Aliexpress Induction Heater

[david]

Hi All,
Is anyone familiar with the little induction heaters sold on Aliexpress?  The attachments show the item and what I believe is the actual circuit.  Not much to it - self-oscillating, (near) zero voltage switching and can have a screw driver glowing after a short time.  A friend gave me one to play with as he's looking for a more modest application of controlled, non-contact heating.
The only way to control the power would appear to be via the supply voltage but there's a couple of things I'm curious about.

1.  Frequency of operation is around 150kHz +/-50kHz.  In spite of the claimed zero voltage switching there would have to be harmonics well up in to the AM broadcast band.  Actual operating time will be quite short but could still cause interference.

2. The capacitors used in the resonant circuit are rated 630vAC 1200vDC.  YouTube has multiple DIYers using much lower voltage and smaller parts such as 250vDC.  Modelling of the circuit with realistic resistances added to the inductors shows the voltage is not so much a problem but the circulating currents can be 10-15A which is a heap of ripple.  I was wondering if a bank of high voltage chip ceramic capacitors could be used - about 1812 size.   Any thoughts on how to obtain about 0.5uF, 250V that's good for 10A?

Cheers,
David

[ken_k]

Hi David
I have found it is hard to produce 100% accurate circuit simulations with this type of circuit.
I'm guessing it could be done with SMD ceramic capacitors. The dissipation factor and calculated temperature rise would need to be considered.
Do you have a method of determining the temperature rise in the capacitors? All the quality chip capacitor manufactures have lots of data available. A few capacitors may be required in parallel. Can you measure the capacitors dissipation factor? It may pay to use large copper pads to heatsink the capacitors. Low ESR caps are also essential for high Q in the LC circuit. If you need power control this may not be the best circuit topology.

 
 

[david]

Hello Ken,
Yes, I've had problems before with similar circuits but I think this one seems realistic with the magnitude of the currents and voltages when resistance values are added to the inductors.
I don't have access to much equipment these days so I would need to rely on manufacturers data.  I was considering a bank of maybe 10 or more 1812 chip ceramic capacitors in parallel but I need to check if even this is enough.  Even name brand MKP capacitors don't usually rate ripple current past 50kHz.
The circuit is awful for power control but it is very simple and oscillates at the resonance of the tuned circuit whereas a driven system would need to be tuned on to frequency.
The price from Aliexpress is very cheap - I can't buy the 2 MOSFETs for that price, let alone the 2 toroids and the high voltage capacitors.  It's main problem is there's no mounting holes or even room to make holes on the board.  I think I'll need to do a little more research and then try the MLCCs.

Cheers,
David