KB6NU finally builds an end-fed, half-wave antenna

I’ve posted about end-fed, half-wave antennas before, but until this weekend, I’d never built one. One of the reasons for this is that most designs are for QRP antennas and not made to handle more than 5 – 10 W of power.

A couple of months ago, I ran across a design rated at 100 W. The design seemed relatively simple to build, requiring only a single toroid and a capacitor made with a short length of RG-174 coax. Well, it just so happens that I bought 100-ft. of RG-174 at Dayton this year, and I found the toroid cores online from the  “Toroid King” for a very reasonable price, so I decided to make my own.

All told, the parts cost about $10, the biggest part of that the plastic junction box I got from Lowe’s for $6.41. Compare that with the $60 that LNR wants for their end-fed antenna.

Sunday was a beautiful day here with temperatures in the 60s, so I decided to put up the antenna. I’d love to report that everything worked perfectly, but this antenna is going to need a little more work. The SWR is 2.6:1 at 14.000 MHz, dropping to about 1.5:1 at 14.900 MHz.

Since the internal tuner on my IC-746PRO is supposed to be good to 3:1, I did use it and made a couple of contacts. The guy in MA even gave me a 599 signal report. I’d be a little more comfortable about using it, however, if I could get the SWR down a bit.

I e-mailed the guy who published the design and asked why he thought the resonant frequency was so high, and he said that all I had to do was add a couple feet of wire to the antenna. He also suggested that adding a turn or two to the coil would bring the SWR down. I did some more reading about end-feds and I’m thinking that perhaps adding a short counterpoise might be something to try, too.

So, while the results so far have been mixed, I’m hopeful that with a little tweaking, I’ll have another antenna to add to my arsenal. It’s been a good learning experience, and I’ve certainly saved a bunch of money over the commercial versions.

UPDATE 11/18/12:
I added 24-in. to the antenna  and it did indeed bring down the SWR of the antenna to below 2:1 in the CW portion of 20m.  I’m happier with this. I still do plan to try a counterpoise. Not so much to improve the SWR, but to see if it makes the antenna a little more efficient.

The box I used for this project is the Carlon E989NNJ, a 4-in. x 4-in. x 2-in. plastic junction box.  This is a very nice box. Not only is it completely enclosed. The screw down cover is gasketed, making it waterproof.

I liked the box so much, I went back to Lowe’s to get a couple more. Not only were they out of stock, when I searched their website for that part number, it came up with no results. It kind of looks to me as though they’re not planning to stack this box anymore.

I Googled the part number and found several places online that had them, but they wanted more money, plus I would have to pay shipping. Fortunately, I was able to find some at a local Home Depot. They wanted $6.88, compared to $6.41 at Lowe’s, but at least I was able to purchase a couple more of them.


  1. Let’s hope you don’t get any nasty RF Burns from those end fed 1/2 wave antennas ?
    Sure, end fed antennas have some advantages is erection, but there always seems to be a lot of RF on the feedlines of them. They are usually power limited, maybe a good thing, with all that stray RF floating around. I know a Ham who has an end fed 1/2 wave antenna for 20 meters he uses vertically, but at QRP power levels. He simply pulls it vertically up a tree limb.

  2. Elwood Downey, WB0OEW says:

    Hi Dan, welcome to EFHWs. I’ve had the best success feeding them with a parallel tuned tank circuit and little or no counterpoise (as per http://www.aa5tb.com). They are quite narrow band but ideal for outdoor portable use when the tuning unit is right next to you on the table with the radio. In search of something more broadband for less fiddling I’ve tried just a transformer with a high turns ratio but that does not seem to work as well. Now even though end-feeding dipoles should in theory function as well as center feeding, I’ve always been a little dubious that feeding at the high voltage point could ever be as effective as locations with higher current. This seems to be born out somewhat by the measurements recently posted by N4KGL at http://www.n4kgl.info/2012/11/measuring-end-fed-half-wave-and-dipole.html. Have fun experimenting!

  3. Bob, KG6AF says:

    Don’t be discouraged; these things do work. My 20-meter Par (now LNR) end-fed is around 1.1:1 at 14.050 MHz, and less than 2:1 across the entire 20-meter band.

  4. Dan – After you experiment and get the SWR down, I hope you will post the results and design tips here in your blog. Also, I would be interested in building one for 75 or 160 meters as I like to work the low bands, so if you have some notes on how to modify this for other bands, it would be appreciated!

    • I certainly will, Tom. I’ll update this post as I try different things.

      I was also going to ask the designer how to go about choosing the L and C values for different bands. Like you, I can see building one for the 80m band.

  5. All this would be a sight more useful if there were some explanation of what these antennae are, besides the tradenames!

