The end-fed half-wave antenna
Transmitter 42. Alright, everyone. Today, we will cover the increasingly popular in fed half wave dipole antenna that has some in the amateur radio world all in a tizzy as they struggle to apply their understanding of antennas to this, quite new to many, feeding technique. We will start with the definition of a dipole, continue with a VHF antenna chamber measurement, discuss some modeling results, and end with an empirical current measurement experiment on a 10 megahertz antenna. Stay tuned for this and more on transmitter 42.
KX4O:Well, hello everyone. John Huggins here, KX4O. I am an antenna engineer. I work for a federal lab, and we gave a presentation not too long ago about the end-fed Half Wave Antenna to a variety of antenna experts in the field. It was well received.
KX4O:I wanted to boil it down for you in this podcast and tell you what we learned. First, we'll start with some of the proclamations that we hear in the popular press, namely the ham radio watering holes typically.
KX4O:No, an antenna can only be fed in the middle.
KX4O:All antennas are through terminal devices.
KX4O:If a second path is not provided for the available RF energy to make simultaneous round trips to and from the antenna system and the transmitter, the antenna radiates no electromagnetic energy at all.
KX4O:While the term other half might be debatable, all end-fed antennas require a ground or counterpoise of some type.
KX4O:The need for some counterpoise or RF ground to return displacement currents is clearly true.
KX4O:All end fed antennas, doesn't matter what they are, require a ground or counterpoise of some sort to push against.
KX4O:If your simulation disagrees with my beliefs, your simulation is wrong.
KX4O:The end-fed half wave is unknown outside of Sammy Hamey circles.
KX4O:Really? Well, let's dig into the details a little bit. The end-fed Half Wave is an obsession of mine. I don't like ambiguities in electromagnetics. Understanding how it works opens the design opportunities for the practitioners, such as me and my fellow workmates.
KX4O:I argue the end-fed half wave does not require a conductive counterpoise. We will go over definition of a dipole, old definitions, the IEEE definition. I will present a measurement of the VHF HT with various antennas in the far field chamber, some simulation results from NEC, and Finite Difference Time Domain method. In experiment, we'll have indirect current measurement of a 10 MHz dipole in 2 configurations, center fed and end-fed, and we'll go over a list of commercial examples that have existed for decades.
KX4O:A note about feed lines.
KX4O:Some people focus a lot of attention on feed line common mode currents. That's great. To see, however, if the in fed half wave requires these currents, some of the experiments completely omit the feed line by design. Turning now to a dipole definition. Here's one from the Internet.
KX4O:Remembering that a dipole is an antenna consisting of 2 length elements with a connecting feed at its center, then how can it be off center fed and still be a dipole? Well, we turn to the IEEE definition, IEEE 100 - 1984, and we read dipole antennas.
KX4O:"Any one of a class of antennas producing a radiation pattern approximating that of an elementary electric dipole. Common usage considers the dipole intended to be a metal radiating structure, which supports a line current distribution similar to that of a thin straight wire so energized."
KX4O:This is important, that the current has a node only at each end.
KX4O:Notice no discussion of feed point placement.
KX4O:Then we move to the more specific IEEE document. IEEE 145 dash 1993. Dipole Antenna.
KX4O:"Any one of a class of antennas producing a radiation pattern approximating that of an elementary electric dipole. Note, common usage considers the dipole intended to be a metal radiating structure, which supports a line current distribution similar to that of a thin straight wire so energized that the current has a node only at each end."
KX4O:That's nearly identical to the the other one. And from a communication standard, the EIA / TIA 329B dipole definition. Here they actually specify the half wave dipole antenna. A dipole whose electrical length is half a wavelength and is formed by a straight metallic radiator, one half wavelength long, whose diameter is small compared to its length, so energized that current has 2 nodes, 1 at each end, producing maximum radiation in the plane normal to its axis.
