End Fed Half Wave Wire Antennas

Updated: 8/7/2022 – A work in progress

End Fed Half Wave wire antennas must be great. Everyone is selling them! Meh.

The antenna engineer’s axiom:

You can have high efficiency (gain), high bandwidth or small size. Pick any TWO.

Real-world EFHW’s have low gain in most directions, they have low bandwidth within the band being used and they are not small…

But who doesn’t like experimenting with antennas?

The “EFHW” can be simple to rig in a portable or field situation. But they are generally my last choice for my primary interest in regional coverage on 80, 60 and sometimes 40 meters. Especially from remote, marginal locations where I need reliable HF Comms. Why? Some theory, basic modeling, ongoing experiments and field evaluations are in order.

A “general purpose” end fed half-wave HF antenna can work OK*.  For me, it’s only real attribute (the shorter, “40 meter” versions anyway) is that it can be simple to deploy in the field, especially if I only have one support available.

Also, as with any resonant length “long” wire antenna, EFHW’s have the handy potential to take power on multiple, harmonically related HF Ham bands, (often with additional impedance matching depending upon your installation and your transmitter’s capabilities).

They can work on other HF ham bands (as on 60, 30, 17 & 12 meters) “out of resonance” with even more lossy reactive help from an ATU of some sort, a further complication in a field deployment.  On those bands, coax losses increase very quickly due to antenna-feedline mismatch.

Performance can be problematic if erected near “anything” in the near-field, including the ground. Some sellers will make a “nod” to this reality but can offer no viable solution. “Try repositioning the antenna.”

The vertical “J-Pole” EFHW does get your VHF/UHF HT into the mid-range repeaters better than that Rubber Dummy Load antenna however.

Where does the signal go? A half wavelength wire in free space will radiate in the classic 3-D “Donut” pattern regardless of where it’s fed along its length; end, center, off center etc.  Consequently, the imaginary “end fed dipole” as some would call it only exists in free space, not in the real world.

* As a practical matter in a portable/field setting, a half wave end fed wire is often going to be deployed as a “sloper” with significant, asymmetrical ground interaction. If erected horizontally, parallel to the ground there will be asymmetrical feedline shield interaction off one end, even if “choked”. Those are significant problems that the typical center fed horizontal dipole does not suffer. In establishing desired, predictable azimuthal and elevation performance symmetry is good!

Slanted at a 45° angle up to a 90′ support and at a 20° “DX” takeoff angle as above, there are four nulls in the pattern including a deep 25 db null to the rear and nearly 9 db nulls off each side.  When that antenna is used on the second harmonic there are many smaller pattern nulls at the 20° DX takeoff angle.  Not great, just be aware.

For an 80° NVIS takeoff angle, and deployed to that 90′ support while operated at the lowest design frequency (half wave) the resulting EZNEC Model of the azimuthal radiation pattern shows it to be within about 2 db of omni directional; usable. 

For a deliberate low-angle fixed DX path, say to the “east coast” or “Europe” you need to try to orient the sloper wire to your 90′ support +/- 45° off that azimuth as seen in the above plot. If not, a lot of your RF goes where it probably doesn’t do you any good. But they will make random contacts on some bands, on some azimuths, at some ranges and at some times of the day.

Below is the same sloper antenna but rigged to a more practical 50 ft support. Azimuthal view, at a 20° “DX” takeoff angle; the support is in the direction of the +Y Axis in both plots.

132′ EFHW to 50′ support

With the far end 50′ high there is now an 8 db null in the forward direction towards your support. It helps to run the wire away from your target azimuth +/- 110° for a 20° DX takeoff angle. Rigged to an even more practical 20′ height that forward null is closer to 10 db. Also, the lower the end support, the higher the ground losses, any gain drops off quickly.

A compromised compromise EFHW deployment could be an Inverted L configuration as I am currently evaluating. Those waste lots of vertically polarized or horizontally polarized energy depending upon your view point, but big waste nonetheless. That’s true even if you install the necessary extensive ground radial system that the vertical element requires.

Additionally, those resulting simultaneous ground wave and sky wave signals from an Inverted L will (according to Thomas Young – of Double Slit Experiment fame) constructively/destructively interfere with each other creating hot and (according to Murphy’s Law) big dead coverage areas. Areas that move, creating self-inflicted, even worse QSB fading due to ionosphere dynamics.

