When two of the waves are superimposed on top of each other with 180 degrees time difference and equal strength, they will cancel each other out. This can be easily demonstrated on a dual trace cathode ray oscilloscope. If the negative and positive voltages are added together, the output should be theoretically zero volts or no signal.

In a practical application, the phase time of an interfering signal is adjusted so that it is 180 degrees out of phase with the wanted signal from the DX antenna. By adjusting the relative levels from the two antennas so they are equal when combined using a splitter/coupler, the unwanted signal should be either completely cancelled out or considerably reduced in strength.

Practical operation

Two antennas are required. One will be the existing DX array and the other will be a phase/interference antenna, which is orientated to give maximum pickup of the unwanted interference signal. Consider the block diagram in Figure l.

The interference antenna is orientated to give maximum pickup of the unwanted interfering signal and is connected to the phase box, which consists of balance and phase controls. The output of the phase box goes into an optional high gain pre-amplifier (preferably tunable - see circuit above), and then into a variable 0-20dB attenuator, before finally going into the coupler/splitter used in reverse. The DX antenna goes into an optional low gain pre-amplifier before finally going into the second input on the coupler. The interference free output of the coupler is fed to the DX receiver.

A pre-amplifier for the interference antenna will not be required if the interfering signal is strong. However, the author prefers to use a tunable low noise MOSFET pre-amp to vary signal levels, hence the variable attenuator is not used. It is important that the optional pre-amplifier in the DX antenna line has a noise figure of 2 dB or less and relatively low gain of typically 10dB.

If we need to phase out local signals, which come from the same location, we could use a fixed phase/interference antenna orientated towards the local transmitters. However, if we have several unwanted local and semi-local signals from different directions, then we need to use a rotatable interference antenna. For best results, the DX and interference antennas must be of a similar type with near equal gain and directional characteristics.  The more directional both antennas are, the more chance one has of obtaining a null.

Phase cancellation nulls will be greater when there are minimal metallic objects in the vicinity of the DX and reference antennas. Other antennas on the same mast, act as passive reflectors, which can cause multi-path distortion, thus degrading cancellation nulls. Ideally, no other antennas should be on the same mast used to support the DX and phase antennas.

The polarization of the interference antenna should be the same as the interference transmitter's polarization. For example, to null a local capital city TV transmitter, use the matching polarization to the local city TV transmitters for the interference antenna.  The polarization of the DX antenna should equal the polarization of the DX transmitter.

The interference antenna is aimed at the local transmitter, in order to produce a ghost free strong clear picture.

When both antennas are orientated toward each other, the antennas should be spaced horizontally no less than one wavelength apart, so as to avoid interaction, and distorted directional polar response pattern problems. Depending on the frequency, that will be no less than 17ft (45-70MHz), 10ft (88-108MHz),and 8ft (175-225MHz). As a general rule, keep the spacing to no less than 17 feet. At the author's location, the interference and DX antennas are mounted on opposite sides of the house.

Practical considerations

Try to keep the antennas away from nearby metallic objects, for example, tin roofs and gutters. If multi-path is present, for example, picture with multiple images or distorted audio, cancellation nulls will be degraded. For this reason, never mount the interference antenna in the roof attic. Some DXers have tried using an indoor interference antenna in the DX shack. However, for best results, the interference antenna should be mounted outside.

Always use good quality low loss shielded coax cable for both antennas. Never use 300-ohm ribbon twin lead.

The splitter used should be a high quality shielded die-cast metallic unit with high isolation of at least 20dB between the inputs. A wide-band ferrite coupler or splitter, for example, (Antiference CS1000 or Tratec ES-02), could be used to combine the two lengths of coax. High quality splitters can be obtained from dedicated antenna installation firms. Do not use standard domestic splitters as sold by Dick Smith, Tandy Electronics, etc. Also do not use a directional coupler since there will be excessive signal loss on one of the input ports. Interconnections between the pre-amplifier, phase box, splitter and attenuator can be simplified by using F-F coaxial joiners.

The author is able to fit the phase unit components and Tratec ES-02 splitter/coupler into a 100 X 60 X 45 mm aluminium box.

Circuit

Several phasing unit designs have been published over the years and most of these use inductors and capacitors for adjustment of phase time. One drawback with L/C designs is that they are difficult to construct and will only work over a narrow bandwidth. Hence, to cover 40-230MHz, at least four different phase boxes would need to be built.

The phasing unit to be described is based around a resistive network, which will enable 360 degrees phase shift over a wide range of frequencies from approximately 15MHz to UHF. I say approximately because the frequency coverage is limited by the characteristics of the F29 ferrite core used in the circuit. If coverage down to medium wave was required, a second unit containing a F14 ferrite core could be built.

The phase shift circuit (Fig.2) uses a 75/300ohm transformer/balun, two capacitors and two 5k ohm linear potentiometers.

