I've got, for the near term, limited resources to put up an HF antenna at my home. One of the options I can manage without excessive cost is a loop, which would enclose the house. I need to transmit on 80m, and likely up to 20m (I understand loops are very tolerant of shorter wavelengths). It wouldn't be as high as I'd like, so would probably perform more like an NVIS, with poor horizontal radiation (given I live in a draw, that's not entirely a bad thing, as less of my energy will go into the ground where it rises).

The only thing is, the space I have pretty much requires the loop to enclose the house, if I'm to avoid entanglements in the dense pine/hardwood mix that covers more than half the lot. How will the steel roof (about 5x20 meters) affect the radiation of an 80m loop (probably hexagonal), mounted close to roof height (roundly 3 meters above ground)?

I'm not certain of the grounding status of the roof panels; it was applied by a "handyman" who seems to do no better work than is expected of him (that is, if you don't know it's wrong, he won't do it right), before I lived in the house.

Here's the property aerial view; the bright white roof with two extensions is mine. The property line follows the fence parallel with the driveway, the near edge of the line of trees next to the plowed field North of the house, and loops into the heavily treed area S and SSE (mix of pines and hardwoods, tangled branches, difficult to run wire into).

Aerial view from Google Maps

As requested in chat, here are photos of the trees that have reasonably open space nearby -- as noted, a mix of pines and hardwoods.

  • 1
    $\begingroup$ It might work well enough to allow you to make some contacts, but let me ask: Do you have any trees or anything else that might serve as an antenna support? Is there room for a guyed mast? A Google Maps satellite view of your land might help us all to see something that you might be overlooking. :-) $\endgroup$ Apr 12, 2019 at 20:22
  • 1
    $\begingroup$ Limited Resources. Masts cost too much. I can support a loop with timber frames made from 2x4 studs if I don't go too high. Those I can afford. $\endgroup$
    – Zeiss Ikon
    Apr 13, 2019 at 15:29
  • 1
    $\begingroup$ Besides the lackluster performance of such a low loop, also consider your proposed design makes an excellent inductive pickup for near-field noise in the house. I suspect you'll end up with something with poor transmit performance and terrible receive performance, and you'll be much better off (better performance, and easier to install) with a mobile antenna on the roof. That metal roof makes a great ground plane: you should take advantage of that. Ground the antenna and you'll only be improving your lightning protection. $\endgroup$ May 10, 2019 at 18:58
  • 1
    $\begingroup$ @PhilFrost-W8II Make that an answer, with data, and you might win the bounty. $\endgroup$
    – Zeiss Ikon
    May 10, 2019 at 19:15
  • 1
    $\begingroup$ @MikeWaters At the time I posted the bounty, that was true; there was only one answer then. Now, with the bounty about to expire, I'm about to review the current answers and likely award the bounty. $\endgroup$
    – Zeiss Ikon
    May 13, 2019 at 10:52

5 Answers 5


Summary: substantial advantage can be realized for every meter increase in the height of the loop. Other parameters, including the metal roof, are much weaker factors in performance.

I simulated a 3.8MHz 1-$\lambda$ hexagonal loop (no wire loss) mounted 3-m above medium-conductivity ground, surrounding a 5m x 20m grid of wires as shown here: enter image description here

Because the loop is so close to the ground (in wavelengths), the 0.2-dBi maximum radiation occurs at the zenith. enter image description here

dropping to -3-dBi at 45$^\circ$ elevation angle. The gain increases less than 0.25-dB when the wire grid is removed, a small change as expected from the low currents reported in the grid wires.

SWR calculations show that the "roof" does have a meaningful effect on the loop's characteristics:

enter image description here

Feeding the loop at the mid-point of a side parallel to the long edge of the "roof" gives somewhat better SWR than feeding the mid-point of one of the "angled" sides, so the loop must indeed be exciting meaningful current in the grid wires. It's possible that the slight gain decrease with the wire grid is due to ground losses, which would couple into the loop, lowering the Q and the efficiency of the antenna. Conversely, feeding at a corner of the loop has minimal impact on the SWR curve, regardless of the presence of the wire grid. The antenna's feedpoint impedance varies considerably as the band is changed; a wide-range antenna matching network will probably be required.

