magnetizing current in transformer

[scottieman]

Dear All,

I need to build a step-up DC-DC converter using push-pull topology. The following link just shows you which topology exactly I am using. (**broken link removed**)

I have read alot of documentation about push-pull converter. However, it seems no one cares about the magnetizing current of the transformer (or the current due to the inductance of transformer).

The following is my problem (please refer to the above link):

When T1 switched on, the secondary side of transformer action forward biased diode D2 and the inductor current iL ramps up. At the same time, the magnetizing inductor is charged through T1 too. When both switches (T1 and T2) are off, the inductor (L) in secondary side is freewheeling with the help of diode D1 and D2.

However, where DOES the current of magnetizing inductor go?? CAN anyone helps me to understand this issue.

Thanks alot
Scottie

 

[armai]

When T1 is off, the secondary side is shorted because D1 and D2 are both conducted. So if you model the transfomer as a inductor paralleled with a ideal transformer, you can see that the magnetizing current is freewheeling in the loop consisted of magnetizing inductor and the primary side the transformer.

 

[scottieman]

Thanks. However, I have the following questions:

If the magentizing current really goes through the body diode of switch T1 or T2 (depends on the previous state) when both switches, T1 and T2, are off, it means there are voltage drop across the transformer primary side. The magnitude of the voltage acorss transformer equals to Vin (without accounting diode drop), and the time for this voltage to exist is D*T (D is duty ratio, T is the period of switching frequency), which is the time required to de-magnetizing the transformer

Since any voltage across one winding of a transformer, this voltage will reflect to the other windings acrroding to the turn ratio. Hence, the transformer secondary sees a voltage pulse with 2*D on-time rather than 1*D.

If what I guess is correct, it makes me confusing @_@

Scottie

 

[armai]

No, that's not true. The magnetizing current never goes throught the body diode of T1 and T2.

When T1 is off, the magnetizing current is freewheeling in the magnetizing inductance and the primary side of the transformer, as I said before.

And at the same time, the leakage inductance and the parasitical capacitance of T1 start to resonant under the voltage Vin, because the transformer is shorted by the secondary side, so Vin is all applied to the Coss and Lleak. And that's why a ringing can be seen on the MOSFET DS.

armai

Added after 49 seconds:

The transformer is de-magnetized when T2 is on.

 

[scottieman]

I get it now

Thx armai

Added after 14 minutes:

Still one question, Armai

as you mentioned before

"When T1 is off, the secondary side is shorted because D1 and D2 are both conducted. So if you model the transfomer as a inductor paralleled with a ideal transformer, you can see that the magnetizing current is freewheeling in the loop consisted of magnetizing inductor and the primary side the transformer."

Since the total current going into a transformer (when the current is normalized by its turns) should be zero, so the freewheeling magnetizing current should appears somewhere else...... THIS MAKE ME CONFUSE AGAIN......

Hope someone can clarify for me

Thanks
Scottie

 

[armai]

The current going to the transformer is zero. This model just helps you understand how the circuit works.

 

[scottieman]

The current going to the transformer is zero. This model just helps you understand how the circuit works.

Yes, I understand model is "model"

At least to me, the inductor currents need to freewheel through a path. And my question is, I cannot figure out THIS path.

Can anyone give me some hint

THX alot
Scottie

 

[armai]

There's no current in the transformer primary winding, and the inductor current, flow through both secondary winding, the current in each winding is iL/2

 

[scottieman]

There's no current in the transformer primary winding, and the inductor current, flow through both secondary winding, the current in each winding is iL/2

Sorry that I did not clarify it clearly. Let me do it again.

The magentizing current (from the magnetizing inductance of the transformer) need to freewheel through a path when ALL the MOS switchs at primary side are OFF. As you mentioned before, the transformer model tells us that the magnetizing inductance are short when both MOS switchs at primary side are off. Hence, the current in the magnetizing inductance can be freewheel.

However, any current go into a port of a transformer must be "transform" into other ports according to their turn ratio. When that current of the magnetizing inductance is freewheel through the transformer primary side, it means that this current must appear at other port of that transformer. However, I cannot figure out where DOES that magnetizing current really go during freewheel.

Can anyone give me some hint

Scottie

 

[jetty]

I think that it is convenient to observe the magnetizing current
of transformers under the condition of no load. I also encountered
the problem during making a push-pull bus converter.Comparing to
upper,the different point is that my magnetizing current is tested
without load.I observed that the current of two primary windings
was commuted through the two primary windings ,the Coss of T1 and
T2 during the deadtime.At the same time, the current of transformers
center-tap existed also. I don't know why? Their waveform is attached file.

Added after 2 minutes:

Attached file

 

[scottieman]

I think that it is convenient to observe the magnetizing current
of transformers under the condition of no load. I also encountered
the problem during making a push-pull bus converter.Comparing to
upper,the different point is that my magnetizing current is tested
without load.I observed that the current of two primary windings
was commuted through the two primary windings ,the Coss of T1 and
T2 during the deadtime.At the same time, the current of transformers
center-tap existed also. I don't know why? Their waveform is attached file.

Added after 2 minutes:

Attached file

Jetty
Really thanks for your sharing on this topic.
do you mind telling me more on your testing condition (e.g. the schematic you used, the type of transformer you are using and what is your meaninng of no-load)

Thanks
SCottie