A Comparison of Closed and Interrupted
Loop Antennas for 40 Meters

L. B. Cebik, W4RNL (SK)

In "The IL-ZX Antenna for 40 Meters", I presented (or resurrected, depending upon one's point of view) a compact interrupted loop antenna for 40 meters. By using folded element wire construction, it provided a coax-compatible feedpoint impedance with no compensating or loading components. Since the overall circumference of the interrupted loop was about 1/2 wavelength, the antenna was very compact, fitting within a 20' wide by 20' High (plus ground clearance) footprint. Fig. 1 on the right shows the essential outline of the ILZX when used vertically. Single wire horizontal versions of the antenna exist. Indeed, in Britain, you may obtain a multi-band version of the antenna.

On the left in Fig. 1 is an antenna that is similar in size and that also uses a vertical orientation. It is a closed loop with a diameter of about 0.127 wavelength, with a resulting 0.4-wavelength circumference. As we move from the region of very small loops with feedpoint resistive components in the 1-Ohm range up to the medium-loop range (circumferences between about 0.25 and 0.75 wavelength), we find some interesting properties. First, the resistive component of the feedpoint impedance climbs so that we no longer need worry as much about the losses of compensating and matching components or the losses of construction joints. Second, the reactance of the closed loop becomes increasingly inductive. When the loop is electrically about 1/2 wavelength in circumference (which for a closed loop is physically larger than 1/2 wavelength), the reactance reaches a peak inductive value only to suddenly reverse to a peak capacitive reactance value with only a slight further increase in circumference. (This phenomenon is familiar to those who have center-fed linear wire antennas that are abut 1 wavelength long.) At the same time, the performance of the loop improves with increasing size. The result is a compromise. When the loop is about 0.4 wavelength in circumference, The feedpoint resistance approaches 100 Ohms while the inductive reactance has a high but manageable value for which we can compensate with a small (low-pF) series capacitor. By tradition--derived from very small loop construction more than from necessity--most closed loops in this arena use fat elements--often copper pipe.

Both antennas are interesting, if for not other reason than the similarity of their sizes. One can square the closed-loop circle or circularize the square shape of the ILZX. However the shapes have little bearing on performance. The closed-loop's circle is convenient for the most commonly used materials, while the wire structure of the ILZX lends itself to the used of non-conductive side supports with rope ties to the corners of the square. Therefore, in the discussion to follow, I shall used the modeled construction shown in Fig. 2, which gives the dimensions for both subject antennas.

In a situation calling for a very compact 40-meter antenna, the structure is likely to be close to the ground. I selected a 5-meter (16.4') bottom height to have a rounded number that accords reasonably well with amateur practice. In both cases, the top height of the antenna is less than 11 meters (or 35') above ground. The radius of the closed loop is 0.0635 wavelength at 7.15 MHz, the selected common test frequency for both antennas. The loop material is 1" copper, a diameter that results in a 98.5% power efficiency according to NEC model reports. (The NEC report does not include losses incurred from the average ground over which I placed both antennas). For an important reason that we shall consider shortly, the dimensional outline of the closed loop does not show the position of the feedpoint or of the required series capacitor.

The ILZX has several notable features. It uses AWG #12 wire (0.0808" diameter). Although the wire is thin compared to the value used in the closed loop, the power efficiency is over 96%. Instead of viewing the antenna as an interrupted loop, let's think of it as a folded dipole with 3" spacing between wires and with the linear elements bent into a square that is 5.5 meters (18.04') on a side. Like a folded dipole, the equal-diameter elements create a 4:1 impedance transformation (regardless of spacing--within limits). Hence, a single wire version of the antenna might show a feedpoint impedance in the 12- to 16-Ohm range. The folded version shows an impedance in the 50- to 65-Ohm range, depending on orientation and height above ground. With the side feedpoint shown, the impedance is about 64 Ohms.

The difference between a linear folded dipole and the bent version in the ILZX is the proximity of the element ends, added to the parallel sections of the "top" and "bottom" sections. The element tips exhibit strong coupling. Therefore, the gap between them becomes an important means of setting the reactance at the operating frequency. Note that the tips come to a point on each side of the gap. If we leave the tips blunt--as we might in a regular folded dipole--the gap dimension becomes very finicky. By bringing the tips to a point, we reduce the amount of reactance change with each unit of physical change in the size of the gap. Such antennas actually go back to the 1930s and sometimes used copper pipe construction (on unbelievably heavy wood frames) with gap extensions that consisted of small plates soldered to screw threads for fine tuning.

