Ken Hough's Website
Amateur Radio

Antennas -- Mag Loops

This section includes the following reports on construction and use of Mag Loops:

What are "Mag Loop" antennas?
Designing Mag Loop Antennas
Mag Loop For 40m
Large Mag Loop For 80m/40m
More experiments with mag loop antennas

What are "Mag Loop" antennas?

Mag Loops (ie magnetic loops) are small tuned loop antennas. They have a reputation for delivering remarkable performance while occupying only small spaces. Mag loops are so named because they are claimed to be most responsive to the magnetic component of electromagnetic waves.

Small tuned loop antennas are defined as having circumferences of less than 0.1 of a wavelength.

In the most commonly used form, a mag loop antenna comprises a continuous loop of thick wire or small diameter tubing having a loop diameter of approximately 1m and with a variable capacitor included in the circumference of the loop. The combined effect of the loop and variable capacitor is to form a high "Q" resonant circuit. RF energy is fed into or taken from the loop via a smaller magnetically coupled loop, or via a conventional gamma matching circuit.

Because of the high "Q" of a mag loop, tuning is quite critical. This means that even small changes in operating frequency will require the loop to be re-tuned. A mag loop can therefore provide pre-selection of incoming signals, possibly reducing interference from strong signals that are higher or lower than the working frequency.

Most mag loop antennas are designed to provide conveniently portable HF antennas that can be used for listening or for QRP transmissions of say 5 to 10 Watts maximum. Variable capacitors will typically have close spaced plates (as used for tuning radio receivers) so that this form of mag loop antenna is NOT suitable for use with high power transmitters.

Mag Loop Antennas For High Power:
These generally operate at significantly higher efficiencies than is the case for lightweight portable loops. Currents around high power loops can reach 50 Amps or so with several thousand Volts being generated across tuning capacitors. High power mag loops are typically made from large diameter copper pipe (22mm or more) and tuning capacitors are most often wide spaced split stator or vacuum variable types.

High power mag loop antennas should always be operated and tuned remotely from a safe distance!

Designing Mag Loop Antennas:

Unlike simple dipole antennas, mag loops can be made to operate over quite wide frequency ranges. Designers have much scope for choice of loop dimensions, power rating, materials of construction, and means of tuning. Choices must be made.

A full theroetical analysis of the working of small tuned loop antennas is beyond the scope of this article -- and probably beyond my capabilities! However, equations that define relevant parameters that are needed to design a mag loop are readily available via the Internet. See:

Loop_Antennas_Glen_Smith_chap05.pdf presents a study of loop antennas in general.
simple.pdf presents simple circuit model of small tuned loop antenna.

Practical Aspects:
the_underestimated_magnetic_loop_hf_antenna_vers1.1.pdf a look at mag loops in practice. includes equations to calculate operating parameters and images of working loops. discusses the principles, uses , and contruction of loop antennas. more on practical apects of mag loop antennas -- includes practical points about tuning capacitors.

Some Practical Designs: presents a working design for a mag loop to operate on 14MHz and 21 MHz. gives a design for a loop that covers 12MHz to 32MHz.
Foldable Mag Loop 40m-17m -- an Elecraft aplication note. -- a list of designs for mag loops

The general equations relating to mag loops have been incorporated into easily used web based calculators and stand-alone calculator programs such as: -- a web based calculator.
aa5tb_loop_v1.22a.xls -- a standalone spreadsheet calculator program. -- includes "loopcalc.exe" and "capcalc.exe"

I have found "loopcalc.exe" to be very helpful when designing mag loops. This program makes it very easy to experiment with various dimensions and to see how changes affect performance of a mag loop. The companion program capcalc.exe provides a quick way to determine dimensions for tuning capacitors.

Some Observations About Designing Mag Loop Antennas:
1. Loop size:
The overall dimensions of a loop will determine the range of frequencies over which it can be used. Increasing values of tuning capacity will tune a loop to lower frequencies, but at decreasing efficiencies. The maximum possible operating frequency of a loop is limited by the self capacitance of the loop. Typically, a loop of 1m diameter will self resonate at around 25MHz or so.

