[TowerTalk] Engineering approach to RF grounding

[Blackburn]

I've been following the recent threads regarding RF grounding with some
interest, and finally can't resist the temptation to throw in my two cents
worth.  The following observations are the result of 25 years of
professional experience in electromagnetic compatibility and communications
systems/facilities engineering, plus almost 30 years of ham radio.

To begin with, there is a lot of confusion resulting from the use of the
term "ground" for what are really a number of disparate and sometimes
conflicting purposes.  For example, "ground" can refer to, among others, a
fault-protection path (the green-wire ground in your house, for example),
lightning downconductor paths, signal reference points or planes, RF
reference points or planes, antenna counterpoises, static discharge paths,
etc., etc., etc.. 

To design an effective "ground" system for a particular installation or
application, one must understand what the purpose of the "ground" is, and
how the various "ground" systems necessary to meet the safety, lightning
and RF requirements interact with each other.  Lets start with a review of
some basic points which may make selecting a particular ground
configuration a little more straightforward:

First, current can ONLY flow in loops, even at RF (remember the grade
school electrical project about completing a circuit between a battery and
a bulb), and at any single physical point the instantaneous sum of currents
entering and leaving the point must be zero (in other words, current is
neither created nor destroyed).

Second, the physical earth is NOT a 'sink'  into which we can dump unwanted
RF or noise currents (it can be part of a current path, just like any other
conductor, although earth is lossy at RF).  As noted above, the net current
entering and leaving any point (even one as large as the earth) must be
zero.  One source of confusion here stems from the fact that at RF, part of
the current "path" may be an electromagnetic field rather than the physical
movement of electrons, with the result that currents induced by RF fields
may APPEAR to flow into or out of the physical earth without an obvious

return path.

Finally, in order to protect ourselves and our families, and to keep the
insurance companies happy, we MUST meet the requirements of the National
Electrical Code, or the local equivalent (how many hams know there is a
section on Amateur Radio in the NEC?).  The NEC is concerned primarily with
personnel shock protection and fire prevention, and it is recognized that
sometimes the NEC requirements may appear to conflict with good RF
engineering procedures.  The NEC also represents a minimum set of
requirements designed to provide reasonable personnel hazard and fire
protection, particularly in the area of grounding.  For example, a
reasonable system for conducting lightning currents to earth without damage
significantly exceeds the minimum requirements of the NEC.

>From an RF standpoint, what we are really trying to do in an amateur
installation is to minimize the risk of RF interference or personnel RF
shock by minimizing RF potential differences (which in the majority of
cases result from common-mode currents flowing through finite impedances)
in the station, and between the station and the surrounding area.  Note
that nowhere in this statement is there anything to construe that the
station must be referenced to the physical earth at RF to achieve this
goal.  There are instances where it is appropriate to provide an RF
reference to the physical earth, but that's a lot harder to achieve than
one would imagine.  Some years ago one of the federal agencies published a
handbook on data processing facilities which included a very interesting
chart showing measured RF impedance of typical ground leads…the values are
a lot higher than one would intuitively expect.

Remember the statement above about current always flowing in loops?  Think
about what happens with an RF transmitter.  RF current flows in one
conductor of the feedline, and an equal and opposite (we hope) current
flows in the other, completing a loop which includes the feedline,  the
antenna, and the radiated field.  In this case, the net (i.e. common mode)
RF current into and out of the transmitter is exactly zero, and except for
potential differences (which we minimize by providing a low-impedance
connection between the various equipments) created by circulating currents
generated within the station equipment,  there is no RF current and
therefore no RF voltage or locally generated RF fields present to create
interference, REGARDLESS of whether the station is "grounded" or not.

Unfortunately, this ideal situation rarely if ever exists in practical
installations.  Common-mode currents are induced by interaction of
wiring/feedlines with the RF fields intentionally radiated from our own
antennas, by poor terminations on the feedlines which induce common-mode
currents (not necessarily related to SWR, by the way) at the antenna, plus
any number of other mechanisms.  One very effective way of minimizing
interference that has been mentioned in previous posts is to use common
mode chokes on the feedlines and other conductors connecting to the station
(for example, I was able to completely eliminate interference to my station
computer by nothing more than the simple expedient of winding 8 - 10 turn

coils in the coaxial feedlines where they enter the house).  Note also that
there is nothing in the NEC (at least that I can find) that can be
construed to prohibit the use of common mode chokes on the power feeders. 
Such  a choke will essentially isolate the safety "ground" from the station
at RF while still providing the necessary fault current path, thereby
eliminating another potential RF current path.  This could be a real
concern on the lower bands, 160 meters in particular.

I suspect that a lot of the experiences reported where interference
problems are solved by improving the station "ground" really have very
little to do with how well the station is actually referenced to the
physical earth at RF frequencies.  I suspect what actually happens in a lot
of these cases is that the improved "ground" system reduces unintentional
RF currents flowing in the station equipment by providing an additional
path for induced common mode currents, or by moving or eliminating
resonances.   Resonances in the system comprised of station wiring, antenna
feedlines, power feeders, and "ground" wires can result in standing waves
in the station wiring, with RF voltage or current maximums appearing at the
station equipment.  The commercially available "artificial ground", which
is actually a tuneable counterpoise, works by changing the resonance of the
system to move the high-voltage or high-current point away from the station
equipment.  To my way of thinking, the use of judiciously placed
common-mode chokes to break up resonances, coupled with whatever other
measures are appropriate to minimize common-mode currents, coupled with
good RF bonding within the station equipment, is a much preferred approach.

First priority in any ground installation must be safety, particularly with
making sure that appropriate fault protection measures are applied.  The
next priority in the design should be lightning, both for safety and to
minimize damage, and because a lot of the techniques necessary to control
lightning current paths (i.e., low inductance connections to earth at the
tower, buried feedlines, single-point entry panels, etc.) also enhance good
RF performance.  Finally, once the fault protection and lightning issues
have been adequately addressed, we can apply an understanding of the
mechanisms by which RF interference is created, along with some good RF
engineering techniques, to control interference.  In a ham station, these
would be primarily techniques such as common-mode chokes, RF bonding
between equipment, and ensuring that antenna feeds do not force common-mode
currents back down the feedlines into the shack.  If you have LOTS of
money, you can use a building-sized equipotential reference plane, with
everything referenced to this plane, like commercial & government
facilities often do.

This post is not intended to provide a specific "ground" design, or to
advocate one particular configuration over another, but rather to point out
some of the physics involved and the mechanisms by which RF interference
problems happen, along with the engineering processes by which we can
arrive at a design which meets all the necessary requirements.  Specific

designs must be tailored to each installation, and providing sufficient
design guidance for areas such as fault protection and lightning, let alone
RF, is way beyond the limits of a reasonable length post.

Remember, unless one has a good understanding of the physics and mechanisms
involved, it is very difficult to generalize application of a specific
technique that may have "worked" in one particular installation.  One tenet
of good engineering is that one does not make technical decisions based on
a single data point, unless one understands the source of the data point
well enough to ensure that all potential variables are accounted for.  The
same approach needs to apply here . . . stick with proven practices based
on sound engineering principles, use the configurations that are
appropriate for your particular installation, and don't be swayed by the
vast amount of folklore, pseudo-science, and just plain dangerous practices
floating around the amateur community.

73,  Keith  KB6B


--
FAQ on WWW:               http://www.contesting.com/towertalkfaq.html
Submissions:              towertalk@contesting.com
Administrative requests:  towertalk-REQUEST@contesting.com
Problems:                 owner-towertalk@contesting.com
Search:                   http://www.contesting.com/km9p/search.htm