit. The concept works great with small mazzili's and ATX transformers, though.

Or, the gain of the BJT's may just be too low for reliable oscillations to start up. They are pretty ancient technology anyway!

Marko

Re: Royer induction heater
Gabriel35, Fri Mar 16 2012, 03:19AM

Interesting, I'm keeping up the progress with Royer too.
Experiencing with diffrent types os tank capacitors, form 1uF to 6uF, to see what happens.

Now I'm using a supply that sags on 38v. and its pretty nice to play with the Induction Heater now.


What bridge rectifier are you using Marko?

I already got two KBPC3510s (35A) Killed (due to high current i think)

And when I was running 6uF on the tank, with a KBPC5010 (50A), The bridge and One mosfet fried, (I Think it's because the high capacitance...)

It really drops in current consumption when the capacitance is a bit high.

If we are going to try a different design what would it be?
Current fed Oscillator like Kim Ladha's one or Voltage Fed like Uzzor's one?

Or let's keep with royer? I think a current control circuit for It would be simply awesome!

Nice to see the topic alive.

Re: Royer induction heater
Linas, Tue Mar 20 2012, 05:34PM

Is it good idea to use these capacitors for high power induction heater ?
i need minimum power of 10kw and i planing to use 4 of them so(20KW max)
it should be around 350Kvar for 4 of them, but does they can hold 50KHz ?

Re: Royer induction heater
Gabriel35, Tue Mar 20 2012, 06:20PM

I think They will work fine if they are Mica or Polypropilene types...

Re: Royer induction heater
Kizmo, Wed Mar 21 2012, 10:54AM

Yes those eurofarads are oil immersed metallized polypropyleme film-foil hybrids just like cde942 series. I have 60kHz drsstc with eurofarad mmc and its perfect :)

Re: Royer induction heater
Linas, Sat Mar 24 2012, 08:57AM

ok, i get 4 of them, so 15uF and 2uH should give me around 29KHz
question is, what kind of driver to use ?
i am thinking about PLL, or maybe i should use phase prediction with inductor ?
I will use SKM400GB125D in final version and for testing SKM200GB125D

Re: Royer induction heater
Gabriel35, Sat Mar 24 2012, 02:13PM

Nice stuff out there!

Re: Royer induction heater
Gabriel35, Thu Mar 29 2012, 08:23PM

Linas, I Think that the PLL is the best choice for you...
Use KimLadha's, Uzzor's or Iam ones...

I'm thinking about that too...

Good luck!

Hey Marko! How's your progress on the royer?
What about the bigger supply and caps that you said you will test?

Let's make a Royer-Melting-A-Screw challenge!
The more I work with Royer, more I get conformed that It will only melt chapes of steel that have a good copling to the workcoil like you said, and that's not the screw characteristic.
Good for melting coins and pieces of aluminum, but would it be capable of turning a screw or nail in hot liquid? (The problem is that when it needs the power (curie point is achieved) it drops down, because the resonance of the circuit changes and it don't have feedback to keep mosfets switching at tank's exact frequency).

What do you guys think?

Re: Royer induction heater
Marko, Thu Mar 29 2012, 09:21PM

Hello

Well, I blew up the original circuit by accidentally soldering a power supply wire into a wrong place. I have a big transformer I made from some scrap E cores, but it's heavy and buzes like hell - and I'm not sure if it'll even be necessary for this little handheld circuit. It's being refitted now with a cooling fan and a handle with a switch so it can be used conveniently for longer times without having the caps overheat.

I did start making an entirely water cooled MMC with caps submerged in water,but gave up on that since RogerinOhio sent me two of those monster eurofarad 100uF 500A caps which I think beat pretty much any MMC I can make.

I made a several attempts with BJT's thinking I might have some luck, but always ended up with the circuit displaying bizzare behavior and eventually blowing up. It doesn't seem like this circuit is achievable with low gain devices.


I returned to studying mosfets and igbt's and decided to find a way to bias them properly so they don't blow up instantly after the power supply is turned on.

I also intend on replacing diode feedback with capacitive or inductive feedback.

Providing stable bias to devices that display severe nagetive tempco proved a big problem.
Firstly I tried using drain resistors in attempt to stabilize them, but this apparently reduced gain and compromised the oscillation in a similar way bjt's did.

I settled on using 5 series 1N4007's as a temperature sensor, bolted onto one mosfet with heat goop. I used their voltage drop to bias the mosfets directly; and after some fiddling I got both feedback methods working!


The circuit as always displays lots of parasitic oscillations which will have to be silenced by copious amounts of ferrite beads (this seems to work so far, though beads get very hot).

