# Coupling factor question

#### airplane100000

Joined Aug 2, 2016
68
I had a Chinese firm construct the auto-transformer (3-pin inductor) seen in the image below.

The inductances of its windings, and its physical dimensions are according to my specifications, but I don't believe that they are necessary with respect to this question.

The primary winding is on the left; and secondary is on the right. Notice the gap between them...
The coupling coefficient is relatively low (not surprisingly), at 0.54. This is likely too low for my application, and needs to be increased.

My question is as follows: What should I instruct them to change in order to increase the coupling between the windings? My intuition tells me that the primary should be wrapped around the secondary. Would this help? Anything else?

#### BR-549

Joined Sep 22, 2013
4,938
It depends on all the details and application. An auto transformer is a tapped coil. One sets the voltage ratio with input and output taps. An adjustable output tap is quiet handy. As a general rule.....if the wanted output voltage is more than 3 times the input voltage........one should consider a conventional two winding transformer.

There are several factors and several adjustments that control the coupling. It will depend on the application.

#12

#### #12

Joined Nov 30, 2010
18,210
My first instinct is to use a toroid or a pot core because they conserve the magnetic field better than a stick shape, but BR-549 is correct when he says you have left out a lot which is needed to find the optimum shape and core. Some of them are frequency, current, and impedance required.

#### airplane100000

Joined Aug 2, 2016
68
It depends on all the details and application. An auto transformer is a tapped coil. One sets the voltage ratio with input and output taps. An adjustable output tap is quiet handy. As a general rule.....if the wanted output voltage is more than 3 times the input voltage........one should consider a conventional two winding transformer.

There are several factors and several adjustments that control the coupling. It will depend on the application.
Thank you for the replies, though perhaps I posed the question unclearly:
The following constraints are given:
Must be an auto-transformer
Dimensions of core cannot be changed
Current turn ratio (as seen in image) must be preserved.

So essentially my question is: How can the transformer in the image be wound differently, to preserve the turn ratio and respective inductances while increasing the coupling coefficient? My theory was to wind the primary around the secondary...

I'm talking about datasheet values here, as obtained by a test in a lab, without regard to any specific application. (Without considering any specifics such as frequency, current, etc...).

#### atferrari

Joined Jan 6, 2004
4,117
I'm talking about datasheet values here, as obtained by a test in a lab, without regard to any specific application. (Without considering any specifics such as frequency, current, etc...).
I know basically nothing to replay to your questions but you ask for details and you don't give details as #12 asked. The application is to be known.

#### R!f@@

Joined Apr 2, 2009
9,734
Yup!
Wind the secondary on top of the primary.

#### airplane100000

Joined Aug 2, 2016
68
Yup!
Wind the secondary on top of the primary.
Thanks,
Why secondary on top of primary? I would think primary on top of secondary, because primary has less windings.
Also would you think that the primary winding (when on top of secondary) should be spread out to cover the whole secondary, or should it be wound tightly, like it is currently?

#### R!f@@

Joined Apr 2, 2009
9,734
Soooo many questions again.
OK..!
Wind Primary over secondary tightly.

#### EM Fields

Joined Jun 8, 2016
583
I had a Chinese firm construct the auto-transformer (3-pin inductor) seen in the image below.

The inductances of its windings, and its physical dimensions are according to my specifications, but I don't believe that they are necessary with respect to this question.

The primary winding is on the left; and secondary is on the right. Notice the gap between them...
The coupling coefficient is relatively low (not surprisingly), at 0.54. This is likely too low for my application, and needs to be increased.

My question is as follows: What should I instruct them to change in order to increase the coupling between the windings? My intuition tells me that the primary should be wrapped around the secondary. Would this help? Anything else?

Why not have them wind a bunch of them with different configurations and then test them and choose the one you like the best?

#### R!f@@

Joined Apr 2, 2009
9,734
The problem with OP is he is keeping the reason a secret.
So I told him what he wanted to hear. But now he wants it the other way.

I dunno the core tyoe or for what it is used for. So I can only say what he wants to hear.
Thus my reply changes how OP wants.

Limited info will get stupid or useless answers. If you want help, answer the questions asked by the members

#### R!f@@

Joined Apr 2, 2009
9,734
And this is the best part.
I'm talking about datasheet values here, as obtained by a test in a lab, without regard to any specific application. (Without considering any specifics such as frequency, current, etc...).

