Feeding Large Loops
Loop antennas have a reasonable feedpoint impedance on all multiples of the fundamental, typically in the range of 100 – 500 ohms (as compared to dipoles that have a high impedance on even multiples). This makes feeding them particularly convenient on harmonically related bands. The harmonic resonances don’t always line up with our ham bands as closely as we might like, but it is possible to get several bands below 2 : 1, and most bands below 3 : 1 or so, depending on the height above ground. When fed with common parallel-conductor feedlines, the impedances in the shack are relatively easy to match.
Feedpoint impedance and resonant frequency
Let’s start by looking at the feedpoint impedance, and resonant frequencies, for the 4 types of 80m loops that we considered in the Large Loop Theory article. Each of these is resonant at 3.55 MHz, and is installed 10 m (33 ft) above ground. A frequency sweep with 8 dips over the HF range is difficult to read with sufficient resolution, so I will resort to tables to display the data. These show the impedance at the resonant frequency when it is inside the the ham band, and at a nominal frequency in the band, with a note of actual resonant frequency.
Note that the exact frequencies and impedances will vary with the specific installation, but these will give you an idea of what sorts of values to expect.
Square Loop fed in one corner
| Frequency (MHz) | Impedance (ohms) | actual resonance (MHz) |
|---|---|---|
| 3.55 MHz | 96 | |
| 7.15 MHz | 90 | |
| 10.1 MHz | 300 – j470 | 10.65 MHz |
| 14.1 MHz | 270 | |
| 18.1 MHz | 450 + j400 | 17.6 MHz |
| 21.0 MHz | 235 | |
| 24.9 MHz | 370 + j300 | 24.5 MHz |
| 28.0 | 210 | |
For the square 80m loop fed in one corner, the impedances on 20m, 15m, and 10m will give an SWR better than 1.5 : 1 with a 200 ohm feed. 40m and 80m have lower impedances (that will increase with height above ground), but still just over 2 : 1. A reference impedance of 150 ohms would improve the match on the lower bands, but may not be worth the additional effort compared with a simple 4 : 1 balun. (A shunt coil across the feedpoint, forming a “beta match” might be useful to raise the impedance on the lower bands without affecting the higher ones as much. Some marginal balun designs might do this accidentally.)
Note that the resonances are at the bottom of the bands on 15m and 10m: this might be good for CW operators, but those using SSB may want to shorten the loop slightly. I have found it most convenient to adjust the loop length for the desired resonant frequency on 20m or 40m, and let the other bands fall where they may.
Square loop fed in the middle of one side
| Frequency (MHz) | Feedpoint impedance | Actual Resonance (MHz) |
|---|---|---|
| 3.55 MHz | 95 | |
| 7.0 MHz | 250 | |
| 10.1 MHz | 250 – j450 | 10.65 MHz |
| 14.15 MHz | 275 | |
| 18.1 MHz | 500 + j370 | 17.6 MHz |
| 21.0 MHz | 410 | |
| 24.9 MHz | 470 + j220 | 24.6 MHz |
| 28.0 | 250 | |
The square loop fed in one side has higher impedances on 40m and 15m, but otherwise a generally similar range. Again, a 4 : 1 balun will give a reasonable match to 50 ohms on multiple bands, or it is suitable for parallel-conductor feedlines.
Triangular loop fed in one corner
| Frequency (MHz) | Feedpoint impedance | Actual Resonance (MHz) |
|---|---|---|
| 3.55 MHz | 90 | |
| 7.0 MHz | 215 + j50 | 6.95 MHz |
| 10.1 MHz | 100 – j320 | 10.5 MHz |
| 14.0 MHz | 290 | |
| 18.1 MHz | 430 + j660 | 17.35 MHz |
| 21.0 MHz | 160 + j120 | 20.85 MHz |
| 24.9 MHz | 450 + j450 | 24.3 MHz |
| 28.0 | 300 + j230 | 27.7 MHz |
The triangular loop fed in one corner also provides a reasonable range of impedances with a 4 : 1 balun. It probably should be tuned up a bit higher in frequency than in this example to move the resonances further inside the band edge on 40, 20, 15 and 10 MHz.
Triangular loop fed in the middle of one side
| Frequency (MHz) | Feedpoint impedance | Actual Resonance (MHz) |
|---|---|---|
| 3.55 MHz | 85 | |
| 7.0 MHz | 260 | |
| 10.1 MHz | 120 – j400 | 10.5 MHz |
| 14.0 MHz | 325 + j60 | 13.95 MHz |
| 18.1 MHz | 610 + j540 | 17.4 MHz |
| 21.0 MHz | 190 + j110 | 20.85 MHz |
| 24.9 MHz | 500 + j510 | 24.25 MHz |
| 28.0 MHz | 430 + j150 | 27.8 MHz |
The triangular loop fed in one side also needs to be shortened a bit to move the resonance further into the 40, 20, 15, and 10m band. The SWR might be over 2 : 1 on 10m using a 4 : 1 balun at the feedpoint, but with proper length adjustment it should provide good multiband operation with reasonable impedances.
summary
All 4 loop varieties have the potential to achieve an SWR of 2 : 1 or better over several bands using a 4 : 1 balun, when the loop length is adjusted properly. While one could choose a particular shape to optimize the match on one or more bands, the radiation pattern (and the positions of the available supports) are likely to be more important factors in that decision. Bandwidth may be somewhat narrow on the higher bands, so an antenna tuner may still be needed, but the SWR is low enough that coax losses shouldn’t be an issue. 300 or 450 ohm parallel lines will work well, too, with a reasonable matching range at the tuner.
If the impedance is too low on the fundamental, then feeding the loop through a quarter wavelength of 150 ohm line (possibly made from dual 75 ohm coax) will step the impedance up to the 200 – 300 ohm range on that band, with minimum impact on the even harmonics.
For non-resonant loops, parallel conductor line to a tuner is an effective approach, as is placing an autotuner at the feedpoint. The most common matching problems will be when the loop uses an odd multiple of 1/2 wavelength of wire, which results in a high feedpoint impedance: this may require some experimenting with feedline length to permit a tuner to match it.