I have also evaluated my experimental 132′ EFHW while deployed as a horizontal “L”, 18′ above ground. With equal length legs, the “vertex” angle is about 73 degrees due to non-ideal supports availability.

On 80 and 40 meters it is omni directional to within 1 db at high “NVIS” angles. The half power points in the elevation plane are +/- 45 and 30 degrees respectively from vertical as expected (EZNEC Pro model). The measured SWR on the targeted CW portions of 80/40 and 20 meters are under 3:1 but much worse in the Phone bands. It’s not great but my military field sets can at least drive it unaided.

As such it has been a useable backup antenna for regional comms out to about 300 miles in NVIS mode with my 10-20 watt field sets, my primary interest. It is a poor performer for long-range DX beyond that however.

It is handy at the home QTH for “general purpose” receive duties. The laws of transmit/receive antenna Reciprocity apply.

Powering an EFHW Antenna:  If you can get power into them they can radiate it (somewhere) well.  That’s because the high current portion (the portion in the middle approximately third of the half-wave wire that does most of the radiating) can be up and away from the radio, infrastructure and some local ground absorption losses.

But that portion needs to be horizontal and low for effective regional NVIS, or horizontal and a half wavelength high for long distance comms.  Or vertical (with an extensive ground system) for either local ground wave or long range omni directional “spray & pray” efforts. (Verticals radiate equally poorly in every direction and essentially not at all straight up for NVIS.)

Unless* the feedpoint transformer is also mounted high up it’s hard to rig an end fed wire that doesn’t simultaneously do all 3 modes like that.  (Resulting in lots of “lost” or wasted RF power that doesn’t get to your target.) Especially when operated as a two or more halfwave antenna on the higher frequency bands; their azimuthal radiation and takeoff angle pattern becomes full of lobes and nulls. * Note: Take a look at W9HH’s good input on elevated feedpoints below in the Comments section.

With a radio designed for “50 Ω” loads fed with coaxial cable, an end-fed halfwave wire will require a matching device.  Those are air, or more typically ferrite core transformers which many people incorrectly refer to as “Baluns”. They are actually “Ununs” with unbalanced coax cable driving the unbalanced end of the EFHW. (Ruthroff Ref. 107).

They are needed to transform that very high antenna impedance into something much lower (and that transformation incurs losses and extra complexity in a portable field deployment).

Transformer designs, EFHW antenna examples and success testimonials abound on the Interwebs. There is a lot of empiricism out there, some of it is well done.  But the more I read about the topic the more I am reminded of the old saw:

The plural of anecdote is NOT data.”  Even if you can get past the Apples versus Lug Nuts comparisons.   Choose wisely.

There are some very good engineering circuit simulations on the Interwebs that explore the transformer design variables and their effects. Some also include tests and measurements of installed systems. Good stuff.

Note that many EFHW antenna users report that the transformer core gets hot at moderate power levels and then the SWR begins to climb with heating towards the ferrite Curie point.  Coax losses also climb correspondingly.  Not good – unradiated power lost – but maybe those losses contribute to the EFHW’s potential to “take power” on multiple bands.  Just as a dummy load is broadband. Refer to the antenna design engineer’s axiom, above.

Hill Billy Heat Test: I wanted to see if this experimental kludge transformer core got hot with 100 watts of short term power on 3550 kc while driving my antenna under test.

Since I could not be in 2 places at once: The Poor Man’s Pyrometer. I put a few shavings from a paraffin candle on the ferrite core as a witness sample; the wax melting temperature is somewhere between 52° and 71° C.

After 60 seconds key down, 100 watts, open to the air, no sign of melting, no SWR changes. The Curie point for Type 43 ferrite is 130° C, so for my purposes, the transformer efficiency with this big FT240-43 core is adequate. “Good to go” for further testing.

Wax Pyrometer

If necessary, the efficiency can be increased by adding more turns to the primary (and correspondingly the secondary to maintain the desired ratio.) My 2-turn primary has adequate inductance at 3.5 mc and is working within the winding space available.

An end fed half wave wire can present a feed point impedance of 2500 Ω or more. That’s a 50:1 impedance mismatch for a 50 Ω output radio (SWR of 2500/50=50:1).  A typical ferrite core transformer as above with a 7:1 turns ratio (49:1 impedance ratio) or more will be “helpful”.

