[RFI] RFI proof speakers

Tim Duffy K3LR k3lr at k3lr.com
Wed Nov 1 11:38:45 EST 2006

At 11:17 PM -0600 10/31/06, Tim Duffy K3LR wrote:

  I really need to find a solution for 40 and 15 meter RFI (caused by my near
  field RF) that gets into my neighbors computer speakers. I have asked this

  before, but I need to find a good set that is close to RF proof. I need to get
them fixed once and for all and a new set of RFI proof speakers is the way to

I can't believe that these speakers can't be fixed by putting _good_ ferrite
common-mode chokes on _all_ of their cables (perhaps as many as five cables: one
the right speaker, one for the left speaker, one for the subwoofer, one for the
wallwart, and one for the USB or audio cable to the computer).

A good choke has _at_least_ 1000 ohms impedance at the relevant radio frequency,
and preferably 3000 ohms.  For a thin, flexible cable such as these speakers
are likely to have, you can make a 1000-ohm choke for 1.8 MHz through nearly 30
MHz by winding 12 turns on a 2.4"-o.d., mix-43, ferrite toroid.  Or you can
wind 3 1/2 turns on a binocular core made of two beads of 0.5" i.d., ferrite mix
31.  The binocular-core choke has the possible advantage of consuming only 18"
cable, whereas the toroidal-core choke uses nearly 36".  The toroidal choke has
the advantage of not requiring you to remove the connector from the cable to
the choke.  The cost is about the same, about $3 per choke.  To get 3000 ohms of
choking impedance

for more info on making such chokes, how much choking is enough, and why.

  I have good ferrite rods (#43) on all leads (audio and power) now and it
helps, but they still squawk.

Rods?!  You get more impedance per dollar cost of ferrite from a toroidal or a
binocular core.  The magnetic circuit of either is closed completely in ferrite,
with no
air in the path.

I doubt that you have enough choking impedance in the cables.  Remember that a
cable just a few feet long and having a free end (as a cable ending at a speaker

has) presents a high (negative imaginary) impedance at HF because it is
"electrically short."  To make much of a dent in its common-mode current you
must insert
much more than this high impedance.  The choke will have more effect if it is
inserted at the end of the cable opposite the free end.  The impedance you
insert should
have a substantial resistive part, so you're not inserting a high-Q loading coil
and making things worse instead of better.  Mix-31 ferrite is better than mix-43
in this

I would replace the rod chokes with binocular- or toroidal-core chokes having
1000-ohm impedance; and I would string two or three such chokes in series to get

2000 or 3000 ohms.  Do NOT try to get more impedance by winding more turns on a
given core; the increased capacitance will give you less, not more, impedance
in the higher HF bands, e.g., 15 meters.  If you're reluctant to make so many
chokes, then double or triple the chokes in just the wall-wart (power) cable,
and then
in the cable to the computer.  I would just choke every cable with 3000 ohms at
the outset, so I wouldn't have to come back.  Enough ferrite is still cheaper
than a
new speaker system, and a new system could be just as vulnerable to RFI as the
old one.  Expensive systems are usually no better than cheap ones, BTW.  RFI is
just not something that manufacturers think about.

Be mindful of the possibility that it's the computer, rather than the amplified
speaker system, that has the RFI vulnerability.  Are the computer's cables
choked?  All
of them?!

  Do you have any ideas?

Is the Pope German?

-Chuck, W1HIS

More information about the RFI mailing list