On Tue,2/10/2015 3:35 AM, Ian White wrote:
I found K9YC's comparative measurements [1] but am struggling to explain either
(a) the shift in resonant frequency or (b) the difference in|Z| magnitude.
The difference between "binocular" style and all in a stack is the
difference in inductance and in the stray C. When the cores are in a
single stack, ALL of the turns pass through all of the cores. When the
cores are binocular, all of the turns go through some of the cores but
one less turn goes through the other cores, thus, less inductance, AND
less resistance coupled from the cores.
When the cores are all in a single stack, the turns can be spread, as I
have recommended for higher frequency chokes, which minimizes the stray
C, which raises the resonant frequency.
To understand this, it helps to understand that the most detailed
equivalent circuit of a ferrite choke is series R and L, which are
coupled from the core, in parallel with the stray C. Some of the stray C
is through the core, but most of it is turn to turn. At low frequencies,
the choke is series R and L, but as it approaches resonance, the series
R and L are converted to their parallel equivalent values, and the
circuit acts like a parallel resonant circuit.
For #31 material, there's a second parallel resonance in series with the
one described above -- it's the dimensional resonance of the core. It is
that dimensional resonance that provides the increased choking bandwidth
below 5 MHz. The only reference I know of for dimensional resonance is
the classic Snelling book, cited in my publications. It's long been
considered "the bible" in the world of ferrite manufacturers and
applications engineers, but is also long out of print. When I was doing
my research on ferrites beginning around 2003, a colleague who was
teaching at the University of Chicago found it in a sub-basement of
their engineering library and loaned it to me.
73, Jim K9YC
_______________________________________________
_______________________________________________
TowerTalk mailing list
TowerTalk@contesting.com
http://lists.contesting.com/mailman/listinfo/towertalk
|