I hope to try scaling this up to a mains powered circuit, hopefully without it blowing up this time!

Cheers,

Marko

Re: Royer induction heater
Gabriel35, Fri Mar 30 2012, 02:35AM

Can't wait to see Royer oscillator running right from the mains!

I've been working with the royer too.

Now I got a serious Tank that is water cooled, with a radiator, and a bank of 32x 110nF 1200v capacitors that get just a little bit warm when operating.

The hell is that ALWAYS when i'm running the circuit something pop, the mosfets or the rectifier bridge...(the rest of the circuit keeps perfectly)
(I'm using 13v zenners instead of 12v ones for IRFP260N's.)

I found on the hard way that, for some reason when the ammeter displays more than 35A of AC current on the transformers side, the mosfets die cold seconds before measuring that current. (I'm using pretty large heatsinks with cooler on them.)

I think that it's because the reactive current of the whole circuit is getting within it's limits (something near 50A) and the mosfets can't handle this. (This is the bad thing about large currents and low voltage)
That is the main advantage of using more voltage to the royer circuit.(Like rectified mains perhaps)

I was running it one minute ago and my rectifier popped, It was a KBPC3510 (35A 1000v) (Measuring 31A on AC side)
Misterious failure (Maybe system latched up?)
What bridge rectifier do you use Marko? With heatsink?

I've found that if I try to run the circuit with 39v (Yes it sags on 39v) and with more than 3,8uF of capacitance on the tank, the circuit become more prone to failure (when I put something big on the work coil, and this object hits curie temperature, the current starts to rise and the mosfets burn...

Now I'm using exaclty 3,48uF and I think it is still so much for the circuit to handle with 39vDC...

Pretty nice Eurofarads! Can't wait to see some photos or them working!

Lets keep the progress and see if we can get a way of making the old and good Royer to run from mains!



What do you think about that misterious failures and things like that? I'm getting my money out of my walet just buying mosfets and bridges...



Cheers!

Re: Royer induction heater
Forty, Fri Mar 30 2012, 03:57PM

Well I don't know much about induction heaters because I haven't gotten around to building one yet, but could the ripple voltage from rectifying that kind of current be causing a problem with the fets or at least some instability somewhere?

Re: Royer induction heater
Gabriel35, Fri Mar 30 2012, 05:14PM

I doubt... I'm using 8x 470uF 400v Electrolytics to filter the voltage... This results in 3750uF...
Do you think I need more smoothing? how Can I do that properly?

Re: Royer induction heater
Forty, Fri Mar 30 2012, 10:29PM

So you've got about 2.2v of ripple per amp that you're drawing on the rectified side. ( and )

That could be pretty significant if you've got components that care about the ripple. But again, I'm not really sure if it affects the mosfets.

Re: Royer induction heater
Gabriel35, Fri Mar 30 2012, 11:07PM

what is the better way to get rid of the ripple? DO you guys think that this ripple can be bad for the circuit?

Adding more capacitors ar a inductor? what do you think?

Re: Royer induction heater
Marko, Fri Mar 30 2012, 11:42PM

I don't think any ripple would represent a problem as long as the voltage never drops into danger zone of <10 volts. I only used like a few thousand uF as well for filtering.

Also very hard to tell why is your circuit blowing up.

Have you tried scoping it? If there is excessive ringing present adding some ferrite beads in series with mosfet source could help.


I find it bizzare your bridge receifier is blowing up. I suspect it a chinese fake. Direct mains short is probably among the worst overloads a bridge rectifier like that can see and I've never killed one before blowing a fuse. So installing a breaker or a fuse on the transformer secondary side might be a good idea too!

I used a 25 amp bridge rectifier and didn't even use a heatsink.

Re: Royer induction heater
Gabriel35, Sat Mar 31 2012, 02:56AM

Oh! Interesting to know that, I think that this bridge is fake so...

But... a more logical thinking makes me think that for some reason is the royer that is ceasing to oscilate (i don't know why) and shorting the bridge... What do you think?

Whe the circuit shorts it takes with it one mosfet or two?

I'm uploading some videos soon...

Thanks!

Re: Royer induction heater
Marko, Sat Mar 31 2012, 10:38AM

But... a more logical thinking makes me think that for some reason is the royer that is ceasing to oscilate (i don't know why) and shorting the bridge... What do you think?

Yes, that's what happens. Some people have blamed parasitic oscillations which get bigger and bigger with increase in supply voltage. If you can scope and confirm them it'd be interesting to see if they can be stopped with some ferrite beads.

