Are both mosfets IRF250P? The connections are strange on the image in post #11. Working coil should be connected to the drain of each mosfet. This is done with two yellow wires. One yellow wire is connected to the right side pin of the mosfet. On another mosfet, it is connected to middle or left side (can't see clearly). Should be the same position.

Also black "GND/0 Volt" wire from the power supply. It should be connected to source of both mosfets. It should be a same pin on both mosfets, but now it seems to be connected to different pins.

Yes absolutely right. I didn't bother looking at the source and drain pins and thought both MOSFETs are in same orientation. But the lower one is vertically flipped. Thanks for looking into this so closely.
Will give update after changing this connections.

 

Inductance is 95 to 100uh not sure about the actual value because i don't have a good LCR meter for this. I used transistor tester which gives a reading of 98uh when connected, can't trust these devices really. And the core saturation current is unknown to me. As said, earlier the core belonged to a online UPS rated for 3kva so it must be a beefy one, i suppose.

Scary. You are just guessing.
Once you saturate the core, the inductance will drop substantially.
This may have failure-inducing effects on your oscillator.

 

Scary. You are just guessing.
Once you saturate the core, the inductance will drop substantially.
This may have failure-inducing effects on your oscillator.

Yes I'm aware of the scary part but i can't afford a good LCR meter right now. It's just a hobby to me and someday in future I'll get one, but until then transistor tester is the way to go.

 

Looks like the Mosfets are wired up Wrong. can you show better pics of the mosfets?!

 

Looks like the Mosfets are wired up Wrong. can you show better pics of the mosfets?!

Yes they were.. i have corrected it and not it makes hissing sound and voltage drops to 1.2V.

 

Updated

 

i can't afford a good LCR meter right now.

There are some reasonable LCR meters available cheaply. They don't have fancy cases but work very well. They are available from many places, for eample Amazon.
https://www.amazon.co.uk/Transistor...argid=aud-857761465096:pla-618756375016&psc=1

 

There are some reasonable LCR meters available cheaply. They don't have fancy cases but work very well. They are available from many places, for eample Amazon.
https://www.amazon.co.uk/Transistor...argid=aud-857761465096:pla-618756375016&psc=1

Yes, the same one i have, in review videos on YouTube they said, these meters are not very accurate but can get you a reading which is not off by +-5%. That's why I said inductance value between 95uh to 100uh.

 

I connected all 3 power supplies in parallel and after 40 to 50 sec, both irfp460n died. Their continuous drain current is only 20amps and what i supplied is way too much. But the coil and the bench seem to vibrate when i power on so might be oscillating and then MOSFETs failed.
I checked the voltage while it was running, i measured 7.6V. so i need more current to make it work.
But i don't understand why this thing is requiring so much current for just oscillating with no load. Also how much current will it require if i want it to work? Anyone have any Idea of that?

 

Yes, the same one i have, in review videos on YouTube they said, these meters are not very accurate but can get you a reading which is not off by +-5%. That's why I said inductance value between 95uh to 100uh.

But there are few LCR meters that can measure saturation current around 30 A - this is MUCH more critical than the measured inductance value.

 

But there are few LCR meters that can measure saturation current around 30 A - this is MUCH more critical than the measured inductance value.

Yes, now I'm leaning towards the inductors getting saturated and hindering the oscillations.
Also read #29.

 

Ideally the Capacitors and Heating coil are tuned to the resonant frequency, your capacitors add up to 99uF , if the inductor is 95uH, that's a frequency of 1.7Khz, which is far too low , I would expect a frequency of 80 to 150Khz..which means lower the capacitance to 1uF to get 15Khz.

 

Ideally the Capacitors and Heating coil are tuned to the resonant frequency, your capacitors add up to 99uF , if the inductor is 95uH, that's a frequency of 1.7Khz, which is far too low , I would expect a frequency of 80 to 150Khz..which means lower the capacitance to 1uF to get 15Khz.

No, i have 3 caps of 1uf each. So total capacitance is 3uf and the inductance which decides the resonant frequency is the inductance of work coil, which i made according to the original author's design, which gives a inductance of 0.5uh.
Putting this in resonant frequency formula gives a frequency of 129 something khz.
This part is right. Its Just the driver which will drive it to resonance is not working.(i think)

Also, i earlier said the capacitor have 1uf 630VDc written on it. Is the dc capacitor making any effect on oscillations?

 

http://kaizerpowerelectronics.dk/general-electronics/royer-induction-heater/

Design I'm using.

 

The one I got from Ebay has 6 capacitors on it. 3 for each inductor, and not looking like the electrolytic ones your using. The ones on mine are MPK box type film caps. Randomly substituting things into a tuned system like this isn't the way to get a working circuit. The same with using an unknown toroidal core, just because it's round. Thins matter.

 

The dc in the marking is it something to consider?

Definitely. If those are electrolytic capacitors they don't like AC and will be damaged by it !

 

The one I got from Ebay has 6 capacitors on it. 3 for each inductor, and not looking like the electrolytic ones your using. The ones on mine are MPK box type film caps. Randomly substituting things into a tuned system like this isn't the way to get a working circuit. The same with using an unknown toroidal core, just because it's round. Thins matter.

Here is the link to the capacitor i used
http://m.elciarcapacitor.com/kt22-film-foil-polyester-axial-capacitor-3908858.html
These are 0.1uf but mine is 1uf and 3 in parallel. Not the electrolytic you assumed.
I don't have exact details of the toroidal core but i have a photo.

The numbers hiding behind the wire is 07. I wrote it into my note before unwinding it, that's how i know.

These were the looks of it

 

Definitely. If those are electrolytic capacitors they don't like AC and will be damaged by it !

They are film foil polyester capacitor. Leads are placed axially. Not electrolytic.

 

I'm going to buy new inductor cores and new capacitors.
What are the things that I have to consider while buying?
I read some articles and found this. For capacitor bank use fkp capacitor instead of mkp, rated for voltage higher than 250vac.
For inductor core I'm not sure what to do. At electronics stackexchange(answering someone's questions) one guy said to yellow colour cores are for filter and such, so to avoid. Black ones are ferrite and are suitable for induction heater.
When i unwound the core, i used angle grinder to remove the hard epoxy and accidentally grounded some of the core. It was shiny and looked more like a iron (saying so because when grinding iron it leaves shiny surface) so most probably inductor cores I'm using are iron power ones not ferrite. E shaped cores are better for high current application but suffer from losses. What should I buy now?

 

Based on the circuit you posted at the beginning of the thread, I have had good results by eliminating D1,D2,R1, and R2 and biasing he MOSFETS through a 470 Ohm (R3 and R4) to a 12V-15V supply. As long as you keep the bias voltage below the maximium MOSFET gate voltage (typically 20V), you don't need the Zener diodes D1 and D2. So the main HV power path goes through L1 L2 ... through the LC and through Q1 and Q2 to GND. The negative swing on the LC oscillator feeds through D3 and D4 to turn OFF the respective opposite MOSFET.