Just thought I'd add a little to W2YR's article.
The purpose of adding ferrites is to put a common-mode impedance in series
with the I/O wiring. Impedance is a function of frequency and the amount
of inductance. Inductance is in turn a function of the core material
(within the frequencies for which it is designed) and the number of turns
in the inductor. For a given core material, inductance can be increased by
wrapping multiple turns through a single core, or by stacking more that one
core in series. I've had reasonable success on controlling noise from
things like TNC's by wrapping as many turns as will physically fit through
a couple of MFJ split cores stacked together, and these type of cores are
relatively easy to get. Core saturation is generally not an issue when
used as common mode chokes, as the intentional currents introduce equal and
opposite fields into the core material.
Part of the problem with obtaining low frequency materials in split core
form is that the FCC (and corresponding European) emissions limits apply
primarily at VHF frequencies. The limits are set primarily to control
interference to TV's; except for AM broadcast where you are typically
dealing with strong signals, the vast majority of the consumer market is in
locations where interference to HF communications systems is not an issue.
Therefore, most of the market for snap-on type ferrite cores is concerned
about controlling VHF emissions.
Finally, make sure the various parts of the computer cabinet are well
grounded to each other, and the computer is well referenced to the same RF
ground as the radio. Common-mode chokes in the computer power leads also
sometimes help, although again it takes a lot of inductance to have any
real effect. Make sure the cores are located as close to the offending
device as possible.
Hope this helps,
> From: W2YR@aol.com
> To: firstname.lastname@example.org
> Subject: Fwd: [RFI] PC keyboard RF immunity?
> Date: Thursday, December 24, 1998 1:01 PM
Administrative requests: rfi-REQUEST@contesting.com