Baluns in antenna systems

Understanding baluns in antenna systems is different to the traditional ‘workbench’ explanation using simple lumped circuit analysis.

This article explores some simple antenna system configurations and the effect of key system components, connections and dimensions on feed line common mode current.

Background

Baluns

Let us define a balun to be any device intended to facilitate or permit a different state of balance on either side of itself. In this article, it is not limited to something that has perfect balance on one side and a grounded leg on the other side, as it is rare that these conditions occur in practical antenna systems.

Differential and Common Mode currents in a two wire line

A pair of conductors in proximity to some other conductors or conducting surface (such as the natural ground) can operate in two modes simultaneously, differential mode and common mode.

Differential mode is where energy is transferred due to fields between the two conductors forming the pair, and common mode is where energy is transferred due to fields between the two conductors forming the pair together and another conductor or conducting surface.

The currents flowing in the two conductors at any point can be decomposed into the differential and common mode components.

Differential current Id is the component that is equal but opposite in direction, it is half the difference in the two complex line currents I1 and I2.

Common mode current Ic is the component of the line currents common to both conductors, it is half the sum of I1 and I2.

• Id=(I1-I2)/2
• Ic=(I1+I2)/2
• I1=Ic+Id
• I2=Ic-Id

So, for example, if I1=2A and I2=-1A, Id=(2--1)/2=1.5A, Ic=(2-1)/2=0.5A.

The feed line and common mode current

Common mode current gives rise to external fields, common mode current on a feed line directly contributes to radiation.

For a feed line to transfer energy from transmitter to antenna without itself directly contributing significantly to radiation, it must have very low common mode current.

There exists a common mode current path with any feed line, whether it is open wire line or coax. At any point along the feed line:

• in a coaxial line operating in TEM mode and with fully developed skin effect, the current that flows on the outside surface of the outside conductor is the common mode current; and
• in a two wire line, the sum of the common common component of the currents in each conductor is the total common mode current.

The common mode current is usually a standing wave, ie its amplitude and phase vary along the line.

The common mode current path in many cases can be represented as a single conductor. This is true even across a transition from open wire line to coax, just the effective diameter of the equivalent single conductor may change.

A series of NEC models will be offered to demonstrate some relevant effects.

Voltage Baluns vs Current Baluns

The terms Voltage Baluns vs Current Baluns differentiate two families of baluns.

• Voltage baluns tend to drive equal but opposite voltages wrt a common terminal on the two output terminals. Common mode impedance is the common mode voltage difference across the device divided by the common mode current through the device, and since the objective is a fixed common mode voltage irrespective of common mode current, the target common mode impedance is very low.

• Current baluns tend to drive equal but opposite currents at the two output terminals. Common mode impedance is the common mode voltage difference across the device divided by the common mode current through the device, and since the objective is very low common mode current irrespective of common mode voltage, the target common mode impedance is very high.

Effective voltage baluns do assure equal but opposite currents in only one situation, that of a symmetric load. Some types of antenna offer fairly symmetric loads, eg a well implemented 144MHz Yagi, whereas a HF dipole is less so due to influences of the built environment, vegetation, soil homogeneity, etc. Some antennas are intentionally asymmetric, eg Off Centre Fed Dipoles. So, whilst a voltage balun is properly quite popular on a 144MHz Yagi (eg the half wave coax balun), they are less suited to less symmetric antennas.

If there is a general purpose balun, it is a current balun as their objective is equal but opposite currents irrespective of load symmetry.

Models

Simple models

An antenna or antenna system cannot be represented well by a single impedance as is so often implied by discussions about antennas. An antenna system might present an impedance between two input terminals, but antenna systems are typically not independent of ground and therefore there will be some impedance between each input terminal and ground.

