Tag Archives: hwef

The Resonance

The reason why I would bother with this topic is twofold: One: I think better antenna systems are actually better; and Two: Technical concepts are central to amateur radio and should be understood and employed as often as possible. I’m discussing these antenna concepts in simple terms in the interest of keeping it accessible. Hopefully it leads to some further investigation and a different way to think about antenna systems.


In the Venn Diagram world it can be said: Not all 50 Ohm feedpoints are resonant, and not all resonant radiators present 50 Ohm feedpoints. I have seen much confusion between “resonance” and “impedance match” on internet forums/platforms/videos. This what drove me to churn out this post.

Resonance in an RF radiator actually does make a difference. I’m a fan of Hemholtz and resonance is a real thing. When I assess a antenna design I start at the radiating element or elements. If that element is not a known and desirable harmonic relation to the fundamental frequency of interest I consider it a non-resonant design. If there is a matching unit connected to a non-resonant radiator I consider it a tuned circuit. When the goal is efficient and predictable RF radiation you have resonant radiators, and then everything else.

Recently there has been an explosion in the use of end-fed wire antennas. I use a few of them and they are an easy way to get on the air, usually on multiple bands. I have made plenty of contacts on these antennas, but at no point was I under an illusion that I was using an efficient antenna system. The now ubiquitous End Fed Half Wave is a strange bird, It utilizes a half wave radiator with a feedpoint impedance of about 3000 ohms, and is operated without a traditional ground plane or counterpoise. The high impedance feedpoint is matched to somewhere near 50 ohms with a 49:1 or similar toroidal un-un transformer. That is not a recipe for efficiency. The power ratings on the matching transformers tell you all you need to know: These matching units get hot, and that heat is your RF not making it out into the world.

PAR Trail Friendly on a Spiderpole

That’s not to say it’s a bad design. It solves a few of the issues that keep many hams from being able to get on HF. You can hang it from a single support, as opposed to needing two or three for a horizontal dipole. It’s also easy to erect on demand, which is great for hams who want to operate portable stations or if they have a compromised (small, HOA, etc…) QTH. I use a 12m Spiderpole and a PAR Trail Friendly when Inwant to get on the air quickly. Also, thanks to the magic of harmonic resonance, it is common to find a match on more than just the fundamental frequency.

Another common design is the random length end fed, usually known by the 9:1 un-un transformer used to feed the non-resonant radiator. The trick here is to find a radiator length that is not resonant on any band you want to work. The hope, and I do mean hope, is that the feedpoint impedance will be somewhere in the 200-800 ohm range, where it can be matched to 50 ohms using a second matching device. Again it has the advantages of ease of setup and multiband capability. The tradeoff is even steeper than the HWEF. This design is force-feeding a non-resonant radiator, has an even lossier (IMO) un-un, and requires a second matching device to get all the way to 50-ohms.

A problem both designs share is the difficulty in modeling a radiation pattern. Even with many analyses and many users over many years nobody seems to be able to say much beyond “omnidirectional”. That’s not too helpful. The main issue in modeling these systems is properly representing the ground/counterpoise. Most installations (mine included) require a common mode choke (CMC) near the radio end of the coax to keep RF from energizing the radio’s earth ground. How efficient does that sound? Do we have any idea what these antennas are actually doing wit the RF that makes it out into the world? Empirically I think we do. Scientifically, predictably, I don’t think we have a good handle on it.

So, how important is the efficiency of a transmitting antenna?

Hams with even some basic experience on HF know that you can make contacts on almost anything. Take a simple transmatch design like the MFJ-901, hook it to a rain gutter, and make a contact. Do we know anything about the radiation pattern, efficiency, or bandwidth? No. We are just interested in forcing something conductive to radiate RF, and hope for the best. The magic of radio waves takes care of the rest. If enough of your RF makes it out in the correct direction you will make a contact. Your signal might even be strong! But this is where it helps to be aware of confirmation bias. The science of radio communications involves being able to control where your signal is going, and control the system design to connect you with the intended destination. Having made some contacts is almost inevitable. Making the contacts you want to make when you want to make them is where the game gets interesting.

In the food-chain of antenna designs these lossy designs are somewhere in the “krill zone”. A good efficient design, properly installed is somewhere in the “cordata zone”, and a very efficient directional design mounted high and in the clear with directional control is “blue whale” territory. At some point the operator is limited by their Effective Radiated Power (ERP) and the weakening of the transmitted signal with the square of the distance from the antenna. Using 10-30% of your RF to warm a toroid is cutting in to your effective range. Every system has limitations, but when running 10, 30, 50 watts at the finals, I feel that making the most of it is important. Every improvement in efficiency and pattern control brings you up another link in the radio food chain.

Mitigating Factors:

Certainly the parallel popularity of end-fed wires and weak-signal digital modes, best represented by FT8, are not coincidental. FT8 allows working at lower signal to noise ratios which is like getting that lost RF back when compared to working SSB or RTTY. More efficient antennas are always an improvement, but an entry-level HF rig, a HWEF, and WSJT-X is a great path of entry to the HF bands. I’ve gone on about this earlier, but there is nothing in-stone about needing to start with CW or SSB. Hams can get on the air, interact with DX, and get some good contacts in the log. I see nothing wrong with that.

