On Fri, 26 Dec 2003 14:50:35 -0800, Cortland Richmond wrote:
>Looks to me like interference at the color burst detector.
>If external video signals don't have problems with interference, this
>pretty much nails it down. Somehow, the HF signal is getting into the IF.
Not necessarily -- this burst detector would also be looking at a
composite video signal (i.e., plain ordinary video on coax).
>This is unlikely to be via the normal mode antenna signal, It is more
>likely to be a common-mode path. Common mode may be broken at the antenna,
>by using back to back 300 ohm baluns and a balanced 300 ohm HPF between
>them, It may also be broken at the AC power, with a sufficiently inductive
>common-mode filter, or by resonant common mode filtering in the bands of
Agreed on all of the above, except that my weapon of choice would be
ferrite chokes using lots of turns around ferrite materials that are
most effective in the spectrum of interest. This generally makes type
43 a poor choice.
>It MAY be due to direct pickup. Direct pickup on the PC board might be
>treated by shielding what is receiving the signal. Internal wiring may be
>the antenna for this, and RF chokes on audio, control and power wires
>should help with THAT.
It could also be what we in the pro audio world call a "pin 1 problem"
-- that is, the shield of an audio, video, or rf cable is not going
straight to the shielding enclosure, but instead goes to the circuit
board where it wanders around for a while before finding its return to
something that current flow on the shield considers "ground." Much of
that "wandering" path will be in common with various parts of the
signal circuitry, so any voltage drop due to current across the
impedance of that path (mostly inductive) will be added to the signal.
Pin 1 problems are a VERY common mechanism of RFI coupling, and the
injected RF often appears at unexpected parts of the circuitry relative
to where it actually comes into the hardware. For example, as part of
my recent research, I injected RF into the shield contact of input #1
or a console or compressor/limiter, and looked for detected RF at the
output (signal return for the generator was the chassis). With
virtually every piece of gear I tested, I found significant RF at the
output when the gain of the input stage I was injecting was turned all
the way down. What happens is that due to this "wandering around," the
RF gets into signal circuitry all through the DUT, not only the stage
where it is injected.
Note that this is quite different from wiring inside the TV acting as
an antenna. But the pin 1 problem IS a very important way that common
mode current can enter electronics.
Jim Brown K9YC
RFI mailing list