Marko

Re: Royer induction heater
Steve Conner, Sat Mar 31 2012, 12:07PM

That's right, but the root cause of the parasitics is the FET that is supposed to be off not turning off fully.

I've seen 6 amp bridge rectifiers sold as 60 amp on EBay.

Re: Royer induction heater
Marko, Sat Mar 31 2012, 01:17PM


Well, not sure about that anymore; it might be just a part of a problem. I've successfully implented capacitive and inductive feedback without diodes which should allow for better turn off (and even driving the gates below zero!) but the drain waveforms still looked like mad mess at 1Mhz. Putting ferrite beads on mosfet sources apparently reduced their high frequency gain and helped huge time. The beads got quite hot though.

Gabriel, I'd suggest you start using an oscilloscope on your circuit to see what's going on. As far as I remember, Roger in Ohio also had his mosfets blowing upafter a certain voltage, but he seemed to have fixed it by placing some steel nuts around some of his wiring.



By the way, I've just realized that the parasitic oscillations are actually common mode between the mosfet drains, with the parallel LC tank cap representing a low impedance path. If that's true, then a common mode choke between the output and the tank might turn out a much better solution than ferrite beads.


Marko

Re: Royer induction heater
Gabriel35, Sun Apr 01 2012, 01:15AM

Nice to know that. Ill try scoping the circuit to see whats realy happening. Thank you so much for your help since now friends! Thats the only reason that I can imagine that is causing the problems...

Re: Royer induction heater
Gabriel35, Tue Apr 03 2012, 12:14PM

Hey guys! I got some videos...

The description of the videos says everything as well as their titles:














The first one shows it melting an steel washer.
There are videos of aluminum melting and cooper heating too.
Proud to present my first water cooler tank! Works pretty nice, I made a radiator for it using two XEON pure cooper dssipators and some cooper coil turns soldered to it, works like a charm.

Hope you enjoy.

What do you think? Clues? considerations?

Re: Royer induction heater
Marko, Tue Apr 03 2012, 06:13PM

Hi gabriel

What was the problem then with your setup before when you complained it doesn't give enough pwoer? You just added more caps? :P

I hoped you would post some scope waveforms, preferably of higher voltage runs where you said your ih tended to blow up. Then try adding some ferrite beads to see if they get better.

Marko

Re: Royer induction heater
Gabriel35, Tue Apr 03 2012, 07:15PM

Hi Marko!
Yes, I really need to get some scope tracings as soon as possible to investigate that misterious failures...
But I don't have a Scope at home, I need to take the whole circuit to the place where I study, they have lots of Digital 100Mhz scopes, So, I'm just waiting to have some free time to do this.

I think about measuring the DC input voltage, to see if ripple is present and if it's significant.
The gate waveforms, the output of the mosfets and the tank waveform.

About Ferrite beads, that's a great idea, I plan to use it as soon as possible since it can only bring good characteristics to the circuit, and nothing that could bring the things worse.

Should I use them just on the gates?

About the power, I got some days ago, a big 15V tansformer from a 700VA UPS... So, I wired it with the previous 24v big transformer that I had...
Now the voltage sags on a very good level!

What do you think about the videos?


Cheers

Re: Royer induction heater
Bluespark, Wed Aug 15 2012, 10:23PM

What happened to Marko's Induction Noobheater circuit?
He had the best design for connecting all the capacitors.
Can anyone come up with a great design for connecting all the caps in parallel
without soldering them?

Re: Royer induction heater
Redstone, Sat Sept 08 2012, 07:41PM

Marko I noticed that the inductors you are using are yellow and that you said that you got them out of a PC power supply, from what I have experienced they are used with DC to filter ripple out and are powdered iron and not ferrite. I believe that is why they get hot when you use them at high power.

Re: Royer induction heater
m4ge123, Sun Sept 09 2012, 12:42AM

See .

Re: Royer induction heater
mohiuddinHimel, Sat Sept 29 2012, 08:26AM

hi marko,

after testing the miniature version of your circuit i wish to build a bigger one. i'll parallel four irfp250. but um stuck with the power supply.....specially with the transformer. i'll order for a big transformer but confused with the rating.

will a 220/35, 20amp transformer be able to give me at least 700 watts??? any suggestion??

Re: Royer induction heater
Sigurthr, Sat Sept 29 2012, 03:32PM

Multiply secondary voltage by secondary current rating to get the VA (or wattage) rating of the transformer.

So a 220V primary, 35V secondary rated for 20A would be 700VA. It is normal for transformers to drop in secondary voltage under heavy load though so depending on the actual impedance of the transformer you may get less than 700VA out due to voltage sag.

