[RFI] Snap on split beads; coax high-pass filter recommendati ons

EDWARDS, EDDIE J EDWARDS, EDDIE J" <eedwards@oppd.com
Fri, 3 Dec 1999 10:30:05 -0600

> -----Original Message-----
> From:	Jon Ogden [SMTP:jono@enteract.com]
> on 12/2/99 9:18 AM, EDWARDS, EDDIE J at eedwards@oppd.com wrote:
> > I have tried 77 mix on audio rectification problems from 14 MHz
> > signals and it had absolutely no effect.  (I actually thought it
> might've
> > been a little worse!)  Whereas the 43 completely eliminated it using
> same
> > size toroid.  
> You're right.  Argh!  I get the numbers mixed up.  I think it was type 43
> that I bought from Amidon and not type 77 (whatever I bought from them
> worked in my house).  According to my chart from Fair-Rite  type 77 is
> only
> good up to about 2 MHz.  I don't know how accurate that is since it says
> type 43 is usable up to 10 MHz!  Oh well.
	[Ed-K0iL]  Be careful with which application that frequency
specified relates to.  Ferrite's (toroid, split, etc.) are used for many
applications.  Not just EMI suppression.  I recall looking at the Fair-rite
and Ferronics catalogs and getting completely confused by their specs
because of the many applications.  They can be used in Low-Level inductors,
Power Inductors, Low Level transformers, Current transformers, Wide-Band
transformers, Power transformers, and Pulse transformers not to mention EMI
suppression.  There are different equations to calculate frequency response
for each application.  I have a feeling those freqs are for a different

	EMI suppression with ferrites isn't as straightforward as one would
think.  I've added some info from my Ferronics Inc. catalog that might help
or might muddy the waters depending on your point of view!

> The material though in the split ferrites and so forth that you can buy
> from
> Radio Shack are not type 43.  They are designed more for attenuating VHF
> energy which would probably make them type 61 or 68 material.
	[Ed-K0iL]  I just called 1-800-THE-SHACk and asked what mix they
are.  The guy didn't know, but he's having someone research it!  They'll get
back to me hopefully today.  I mentioned there's a couple hundred guys
(customers) out there on this mailing list waiting to find out.  

> All these doggone numbers!
	[Ed-K0iL]  No kidding!  And it's not just numbers but letters as
well!  Well, here's the info from Ferronics (sri about any typos):

	"Ferrite beads provide a simple, economical method for attenuating
unwanted high frequency noise or oscillations.  By slipping a bead over a
wire an RF choke or suppressor is produced which possesses low impedance at
low frequencies and relatively high impedance over a wide high frequency
band.  The effectiveness of this impedance in reducing RFI/EMI depends on
the relative magnitude of the source, and the suppressor and load
impedance's.  Beads are also available fixed on a wire, taped and reeled for
automatic insertion." (So there's no excuse for mnfrs not installing them on

	At high frequencies the permeability and losses of ferrite vary with
frequency.  The permeability declines while the losses rise to a broad peak.
The equivalent circuit and curves in figures 1 & 2 show how this property
can be used as a broad band filter."

	Figure 1 shows a series circuit containing a source Es and its
impedance Zs and a load impedance Z-load.  The bead appears in series as an
inductor L and AC-resistance R.

	Figure 2 is the graph (this is tougher to describe).  It shows 3
variables L, X, and R.  The X axis is frequency increasing, and the Y-axis
is amplitude of L, X & R.  L (inductance) is max at low freqs and level
until some point then it drops off to zero(?) or min.  X (mostly inductive
reactance) starts at a low level at low freqs and then climbs with
increasing freq peaking just before L goes to min (or zero?) then starts
dropping rapidly after L is min (or zero?).  This is the one that gets me: R
(AC-resistance) starts at min (or zero?) and very gradually climbs with the
increasing freq until it peaks at some freq well above the point where L and
X go to min (or zero), then it gradually declines.  I believe this is due to
skin effect.  With the ferrite bead over the wire, the permeability (mu)
decreases with increasing freq causing the conductive skin depth (which is
the Sq Root of mu*pi*conductivity*freq) to decrease even though freq is
increasing.  As skin depth decreases there is less area for electron flow;
the resistance will therefore increase proportionally.  Now, back to the

	"Ordinarily beads of ferrite are slipped over a wire producing a
one-turn device.  To low frequencies the component presents a small
inductance whose reactance can often be neglected, while to high freqs the
device presents a higher series resistance with near zero reactance.  Since
this resistance is a result of material losses it is a true dissipative
element.  Furthermore, since the reactance is low there is little chance for
resonance with stray capacitance which would spoil the suppression."

	So there ya go.  I noticed Ferronics uses "Letters" to represent the
material type so apparently the numbers like 43, 77 an so on are not
universal.  B & T material is used for noise (signal) suppression at 0.5-100
MHz while J is 5-500 MHz.  Also K for above 20 MHz (close to 43?) and P for
above 80 MHz (close to 77?).  

	Remember that you are probably needing to suppress a few harmonics
as well as the fundamental when it comes to audio rectification so chose the
material accordingly.  Also lower fundamentals (80, 40 & 20 mtrs) may or may
not cause audio rectification in audio circuits depending on circuit design
or perhaps due to length of speaker wire becoming more or less resonant with
a signal (fundamental or harmonic).

	I'll post mtl mix info if Rad Shack gets back to me!

	de ed -K0iL

Submissions:              rfi@contesting.com
Administrative requests:  rfi-REQUEST@contesting.com
WWW:                      http://www.contesting.com/rfi-faq.html
Questions:                owner-rfi@contesting.com