As an example, on 80 meters, with the radio or feed point on or near the ground, the wire end might be suspended to a support 90+ feet high; the slope angle is then 45°.  EZNEC computed Z=1525 + J 2064 Ω. 50 Ω SWR = 51:1

Note that “antenna SWR” at the feedpoint does not effect radiation.  High SWR on the feedline (if there is one) just causes feedline losses and losses in the transmitter/ATU.

RF Transformer Internet Folklore says that the primary and first 2 turns of the secondary “should” some how be twisted together.  This would form a distributed “gimmick” capacitor between the primary and part of the secondary; that capacitance becoming an element of a high-pass filter. 

Folklore also recommends shunting the 50 Ω terminal to ground with a discrete ~100 pf capacitor, thus becoming part of a low-pass filter. That capacitor also forms a capacitive voltage divider along with the gimmick capacitance. That supposedly increases the bandwidth to higher frequencies by compensating for any series leakage inductance. 

I tried that in my test antenna setup. I found that the 100 pf cap has little measurable effect, at least on the 80/40 meter CW bands that I am mainly interested in. (Note: the HV rating of the doorknob capacitor is not necessary, the primary voltages are fairly low, it’s what I had handy during testing.)

49:1 Z Transformer – “Conventional Wisdom” Design

Also, folklore has it that one should “cross” the core for the second half of the secondary winding.  This presumably to reduce the stray capacitive coupling between the high impedance output end and the low impedance 50 Ω terminal, which would partially bypass the transformer.  It would have some kind of minimal effect in the lower HF. I’ll give it a try.

Below is an experimental QRP version for the portable field antenna kit if used: A 43 Mix toroidal 49:1 (Z) impedance transformer for a high impedance EFHW antenna. Mix 43 is appropriate for the lower HF bands, my primary interest. The transformer helps, if you must.

QRP 49:1 Z Transformer

Above: Mounted in a plastic 35 mm film can (remember those?). The toroid does not get noticeably warm to the touch with my 10 watt military field sets operating on 80 and 40 meters.  There is sufficient core cross-section to be adequate with low power sets.  The compact packaging is Good to Go for Bush Ops.

This image has an empty alt attribute; its file name is IMG_5739.jpg

Film Can 50 Ω – Hi Z Transformer

One way to fine tune the transformer turns (impedance) ratio is to terminate the output with a non inductive resistor or a 5-10K non inductive pot (rheostat).  Add or delete secondary turns from “design” to yield the best match versus the resistance presented by the resistor/pot.

This image has an empty alt attribute; its file name is Testing-EFHW-Transformer.jpg

E.g. Using a 49:1 transformer shoot for the best 50 Ω match @ appx 2450 Ω load over your working frequency bands.  Moving the pot up and down gives a feel for the SWR versus antenna impedance sensitivity. 

Then verify it with your installed antenna and SWR measurements. Tweak the antenna length if necessary and then maybe adjust the turns ratio again. Then –

The antenna may also require a “compensation” inductor(s) and/or a series capacitor somewhere (?) along its length to get it to perform as well as it can on the band(s) you are most interested in. Then changing the height/placement will probably nullify those specific tweaks…Drop, tweak, hoist, measure, repeat… (Groan…)

It may also require a “choke” inductance somewhere back along the coax since the coax shield is also a part of (and “completes”) the antenna as it heads to your operating position. The length of the coax itself might then need changing to make the impedance the transmitter “sees” acceptable.

132 foot end fed wire antenna

This 132′ wire end can be plugged into the transformer output with a banana plug or a pigtail, the radio connects to the BNC connector. 

The GRC-109, TRC-77, RS-1 and RS-6 sets that I use in the field will deliver full power to low-to-medium impedance antennas (typically up to around 1200 ohms) unaided.  When trying to drive a high Z antenna like the EFHW, their output power indicator lamps (series antenna current) won’t glow, therefore precluding any PA tuning adjustments – which wont have any effect in any event. Not good.  These transformers are an engineered attempt to get around one of the EFHW’s systemic problems.

The GRC-9 however can work adequately with higher impedance end-fed wires as part of their issued antenna kits. (Those kits also include a substantial ground radial system.) In the “REEL” antenna selection mode, their output coupling network is designed to efficiently handle higher impedance, end-fed antenna loads relative to the high plate impedance of a vacuum tube PA.