For example I have a 120:12 2A transformer which under 2A load drops to only 9V. The transformer is only capable of 18VA despite appearing to be a 24VA transformer.

Re: Royer induction heater
mohiuddinHimel, Sat Sept 29 2012, 06:30PM

"It is normal for transformers to drop in secondary voltage under heavy load though so depending on the actual impedance of the transformer you may get less than 700VA out due to voltage sag"

So should i increase the current capability to achieve 700 VA or more under loaded condition???

Marko's heater draws huge power when loaded. The voltage sags as well(last time it went to 17volts from 32 volts ). To achieve more than 700 watts what should be the ideal transformer rating??

Re: Royer induction heater
Gabriel35, Mon Oct 01 2012, 08:27PM

I Think a 43VAC (rectified becomes ~65v) Will be good for that application.

Re: Royer induction heater
mohiuddinHimel, Tue Oct 02 2012, 07:03PM

Thanks gabriel....i've already done with the transformer....its 35Vac...rectified and it will give me roughly 50VDC. Capable of 30 Amps. I've also got a variac which i plan to adjust the voltage if it sags too much.


Um now really confused with the capacitance. Marko said for lower power you need larger capacitance (like 4uf) and for higher power you should use high voltage and less capacitance. So, to get maximum power output at the work piece what capacitance should i choose?

In my country MKP/FKP capacitors are pretty hard to get. Will MKT X2 type capacitor work for these application? If not then any recommendation? Last time i used 100 ceramic caps (parallel on a bread board n connected through wire) and the heating was too little.

Thanks in advance. :)

Re: Royer induction heater
Gabriel35, Tue Oct 02 2012, 08:13PM

Hey!
What's your country?

MKTs will works, but Polypropilene ones will perform much better... This doesn't means that it can only be the X2 type, any well rated poly will work well.

I recommend something like 26x 220nF - 1000v caps...


The idea of using a variac in case the transformer sags too much is good, and surely it will sag, and you will ned the variac to crack the thing up...

45v already sagged, plenty of current is a good number, then you'll see the limitations of your circuit/traces/cooling... And this is pretty nice to do.

Probably if you have Well cooled Mosfets, and tank circuitry, and everything well built, you'll be able to push almost 1800w from the royer if using pairs of paralleled mosfets, ans something around 1200w I belive, using single IRFP260N's.

Keep us informed of your progress.
I hope to see photos of your work soon!

Re: Royer induction heater
mohiuddinHimel, Tue Oct 02 2012, 10:44PM

hi
Um actually from Bangladesh. I was planning to connect a large capacitance(something like 40,000uF 100VDC) as the filter cap because high current output of the transformer. I'll also connect a 50A relay between the large capacitor and dc bus. So that all that capacitance can get charged. When i'll press the trigger the relay will connect and the current will resonate.
After couple of minutes if it sags to much i'll release the trigger for couple of seconds allowing the filter cap to be charged again. :)

I've made two pretty large inductors from ferrite E cores. 19 turns each. Photos are coming soon.

Btw thanks a lot for the information. I was badly in need of those suggestions. To make it rough i was planning to use 40X0.1 uF X2 1000 volts capacitors as parallel bank...if you suggest i can go with 26x 220nF - 1000v caps too.

Mosfets will be water cooled directly connected to the copper pipe as Marko's design.

Looking forward to your opinion.

Here are the inductors i made out of ferrite E core and thick copper wire. 19 turns each.





got lucky and managed to get 38X100nF MKP r76 capacitors. and 4X56nF caps.

Re: Royer induction heater
Gabriel35, Wed Oct 03 2012, 06:54PM

Nice pics! Keep the progress...

Water cooled mosfets is a good idea too...
You can make the tank with 40x 0.1uF, this should work even better, since the currnt will be shared between more capacitors...


BTW Nice caps too!

Re: Royer induction heater
mohiuddinHimel, Fri Oct 05 2012, 09:21AM

i electroplated the PCB board to increase copper thickness and current carrying capability. Then made a sandwich with two PCB pieces and a piece of wood in between them. Total capacitance is 4 uF. I'll bolt two extra thick copper bars on each side to mount the cap-bank with work coil.

I really wonder will this pcb copper layer be able to carry 200/300 amps to the work coil :/ :/

any suggestion ??

