Royer induction heater

[afk]

I HAVE REWORKED THE CIRCUIT PLEASE CLICK ON THIS LINK TO HEAD TO THE NEWEST CIRCUIT: https://highvoltageforum.net/index.php?topic=46.msg540#msg540

Hi guys,

First time coming here, though I have been in Kaizerpowerelectronics blog before.

Since it will be pointless to make another thread just for the same induction heater, I'm borrowing kamelryttarn's thread to post about my machine. If you don't mind, that is (I'll do another thread if you do mind).

First of all I have to thanks Mads for his advice on using multiple pairs of MOSFETS to deal with high-power circuit. I have made a prototype, and it worked out well:


However, the consumption rate is bad. Even at no load, it consumed approx. 16A for 15.7V input (dropped from 18V). When I put a screwdriver in, the input current rose to  17.47A. Putting in a piece of stainless steel (3x6 cm²) it became 24A. Basically, the wasted energy into heat is just too big.

I assumed that the soldered joints and copper traces burn out a lot of energy, so I decided to redo the circuit (simpler, and wider traces for high-load). Also, getting a bigger work coil.

But for this time, it does not oscillate. I fried 3/4 MOSFETs when it failed to oscillate, and had to redo the circuit (again). Still no oscillation. Just in case, I only make the newest circuit to run under voltage for around a few seconds to observe the problem.

It seems that the input voltage has dropped very low, around 7.5V (from 18V) at 20A consumption. Something must happen badly with the power supply, so I'm posting it all here as well.

The transformer is one I wound by myself. It is very big (enough to be put inside a computer case) and personally I think it can support at least 2kW:

Little final touch:


Assembled with rectifier diodes. You can see below the two square components (the bridge rectifiers) that can deal 1kV-50A each. I used two so that it can work well to 100A (I'm using the transformer at 40A as my aim, btw). To protect the transformer I also added a 20A anti-leakage circuit breaker for the primary coil.


Basically I'm confident that my transformer can support a lot of current even for 18V secondary. I'm, however, doubting the filter I used.

I'm using capacitor-input filter (pi-filter with 1 capacitor bank + 1 coil + 1 capacitor bank) to increase the filtering efficiency. You can see the big-ass coil of 48mH underneath the capacitor bank (actually two banks of around 7000µF each I combined on one circuit board).


I assume that with bigger coil the circuit will need stronger power spike to start up, and with that it overloaded the filter. Do you think that it is the case with mine?

EDIT: To OP: Where did you buy (or find) the choke? I'm trying to find a big one like that but it is hard to get one. If you bought it, how much did it cost? And btw, you are using the two chokes in the induction heater, or they are for the power supply?

[Dylan Burger]

Hi guys,

First time coming here, though I have been in Kaizerpowerelectronics blog before.

Since it will be pointless to make another thread just for the same induction heater, I'm borrowing kamelryttarn's thread to post about my machine. If you don't mind, that is (I'll do another thread if you do mind).

First of all I have to thanks Mads for his advice on using multiple pairs of MOSFETS to deal with high-power circuit. I have made a prototype, and it worked out well:


However, the consumption rate is bad. Even at no load, it consumed approx. 16A for 15.7V input (dropped from 18V). When I put a screwdriver in, the input current rose to  17.47A. Putting in a piece of stainless steel (3x6 cm²) it became 24A. Basically, the wasted energy into heat is just too big.

I assumed that the soldered joints and copper traces burn out a lot of energy, so I decided to redo the circuit (simpler, and wider traces for high-load). Also, getting a bigger work coil.

But for this time, it does not oscillate. I fried 3/4 MOSFETs when it failed to oscillate, and had to redo the circuit (again). Still no oscillation. Just in case, I only make the newest circuit to run under voltage for around a few seconds to observe the problem.

It seems that the input voltage has dropped very low, around 7.5V (from 18V) at 20A consumption. Something must happen badly with the power supply, so I'm posting it all here as well.

The transformer is one I wound by myself. It is very big (enough to be put inside a computer case) and personally I think it can support at least 2kW:

Little final touch:


Assembled with rectifier diodes. You can see below the two square components (the bridge rectifiers) that can deal 1kV-50A each. I used two so that it can work well to 100A (I'm using the transformer at 40A as my aim, btw). To protect the transformer I also added a 20A anti-leakage circuit breaker for the primary coil.


Basically I'm confident that my transformer can support a lot of current even for 18V secondary. I'm, however, doubting the filter I used.