KX4O:Again, no specifics where it's fed. With this definition, we can clearly see that a dipole fed in its center, offset, or at the end are all dipoles by definition and analysis.
KX4O:Throughout this presentation, I'll be using the terms Hi z for impedances between 1,000 to 5,000 ohms, typically found at the high voltage points along a dipole antenna, and Lo z to be the 1 to 100 ohms, say 50 to 75 ohms typical at the high current point. The word counterpoise in my presentation will be any conductor from the feed point be it mast, feed line, any conductor other than the actual antenna radiator.
KX4O:Now, we turn our attention to a end fet half wave measurement in an antenna chamber, at VHF.
KX4O:We use an HT transceiver for VHF. It's a ham radio, Yaesu FT1D. It is an APRS rig that can be set to beacon. It's perfect for our use. We set it to beacon once every 30 seconds in APRS beacon at about a 135 milliwatts.
KX4O:And that was nice because we didn't have to attach any test cabling to it. Therefore, we have just the radio and whatever antenna is attached to it in a far field chamber that is an RF and anacoic chamber. And we just pick up the energy from a distance and measure the effective radiated power. There are 4 antennas used with this radio: a laboratory reference dipole, the stock antenna that came with the FT-1D, one of the very popular quarter wave add on WIPs that you can buy to substitute for the stock antenna, and then an MFJ end-fed Half Wave VHF antenna about a meter long that has a transformer at the bottom and a full half wave length radiator above that. The reference antenna is a laboratory dipole set to 146 Megahertz.
KX4O:It doesn't get any better than this. We had the HT attached to the feed point of this antenna, and we swept it from 0 to 180 degrees. And we measured the effective radiated power at all these angles. But what I'll be reporting here is the broadside gain as received by the laboratory gear. We calibrated the results to this reference dipole.
KX4O:We had 135 milliwatts of transmit power. That's 21.3 dBm, and of course the gain of the dipole is 2.15. That gives you about 23.4 dBm. There's a nice curve here that you can go see on hamradio.me.
KX4O:There's an article there. You can basically click on the tab EFHW, and you will find all the results for everything I'm telling you here. So 23.4 dBm was the reference. Again, power plus the gain of the dipole. So now, we have a reference that we could use to measure all the other antennas against and the first one was the stock antenna.
KX4O:The Yaesu FT1D model is comes with a 7 inch or so rubber ducky whip and, yes indeed, it has less gain than the reference dipole. Not really a surprise. It's a smaller antenna. That is always going to be the case. It was down about 8.7 dB from the reference dipole.
KX4O:Then we replaced the whip with a popular quarter wave whip antenna. Everyone assumes that that is a better radiator. Let's find out. Well, it didn't turn out so good. It's 17 dB down from a reference dipole.
KX4O:The stock antenna actually radiates better at 146 megahertz than the so called improvement quarter wave antenna. Then we put on the half wave end-fed antenna made by MFJ... it's the model 1714 with SMA. It's relatively unknown. It's one of the extendable telescopic antennas. Very fragile in my opinion, but nonetheless a perfect test sample for this test. We did tune it up a little bit on the laboratory VNA just to get it perfect at 146 and put it on the HT.
KX4O:It has a transformer that changes 50 ohms from the feed point to the thousands needed for feeding the end of a dipole. And this one comes in at 2.9 dB less than the reference dipole. Still not perfect, but pretty pretty good pretty good match. There's always gonna be a loss in that transformer. And as small as it is, I'm not terribly surprised.
KX4O:But, anyway, it's admirable performance. It's the best of the bunch. So some conclusions from this antenna chamber test. The quarter wave upgrade whip is not so good. The Yaesu stock antenna? That's actually not too bad for its size. It certainly beats the so called improved quarter wave whip. The half wave end-fed antenna, while unwieldy performs closest to the ideal dipole. So to make this test a little more interesting, we added a quarter wave "tiger tail" connected to the shield connection of the antenna to all 3 antennas and repeated the tests. The stock antenna was actually worsened by the addition of a quarter wave additional conductor.
