Toroidal transformer winding & steel vs iron core material

Thread Starter


Joined Oct 10, 2014
I may end up winding some transformers and was considering validator formation as I have some material that may work well for the iron core. I'm considering using Olympic specification weight plates which have a 2" (or there abouts) center hole for the barbell. My plates are made of cast iron but I can also get plates made of steel in 1.25, 2.5, 5, 10, 25, 35 45 and 100lb increments. I'm not planning on anything larger than maybe the 10lb plate (about 7-8" diam) if I decide to make a large transformer.

Anyone who has looked at how transformers are made can see that toroidal are more difficult to make than using the E form plates and a bobbin. From what I have seen, it looks like many lengths of wire are used in serial to form each winding and they solder them together.

Here are some videos on how it can be done:

Machine process

Manual - This guy has a neat idea on the process!!

What I am wondering is whether cutting a slot from the center of the plate to the outer edge so the wire can pass through, allowing the winding to happen much easier, will have an effect on the transformer performance. The width of the cut would only be as wide as necessary for the wire to pass through so I would say 1/8" at most probably more like 1/10" inch. After winding is finished a piece of steel could be placed in the gap and tack welded in place IF TOTALLY necessary but since it could damage wire insulation/enamel I would really like to avoid this.

Another question is whether steel of cast iron is better. IDK what kind of steel it is, I suspect it is low carbon and pretty soft (in comparison to tool steel) as it is made for being a weight, not as a high speed cutting tool. It should be made to not chip when hit against other plates and such, so it should be soft.

how much iron or steel is necessary? I've read that microwave oven transformers reach saturation point because they don't use enough steel/iron and I'm not sure what effect this has and where more iron is needed to make the transformer "better" - middle, sides/surrounding, etc. Does more iron allow more wattage handling/production?

The reason I a looking at using the weights is because the total weight of the transformer isn't really an issue and in a could cases, the more heat it can dissipate the better. The last pic of the had dumbell I was wondering if something like that could be used as the iron core with the primary wrapped around the handle and then the secondary wrapped around the finished primary. My only concern is heat in the primary and what purpose the weight ends will have on the transformer. weights1.jpg weights2.jpg weights3.jpg weights4.jpg

weights1.jpg weights2.jpg weights3.jpg weights4.jpg

Thread Starter


Joined Oct 10, 2014
Thank you for explaining the lamination! That makes sense why things such as this aren't used. I figured the laminations and such were to allow for basically infinite sizes by adding/subtracting layers. So, I see this would not be a good idea. Thank you!


Joined Nov 23, 2012
Transformer cores need to be laminated to minimize eddy current losses.
Solid iron, such as the barbell weights, would have high losses and make a poor transformer core.
Typically, torroidal cores are made of iron powder or ferrite powder. Iron powder is rendered non-conductive by a phosphating surface treatment - no conductivity and very small powder means almost no eddy currents and no heating.


Joined Jun 17, 2014
Hello there,

Also, cutting a gap 1/8 inch wide in a toroid core or any other core for that matter means reducing the permeability by a substantial amount, and that means adding many turns to make up for that loss.
A core like that with metal like that probably has a permeability of 100 which is considered very low. That would require a lot of turns to begin with so making a gap just makes it harder because even more turns would be required than. Granted not as bad as with a high permeability material, but still doesnt help much.
The shape is not ideal by any measure, being long across one side and very short across another side. The ideal cross section for a toroid is round, but square is also used for simplicity. The cross sectional shape affects the efficiency of the windings. Round is best followed by square followed by rectangular which is almost a square followed by long rectangular which is about the worst of the simple shapes.

However, a test is the ultimate decider. You would wind some turns on one of your 'cores' and then do a few tests at low AC voltage. That would tell us what you can and cant get out of using one of these. I've used steel binder clips to aid in magnetic field sensing for example because i did not have anything on hand of the right shape at the time. That type of material is non ideal while the right metal would have worked much better, but it did help anyway for low field strength levels.

Normal transformer material is made specifically to have certain properties and those properties are sometimes different depending on which direction you look in within the metal itself. So a lot of thought goes into making a transformer core whereas a barbell is just a cast piece of steel with no attention to how it reacts to magnetic fields.
Last edited:


Joined Jul 18, 2013
Typically, torroidal cores are made of iron powder or ferrite powder. Iron powder is rendered non-conductive by a phosphating surface treatment - no conductivity and very small powder means almost no eddy currents and no heating.
Although the mains (LF) versions are typically wound with a coil of sheet/laminated nickel-iron alloy or silicon iron, similar fashion to a steel tape before extending.
Toroids consist of different core materials to meet the basic needs of ... Cores are produced from either thin laminations or metallic materials with a powder coating often widely utilized toroid and transformer winding components.
The powdered iron core is not ideal for transformers as it is implicit in them. Carbon-based (grain-oriented) silicon steel used to be a fundamental material for modern processors. The failure of both the center and the winding in power applications is the greatest way to trace the B-H curve of the toroid transformer.


Joined Jan 23, 2018
I have some cores that were tape-would, bonded, and then cut, and the cut edges lapped so they are flat. THAT is the way to make it easier to wind the coil. So while just cutting one slot is a bad idea, a cut and lapped joint can work fairly well.
To evaluate the effectiveness of any alloy for transformer use, toroid or other, you need to look at the B/H curve. That relates magnetization to magnetizing force. That will let you know about the amount of losses.