For the first time of my life, I am trying to do a low frequency transformer. I’m used to seeing low frequency transformers built with, what I think to be, laminated silicon steel.

I have some difficulties to find a graphic which compares low frequency core material in function of the core losses. I think that at those frequencies (50/60 Hz) the core losses are really low. Even if it is the case I would appreciate to have some datas to convince myself.

If you know where I can find a graph which compareres the different magnetic material at those frequencies or any tips to give, it would be a pleasure to hear it :)

Thank you very much and have a nice day :)

  • \$\begingroup\$ What’s low frequency in your case? Give a value. \$\endgroup\$
    – winny
    Commented Jul 24, 2021 at 8:00
  • \$\begingroup\$ It is written :) \$\endgroup\$
    – Jess
    Commented Jul 24, 2021 at 8:01
  • \$\begingroup\$ “I think that at those frequencies (50/60 Hz)” isn’t a specification. Your question and what you try to solve is unclear. \$\endgroup\$
    – winny
    Commented Jul 24, 2021 at 8:03
  • \$\begingroup\$ I am trying to have datas for comparing the core loss of different core materials operating at 50/60 Hz \$\endgroup\$
    – Jess
    Commented Jul 24, 2021 at 8:05
  • \$\begingroup\$ The issue you face is that for low loss, the primary inductance needs to be very high (Henries not millihenries) and doing that without excessive copper loss, drives you to use high permeability materials, i.e. something very like silicon steel; ferrites just won't cut it. Then you have to balance losses from saturation, eddy current in the laminations, copper losses etc : the answers give you a start there. \$\endgroup\$
    – user16324
    Commented Jul 24, 2021 at 12:37

2 Answers 2


Core flux losses decrease with thinner silicon transformer steel and were originally graded by a number that represented the loss in W/kg. Then as they were able to produce very thin materials of cold rolled grain oriented steel or CRGOS the numbers did not always correlate to precise losses, yet were able to reduce losses from 0.9 W/kg down to 0.3 and even less.

Iron and hot rolled steel have far greater losses, while ferrite has far lower permeability.

With the key words in this answer you will be able to search and find suppliers with all the ratings and graphs with the tips you are seeking.

The losses are a function of eddy currents in the thickness of the material. Each layer is coated with a very thin silicate coating for an electrical insulation which gives each layer high capacitance with a tiny gap yet very high permeability to increase saturation levels of layers and greater linearity. The LC equivalent circuit of the core thus makes it suitable only for low frequencies.

MOT units are also made this way but welded across the seams to eliminate hum but that shorts all of the insulated laminations and makes them far less efficient as a power transformer. For some high voltage cores it is important to remember if you want higher performance to keep the edges clean of magnetic conductive dust particles, which is another topic of insulation breakdown from contaminated partial discharge.

This is an answer with clues rather than what I would do to explicitly hand you the answer from a web search to refresh my experience.

  • \$\begingroup\$ Thank you for your answer ! It helps ! :) If ferrite have lower permeability it meanns that for the same number of turn and the same core geometry I will have a lower magnetizing inductance and so higher losses, isn't it ? And magnetizing inductance comes from the fact that the permeabillity of the core is not infinite. So it can actually be attached to core loss too, isn't it ? \$\endgroup\$
    – Jess
    Commented Jul 24, 2021 at 8:22
  • \$\begingroup\$ An other point, I do not get how I could have higher permeability and greater linearity. Generally, if I introduced some non magnetic material, as air, or insulation material, I would reduce the permeability of the material and so increase the linearity, but it would be at the expense of permeability. As the permeability is equal to the slope on the BH curve, I can 't get higher permeability and greater linearity ? I may miss understand your answer .... I think your english is a true english, isn't it ^^ ? \$\endgroup\$
    – Jess
    Commented Jul 24, 2021 at 8:29
  • \$\begingroup\$ Interesting point "CRHOS" :) \$\endgroup\$
    – Jess
    Commented Jul 24, 2021 at 8:46
  • 2
    \$\begingroup\$ @Jess Low magnetising inductance does not increase core loss, but increases primary current, and so increases copper loss. \$\endgroup\$
    – Neil_UK
    Commented Jul 24, 2021 at 10:00
  • 1
    \$\begingroup\$ Ferrite has 10% of the mu of CRGOS and lower Tesla saturation levels from distributed magnetic particles in ceramic insulation. CRGOS yields your answer in 10 seconds on google. spacematdb.com/spacemat/manudatasheets/crgo.pdf \$\endgroup\$ Commented Jul 24, 2021 at 13:26

Wikipedia gives you this in their electrical steel page.

The magnetic properties of electrical steel are dependent on heat treatment, as increasing the average crystal size decreases the hysteresis loss. Hysteresis loss is determined by a standard Epstein tester and, for common grades of electrical steel, may range from about 2 to 10 watts per kilogram (1 to 5 watts per pound) at 60 Hz and 1.5 tesla magnetic field strength.

This wide range is such that you would be best to look at the properties for individual manufacturers' steels, as low loss steel is a premium product, and so many price/performance points are available.

It's worth comparing low frequency steel transformers to high frequency ferrite transformers. In the latter, the flux limit is normally driven by core losses rather than saturation. In mains transformers, core losses are never the limiting factor, and saturation controls the maximum permitted flux.

  • \$\begingroup\$ Thank you for your answer :) Do you have in mind some steel transformer manufacturers ? \$\endgroup\$
    – Jess
    Commented Jul 24, 2021 at 8:37
  • \$\begingroup\$ For those interested : nicore.com.cn/Ultra-thin-silicon-steel-core-1 \$\endgroup\$
    – Jess
    Commented Jul 24, 2021 at 9:46
  • \$\begingroup\$ @Jess, some manufacturers: Thyssenkrupp (Europe), NLMK (Russia), Aperam (Brazil)... \$\endgroup\$ Commented Jul 24, 2021 at 9:48
  • \$\begingroup\$ Lamination thickness is chosen to minimize core losses/kg and minimize mutual coupling losses from high mu leaving the saturation limitations as the natural limitation of flux density demanding more material with kVA ratings. \$\endgroup\$ Commented Jul 24, 2021 at 13:35

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.