Increasing the diameter of the wire or tube used to make up a loop will decrease it's RF resistance and therefore increase operating efficiency.

2. Tuning capcitors:
For operation at high power, loop tuning capacitors must be able to withstand voltages of several thousand Volts at currents of 50Amps or so. Rotor wiper contacts of single stator variable capacitors are not suited to this duty. Wide spaced split stator or vacuum variable capacitors are recommended for operation at high powers.

3. Safety:
Care should be taken to ensure that personnel (including children and pets) cannot come into contact with the very high voltages that can be generated by mag loop antennas.
My Initial Experiments With A Mag Loop Antenna:
I made up a 0.9m square loop using 15mm copper pipe and standard 90deg connectors. All joints were soldered. I used a twin gang tuning capacitor that had been removed from an old AM/FM broadcast radio. This effectively gave me two split stator variable capacitors of 15pF + 15pF or 315pF + 315pF -- useful for easy tuning of high frequencies or low frequencies respectively. I used a simple gamma match circuit to connect 50 Ohm coax into the loop.

This loop was initially intended to be used for listening only. With the antenna standing indoors at near ground level, results were impressive! Sensitivity was very good and with the addition of fixed capacitors, I was able to cover all of the HF range up to approximately 25MHz. Selectivity was also good. The antenna performed well as an RF pre-selector. My only problem was that the tuning capacitor did not include a slow motion drive, making tuning fiddly.

I replaced the original tuning capacitor with a wide spaced split stator variable capacitor of approximately 50pF + 50pF, added sufficient fixed capacitance needed to be able to tune to over the 40m band, and arranged to drive this remotely. This might seem to be easy, but it was quite a project -- see below. The antenna was then moved up into my loft space where I could compare it's performance with that of my 40m/20m dipole.

Listening tests were again impressive! Depending on location, S meter readings from the two antennas were similar!

The finalised version of this mag loop (see below) was capable of handling sensibly high powers. While using this antenna at 50Watts on 40m, I received a 5 and 8/9 report from a station on the Isle of Sky, over 300 miles away.

A Working Mag Loop For 40m:

This mag loop was based on the same loop that is described above -- ie 0.9m square/15mm diameter copper pipe. I decided that it would be optimised for operation on the 40m band and be capable of operating at up to 100Watts.

This antenna was to be installed in my loft space, so a remotely controlled tuning system was needed.

1. Two compression type connectors have been included in the vertical sections of this loop. Basically, the loop was cut in two to be able to take it up into the loft. Signal reports suggest that these connectors have a negligible effect on the overall RF resistance and performance of the loop.

2. Large coax choke balun used in feeder cable.

Tuning Capacitance:
"loopcalc.exe" showed that at a total tuning capacitance of approximately 140pF would be needed to tune to the 40m band. At 100Watts, voltages of up to 4.2kVolts could be generated across the tuning capacitor, so a wide spaced capacitor was needed. Initial experience showed that tuning on this loop was very sharp. I decided to use a low value variable capacitor in parallel with a larger value fixed capacitor. This arrangement could provide easy tuning over the 40m band.

Variable Capacitor:
I obtained a single stator wide spaced variable capacitor of 150pF. This item was made using a "bolt-together" construction, so it was a fairly easy matter to re-build it as a split stator capacitor of 50pF + 50pF.

Note: Ideally a tuning capacitor for a mag loop should use a fully welded structure.

Fixed Capacitor:
This item proved to be more of a problem.

Given a tuning capacitor of 50pF + 50pF (ie an effective capacitance of 25pF), an additional fixed capacitance of approximately 125pF would be needed to tune the loop over the 40m band. Not having, or wishing to pay for, a high voltage vacuum capacitor, I decided to make up a fixed capacitor using double sided epoxy/glass PCB.

