Purpose This paper aims to present the proposition of a new experimental method for obtaining very crucial data of the structural steel that is used in the tank of oil filled power transformers, namely, the volumetric losses and the magnetic permeability, both in function of the density of magnetic flux. Although these data are not usually available, they are fundamental for helping the transformer designer in avoiding the occurrence of hot spots in the transformer tank. The adoption of a conventional Epstein frame has restrictions because of the incompatibility between it and the samples of the steel. Design/methodology/approach The basis of the proposition is the same as that of the Epstein frame, with significant attention paid to the additional losses in the winding that creates the magnetic flux to the samples in the core. These losses can be significant and are created by the harmonics of current along the windings and are summed to the ohmic losses. For separating these winding losses from the magnetic losses, each sample is made as being the core of a toroidal 1:1 transformer. Thus, two tests with two identic of these toroidal transformers are necessary. Findings The proposed methodology is simple, because it is very similar to the classical tests of transformers (no-load and short-circuit tests). The process of separation of losses requires only a numerical fitting of curves for adjusting the winding losses as a function of the current amplitude, and the obtained results are coherent with the expected behavior of the magnetic losses and the magnetic permeability of a structural steel. Research limitations/implications The method gives very approximate results in comparison to those obtained using the Epstein frame. The influences of the temperature and/or of the skin effect have not been evaluated. Practical implications Practical, real and thus confident data of structural steel, such as the magnetic permeability and the volumetric losses (hysteresis and Foucault), become available for the transformer designer to take actions for not only reducing the tank losses but also for avoiding the occurrence of hot spots through computer simulation. Originality/value The proposition is very new, as it allows to test steel samples with a size that does not fit to a usual Epstein frame. It takes into account the real influence of harmonic of currents in the losses along the winding of a classical Epstein frame, which has not been so far mentioned. It allows obtaining data of structural steel that had not been considered important until now.