SEPIC converter: worst case ripple curent in coupled inductor primary and secondary?

[treez]

Hello,
The following article states how augmenting the leakage inductance of a coupled sepic transformer can reduce the coupled inductor's rms currents.

**broken link removed**

Supposing I augment the "leakage" inductance by adding an extra inductor in series with the primary, then the ripple of the primary gets "steered" to the secondary.
In such cases, how do I calculate the now increased ripple in the secondary coil?
(consider a 1:1 sepic inductor)

 

[FvM]

High ripple SEPIC ripple currents occur with closely coupled inductors and a switching frequency near to the resonance frequency formed by the leakage inductance and the SEPIC capacitor.

As I already mentioned in your previous threads, I don't believe that this is likely to happen in a real SEPIC converter, but some authors managed to make the leakage inductance unusually low to get the effect. We have heard different opions if a coupled inductor design can be considered as original SEPIC at all, but the majority of designers apparently approves. At least it's a popular concept.

Technical dual winding inductors made without particular attempts to minimize the leakage inductance get a coupling of 0.9 to 0.95, which should be fine for a coupled inductor SEPIC. So why should anyone "augment" leakage inductance with external inductors, except for pure academic purposes?

Technically, there's no essential difference between external and internal leakage inductance. Why do you expect increased ripple?

 

[treez]

Technical dual winding inductors made without particular attempts to minimize the leakage inductance get a coupling of 0.9 to 0.95, which should be fine for a coupled inductor SEPIC. So why should anyone "augment" leakage inductance with external inductors, except for pure academic purposes?

If bifilar wound, then the msd1583 coupled inductor shows that k>0.98 is usual....

MSD1583 coupled inductor datasheet
**broken link removed**

Technically, there's no essential difference between external and internal leakage inductance. Why do you expect increased ripple?

.......I agree there's no difference, (but please see * below) but bifilar winding gives k>0.98 and bifilar winding is the cheapest way to wind. So we use bifilar winding, and then we get too tight coupling, and so need to add an extra external inductor in order to bring down the resonance frequency of C(sepic) & "L(leakage)".

...when this extra inductor is added, say in series with the "primary", then you find that "primary" current ripple significantly decreases, and "secondary" current ripple increases...how do you calculate the actual increased ripple of the secondary when this extra inductor is added?

I am just wondering if one can consider the design using the same coupled inductor but used in flyback topology....and then one can say that whatever the "primary" and "secondary" current ripple was with the flyback, then the sepic's "primary" and "secondary" ripple currents would always be less then this?........basically I am trying to calculate I^2.R losses in the primary and secondary of a coupled sepic inductor when there's an extra inductor added in the way described.


* With coupled inductor sepics, adding an extra external inductor in series with the "primary", decreases the input current ripple, but adding the extra inductor in series with the 'secondary', decreases the output current ripple........basically , whichever coil you add the extra inductor in series with , means that the ripple that would otherwise be in that coil, will get "steered" to the other coil....internal leakage inductance amounts to adding an inductor in series with *both* primary and secondary, so doesn't give you this ripple "steering" effect.

 

[FvM]

If bifilar wound

Yes, that's what I call "particular attempts to minimize the leakage inductance". You won't do it to make a SEPIC inductor.

bifilar winding is the cheapest way to wind

No.

...when this extra inductor is added, say in series with the "primary", then you find that "primary" current ripple significantly decreases, and "secondary" current ripple increases...

How do you find that? Surely not unless the coupling factor becomes very low. So it's not a meaningful scenario.

 

[treez]

if you try it on the simulation you see its true...its called "ripple steering"....

**broken link removed**
..please see the explanation of ripple steering in the block of text just below figure 7.


Also, I cannot understand why you would not make a coupled sepic inductor with a bifilar wire?

 

[FvM]

..please see the explanation of ripple steering in the block of text just below figure 7.

I think the explanation is making the matter unnecessarily complicated.

Alternately, it is possible to construct a coupled inductor where the leakage appears on only one winding.

In simple words, it's a transformer with leakage and non-unity winding ratio.

 

[treez]

i'm speaking about the case whereby there is a k>0.98 coupled inductor (1:1) and there is an extra inductor added in series with the primary.....so its a 1:1 coupled inductor with the extra external inductor in series with it.

 

[FvM]

Just two equivalent circuits.