Initial tests run at a power input of 5 Watts showed the loop to tune well across the 40m band and for tuning to remain reasonably stable. However, when power levels were raised towards 100Watts, tuning drifted significantly off frequency due to dielectric heating of the epoxy/glass PCB material used to make up the fixed capacitor.

Double sided PTFE based PCB is expensive and difficult to obtain, so I decided to try making up a capacitor using RG213 coaxial cable. RG213 presents a capacitance of approximetely 100pF/metre. For convenience, I cut two pieces of RG213, each measuring just over half a metre, and connected them in parallel across the tuning capacitor. Using this arrangement, tuning over the 40m band remained quite stable while running at 100Watts. Problem solved!
Motor Drive:
The mag loop was to be installed in my loft space, so some form of remotely controlled motor drive system was needed to operate the tuning capacitor, preferably one that could show the position of the tuning capacitor at the control site. ie in the shack

This proved to be the most time consuming part of the project. Three possibilities came to mind:

1. A DC motor controlled from a variable voltage supply. The motor would need to be heavily geared to be able to drive the tuning capacitor with any degree of finesse. In addition, some form of position indicator would be needed that was electrically coupled back to the control position.
2. A stepper motor. This would need an appropriate stepping controller. Again, some form of position indicator would be needed, although with a stepper motor, this could simply be a micro switch that could be used to detect/set a correct startup position.
3. A servo motor unit. Integrated servo units are available that will adopt a certain position depending on the pulse width of a continuous pulsed signal. No other position indicating device is needed.

I decided to use a servo motor drive.

About Servo Motor Drives:
Servo motor drives are mass produced for use in the field of model making and so are reasonably cheap. They include a small DC motor that is geared so as to produce high levels of torque. External rotation is typically limited to 180 degrees or so, and this is linked to a position sensing potentiometer. Internal electronics allow position to be set depending on the pulse width of a train of square wave pulses.

I used a small servo unit that was obtained from -- part No. FS35Q . This unit requires a supply voltage of 5V. The control signal must provide a pulse repetition rate of approx 18msec with a pulse width that can be varied from 0.75msec to 2.25msec. I used the circuit given by Rob Paisley, but with timing components chosen to suit the servo that I used. The modified circuit is shown below.

An oscilloscope can be helpful when setting up this circuit. If timing is incorrect, the servo might not work reliably or might simply "jiggle about".

No additional position indicating circuitry was necessary, because this could be inferred from the position the knob attached to the shaft of R2.

RF Interference:
Initially, the servo was badly affected by the very high RF fields close to the tuning capacitor. It was necessary to wrap the servo in aluminium foil and to wind several turns of the servo power/control cable onto a toroidal core. After doing this the servo could be controlled very well -- BUT only at low power. I adopted the procedure of tuning to frequency while using only 5Watts of RF and then switching off the controller so that the servo would remain at the set position. I could then increase RF power to 100Watts without any problems.
Coupling signal to/from the loop:
I retained the gamma match connection into this mag loop. It worked well for the original version and there's no reason why it should not continue to work for the updated higher power version. Of course, I did include a coax choke balun in the 50 Ohm coax feeder to the loop.
Operation and Performance of the loop:
This mag loop was installed in my loft space at a similar height and location to that of my 40m/20m dipole antenna. I manage connections to my HF antennas via a switch box that is situated in my shack, so I can easily conduct comparative tests between the two antennas.

Allowing for the different radiation patterns of the two antennas, listening tests have shown only small differences in sensitivities of the two antennas -- the mag loop being less than one "S" point down on the dipole. This is quite a small mag loop for use at 40m, so this sensitivity seems to be quite reasonable.

On transmission, both antennas perform quite well. For example, during a comparative test done during a QSO with a station over 300 miles away in the Isle of Sky I received signal reports of 5 and 8 to 9 for both antennas while using 50 Watts.