On 80 and 40 meters, and as-rigged in the field, these transformers can deliver antenna power depending upon the height/near-field environment.  Who knows which directions or takeoff angles the resulting RF goes when you have a specific target (see models) but some of it comes down somewhere.  It does make contacts to random places if it’s all that I have, or all I am interested in at the time.

An EFHW antenna also presents a high impedance on n-Even harmonics. Since you already have a 50/HiZ transformer feeding it, the system will also “look like” a low-enough impedance load to the transmitter on those higher harmonic frequencies. Hence the system is “multiband” for that reason – it will take power, possibly with additional ATU assistance if the transmitter needs it to deliver rated power.

Note that the EFHW has little merit when considered as an antenna for today’s military communications. Its “multiband” capability is useless in military service since military frequency assignments are not made on any harmonically-related basis.

For Hams operating ALE (Automatic Link Establishment) nets confined to harmonically related Ham bands the EFHW can find some utility. At least as far as getting power into one. That is a whole other discussion.

Here’s another experimental 49:1 Z version, this one with a larger FT140-43 toroid.  This one will handle more power.

FT40-43 50 Ω/High Z Transformer

BNC connector to the radio, antenna Banana jack to transformer secondary, alligator clip for ground radial. Awaiting field packaging but it may be a little too big for my portable field kit.

Below during a recent field Op. The transformer now sitting on my tire serving as a high voltage insulator. Improvise, adapt, overcome…Still awaiting field packaging!

Shown below is this more-portable 132′ antenna made with WD-1A/TT infantry field telephone wire, AKA “slash wire”.  By design, the wire is thin, olive drab, dull and hard to see in the trees. This light weight wire (NSN 6145-01-047-4345) has a breaking strength of over 200 pounds.  It is made of 4 tinned copper wires for low resistance and 3 galvanized steel wires (for strength) in each conductor, “zip cord” style. The 2 conductors are tied together at each end.

The insulation is rated at 1000 volts RMS. It is also “slippery” with its vinyl jacket; it slides over tree branches very well.  The 3 steel strength wires make it somewhat “springy” but it’s good stuff for any wire antenna, especially in the field. Also OK for a permanent installation but bigger AWG copper would be somewhat better losses-wise. The receiving guy would not even notice.

I recently deployed it at a campsite in an Inverted L configuration 25′ over ground radials. The SWR (at the BNC connector) with this 49:1 transformer was around 1.3:1 on 3550 and 7050 kc, my design targets.

However the SWR nulls within the 14, 21 and 28 mc bands were all well over 4:1, worse in the CW portions, not good.  My experimental EFHW antennas have demonstrated rather narrow tuning between the in-band 2:1 SWR points. This indicates relatively high Q. But measured SWR is a small part of the overall performance picture.

It is usable at camp sites for intended regional NVIS contacts with my buddies or home QTH on 80 meters. I’m still evaluating its AZ/EL/Range coverage compared with long experience with my low dipole which is still my preferred field antenna.  Especially since those “camp skeds” are usually to specific locations on that single band. The simple dipole excels but this system will be getting a good field evaluation in various configurations. .

EFHW 80 Meter antenna with transformer

On that camping trip I had this WD-1A/TT antenna rigged as an inverted L over a ground radial wire. It worked OK on 80 meters CW to a buddy 200 miles away but I was unable to run any EFHW/dipole comparison testing as I would have liked during this deployment.

While running CW and radio teletype, the PRC-174 transmitter output indicator showed power being delivered to the antenna system on both 80 and 40 meters. That TRC-77 also delivered full power to the system. A pretty hasty campsite setup, the experimental transformer is still awaiting “field packaging”:

Field Ops with the PRC-174 set up for Signalink RATT Teletype Ops

During transit the banana plug somehow got lost. So a Mk 1 Mod 0 Field Expedient Banana Plug was jury rigged with a cooperative twig jammed into the female connector along with the antenna wire. Improvise, Adapt, Overcome.

This image has an empty alt attribute; its file name is IMG_6222-2048x1536.jpg

That’s overnight snow in the background.  The soil was pretty wet at this camp site where it had melted.  The alligator clip grabs the ground radial.

