AC Losses Transformer Foil WInding

Thread Starter

Tom32

Joined Dec 5, 2018
10
Hello.
I have a question regarding AC losses from transformer windings. I have a transformer with a winding ratio of 1:1 and realize the primary and secondary side each with a copper foil.
To eliminate the skin effect I want to divide the foil winding of the secondary side into 4 parallel foils. I lay the 4 foils on top of each other and connect them in parallel (secondary winding remains 1). How do I determine the number of layers for the secondary side? Do I now have 1/4 layer per parallel foil and thus 1 layer in total or does the secondary winding now consist of 4 layers?
I would like to determine the Rac/Rdc ratio, but have no idea how the number of layers (p) is determined in the parallel arrangement.
I use the formula from the attached graphic. Where Delta is defined as: Thickness/Skin depth
 

Attachments

shortbus

Joined Sep 30, 2009
7,685
To eliminate the skin effect I want to divide the foil winding of the secondary side into 4 parallel foils
You don't say what frequency your using on this, but regular 50 - 60Hz doesn't have enough skin effect to bother about.

Read here about half way down the page to find the mitigation techniques - https://en.wikipedia.org/wiki/Skin_effect

From the same link -

Skin depth in copper
Frequency Skin depth (μm)
50 Hz 9220
60 Hz 8420
10 kHz 652
100 kHz 206
1 MHz 65.2
10 MHz 20.6
100 MHz 6.52
1 GHz 2.06
 

Thread Starter

Tom32

Joined Dec 5, 2018
10
In the higher frequency range up to 100kHz.
Yes, I know the skin effect. I'm more interested in the parallel distribution of the copper foils. The problem is that the thickness would be too big, so I want to connect several foils in parallel. But I would like to judge my AC/DC ratio first. Unfortunately, I'm not sure how to determine the number of layers correctly. In the literature it is always assumed that the windings are divided and the respective foils are connected in series.
 

shortbus

Joined Sep 30, 2009
7,685
Don't know or understand what you're trying to do or figure out. What is this foil wrapped transformer used in or for? Never saw one or even heard of one. High frequency transformers are usually wound on ferrite cores and don't know how you would do this with foil.
 

Thread Starter

Tom32

Joined Dec 5, 2018
10
Don't know or understand what you're trying to do or figure out. What is this foil wrapped transformer used in or for? Never saw one or even heard of one. High frequency transformers are usually wound on ferrite cores and don't know how you would do this with foil.
The Winding is build with copper foil oder copper sheet (sandwich winding) on a ferrite core. The Transformer is used for LV High Current applications.
 

shortbus

Joined Sep 30, 2009
7,685
The Winding is build with copper foil oder copper sheet (sandwich winding) on a ferrite core. The Transformer is used for LV High Current applications.
Can't see how that can even be, if LV stands for 'low voltage'. High current takes big wire or conductor, and lower voltage takes an even larger one for a similar current. Please give an example of where this transformer is used.
 

Thread Starter

Tom32

Joined Dec 5, 2018
10
Yes, LV stands for low voltage. The Secondary winding is made with copper foil/sheet and used for high power DC/DC converters.
 

shortbus

Joined Sep 30, 2009
7,685
The Secondary winding is made with copper foil/sheet and used for high power DC/DC converters.
You must have a different idea about high power. Most DC-DC converters I've looked at are less than ~5Amps, not what I would consider high power. But you may be different.
 

MisterBill2

Joined Jan 23, 2018
6,065
What is the motivation for the change you are asking about? I predict NO SUCCESS for the rewinding operation, and not even any theoretical benefit even at 400Hz. I have wound transformers with 1 inch wide tape foil and they worked very well at 60 Hz.. Normally, to reduce transformer losses at higher frequencies a material known as LITZ WIRE is used, and where needed it serves very well.
Is the transformer getting very hot? That is an indication of handling a lot of power, but they do get hot because efficiency is never 100%. So there are a lot of questions to be answered.
 

MisterBill2

Joined Jan 23, 2018
6,065
One thing not previously mentioned is that there will be the same surface area when the ribbon is divided into four ribbons, and so the skin effect losses will not be reduced. BUT the connection losses will be increased and the thermal resistance will also be increased. Thus all of the effort spent modifying the transformer will produce no benefit at all, if you are successful at completely rewinding the transformer. The change that may be of some value would be to replace the present laminations with some of a better magnetic quality, if the present laminations are not of the best quality already.
But I am still asking why you think that the transformer needs to be improved. I have not seen any statements as to what is observed that would indicate that changes need to be made to reduce skin effect.
 

Thread Starter

Tom32

Joined Dec 5, 2018
10
The problem is that the copper sheet has a thickness of 2mm. At a frequency of 100kHz the penetration depth is about 0.21mm, so the current density at the edges would be very high. I hope to reduce the skin effect by dividing it into thinner sheets. The sheets are insulated against each other with Kapton foil.
 

MisterBill2

Joined Jan 23, 2018
6,065
There is more to the design of a transformer than just skin-effect losses. The rest of the copper serves as a heat spreader for the heat generated in the effective resistance. Also to be considered are capacitance and heat transfer. So my suggestion is to consider creating a new transformer with high frequency steel laminations, and leave the existing transformer be. And another consideration is the frequency, which 100KHz seems quite high a frequency for a larger transformer. I would consider a much lower frequency to avoid a lot of switching losses.That may offer a greater improvement for a lot less effort.
 

Thread Starter

Tom32

Joined Dec 5, 2018
10
Yes I know, but I would like to look at different design possibilities. Also, ways to reduce the skin effect and later compare the results.
 
Top