Topband: 160M Vertical tuning advice needed

[Doug Grant]

My 160M antenna at my island QTH in Maine seems to work pretty well.
But the SWR changes quite a bit depending on the direction chosen in
the Comtek PVS-2 box, and I need to keep retuning the amp when I
switch directions. I suspect I can pick up another dB of gain and
better F/B if I get things tuned right, and avoid damaging my
amplifier from mistuning. The Comtek box is OK, verified by installing
50-ohm loads on each port, and switching directions.

The antenna is a 2-element phased vertical array. Each element is
1/4-wave, and consists of about 50-60 feet of vertical wire, with the
rest of the length sloping up to 80 or 90 feet on the tower in the
middle. I had considerable trouble trying to get each element cut for
1830...my MFJ259B gets hit pretty hard by AM BC interference, and the
MFJ add-on filter did not seem to help. I had to sort of guess at the
lengths when I first installed them.

I spent some time in the woods this week with a borrowed GR RF
impedance meter. It is immune to BCB pickup and makes very accurate
impedance measurements that make sense and that I believe. The unused
element was floating when I made the impedance measurements. The
MFJ259B made a great source, and my IC706 was a fine detector.
Off-null, the S-meter was pinned. At null, the signal was at the
bottom of the S-meter, and in some cases, undetectable by ear.

One of the verticals measures about 35-40 ohms resistive, and X varies
from about -50 to +50 ohms across the band. The plot of R and X vs.
frequency is textbook-example-perfect. It is, however, resonant at
about 1925. I'll be replacing the wire this weekend, and will make the
new one a bit longer, so I get it resonant around 1830. There are
about 25 radials on this one, mostly 1/4-wave long, and covering most
of the azimuth angles. This one is well-behaved.

The other one behaves differently. It is a little closer to the tower
(I have tuned the tower to break it up electrically with a shunt wire
and capacitor, adjusted for maximum current). This vertical had about
15 full-length (1/4-wave) radials and about 20 shorter ones (more on
this later). It is resonant around 1862...X runs +/- about 50 ohms
across the band, just like the other element, but the R part is around
55 ohms...varying maybe +/- 5 ohms across the band. This is obviously
higher than one would expect, and considerably higher than the other
element in the array.

My experience with my 80M verticals was that adding more radials
(always a good thing anyway) lowered the impedance quite dramatically;
going from 15-20 full-length radials to about 40 lowered the R part of
the impedance from 70-90 ohms down to a more-reasonable 40 or so. The
obvious thing to do on the 160 vertical was add more radials.

Today I added 2000 feet of wire in the form of about 20 more radials
to the misbehaving 160M vertical, expecting to see a lower value for
R, closer to the other vertical. Re-measuring the feedpoint impedance
showed no change. Argh. All those mosquito bites and scratches from
the underbrush for nothing.

The radial placement for this vertical is not symmetrical...the
feedpoint is only about 40 feet from the street, about 60 feet from a
property line, and about 30 feet from the lawn (the lawn is going to
be a tough place to add radials for several reasons). There are now
about 30 full-length (1/4-wave) radials where they fit, and about 20
shorter ones in the directions where I can only install short ones.
The full-length radials are mostly concentrated in about a 180-degree
section of the azimuth. There is one 90-degree quadrant with no
radials. The other has radials of 40-60 feet. In other words, looking
down on the radial field, from 0 to 180 degrees, lots of full-length
radials; from 180 to 270, no radials; from 270 to 360, a bunch of
40-60-foot radials.

I am kind of stumped here. I tried creating something that looks like
my antenna in EZNEC5, with varying sets of radials, but it won't give
me a 55-ohm impedance no matter what I do to the radial field in the
model. It also does not give me a higher impedance if I add something
that looks like the tower.

Any suggestions?

73,

Doug K1DG