Abstract
We previously reported that the estimated eddy current losses agreed comparatively well with the measured ones for very thin 3% Si-Fe strip-wound cores in the 500Hz∼5 kHz frquency region. The eddy current losses were estimated by applying domain wall numbers to a modified Pry and Bean model. We observed the domain wall numbers on the outermost surface by using an SEM. Here, in order to investigate the difference between the measured eddy current losses and the estimated ones below 500 Hz, we removed the layers of a 3%Si-Fe strip-wound core one by one and examined the relation between the domain wall numbers and the eddy current losses. In this experiment, the cross-sectional area of a strip-wound core had to be large enough to maintain a sinusoidal flux condition in core loss measurement after several numbers of layers were removed. We also examined the use of continuous annealing to fabricate along and very thin grain-oriented 3%Si-Fe strip.
