Turns ratio question for a custom transformer

[FusionITR]

For a power supply i'm building, I need a primary inductance of 32.82uH. If I use a Ferroxcube E30/15/7 core with 3C81 material, I get a primary number of turns of 3.62 given (rounded to 4 turns) the formula N = sqrt(L/Al), given the Al value of the ungapped 3C81 material.

Data here: https://www.electro-tech-online.com/custompdfs/2006/06/PSG2003.pdf

Given I need to step down ~20V to 5V, isnt it unstable to have so few turns when actually building a transformer? Do I required to get a gapped core to lower the inductance factor so I can have more turns?

 

[Hero999]

I don't know, I suppose you could just use a different core.

There again will your power supply circuit be alright with a higher inductance?

That way you could easilly have 8:2, or even 12:3.

 

[Russlk]

Why do you need such a precise primary inductance? That is probably the minimum value to use. You can't have less than 1 turn on the secondary, but since 4:1 is the ratio you want, it should work, but I would use 8:2 or 12:3 as Hero999 suggests.

 

[FusionITR]

Why do you need such a precise primary inductance? That is probably the minimum value to use. You can't have less than 1 turn on the secondary, but since 4:1 is the ratio you want, it should work, but I would use 8:2 or 12:3 as Hero999 suggests.

The 32.82uH inductance I need for the primary is the minimum inductance needed. This is a flyback transformer for a SMPS, so is it as simple as just increase the number of primary turns to whatever value I want and my the secondaries propotional?

Like in a general case, I make the calculations for the core I have and I need 4 turns to reach the minimum inductance needed, I can just wind 20 turns and make the secondaries propotional to the 20 turns in the primary?

I hope this makes sense.

 

[Russlk]

More turns equals more inductance, and it will take longer to charge the magnetic field, which limits the frequency. E=L*dI/dT. Or, dI/dT =E/L amperes per second.

 

[FusionITR]

Well in a SMPS power supply freq is very important, so I dont think I can play around with the turns too much.

If I calculate I need 4 turns for a primary inductance of 32.82uH, but I use 12 turns instead, how much does that effect the maxium frequency the transformer can operate at? Is there a formula or guideline I can use?

Thanks for the information, I'm pretty new to power electronics and am trying to learn.

 

[Optikon]

Well in a SMPS power supply freq is very important, so I dont think I can play around with the turns too much.

If I calculate I need 4 turns for a primary inductance of 32.82uH, but I use 12 turns instead, how much does that effect the maxium frequency the transformer can operate at? Is there a formula or guideline I can use?

Thanks for the information, I'm pretty new to power electronics and am trying to learn.

I think you are approaching your design problem incorrectly. I don't have time right now to outline a flyback transformer design procedure so my best advice is to get the book "switching power supply design" by Pressman. Everything you need to know about this is in there.

 

[FusionITR]

I think you are approaching your design problem incorrectly. I don't have time right now to outline a flyback transformer design procedure so my best advice is to get the book "switching power supply design" by Pressman. Everything you need to know about this is in there.

I already have a SMPS design book that I am learning from.

What is says to do is calculate a minimum primary inductance based on the input voltage, switching frequency, and peak current.

When this minimum inductance is calculated, choose an appropriate core (based on power requirements), then calculate the number of turns in the primary to achieve the minimum inductance. The N^2 = L/Al formula (which is in every ferrite calatlog ive seen).

The inductance factor (Al) is straight from the catalog, so when calculating the number of turns needed to reach that minimum inductance needed for the primary thats how I got the low number.

This design proceedure seems correct to me, where am I going wrong?

 

[Optikon]

I already have a SMPS design book that I am learning from.

What is says to do is calculate a minimum primary inductance based on the input voltage, switching frequency, and peak current.

When this minimum inductance is calculated, choose an appropriate core (based on power requirements), then calculate the number of turns in the primary to achieve the minimum inductance. The N^2 = L/Al formula (which is in every ferrite calatlog ive seen).

The inductance factor (Al) is straight from the catalog, so when calculating the number of turns needed to reach that minimum inductance needed for the primary thats how I got the low number.

This design proceedure seems correct to me, where am I going wrong?

NOt that you are going wrong. It is good you have a text to help you because these designs are not as straightforward as the theory would suggest.

Usually, the core is chosen first based on power requirements (as you know) as well as geometrical concerns (like can it fit where it needs to go) and noise(core geometry & winding methods affect noise performance)

Also, do realize that you will more than likely get within only 10% of your design inductance. The formulas and the non-linear properties of the core material (as well as the variability) only make it practical to design for an approximate value of inductance. You also have not stated what flux density you will be running the core at and what temperature range the material will experience. These both will change your actual inductance.

What do you mean by "unstable" with building a tranformer with a few turns? I have designed ONE turn xformers in the past and do not understand what your concern actually is.

The number of turns you end up with and the way they are would will affect parasitics as well like leakage inductance and coupling capacitance