On Sun, 29 Feb 2004 14:47:16 +0100, Franki @ ON5ZO.be wrote:
>Could it be that the root of the problem is the amplifier/surround
>decoder?
Where have ferrites been tried, what specific types, and how have they
been used? The basic function of a ferrite in RFI suppression is to
turn the cable that you use it on into a choke. At higher frequencies,
the resistive loss component dominates the inductance. But what you are
trying to do is stop RF current flow on the cable. In other words, the
equivalent circuit of a wire with ferrite on it, or a cable wrapped
around a ferrite core, is series R and L. From an RFI suppression point
of view, you don't care which dominates, you just want as high an
impedance as possible. Remember that the L increases as the square of
the number of turns.
There are many ferrite materials, and they behave differently with
respect to frequency. The more common materials targeted at VHF RFI are
generally less effective in the HF spectrum (the #43 material made by
Fair-Rite, for example). The peak rejection frequency of a ferrite is
moved down in frequency, and the impedance of the ferrite is increased,
by wrapping the cable multiple turns around the core. The reason the
frequency moves down is, I believe, largely due to stray inductance
across the winding, but there may be other reasons.
I've been able to achieve good levels of impedance and rejection of HF
RFI by putting a lot of turns around a large toroid. A visit to the
Fair Rite website will yield a pdf of their 100-page catalog with
excellent applications notes and data on their products. I've found it
quite helpful in the research I've been doing on RFI into audio
systems. (I have no connection to them).
In general, the chokes need to be as close to the detection point as
possible. Don't forget loudspeaker cables, which are a VERY common path
for entry into power amplifiers. Are the power amplifiers separate
units? Are they built into the loudspeakers? Amplifiers built into
loudspeakers tend to be VERY RFI prone. I suspect poor shielding, but
also pin 1 problems.
Another common problem is parallel loudspeaker cable (what we call ZIP
cord in the US) -- twisted pair wiring makes a system FAR less
susceptible if coupling through the loudspeaker outout terminals is the
problem, and lots of RF problems are elminated by replacing zip cord
with twisted pair. It's funny, in a perverse way -- the purveyors of
the esoteric cables that one of my engineering friends likes to call
"audio jewelry" don't know any real physics at all, so they don't know
that the huge parallel-wire loudspeaker cables they are selling
contribute to this problem. Sure, the root cause is the electronics,
but the wrong cable makes it worse.
The "pin 1 problem" is another name for improper termination of cable
shields within equipment. Cable shields SHOULD go directly to the
shielding enclosure, but in lots of equipment, they go to the circuit
board first. When they do, any current on the shield (like that induced
by your transmitter) wanders around the circuit board until if finds a
return path (remember, even though your friend calls it an audio cable,
mother nature calls it an antenna). The path on the circuit board has
impedance (mostly inductance), and the voltage drop across that
impedance gets coupled into the audio circuitry. By the way -- the pin
1 problem can also be a factor on output wiring, including loudspeaker
wiring (even though it's not shielded).
Another important point. When a pin 1 problem is the path into the
equipment (what Neil Muncy calls "letting the fox into the hen house),
it is VERY common for the RF to get into some part of the circuitry far
removed from the wiring where it is being coupled. That's where the
"wandering around" comes in -- what circuits it gets into depends on
the path to "ground." In other words, the RF might be coupled on
output cable #1, but the RF can be getting into input stage #2, or even
the mix buss.
There are two fixes to the pin 1 problem. 1) Connect the shield
straight to the chassis. 2) Stop current flow on the shield. That's
where the ferrite chokes come in.
If your neighbor would be impressed by some official-looking stuff to
read on the topic, you can find four papers that I've written about it.
Two were presented in Amsterdam last May, two more in New York last
October. Go to aes.org, then publications, then convention preprints,
then search on my name. A pdf download costs $10. I don't get any of
the money.
Jim Brown K9YC
_______________________________________________
RFI mailing list
RFI@contesting.com
http://lists.contesting.com/mailman/listinfo/rfi
|