If the phase box is used with a Tandy 0-20dB variable attenuator, use a .01 µF ceramic capacitor from the center pin of VR2 to the output coax socket. This will prevent any possible resistive interaction between VR2 and the 0-20 dB attenuator.

The connecting leads between VR1 and VR2 should be no more than one inch in length. (25mm). It is important to keep all leads as short as practicable.  The circuit consisting of T1, R1 and R2 needs to be as symmetrical as possible. VR1 and VR2 are mounted on a double-sided copper board bolted to the lid of the metal box. T2 is a 75ohm unbalanced to 300ohm balanced balun, which can either be constructed or salvaged from the back panel of old television receiving sets. 75ohm to 300ohm baluns can be also purchased at any Tandy (Radio Shack) store.

Operation

After connecting the units as detailed in diagram (Fig.l, above), adjust the attenuator (or tunable pre-amp) for maximum gain from the interference antenna. Then alternately adjust controls R1 and R2 for the deepest null. Next adjust the attenuator / tunable pre-amp until there is a further deepening of the null. Continue this procedure until the deepest interference null is obtainable. With practice, a null should be achieved in less than 30 seconds. The author uses the signal strength meter on a Icom R7000 receiver to facilitate adjustment of signal nulls. When the maximum null is obtained, the settings on the Phasing unit will be quite sharp.

If it is not practical to use two separate antenna masts, it is possible to vertically mount the interference and DX antennas on to one mast. Some DX enthusiasts have tried this with a degree of success, but it is doubtful that results will equal the two-mast method.

1. Position the DX antenna at the top of the mast in the normal way.  The lower interference antenna needs additional holes drilled in the boom, so that the boom to mast clamp sits in a position, that causes the lower antenna to mount on the mast one-quarter wavelength (90 degrees) ahead of the top antenna.
   
                         
   
   The vertical stacking distance should be as wide as practicable, or at least 4ft. One disadvantage with this is that it is impossible to aim the two antennas in different directions. In summary, common vertical staggered antennas on a common mast will enable nulls to be obtained. However, the results obtained will not equal the use of two masts spaced horizontally at least one wavelength.

2. Due to the remote possibility that both antennas are in phase with respect to antenna spacing, it may be worthwhile adding approximately 5-30 inches of coax to the interference line. If nulls are better, leave the length permanently in line, and if nulls are the same, leave the extra length out.
   

Results

The author has found null depths range from approximately 20-45dB on the interfering signal.  Null depths are influenced by various factors: Multi-path, due to aircraft, local metal objects, terrain, strength of the local interference, etc. The null may tend to 'drift' if you are either near the flight path of domestic aircraft or if the interference is very strong.  As a general rule, high powered FM signals at 50 miles and over can be reduced to white noise.  Low powered community FM transmitters at over 15 miles can usually be nulled. TV signals over 30 miles away can be reduced to either white noise or at least to the extent where moderate strength DX signals will over-ride.  The author's local TV transmitters are approximately five miles distant, which means that although considerable reduction is possible, complete cancellation is impossible, due to multi-path/ghosting. Strong Sporadic E TV signals can occasionally over-ride my local area Sydney channel 2 transmitter.   Using an ICOM R7000 receiver, it is also possible to hear the meteor shower pings of TV video carriers plus or minus 10kHz of the local video carrier. This of course would be impossible without the signals from the local transmitter being phased.

When a local TV signal video carrier has been reduced to S3 level or less, you will notice a low-pitched hum noise.  This is similar to the effect that is obtained by nulling a local medium wave radio signal, when a tilted altazimuth loop is used.  What you are hearing is residual multi-path scatter, which is audible because the main carrier has been reduced by at least 40 dB.  The author has found that null depths are greatest, when the interference antenna is aimed toward the interference.  Acceptable results can also be obtained, when both antennas are aimed in the same direction, provided one has sufficient signal level from the interference antenna.

Conclusion

The use of phase cancellation has been of great benefit with receiving DX stations over local and semi-local signals that otherwise would have been impossible. It is hoped that the above information will be of assistance to DX enthusiasts and will allow them to receive additional DX signals to their present quota.

Parts list

1) One Aluminium box (mm) 100 X 60 X 45 (Dick Smith Part No H-2305
2) Two .0luf ceramic capacitors. (Dick Smith Part No R-2321).
3) Two 5k ohm linear potentiometers. (Dick Smith Part No R-755
4) One 75 ohm to 300 ohm balun (F29 ferrite) available at Tandy or Dick Smith.
5) Two 75 ohm panel sockets. (Dick Smith Part No P-2046).
6) Two Double female coax joiners (Tandy Part No 150-9516)
7) One 75 ohm variable attenuator (Tandy Part No 150-0578).

      

Fig 6., Suggested layout for phaser box.

Specifications

    Frequency coverage:       15-860Mhz (with F29 ferrite balun)
     Phase shift:             360 degrees.
     Interference null   typically 20 - 45dB.
     DX signal loss:       ~4dB (0dB with optional preamp in DX line).

References