Large Mag Loop For 80m AND 40m:
I have recently built a larger mag loop antenna designed specifically for use on the 80m and 40m bands This comprises an octagonal loop made from 22mm copper pipe, tuned via a 550pF vacuum variable capacitor. Each side of the loop is 1m long, giving a loop circumference of 8m. According to "loopcalc.exe", at 40m this new mag loop should be approximately 7 times more efficient than my old 1m square loop. At 80m the difference should be nearly 20 times!

Why did I decide to restrict myself to 22mm copper pipe? Surely, larger diameter pipe would provide higher efficiency. Yes, but.......

Calculation showed that at 80m and 22mm pipe, the operating bandwidth of the loop would be only 3kHz or so -- just about wide enough for SSB operation. Larger diameter pipe would have resulted in even narrower bandwidth. Operating experience has shown that 22mm pipe was a good choice.

Although I have described this as a "Large Mag Loop", at 80m it still (just) qualifies as a small loop. ie the circumference is equal to 0.1λ.

Construction of the loop was fairly straightforward, but with a width of 2.4 metres this loop was too big to manage directly. A more easily manageable support structure was needed. This was built from square section GRP tube with the various pieces fastened together by means of cross lap joints and glass fibre/resin.

I had decided to use 38mm x 38mm square section GRP tube with 3mm wall thickness, but the supplier could only offer me 5mm wall thickness at the same price. The extra stiffness of the 5mm wall thickness has turned out to be a better choice.

The loop was clipped onto the cradle by means of ordinary plastic pipe clips, which also serve as stand-off insulators. Here's a view of the a layout of the parts for the loop, support frame, and weatherproof box to contain the tuning system.

Eight 1m lengths of 22mm copper pipe were cut to make up the loop. These were joined into a complete octagon using 45deg capillary connectors. I used ordinary lead/tin solder to ensure joints were completely filled. I then cleaned back all exposed solder so as to leave the minimum "length" of solder across each joint.

There are many references on the Internet about the need to use silver solder to make joints in a mag loop. In short, I don't believe them and offer the following to support my belief:

Construction of the tuning unit was without doubt, the most complicated and time consuming part of this project. Whereas small/portable mag loops can be tuned fairly easily, the very narrow bandwidth of a high effiency/high power loop requires a far more precise tuner that can withstand several killoVolts at many Amps.

The optimum component for this duty is a vacuum variable capacitor. Expect to pay £300 pounds or so for a new one. I bought a second hand one via Ebay for significantly less. Loop currents can reach 30Amps or more, so it is important to keep loop resistance as low as possible. The tuning capacitor was connected into the loop using heavy copper straps. See below.

The vacuum variable capacitor includes a drive shaft requiring 15 turns to move the capacitor from minimum to maximum setting. This is a helpful slow motion drive, but still nowhere near enough to provide the fineness of tuning needed for this loop.

The drive shaft of the variable capacitor was driven from a DC motor via a speed reduction train comprising two series connected ball race type reduction drives plus a pully/belt drive onto the motor, giving an overall speed reduction of 1800 : 1. This has turned out to be just about optimum.

This mag loop was to be sited and operated remotely from the shack, so I needed some means of indicating the position of the tuner. This was done by including a 10 turn (50kΩ) potentiometer that was coupled (via 1.5:1 gearing) directly onto the drive shaft going the tuning capacitor. The position of the potentiometer, and therefore of the varicap, could then be read remotely via the tuner control box (see later).

The loop was to be sited outdoors, so a weatherproof housing was essential for the tuner unit. The picture below on the left shows the upper part of the loop and the means of clipping it onto the cradle, and the weatherproof box containing the tuner system. A close up view of the tuning system contained in the box is shown on the right. Note the use of a nylon drive shaft to provide reliable electrical insulation between the tuning capacitor and the drive mechanism.

Please excuse the elastic band shown coupling the motor to the drive train. It has since been replaced by a proper rubber drive belt.

Gear wheels and motor were salvaged from old printers. At 12Volts, the no-load current of the motor is approx 50mA. Stalled current is approx 500mA.

(To see more detail use 'View Image' in your browser)

A reversible DC supply was used to drive the motor, as indicated on the left below. The resistor R was selected by trial to give the slowest reliable drive speed.

