primary winding for push-pull converter @100amps, 100kHz

[vena2a]

Hello,
I am new in here, but I want to show you something. It is a 100 amps primary winding for push-pull converter which should work at 100 kHz.



This winding is placed on ETD49 core and is designed of 0.5 mm thin copper plate. I decided for this solution because depth of penetration is roughly 0.2 mm (at 100 kHz) and I wanted to reach small resistance of winding.
As you can see on the picture, the winding consists of two similar parts (each part makes two turns with center). These two parts are connected in parallel into one winding.

So, what do you think bout this design.

[jamesrnz]

interesting, do you split the bobbin?

do the phase a phase b winding go in opposite directions?


are you insulating the foil completely?

[areza]

winding a transformer is challenging task , using foil to wind is more challenging, this one really looks like most challenging,

bobbin seems not ok with foils, need some mods,

those connecting foils coming out from winding need to be replace with round wire, so they can be bent proper direction with out breaking ,

[vena2a]

to jamesrnz:

No I won't split the bobbin. This winding can be made of E-shaped 0.5mm copper foil (the foil will be turned on the bobbin).

Of course, the winding will be insulated, except the ends and center of winiding. The ends and center of winding will be soldered on PCB ... sometring like SMD.

The phases windings (which are in fact connected in parallel into one bigger winding) are the same. The windings have the same directions ... they are little bit (about 40 degrees) turned each other. You can see that on the picture. The first turn has red pin on both of windings ... so the directicrion is the same.

[sawreyrw]

If the skin depth is .2mm, a .5mm winding is thicker than necessary (or effective). Winding a thiner tape should be easier. The width of the tape needs to be sufficient to keep the current density (and I^2 R losses) low enough to prevent excessive heating of the winding.

[jamesrnz]

does not the skin depth penetrate on both sides of the tape 0.2 x2=0.4 so 0.5 wouldn't be far off?

[sawreyrw]

The relatioship between skin depth and Rac/Rdc is addressed in this paper. While the current wants to flow on both surfaces of the winding, proximity effects complicate this.

http://focus.ti.com/lit/ml/slup197/slup197.pdf

[Eva]

Have you considered one or more small toroids, say 28mm in diameter, with as much strands of 0.5mm magnet wire as required? This seems much easier to manufacture since only conventional parts and methods are used. No wonder it's the preferred approach in car audio amplifiers. E and ETD cores are seldom used for 12V push-pull applications.

[homemodder]

Hi Eva

I understand you know a thing or two about switchmode supplies. Could you possibly direct me to what you would call a excellent diy design here on these threads for a car supply. Ive built plenty on the design as on rod elliots pages, not exactly the same but as per schematics given in elektor 1994 issue. Im looking at high efficiency and higher power. Also would one get more effenciency from a torroid than from a etd type former?? I used litz wire on my etd formers but some time back I read somewhere on the net that its better to use a larger diameter wire because of the skin effect, this confused me as its contradicting what is said in the elector article.

[Tahmid]

Hi Eva,

E and ETD cores are seldom used for 12V push-pull applications.

Just a query: Why are E and ETD cores seldom used for 12v push-pull applications?
If you please explain, it will be helpful to me.

Thanks.