Re: Royer induction heater
Sulaiman, Fri Oct 05 2012, 12:39PM

I have noticed a few attempts at off-line zvs/royer/cfpr inverters failing,
one common problem is that high voltage igbts tend to drop significant voltage Collector-Emitter (Drain-Source for mosfets)
this means that the circuit where a diode from one collector (drain) turns off the gate of the other device does not work at higher currents.
You need a reliable gate drive scheme for high voltage transistors carrying large currents.
Just thought I'd mention it.

Re: Royer induction heater
mohiuddinHimel, Fri Oct 05 2012, 06:07PM

Thankyou sir

Right now i dont have any scope as my vacation is going on and i cant access my University Lab for the time being. So i need something very simple yet powerful.

Can you give some designing tips for gate drive schematic???? The FET i'll use is IRFP260N in pairs. Water cooled.

I've come a long way so its kind of impossible for me to change the whole circuit.

Re: Royer induction heater
Gabriel35, Fri Oct 05 2012, 08:40PM

Nice capacitor construction... Looks Heavy duty...

What Sulaiman said is true! it is a challenge to get a royer with dedicated drive circuit right from the mains! But it's not impossible and it would be very nice to see it happening!

Keep your progress...
Remembering that a well made IRFP260 based, original Royer, can do do serious damage to metal pieces and it can be ferrous, or non-ferrous materials... =D If you want to go over 1000w or 1500w, forget it and star designing a dedicated driver circuitry...


Good luck! Keep us informed!

Re: Royer induction heater
mohiuddinHimel, Fri Oct 05 2012, 09:53PM

hey thanks for the feedback.

1000W is like overkill for me. I'll be very happy if i can deliver 700-800 watts to the work-coil. My goal is to make a 16 mm dia 1" long iron rod bright hot in 10-15 seconds. Is it possible to achieve with my existing configuration????

And of course i'll make the whole construction neat and heavy duty ;)

Re: Royer induction heater
Sulaiman, Sat Oct 06 2012, 09:15AM

Maybe an example will help, IRFP260 was mentioned so :-
Vds <200V so Vdc<64Vdc,
at Tj=150C Rds(on)=2.3 x 0.055=0.1265 Ohms

To turn off the other device Vds(on) <= 4V - Vdiode..say 3.5V (good diode)
So Ids(on) < 3.5V/0.1265 Ohm = 27.7 Apk=27.7 Adc=1770W
not counting for any switching losses, each transistor would have to dissipate
1/2 x 3.5V x 27.7A = 48.5W
which needs a heatsink temperature of less than 150 - (0.7 x 48.5) = 116C
... = quite large or a cpu cooler etc. per transistor

Be aware that the current in the resonant L//C will be many times 27.7 A.

At this power level every parameter is at its limit
... the slightest departure from 'ideal' will cause failure.
I'd say 50 Vdc at 20 Adc = 1 kWdc is a reasonable target.

Now do similar calculations for your chosen high voltage transistors.
Remember that schottky diodes rated at 1.2 kV aren't common

Re: Royer induction heater
mohiuddinHimel, Sat Oct 06 2012, 12:04PM

Thankyou sir for making me understand the behavior of FET more deeply.

The calculation above is for a single transistor. As i am using two of them in parallel, total 4 for the circuit, the current will be devided. So to achieve 1000Watt each of the FET(mosfet) has to conduct 14 A max (i assume). Each of the FET will be placed on a thick copper bar welded directed to cooling pipe. I'll be able to keep the temperature near 80-85 C.

The diodes will be UF4007 fast rectifier diodes 1KV two in parallel. Seems safe to me :)

Please correct me if i am wrong.

just hope this transformer will deliver 1000watts. Still disappointed with the size. I thought it would be bigger. Anywayz finger crossed

Re: Royer induction heater
mohiuddinHimel, Tue Oct 09 2012, 09:42PM

Here is the parallel ZVS circuit originally Marko's design with a little scale up. Is the circuit "All right" ??

Re: Royer induction heater
Marko, Wed Oct 10 2012, 08:31PM

Hello -

I don't have much experience with paralleling mosfets in this circuit, but I fail to see the purpose of having portions of your tank capacitance across every mosfet. I would pretty definitely leave those out.

Also, there is no need for paralleling zeners, 10k resistors and diodes; you can use single zeners (of course with adequate power rating which you need to calculate according to your supply voltage). Also, single UF4007 diodes should be enough.

If you need to make up higher power zeners from several small ones, I recommend using a series of several lower voltage rated zeners. Paralleling diodes is generally a no-no and may result in explosions.


Finally, whenever you parallel mosfets it may be good idea to decouple mosfets from each other to prevent parasitic oscillations. A ferrite bead along with 10ohm resistor in series with each gate may be a good start. I think someone here even used steel nuts over the wires instead of ferrite beads for this purpose!