I'm using capacitor-input filter (pi-filter with 1 capacitor bank + 1 coil + 1 capacitor bank) to increase the filtering efficiency. You can see the big-ass coil of 48mH underneath the capacitor bank (actually two banks of around 7000µF each I combined on one circuit board).


I assume that with bigger coil the circuit will need stronger power spike to start up, and with that it overloaded the filter. Do you think that it is the case with mine?

If it is drawing an insane amount of current at no load - you have too little turns on your work coil and/or your tank cap is too small/big.

Mads should be around soon to tell you about the thread thing.

If you'd like you can take a look at the troubleshooting guide at the bottom of my thread on the mazzili driver. Best of luck! and let me know how things go

[RocketScienceSmurf]

afk, that is one HUGE toroid transformer. I'm amazed that you wound it yourself! Nice work.

[afk]

afk, that is one HUGE toroid transformer. I'm amazed that you wound it yourself! Nice work.

I got the core for around $44 (biggest spending for my project), and also need quite a lot of wires to wind it. It is a shame that I don't have much experience in transformer myself, so I did it mostly by consulting on Internet or people around me.

Winding it is a pain as well. The primary is still okay-ish (2.3mm-diameter wire), but for the secondary, it is 2.5x7mm² square wire, which is 1) thick af and 2) insulated by fiberglass, which can cause you a hell of itch. I added another secondary output at 14V for fan if my machine needs one for cooling.

If it is drawing an insane amount of current at no load - you have too little turns on your work coil and/or your tank cap is too small/big.

Mads should be around soon to tell you about the thread thing.

If you'd like you can take a look at the troubleshooting guide at the bottom of my thread on the mazzili driver. Best of luck! and let me know how things go

Hmm... I'm trying to tackle with my coil a bit... Still, I'm having a 6-turn, 10-cm-diameter work coil but it is currently not running. The old coil that worked had 5 turns with 7 cm of diameter, but the consumption rate was bad.

It is hard for me to find better tank cap. The best one I can find here is MKV cap (which is, honestly, heating up fast and that worries me). I don't have much experience in finding good cap myself... There is a retailer here that sold power cap (used for actual induction heater) but apparently it was not selling well, so currently they did not stock up the cap...

[Dylan Burger]

Hmm... I'm trying to tackle with my coil a bit... Still, I'm having a 6-turn, 10-cm-diameter work coil but it is currently not running. The old coil that worked had 5 turns with 7 cm of diameter, but the consumption rate was bad.

It is hard for me to find better tank cap. The best one I can find here is MKV cap (which is, honestly, heating up fast and that worries me). I don't have much experience in finding good cap myself... There is a retailer here that sold power cap (used for actual induction heater) but apparently it was not selling well, so currently they did not stock up the cap...

First off- if anything fails in this driver - it would be a good idea to check all other components too. Your transformer is probably fine. Just make sure your Zenner diodes arent shorted and that your ultrafast diodes are ok. Next step would be mosfet replacement. Then you can test tank cap - charge it up to 30v and see if it starts to discharge quickly. If it can hold its charge (and its capacitance always reads the same) then its good.

Alright , It may be that your frequency is inappropriate for your switches or if your driver isn't working at all anymore your zenner/fast diodes have blown. If that is the case your FETs will also be damaged. a multimeter on diode test can quickly ensure that everything is working and not shorted out.  If it only works with one coil, then i would check that it is not pushing the frequency up too far.

I see you have ~80Khz resonance - Thats not bad. Sometimes a bad cap can start to blow up mosfets. What size is your capacitor and how much capacitance does it have ?

[afk]

First off- if anything fails in this driver - it would be a good idea to check all other components too. Your transformer is probably fine. Just make sure your Zenner diodes arent shorted and that your ultrafast diodes are ok. Next step would be mosfet replacement. Then you can test tank cap - charge it up to 30v and see if it starts to discharge quickly. If it can hold its charge (and its capacitance always reads the same) then its good.

Alright , It may be that your frequency is inappropriate for your switches or if your driver isn't working at all anymore your zenner/fast diodes have blown. If that is the case your FETs will also be damaged. a multimeter on diode test can quickly ensure that everything is working and not shorted out.  If it only works with one coil, then i would check that it is not pushing the frequency up too far.

I see you have ~80Khz resonance - Thats not bad. Sometimes a bad cap can start to blow up mosfets. What size is your capacitor and how much capacitance does it have ?

As for the first try on the new coil, I burnt through 3 FETs and a few diodes. Most of the rest were still safe, but to be sure, I replaced all diodes and the remaining MOSFET as well.