How can a data sheet be taken without values and application in which it will be used for ?

#### BR-549

Joined Sep 22, 2013
4,938
AN auto transformer is a different animal than a regular transformer.

A tapped coil does not have a coupling coefficient. A coupling coefficient is between two separate windings inductance's.

It's the interaction between two fluxes. An auto transformer only has one flux. Therefore no flux interaction.

One can vary the coupling co-efficiency between two windings thru several different methods.

It depends on the application.

Last edited:

#### EM Fields

Joined Jun 8, 2016
583
I'm talking about datasheet values here, as obtained by a test in a lab, without regard to any specific application. (Without considering any specifics such as frequency, current, etc...).
Since frequency, current, and etc. affect the inductance, how do you propose to run a proper test without considering them?

#### airplane100000

Joined Aug 2, 2016
68
@R!f@@ Thanks for telling me what you thought "I wanted to hear" rather than what you believe to be helpful. You're a real godsend.

@BR-549 I appreciate the serious reply, I am not holding back information by some misplaced paranoia. I'm bound somewhat by a NDA; to me it seems that this questions pertains to a matter of theory irregardless of a specific application. Otherwise I would not have asked it, though perhaps I was mistaken.
Do you mean that an ideal autotransformer has no flux interaction? That makes sense, the problem here is that the component in the image is not ideal. The coefficient between the left and right winding is 0.54. You mention methods to vary the coefficient, if you can please elaborate that may be helpful. The application is essentially a flyback converter.

@EM Fields My question was not how I can run a proper test. Clearly I've considered these values, and the manufacturer runs a "proper test" by implementing of these values, and sends me a datasheet. My question pertains only to improving the coupling between the two windings, and thus reducing leakage inductance.

Again apologies if yet-still the information is insufficient. If so, than I appreciate the effort nonetheless.

#### R!f@@

Joined Apr 2, 2009
9,734
So it's a flyback converter.
May I ask why are trying a non isolated transformer.
Here is something that can help

Do you have a diagram by any chance ?

#### frpr666

Joined Feb 2, 2010
50
Is this transformer used for broadcast the E.M. field to the air ? Did you consider the EMC compatibility ?

#### airplane100000

Joined Aug 2, 2016
68
So it's a flyback converter.
May I ask why are trying a non isolated transformer.
Here is something that can help

Do you have a diagram by any chance ?
The only reason I am using this type of transformer is because of the physical dimensions. It is flat, that's it. Can't find any isolated transformer (or bobbin for that matter) with high enough inductance, with this step-up ratio that's under 8mm height.

Anyway I just told the manufacturer to rewind it, this time with one winding on top of the other. I suspect that will improve the coupling, if not then I've got some studying to do

#### BR-549

Joined Sep 22, 2013
4,938
Ok....airplane.....I am no power designer(or any designer)....but in order to understand this contraption....I need to see the taps on that auto transformer. I can not see the taps in that photo. How many connections does the unit have. What is the turns ratio. Which taps of the coil is the input and which taps are output.

Is the center tap used as ground. Now I'm thinking ignition coil. Anyhow all this little crap matters.

Use toroid core.

Adjust core mu for working current.

Wrap secondary over primary.

Watch out for reactance and kickback.

Consider a different strategy.

Coil-less switching......quasi resonant switching.

#### R!f@@

Joined Apr 2, 2009
9,734
You can wind the primary evenly to the length you require through out the core and then wind the secondary over the primary covering it evenly.
This way primary is spread and secondary is coupled better than before

I dunno about the size of your tx but there are low profile flyback transformers.
The UI or EI types are better at coupling, I think.

#### EM Fields

Joined Jun 8, 2016
583
The only reason I am using this type of transformer is because of the physical dimensions. It is flat, that's it. Can't find any isolated transformer (or bobbin for that matter) with high enough inductance, with this step-up ratio that's under 8mm height.
If you could describe the transformer in terms of the space into which it must fit, the respective required inductances of the primary and secondary, the transformer's required excitation voltage, current, and frequency, its load impedance, etc, etc, that would be extremely helpful.
Anyway I just told the manufacturer to rewind it, this time with one winding on top of the other. I suspect that will improve the coupling, if not then I've got some studying to do
An excellent empirical first step!