PS. I'm also seeing you intend to use ferrite cores for your inductors. In this case, make sure they are properly gapped so they can withstand the required currents without saturation. It would be wise to do some math on this as well before firing the oscillator up.

Marko

Re: Royer induction heater
mohiuddinHimel, Thu Oct 11 2012, 07:33PM



Hey thanks for the suggestions!

Those capacitors(snubber) were used to suppress high voltage spikes.

Anyways simpler is better. Revised schematic>>




each mosfet with 10 ohm resistor. I'll also use ferrite bead around each mosfet's gate. Any final tip????

Re: Royer induction heater
Mads Barnkob, Sun Jan 20 2013, 01:25PM

Time to revive a good old thread :)

I built this Royer IH, the circuit:

I made some measurements on it, 35VDC in, 100V tank voltage, 10A inverter current, 200A tank current, 100kHz.

Re: Royer induction heater
klugesmith, Sun Jan 20 2013, 08:50PM


I like your electroplating work. Any idea how much copper you added? (for example, using before-and-after measurements of weight, thickness, or electrical resistance)

It would not be hard to measure the sheet resistance (Google that) of your PCB square, with or without capacitors, using a four-terminal method.
Run a known, constant current of a few amperes from one corner to the opposite corner. Then with a voltmeter on a millivolt range, measure the voltage difference between two places on the board.
It's educational to find and mark lines of constant potential using one voltmeter probe, while the other probe is connected to a reference place such as a third corner. The constant potential lines will approximate circular arcs in the neighborhood of your current source and sink points.

Re: Royer induction heater
Dri0m, Mon Jan 21 2013, 05:24AM

How much can those two inductors vary? I made identical coils but one has 144uH an another one has 131uH... It's kinda hard to rewind it because the wire is 2.5mm in dia and i have only few meters

Re: Royer induction heater
Mads Barnkob, Mon Jan 21 2013, 09:09AM


I measure my inductors to be within the same tolerances as yours.

I will try to upgrade my circuit with some 300V MOSFETs, have to source a new power supply and much better tank capacitor too.

Re: Royer induction heater
Gabriel35, Wed Jan 23 2013, 12:41PM




That's a good idea! Lets keep working to see the limits of the godd mazzili driver!

Re: Royer induction heater
Redstone, Sun Feb 03 2013, 08:26AM

I believe the problems that most people have with a ZVS driver is that they do not have a powerful enough power supply, If the voltage drops below 10 volts they can latch up and burn out a mosfet. if you power it with a little wall wart and then load it too much, it will stop oscillating, but the mosfets will "survive". if you do the same thing with a better power supply, like a good supply from a computer (some are not good at all) you will fry mosfets. assuming all components you have in the schematic are good, this strategy should work.

I noticed in earlier posts marko mentioned over current detection, I think under voltage would be better, the real dangerous problem is if it stops oscillating. Feed the beast every thing or it will stop oscillating. just don't over load it.

Markos schematic operates around 50 khz which is too slow for small parts like nails and screws, but would work better for larger items like a solid 3/4" bolt and up. that is why some people here can't melt anything with it, it gets hot but will never absorb enough power to melt it effectively. I think markos inductor coil is too small to work as effectively as it could, more turns maybe try slightly smaller cap value also.

Re: Royer induction heater
Gabriel35, Mon Feb 04 2013, 06:57PM



Pretty Interesting, Didn't know that!

Re: Royer induction heater
JoeDarkon, Wed May 22 2013, 07:25PM

First, thanks for making this circuit available!

I have been collecting components to make an induction heater based on Markos circuit but I just realized I got mks capacitors instead of mkp capacitors (these) Is that going to be a problem? I looked at the specs and it looks like the mkp ones are better for this type of application but Im not sure.

I am planning on first making a demo unit but then I intend on one that a little more robust as I would like to use it for some forging. Would the MKS caps work this its application?

Re: Royer induction heater
arbartz, Fri Jul 25 2014, 12:56AM

So I'll be honest, I didn't read every post in this thread. If the answer is in here I'm sorry for asking.

I built the circuit pretty much exactly as it is shown in the first post. Here is the list of components I used:

2x IRFP250M MOSFETs
27x Kemet R46KN410050N1K Caps
2x 10kR Resistors
2x 2W 470R Resistors
2x ON Semi MUR8100EG Diodes
2x MCC 1N5349B-TP Zeners
2x 10A 100uH Inductors

The circuit seems to work fine when I'm providing it with a stable 12V, but as soon as I hook up 24V the MOSFET's pop. Taking a meter to it shows a dead short between ground and power. I replace the MOSFETs and it works fine again. So I'm not sure why it would pop almost instantly under 24V, but run just fine at 12V. Does anyone have any ideas? 12V doesn't really get me any melting power, only glowing power...