With the second try, it is still not oscillating, but everything still looks okay (I did a quick measurement of resistance between important nodes and compared the values before and after the test). Within 5 seconds running, I checked that the source voltage dropped to approx 7.5V from 18V (before it was 15.7V), drawing approx 20A of current. I assumed that the problem lies in CLC filter, notably the capacitor banks could not hold the voltage high enough during the startup, so it fails to oscillate. I used 17 electrolytic capacitors of 1kV-1mF (though they aren't at good quality: most of them are 750µF and 900µF), creating two banks of around 7mF for the CLC filter. I'm planning to buy even more capacitors as well.

Tank cap is 16x0.22µF MKV cap and I measured 4.33µF (I assumed one cap got disconnected from the bank, I'll resolder the joints later). Since with this tank cap I was able to make the circuit oscillates and the capacitance didn't drop so much, I think it is still working. Though, I'll need better cap if I'm aiming for 2kW heating output.

[Dylan Burger]

I don't have any scientific knowledge on these drivers, only knowledge from experience. I can say that if the tank cap is heating, and your getting a high frequency from such a large tank capacitor, then your driver is basically focusing on pumping power into the cap. I theorize that it has to do with impedance. You want an equal load across caps and coil, so that you get a good resonant rise and so that your caps (or coil) doesn't suck up all the power.  That's just my theory anyway, so anyone who has proper knowledge is more than welcome to completely discredit what I have said.

If you are pulling 83 KHz from a 4.3uF tank cap, then you have a very low inductance primary coil. It makes sense that current draw is that high. If you would like to lower the current draw you can add a turn or two.


If 16x0.22 uF capacitors are heating, then there is a lot of tank current flowing (which makes sense since you are drawing 20A from your supply).  I would try adding a turn or two and see how that goes. You want it to not draw much when nothing is in your work coil (maybe 5A to 7A). Also keep an eye on your tank voltage- higher tank voltage means a stronger electric field in your coil. Which also means easier heating of objects.  For 2Kw induction heating you'll need low resistance caps (low ESR rating I think?) , or high RMS current rated caps. But I think if you can balance your tank current out between the work coil and  the caps you'll be fine.


Also, if zenner diodes keep blowing up, make em 18v 1W zenners. I've never had a zenner fail since

[RocketScienceSmurf]

Are you sure your psu choke isn't part of the problem? 48mH is a LOT! My ginormous choke has an inductance value of 1.8mH. Maybe the DC-resistance in your choke is too high causing a voltage drop when you start to load it?

A choke that has 26 times the inductance of the one I got but is one tenth in size has got to have a lower current rating, thinner wires and therefore much higher voltage drop at high current. (I think... I'm no expert)

[Dylan Burger]

Are you sure your psu choke isn't part of the problem? 48mH is a LOT! My ginormous choke has an inductance value of 1.8mH. Maybe the DC-resistance in your choke is too high causing a voltage drop when you start to load it?

A choke that has 26 times the inductance of the one I got but is one tenth in size has got to have a lower current rating, thinner wires and therefore much higher voltage drop at high current. (I think... I'm no expert)

This is a very valid point. I did not consider that the resistance might  be inhibiting the supply power.

[RocketScienceSmurf]

Even a DC resistance as low as 0.5 Ohm (500mOhm) constitutes a voltage drop of 10V at 20A. It doesn't take much when you are working with currents like these.

[afk]

Hmm? The choke has to be around 45mH or else I'll have to buy even more capacitors for the CLC.

The cutoff frequency of my CLC filter is approx 9Hz with approx 45mH choke and approx 7mF cap banks. The grid is at 50Hz and I used full-wave rectification.

I can actually liberate a few turns, but then I have to invest more on capacitors. Reducing inductance by a half, I will have to compensate by doubling the amount of capacitors. It is pretty hard to even find low-quality capacitors with high capacitance (as you can see, I bought 1mF caps, but their actual capacitance is only at 75-90%).

I wound it with 2.5x7mm² square wire (the choke is beneath the capacitor banks in the last image), so it is plenty even for 50A usage. Wires are insulated by fiberglass which is fairly resistant to heat, and also, since it uses a lot of E-T sheets the heat dissipation should be sufficient for a dozen minutes working.

I prefer to reduce the number of turns, though. Less turns will reduce inductance loss at high current (or so I read from various forums and documents). But then I don't want to spend so much money on capacitors. Even the wires I used for the choke are the leftover of my transformer. It is such a dilemma, haha.

[RocketScienceSmurf]

You seem to have given the psu filtering a lot more thought and attention than I have so far, but ripple frequency was one of the reasons I opted for a three phase setup for my power supply. You can get away with a lot smaller capacitor bank that way. I will build a LC filter with my choke and a suitable capacitor bank.

