Will Matney has exposed the hype about Hipersil.
Very well presented Will.
Norm N8NH
At 11:01 PM 4/17/05 -0400, you wrote:
>Hipersil, the myth and the truth.
>
>I was asked about a transformer using nothing but a Hipersil transformer.
>I offered an EI type, M6 material transformer of the same specs, but it
>wouldn't do. Well folks, this gave way to me writing this for all on what
>exactly is a Hipersil transformer.
>
>Hipersil is a trademark owned by Westinghouse Electric for a type of
>transformer core and material they manufactured at one time. Hipersil, the
>material, is mearly a cold rolled, grain oriented, silicon steel with
>about 3.5% silicon added. That's it, nothing more. M6 material is the same
>material offered by several steel manufacturers, and is what Westinghouse
>bought in large coils. M6 is the AISI (American Iron and Steel Institute)
>number for 3.5% silicon steel which has been cold rolled, with the grain
>oriented in the rolling direction.
>
>
>What is cold rolled, grain oriented (CRGO)?
>
>This is a type of steel, which after being rolled out to a strip, or
>sheet, is cold rolled again to where the grain of the steel orients itself
>in a certain direction. What this does is drop the reluctance of the steel
>to the magnetic flux, which in turn lowers its losses, nothing more.
>
>
>Why add silicon?
>
>Silicon is added to control residual magnetism (core staying magnetized ).
>The thinness of the laminations controls eddy currents. There is different
>grades of silicon steel, with varying amounts of silicon. On the lower end
>is non oriented silicon steel (CRNO) which wasn't cold rolled for the
>grain to line up like M50. On the upper end is M6 which is about the top
>of the heap for CRGO steel without other alloying agents being added such
>as cobalt ot nickel.
>
>
>What is a Hipersil core?
>
>Well, there are many "Hipersil" cores if you want to call them that,
>including EI cores, but the steel manufacturers just use the "M" numbers
>to designate the steel. The proper question is what is a "cut core" or a
>"C-core"? A C-core is made by winding a thin strip of CRGO or CRNO steel
>around a rectangular or square mandrel in a lathe. The steel strip has an
>adhesive on it so it will bond together after it is rolled to a certain
>thickness. This strip can be several thicknesses where 14 mils is a very
>common one. The very thin ones like 2 mils is for higher frequency work in
>the audio frequency range. 14 mils is commonly used in power transformers
>for 50 and 60 Hz.
>
>After the core has been wound to the proper thickness, it is generally
>annealed to improve its properties, or some are pre-annealed. It is then
>bonded so it cant come apart. Next, the core is transfered to a band saw
>and cut in half. Each half is marked and then the cut faces are either
>machined or ground for a close fit when put back together. This helps
>eliminate the air gap if any.
>
>
>Why is a Hipersil core touted to be better than an EI core?
>
>Well, there's a lot of myth and mis-information here. The only difference
>between a cut core and an EI core using the same material is weight. This
>weight savings is due simply to the corners being rounded off where an EI
>core is rectanglar overall, that's it, nothing more. The weight savings
>you'll get is from 15% to 20% just because the corners are rounded off. No
>matter what, each type core has to have the same core area (A) in each type.
>
>A cores power handling ability (in watts or volt amperes) comes from its
>ability to cram all the magnetic lines of force (flux) into a small core
>area (flux density). Every core has a maximum flux density (Bmax) which is
>determined by the cross sectional area of the core (A) and the material
>the core is made from, nothing more. If this limit is exceeded, the core
>goes into saturation. This means that any more increase in current does
>not cause any more flux in the core. It also means that the cores
>permeability drops off sharply at this point. The waveform becomes
>distorted at the point of saturation and beyond too. A core is designed to
>not saturate when maximum power is drawn from the transformer (this is not
>the case in an inverter). Generally, a flux density is selected which is
>just below Bmax at design time. Lets say the material gives a Bmax of
>14,500 gauss. A working flux density of maybe 14,000, or 13,500 might be
>used. However, this increases the weight of the core
> because more iron has to be added to it to keep the flux density down.
> This is why when selecting a transformer, you should figure the maximum
> current in amperes to be drawn for each winding including losses. Losses
> in most filtered power supplies run around 5% or so. The reason for this
> is the designer trys to make the transformer as light as possible and
> this in turn means they design the transformer to run as close to
> saturation as possible without going over. On designing power chokes, the
> DC portion has to be considered along with the AC to keep away from
> saturation. The core can be gapped to correct some of this, but that is
> more than what can be covered here at this time.
>
>What are the disadvantages to a C-Core?
>
>First and foremost is cost. The cost of pre-made cores is expensive unless
>buying a large quantuty of each size. The mounting hardware costs more
>than does on an EI core of the same size. The last, and an important one
>is that a C-core runs hotter than an EI core of the same size. This is due
>to the mass of iron being smaller than on an EI core, plus the fit of the
>bobbin to the core. The only plus to a C-core is labor savings, and a
>minor weight savings. It doesn't take as long to put together a C-core as
>it does to stack an EI core. However, an experienced builder can go pretty
>darn fast stacking the lams of an EI core, even being interleaved.
>Actually, a tigther fit can be achieved of the coil to the core on an EI
>core than a C-core, because of adding the lams, one at a time to the exact
>thickness needed. With a C-core, you get the one thickness, that's it. To
>make it fit, extra paper or wedges is used a lot. This also makes one heat
>up more if the bobbin doesn't have a ti
> ght fit to the core. This raises the cost of the mandrel used to wind
> the coil as it's tolerances are tighter.
>
>
>Is an EI core better than a C-core?
>
>In my opinion, Yes! The reason being is the heat. Heat is the killer of
>all electrical and electronic components. I wouldn't care if the core
>weighed 20% more as long as it lasted longer.
>
>
>The above is just my opinion on the subject, and hope more have been
>brought into the light on the differences between these two types of
>transformers.
>
>Best,
>
>Will
>
>If you would like a PDF copy, please e-mail me.
>
>_______________________________________________
>Amps mailing list
>Amps@contesting.com
>http://lists.contesting.com/mailman/listinfo/amps
_______________________________________________
Amps mailing list
Amps@contesting.com
http://lists.contesting.com/mailman/listinfo/amps
|