KX4O:Which should not be too surprising, the quarter wave whip was greatly increased almost 10 dB, showcasing the idea that you've got to have a whole half wave antenna to be the most efficient. Quarter wave alone with no additional counterpoise is not going to work that well. The half wave antenna with the additional quarter wave tiger tail counterpoise showed no changes whatsoever. It was utterly immune to the effects of additional conductors. It showcases the notion that the half wave antenna is completely self contained. It requires no counterpoise , and where one exists, it doesn't have any interaction with it, at least at a quarter wave or none. So some conclusions: The Tiger Tail greatly improved the quarter wave whip turned it into a half wave antenna essentially. The a Yaesu stock antenna is slightly worse, half wave infinite antenna, little to no change. The end-fed half wave seems unaffected by counterpoise or lack thereof. Now we turn our attention to some simulation.
KX4O:For this, I used the finite difference time domain method. I used a wire model with a feed point placed at 4 different points, the 50% midpoint, 30%, 10%, and very close to the end at 2%, noting that you cannot put a source at the very end of a wire in a simulation, but you can get pretty close and certainly see the effects. So I ran 2 tests. I calculated the H field and the electric field around the dipole. The H field of course is proportional to the current through the wire, and in all cases 50%, 30%, 10%, and 2%.
KX4O:The H field around the antenna is pretty much the same with a slight difference at the 2%. But effectively, what we're looking at here is the dipole is a dipole no matter where you feed it as far as current goes. Now, if you look at the E field, you, of course, see a minimum in the middle of the antenna, and it's maximum at the tips. And in in the 50% case, that's a nice balanced looking, antenna. 30%, it's almost clearly the same, but you could start to see a voltage discontinuity creating an E field discontinuity where the feed point is.
KX4O:It's even more pronounced at 10%, and it's greatly pronounced at 2%. So when the feed point is at the end of a dipole antenna, we now know the penalty you pay, which is very high voltages that you have to deal with, as well as reactances. So the simulation suggests that it is completely possible. All the test patterns show that the energy creates the same basic pattern of a dipole, but the feeding technique will be more difficult.
KX4O:I also did some tests with NEC using a transformer that I designed basically with wires. So it's a design with a wire equivalent to the coil, as well as the half wave whip. And of course, that produces a reactive component that you have to trim out with a in series capacitance. But if you shunt tap the coil at the right spot, and add a little bit of capacitance, you wind up with 50 ohms that you can feed with a typical source. And of course, this is perfect. There is no transmission line.
KX4O:The source is literally right on the transformer. It's a complete wire model, so there's no assumed perfect transformers or ideal transformers like there is in some cases. This is literally wires representing the coil and simulating the coil's effects. Again, you can see all of this on hamradio.me.
KX4O:I use NEC 4, but you can do this in NEC 2. And sure enough, you wind up with an antenna that radiates very well into the halfwave of a wire, and you see a perfect current profile from top to bottom at a 50 ohm feed point. The coil is a 20 to 3 auto transformer. And then I attached a wire to the base of the coil and extended it out at a right angle to the radiator. The idea here was to see how much current gets pulled away from the feed and to see what lengths of counterpoise cause an effect, if any.
KX4O:We found a case where if you have a half wave of a wire in the counterpoise, this is the case where the tip of the wire is, of course, in open air, so that's a high z. The middle point of that is low z. And then where it connects to the transformer, it's high z again through transformer theory. Well, this causes problems. This starts to draw a lot of current.
KX4O:If you remember the tiger tail is only a quarter wave long, so the tip of it is at a high z but a quarter wave in it's low z. So when it presents a low z to the antenna transformer no current seems to be flowing. That seems a little counterintuitive, but if you think of this as a power impedance theorem, where like impedances transfer the most power, it begins to make some sense. So where the ground presents a like impedance to the antenna's native impedance, in this case high because it's an infinite half wave, you will draw some current.