Raising the loop just 1 meter increases the gain at the zenith by 2dB. Another meter increase adds another 1.5dB. The gain at 45$^\circ$ elevation increases at about half the rate of the gain at the zenith.

At the request of another SE member, here are 3D plots that show how the pattern evolves as the frequency is successively doubled:

7.15MHz - maximum gain ~0dBi, impedance 152.6 - J 423.1 ohms

enter image description here

14.2MHz - maximum gain 5dBi, impedance 439.3 - J 1106 ohms

enter image description here

  • $\begingroup$ How would it affect the results if the roof were a little higher; say, 400mm above the loop? $\endgroup$ May 7, 2019 at 10:23
  • $\begingroup$ I thought that the OP wanted to make a rectangular loop slightly larger than his roof. I was going to model that today. (But I now have almost 4000 rep, and you have 2400. You need the +50 rep more than I do. ;-) I suggest you also model the loop like that. $\endgroup$ May 7, 2019 at 15:25
  • 1
    $\begingroup$ @MikeWaters OP says, "How will the steel roof (about 5x20 meters) affect the radiation of an 80m loop (probably hexagonal), mounted close to roof height..." $\endgroup$
    – Brian K1LI
    May 7, 2019 at 15:27
  • $\begingroup$ Okay. I also thought he was interested in DX, but perhaps I'm thinking of another question. Anyway, can you insert the 2D azimuth and elevation plots for 80, 40, and 20 in your answer? That would more clearly show the gain vs. elevation angles. $\endgroup$ May 7, 2019 at 15:53
  • 1
    $\begingroup$ Actually, Mike, I was originally planning an oblong hexagonal loop, much like what Brian modeled above. $\endgroup$
    – Zeiss Ikon
    May 7, 2019 at 16:37

As other answers have established, the presence of the roof doesn't affect the loop pattern too much. However, the proximity to ground does make the antenna quite inefficient, and radiation is mostly upward.

Moreover, I suspect the noise performance of such an antenna will be especially poor. A loop antenna around the house will make a very effective inductive pickup for every noise source inside the house.

So let me propose an alternative design: a mobile antenna on the roof. It would be easy to install and cheap, probably more so than the loop.

The metal roof will make an excellent ground plane, and a vertical can take advantage of that. The roof need not be grounded in any particular way to be effective as an antenna: it merely needs to be conductive.

If the antenna is grounded, the antenna can only reduce your risk of lightning related fire and structural damage. Lightning rods are installed on fire-prone buildings (like barns) so the bulk of the surge current travels through the lightning rod, ground rod, and conductor between. The result is some warm copper. Without a lighting rod the current may go through wood, bales of straw, or other flammable and not very conductive materials, the result is fire.


The space I have pretty much requires the loop to enclose the house [...] How will the steel roof (about 5x20 meters) affect the radiation of an 80m loop?

Although such antenna may work, personally I wouldn't expect it to have an overwhelming performance. From what I know a loop antenna radiates perpendicularly to the plane of the antenna which means it will radiate 50% of the energy to the ground and 50% - to zenith. Thus on 80m only short distance communication will be possible. Also I hope you are aware that a multiband loop antenna requires a tuner (you didn't mention whether you have one).

If you have a tuner my advice would be to give up 80m (who needs 80m anyway? :D) and build a 40m vertically polarized vertically oriented delta loop. Unlike horizontally polarized delta, vertically polarized delta has to be placed only ~2m above the ground to be suitable for long-distance communication.

I've tried such an antenna not a long time ago, here is a scheme:

enter image description here

It works on HF bands from 80m to 10m, although it's not very efficient on 80m. I used a 10m fiberglass fishing rod as a mast which costs ~45$ on eBay. As I understood you have trees near your house thus probably you don't need a mast.

(Note: For 80m double the length of the loop, the distance from the ground and the mast height.)

Another option that doesn't require a tuner is to build a trap dipole. I've built such a dipole for 20, 40, 80m bands once. The antenna turned out to have total length 27.5m. It also worked on 10/12/15/17/30m bands with SWR about 2. It's enough to place the balun about 6-7 meters above the ground, although for better performance on 80m it's better to place it higher.

enter image description here

Hope this will help. Good luck and 73!

