ECLECTIC AETHER - Adventures with Amateur Radio

Moebius loop antennas

Some initial notes relating to the performance of muti-turn Moebius loop antennas.

Main focus is based on the Youloop design published on the Airspy forum, but the versions used for these tests are not an actual commercially produced Youloop

Single turn of coax forming the loop with a crossover at mid position.

As a test, I built three loops of 500mm diameter, to dimensions as shown in the original Youloop drawing, but using 75 Ohm WF100 satellite TV coax consisting of a 1.00mm solid plain copper conductor, foam polyethylene dielectric, copper foil tape and a plain copper wire braid and a 1:1 transformer.

The loops were constructed as three different types, but as much as possible was kept the same in each loop.
  • Moebius type similar to the Youloop design with a coax crossover at mid point
  • Single turn screened loop with split in screen at centre point
  • Single turn loop just using inner of coax with screen removed

The loops were energised with a 1m diameter square wire loop mounted "face to face" with the loop under test.

All loops were mounted in the same position in the near field and were predominantly flux coupled.

Varying the spacing between the loops from 0.3 to 2m produced similar results.

There are some lumps and bumps in the response curves as I measured the loops indoors, and although I attempted to minimise interaction with other nearby objects as much as possible, some is still present, although it was noticeable that the single turn screened loop with split in screen at centre point gave the most consistent results.

Note that these results depend upon the near field measurements being an accurate representation of what is also occurring in the far field, and this may not be the case.

The test set-up is shown below.


The amplitude response of the whole chain is pretty flat (+/-0.5dB) from 1MHz to 40MHz

The transformers have less than 0.5dB insertion loss (typically 0.3dB) from 1MHz to 40MHz and a return loss of 20dB over the range 1MHz to 36MHz.

Ignore the absolute dBm values on the vertical scale, I have not used a set calibration point and the dBm values changed with loop spacing.

All plots on each graph were taken with the same settings and as close as possible to the same physical loop position and spacing.

The first plot shows the measured levels of the three loops. The single turn loop with split screen gave the best average level across the frequency range, and the single wire loop was close behind. The Moebius type was typically about 3dB lower in level.

The second plot shows the Moebius loop relative level when the 1:1 transformer is swapped for a 8:1 transformer presenting a 400 Ohm load to the loop.

The next plot shows the relative levels of the loops terminated in 50 Ohms and additionally the Moebius type when terminated in 400Ohms.

Overall it looks like the Moebius type would possibly benefit from a different transformer impedance ratio, although it has been pointed out to me that using 75Ohm coax for the loop may cause a different impedance transformation ratio and various additional impedance mismatch 'ripples' to occur.

The next step was to build some loops from 50 Ohm coax, and make some signal level measurements using off-air signals as the source, in order to try and validate the near field measurements.

If you wish to experiment using transformers with different turns ratios, here is a chart showing measured secondary impedance vs. turns on an BN73-202 binocular core.

The solid lines show the predicted values and the crosses show the measured values.

Following on from the previous tests I decided to build and measure some more loops.

This time I thought I would try them indoors in order to investigate the differences in gain and S/N ratios.

I used the AirSpy HF+ receiver 48ksps, Pre-amp on & AGC off.

Antennas fed via 2m of RG 223 double screened 50 Ohm coax.

I used a selection of offf-air signal sources and tried to ensure that the received Signal to Noise ratio was >30dB whenever possible.

All measurements were made on each frequency within a few seconds of each other and peak values of signal land noise floor were measured over a duration of 2 minutes in order to allow for changes in propagation.

Note:- In the following section I have stated 2m diameter loops - this is a mistake it was 2m circumference - I'll correct this later.

The Loops I compared were

Home built YouLoop using 2 x 1m lengths of 1/4 inch 50R Andrews Heliax semi rigid coax connected to Coil-Craft WBT1-6 center tapped 1:1 transformer switchable between 50R or 200R loop termination

2m diameter 1 turn loop using outer of 1/4 inch 50R Andrews Heliax semi rigid coax connected to Coil-Craft WBT1-6 center tapped 1:1 transformer

Existing home built tuned loop 1.4m diameter using outer of 1/4 inch 50R Andrews Heliax semi rigid coax

The loops were alternately mounted in the same position of my first floor workshop surrounded by various items of powered electronic equipment which contributed to the overall noise floor.

Using the 2m diameter 1 turn loop using outer of 1/4 inch 50R Andrews Heliax semi rigid coax connected to Coil-Craft WBT1-6 center tapped 1:1 transformer as the reference level.

The small tuned loop provided the highest overall signal level, typically 10dB better than the reference, although the Q and gain drops gain  off rapidly below 2MHz, as a very large value of capacitance is required to bring it to resonance.

The YouLoop terminated with 50 Ohms provided approximately 6dB higher signal level on frequencies below 1.5MHz and 6dB lower signal level on frequencies above 5MHz

The YouLoop terminated with 200 Ohms provided signal levels that were similar to the reference 1 turn loop and peaking to have about 4dB more gain at 9MHz and 8dB less gain at 18MHz.

In all cases the S/N ratios were very similar as the ambient noise floor was high enough to be detected on all frequencies with all antennas, but this is not surprising as the loops were mounted adjacent to a LCD computer monitor.

Overall the You Loop seems to follow the same trend as demonstrated by the previous near field measurements.

My feeling remains that the multi-turn Moebius loop form of construction is an added complication, that seems to require a lot of additional effort in order to achieve a level of performance that is comparable with that of a single loop.

In all cases these small low gain broadband loops are really only suitable for use with low noise figure receivers for example the AirSpy HF+ and Discovery. 

Otherwise it may also be necessary to add a broadband pre-amplifier in order to bring the weaker signals up to a more usable level.