ringing output of transformer

[franticEB]

Hi i have a problem with this circuit:



on output of the first transformer (OUT1) i measure an important ringing in corrispondence of the closing of the mos as in figure.



The problem is that i want to measure a DC voltage of 4 Volts (result of transformer ratio) where i draw a red arrow (OUT4), but i measure the voltage given by the peaks of the square wave.

The switch is controlled by microcontroller: when open through an A/D tre micro reads 0 of pull-down 10k resistor, when closed 7.5 Volts.

I tried to solve the problem putting a SNUBBER circuit on primary of both transformers but it doesn't delete the disturbance. 8-O

How can I do in order to avoid this problem?

 

[jiripolivka]

I do not see the purpose of the rectifiers between the two transformers. I only pushes a strong DC components through them, and their cores may saturate.
Overshoot with ringing is usually due to a reduced circuit transient damping. If a snubber across the first transformer did not help, then try to connect a variable potentiometer across the second transformer primary, or remove the two diodes and replace them with the variable resistor.
Long connections can also contribute, as well as the parameters of the two rectifiers between the two transformers. Still unknown their purpose?
I am missing a capacitor across the Vcc for the MOSFET switch. Is there a magnetic coupling between the two transformers? How did you measure the output voltage? Your oscilloscope probe can pick up transients showing up as ringing.

 

[FvM]

Even the winding direction isn't obvious from the schematic. I would expect forward operation of the output side, but it's not clear.

Ringing at the forward edge is a matter of leakage inductance in combination with transformer self-capacitance. It can be avoided by making a better transformer.

A better snubber effect would be expected for a snubber placed at the secondary in this case.

 

[franticEB]

I do not see the purpose of the rectifiers between the two transformers.

The purpose is explained in this figure



i think that the purpose is the decoupling of two transformers
the circuit is a forward converter that feeds a circuitry on other board.
i draw a red cross because the effect of the third transformer is ininfluent to the overshoot (i try to remove this connection and the peaks remain)

I only pushes a strong DC components through them, and their cores may saturate.

I don't understand...

- - - Updated - - -

It can be avoided by making a better transformer.

A better snubber effect would be expected for a snubber placed at the secondary in this case.

I try it also but not very good results....

 

[FvM]

A possible problem of the circuit and it's operation mode is that the you don't allow the inductor to demagnetize before the next pulse is applied. Thus the flyback diode's reverse recovery current step happens during rising pulse edge and might cause the observed overshoot.

You either need < 50% duty cycle or n2 > n1.

 

[jiripolivka]

I think the circuit is exceedingly complex for a given purpose to generate 4V DC. The user does not understand how the intermediate part works ... neither do I.
I would suggest to use only the first transformer, and adjust the secondary winding count to get the output voltage desired.

Again, check the oscilloscope probe, it may pick the transients by itself. Switching at 100 kHz generates a lot of harmonics over 10 MHz.

 

[franticEB]

A possible problem of the circuit and it's operation mode is that the you don't allow the inductor to demagnetize before the next pulse is applied. Thus the flyback diode's reverse recovery current step happens during rising pulse edge and might cause the observed overshoot.

You either need < 50% duty cycle or n2 > n1.

Tha waveform on the Drain of the mos, so on the primary winding of the transformer, is quite good and doesn't show any oscillation.
So i put the probe on the cathode of the diode of recovery winding and there is a little oscillation that becomes bigger on the secondary.

I try to make some simulations and i noticed that i can replicate this behaviour if i put a capacitor (200-300pF) in parallel with this diode....

Any idea? PCB? Thanks

 

[FvM]

My explanation in post #3 still holds. Ringing is more or less normal behaviour if the transformer has some amount of leakage inductance.

 

[franticEB]

You either need < 50% duty cycle or n2 > n1.

Is there a way to calculate the number of turns of N2 in order to avoid the problem described before?

 

[franticEB]

A possible problem of the circuit and it's operation mode is that the you don't allow the inductor to demagnetize before the next pulse is applied. Thus the flyback diode's reverse recovery current step happens during rising pulse edge and might cause the observed overshoot.

You either need < 50% duty cycle or n2 > n1.

in order to demagnetize the inductor have i to increase or decrease the number of turns of recovery winding? if increasing the value of the inductance does not increase also the recovery time?
Is it correct to make N2 half the value of N1?
Thanks