                Is    ___________         Z1    I1
               -->   |           |        ___  -->
           +  -----a-|           |-c-----|___|------.
          Vs         |  BALUN    |        ___       |
           -  --.--b-|           |-d-----|___|------.
                |    |___________|             <--  |
                |                         Z2    I2  |
               GND                                 GND

In discussing baluns, (Lewallyn, 1995) gives a simplified model of a balun's environment as in Fig 0. This two component model tries to address the deficiency in the single impedance model, but it ignores the impedance that exists between terminals independently of ground.

For example, some have interpreted Z1 and Z2 in Lewallyn's model to be each around 25Ω, and based on Lewallyn's statement that to achieve good current balance requires only that Zw be much greater than Z2 that common mode impedance Zw needs only to be greater than 10x25=250Ω, the makings of a simple Rule of Thumb (RoT). They are wrong, but these numbers seem embedded in the folk lore of ham radio, they are indeed ROT.

Notions developed around the one and two component simplified networks are likely to lead to wrong conclusions because those models are incomplete.

(Schmidt, Sec 2) offers a more complete model, a three terminal Y equivalent for a two wire transmission line with differential and common mode load. It should be remembered that this model is specific to a certain point, and that it is likely to be different closer or further from the load.

The choking effect of a current balun is independent of the differential Zo of the line used to make the balun, or adjacent to it on either side. Nevertheless,  Rules of Thumb that relate minimum common mode impedance (Zcm) to differential Zo (Zcm>>Zo) are often cited, more ROT.

(Witt 2003) proposes another metric, IMB, which is essentially total common mode current divided by differential current (I2Ic/Id|) for a 11 different two component load resistances, each with each of three points grounded, so 33 combinations in all at each frequency of interest. In WItt's tests, IMB can never be greater than 100%, but in real world antenna system, IMB is not limited to 100%. Witt's result dataset is not directly usable in a model of an antenna system to predict common mode current. If the system is linear, the dataset can be used to calculate Zcm (which of course is frequency dependent) which can be used in an NEC model of an antenna system. Some have suggested that IBM of -20dB or lower (ie <10%) is acceptable without justification, and ignore the fact that IMB is likely to vary along a transmission line.

Others make recommendations for minimum common mode impedance without real explanation, for example (Maxwell, 2001, 21-7) recommends a common mode impedance magnitude of around 750Ω as adequate.

There are other simple models that are so flawed as to not warrant discussion here.

The balun examples used in this article use a moderately high common mode impedance of 1000+j1000Ω. This is readily achievable mid band with a toroidal ferrite cored current balun, but difficult to achieve over an extremely wide band. This impedance is chosen to be sufficient to clearly demonstrate effects, but at the same time be practically achievable.

NEC models

This article is based on NEC models of a simple antenna system, the nominal radiator and the feed line common mode current path. This approach captures the electromagnetic coupling between all of the conductors in a flexible way that allows exploration of configuration changes to demonstrate their effect.

The following models are chosen to draw out particular issues that demonstrate the 'balun problem'

Model 1

The key characteristics of Model 1 are:

• half wave dipole for the 40m band;
• 10m height (λ/4) above average ground;
• feed point is 5% off centre to expose the model to some practical outcomes of less than perfect symmetry;
• feed line is coax, and it runs from a physically small transmitter located at ground height directly up to the feed point;
• lower end of the feed line and transmitter are isolated from ground; and
  
• no discrete balun or common mode choke.
  

Fig 1:

Fig 1 shows the magnitude of the conductor currents which contribute directly to radiated power. The common mode current path is the outside surface of the outer conductor of the coax, and is the vertical conductor in Fig 1.

Note that the common mode current flowing on the coax is almost as great as that flowing on the right hand dipole leg, and much greater than that flowing on the left hand dipole leg. In this case, the coax contributes more to radiation than the left hand half of the dipole.

