Hi Guy, and thanks for the reality check. My friend has a 130’ tower with 2
StepIR Yagis on top and he’s been running the reverse- feed for a while. I was
operating the first 6 hours of ARRL 160 contest and we didn’t get skimmer
Spotted into EU for many hours (and weak 02) when other locals were working Eu
from the beginning. It was really evident that serious improvement was needed.
My question was to see if we should make the reverse feed work or go with a
standard loaded tower. The latter is the sanest choice
Regards , Kenny K2KW
> On Dec 3, 2018, at 8:08 PM, Guy Olinger K2AV <k2av.guy@gmail.com> wrote:
>
> Hi Kenny,
>
> You set your curiosity on a real toughie. Tread where angels fear to go, and
> all that.
>
> The N4KG feed is extremely situational in its efficiency, or some efficiency,
> or no efficiency.
>
> It can be modeled in NEC 4, the version of NEC which allows you to have
> conductors both above and in the ground. You need to explicitly model the
> base of the tower down into the ground to see all that is going on, including
> driving RF power into resistive ground, which is why you need to model it.
> The model must also contain explicit description of the tower and all the
> aluminum up the tower.
>
> Then it becomes easy to see what is going on. And why it works over here, and
> can be a dreadful failure over there.
>
> At the point where the coax shield is attached to the tower, the current will
> attempt to go both ways. The current off the braid will be divided between
> going up the tower and going down the tower by Ohms law for impedances at the
> operating frequency. The up and the down paths are not in series, they are in
> parallel. And The down current will be driven into the ground which appears
> as this big ugly lossy resistor.
>
> If the path up the tower, a combination of wires on the tower, yagis yada
> yada looks like a 150 ohm part R part X very miscellaneous impedance, AND the
> path down the tower into the ground looks like thirty-something ohms
> resistive, most of the RF current goes into the ground, wasted. Noting that
> the division of power between the two is proportional to the square of the
> division of current, in that case less than 10 percent of the power goes up
> the tower and the thing performs like a dummy load. Even if you do some stuff
> to compensate, every time something is done on the tower, you risk undoing
> the compensation.
>
> You might try to fix that nasty tower-base-to-dirt resistor by putting down a
> good radial field, to reduce the resistance of the down path to 1 or 2 ohms.
> But then you've just done the hard part of setting up a stoutly performing
> loaded tower, and the gamma/omega match actually can be adjusted to present a
> 50 ohm load to the coax. The whole idea of the N4KG scheme was after all, not
> to have to do ground radials.
>
> All in all, this is why you see and hear of many gamma and omega-matched
> towers that perform well, AND 24 years after the N4KG article you have to ask
> around, and discover only scattered instances of one working well,
>
> There is a way to accomplish what was intended by the N4KG method, if you
> can't actually produce the tower plus antennas above the feed that actually
> IS an electrical quarter wave:
>
> 1) use an inverted L with the bend supported by the tower as the directly
> powered radiator. This takes the natural wide variation and uncertain
> behavior of tower top stuff out of play.
>
> 2) use an FCP for your counterpoise. Plus and minus 33 feet on 160. This gets
> you out of the need to find a 125 foot radius for (in the KG case) for four
> resonant 1/4 wave elevated radials.
>
> 3) have an isolation transformer to a) insure that the initial direct driving
> power only goes to the L/FCP and nothing else, b) the L/FCP is completely
> isolated from any conductors except the supported tower [see 4)].
>
> 4) now, since the L will induce the fool out of the tower which is just a few
> feet away and parallel to the L's vertical wire, counter the induced voltage
> in the tower with an intentionally placed horizontal shorting wire from the
> vertical wire to the tower.
>
> 5) even if you can only reduce the current in the tower to 1/3 of that in the
> L/FCP, by the square law, that means more than 90 percent of the radiation is
> from the L and less than 10 percent of the power is subject to reduction by
> being driven into the ground at the tower base.
>
> This scheme is explained on k2av.com, in the section "65+ Tower L Bend", and
> illustrates what goes on with the RF current down at the base of the tower,
> which is the same issue with the N4KG feed. Even if you could care less about
> an L/FCP, the issue at the bottom of the tower is the same for both, and the
> documentation for a floating L/FCP supported by a tower will explain the
> issue and why you can't ignore it. .
>
> Doncha love this stuff?
>
> 73, Guy K2AV
>
>> On Mon, Dec 3, 2018 at 1:02 PM Kenny Silverman <kenny.k2kw@gmail.com> wrote:
>> I meant to ask if the antenna efficiency (dB) was better with either feed
>> method.
>>
>> Regards , Kenny K2KW
>>
>> > On Dec 3, 2018, at 10:15 AM, Kenny Silverman <kenny.k2kw@gmail.com> wrote:
>> >
>> > Has any analysis even done comparing the N4KG reverse-fed tower feed
>> > method with elevated radials vs a shunt or omega feed with ground radials?
>> >
>> > This is a tall tower and I think it will be close to self resonance or
>> > longer.
>> >
>> > Regards , Kenny K2KW
>> _________________
>> Searchable Archives: http://www.contesting.com/_topband - Topband Reflector
_________________
Searchable Archives: http://www.contesting.com/_topband - Topband Reflector
|