KX4O:Where it presents a low impedance, you won't. So sure enough, if you take this wire and you keep extending it forever, you wind up with a graph that shows you that every time there's a Hi Z presented to the base of the transformer, you start to rob current from the main aerial. And this is where a counterpoise goes wrong. So key takeaways of all this certain lengths perturb the operation of the enf-fed half wave. For example, a quarter wave of grounded wire will definitely mess with the antenna's reflected power, and show a different pattern.
KX4O:Many lengths have measurable, but nonetheless minimal impact on the antenna's reflected power. It's true that only in about 1 in 6 or so of the cases does the impedance match well enough to actually cause a problem. So the moral of that story is very often you'll have the case where nothing goes wrong, the grounding system or the feed line length or both doesn't cause a problem. It's when the impedances happen to match we have the most issues. We get a false sense of security because most of the time, happenstance creates no problems for us.
KX4O:But we have to be very vigilant, and not create a situation where if you change one thing in the grounding system, or the mounting system, or the feed line length, we have a problem. And, of course, because all of this is because of transformer action along the wire, these conditions occur every half wavelength where they do occur. Now of course, a manufacturing example of this is my Diamond Antenna 770 end-fed Half Wave 2 meter / 440 mobile whip. This whip comes with instructions that specify not to ground the mount of the antenna. In fact, they give you a nice piece of rubber to wrap around it so that when you clamp it to your luggage rack, there is no conduction made to that luggage rack.
KX4O:All you have is a piece of coax. It goes from there into the the body. It's RG 316. It's very narrow, very thin, very easy to work with. Highly recommend it. But it happens to be a length that's kind of peculiar.
KX4O:It's not 12 feet, not 15 feet. It's 13.5 feet. So one can ask why such an odd length for a 2 meter mobile antenna that is an enf-fed half wave? Well, if we go back to the graphs on hamradio.me, in the case where the transmission line is grounded to the radio, which is of course is grounded to the car, 13.5 feet is right smack in the middle between 2 problem spots 2 problem lengths. So 12 15 feet are too close to points where we can have a problem with a feed line presenting a high impedance to the base of the antenna, which would force conduction.
KX4O:But 13.5 feet? It's right in the middle. Is this the reason why they have 13 and a half feet? I don't know. But the mathematics certainly holds up, and it's certainly my best guess as to why they do that.
KX4O:Conclusions: Simulation is useful to examine infinite halfway behavior and manufacturers seem to understand very well what's going on in the infinitive half wave world and have for quite some time.
KX4O:Finally, I'll highlight an experiment I did in my backyard with a 10 MHz dipole antenna. Just a piece of wire with supports at the end. It was not hoisted very far off the ground to enable me to get access to the feed points where I attached a Raspberry Pi 0, which contains one of these WSPR transmitter programs, as well as an adapter from TAPR that amplifies and properly filters the 10 MHz signal so you don't generate need harmonics. That works exceptionally well.
KX4O:It's wifi controllable. And to this, I also created a sensor that I can hang on the wire anywhere I want. It uses an off the shelf current measurement device for e m c work made by BeeHive Electronics, which generates a voltage which is then converted into a higher voltage, which then goes to an a to d board on top of another Raspberry Pi 0 and also wifi controlled. So I have 2 devices that I could put on the antenna that have no wires attached whatsoever to spoil the results that are both controlled by wifi. Again, the test transmitter and the H field measurement device.
KX4O:The point of the H field sensor is to, again, just like in simulation, where there's more magnetic field, there's more current. So it's an indirect measurement of current along the wire. And as you would expect, current is maximum in the middle and is less at the ends. All of this was combined into a PC and a wifi network. I had a shell open to each unit and I was able to energize the antenna and take a measurement from the H field probe every 1 foot along the antenna for the various scenarios.
