I guess the answer depends on what you consider to be a significant loss.
For an elevated radial vertical, the amount of loss you get from a
ground connection varies depending on the exact configuration of the
antenna system, number of radials, the ground quality, and height of the
radials. Some who have analyzed N4KG's inverted fed 80M vertical, which
is an elevated radial system with a tie to ground thru the tower, have
calculated a loss of 0.8 dB using NEC4. I analyzed it with NEC2 and got
-1 dB, but I would expect the NEC4 analysis to be closer. The answer
depends a lot on what makes up the ground system. If the ground system
has a set of radials in addition to ground rods, the gain actually goes
up instead of down.
The amount of risk you are willing to take by not grounding your
antenna, in return for a dB of gain is a matter of personal judgment.
There are some schemes that can reduce this risk such as gas discharge
tubes, spark gaps, etc., but they require a ground system.
You will get some degree of pattern distortion from this connection if
the ground system that you are connecting to is not symmetrical. Pattern
distortion with an elevated radial vertical is very difficult to
control. Differences in radial length of only a few inches can create
significant pattern distortion. The height above ground of each of the
radials, their spacing from each other, the ground quality under each
radial, and the way in which you connect the radials together at the
center all contribute significantly to pattern distortion. N4KG
connected his radials together INSIDE the tower to help reduce pattern
distortion. If you are very concerned about pattern distortion, a buried
radial system would be a better choice. In either case, the largest
contributor to pattern distortion is usually other antennas in the
immediate vicinity of the vertical. Even non-resonate towers contribute
a significant amount.
You are indeed fortunate if you live in a location that can produce a
gain of 2.7 dBi from an elevated radial 1/4 wave vertical. That usually
requires a ground dielectric constant of 20 or more. For most of us, a
low band single 1/4 wave vertical with elevated radials usually produces
close to 0 dBi gain.
An accurate calculation of ground losses has always been a problem.
There are lots of variables with ground quality being a major factor.
All of the software programs have various degrees of inaccuracy when
computing ground losses, so you have to temper these results with
studies and measurements made by qualified individuals. These do not
always agree in value, but all agree on the trends.
Jerry, K4SAV
Dennis OConnor wrote:
>There is considerable research that indicates that grounding the elevated
>radials for RF harms the pattern more than just a little... At -1dB hit, and
>likely more... If the vertical is only +2.7dBi gain to begin with a -1dB hit
>is significant... At my station another dB is roughly $2000 so losing a dB
>is significant to me...Worse than the loss is the distortion of the expected
>pattern... When working weak signal DX 1db is often the difference between
>making the Q and not making it... For rag chewing with the locals on 75
>meters 1dB is meaningless. and you might as well ground everything.
> If DC grounding the radial system is important to you then do it through a
> husky parallel tuned circuit so that the RF on the radial is maintained above
> ground potential.. And put a common mode choke on the coax at the feed point,
> and if necessary repeat the choke 1/8 wave back from the antenna...
>
> denny / k8do
>
> __________________________________________________
>Do You Yahoo!?
>Tired of spam? Yahoo! Mail has the best spam protection around
>http://mail.yahoo.com
>_______________________________________________
>
>
>
>_______________________________________________
>TowerTalk mailing list
>TowerTalk@contesting.com
>http://lists.contesting.com/mailman/listinfo/towertalk
>
>
>
_______________________________________________
_______________________________________________
TowerTalk mailing list
TowerTalk@contesting.com
http://lists.contesting.com/mailman/listinfo/towertalk
|