How to make small Toroidal Flyback Transformer using Mu-Metal tape for SMPS circuit?

[Unixxx]

Hello,

I am trying to find some information on building a Toroidal Transformer from scratch, so that I have exact parameters needed and something I would like to learn as a hobby and skill. Please don't say to go buy one, I could if I wanted to but will never know how it is made and why it works.

From my research "Metglas" is a good core material for toroidal transformers. It's impossible for a consumer to get hands on that and very expensive but not much is needed for SMPS circuits, since SMPS does not need to use large and heavy transformers.

So second best core material seems to be "Mu-Metal". I can easily buy this from online ready to be shipped to my doorstep. It is also relatively expensive but as again not much is needed, I don't think the Toroidal Flyback Transformer/inductor I need would be any larger in diameter than the size of a thumb.

I am also considering to buy electrical metal (silicon steel) instead of Mu-Metal since it is cheaper and is widely used in making tape wound Toroidal Transformers, but no one sells silicon steel rolls online that can be purchased easily. One would need to contact a manufacturer to get their hands on silicon metal, most likely from Alibaba and would need to buy a massive quantity for the manufacturer to even make it worthwhile for them to ship it to me.

So how does one exactly make a toroidal Flyback transformer using Mu-Metal?
What are the math, formulas, parameters, calculations and books needed for designing, building and constructing a toroidal transformer?
I searched entire youtube and there is only one person who is successfully making toroidal transformers using electrical metal (silicon steel), here is the video, you can check his channel and there are videos of him making toroidal transformers:

It seems the toroidal transformer needs to be sintered which is not hard to do at home can use a homemade forge furnace for that or just build an electric furnace.

Basically I want to design a SMPS with the following parameters: Input 125 volts, Output 12 Volts, 15 Amps, Topology Flyback/Buck Converter. Not sure about the switching frequency yet.

Yes, I know I can buy a pack of assorted inductors/transformers and can make the SMPS, but I just want to learn the process of making a toroidal transformer from scratch.

Thanks.

 

[Buk]

From my research "Metglas" is a good core material for toroidal transformers. It's impossible for a consumer to get hands on that and very expensive but not much is needed for SMPS circuits, since SMPS does not need to use large and heavy transformers.

These guys will sell you metglas ribbon at a reasonable price: 0.025 to 0.018mm thick, so it takes lots of layers for even quite small cores.

It seems the toroidal transformer needs to be sintered which is not hard to do at home can use a homemade forge furnace for that or just build an electric furnace.

Sintering is (usually?) reserved for powered metal parts. eg. ferrite and similar. And it also requires compression as well as heat.

Metglass ribbon cores are usually potted in epoxy.

 

[fourtytwo]

Your desired output power is 180W, probably beyond flyback you will probably need a half-bridge or forward converter.
What is your intended frequency of operation ? Steels of any kind are not usually found in SMPS transformers as the frequency is to high, you will probably need ferrite such as the popular ETD cores. If this is offline (120/240V) you might be better buying an SMPS rather than trying to build it as not only is it dangerous but the inevitable blow ups from mistakes & miscalculations will be very expensive, time consuming & soul destroying if you are starting from nill experience.

 

[shortbus=]

What your saying doesn't make much sense to me. From what I know Mu-metal is nonmagnetic and used to shield things from magnetic signals and such. And then you come to a transformer core which needs to be magnetic, so how does Mu-metal make a good core? Most cores like your talking about are Ferrite which is magnetic to a slight extent but less than steel.

 

[Buk]

From what I know Mu-metal is nonmagnetic

Mu-metal is a soft magnetic material specifically designed for use as a core material.

From Wikipedia:

Special alloys​

A family of specialized alloys exists for magnetic core applications. Examples are mu-metal, permalloy, and supermalloy. They can be manufactured as stampings or as long ribbons for tape wound cores. Some alloys, e.g. Sendust, are manufactured as powder and sintered to shape.

Many materials require careful heat treatment to reach their magnetic properties, and lose them when subjected to mechanical or thermal abuse. For example, the permeability of mu-metal increases about 40 times after annealing in hydrogen atmosphere in a magnetic field; subsequent sharper bends disrupt its grain alignment, leading to localized loss of permeability; this can be regained by repeating the annealing step.

The reason it works as a shield material is that it acts as a high permiability 'wave guide' giving magnetic fields a low reluctance path around the shielded entity.

Its BH curve:

 

[crutschow]

Mu-metal is nonmagnetic and used to shield things from magnetic signals and such.

If it's nonmagnetic, then what property does it have that it can shield magnetic signals?

 

[shortbus=]

If it's nonmagnetic, then what property does it have that it can shield magnetic signals?

It was a senior moment on my part.

 

[schmitt trigger]

Shortbus;
Don’t worry. Most of us old farts have these “senior moments.”
Just the other day I burned my fingers while attempting to solder a component with jumper wire.

BACK TO THE OT: Flyback transformers have significant DC bias and thus an air gap is mandatory. This is difficult if not impossible on toroidal cores, unless it is powdered iron with its distributed air gap.
If you really, really want to use Metglas in this application, you’ll have to use C cores, and apply the gap between both halves.

 

[shortbus=]

Most of us old farts have these

I resemble that remark.

Again I may be off the mark on this, but if Mu-metal is really soft and overly magnetic wouldn't that be bad for a flyback transformer? Aren't they usually high(or higher ) frequency transformers? At a high frequency and high magnetic value wouldn't one made with Mu-metal saturate fast and not give the response that a flyback needs?

 

[Buk]

if Mu-metal is really soft and overly magnetic wouldn't that be bad for a flyback transformer?

It is a fairly soft (ductile) metal, but the softness in 'soft magnetic' (as opposed to hard magentic) means that it doesn't retain its magnetism once an external field is removed. Hard magnetic materials are magnets. Soft magentic materials have high magnetic permeabilty, but can never be magnetized in their own right. ie. Not magnets.

Thus soft magnetic materials include ferrites, zinc and vanadium alloys and silicon steel.

At a high frequency and high magnetic value wouldn't one made with Mu-metal saturate fast and not give the response that a flyback needs?

Mu-metal has low hysterysis, so is good for high frequency -- it can change magnetic orientation quickly -- but it saturates quite quickly so is not good for high density current.

 

[unclejed613]

i've seen toroid transformers made with metal tape, but they operated at 60hz, not at tens of khz.

 

[shortbus=]

i've seen toroid transformers made with metal tape, but they operated at 60hz, not at tens of khz.

I have one in a power supply for an induction heater I made that has a metal core. It's a big heavy piece of iron that's for sure.

Seems like the TS has bowed out of this.