In the RFI/EMI world the maximum length
of grounding conductors has been found
to be 1/20 of a wavelength. Conductors
longer then 1/20 of a wavelength become
effective unintentional radiators. It is true
that we consider a 1/4 wavelength to be
the most effective and desirable length.
But significant radiation does occur from
even cables as short as 1/20 WL.

In the typical installation even shorter conductors
can become electrically longer due to the bends
a normal conductor experiences. Very few, if any,
real world grounding conductors run in a straight
line. And the typical grounding conductor in a residence
is anything but short and straight. Each bend adds
inductance that increases the electrical length of
a conductor.

Ignoring bends, for various frequencies, 1.20 of a
wave length equates to:
100KHz 150M
500KHz ï€*ï€* 30M
1MHz ï€* 15M
5MHz 0.ï€*3M
10MHz 0.15M
15MHz ï€* 1.0M
20MHz ï€*0.75M

For RFI with frequencies at 5MHz and higher it should be
clear that it is very easy to exceed the magic "1/20�" rule.

This is why efforts to reduce Differential Mode noise must
be applied as close to the noise source as possible.

Other thing to keep in mind that in the normal residence, by NEC,
the ground and neutral are bonded at the meter base, but the house
grounds are bonded to each other in the breaker box and may be
bonded to each other at other points. This creates an wiring maze
that defies any thorough analysis. Grounding conductors can form
an unbelievably effective antenna! This was experienced during a test
at a wired new home that was not connected to any power sources.
In the bedrooms furthest from the breaker box, attaching any of the
power conductors to the antenna connector on a DX398 and battery
powered R2000 proved to be a learning experience. A ground wire
that is an effective antenna will certainly radiate RF noise
effectively.

The house grounds were connected to what the NEC would recognize
as a "meter base" even though the home will operate from a PV array
with a propane powered back up genset. The exterior grounds are much
better then what the NEC calls for.

Terry

wrote in message
ps.com...
In the RFI/EMI world the maximum length
of grounding conductors has been found
to be 1/20 of a wavelength. Conductors
longer then 1/20 of a wavelength become
effective unintentional radiators. It is true
that we consider a 1/4 wavelength to be
the most effective and desirable length.
But significant radiation does occur from
even cables as short as 1/20 WL.

A 1/4 wavelength long ground wire is the WORST possible length. A 1/4
wavelength line acts as an impedance inverter (just as in transmission
line). So the low Z at the ground rod gets transformed to a hi Z at the
radio- making the ground totally ineffective.

Dale W4OP

On Oct 1, 3:09 pm, "Dale Parfitt" wrote:
wrote in message

ps.com...
In the RFI/EMI world the maximum length
of grounding conductors has been found
to be 1/20 of a wavelength. Conductors
longer then 1/20 of a wavelength become
effective unintentional radiators. It is true
that we consider a 1/4 wavelength to be
the most effective and desirable length.
But significant radiation does occur from
even cables as short as 1/20 WL.

A 1/4 wavelength long ground wire is the WORST possible length. A 1/4
wavelength line acts as an impedance inverter (just as in transmission
line). So the low Z at the ground rod gets transformed to a hi Z at the
radio- making the ground totally ineffective.

Dale W4OP

Hi, Dale. On your SW insulated wire antennas, do you then recommend
using a ground at all, or just go with the ungrounded cable-to-wire
junction?

I am seriously thinking about getting two of these gems and phasing
them, if I can make the geometry work on my constricted lot...if I
don't have to ground them, I'd be way ahead of the game for other
reasons. If grounding is better, would separate ground rods be
appropriate?

Thanks!
Bruce Jensen

"bpnjensen" wrote in message
ps.com...
On Oct 1, 3:09 pm, "Dale Parfitt" wrote:
wrote in message

ps.com...
In the RFI/EMI world the maximum length
of grounding conductors has been found
to be 1/20 of a wavelength. Conductors
longer then 1/20 of a wavelength become
effective unintentional radiators. It is true
that we consider a 1/4 wavelength to be
the most effective and desirable length.
But significant radiation does occur from
even cables as short as 1/20 WL.

A 1/4 wavelength long ground wire is the WORST possible length. A 1/4
wavelength line acts as an impedance inverter (just as in transmission
line). So the low Z at the ground rod gets transformed to a hi Z at the
radio- making the ground totally ineffective.

Dale W4OP

Hi, Dale. On your SW insulated wire antennas, do you then recommend
using a ground at all, or just go with the ungrounded cable-to-wire
junction?

I am seriously thinking about getting two of these gems and phasing
them, if I can make the geometry work on my constricted lot...if I
don't have to ground them, I'd be way ahead of the game for other
reasons. If grounding is better, would separate ground rods be
appropriate?

Thanks!
Bruce Jensen

Hi Bruce,
I copied this from the FAQ portion of our site:
What is the best grounding arrangement for the EF-SWL?

Unfortunately, there is no one answer. That is why we have provided separate
grounds on the matchbox. Lug 1 is the ground end of the antenna side of the
transformer while #2 is ground for the coaxial shield. Depending on the
source of noise, it may be advantageous to: 1. Leave the factory shorting
strap in place and attach either lug to a suitable RF ground at the antenna.
2. Remove the strap and ground only lug 1 at the antenna site. Then, the
receiver may be grounded or left ungrounded- whichever results in lower
noise.

So, there is no single answer, but some sort of short ground for the
matchbox will yield less noise than no ground. With no ground at all, the
outside of the coaxial shield is free to conduct noise.
Please note, that we have had to suspend all amateur/SWL mfg for October
while we deal with an unprecedented level of military and aerospace work.
Thanks for your patience.

73,

--
Dale W4OP
for PAR Electronics, Inc.