The effect of the common mode current in this case is to:

1. distort the pattern from that of the dipole alone, possibly reducing antenna gain;
2. result in higher field intensity immediately adjacent to the feed line, and increased risk of Electro Magnetic Compatibility issues (ie interference to other electronic equipment in the premises and in neighboring properties);
3. couple more noise and other unwanted emissions from power wiring and equipment in close proximity to the feed line to the receiver; and
4. to cause very high charge at the transmitter (ie very high voltage wrt ground).

For some applications, (1) pattern distortion and loss of antenna gain are critically important concerns, and for others, they are relatively unimportant.

EMC (2) is potentially important in most applications, especially where interference escalates into neighbor disputes.

Increased receive noise (3) is relevant in any noise limited applications. For many, it might not make much difference working the local repeater, but for any challenging contacts, noise is important.

Excessive charge at the transmitter (4) creates health issues (wrt maximum permitted exposure to electromagnetic fields) EMC issues in shack (computer interference, tx and power supply malfunction etc) and personal injury or discomfort from RF burns from microphone etc.

This is a really poor configuration, it suffers from all the likely adverse outcomes of common mode current. This type of configuration is often proposed to new hams on online fora, and measures such as effective station grounding and effective baluns as unnecessary and ill advised.

Model 2

The key characteristics of Model 2 are:

• half wave dipole for the 40m band;
• 10m height (λ/4) above average ground;
• feed point is 5% off centre to expose the model to some practical outcomes of less than perfect symmetry;
• feed line is coax, and it runs from a physically small transmitter located at ground height directly up to the feed point;
• lower end of the feed line and transmitter are effectively grounded; and
  
• no discrete balun or common mode choke.
  

Fig 2:

Fig 2 shows the magnitude of the conductor currents which contribute directly to radiated power. The common mode current path is the outside surface of the outer conductor of the coax, and is the vertical conductor in Fig 2.

Note that the common mode current flowing on the coax is very low. In this case, the coax contributes very little to radiation compared to the dipole legs.

The reason for the reduction in common mode current from Model 1 is:

• the feed line is a quarter wave in length; and
• the feed line is grounded at the lower end.

Both of these factors means that the upper end of the feed line common mode conductor offers a high impedance to common mode current, in fact the grounded quarter wave feed line acts as a balun.

Model 3

The key characteristics of Model 3 are:

• half wave dipole for the 40m band;
• 20m height (λ/2) above average ground;
• feed point is 5% off centre to expose the model to some practical outcomes of less than perfect symmetry;
• feed line is coax, and it runs from a physically small transmitter located at ground height directly up to the feed point;
• lower end of the feed line and transmitter are effectively grounded; and
  
• no discrete balun or common mode choke.
  

Fig 3:

Fig 3 shows the magnitude of the conductor currents which contribute directly to radiated power. The common mode current path is the outside surface of the outer conductor of the coax, and is the vertical conductor in Fig 3.

Note that the common mode current flowing on the coax is less than in Fig 1, but more than in Fig 2. In this case, the coax contributes a moderate amount to radiation compared to the dipole legs.

The reason for the increase in common mode current from Model 2 is:

• the grounded feed line is increased to a half wave in length.

This means that the upper end of the feed line common mode conductor offers a low impedance to common mode current.

Model 4

The key characteristics of Model 4 are:

• half wave dipole for the 40m band;
• 20m height (λ/2) above average ground;
• feed point is 5% off centre to expose the model to some practical outcomes of less than perfect symmetry;
• feed line is coax, and it runs from a physically small transmitter located at ground height directly up to the feed point;
• lower end of the feed line and transmitter are effectively grounded; and
  
• a discrete common mode choke of impedance 1000+j1000Ω at the feed point.
  

Fig 4:

Fig 4 shows the magnitude of the conductor currents which contribute directly to radiated power. The common mode current path is the outside surface of the outer conductor of the coax, and is the vertical conductor in Fig 4.

Note that the common mode current flowing on the coax is less than in Fig 3. In this case, the coax contributes less to radiation than the feed line in Fig 3.

