Rick Karlquist N6RK wrote:
>If you put an ampere turn or so through a high permeability
>ferrite, it will magnetize and permanently lose a lot of inductance.
>You can of course fix this by heating above the Curie temperature.
You are absolutely correct, and I knew it when I wrote what I did,
which is why I hedged by writing that they do not retain ***much***
magnetism (in comparison with "hard" ferrites). I probably should
have been less vague.
I have experimented with ferrite cores, deliberately putting DC
ampere-turns through them to see what it did to their incremental, or
small-signal, permeability. Just as you say, they magnetize
permanently and much of the "choking" inductance is lost. How much
is lost depends on the material. You can find B-H hysteresis curves
in the Fair-Rite catalog.
Whenever I had magnetized a core I always marked it so that I would
not use it later without having demagnetized it.
>I cringe when I see people putting ferrite beads on one
>leg of a 20A power supply. The 20 ADC magnetizes the bead
>into saturation the first time you turn it on. What you want to do
>is run BOTH wires through the bead so there is no net DC.
I agree completely. If you must put a ferrite bead on a wire or a
cable carrying a significant net (common-mode) current, whether it's
DC or 60-Hz AC or whatever, you must use an appropriate type of
ferrite. The Fair-Rite catalog will help you select one.
Unfortunately, such ferrites don't have huge permeabilities.
BTW, I learned while experimenting with 60-Hz AC that high-mu
ferrites such as type 64 are slightly piezoelectric or piezomagnetic
or magnetostrictive, or something. When the magnetization reverses,
the core clicks. I think that it was the core itself clicking, not
the wire (which of course experiences an electromagnetic force).