Inverted-U dipole

I have tried a number of dipoles for the 80 mtr band on my terrace. Given the size of my terrace, a full-size dipole basically out of the question. For the same reason, installation at a decent height ( > 0.2 λ) is also impossible. My apartment is on the top floor of a 7-storey building. So why not install a "long" dipole across the terrace, and let the excess of the legs dangle down the outside wall of the building? Indeed, why not? I can't leave such an inverted-U dipole antenna (ref. 1) installed permanently, but the experiment is worth a shot!

From other experiments I had a 2 x 15 m dipole, hidden in the "antenna cemetery" corner of my wrap-around terrace. I installed the 2 x 6 m center part of it across the terrace, about 2 m off the ground. I attached a weight at the end of the remaining 9 m of each leg, and lowered it along the outside walls. In other words, the antenna is a "bent dipole" that straddles the building. My automatic antenna tuner/coupler (ATU) is located directly at the feedpoint of the dipole.

The 2 x 15 m dipole in "inverted-U" configuration

Diagram of installation at my apartment - top view

Position of the vertical wires - the antenna straddles the building

(source: "Electron" magazine,VERON, September 1999, p. 386)

Obviously, about 60% of the wire length is now directly against the wall of the building, with all its steel-reinforced concrete. And on one side of the building, the wire is even up against the steel rainwater downspout! Despite all this, I was impressed by the results. No, QRM on 80m has not gone down, but the antenna appears to be effective for transmitting! I checked this with an on-line Web-SDR receiver that is located about 1000 km (600+ miles) to the north of my location (my desired direction for 80 mtrs DX). I transmitted a 60 watt carrier for several seconds, and checked the waterfall display of the receiver. Yes!

Waterfall trace at remote receiver - 80m

(9-Feb-2012, 18:30 local time, 60 watt, local sunset 18:15, sunset at RX 17:40)

As the antenna is connected to my ATU, I also did a test on 40 m, where the dipole is about 0.4 λ long: again - success!

Waterfall trace at remote receiver - 40m

(10-Feb-2012, 08:30 local time, 60 watt, local sunrise 08:00, sunrise at RX 08:02)

For fun, I made a "quick-and-dirty" 4NEC2 model of this antenna (ref. 2). Note that I did not model the building or the heavy steel pergola on my terrace!

Geometric model (left) and vertical cross-section along the Y-axis of the radiation pattern

(building etc. is not part of the model; common-mode suppression assumed at feedpoint)

Simulated 3D radiation pattern of my inverted-U

(the building etc. is not part of the model!)

Note that the remote receiver mentioned above is located in a direction about 25° off the y-axis in the radiation pattern diagrams. I.e., not in the direction of maximum radiation.

REFERENCES

Ref. 1: "The Bent Dipole - Do you lack the open space for a wire dipole antenna? Why not bend it to fit?", by Nizar Mullani (K0NM); also appeared in QST, May 1997, pp. 56-57
Ref. 2: NEC-model of my inverted-U dipole (without simulation of building), Frank Dörenberg (N4SPP), September 2015