Asymetrical Interleaving Winding Technics, Advice Needed!

[Ultima Thule]

Hi,

this is pretty short question to any of our SMPS pro's on this board, I'm going to design a transformer and would like to know if I can instead of ordinary winding technics use an asymetrical interlieving?

Traditionally it's made up something like this:

Pri: 40 turns ::: Sec: N turns ::: Pri: 40 turns


but I want to do something like this:

Pri: 10 turns ::: Sec: N turns ::: Pri: 70 turns


The idea is to keep secondary windings as close to the "cold" side of the primary winding without loosing to much magnet decoupling between the primary and secondary winding while gaining decreased EMI levels to the secondary.

Cheers Michael

[mzzj]

Originally posted by Ultima Thule
Hi,

this is pretty short question to any of our SMPS pro's on this board, I'm going to design a transformer and would like to know if I can instead of ordinary winding technics use an asymetrical interlieving?

Traditionally it's made up something like this:

Pri: 40 turns ::: Sec: N turns ::: Pri: 40 turns


but I want to do something like this:

Pri: 10 turns ::: Sec: N turns ::: Pri: 70 turns


The idea is to keep secondary windings as close to the "cold" side of the primary winding without loosing to much magnet decoupling between the primary and secondary winding while gaining decreased EMI levels to the secondary.

Cheers Michael

Sounds like bad idea, proximity losses etc. increase.

Why not use screening(copper foil) between pri-sec connected to gnd instead of this?

[Ultima Thule]

mzzj,

can you explain what "pxoimity losses" means, and what are those other "etc." losses?

Do you really have practical experiences of asymetric interleaving and have made comparsions with symetrical interleaving?

Cheers Michael

[Eva]

This asymmetrical technique is used sometimes to achieve a given value of leakage inductance, since its value decreases progressively until it gets halved as one of the windings is progressively splitted until two equal halves are obtained.

I can't figure out any other useful purpose for that technique.

[Danko]

If it would be a push-pull SMPS, like the ones in car power-amps, then the transformer may saturate, becouse ofg the asymmetric currents in the primary. The outer half is 70turns, seven times more, than the other part (thus seven times more resistance), and - becouse the turn's radius is bigger - then one single turn's resistance is bigger, than the resistance in the other half.
Once, I wound a SMPS trafo, like you write: pri:::sec::ri, the primary windings were equal, but the mosfets weren't symmetrically switching. Here's the picture, tagen from the MOSFETs GATE's:
http://sziget.mine.nu/~danko/aramkor...d/12/szkop.jpg

Luckily, the trafo wasn't saturating, becouse I used a current mode controller IC, and the IC can do cycle-by-cycle current limiting.

[Ultima Thule]

Hi Danko,

I was probably not clear with my idea, the both primary windings are connected in series and the middle connection point is floating. Your talk about push pull it's clear case, there even fully symmteric is not enough without DC supervision.
BTW, as a note to your post, it's not so much the resistance that is the problem but the inductance asymetry the increase by the square, for instance 10:70 gives 1:49 ratio in inductance difference.

Hi Eva,

yes it's true that the leakage inductance will be affected especialy if no interleaving is used at all, eg. sec winded adjecant to one side of the primary winding.
My idea is that as long as we have at least one layer of primary the secondary is inside the primary winding and thereby the leakage inductance wont change so much.
My speculation is as soon as the secondary is only on one side of the primary winding the magnetic decoupling is exponentialy decreased and leakage inductance increased.

Like the gapped legg of a ferrite core, once we introduce a very small gap Al dropps dramaticaly compared to no airgap, incresing the gap even further and Al wont change much anymore, the same idea I had with the asymmetrical interleaving winding technic. I don't believe it's linear function of how we inteleave secondary between the primary, but this is also my question to all ya pro's!

I prefer to not use any grounded foils even if that is one technic to use.

Cheers Michael

[mzzj]

Originally posted by Ultima Thule
mzzj,

can you explain what "pxoimity losses" means, and what are those other "etc." losses?

Do you really have practical experiences of asymetric interleaving and have made comparsions with symetrical interleaving?

Cheers Michael

Proximity effect/losses result when your winding section is a lot more thick than skin depth. Check TI seminars on magnetics/winding desing for detailed info.http://focus.ti.com/docs/training/ca...?sku=SEM401014


No, not any practical experiences as i dont see much point doing that. And what is your problem with screening foils?

[Ultima Thule]

Hi mzzj,

thanks for the link, the material is known for me but a bit forgotten.. I actually attended TI's seminar in Helsinki, I think it was a year earlier or so, when they took up that material for the first time.

But that's the material I had in my mind in my first post and thats what a meant by "magnetical coupling"(not decoupling as I wrote first) between wires.
There seem to be some penalties if I move secondary into an asymmetrical interleaving, but it seem to worsen exponentially so it's probably much better to have at least one layer of primary windings on one side comparing to not interleaving!

I guess I need to make some practical tests.

Cheers Michael