[afk]

I have a small question regarding the toroidal cores for the two chokes in the driver.

I just read a few articles on the net and I found out that the mix in the core can affect the performance of the chokes. And you can determine the freq band that the core can work based on its paint color. I'd like to know more about this, in case if it is true.

I'm having two green cores, but they are quite small so I'm thinking of purchasing something bigger, so that I can use bigger wire.

[RocketScienceSmurf]

AFAIK there is no actual color standard that every manufacturer adheres to. Micrometal for example has a color standard so as long as you buy an original Micrometals core you can determine the core material based on the color but you can not necessarily use the same information to decide the composition of a core from another manufacturer. I bought two Micrometals T300-60D-cores. I would have preferred T400-60D but they were harder to find and quite expensive. The T300 was a good compromise that I think will be sufficient for my experiments. At these power levels the T300-60D is pretty much the smallest core you can use according to their software.

http://www.micrometals.com/downloads/MicroRelease_March2010.exe

[Mads Barnkob]

You do have a oscilloscope, so you should wind yourself a cascaded current transformer (two ferrite ring cores with each 33 turns, 1 looped into the other and that gives 1:33:33 ~ 1:1000 turns ratio, so 1000 A through the first gives 1A out, then you can adjust shunt resistor for the voltage desired) to measure the primary circuit current, it could be that you actually have 200 Ampere flowing there and that explains your large current draw from the power supply.

It sounds like kamelryttarn has a good point in the large inductance of your filter, I never used a L part of the filter, just 4x 10000 uF in parallel. This circuit is paramount to fail at too low supply voltage and that is the true killer and also why this circuit sucks, you never know what happens at startup and you risk blowing it all up

About ferrite markings, there is no standard at all and there are numerous different names for the "same" materials as different manufacturers have copyrighted their own branding names. Micrometals does a two colour identification but mostly I have uncoloured cores that just has a small writing with the material type like "3C85" or "3F3".

I have a lot of ferrite cores for sale, check my thread that I'll make tomorrow

[afk]

Since I'm still looking for filter cap and better toroid for the two chokes, I'm posting the schematics of the driver circuit:



I'm using double MOSFETs schematics so that it can support higher current consumption. I have tested its functionality before with 12V and IRF250N before and it did work, but just for confirmation whether it needs some tweaks.

[petespaco]

Are you actually using 3 different power supplies?

Do you plan on running the device on only 24 volts?

What maximum amperage/wattage are you expecting?

What Fres are you designing for, and why?

Pete Stanaitis
--------------

[Mads Barnkob]

The schematic looks good, this is the same way that I made it once, it was however not that successful, I kept burning one particular MOSFET in the setup and I can only think of  bad balance in the current sharing as the culprit. It was using two power supplies, 30VDC for the power circuit and 15VDC for driving the MOSFETs.

I think he just has 3 power supplies in the spice simulation, for the simulations/overviews sake.

[afk]

Are you actually using 3 different power supplies?

Do you plan on running the device on only 24 volts?

What maximum amperage/wattage are you expecting?

What Fres are you designing for, and why?

Pete Stanaitis
--------------

As Mads said this is just a simulation schematic. To view it easier, I modified the schematic a bit.

I'm expecting to use this at 50-60V for a consumption of at least 2kW. Since I'm still fiddling around with the work coil, I can't determine its inductance yet (my RLC meter isn't precise enough for this matter), but I'm aiming the freq to fall between 70-200kHz. It is used to melt copper and silver for metallurgy. Since I'm working with powdered metal, that frequency should be enough.

[petespaco]

"but I'm aiming the freq to fall between 70-200kHz. It is used to melt copper and silver for metallurgy. Since I'm working with powdered metal, that frequency should be enough.":
    This will be interesting to me, since the non ferrous metals that I have attempted to heat have done VERY POORLY so far with my Chinese 1000 watt device at anywhere from 43KHz up to about 113 Khz.  I can only get about 1 amp of NET current into a piece of 1/2" OD copper tube.  Same for a 1/2" X 5/8" solid tin bar.
So, at 48 volts input, that's only about 50 watts.
   I do see that people are melting small quantities of solder and aluminum, but they are using graphite crucibles.  In that case, I think it's the crucible that gets hot, imparting its thermal energy to the work.  There is one guy, "The Radio Mechanic", who seems to be successfully heating a small  pot of solder using a ceramic crucible.  He even has a cheap PID controller and thermocouple setup to control the induction heater's output in an on/off mode.
  From what little I have read, non ferrous metals need much higher frequencies and a LOT of power.