[TowerTalk] FW: The PERILS of single point grounding

[Jim Jarvis]

We've beaten to death the notion of 'common' and 'gnd' being tied
together at one point only...the service entrance.  NEC is clear
on that.

What we haven't done is carefully examine what happens with our RF
SPG, as it relates to the power ground, and the chassis grounding
of our equipment.

Picture this installation... we install a perimeter ground around the
house using strap.  We bond to 8' ground rods at each corner of the house,
and extend strap radials outward from the perimeter, as we can.  Then,
we make ONE connection to the electrical system ground, at the service
entrance, and add another 8' rod at that point.  So far so good.

Then...we run all our coaxes and control lines through a grounded panel,
which may or may not be adjacent to the electrical service entrance.
Then...we plug our equipment into the power system in the shack.

In the event of a lightning strike,  the surge impedances of the various
paths to the equipment via RF gnd and power gnd will be different.  Maybe
a little, maybe a lot.  A potential difference will be generated between the
grounds, and across some aspect of the equipment.

Can you really HAVE a single point ground?  Even if the RF and power service
entrances were immediately adjacent, and tied together with 3" strap...the
drop circuits from each will follow different paths, and may have induced
current on them from the primary strike...with different potentials due to
their geometries.  I still disconnect everything in the main shack.

In a 24/7 broadcast installation, or amateur repeater, you clearly can't do
that.  You take reasonable precautions, and hope for the best.

Thoughts/comments?

n2ea
jimjarvis at ieee.org