[Amps] Cut core and EI core loss diferences

[Will Matney]

Mike,

Thanks for the comments. I happened across that book after Tom Aldridge and myself discussed the possibility of using a saturable reactor to control inrush current in our amps, which by the way will work. I did a good bit of research on the subject and needed research material. I just happened across the book on eBay and it surely fit the bill. One thing I have found is that the older books on magnetics from the 1920's to say the 1960's have way more good useful information than anything new in print. That also gave me the idea of writing a book on the subject using the newest construction materials and practices. I have started it, but it will be some time to complete, as I only work on it when I have time, or really feel like it. What I want is a good in depth technical book on the subject but also one any layman can understand in the process.

The main thing is, I wanted to put a stop to the mis-information that has been fed to amateur operators about their equipment over the years. What I assume happened was when Hipersil first hit the scene, it was true that it had less iron losses than what was then known as silicon steel. Silicon steel at the time was only cold rolled NON-oriented steel. Hipersil was the first cold rolled GRAIN-oriented steel, and was only available in cut cores through Westinghouse for some years. I figure after the patents ran out, and others began to make it in sheet, or coil, they began using it in the EI core types. When this happened, it threw out the toe hold that Hipersil had as being the best. The problem is that this kept on by word of mouth and was hardly ever talked about because nobody thought to look at it. I only noticed it when looking over the watts per pound loss curves between the core types using the same steel about two years ago. Yet, I figured it was only because of the cut gaps on the cores which caused it until I read the part in the book. Other books just skimmed by the subject without coming out and saying it. Some only mentioning what a Hipersil, or cut core was, and showing a few pictures.

Electrical steel has advanced a huge bit over the years (especially in the last 40) where several types have higher permeabilities, and some higher flux densities than M2 grade CRGO steel. When looking at any core type, the central core inside the windings is the most important part and really is what determines everything besides adding an air gap. The area of the cross section is what determines its power handling capability, and is definitely where you would want to control what losses you have the most. The legs, and yokes, do count too, but not as much as the central core itself. The ones I'd think with the lowest losses would be ones wound like the spira-core by GE which has no cut air gaps, just the ones between the strip. It is wound as a toroid on the winding itself. They also make strip wound toroids too and are being used some today I see in some European amps. The only problem with these is needing to use a toroidal winder, and are slower to wind than an open coil in a lathe. I forget the manufacturer I seen using these, if anyone knows, please let us know.

Best,

Will


*********** REPLY SEPARATOR  ***********

On 4/21/05 at 9:16 AM Mike wrote:

>Hello Will,
>
>Thanks for the very complete posting - look forward to your well thoughtout
>comments.  
>At 11:08 PM 4/20/05 -0400, you wrote:
>>Ok, one last word. I finished my research into this as I myself wanted to
>find exactly why cut core iron losses were higher than EI cores. I figured
>it was just the air gaps on the two C sections but there was more. I
>reference a book below named "Saturating Core Devices" by Leonard R Crow
>where this very thing is mentioned. The book is about saturable reactor
>design but discusses transformer construction also as this is all a reactor
>is, a transformer with a DC coil added to control it.
>>
>>Quote;
>>
>>"The assembly as described [1] is necessary if Hipersil is to be used in
>the saturable reactor. It affords the advantages of ease of fabrication,
>lower core loss [2], and saving in weight and space.
>>
>>Nevertheless it has disadvantages of expense and the problem presented by
>the fact that although the mechanical construction would lead one to
>believe that it is equivalent to the standard 3-legged core [3], actually
>the ac flux path is different and inferior. Fig. 2-13 illustrates how the
>C-core is traversed by the ac flux, and, consequently, the iron losses are
>greater than would be the case if the ac flux path were identical to the
>3-legged core. For the Hipersil assembly shown in Fig. 2-13 [4], the ac
>fluxes balance in the adjacent central legs, but these fluxes must travel
>through the middle section because of the magnetic discontinuity between
>the cores.
>>
>>This magnetic discontinuity between the cores is caused by the air gaps
>between the individual turns of the Hipersil steel ribbon; these air gaps
>are composed of insulative materials in this type of construction. Fig.
>2-14 shows a set of two cores wound with Hipersil ribbon. For clarity, the
>cross section of each ribbon turn is enlarged or magnified as is also the
>thickness of insulation on and between the ribbon turns. the two
>broken-arrow lines through the top of the core sections and the one
>broken-arrow line at the bottom of the core sections represent flux lines
>through the core air gaps. Let us assume that instead of having four turns
>of Hipersil ribbon in each core that each core has four hundred turns and
>that the insulation gap between turns averages 0.001 inch. Then for the two
>cores we will have 800 X 0.001 inch = 0.8 or a total of 800 small air gaps
>equivalent to a single large air gap of 4/5 inch through the cross section
>shown by broken-arrow line X. A much grea
>> ter air gap reluctance is offered for the flux path shown by line A,
>since here a large additional air gap is offered by the air gap reluctance
>between the curvature of the two core corners. The same reluctance theory
>for this type construction, of course, would hold true with either silicon
>or Hipersil ribbon".
>>
>>Footnotes;
>>
>>[1] Using two C-cores as a shell core as in an EI core.
>>
>>[2] Lower core loss as compared to cold rolled non-oriented steel.
>Hipersil being CRGO, and at the time (1949), only available in C-cores. M6
>material is now available in EI cores on up to M2 which are all CRGO steel.
>>
>>[3[ 3-legged core is referring to a EI core in the book.
>>
>>[4] Fig. 2-13 shows two C-cores being used as a shell core which looks
>like an EI core.
>>
>>The magnetic discontinuity spoke of is in between the two core sections
>where they butt together down the center of the core. The text don't
>mention this but the gap is also increased by the banding on each C section
>holding each together. In other words, the flux has to jump across this gap
>from one half core to the other! In an EI core, this is not the case. Plus,
>the fit on the cut gaps for a C-core are about a 0.001" to 0.002" fit which
>another gap is presented here. Grinding and lapping will only get it very
>close but a gap of a few microns still exists. reference a previous post of
>mine about a test that was made.
>>
>>One last point of interest. the name Hipersil came from adding three words
>together.  HIgh PERmeability SILicon.
>>
>>Best,
>>
>>Will
>>
>>_______________________________________________
>>Amps mailing list
>>Amps at contesting.com
>>http://lists.contesting.com/mailman/listinfo/amps
>>
>	73,
>	Mike,K4GMH