You don't mention what kind of array. I think it matters a lot.
I was chasing a concern about radial current asymmetry (significant
pattern distortion predicted by several authors) for a top loaded 160m T
with 8 elevated radials which have a 3.5:1 measured current imbalance
highest to lowest due to proximity to a tower for one of them and a
metal building for another. So the coupling even at 15' separation was
significant. I was told this would significantly distort the pattern,
but modeling only shows ~.4db of pattern asymmetry. I modeled measured
radial current asymmetry by putting a current source in each radial at
the measured value (the results look ok, AGT=1.035, but I'm not sure
this is the best way to model).
Then I was told that the I^2 * R loss in the high current radials would
cause excessive loss, but the skin loss calc for 13ga aluminum was 3
watts at QRO in the highest current radial.
From what I've been told/read radials need to by VERY symmetric to
develop the pattern for 4 squares. Here, I think the wisdom is correct
to preserve the phase accuracy from each antenna to the phasing
network. For a dxpedition I built 160m top and inductor loaded resistor
swamped receive 4sq (no amplifiers) per the W8JI ideas and he emphasizes
having an on the ground/buried symmetric radial layout with no cross or
otherwise connections. It worked. Also, K3LR advised me that my DXeng
permanent QTH high impedance receive 4 sq should have identical radials
(length, #, and azimuth) on each antenna. Mine aren't and it works well
but shows one pattern peculiarity, so adding the one needed radial per
antenna is on the to do list. (or removing all of then since the antenna
support ground rod RF ground is likely good enough for the high input Z
amplifier). However, I think all 4 squares are very pattern sensitive
to nearby metal stuff - buildings, towers, other radials and antennas.
ON4UN discusses this in the section about 4 squares.
So depending on the array design, cross connections may matter a lot.
Another wisdom, correct I think, is that wires that cross don't couple
much but close parallel ones will so a physical connection is not
required. So that is something to consider in your radial layout.
Another factor IMO is that buried radials are swamped by the ground
losses so their sensitivity to nearby structures are a lot less than
elevated ones.
It may also be helpful to peruse antennasbyn6lf.com, W8JI.com and
W8WWV's web sites.
I think 4 sq builders have a lot more to say re this topic.
Grant KZ1W
On 5/31/2018 5:39 AM, john@kk9a.com wrote:
In my very limited free time I am installing ground radials for a vertical
array. Some of these radial wires (not elevated) will be near a tower.
Should I attach them to the tower lightning ground system or should they
be isolated or does it matter? The vertical's coax shield will be grounded
to this tower.
John KK9A
_______________________________________________
_______________________________________________
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
|