KX4O:So I did this for both the center fed case as well as the end-fed case. The enf-fed case, the first transformer I used was from Par Electronics. It's a portable unit for, say, summits on the air work, low power. And the results show the center fed dipole and end-fed dipole both show a good current profile, maximum in the middle of the dipole and trailing off at the ends, just like you would expect. The end-fed dipole is in fact about at 1.6 dB less power than the center fit dipole, but it certainly is working.
KX4O:So, 1.6 dB difference and identical shape. The results of the center fed and end-fed tests show identical shape with the end-fed being about 1.6 dB less power than the center fed case. I put on a different transformer, also made by a PAR, but it's a high power version for another product of theirs. And in this case, both end-feds overlay each other, and they are both 1.6 dB less energy than the center fed case. So conclusions.
KX4O:Both the end-fed and center fed dipoles operate just fine without any additional conductors or feed lines. The end-fed energy is less than the center fed. If you do the math it emits about 1.6 dB less power than the center fed. Transformer loss likely explains the 1.6 dB difference. Center fed, end-fed, or pretty darn close to end-fed if you like.
KX4O:You can clearly see the dipole operates well from either feed point position and without benefit of a counter poise. Now some people suggest you need a .05 wavelength additional piece of wire to serve as a counterpoise for the end-fed case. While my previous test shows that that is not a requirement, I did repeat the test with a little wire attached to the end. And I measured current along it as well. And there is a little bit of current on it, but it's very small.
KX4O:It's lower than the end of the dipole section. And the current profile of the main is a little bit less. It actually worsened things a little bit. All these various situations effectively energize the dipole antenna with no particular problem whatsoever. This shows that there's little to no need for a counterpoise to make the antenna function.
KX4O:Finally, we'll address the statement that suggests these antennas only show up in Sammy Hammy circles. Well, the marine industry has been doing this for a long, long time, decades, because fiberglass boats don't provide a good counterpoise. Most of their antennas are end fed half wave antennas or variations thereof. In the land mobile radio industry, we have 2 products. 1 by Laird, 1 by Larson.
KX4O:The Laird b 1322w is in fact an end-fed Half Wave Antenna. Now most people mount that onto a metal car, and sure that has the benefit of a counterpoise there, but I have actually experimented with that in a portable backpackable unit and I can make it work without any particular concerns whatsoever about a counterpoise length. The military of course has had their say on this. There is a document from the Marine Corps. They don't explain much about the radio system, but they do highlight the fact that there is an end-fed half wave and how to use it.
KX4O:And finally, there is documentation dating all the way to 1928 that talks about the end-fed Half Wave.
KX4O:Okay. Some final conclusions. Well, you can feed a dipole anywhere you want it with proper design. The end-fed Half Wave doesn't need the other half of the antenna to work.
KX4O:Don't be misled by the popular press as it re-tells the same myths over and over because the end-fed half wave could be a good solution to your future antenna needs. Understand an end-fed half wave feed line grounded near the feed point doesn't necessarily mean large currents will flow to ground. Understand as well a counterpoise that presents a high common mode impedance to the high impedance feed seems to eagerly flow power. So there are counterpoise lengths that can cause problems, but short ones don't.
KX4O:Some people have commented, especially on my 10 megahertz test, and they suggest I repeat it with a much higher elevation. That's fine. I have no problem trying to do that. It does take a bit more of a crew to do this in a meaningful way. They're citing that it might be capacitively loading to the earth. Also, to take the counterpoise wire that I added, the 0.05, and make it much much longer and trim it foot by foot to see if I can trigger the point where it conducts a lot of energy.
KX4O:I should do that and I plan to. So The end-fed Half Wave Don't poo poo it. It is a fine antenna. You lose a little efficiency in the transformer, but sometimes feeding an antenna at the end is far more convenient for the various situations that might present yourself. Summits on the air would be a great example.
KX4O:I have used them many times for such endeavors. I hope this helps to sway some of the things you may be hearing in the various watering holes of the ham radio community. End-fed Halfwave. Go for it.
Creators and Guests
