Getting Loopy

Getting Loopy – by John Keating AI6LY / November, 2019

How the hunt for the source of RF interference led to the realization that the resonant frequency of my small transmitting loop antenna moves inversely with ambient temperature.

Why I got interested loop antennas

In 2017 I experienced particularly disruptive QRM in the evenings on 40M. (See nasty bits recurring every 20 kHz in the panadapter image below). Based on rotating my HF antenna, I determined the source was directly south from my QTH. I had seen W0IVJ’s  article, “Locating RF Interference at HF,” in November 2014 QST and decided to build the small receiving loop he described, pair it with a small shortwave receiver, and walk up and down the street.

I tried a few times to locate the exact source, but the antenna really didn’t have sufficient directivity to identify the culprit. So, just for amusement I tried to use it for transmitting.  Predictably, the air variable tuning capacitor arced at low power – about 10W if I recall correctly – so I figured the whole thing needed to be scaled up. But first, I acquired a Pixel Technologies/DXE RF-PRO-1B receive-only loop to investigate claims about its directivity; it was interesting for the obvious lateral nulls, but I found its gain and SNR didn’t compete with my EX-14 tribander.

What is a “Magnetic Loop” antenna?

For quick background, a small transmitting loop antenna consists of a loop whose circumference is between one-eighth and one-third of a wavelength. It is essentially an inductor which can be resonated with a fixed or variable capacitor to form a high-Q tuned circuit. It radiates a figure-8 pattern in the plane of the loop, and exhibits deep nulls broadside. Reasonable performance can be achieved when operated at heights that are only a fraction of a wavelength – slightly more than one loop diameter – above ground level. RF current through the loop creates a strong magnetic field which, in turn, generates an electric field; hence, the name “magnetic loop.” 

The scale-up

Over the course of many months, I home-brewed several loops using various sizes of aluminum and copper tubing before settling on a roughly 1 meter diameter radiating loop of half-inch copper tube, tuned by a large vacuum variable Russian surplus capacitor (see photo), all mounted on a PVC pipe mast.  The 5-250pF 5kV variable capacitor has enough range to tune from 10 to 40 meters and sufficient voltage rating to withstand levels generated at transmit power up to about 300 hundred watts. The capacitor shaft is coupled to a gearhead reduction motor driven by a PCM speed controller whose low frequency control signals are multiplexed on the coax feed line. The feed loop is approximately 1/5 the diameter of the radiating loop, and there is an RF choke at its feedpoint. During early testing, I had to experiment with the vertical position of the feed loop to determine the location for optimum SWR, now typically 1.05:1. I also tried several different variable capacitors.


AA5TB has a nice excel file that generated the following predictions for my design operated at 20m:

Bandwidth = 28.806 kHz (-3 dB points)

Efficiency = 28.806 %

Wavelength Percentage = 13.163 % λ

Loop dc Inductance = 2.407 μH

QL (Quality Factor) = 511.686 

Total tuning Capacitor = 52.736 pF

Capacitor Voltage = 3306.179 V rms (at 100W drive)

Initial testing confirmed the expected high Q/narrow bandwidth characteristic; on 20m it was 23 kHz (3:1 SWR) and 13 kHz (2:1 SWR), with a measured Q of 744.  The tuning capacitance of 49pF (nominal) was very close to the calculated result.  Resonant frequency was very sensitive to changes in capacitance. So, when I set the antenna up for operation and saw a lot of drift, I suspected thermal effects and decided to investigate.

Characterizing the temperature issue

With the antenna set up in the backyard and tuned to an arbitrary point in the 20m band, I logged the resonant frequency and ambient temperature over a period of about 36 hours (without transmitting during that period).  The data, shown in charts below, clearly indicate that resonant frequency varies inversely with temperature. On average, the change was about -1.6 kHz per degree F. The steepest rate of change was 500 Hz/minute which occurred at about 3PM on the second afternoon as the temperature declined from 82F with antenna in direct sun to 75F with antenna in shade.  Given the robust construction of the copper elements, I attribute the drift to the glass-encased capacitor becoming a mini-greenhouse. I plan to perform additional testing with a shade on the capacitor, and also to find the impact on resonant frequency from transmitting at different power levels.

Thermal issues aside, is it a good antenna?

In limited daytime on-air testing, the antenna “did what it is supposed to do.” I QSO’d with a station in Hawaii whose signal was about four S-units weaker on the loop than on my EX-14 tri-band yagi. His signal dropped six S-units when I turned the loop broadside to Hawaii. There was little difference in band noise between the loop and tribander, not surprising since the noise level was very low on the day of the test. Band noise on the loop was two S-units higher when turned broadside (i.e., orientated for roughly North/South figure 8 pattern), while noise increased only about one half S-unit with the tribander aimed similarly, due to its front-to-back characteristics. In transmit, the signal received by the Hawaiian station from the loop was two S-units weaker than the tribander with both aimed at Hawaii.  When I turned the loop broadside, the signal received in Hawaii was seven S-units lower than when the loop was aimed direct. Signal reports were measured during SSB phone exchanges on FTDX-3000 radios at both stations. 

  Loop EX-14 Loop vs EX-14
Signal received at AI6LY, RX direct S-5 S-9 -4
Signal received at AI6LY, RX broadside S-1 n/a  
Signal received in Hawaii, TX direct S-7 S-9 -2
Signal received in Hawaii, TX broadside S-0 n/a  
20m band noise, East-West S-1 S-1  
20m band noise, North-South S-3 S-1.5  

There are many fine articles about small loop antennas. However in my reading I had not seen the thermal issue raised per se. TF2LJ has posted an interesting page on a DIY arduino-based stepper motor-driven autotuner for loop antennas, citing the high-Q as the cause for retuning the antenna if you move your operating frequency by more than a few kHz, but without mention of the temperature sensitivity. In email correspondence to me, VK5KLT made the amusing comment that not only does the small loop make an excellent ambient temperature sensor, but when placed horizontally it also couples to the proximate ground and makes an excellent ground moisture transducer / sensor as well….

To conclude, the antenna works. It is directional (for vertically polarised sky-wave signals arriving at very low elevation angles), and is generally insensitive to ground effects (since the ground does not form the “missing half” of the antenna). Due to the high-Q of the resonator, one must occasionally adjust the tuning to keep the SWR in an acceptable range as it shifts due to changes in humidity and temperature.  The performance limitations and the necessity of retuning are the tradeoffs for utilizing a small footprint antenna.