The sketches shows the ILZX at a relatively low height, vertically oriented, with a side feedpoint and a side-gap position. This orientation yields the best low angle patterns that we can obtain from the antenna. In contrast, most common implementations of the closed loop have chosen a bottom position for the feedpoint and reactance-compensating capacitor. In fact, the closed loop has properties sufficiently like a very small loop to allow us to position the feedpoint and the required capacitor almost anywhere along the circumference, and not necessarily at the same place. Each selection has consequences that we may accept or reject according to our needs. For example, a very small loop has a current magnitude and phase that remain virtually constant along the length of the loop. In the 0.4-wavelength circumference loop, the current magnitude changes by no more than a 3:1 ratio of maximum to minimum. This change is small compared to the current levels that we find along a linear element. As well, it is small compared to the ratio of maximum to minimum current in a full 1-wavelength loop.

Let's assume that the terms "top," "bottom," and "side" have conventional meanings relative to the ground. We may place the feedpoint at any one of these positions. Likewise, we may place the series capacitor at any one of these positions. The following table shows what happens to the maximum gain, the elevation angle of maximum radiation, and the feedpoint impedance for various combinations. In all cases, the compensating series capacitor value remains constant and represents a reactance of -j2417 Ohms at 7.15 MHz. As well, the closed loop remain physically constant.

With the feedpoint and capacitor both positioned at either the top or bottom, the pattern for the relatively low and vertically oriented loop is mostly straight up. The dominant polarization is horizontal. Fig. 3 shows the broadside and edgewise elevation patterns for some of the cases. The left pair of elevation plots yield the most NVIS-like upward patterns at a reasonably good gain level. The right side of Fig. 3 shows the elevation patterns for the use of a bottom feedpoint and a top-positioned capacitor. However, the patterns also apply to the case in which the feedpoint is on the side and the capacitor is at the top. The top-mounted series capacitor pattern has a significant lower angle component, but only edgewise to the plane of the loop. These cases appear to illustrate the fact that the position of the series capacitor has a stronger bearing on the pattern shape than the feedpoint position. For example, the table suggests that the feedpoint at the bottom with the capacitor on a side yields patterns very much like those where both the source and the capacitor are positioned on a side.

Only two of the options present a highly workable feedpoint resistance: bottom-bottom and side-side. The side-side position combination does require adjustment to the capacitor value to 9.31 pF to null the remaining loop reactance. The very small amount of required change (0.1 pF) suggests that tuning the loop can be very finicky without either special components or excellent ingenuity.

Fig. 4 compares the elevation plots of the side-side closed loop and the ILZX. In the configuration shown in Fig. 2, the ILZX shows a maximum edgewise gain of 0.05 dBi at 24 degrees. The maximum edgewise gain is -0.26 dBi at 25 degrees. The average gain of the two antennas is almost identical, while the ILZX exhibits a slightly more circular azimuth pattern. (With the ILZX fed at the bottom and the gap at the top, the resulting patterns are similar to those for the closed loop in the bottom-bottom configuration.)

When oriented at relatively low heights, both the closed loop and the ILZX benefit from side feeding to yield low angle patterns that benefit HF communications. Indeed, their patterns are not sufficiently different to be detectable in ordinary operations. The remaining question is whether there is a more decisive factor to separate the two antennas for amateur operations. There might be, if we assume that most amateurs prefer wider operating bandwidths from their antennas.

Fig. 5 presents the SWR sweeps for the closed-loop and the ILZX from 7.0 to 7.3 MHz. In each case, the curve is references to the resonant impedance of the individual antenna. For the ILZX, the reference impedance is 64 Ohms. The 98.5-Ohm reference impedance of the closed loop includes the use of a 9.31-pF series capacitor at the side feedpoint. The 2:1 SWR bandwidth of the closed loop is 60-70 kHz. In contrast, the 2:1 SWR bandwidth of the ILZX is about 150 kHz. As well, even without 50-Ohm matching at the feedpoint, the rate of SWR change for the ILZX is low enough that the internal tuners that come with many current transceivers could easily handle the matching task. At 40 meters, the losses of coaxial cables larger than RG-58 would not be troublesome for most operations. Nevertheless, for maximum 40-meter QRN reduction, the narrower bandwidth of the closed loop may serve a useful purpose.

When we lay out the physical and the electrical properties of both antenna types, each has advantages and disadvantages. The point of these notes is not to recommend one over the other, but to make the relative properties of each more readily apparent. Perhaps the only general conclusion to these notes is the fact that if we construct either antenna in a vertical plane and at relative low heights, then side feeding is generally highly beneficial for long distance operations, altyhough bottom feeding can create a compact NVIS antenna. Enjoy the interesting conundrum. . .

Return to Amateur Radio Page