Below: Now testing with a larger FT240-43 ferrite core and mounted at my “Antenna Test Range”. The 49:1 Z transformer is at the base in its weather proof box, the coax and radials are buried:

N6CC EFHW Inverted L Base Transformer Housing

(Below is another version; nominal 600 Ω transformer for a higher power transmitter.  This one would be good for a vertical half-rhombic or other medium-Z wire antenna such as a non-resonant end fed wire):

50/600 Ohm Ferrite RF Transformer

Below, bench testing the smaller version with a non-inductive resistor or rheostat (Pot) load, set to 2450 Ω.  (The meter Resistive component of R+jX is out of calibration, it shows 40+j0 Ω with a 50 Ω dummy load. The measured SWR and frequency are OK.)

FT140-43 Transformer load test. Just moving your hand near the output terminal causes excursions in the measured SWR while testing. This is indicative of the sensitivity of the transformer to stray/shunt capacitance appearing at the output – as presented by an actual antenna.

The resulting impedance seen at the input is quite sensitive to the antenna’s immediate, near-field environment.  That is part of the source for the EFHW’s widespread reputation for being “finicky”, “won’t load”, “position sensitive”, “still needs a tuner”, “RF in the shack”, “so I returned it”, etc..

The following attempts to simplify the complex “end effects” involved with these systems.

The system requires a transformer and the cold end of that transformer secondary must be tied to a reasonably good “ground” for consistent results. Even though the current flow in the antenna wire is relatively low at the feedpoint.  It is NOT zero because the antenna impedance is NOT infinite; Ohms law applies.

Why the grounding requirement? Because RF current flow in a conductor in concert with the oscillatory electric field is what causes Electromagnetic radiation (ala Mr. Maxwell) and the transmitter does not “create” current from nothing (ala Mr. Kirchhoff).

With an unbalanced end-fed antenna like the EFHW the power amplifier/output network essentially “pushes” current into the antenna by “pulling” an equal current from the ground system. This creates a complete circuit for the sinusoidal RF current to flow in the antenna. The resulting E/M field induces current flow back into the ground system. Repeat. This is why ground losses must be minimized – they waste transmitter power because ground losses are in series with the antenna radiation resistance (ala Mr. Ohm).

Some commercial EFHW antenna manufacturers (salesmen) state that a ground/counterpoise is not needed (except for safety/lightning protection; true).  They rely on the uncontrolled, random capacitance between the coax feedline shield (or your body/radio) and the lossy soil to complete the circuit.  Not real good.  Why would you even consider doing that? (Better in a permanent installation with buried coax cable.)

Also, providing a good DC/RF ground at the transformer secondary will also quickly bleed off any precipitation, snow or blowing sand static charge.  Bringing that into your shack to find a decent “ground” via the coax shield is not a good idea.

Note that voltage, including “high” voltage, is measured between two points: The EFHW antenna feedpoint and the chassis, ie: ground (and hopefully not via YOU, capacitively or otherwise.)

The end fed half-wave antenna will necessarily have high voltages at its far end and its feedpoint – and the matching transformer must withstand every volt.

Eg:  A transmitter delivering 100 watts to the end of a 2500 ohm EFHW antenna will see about 500 volts (700 volts peak) at the output of the matching transformer to chassis ground.  Generic 600 volt rated hookup wire is not very suitable.  How about that uninsulated antenna wire? Something to worry about when using a “high powered” transmitter (or fingers).

One other problem with an EFHW antenna in a portable, temporary or field setting is purely mechanical/practical.  My 80 meter CW band half wave wire is 132′ long.  When deploying mine in the woods it seems like it is a VERY long 132′.

In the forest it can be cumbersome to unwind/erect versus say, 2 quarter wave dipole halves.  Those can be individually launched and then connected to the center insulator/coax and hoisted up rather easily, even as an inverted Vee if only 1 support is available.

A continuous 132′ wire will often require an intermediate tree(s) to keep the sag minimized. Sag may cause SWR excursions in the wind; can your transmitter keep up? Throwing an additional weighted messenger line through a mid-span tree can be difficult and frustrating. Often a 2-man job.

As with all antennas, the EFHW is a compromise. They can be further compromised at installation by available real estate and supports geometry while trying to favor a particular path. They are not particularly efficient at either short or long ranges. But they will produce contacts as any wire antenna of similar length and height will.

For some more ideas on portable field antennas take a look here: https://www.n6cc.com/field-antenna-kit/