On 6/7/2013 4:12 PM, Jim Garland wrote:
Jim,
I think your writeup is excellent. The only point I would disagree with you
about is your assertion that "ground loop" refers only to magnetic coupling.
I don't think "ground loop" applies to ANYTHING. The word "loop"
properly applies to magnetic coupling.
This is just a matter of nomenclature, of course, which can mean different
things to different people, but to me the phrase includes any situation
where current can flow between different pieces of equipment because of a
difference in AC (or DC) potential. This can happen when there are two or
more ground paths connecting the equipment.
The situation certainly includes magnetic induction, but can also be caused
by other things. The most extreme example for hams probably is caused by not
using a single point ground for their station equipment. If lightning
strikes their tower, then destructive currents can flow along the coax
shield from their antenna, to their $5000 transceiver, into the green wire
on their power cord and to their service entrance ground. A single point
ground, by contrast does not have this problem, since there is no path for
the return current. The entire station just rides up to high voltage during
the strike, but no damage is done.(That's the theory, anyway. When hundreds
of KV are involved, all bets are off!)
The problem with this analysis is that it is simply not possible to bond
things together without somehow forming a lot of loops. For example,
consider any collection of gear in a shack, all interconnected with
shielded cables, AND with every piece of gear with its own wire to some
magic single point. If you have only two such pieces of equipment, you
have three loops formed by the "ground" wires and the cable shields. As
the number of pieces of gear increases, the number of loops increases
exponentially.
In general, those who are the authorities on lightning protection call
for more interconnections, not fewer. At lightning frequencies,
impedance is dominated by inductance, not resistance, so more
inductances in parallel are viewed as an advantage, not a disadvantage.
More parallel paths, and more bonds between interconnected equipment,
tends to minimize the difference in potential between equipment. Yes,
there is magnetic induction from lightning current, but the authorities
consider more bonding to be the better compromise.
The true interpretation of "single point grounding" is that all grounds
must be bonded together by the lowest practical impedance. The "star
ground" technique, where each piece of gear has a 6 ft wire to a
grounding panel, for example, means that there is 12 ft of copper
between those boxes, whereas if the equipment is bonded directly the
length is typically 3-4 ft. The difference in buzz due to IR drop is
10-12 dB!
Henry Ott, a major authority on EMC and a consultant on EMC to most
major corporations, talks about the hidden schematic lurking behind the
ground symbol. He talks about keeping track of where the current flows
-- ALL the current, not only the intentional current. We may have a
simple "audio cable" connecting two pieces of equipment, a "power cable"
connecting equipment to the power line, a "speaker cable" connecting a
speaker to its power amp, a video cable connecting a monitor to a
computer. But Mother Nature knows that every one of these cables is an
ANTENNA, so it can carry RF current, and that those cables connecting
equipment can be carrying leakage current.
It is all of those unintended currents that we must keep track of, and,
in general, we want them OUTSIDE the box, not inside. It is the Pin One
Problem that lets them inside the box. And we avoid Pin One Problems by
connecting cable shields directly to the shielding enclosure.
73, Jim K9YC
_______________________________________________
Amps mailing list
Amps@contesting.com
http://lists.contesting.com/mailman/listinfo/amps
|