Thanks!

Re: Royer induction heater
Fractal Universe, Sun Dec 14 2014, 11:29PM


how big is your work coil ?; how many turns?

i looked up the part # on your capacitors and they are 1 microfarad each, are they connected in parallel; if so that s 27 microfarads, way more than the 4 microarads in Marko's design, that seems like a lot

if you plan on running it on only 24 v (30v max), then 100 Vds MOSFETS are probably a better choice than irfp250's due to their lower Rds ON; Marko recomended that the MOSFETs used have an Rds ON of less than 50 milliohms

Re: Royer induction heater
StevenRS, Thu Jan 29 2015, 06:48PM

So, I decided to jump on the induction heating bandwagon finally. I actually *need* one to quickly heat up a thin steel band, 1" wide and .015" thick so I can quench it as I draw it through the heater. This is the schematic I've come up with, with modifications to deal with my particular PSU.



I already have all the parts, and have tested discrete elements of it. Before I hook it all up, does anyone have any concerns/criticisms? I could replace the mosfet that switches the circuit on once the cap is charged with a relay, but it should only dissipate about 10 watts.

I also am thinking about adding a simple overcurrent protection circuit to it, but not sure about the best way to do that. Ideally I would like to detect a latchup and switch it off via the capacitor charge detector mosfet.

What I would really like is a way to drive the mosfet gates down faster, because it looks like the vast majority of their dissipation happens because they both enter linear operation simultaneously, with once switching off and the other on. I wonder if a slightly asymmetrical design could make sure one switches faster than the other and reduce dissipation while preventing latchup without blowing everything to pieces. Will experiment, and post pictures soon.

Also, if anyone wants it I attached the spice schematic I've been playing with. I changed the filter cap's value to 1000µ so the simulation starts in a reasonable amount of time, otherwise it takes AGES to start oscillating (It does, though). You will need to place the two files for the fast diodes and irfp260's into your LTspice top dir for it to run.




]royer_induction_heater.zip[/file]

Re: Royer induction heater
hen918, Fri Jan 30 2015, 04:27PM

Are you sure this will be powerful enough?
You may need to double up on the MOSFETs and I would make the coupling inductors out of ferrite cored pipe, and the cap banks out of many, many, small caps(I like polypropylene tubular audio caps 0.1uf)
You'll have to tune it to get close coupling with your work-piece.

Good luck! (I've made Royer oscillator type IHs and never got them very efficient.)

Re: Royer induction heater
Experimentonomen, Sat Jan 31 2015, 12:53PM

You might wanna look into Neon John's "ROY" circuit which is a improved version of the royer/mazzilli circuit.

Re: Royer induction heater
hen918, Sat Jan 31 2015, 02:39PM


Please note that the "ROY" circuit is designed for rectified 110v AC not 240v AC

Re: Royer induction heater
StevenRS, Sat Jan 31 2015, 04:32PM

Forgot to mention that my cap is made of I think 18 or 19 polypropylene film caps, soldered directly to a chunk of bus bar.

I'm going to assemble it today, we will see how it goes!

Re: Royer induction heater
mister_rf, Sun Feb 01 2015, 04:42PM

Hi, I was able to test this Royer Oscillator design.
Instead of IRFP260N I have used the IXFK140N30P MOSFET transistors. No additional circuits for latch-up protection, just a small switch to connect the power to the gates after the main power on. Main capacitor =4.4uF, made from a bunch of 20 polypropylene film caps, the 940C series. The power supply used for the test provide 50V and the current up to 40-50 amps. I used to measure the current by a small 0.1 ohm shunt resistor.
For the no load condition, I=6-7 amps and on load the current rise up to 30 amps. Please see the attached video.

Re: Royer induction heater
Claude, Mon Feb 02 2015, 02:40AM

Hello I am able to obtain a large quantity of wima 0.1uf MKC capacitors. I am wondering if these will be ok for a small induction heater project as they are metallised polycarbonate.

Re: Royer induction heater
Mads Barnkob, Mon Feb 02 2015, 06:33AM


You need it the other way around. Without power on the gate, you have no idea if they are open, closed or somewhere in between.

