[Amps] Wound core transformer losses

[Will Matney]

Since the latter posts, I got to reading a little more last night on losses in strip wound cores and it hit me that I had forgot yet another reason for the losses. When those cores are wound, one must remember they are wound in a spiral. Well, in nature, those magnetic lines of flux dont flow in a spiral but a circular or oval pattern following the shape of the core. The thing is during a cycle of current, lets say the line starts in the center of the core, and begins to go out through the legs. As it gets into the legs it wants to make a straight path downward before it gets bent again. Since the line wants to go straight, and the core is spiral wound, as the line gets lower, it starts moving from the middle of the lam to the skin. After it gets so far, it has to jump over to the next lam so as to make a straight line. Now, where it jumps is only one thing, a gap! Thus, the more tuns on the core, the more gaps there is adding up, besides any from a cut where the core goes together. This even being on a toroid that is spiral wound. Another thing, this could cause more than one gap between each lam or the flux may have to jump over more than once per layer to keep in a straight line ( I would think maybe twice). That is according to the design of the core and its shape.

If one wanted to make one with the very least amount of losses, one would need to make rings out of the strip and butt weld each together. Each ring then would progress in size. Then each stacked over the other until the proper thickness was made. That way, the flux would have a straight path with no breaks around the core, and not have to hop over any gaps.

This kind of throws a wrench in my earlier post about strip wound toroids being the best. However they probably are over the other types now available, but maybe not over an EI lam as it would be according to all the gaps adding up, and which had less.

One other thing, which is just a theory, is about the flux moving in a CRGO sheet used in an EI section. A perfect 100% grain could hurt more than help in this case I would think. The reason being, in a core using CRNO steel, the flux can move easily in any direction. In CRGO, it moves easily in only one direction. The problem is that the flux has to bend around at the top and bottom of the core into the legs. Using 100% oriented, it would cause a greater reluctance in these areas than using say an orientation of maybe 75%. This is just a theory of mine of course. The thing is, the steel manufacturers tests are done on a straight piece of steel for the losses. They show 100% with less losses than something of 75%. Any thoughts on this would be welcome.

Best,

Will