On 6/7/2013 12:54 PM, Roger Parsons wrote:
I think that a technically correct generic name for the 'Pin 1 Problem' would
be unwanted 'Common Impedance Coupling', and a bit of a search shows that this
term has been quite frequently used.
No, it is NOT the correct description.
Joe hit the nail on the head, and Paul got a lot of it right.
Old time broadcasters who "never saw these problems in big consoles"
were almost certainly working on gear built in the old days, when a jack
was screwed down to the chassis, and the shield went straight to the
chassis. Pin One Problems began to abound when mfg got too efficient
for its own good, with connectors that mounted to the circuit board, but
forgot that shields had to make contact with the chassis.
Now I'm going to really arouse folks -- there is no such thing as a
"ground loop!" When power line "buzz" couples into unbalanced wiring,
it is because the chassis of two pieces of equipment being connected are
at different potentials, and that potential difference is added to the
signal. The fundamental cause of the potential difference is nothing
more or less than LEAKAGE current from the AC power system. The path
for this current is capacitance between the "high" side of the AC power
line and the chassis, which must, by law, be connected to the "green
wire" of the AC line, that current returns to its source (the power
system) via the green wire, and the current produces IR drop in the
green wire. That IR drop is what we hear.
The leakage current consists primarily of the HARMONICS of 60 Hz, which
is why we hear "buzz" rather than "hum" (pure 60 Hz). Why harmonics?
Because nearly all equipment we connect to the power line as a load has
a capacitor-input power supply, which causes current to flow in
relatively short pulses at the top and bottom of each cycle, which in
turn makes the current rich in harmonics.
As if that weren't enough, in three-phase systems, "triplen" harmonics
(any harmonic number divisible by three) will ADD both in the neutral
and in the ground. Few of us have 3-phse in our homes, but the vast
majority of homes are fed by 3-phase distribution systems. As a result,
almost all of us will see a lot of triplen harmonics if we do an
spectrum display of the power line, and of the voltage between one
chassis and another of equipment plugged into different outlets.
So what we incorrectly call a "ground loop" and do stupid things to
avoid, is really AC leakage current, and the cure is VERY simple.
1) Minimize the voltage difference between one chassis and another by
getting power for everything that will be interconnected from the same
AC outlet, or from outlets that share the same green wire, or from
outlets that are bonded together by short fat copper.
2) Bond with short, fat copper from chassis to chassis of every piece of
gear that has an unbalanced connection.
#1 takes the IR drop of the long "green wire" from the outlet to the
panel out of the picture, so the remaining IR drop is only the short
(typically 6 ft) line cord. #2 helps two ways. First, shorts out the
remaining potential difference by brute force. Second, it diverts that
leakage current (and most of the RF that might be picked up on a short
cable) to the chassis, so it also cures any Pin One Problems that might
be present.
In the tutorials I do for pro audio/video trade shows, I call this
technique "local bonding," emphasizing the fact that it only works where
cable lengths are short enough that the resistance of the bonding
conductors is low.
As I see it, there is one, and only one, good time to use the word
"loop" in a discussion of noise coupling, and that is when the coupling
mechanism is the MAGNETIC field. With all magnetic coupling, the
strength of both the field and the sensitivity of a system to a magnetic
field is the LOOP AREA of the coupling path. In our stations, we
commonly see magnetic coupling in three situations.
1) Leakage flux from big power transformers close to our equipment
and/or its wiring.
2) Wiring errors in the power system that establish a magnetic field.
Perhaps a neutral is bonded to the green wire at more than one point, or
a load is connected between "hot" and "green," or an outlet is mis-wired.
3) We feed a long wire antenna that ends in the shack, and thanks to the
length of the antenna, there's a current peak near our equipment. That
current produces a strong magnetic field. As has been observed, a cable
shield provides NO magnetic shielding at power frequencies, and TWISTED
PAIR is VERY effective at rejecting magnetic fields.
73, Jim K9YC
_______________________________________________
Amps mailing list
Amps@contesting.com
http://lists.contesting.com/mailman/listinfo/amps
|