Re: Royer induction heater
mister_rf, Mon Feb 02 2015, 09:16AM

I need to give all details about the schematics I have used for.
That’s the one including some additional resistors and diodes:

Re: Royer induction heater
hen918, Mon Feb 02 2015, 09:25PM

I would advise pulling both gates down via diodes to turn it off. works very well and means that the MOSFETs do all of the switching, which they are very good at!
btw this was Uzzor's idea, not mine

Re: Royer induction heater
Fractal Universe, Tue Feb 03 2015, 07:39PM

So i finally decided to post my version of this device, i designed it to be an add-on to a MIG welder, which has a DC output of 38v max, so i went with 150 v MOSFETS and 150 v schottky diodes. The capacitance is twenty four 0.1 microfarad 400v mkp caps in parallel, although i plan on doubling it, the total power in was approx 1400 watts

the second pic is about 1 min after the first, the 1/2'' id hex nut is melting into a puddle

Re: Royer induction heater
Claude, Sun Feb 15 2015, 09:48AM

Hi fractal universe. What is on the circuit board to the side of your royer induction heater. Is it a separate gate drive?

Thanks
Claude.

Re: Royer induction heater
hen918, Sun Feb 15 2015, 08:14PM

Ohhhh, Steve, where did you get those big toroids? I could do with some.

Re: Royer induction heater
Fractal Universe, Sun Feb 15 2015, 09:35PM

hello, i got those on ebay 3 months ago for $8.95 each, they were labelled as RFI/EMI filters used on large 2000 watt dimmers for incandescent lamps in theatres, the seller had only 10 of those toroids i bought two of them for this IH. unfortunately the seller is now sold out,

i have no idea what the exact specs of those toroids are, on whether they are powdered iron or ferrite and what mixture (powdered iron is superior for the Mazzili IH because it does not saturate easily, even with a large DC component of current flow, like when the IH is heating a large metal piece). they are 3.5 inches O.D. wound with 60 turns of #11 magnet wire,
i have tested many toroids on the Mazzili IH and it looks like micrometals type 26 or type 52 powdered iron works very well, of course the bigger the better:) you can get those at amidoncorp.com

amidon only sells the bare toroid cores, if you choose to go this route, you will also need some wire to wind them yourself.

hello, it is the gate drive circuit; the same as the classic Mazzili circuit with minor modifications, no separate power source.
the modifications are: the positive and ground terminals feed into the GD board through ferrite toroids and a 100v 2.2 microfarad film cap connected from positive to ground, these filters protect the GD circuit from any electrical noise. the zener diodes i used are 18v 5 watt 1N5355's. the resistors are 270 ohm 5 watt metal oxide resistors. instead of using fast diodes i used MBR2015CT 150v 20a Schottky diodes; less than 0.5 forward volts (TO220 packages), using Schottky diodes instead of fast diodes makes a difference.

the MOSFETs were four (2 in parallel on each side) FQA46N15 150v 50a; which failed when i tried to melt a lugbolt, i then upgraded to four IRFB4115PBF (150v 100a ea) and those work better

since each side has 2 MOSFETs in parallel, the gates of each device is connected to the GD circuit with a ferrite bead(pulled from TV's) in series to dampen parasitic oscillations

Re: Royer induction heater
Claude, Mon Feb 16 2015, 07:55AM

Thanks for the explanation to the modifications of the gate circuit.
Claude.

Re: Royer induction heater
hen918, Mon Feb 16 2015, 09:44PM

Thanks Steve, found some nice toroids here:
I think these are the same as Amidon ones.

Re: Royer induction heater
Fractal Universe, Tue Feb 17 2015, 05:52AM

those are exacly the same; micrometals cores, amidon is only a distributor.

Re: Royer induction heater
tanc, Wed Oct 25 2017, 01:52PM

hi everyone, i too am trading the marco/mazilli IH path and loving the process (unlike several of my RIPFETS!). I am currently on version 3 and I have 2 questions.

1:how do i limit the current of a rewired MOT? can this be achieved with a phase control thyristor circuit?

2:can the capacitors from microwaves be used in a tank circuit as they claim to be good for over 1500VAC?

hoping this thread has not gone cold

Re: Royer induction heater
hen918, Wed Oct 25 2017, 03:54PM

If you are talking about using a rewound MOT as a transformer in a mazzilli circuit, you're out of luck: MOTs (or at least the old style ones we tend to call MOTs) are designed for 50 or 60 Hz, not 50 or 60 kHz, so they would be incredibly lossy even if you could get them to work.
Microwave oven capacitors are also designed for low frequencies, and I would assume (I haven't tested them) that they would also have high losses at high frequencies.