(published in Electron #4, 2007)
Baluns have been around from the beginning of wired and wireless communication. Baluns in radio antenna systems in particular have been discussed from the early radio-days and are may still being found in current radio magazines. Although the balun is a simple and straight forward circuit component, the design and application has often been surrounded by some sort of 'magic', with magistrates riding hobby-horses. Despite the long balun history, or may be just because of this, the subject of baluns is still popping-up regularly 'on the waves' and therefore it seems a good idea to put various aspects around these components a little bit into perspective in a short series of articles.
Let's first take a short look into some balun applications in radio-communication.
The word 'balun' is an assembly of balance to unbalance transformer, a transformer to connect a balanced system to an unbalanced system without influencing parameters in either system. Balanced to unbalanced transitions are part of our everyday radio-life as in balanced microphones for more dynamic range, at inputs and outputs of balanced amplifiers either base band audio or at HF, feeding balanced antenna systems etc. Baluns therefore may be discovered at various places and it is clear the same functionality to look somewhat different at different parts of the systems.
- Around a balanced microphone a low-power balancing voltage-transformer may be an optimal component.
- At the input of a balanced HF FET amplifier, both branches should be fed with a symmetrical voltage.
- At the output of the same amplifier currents should be balanced for symmetrical loading and canceling of even harmonics.
- At a symmetrical antenna, both halves should be fed with an equal and opposite phase current to obtains a symmetrical radiation pattern.
These are only a few of many situations requiring a balun at the transition point. It is clear each situation in the above examples to require a somewhat different approach and before we start thinking about baluns, it is obvious we should first recognize the particular 'problem' we are trying to solve and the voltages, currents and impedances around the application.
In almost all modern commercial transceivers we find a balanced, semi-conductor amplifier at the transmitter output stage, that is coupled to an a-symmetrical output terminal by means of a wide-band balun. Various options are open to connect this a-symmetric output to the antenna, of which one is to apply an a-symmetrical feed-line to a balanced dipole antenna, again through a wide band balun. This we recognize as a double 'transformation' of which the last part has to be constructed by the owner of the system.
This double transformation did become popular after the last world war, when much ex-army material became available to destitute European radio-hams. Part of this army dump was consisting of loads of coaxial transmission- lines since this had shown to be a reliable means for quickly rigging up transportable transmitter stations in the field. Before this period almost all transmitter stations where connected to the antenna through symmetrical feed-lines, that not only exhibited very low loss but were also easy to mach to the high(er) impedance of the tube transmitters at that time.
Pro's and con's of various antenna feeding methods have been discussed more extensively in the chapter "Where does the power go", and appear to be closely related to the impedance (range) of the antenna system.
In this article-series baluns will be discussed mainly for connecting a-symmetrical feed-line to balanced antenna's, to ensure equal but opposite phase currents in each dipole half and no parasitic currents outside this circuit.
When designing symmetrical antenna's in an antenna design environment (e.g. Mmana or EZNEC), we ensure the design to be perfectly symmetrical and find radiation patterns to match this symmetry, so why should the practical antenna currents deviate from this perfect symmetry?
In practice dipole antenna's rarely are perfectly symmetrical because of all sorts of obstacles around or near this antenna, (a-symmetric) vicinity of trees, wires to couple to metal parts (roofing, drainage), different soil-types or soil humidity etc.
A second important factor is the feed-line itself. When an a-symmetric feed-line is connected to a symmetrical antenna, both antenna halves 'see' the impedance of this feed-line, with one halve to also 'see' the outside of the a-symmetric feed-line. If this additional wire exhibits a high impedance as compared to the antenna at the feed point, relatively little current is 'leaking' to this additional antenna. This parasitic antenna however may also exhibits a low impedance depending on the length of the feed-line. In this latter situation the out-side feed-line may become the main antenna with a completely different radiating pattern than designed for. This will not only be noticed at your distant communication party since the parasitic antenna is 'stealing away' communication power, but also at your neighbors when they unwillingly are listening to your radio-contacts through the stereo equipment. It may even be apparent at your position at the controls when HF feedback creeps into your equipment or make your transceiver 'hot' to the tough.
When receiving reciprocal problems arise with the unwanted additional antenna to tap into the 'electro-smog' around your house and your neighborhood, since this type of noise usually is vertically polarized.
All in all it may be clear that balanced (antenna) systems should remain balanced and should be isolated from unbalanced feed-lines by means a proper balancing device. It is also clear this device should be of the current balancing type, also known as a 1 : 1 current transformer or sleeve choke.
In a current transformer, conductors are tightly coupled; a current in one conductor will have a current flow in the other of equal magnitude but opposite phase, a schematic example may be found in figure 1.
This 1 : 1 current transformer consists of windings of equal turns that are tightly coupled. Current 'i1' is entering the transformer at terminal A and is leaving at terminal C. The current flowing in winding A-C, will induce a current in winding D-B of equal magnitude, that is leaving as 'i2' at terminal C.
This current will introduce a voltage across load resistor Rb of magnitude u2 = i1 ( = -i2) x Rb. Since this transformer has equal number of turns for both windings, and therefore equal impedance, an equal voltage will appear across terminals A-B, making u1 = u2. This current transformer also forces currents i1 and i2 to be equal and equal to the current through the generator, provided no other currents than i1 and i2 can flow anywhere in the circuit.
This description is similar to what is happening inside a transmission-line. Again the conductors making up for this line will be carrying currents of equal magnitude and opposite phase because all of the electromagnetic filed is forced to stay 'inside' the transmission-line. Let's look at figure 2 of the transmission-line as a 1 : 1 current transformer.
transfered power with no reflection of energy back to the generator. This also requires a voltage at the output of the line to be equal to the voltage at the input (for an ideal transmission-line). Again transmission-line currents are equal and of opposite phase and equal to the current through the generator. As in a current transformer this is true as long as no other currents will flow.
Since the output of the transmission-line is 'free floating' this may be connected to any other position in the circuit. If, for some reason the bottom part of the load resistor Rb is at a voltage U with respect to ground (e.g. when connected to the top of the generator), an additional current will flow through the outside of the transmission-line, that is no longer coupled to the currents at the inside. Since this current also has to be delivered by the generator, this leakage current is lost from the load. In terms of a transceiver circuit, power as delivered to the antenna system by the transceiver (generator) will no longer be only absorbed by load Rb, the radiation resistance, but also by a (dummy?) load Zsleeve. In the receiving position energy as received by the antenna will no longer be delivered to the receiver only.
The amount of additional current will depend on the impedance of the outside of the transmission-line, depicted in figure 2 as the sleeve impedance Zsleeve. At very long transmission-lines, this current will be very small, but so will the currents through the transmission-line by means of internal loss. At short lines we may enhance the ratio of transmission-line current to sleeve current by enhancing the sleeve impedance. This will make no difference to the transmission-line currents, since these are 'locked-up' inside the line and are running as a differential-mode current. So either in a straight line, curled up or made into an inductor makes no difference to the transmission-line current, because no differential current will be generated as the outside current is common-mode only. This common-mode current may be further diminished by winding around a high permeability ferrite core (the dashed line above the transmission-line). With this construction we created a common-mode choke to block off any outside current, again approaching the ideal current transformer of figure 1.
Let's look at an antenna system with the above transmission-line current transformer as a balun to connect a symmetrical dipole antenna to an a-symmetrical feed-line, For this model the antenna radiation resistance has been split-up in two over each dipole half. To close the loop, the ground return current has been depicted by the ground symbol at each end of the antenna and the ground symbol at the transceiver. This model may be found in figure 3.
The generator of figure 2 is the transceiver in figure 3 with the load depicted as the two halve radiation resistances. Current delivered to the antenna will create voltages across each half of the dipole, creating a voltage gradient at the outside of the feed-line from the transceiver to the antenna. This voltage will create an additional feed-line current that will be 'subtracted' from the antenna current and will therefore not contribute to the dipole radiation. This parasitic current may be minimized if the sleeve impedance is maximized, presented in figure 3 as the impedance Zsleeve.
The effectiveness of the sleeve impedance in minimizing this parasitic current depend on the ratio of Zsleeve to Rant; the higher Rant, the higher Zsleeve should be. This is especially important when the (dipole) antenna is driven outside resonance and Rant is relatively high. To still be effective at minimizing parasitic currents, Zsleeve should be higher still, a lesson that is often forgotten when applying a balun to an antenna.
When testing balun effectiveness it should be checked at the highest operational load impedance and in a maximum unbalance situation, with the Zsleeve to carry the same voltage as the load. At these test conditions insertion-loss and SWR should still satisfy the specifications to qualify. In all test and measurement conditions in the following chapters, baluns will always be tested under these maximum unbalanced conditions to ensure the antenna system to operate in all practical situations.
So sleeve impedance should be high, but how high is high enough? A rule of practice is the parasitic current to be lower than 1/4 of the load current to be insignificant enough. Since power is scaling with current squared, the amount of power lost under this parasitic current regime will result in an insignificant deviation of the S-meter at the receiving side as compared to an ideal no-loss condition.
To reach this current condition, the impedance of Zsleeve in practice should be in the 150 - 300 Ohm range when the balun is applied with a resonating dipole antenna. This impedance value is not hard to obtain with various types of coils (turns of transmission-line), either with or without a ferrite core.
Analyze your antenna system
Baluns will be applied at various antenna systems and frequency range . At designing or acquiring these components however, one should be very much aware of the impedance range the balun is specified for related to the range of impedances that will occur in your specific situation. Many such baluns on the market are designed for 50 Ohm system impedance and will operate according to specifications (power, frequency range) only at that impedance. Under practical conditions this specified impedance range may easily be violated, especially when an antenna is driven outside resonance with voltages to easily become much higher than the component has been designed for.
As an example: A balun that is specified for 250 W. in a 50 Ohm environment will already by overstressed by a factor of two when driven at 100 W. in a 500 Ohm environment, which is still a low value for a dipole out of resonance. Since an antenna tuner usually will translate this 500 Ohm impedance nicely to 50 Ohm again, this condition may easily go un-noticed. Burned baluns and / or damaged transceivers may be the result.
A good analysis of the tuner / balun situation at various impedance levels has been published by Kevin Schmidt, W9CF, in a paper called: 'Putting a balun and a tuner together'. In this analysis, system impedances are discussed under various load conditions, as are requirements to the balun in a balanced feed-line situation. Also a method is being presented to calculate and measure these requirements.
The next articles in this series will discuss various types of (antenna) baluns with and without core material. I would appreciate your comments, questions and remarks to this and any of the other balun articles.
Bob J. van Donselaar