The potentiometer was fed from a stabilised +5V supply, and it's position was read using the circuit shown below on the right.

I had intended to replace this simple motor control system with something better (eg PWM drive), but it has continued to work satisfactorily as it is.

RF coupling into the main mag loop was achieved by means of a screened Faraday loop shown here on the left. This arrangement minimises he pickup of electrostatic noise (local QRM) that might otherwise pass back to a receiver. RG213 coax was use to make this Faraday loop as it was stiff enough to retain shape with minimal support.

I applied the recommended practice of using a coupling loop of 1/5 of the diameter of the main mag loop, and this proved to be correct. At 80m and at 40m, optimum tuning results in SWR readings of 1:1. At higher frequencies, SWR increases slightly, but can be optimised by flattening the loop somewhat.

Visual impressions might suggest a poor/lossy coupling between the two loops, but at resonance this arrangement works very well indeed, and has advantages in that:

--there is no direct electrical connection between the loops
--no breaks are made in the main loop to insert a toroid transformer which could adversely affect RF resistance of the loop
--no need to include a toroid core around the main loop conductor during construction

Connections within the Faraday loop were made weatherproof by wrapping with self amalgamating tape, followed by over-wrapping with ordinary PVC tape.
This antenna is REMARKABLY QUIET, with practically no signals at all being heard off resonance. When I first tested this antenna, it was so quiet that I thought that there must be a serious fault. Eventually, I discovered that I was tuning too quickly and was simply missing the very sharp resonance.
The very sharp resonance of this loop was confirmed by using the "spectrum scope monitor" facility on my Yeasu FT897D. I simply tuned across the resonant frequency of the loop while the receiver gave "S" meter readings of the radio background noise. As predicted by 'loopcalc.exe', bandwidth was indeed very sharp!

This mag loop must be re-tuned whenever the operating frequency is changed, but using the system described above, this is not difficult. My procedure is:

Using my Kenwood TS590S:

1. Tune radio to required frequency
2. Using the coarse/fast setting, sweep the loop tuning across this frequency while listening to background/signal noise. Adjust loop tuning for maximum signal/background noise signal.
3. Switch TS590S to AM mode with power setting at only 5 Watts. Then press PTT and observe SWR reading. Using fine/slow setting to tune loop to minimum SWR (normally 1:1).
4. Switch TS590S to preferred mode of operation (normally SSB) and operate normally.

Unlike the Yeasu FT897D, the Kenwood TS590S retains pre-set power settings when changing operating mode. This means that I can leave the AM setting at 5 Watts (to minimise interference to others when tuning up) while retaining a higher power setting when I switch over to SSB mode.

Tuning range of 2.4m diameter/8m circumference mag loop
This mag loop can be tuned from below the 80m band up to JUST include the 20m band. Of course, it can be tuned to ALL of the amateur bands between these limits. At 20m, the vacuum variable tuning capacitor is at the extreme limit of it's range. If I were to re-build this antenna, I would make the loop slightly smaller so as to more easily include the 20m band.

Location of the loop and some practical aspects.
The 2.4m diameter/8m circumference mag loop is shown on the left located in my very restricted back garden/yard. The structure is held upright by a length of plastic pipe fixed (and hinged) between an adjacent wall and the upper part of the support frame. This allows the loop to be oriented to suit preferred directions of propagation.

A (white coloured) five core cable can be seen descending vertically from the box containing the tuning system to meet with the coax cable that feeds into the coupling loop situated near to the base of the main/tuned loop. From there the two cables pass on to a weatherproof box, and from there cables lead indoors.

A closer view of the Faraday coupling loop is shown below. This is supported on two lengths of plastic pipe that are fastened to the main support using plastic cable ties.

