TopBand: EWEs

[Tom Rauch]

> Date:          Wed, 01 Apr 1998 15:49:39 -0500 (EST)
> From:          k6se@juno.com (Earl W Cunningham)

Hi Earl,

I appreciate your comments. I just wanted to point out Eric was dead 
on about a few things, at least from what I observed here and others 
report.

> I agree with the 180-spacing statement, but I'm unsure how the vertical
> sections are 180 out of phase.  Is this because one vertical element is
> upside-down compared to the other?

Yes. One is base fed, the other top fed. That provides crossfire 
phasing, which when the element spacing is less than 1/4 wl always 
fires towards the feedpoint end of the array. (Like in a log 
periodic.)

> The horizontal section radiates broadside to the wire, but to a much
> lesser degree than the vertical sections.

Especially less over a large near perfect ground. If the ground is 
large and near-perfect the velocity factor of the top wire (acting 
as a transmission line) becomes closer to free-space (ideal is a bit 
faster than freespace, not slower) and the current taper is reduced 
because loss due to radiation and ground losses below the wire are 
reduced.
    
> "K9AY loops are the same way.  Stick them on a crummy ground, and they
> very likely won't work as well as they can."
> 
> In modeling, I can't get the K9AY antenna to perform anywhere near as
> good as the plots in his article in QST, no matter what the ground .  My
> modeling shows the Ewe to be superior to the K9AY loop.

That makes sense, since the K9AY system has two VERY closed 
spaced "elements". The signal sensitivity is poor, while the close 
spacing makes it very susceptible to flaws in the system. 

They don't seem to work *well* here also, no matter what I do. The 
one I tried did work better over a ground screen, but no better than 
a small regular old loop as far as I could tell.

> Modeling shows that a 4-el endfire/broadside has as good or better
> pattern than a good Beverage, BTW.

They play that way here. My 4-el transmitting vertical array almost 
always works better than the Beverages for receive, unless I phase 
two Beverages together. In Cleveland Ohio the best RX antenna was a 
four element array of ten foot verticals spaced 350 feet broadside 
and 70 feet endfire. It blew the Beverages and in line loops away.

> Travelling waves?  I thought that applied to Beverages, not Ewes.  Please
> elaborate.

If current is uniform or slowly tapers (due to loss from radiation 
and resistance) in the system and contains no or greatly 
suppressed "end reflection", it has no standing waves and can be 
called a traveling wave system. 

The "waves" travel to the terminated end where they are
dissipated, preventing reflections that cause standing waves. A 
dipole has standing waves, a terminated rhombic is a traveling wave 
system.

> How can the currents in the two vertical sections of an Ewe possibly be
> equal?  Modeling shows up to over 70dB F/B.  It also shows no change in
> results (other than feedpoint lack of reactance) in an inductively loaded
> Ewe vs. no loading..

The ideal current distribution is equal. Less than ideal still works. 
Check your yagi out, for example. Ideal is two equal size driven 
elements with equal currents (two element case), yet a less than 
perfect yagi still works great.

Re-thinking what you are saying, I see what you described now. (I'm 
at a disadvantage because I modeled the antenna in my head, and the 
software was flawed). The inductor at the base of the fed section 
can't hurt anything because the feedline exciting the second vertical 
element is the horizontal wire at the top, and that isn't disturbed 
by the additional coil at or near the bottom of the fed element. 
Neither is the current distribution in either radiator. Very good!

> Ewe modeling results must be close to the real world, otherwise there
> wouldn't be so many happy users of Ewes.

I'm positive no one ever said they "don't work". I would still be 
cautious relying on any model of something as complex and 
non-homogenous as real earth, especially when there is little or no 
verification of the model. 

How the person and program models the system is another problem. It's 
easy to overlook simple things in the model since the system can be 
made "perfect" far beyond what the real world allows, and design or 
model a system that can't be built even though it looks simple.



73, Tom W8JI
w8ji.tom@MCIONE.com

--
FAQ on WWW:               http://www.contesting.com/topband.html
Submissions:              topband@contesting.com
Administrative requests:  topband-REQUEST@contesting.com
Problems:                 owner-topband@contesting.com