[Amps] Hipersil, the myth and the truth. (Updated)

[Will Matney]

Rich,

Thing is, The old Hipersil literature (engineers books) say it's the same material as M6 as does AISI. Most conventional transformers using EI cores are made using something like M19 or M22 which is still a CRGO steel but has higher losses. I don't know of any that wind transformers with M6 unless it's a special because of the cost of the lams. M6 costs about X3 what M22 costs and about X2 what M19 is. Some even wind the cheapie transformers with M50 or plain old cold rolled steel. The engineer at Magnetic Metals and Tempel Steel both told me there was not one bit of difference. Magnetic Metals has a trade name for their version of steel on the C-cores but I cant remember the name right off. Magnetic Metals sold out their line of EI lams to Tempel who is pretty much the sole steel provider anymore besides Allegheny Ludnam in the US. The original Hipersil line was sold to a European maker, I forget the name at the moment. Right now, in the U.S., the two largest C-core makers are Magnetic Metals, and Core Tran down in Texas. I think they're may be one other doing them in large volume. Core Tran will wind them using any steel you want.

The core has losses as it goes up in numbers. M6 being the lowest without any other alloying and M50 being the worse case. The next after M50 is plain cold rolled steel. After M6, they add either cobalt or nickel like for M2 which is the very top of the heap. Then there's the Permendurs, etc. after that. I suspect, any tests was done against maybe M19 or M22 which I would totally agree have more losses.

I have to disagree with Pete as there's NO cores with a zero loss in the iron even in the Permendures. Especially when it has a miniature air gap as compared to an interleaved stack which practically kills any gap. If they had checked it against an EI stack that had a butt gap, then we could compare apples to apples because the but stack has a gap and is used in chokes. There's another reason for me here, they done a test using a highly lapped gap on a C-core to see about maybe helping with saturation. It did work some, but were talking a surface finish of a few microns which is very expensive to do. To be honest, some of the new C-cores aren't even ground, just saw cut with a 100-125 microinch finish. Another problem after the core is cut is having each lam to match back up with the one it was cut from. this sort of shorts out the lams from one to the other and causes eddy currents at the gap, which again, the losses go up.

There was some C-cores designed by I think GE which had a staggerd pattern to the lams where they met. Their theory was that this would help stop any gap similar to an interleaved stack in an EI core. Then they came up with one which you wound the coil first, and mounted it in a core winder. It worked similar to a toroid winder where it passes the steel strip through the coils center and wrapped it around forming a donut core without any gap. That did prove out and was used on pig pole transformers along with the first mentioned type. The reason for this mention is just to show that GE understood that getting rid of the gap was better.

 What tests I've seen between an interleaved stack EI core compared to any with a gap is that the interleaved stack wins out hands down. The stack using M6 material can have as high as a 98% efficiency and some claim 99%. This might be what Pete is talking about (because of the material) but I cant see a C-core having a 99% efficiency, neither does any of the 20-30 text books on transformers I have read. It might have something close in a cold state, but when it heats up more than an EI type with the same core area, the reluctance has to go up. That magnetic current in the iron is directly proportional to the electrical current and as you and I both no, when something heats up, its resistance or reluctance goes up. Then, there's that stinking little gap showing its face again in the mix.

If you ever get a chance to look through the old Magnetic Metals catalogs (I think may still be some available for the EI cores) of the EI cores and the C-cores, check out the watt per pound loss curves between the two and you'll see where I get my info. For interleaved EI stacks using M6, you can see very low watts per pound loss as compared to a C-core with the same core area.

One last thing I should have mentioned earlier when I first wrote about this. Hipersil touts running those cores with a Bmax of 17 kilogauss. Over 15 kilogauss, the losses are really great compared to the savings and they really drop off afterwards. If the core was designed around 17 kg instead of 15 kg, be leery because it sure is gonna run hot as hell. The thing is, that's for M6 too and M19-M22 is only 14.5 kg. so you really don't save that much (a whopping 500 gauss?), just the watts per pound loss and it's not much greater either!  When designing a core, the C-core uses a slightly different formula which takes into count the window area with the core area or the WA product. However, no matter what, the core area itself HAS to be a certain size in order not to saturate, so the standard old EI formula still works. The WA product was invented to try to help speed up the design process by cutting back on the trial and error process of having enough window for the coil. So, if you use the old EI formulas, and you see you need a certain size core in square inches, or centimeters, then get one smaller, send it back as it's simply built wrong.

But here's the root of things and what I'm saying. We can have all these neat different designs in the world, even save a few watts on the pound, but would you rather have one that runs cooler or hotter? This ain't something I made up in my head, the information is available from most any tranny manufacturer or core supplier, a C-core runs hotter than a comparable EI core of the same core area. No ifs's, ands, or buts about it. I think I'd prefer the cooler one which lasts longer.

I'm not gonna spell check this as it's too long and I don't have time to make it perfect so please look over it.

Best,

Will

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

On 4/18/05 at 4:27 AM R.Measures wrote:

>I have observed that Hipersil core transformers have substantially more 
>volts per turn than conventional core transformers.  This results in 
>them needing to use less wire to achieve the same secondary potential - 
>which means less ESR and less heat.  Also, Pete Dahl told me that with 
>a conventional core, roughly half of the loss is Cu-loss and roughly 
>half the loss is core-loss, but with a Hipersil-core, virtually all of 
>the loss is Cu-loss.
>
>On Apr 18, 2005, at 3:13 AM, G3rzp at aol.com wrote:
>
>>
>> In a message dated 18/04/2005 04:03:29 GMT Standard Time, 
>> craxd1 at ezwv.com
>> writes:
>>
>> Is an EI  core better than a C-core?
>>
>> In my opinion, Yes! The reason being is the  heat.
>>
>>
>> All very interesting, and useful info. But for some reason, nearly all 
>> the
>> military stuff over here has, for years, tended to use 'C' cores. I 
>> can see the
>>  weight saving for aircraft stuff, but why would the Navy be so keen 
>> if there
>>  weren't other advantages? Or the Army, bearing in mind that 
>> 'portable' to
>> them  tends to mean 3 men and a truck?
>>
>> I vaguely remember somewhere the claim that because the C cores are
>> precision ground, the overall magnetic reluctrance is less.
>>
>> 73
>>
>> Peter G3RZP
>> _______________________________________________
>> Amps mailing list
>> Amps at contesting.com
>> http://lists.contesting.com/mailman/listinfo/amps
>>
>>
>
>Richard L. Measures, AG6K, 805.386.3734.  www.somis.org
>
>_______________________________________________
>Amps mailing list
>Amps at contesting.com
>http://lists.contesting.com/mailman/listinfo/amps