  • $\begingroup$ "Who needs 80m, anyway?" Well, I need it. And more importantly, so does the OP. $\endgroup$ Apr 16, 2019 at 15:44
  • 1
    $\begingroup$ That's nice use of existing structure. Sadly, I live in a mobile home; no part of the house (even the ridge line of the roof) is anything like 6m from the ground (max is more like 3.5m). And in order to reach, say, Europe (when the sun cooperates), I'd have to orient that more or less correctly between trees or poles, because as noted, it radiates mostly from the faces of the loop. Don't know much about wave traps at all. $\endgroup$
    – Zeiss Ikon
    Apr 16, 2019 at 15:50
  • $\begingroup$ @ZeissIkon you can just rotate the loop 180 degrees and nothing will change. In this case you will need one mast or tree 6-7 meters high for 40m. For 80m double the length of the loop, the distance from the ground and the mast height. $\endgroup$ Apr 16, 2019 at 16:40
  • $\begingroup$ Right, but I need to get within 30 degrees or so of the direction I'll want, or they'll get half or less of the signal they'd get if they "see" the loop face on. And I still don't have anything but masts or trees anywhere close to 6m. $\endgroup$
    – Zeiss Ikon
    Apr 16, 2019 at 17:11
  • $\begingroup$ @AleksanderAlekseev-R2AUK Do you have pattern plots for that antenna, or better yet, an EZNEC file? I'd like to play with a model. A design like that might be just what he needs. $\endgroup$ Apr 16, 2019 at 17:19

A horizontal loop of 283Ft circumference will work brilliantly. Feed it at any point with 300ohm flat twin window (slotted ) feeder and ideally use 60 feet of this back into a balanced tuner in the shack. Height will not be overly important. You will have high angle radiation on 80m but that will still work long distances after dark. The angle of radiation will drop for 40m band use and get progressively lower still as you work your way towards the higher bands like 20 15 and 10m. At 10 metres you should have eight daisy petal lobes radiating as horizon kissers. It can be a rectangle, circle or heptagon as you like, but aim to keep as open a shape as possible. (Avoid "Letter box" slot shape for example !) Mine was at an average height of 18 feet and was the "best antenna I ever had". It won't break the bank either. Good luck ! Mike EA7JVZ.


To add to Brian's fine answer, I'd like to share some two-dimensional plots to complement his answer and 3D patterns.

To me, 2D plots are also necessary, because they show the actual takeoff angle vs. gain. Generally speaking, the lower the takeoff angle, the farther away that you can work DX.

75 Meters

As was expected by most of us who have decades of experience on the HF bands, this is not an optimum antenna for DX on this band because it is so low in terms of the ratio between its height and wavelength. On 75/80m, it's pretty much omnidirectional and most of the radiation goes straight up. It's an NVIS antenna. Depending upon the time of day, it can work well for stateside work within, say, 750 miles.

75 meter 3D radiation pattern

75 meter 3D pattern

75 meter 2D elevation pattern, gain vs. takeoff angle

75m elevation pattern

75 meter 2D azimuthal pattern, looking down on the antenna

75m azimuth pattern

40 Meters

40 meter 3D radiation pattern 40 meter 3D pattern

40 meter 2D elevation pattern, gain vs. takeoff angle

40m 2D elevation plot

40 meter 2D azimuthal pattern, looking down on the antenna

40 meter 2D azimuthal pattern

20 Meters

20 meter 3D radiation patterns

20 meter 3D radiation pattern20m 3d plot rotated

20 meter 2D elevation pattern, gain vs. takeoff angle

20 meter 2D elevation pattern plot

20 meter 2D azimuthal pattern, looking down on the antenna

20m azimuthal pattern plot

These pattern plots are based on the same EZNEC+ model that Brian was kind enough to share with me.

An important word about the gains

Notice that the gains in these pattern plots are in dBi. That means the gain over a theoretic isotropic antenna. A dipole has a peak gain significantly higher than an isotropic antenna, at the expense of nulls off the ends.

In summary, a ham should have more than one antenna. This loop would be good on some bands, and the vertical as Phil Frost suggested would be good on other bands and under some conditions. The same would be true with a dipole.


You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .