This article originally appeared in QRP Quarterly, July 2001 issue, pg 34-ff.

 

Antennas, Transmission Lines, Tuners - Myths, Mysteries and Qualifiers.

By Don Wilhelm W3FPR

E-mail: [email protected]

When Amateurs get together, there is a lot of talk about antennas and how to get the ‘soup’ from our transmitters to the antenna.  In the process of such talk, a lot of stuff creeps in that I would call myths.  It is not that such myths are untrue, but rather some things have potential for being universally accepted when in truth, they apply only to certain specific situations.  A few examples:

 - Stay away from quarter-wave transmission lines

 - Keep the SWR low

 - Always use a balun

 

There is yet another group of statements often made, that are only true in a few situations, and seem to be perpetuated as absolute truths due to mis-information with sometimes a bit of advertising hype thrown in.

 - Baluns reduce SWR

 - Coax is better than parallel feedline

 - SWR will burn up your finals

 

Here, I will attempt to convey a few concepts and principles to help you better understand what is going on in an antenna system, particularly on the transmission line and through an antenna tuner.  I am not going to provide formulas for you to work from; there is plenty of good reference material for that ranging from detailed study materials to practical information presented in literature like the ARRL Handbook and the ARRL Antenna Book.  My goal is to help organize a picture in your mind that will assist you in knowing when and where to apply the many tools and formulas that are available dealing with antennas and transmission lines.

 

A complete and thorough analysis of the full properties of radiated RF energy is quite complex, and I would ask that you keep that in mind when approaching the subject.  What we are really dealing with is a four dimensional relationship; not only do we have a directed point in space, but the time position of that point.   Since we usually deal in a three-dimensional world, we lack the natural ability to properly visualize all the variables that are present.  We can analyze antennas easily by freezing one or two of the dimensions in the problem, but we need to realize that is what we are doing.  We cannot apply an answer that was produced by freezing those variables to another situation where they vary - we intuitively know that, but the number of variables in dealing with antennas can be large, and it is easy to lose sight of some.  I believe that is why we end up with so many myths when dealing with antenna systems, and my response is “It ain’t necessarily so”.  I encourage always asking; “How is this system or situation different from the last” and also how is it similar.

 

There are several solid facts though that can be used to advantage.  The first is that the current at the end(s) of an antenna is zero and the voltage will be at its highest there.  Sort of like in a DC circuit, no current can flow from an open, but unlike a DC circuit, as soon as you move away from the end of the antenna, the current will increase, so don’t take my DC comparison too far - it only applies at the antenna ends (an  example of use caution when extending an analysis).

 

The other useful fact is that along an antenna wire, the current will increase (and the voltage will decrease) until you reach a point 1/4 wavelength from the end and decrease again, also the direction of the current will reverse every half wavelength.  Look at antenna books and you will see plots of antenna current and the direction vectors. 

 

Everything works out nicely with a center fed antenna that is a half wavelength long, or an odd multiple of that.  For simplicity, I’d like to restrict my discussion to antennas that are center fed or end fed.  Those with off center feed require more analysis than I will present here.  But what if the antenna is non-resonant as in the case of a multi-band antenna?  I assure you that the same relationships hold true, and as you follow the antenna in from its endpoint, you encounter a feedline at some point where the current is not necessarily at a maximum. The current must continue and will follow down the feedline until it reaches a point where it is a maximum and the voltage is a minimum.  For a center fed dipole, when we do this ‘current-wrapping’ from each end of the antenna, we will find that the currents on each side of the transmission line are of equal amplitude, and will be have opposing directions.  So once we get on the feedline, the condition necessary for a true feedline - equal and opposite currents exist.

 

That brings us to the real substance of this discussion, which is feedline behavior.  If we continue the current and voltage relationship moving from a minimum to a maximum and back to minimum again, we can repeat this current/voltage picture until we reach the transmitter end of the feedline.  All along the way, these currents and voltages will be equal and opposite.  Fact time - if the voltages and currents on your feedline are not equal and opposite, it is not acting as a feedline alone, it is also behaving as something else - perhaps as an antenna or a resistor.  To keep the feedlines being feedlines we can do our best to run the feedline away from the antenna at a right angle so the feedline is in what I would refer to as the antenna’s RF ground plane which has a point in the center of the antenna feedpoint and extending perpendicular to the antenna in all directions.  Put a pencil through a sheet of paper, the pencil is the antenna and the paper is all the points at RF ground potential with respect to that antenna.

 

Keeping the feedline balanced

Once we have established balance at the antenna end, we need to keep the feedline balanced.  If it is parallel feedline, we can observe these rules of thumb; support it away from conducting surfaces by a distance three times the spacing of the feedline. If we have to cross a conductor, run the feedline at right angles to that surface or straight through it, and if we must make turns, keep them gentle, like a radius of 10 times the conductor spacing.  Coax is easier because it is shielded, and can be coiled, run next to conductors, laid on the ground and other nice things.  We speak of coax as an unbalanced line, and as far as the outside of the coax shield is concerned, it is unbalanced, however inside the coax, the center conductor and the shield will have equal and opposite currents and meet that qualification for a balanced line  - so as far as the proper RF is concerned, even coax is balanced.  Due to skin effect, the outside of the coax can (and should) carry a ground potential without effecting the balanced condition going on inside the coax.

 

SWR as a friend

We know that feedlines come in different impedances, and that is the characteristic impedance of the line, and we also know that if we terminate a line with a resistor equal to its characteristic impedance, we end up with a flat line or SWR = 1:1.  Under these conditions, the current along the feedline will be equal at all points.  That is nice if we can arrange the antenna to provide a resistive feedpoint equal to the impedance of the line we want to use.  Life is not always so easy, and multiband antennas (except for trap antennas) can provide that condition for only one band at most.  SWR will be something other than 1:1 for the usual multiband antenna. 

 

When we have SWR on the feedline, it means that the voltage and current on the conductors will change as we travel down the line.  If we measure the current at all points along the feedline, we can find a point where that current is a minimum (it will not be zero), then an electrical quarter wavelength further down the line it will be at its maximum.  We will also find that where the current is maximum, the voltage will be at its minimum, and vice versa.  The ratio of the maximum current to the minimum current is the SWR ratio.   Also, at all points along the line, the currents in the two conductors will still be equal and opposite, so the line is still balanced.  In a real transmission line with SWR, the higher voltages will leak a bit more across the dielectric, and the higher current points will cause more resistive losses. With high quality feedlines, we can minimize these losses, but we can’t eliminate them.

 

We have established that the line has higher voltages and higher currents, so what about the old tales that the reflected power comes back down the line and zaps the transmitter.  It doesn’t do it that way!  What really happens is that the output stage has to deal with the increased voltage and/or increased current.  It is usually the voltage that gets to solid state devices, and if high enough, will punch a hole in their substrates and ‘let the smoke out’.  Most modern output stages are designed to withstand a certain level of additional voltage and a 3:1 SWR can be handled by most of them.  Do check your transmitter specs to be sure.

 

How much power will get to the antenna if a mismatched line is used?  All of it, except for the power turned into heat by the line losses.  Yes, in any given wavefront approaching the antenna, only a portion of that power moves out onto the antenna, and the remaining portion is reflected back into the transmission line, and it travels the ‘wrong way’.  When it reaches the transmitter, it turns around again, and travels back to the antenna, where a similar percentage is radiated - this same thing happens again and again.  Well, sounds like a losing battle here, but no, remember that we only did this for a single wavefront, and there is another one coming right behind it on the next RF cycle (not exactly correct, but good enough for understanding), and the sum of all these wavefronts getting transmitted, and partially reflected again and again is equal to the power that the transmitter is generating.  Nothing is lost by the relections per se; any additional loss is the result of the higher voltages and currents that occur.

 

OK, so we have SWR on our transmission line - how can that help us?  It can help us match our antenna impedance to the transmitter’s output impedance.  The impedance at any point along a mismatched transmission line has a specific voltage and current relationship, and that relationship can be translated into an impedance (Z=V/I).  The impedance will be a maximum at those points where the voltage is highest and the current is lowest (these conditions will occur at the same place along the line), and will be a minimum at the point where the voltage is a minimum.  By changing the length of the transmission line, we can find a feedpoint impedance that matches the requirements for our transmitter output.  In fact, Cecil Moore, W6RCA has designed a ‘No-Tuner’ multiband antenna system that works by switching in additional lengths of feedline to adjust the impedance seen at the transmission line feedpoint.  The statement that “My feedline tunes my antenna” is a true statement - and it is because of SWR that it is possible.

 

Quarter wavelength stuff

If any length of feedline with SWR can be used as an impedance transformer, then what about all the quarter wave sections that we have heard about?  There is no magic in a quarter wave transmission line, but if we are using numbers and formulas, the impedance transformation for a quarter wavelength is a lot easier to calculate than some random length.  If we have low impedance at one end, the opposite end will have higher impedance.  The transformation formulas are in the reference books, so I’ll not repeat them here.  I would point out that if the impedance at one end is purely resistive, the impedance at the other end will also be resistive, and a random length section can be resistive at one end and yet have a complex impedance at the other end.

 

As far as avoiding quarter wave multiples in your feedline length, I am uncertain why this has come to be a guideline.  In a real world installation, it may help sometimes if the feedline is located in the field of the antenna rather than at its centerline plane where the feedline picks up radiation from the antenna and acts like a counterpoise with a resulting high voltage at the open end.  I know of no other generalized reason to avoid a quarter wave feedline.  The origin of this rule of thumb may have been from the times when hams fed antennas from a link on the transmitter output tank or a direct tap, and found that some antennas with quarter wavelength feedlines transformed the feed impedance to a value that was difficult to feed with that output arrangement.  So don’t be concerned if your feedline is a quarter wave or multiple on some bands as long as you have the proper tools for dealing with the impedance at the shack end of the feedline like a good antenna tuner, and orient the line for minimum pickup from the antenna.

 

Balanced or unbalanced and Baluns

I have mentioned that coax feedline may be physically easier to work with than parallel line.  You can toss it most anywhere and it will be OK, not so with parallel line which must be properly arranged if it is to work as a feedline rather than an antenna or a resistor.  A properly installed balanced feedline can give the advantages of multiband use and relatively low loss with the variety of SWR that result from a multiband antenna.  For single band operation, coax feed is usually the easiest to deal with and I highly recommend it for that application.  If operated with SWR less than 3:1, good quality coax has only small loss.  Reducing the SWR below 3:1 may not be worth the effort if the payback will be in loss only.  Keeping the transmitter happy and providing its full output is another matter - some will provide full output with a 3:1 SWR while others will not.

 

With coax feed on a dipole, due to the surface effect of RF current, the equal and opposite currents in coax are confined entirely to the inside of the cable, and RF will not flow on the outside - at least not from the source (your transmitter).  Now, look at the junction of the antenna wire and the coax shield.  When the RF current gets to that point, it has two conductors to look at, 1) the antenna wire, and 2) the outside of the coax shield.  The electrons dutifully split here with half of them going out along the antenna wire and the other half traveling along the coax shield.  A balun will stop this action, and keep all the RF on the antenna.

 

If we chose to use a balun, which one?  We as QRP folks tend to use small toroid cores for our baluns just because they are small, and “we don’t need anything big for QRP”.  I agree with that based on the capabilities of the balun core itself, but what about the wire?  If we use a small core, we also must use small wire to fit all the turns into it.  Unless we really need it compact, perhaps we should re-think that one, larger wires have smaller resistance, and therefore smaller resistive losses.  Why operate QRP and shoot ourselves in the foot by turning part of our signal into heat?

 

If the installation of a balun should make your antenna bandwidth better, or make your apparent SWR go down, then you should be looking at another balun - the ones that make the SWR or bandwidth better are doing so because they are turning precious watts into heat, and you can do better than that.  A simple resistor across your antenna terminals will increase the apparent bandwidth and lower the SWR too, but it radiates only heat, not RF, and a balun that produces the same effect is just making heat too.

 

Let’s say we fail to install a balun at the feedpoint, is it important?  I tend to agree with the late Doug DeMaw on this one - maybe and maybe not.  If we leave the balun out, we will end up with the feedline actually becoming a part of the antenna.  It will radiate, and in most installations that radiation will be vertically polarized.  What harm is there in that, the RF gets radiated and we can work stations, and sometimes the vertical radiation even helps if propagation conditions and distance work out right.  But, when we are trying to create a particular antenna pattern, like for a beam or other directional antenna, this radiation most likely will be in the wrong direction and wrong polarization to aid our desired pattern and the use of a balun to eliminate it is prudent.  In this article, I will not attempt to counter those who fear that this extra RF on the coax shield will travel back to the shack and do all kinds of damage.

 

 

Antenna Tuners

 

Antenna tuners come in two major flavors - the kind that act as a transmission line, and the kind that act as a tuned circuit.  Let’s talk about the ones that act like a transmission line first.

 

We already discussed how a transmission line itself has the ability to change its input impedance by changing the length.  Well, we can electrically represent a transmission line as a collection of inductors and capacitors too.  Not only represent it that way, we can build a physically short transmission line with lumped inductors and capacitors.  The “L” network tuners, the “T” network tuners, and the “PI” network tuners all fall into this category.  They all do the task just fine, and each one has its own limitations.  The PI network with a single inductor is also a low pass filter and can be very efficient, but at low output impedance, the output capacitor has to be very large for a match, so it may not be a good choice for the low frequency bands given the choice of available variable capacitors in today’s world.  The L network can be quite workable over a large range, and can take the form of either a high pass or a low pass design, although some compromises may have to be made due to the size of available components.  The T network with a single inductor is the one most commonly used, is a high pass design, and has become the favorite of most manufacturers.  It can cover a large range with its three-variable elements.  With its wide range we have another problem - an apparent match can be achieved with several combinations and some of them produce high losses.  With the T match circuit, use the setting with the greatest capacitance on the output side and the smallest inductor setting that will provide a match.

 

The Tuners that use a tuned circuit usually look like a parallel tuned LC tank and are tuned to the frequency of operation (any capacitive or inductive reactance at the feedline terminals may also be a part of this tuned circuit), and sometimes will use a tuned series circuit rather than a parallel tank for matching to low impedances.  The coupling to the transmitter may be obtained by a tap on the coil, or more commonly by a  ‘link’ which is another coil placed in the magnetic field of the main coil.  There may be a capacitor in series with the link to increase the range of the tuner.

 

These tuned tank circuit tuners provide a bandpass filter characteristic, and a link coupling offers some immunity from static noise especially on the lower frequencies.  The side that connects to the antenna feedline can be configured as either a balanced or an unbalanced circuit.  In The parallel LC tank configuration, the feedline can be connected to taps on the coil, a very wide range of impedances can be matched with high efficiency - attach the feedline toward the outer ends of the coil for higher impedances, or closer to the center for lower impedances.

 

Link coupled parallel tank circuit tuners are by far my favorite.  Set the output tank tuning somewhere near resonance at the frequency of interest, and start with the taps about mid-way out on the coil, and then fine tune for minimum SWR on the input side.  If it does not go down to 1:1 set the taps to the feedline at a different place and try again.  Once you have established the proper tap point for any particular band and antenna, tuning to a 1:1 SWR is fast and easy.  There are no false tuning points to be troublesome, and the balance for parallel feedlines is excellent if the components are arranged symmetrically.  The down side is that they do not lend themselves to easy bandswitching, especially in the balanced configuration, and with commercial coil stock becoming scarce and/or expensive as well as dual section variable capacitors becoming scarce, these simple but excellent tuners may become a thing of the past.

 

Losses in antenna tuners and feedlines

 

As QRP oriented folks, we should be interested in getting as much of our transmitter output as possible radiated into the air as RF energy, and minimize the energy radiated as heat through our feedlines and antenna tuners.  For portable operation, we can find an antenna that can be easily erected in the field, and feed it with a lightweight feedline.  We must accept some compromises for the sake of portability and ease of use.  Say we chose a center fed dipole and we operate on a single band - we can cut the dipole so its feedpoint impedance is close to 50 ohms and feed it directly with coax, no tuner required.  Similarly with a single band vertical, we can get it tuned to match the feedline at the home location, carry it into the field and set it up with confidence in a short time and spend our time operating instead of tuning antennas.  We don’t even need to carry an SWR meter along unless we expect something to change.  If we anticipate multiband field operation, or will be operating from strange and difficult places like a hotel room, more thought must be placed into the antenna situation, a tuner is a necessity, and we will likely have to face more compromises.   In general, use the fattest wire practical in a tuner’s inductor as well as the feedline so the resistive losses are smaller, and accept the compromises presented by the choices we are forced to make.

 

At our permanent operating station, we don’t have to make as many compromises (ignoring the limitations created by CC&Rs and other situations requiring ‘stealth’ operation).  I believe we should strive for the least loss consistent with our choices of operating convenience.  What does that mean?  It means that we use fat wires for our inductors in the tuners, and that may also mean physically larger tuners for lower loss.  It is a simple matter of lower resistance that makes a high power rated antenna tuner more efficient even for QRP operation.  Those small signals will not get lost in that big box.  Alternately, we can build our own antenna tuners using physically large inductors but use variable capacitors having lower voltage ratings for QRP and gain the advantages of the high power tuners without the high cost of high voltage variable capacitors.  We can often ‘have our cake and eat it too’ by just re-thinking things with a small bits of similar information.

 

I have only hit on a few of the highlights of transmission lines, and have made no attempt to state how well any particular antenna will radiate - I leave that to the antenna experts.  I have attempted to give you a slightly different way of thinking about how we move the power from our transmitters to our antennas.  If I have encouraged you to get out the books and apply some of the formulas and ‘rules of thumb’ in a meaningful way, I have achieved my goal.  Put up the best antenna that you can muster under the constraints that life and circumstances have handed you and squirt some RF into the air, for that’s what this stuff is all about.  Then you will have done your best, and perhaps learned something along the way that you can use to do it better the next time.