I have a system with two gamma matched towers that works pretty well. I
did all I could to make them appear to be similar. The gamma dimensions
and capacitance are identical - and I used a loading wire on top of one of
them - making it longer and longer until the match was similar.
After that - I used current probes to check the current levels in the tower
- and they looked fairly close. Then I went driving around a few miles
away checking the pattern and found it to be very nice. I see probably 15
or 20 db F/B ratio from the end fire pattern. It has some gain in the
broadside pattern as well. For awhile - I had relays setup so I could
detune one of the towers and compare signals with just one tower fed. Once
I had confidence in the system - I removed that feature.
K6SE was the first guy I saw doing this - with much shorter towers than
mine.
So - it can be tricky - but if your towers are pretty similar in dimensions
and you can use identical parameters for the gamma match - then you might
be surprised how easy it is to get working.
Tree N6TR
On Thu, May 30, 2013 at 3:53 PM, Tom W8JI <w8ji@w8ji.com> wrote:
> I did try it with two shunt fed towers on 160 meters 1/4 apart using a
> Wilkinson power divider and had little success >>>
>
> Because of mutual coupling, antennas with unidirectional patterns have
> grossly dissimilar impedances at each element. The elements, provided they
> have reasonable loss, have nowhere near similar impedances and nowhere near
> the impedance as a single element.
>
> Elements in two-element arrays require equal currents to have deep nulls.
>
> Since the impedances are grossly different, there isn't any divider that
> will supply the design phase and current ratios to low loss elements. It
> takes a phasing system custom designed for the actual impedances involved.
>
>
> <<and then read in ON4UN’s Lowband book that phasing shunt fed tower was
> very difficult if possible at all. >>
>
> That is not correct.
>
> I successfully phased a G5RV (100 ft high as a "T") against a 130 foot
> shunt fed tower. It had extremely deep nulls with three patterns, and I
> had four patterns. I had unidirectional NE, unidirectional SW,
> bidirectional NE/SW with deep nulls NW and SE, and broadside (which was
> almost omni because of the close spacing).
>
> Shunt fed towers do not behave like normal tower because the shunt wire
> acts like an additional length of transmission line. This alters feedpoint
> requirements, and the elements no longer require equal currents at the
> feedpoint for deep nulls like a current maxima fed element, or equal
> voltages like a voltage maxima fed element. They require something between
> a voltage fed element's requirements of equal voltages and a current fed
> elements equal currents, the exact requirement dependent on the shunt
> characteristics as a transmission line. The shunt also introduces phase
> shift that must be allowed for.
>
> While it is not an easy cookie cutter task, it is not nearly impossible.
> It can actually be pretty fast to set up, if the phasing system is
> adjustable in delay and ratio.
>
> Wilkinson's and similar only supply the required ratios and phase when
> terminated in the design impedance, and that does not happen with
> unidirectional patterns in small endfire arrays (unless significant loss is
> added). When the division and phase errors caused by mutual coupling
> creating dissimilar impedances is combined with the shunt feed adding phase
> shift, and the shunt changing required ratios in voltage and current, it
> probably would not work very well. That doesn't mean phasing shunt systems
> cannot work, or even that the job is difficult. It is just different.
>
> 73 Tom
> All good topband ops know how to put up a beverage at night.
> _________________
> Topband Reflector
>
All good topband ops know how to put up a beverage at night.
_________________
Topband Reflector
|