(To see more details use 'View Image' in your browser)

In calm weather, this arrangement works very well. However, strong winds caused the base of the support to "walk", which in turn caused the assembly to fall down. Also, there was a tendency for the loop to slip out of the plastic pipe clips. Both of these problems have been solved as follows:

1. Guy ropes have been fitted to the ends of the upper cross bar of the support frame, and these have been tied to suitable fixing points. This means that if the orientation of the loop is to be changed, then the guy ropes must be reset -- a minor inconvenience, but effective.

2. Short lengths of nylon cord have been tied close to the pipe clips between the main loop and the support frame to prevent the loop from becoming detached from the clips.

These measures have enabled the loop to withstand all weather conditions so far.

So how does it perform?
My first tests compared the sensitivity of this loop with a 40m/20m inductively loaded half wave dipole and with my smaller 0.9m square "experimental" loop. Results were very encouraging. At 40m, the big loop was about as sensitive as the dipole, and was noticeably more sensitive than my smaller experimental loop.

I don't have any other antennas that can tune at 80m, so cannot do similar tests at 80m. It does load well to give SWR = 1:1 and can handle 100Watts. Comparisons of signals received via this loop with reports given by other stations suggests that the sensitivity of this antenna is similar to most practically mounted dipole and dublet based systems.

I typically receive 5 and 9+ or even 9++ reports when using this antenna. The local Cumbrian fells can completely block ground wave propagation in some directions, so that for some local QSOs I have to rely on propagation via NVIS. However, I am often able to receive (and give) 5 and 9+ reports for these QSOs at linear ground distances of only 30 miles or so.

This mag loop works well for normal DX propagation and for NVIS propagation.

More experiments with mag loop antennas

Major factors effecting the efficiency of mag loops are:

1. Material of construction. This is typically either copper or aluminium with copper generally accepted as being the better choice.

2. "Width" of the loop conductor. In the case of circular tubing, this is the circumference of the tubing, and for a thin flat conductor might be expected to be (?) equal to twice the actual width of the conductor.

3. Frequency of operation. Skin effect resistance increases approximately as the square root of frequency. At high frequencies this effect seriously reduces the efficiency of mag loops. (more-on-the-skin-effect)

To learn more about how significant these aspects might be, I built and tested the following two mag loop antennas:
A loop 8m in circumerence made of a 100mm wide strip of aluminium foil.
This antenna was constructed to compare Andy Choraffa's design (ref: “Making an efficient loop aerial using aluminium foil”, RADCOM, page 54, Feb 2014) with the 8m copper pipe loop described above. The loop was made from a single 100mm wide length of ordinary cooking foil, and was tuned using a manually operated split stator variable capacitor. It was situated indoors but at a similar height to that of the 8m copper loop that is situated outdoors.

It was suggested that the large surface area of the aluminium foil would provide reduced RF skin resistance and hence improved efficiency.

Comparisons of signals received via the two antennas showed the foil based antenna to be LESS sensitive by approx 1 S point and also (consequently?) to have broader tuning than the copper based 8m loop. It seems unlikely that the use of 100mm wide aluminium foil would give any advantage in sensitivity over 22mm copper pipe. In any case, the fragility of the foil based loop makes it unsuitable for use outdoors.
A 1.2m circumference square loop made from 22mm copper pipe.
This loop was tuned via a 90pF + 90pF wide spaced split stator variable capacitor to cover the 10m and 6m bands, to explore how well a mag loop would perform on these higher frequency bands. A small DC motor with worm reduction gearing was used to drive the tuning capacitor. RF coupling was achieved via a screened Faraday loop.

At these higher frequencies, RF losses in the loop due to skin effect become very significant, so tuned bandwidth was expected to be wider than for the high efficiency loop described above. This meant that tuning should be correspondingly less critical, which did prove to be the case. At 10m/6m, atmospheric noise/QRM is relatively low, so that the small amount of local QRM generated by the DC motor provided a helpful signal for tuning purposes.

Sensitivity of this antenna was found to be significantly less than that of a 15m/10m/6m fan dipole, particularly at 6m. Unless available space was VERY restricted, I would not recommend mag loops for use at these higher frequencies.