Some More Observations regarding the EWE Antenna – by Bill Marsh
I have recently been doing some design work on the EWE antenna to better understand its characteristics. I have been using “Modelling Software – EZNEC”, the source courtesy of Peter Mott.
Whilst the EWE can be built to any dimensions the larger the area of the resulting loop the higher the signal pickup will be. From the modelling software it has been realised that for any given length of the two “Vertical” elements there is an optimum length for the “Horizontal” element and visa versa. This optimum length for the “Horizontal” element determines the depth of the null for signals off the back of the “EWE”. It is possible to achieve nulls of 50 to 60 db. Having the best null can mean the difference between hearing a weak DX signal and not hearing it due to an offending station from the opposite direction. My own EWE is very good at hearing North American DX and interference from Australian signals is almost non-existent. I have now realised that there is still some room for improvement at home, but town-house section constraints will limit some of the improvements.
In order to fully test the “Modelling Software” I turned to my portable EWE to model the optimum horizontal length. My portable EWE is used in remote locations where there are no constraints. See my article: “A Portable EWE – Bill Marsh Build” on our website. A Portable EWE – Bill Marsh Build In this build the “Vertical Elements” were made from 4 m (13.12 ft) bamboo poles chosen to suit the vehicle I was driving at the time. I had purchased a 50 m (164 ft) roll of 32 strand appliance wire for its extreme flexibility. I firstly cut 2 x 4 m lengths plus terminating tails for the vertical sections and cut the remainder in half for the 2 x horizontal wires. This resulted in 2 x 15.8 m (52 ft) detachable wires. The end result therefore was a 4 m x 15.8 m EWE with a ground wire. The terminating resistor for this EWE was nulled at 1.12 mHz.
Using the modelling software it was found that the optimum length of the horizontal elements should have been 21m (69 ft) for the deepest null possible. Note greater than 21 metres the depth of null reduced. It is to be noted that in any EWE constructed the terminating resistor value changes with frequency. I chose 3 points on the medium wave band namely 720 kHz, 1120 kHz and 1520 kHz to measure values. In the case of the optimum EWE above the terminating resistor values are 991 ohms at 720 kHz, 1002 ohms at 1120 kHz and 1019 ohms at 1520 kHz. For the average EWE it is therefore recommended that the terminating resistor be chosen (adjusted) for mid band i.e. 1120 kHz. Some loss of depth of null will result at higher and lower frequencies as this is a compromise.
The alternative to the compromise above is to adjust the termination resistor remotely. This requires the installation of 2 identical matching transformers at each end of the EWE and a coax from each end back to the shack. The advantage of this is the null can be fine tuned for any frequency but more importantly the favoured direction can be reversed 180ᵒ. i.e. a SE EWE can be reversed to become a NW EWE. All from the shack.
The changes from a 15.8 m horizontal element to a 21 m long element provided a 2 dBi antenna gain and an extra 22 dB of notch off the back end of the EWE.
A summary of findings along the way:
- There is an optimal ratio of vertical to horizontal length, to gain the deepest null.
- The larger the area of the EWE the more signal is captured.
- The input matching transformer impedance ratio is not particularly critical. 450 ohms to 50 ohms for and on ground EWE and 940 ohms to 50 ohms if raised above ground to behave like a FLAG.
- The terminating resistor is critical for maximum null of signals from the reverse direction.
- Type 73 binocular cores should be used for matching transformers to obtain best results from LW through SW. Type 43 Toroids tail off below 1 mHz. When using binocular cores it is suggested separating the windings using plastic tubing. Research suggests this cuts down the inter-winding capacitance and supposedly noise levels. I do not have the necessary test equipment to prove this but the research is from wiser engineers than myself.
I have often wondered why the Bay of Islands SDR and Delta antenna stands out as one of the best in the world at present. Accordingly I modelled Peter’s Delta antenna for the answers. The single most significant difference is the average “Elevation Angle”. For any EWE antenna the angle of elevation at 1120 kHz is 30ᵒ whereas it is 22ᵒ for a Delta. It is for this reason that the Delta is more likely to capture more DX than a EWE antenna.
In order to verify this I have erected a Delta antenna in the same direction as my main EWE for a comparison of results. The Delta was constructed fence to fence as constrained by my town-house section which naturally determined the height of the centre pole. The modelling software showed that the gain of this Delta would not be as good as my elevated EWE due to a smaller capture area. The front to back of the 2 systems was found to be on a par. I also found that the variation in value of terminating resistor was considerably less with frequency with the Delta.
What did I find in practice. During daylight hours the weaker NZ stations were down a couple of dB on the Delta. At night however when North American DX is present the gain difference between the Delta and the EWE was barely noticeable. The most noticeable difference however was the Delta was quieter and audio clearer.
My dilemma now is which of the 2 aerials to keep. Won’t be making any choices on this for at least a couple of months of trials.
