Multiple output flyback SMPS and primary leakage inductance

[treez]

Multiple output flyback SMPS and prmary leakage inductance

Page 5 (the top left of page 5) of the below document (AN1679/D “How to deal with leakage elements in flyback converters”) says that where a flyback transformer has multiple outputs, then when the outputs are more closely coupled with each other, the leakage inductance as “seen” from the primary actually increases.

However, the diagram of a multiple output flyback transformer that they show (figure 9 , page 4) actually has three “Limbs”. Does the above statement still apply when all the secondaries and the primary are wound round one single limb?....(as is the case with most low power flyback smps transformers.)

AN1679/D:
https://www.onsemi.com/pub_link/Collateral/AN1679-D.PDF

 

[FvM]

Re: Multiple output flyback SMPS and prmary leakage inductance

Three-limb cores are used in three phase transformers and transductors. I didn't yet see it in any standard SMPS. Figure 9 and the related comments make me really doubt about the quality of this application note.

 

[treez]

Re: Multiple output flyback SMPS and prmary leakage inductance

Three-limb cores are used in three phase transformers and transductors. I didn't yet see it in any standard SMPS.

Thanks yes, the three limb core is virtually never seen in low power , offline flyback SMPS transformers, so it is surprising that a diagram of a three limb transformer appears in an app note about flyback transformers.
Though do you then disagree that closer coupling of the secondaries means more leakage inductance "seen" from the primary?

 

[smijesh]

Re: Multiple output flyback SMPS and prmary leakage inductance

The paper says
This paper reviews the magnetic and electric models of the two winding and three winding transformers

They considered C core winding (Cross connected) is two winding transformer and E core winding is three winding transformer.
There is nothing wrong on this paper. They thematically studied/compared the leakage inductance of the above two methods.

Three winding method is good for low leakage application but no one preferring such winding because
1. More cores section area required compared to center limp winding(Conventional)
2. Low magnetic shielding effect

Three phase transformer cores are different stuff, its all three limps are equal in cross section.

 

[FvM]

Re: Multiple output flyback SMPS and prmary leakage inductance

There is nothing wrong on this paper. They thematically studied/compared the leakage inductance of the above two methods.

Nothing wrong, but not related to commonly used SPMS transformer geometries.

Three phase transformer cores are different stuff, its all three limps are equal in cross section.

Different stuff, yes. If you ever modeled three phase transformers or inductors, you'll know why I brought the comparison.

The three limb transformer core with separate windings on each limb has a very specific behavior (doesn't matter in a first order if equal or unequal cross section). There may be cases where this specific behavior is exactly wanted in a SMPS design. But bringing it as example why leakage inductance in multi-output transformers would be unintentionally higher than in single output is just misleading.

Three winding method is good for low leakage application

May be I missed the point in the paper that claims that the three windings on an E core topology is there to achieve low leakage. Can you guide me to find it?

 

[treez]

Re: Multiple output flyback SMPS and prmary leakage inductance

i must admit it is a little bizarre having a diagram of a three limb transformer in a paper about flyback smps transformers.
It makes me wonder if the bit about leakage as "seen" from the primary really being more if the secondaries are more closely coupled, is true or not (in a single limb, eg ETD, flyback smps transformer)

 

[FvM]

Re: Multiple output flyback SMPS and prmary leakage inductance

The lower bound for leakage inductance is the inductance of the respective air core transformer. A core can only increase it.

From this consideration, it's quite clear that winding separation generally increases leakage. Bringing windings close together reduces it, interleaving partial windings can further decrease it. Some geometric formulas are derived in Snelling Soft Ferrites.

 

[treez]

Re: Multiple output flyback SMPS and prmary leakage inductance

Thanks, supposing I have a flyback smps transformer on say an ETD core.
And I have three equal but separate secondaries and one primary (but the primary is of two halves connected in series for interleaving purposes.)
Which gives least leakage as “seen” from the primary?................
P_a / S1 / S2 / S3 / P_b
or….
S1 / P_a / S2 / P_b / S3
……According to Basso, the top solution gives the most leakage as seen from the primary, if I have interpreted him correctly in the article in the top post.
Which one do you think gives the most primary referred leakage inductance?

 

[FvM]

Re: Multiple output flyback SMPS and prmary leakage inductance

According to Basso, the top solution gives the most leakage as seen from the primary

I expect the same. There's a possible ambiguity if you look at average leakage inductance or maximum leakage inductance to individual secondary windings.

In practice, we'll often prefer the first solution though, because it has only two compared to four HV isolation interfaces if high voltage is only between primary and secondary.

 

[treez]

Re: Multiple output flyback SMPS and prmary leakage inductance

n practice, we'll often prefer the first solution though, because it has only two compared to four HV isolation interfaces if high voltage is only between primary and secondary.

Thanks, i see your point, because it is mains connected (100-265VAC) though i am thinking that if we can do the primary in TEX-E triple insulated wire then the boundary will be taken care of. (the secondaries would be in plain old enamelled copper wire)
Each secondary is "separate", but they are 'stacked' after the diode to give a 340v total output. (each secondary is 113.3V).