The reason for the decrease in common mode current from Model 3 is:

• the discrete common mode choke (balun) added at the feed point.

This means that the upper end of the feed line common mode conductor offers a higher impedance to common mode current.

Model 5

The key characteristics of Model 5 are:

• half wave dipole for the 40m band;
• 10m height (λ/4) above average ground;
• feed point is 5% off centre to expose the model to some practical outcomes of less than perfect symmetry;
• feed line is coax, and it runs from a physically small transmitter located at ground height directly up to the feed point;
• lower end of the feed line and transmitter are isolated from ground; and
  
• a discrete common mode choke of impedance 1000+j1000Ω at the feed point.

Fig 5:

Fig 5 shows the magnitude of the conductor currents which contribute directly to radiated power. The common mode current path is the outside surface of the outer conductor of the coax, and is the vertical conductor in Fig 5.

Note that the common mode current flowing on the coax is much less than in Fig 1. In this case, the coax contributes very little to radiation compared to the dipole legs.

The reason for the decrease in common mode current from Model 1 is:

• the discrete common mode choke (balun) added at the feed point.

This means that the upper end of the feed line common mode conductor offers a higher impedance to common mode current.

Model 6

The key characteristics of Model 6 are:

• half wave dipole for the 40m band;
• 20m height (λ/2) above average ground;
• feed point is 5% off centre to expose the model to some practical outcomes of less than perfect symmetry;
• feed line is coax, and it runs from a physically small transmitter located at ground height directly up to the feed point;
• lower end of the feed line and transmitter are effectively grounded; and
  
• a discrete common mode choke of impedance 1000+j1000Ω at one third the height of the feed line.

Fig 6:

Fig 6 shows the magnitude of the conductor currents which contribute directly to radiated power. The common mode current path is the outside surface of the outer conductor of the coax, and is the vertical conductor in Fig 6.

Note that the common mode current flowing on the coax is more than in Fig 4. In this case, the coax contributes more to radiation than that in Fig 4.

The reason for the increase in common mode current from Model 4 is:

• the discrete common mode choke (balun) has been moved down the feed line to about one third height.

Note that there is common mode current on both sides of the balun, and the common mode current on each side is approximately equal. The common mode choke has changed the common mode current distribution, but it does not choke it off, nor contain common mode current to only one side of the choke.

Model 7

The key characteristics of Model 7 are:

• half wave dipole for the 40m band;
• 10m height (λ/4) above average ground;
• feed point is 5% off centre to expose the model to some practical outcomes of less than perfect symmetry;
• feed line is coax, and it runs from a physically small transmitter located at ground height directly up to the feed point;
• lower end of the feed line and transmitter are isolated from ground; and
• a discrete voltage balun at the feed point.

Fig 7:

Fig 7 shows the magnitude of the conductor currents which contribute directly to radiated power. The common mode current path is the outside surface of the outer conductor of the coax, and is the vertical conductor in Fig 7.

Note that the common mode current flowing on the coax is a little more than that in Fig 1. In this case, the coax contributes a little more to radiation than that in Fig 1.

The reason for the increase in common mode current from Model 1 is:

• the discrete voltage balun at the feed point is not just ineffective, it has actually promoted common mode current.

The real world

Real configurations

These examples have been chosen to show effects of configurations and measures. They do not represent any specific practical real world system, but provide an insight into the way that antenna systems behave.

In a real world implementation, the feed line probably follows a more complicated path from feed point to transmitter, and then there may be a path for common mode current from transmitter to ground.

The real world scenario is more complex in having many equipments, connections for power, control signals, telecommunications, signals etc.

Anecdotal evidence

Models must always be reconciled with the real world, model calibration is part of the modelling process. So, do the models above calibrate with the real world?

Adding an effective balun may alter the gain pattern of an antenna. That may alter the level of received signals and noise for better or worse,  depending on the directions from which energy comes. A single experience that a balun did or did not make a significant difference does not prove that baluns work or don't work.