Compact, Broadband, Efficient: Pick Two

In the HWEF and 9:1 design cases the user is giving up efficiency for a more compact, more broad-banded (lower Q) device.

The HWEF, thanks to its resonant radiator, is somewhat more efficient, but the price is being paid for getting wide frequency coverage on a single radiator. Some designs find ways to better balance the equation by sacrificing some bandwidth and band flexibility to increase efficiency. One QRP design I use is the PAR 102040 Trail Friendly. This design uses three techniques to get a three-band end-fed vertical into a single 41′ wire radiator package. Trick 1: It uses a trap to isolate the 20m half wave element from the 40M extension. Trick 2: It uses the characteristic electrical lengthening effect of the trap to keep the 40m extension (and overall length) shorter than a simple 40m half wave radiator. Trick 3: using the second harmonic of 20m to realize some usable bandwidth on 10m. The transformer is pretty much handling the design goal of representing a 3000 ohm feedpoint as a 50 ohm load to the rig. No additional trans match is needed if you have trimmed the 40m extension properly and are ok with the tuning points. It isn’t brilliantly efficient, but it’s not horrible.

Chameleon MPAS Lite – Mini Review:

In early 2021 I picked up that Chameleon MPAS Lite and wanted to give it a run as a portable antenna system for POTA. This unit is extremely compact and is comprised of a ground spike, a un-un, a heavy duty stainless whip and 65’ of wire to use as a counterpoise. With the engineering triangle in mind I knew I was giving up a lot of efficiency for a very compact, very broadband system. And that’s what I got. It works well enough, especially if you are calling CQ and self-selecting stations who can hear you. I found that replying to anything but the loudest stations was a bad recipe for success. In a Chameleon forum I made the comment that these antennas aren’t good DX setups. I stand by that. It isn’t that you won’t work any DX, you would just work more DX more easily with a better DX design. For mid-range and NVIS work I think it is a solid solution. That assumes you are able to make the contacts you need to make. Also, watch the power rating because the un-un will get warm, and will fail under excessive power and duty cycle.

The MPAS system uses a 5:1 un-un design to feed anything from their very nice stainless whip, MIL Whip system, a longwire, or whatever conductive item you decide to use. The 5:1 takes a conservative approach to the non-resonant radiator problem, and a second transmatch (internal or external unit) is required unless you get lucky and dial in a 250 Ohm (or 10!) feedpoint impedance for the Un-Un. Chameleon publishes a perfect omnidirectional pattern for their antennas. That’s a good assumption for the operator since non-resonant systems are difficult to model accurately. Just assume your RF is going in all directions equally. Might as well! In practice I have made contacts and successfully activated a few parks using the MPAS Lite. Chameleon builds a beautiful and rugged product, and supports their users well. II feel like it is a good system and the user will get the best out of it by realizing it’s strengths and weaknesses.

One concern I have from following a few antenna forums, including a Chameleon product support forum, is the number of users claiming to have “resonant” solutions. An actual resonant radiator would provide a terrible match with a 5:1 un-un. What users are doing is finding a radiator length that presents a 1:5 feedpoint impedance to the CHA transformer so the radio sees 50 ohms(ish). That approach may be usable on that band with no additional matching, but is likely to be worse on other bands.

Another constant question on these forums regards radials and counterpoises. Strangely the solutions discussed are often none, one, or several short radials. All of these are poor solutions. However, when the central design concept is to never be resonant, not require radials, and use almost anything as a radiator, I don’t know what a good solution looks like. The Chameleon counterpoise kit is 65′ of very hunky insulated wire and can help you play with the counterpoise dimensions/layout and maybe help in finding a match. That said I feel like users are trying resonant counterpoise lengths and I think that does more harm than good. With a single counterpoise, stick to the game plan. Why make the two impedances different?

The Chameleon 50′ RG-58-ish feedline with a ferrite bead choke on one end is a decent way to control the stray RF. And it is necessary. Every time I have used it I had to control RF that wanted to energize the radio’s earth ground.

The best application I have seen for these matching units is Chameleon’s own Tactical Delta Loop which uses the CHA as the feedpoint in a system where 5:1 isn’t a bad design value. I will be cobbling a test setup soon out of various Chameleon and Wolf River parts.

Wrap Up

If it sounds like I am down on end-fed designs, I’m not. They have a proven track record and get hams on the air. What I would like to get across is that I hope ops get some time on other designs. Building a nice 1/4 wave vertical over a good set of radials is very easy and cheap. Think of a it as a single-band DX Commander style build. I think it is worth the time and small expense to see how a resonant design works and maybe make a few on-air comparisons.

If you stuck with this post I owe you a pint. I don’t write these to be concise blurbs. I like to compose my thoughts on these topics and I figure sharing it is part of the experience. Let me know what you think! Reply on the blog, or look me up on QRZ and drop me an email.

Get on the air, and always have fun. 73, Pete N1QDQ