I wish I had a dollar or an euro every time someone comes with a "my circuit does not work", but fails to provide any meaningful information.
All I can say at this point of time, in my previous company, which at one point was the world's largest power supply company, we used the 3842 everywhere. We literally used millions of those devices a year. It has its own idiosyncrasies like any IC, but it is relatively straightforward to use and very reliable.
thank to all i burn about 16 transistor :-x and 6 mosfet :-(
the transistor is burn in less than 1 second when i test it with 220v.
but the circuit work well when i work only with 24v opcorn: .
i use current limiter when i work with the main power 220v "40w lamp as resistor"
i powered the 3842 by 24v "external power source"
i use about 100 turn primary when i work with 220v.
Without a load, high flyback Vce can easily kill the transistor. The snubber is only good to absorb the leakage inductance energy, not the full flyback energy. The other point is that the 1K base resistor value will put the transistor into linear operation with huge losses.
I don't know which test is intended with this circuit, but it should consider the transistor's safe operation area. Definitely no "3842 problem".
thank you i was think the snubber circuit for high flyback Vce.
but when i work with 220v it act as short circuit are this duty cycle problem. and with 24v it work well.
thank you i was think the snubber circuit for high flyback Vce.
but when i work with 220v it act as short circuit are this duty cycle problem. and with 24v it work well.
From my experimental simulations, you need a larger capacitor and smaller resistance. The snubber capacitor must absorb an energy spike quickly, and the resistor must dissipate it gradually.
My simulation is only a rough match to your setup. It's to show how the DCR snubber acts.
The load gets 40W.
Notice that voltage on the capacitor is kept within a manageable range. The resistor dissipates peak power of 8.64W.
Without any feedback, how do control the duty cycle?
Do you realize that the 3842 is a current mode controller, and that the peak switch current is controlled that way?