Likewise, troublesome RF interference in equipment in the shack, in the house or in a neighbor's house may or may not be sufficiently improved by a balun to fall below the problem threshold. Again, no single experience is proof of effectiveness or otherwise in the general case.

Much of the anecdotal evidence that balun's don't make a difference is based on experience with voltage baluns. Voltage baluns are known to be ineffective in reducing common mode current in many applications, Model 7 above is just one example. Nevertheless they are widely recommended to newcomers.

To give value to anecdotal evidence, one must understand the context and try to understand the reasons for the outcome to a greater depth than that the device 'worked' or didn't. Often, anecdotal evidence is more a qualification of the supplier than credible scientific evidence.

Application factors

The application influences which design criteria are more important.

For baluns used in a highly symmetric environment and where the feed line is bound to or even bonded to a neutral metal mass (eg VHF Yagis), a balun with low insertion VSWR and choking impedance of as little as 500Ω might reduce common mode sufficiently.

For a multi-band dipole with open wire feeder for HF, insertion VSWR is relatively unimportant because it will be used with an ATU, but high common mode impedance (perhaps greater than 1000Ω) and high voltage withstand are priorities. Highly asymmetric antennas such as an OCF dipole may warrant very high common mode impedance (perhaps greater than 2000Ω).

A vertical monopole with elevated radials or with ground level radials may benefit from a current balun of medium to high common mode impedance to minimise the common mode current on the feed line. It is naive to think of the base of a ground mounted vertical as grounded and the transmitter as grounded and therefore there is no prospect of common mode current. RF grounds are not that clear cut.

Buying a balun

If you intend buying a balun, think about:

• the purpose of the balun;
• common mode impedance objective;
  
• the frequency range;
• importance or not of low insertion VSWR and the nominal differential impedance;
• voltage withstand;
  
• power dissipation;
• weatherproofing, connectors, packaging etc; and
  
• cost.

Prioritise these criteria for your own application.

Conclusions

We can draw some conclusions:

• Antenna systems have potential for undesirable common mode current on the feed line.
• The common mode current distribution is usually a standing wave and is constrained by the same boundary conditions as apply to the nominal radiator. At an open circuit, current is zero (voltage may be high), at a short circuit, voltage is zero (current may be high).
• Common mode current can alter antenna gain pattern, exacerbate Electro Magnetic Compatibility problems, increase pickup of local noise.
  
• One and two impedance equivalent circuits of antenna systems are inadequate models for analysing the behavior of baluns, and the Rules of Thumb derived from them are ROT.
• Common mode current can raise the potential of the transmitter chassis.
  
• A transmitter is unlikely to be at true ground potential, but more importantly, there are ways to minimise the potential difference between the transmitter and objects nearby.
• A station where no attempt is made to ground the station, or equipotential bond the station ground with nearby objects is likely to impress uncontrolled and possibly disruptive RF voltages across equipment interfaces.
• Baluns are one measure that may help in reducing common mode current.
• A balun with high common mode impedance is more likely to reduce common mode current.
• Voltage baluns inherently have very low common mode impedance.

Each of the model provides evidence for some specific conclusions, and they are listed in Table 1.

Links / References

• Lewallyn, Roy. Nov 1995. The 1:1 Current Balun: http://eznec.com/misc/ibalun.txt
• Maxwell, Walt. 2001. Reflections II: Worldradio Books
• Schmidt, Kevin. nd. Putting a Balun and a Tuner Together: http://fermi.la.asu.edu/w9cf/articles/balun/index.html
• Witt, Frank. Sep 2003. Evaluation of Antenna Tuners and Baluns–An Update In QEX Sep 2003. Newington: ARRL.
• Baluns on VK1OD.net
  

Changes

© Copyright: Owen Duffy 1995, 2021. All rights reserved. Disclaimer.