[TowerTalk] Radials

[Wilson Lamb]

Nothing to do with towers, but interesting...
Last spring I went through this, with the help of W3AHL, who is an EZNEC whiz.
I read and reread the ON4UN Low Band DXing chapter AND the ARRL chapter and have no problem with the basic theory of what's going on.
The obvious best possible answer, from an effort level and practicality standpoint, is four elevated radials, with some loading, if needed.
That "if needed" is a little fuzzy for me now...
Four radials, each tuned against the vertical element, are as good as many more ground radials.
The only question is whether you can have four clotheslines running across your yard.
I'm not going to recap the whole subject, since I'd do a poor job anyway, but I will describe our results, with which I'm quite pleased!
A friend sent me a design of a 160 vertical with drooping top loading wires, looking like an inverted v, using window line as the vertical member to increase BW slightly.
It took me about three seconds to realize that not only did it look like an inv v, but that it could actually be one, if the top end of the window line were not shorted!
Cutting to the chase, here's what we built, after I pulled a line between two of my tallest trees, with a tape measure hanging down and reading 65+ feet to the ground.
57' vertical window line, 132' T Top loading wire, four 75' radials elevated 7', a few uH of coil between the junction of the radials and the coax shield, no local ground at all
.
The results are great! 
 Z=35 Ohms at 1825 or so, no tuner needed.  It easily works EU with 100W, CW.  Reports from VK are good, with 800W.  I even got a RX report from Tasmania, but his xmit setup was not readable by me.

Then the bonuses:
I don't remember the numbers, but with an L network at the base and 150' of coax, Europe is easy on 80m with 100W, even though it's 45 deg off axis (T Top is E-W).
The dipole idea is very successful.

Another L network tunes it as an 80m vertical, but I haven't pursued that one in operation.  I did build the network and it tunes fine.

On 40, the dipole pattern is four lobed, but not bad.  I haven't gotten around to trying it on the air.

Anyone interested in welcome to the EZNEC files we (Steve) made.

Lessons learned:
Being elevated and resonant, the radials capture most of the current from the vertical element, removing ground resistance and loss, mostly.  It would take many more to make noticeable improvement.

Tuning is necessary and critical.  Each radial was trimmed against the vertical.
The predicted voltages on the radials are no joke.  I made a big black char on a tree when the radial got too close, picture available.

T Tops are great, if you have the space, since the pattern with them is very uniform, no lobes at all.
Shorting the top of the window line is not necessary, making dipole operation possible, verified in EZNEC and on the air.

This antenna is so good one really needs Beverages for receiving, because many people hear it that I cannot hear at all, or barely.
I felt awful when I got replies to CQ but couldn't make out calls.

Of course when we did this, mostly to work Bouvet Is., the DXpedition fell apart and 160 propagation mostly went away, but it was the best project I've had in a while!
Working with a couple of good friends to shoot lines, build and tune the antenna, and whine about Bouvet made it all fun.

I know the OP asked about a Steppir in his yard, but I hope he'll be interested in my story.  I expect four tuned and elevated radials would be a good solution for him, if he can get the radials in the air.

The dipole idea was original to me, but I doubt I'm the first to have it.  

If you want to know more about my experience or to see the EZNEC stuff, let me know.
73,
Wilson
W4BOH