    The main problem with the end fed is that the fed end is near the ground! I’ve used them to a KW and more, but sometimes I could draw an inch of arc off the capacitor in the shack.


    • Dan KB6NU says:

      Wilson, the purpose of this post is to describe how I built the antenna. The plans are at http://earchi.org/proj_homebrew.html. For a technical discussion about end-fed, half-wave antennas, go to http://www.aa5tb.com/efha.html.

      You say “The main problem with the end fed is that the fed end is near the ground!” Well, as far as I can see, there’s no technical reason that it has to be. You can raise it up as high as it needs to be.

  6. Isnt an Enf Fed 1/2 antenna basically a 1/2 vertical antenna laying on its side with no counterpoise ?

  7. Wouldnt mind something like this for 80M DX work. I need to build something for 80 and 160 so I can start DX’ing on those bands more.
    Also, how do the End Fed designs get along with older tube transmitters ? Any concerns ?

    • Dan KB6NU says:

      You can orient an end-fed, half-wave antenna vertically, if you have the space, and it probably would have some interesting characteristics.

      Some folks say that you should have some sort of counterpoise. Mine doesn’t currently, but I intend to experiment a little bit with one to see if I can achieve a better SWR and/or make it more efficient.

      I haven’t heard of any problems using them with tube-type transmitters. I wouldn’t think there would be any problems, but I’m certainly not an expert.

  8. I’ve used the PAR/LNR end fed half waves with great success on 40, 20, 17 and 10.
    I also use the triband version (40/20/10). They all work just as well as a center fed dipole. Nothing unexpected save for they build the very nicely. I have built my own prior to that and I can say they did work and the key is getting the inductance right
    and the capacitance right. The best way to tune them is use a load resistor of say
    3300-4200 ohms (not that critical) and tune if for 1:1 with that value in place then take it out doors and add the wire and tune that for the same swr you got with the resistor. FYI rather than AA5TBs parallel tuned circuit you can use a L network
    with the correct values (source 50 ohms, load about 3800 ohms) less complicated.

    Feed point resistance at resonance is dependent on wire (or tubing for free standing vertical) diameter and height above ground (for horizontal wires) with some variation with inverted V and inverted L or sloper installations. Again if you use a median value of say 3000 ohms for the load a 2:1 variation is anything from 1500
    to 6000 ohms and that a is why if tuned right they get less fussy.

    Those that get excessive return currents (hot coax) failed to get the antenna wire
    resonant at the desired frequency and are actually running a random length wire
    with an attendant high reactance and lower feed point resistance which will definitely put RF on the coax as it then needs a solid ground. NOTE, resonance for
    a half wave is R of some value (varies with feed point) and X reactance of zero (0).

    High voltage on the ends, all half wave center fed dipoles have that too. All feed points and antenna ends should for safety sake be out of reach.


    • Dan KB6NU says:

      Great comment, Allison. I hadn’t thought about tuning the matching unit before hooking up the wire to it. Thanks.

  9. Allison says:

    Some followup notes:

    A end fed half wave is the same thing vertical or horizontal. Its electrically the same as a half wave dipole with the difference that rather than center fed its fend from the end.

    If anyone has worked with off center fed dipoles you know as you get closer to the end the feed point impedance goes up till you reach the very end.

    All antennas that are not self contained (center fed dipoles, Loops) have some form of counterpoise or ground, an end fed has such a high impedance feed point that ANY ground/counterpoise is enough to work against. Most of the commercial units the coax shield is that counterpoise.

    For 80M and 160M its doable using an L-network (note the wires will be 130 and 260ft giver or take). This is usually done in the form of an inverted L due to the
    length (unless you have tall wide spaced trees).. And yes you can bend any half wave antenna (dipole or end fed within limits) into a V, Inverted V, vertical or inverted
    L to name a few. I’ve run an half wave inverted L on 75M and it was very effective
    despite the vertical portion being only 30Feet and the rest horizontal at that height.

    End fed half waves as horizontal or sloper are sensitive to the height above ground just like the center fed version. Bother work best a height (for pattern and termination resistance) however the end fed can originate near the ground and go up (vertical, sloper to high point, or inverted L) with less troubles.

    RF burns from end feds are about the same as dipoles, the end have high voltages.
    IF there is significant RF on the coax then something is wrong. Likely causes are the
    wire is not a half wave (detuned by trees or close to ground) or the feed point match box is maladjusted, there are magic cable lengths that will both pick up RF and put voltage peaks near the operator. Solutions for the later case is longer or shorter coax (assuming the antenna is tuned) or a short counterpoise at the antenna feed point (try about .05 wave length long). I’ve not had a problem at the 130W level on any band.

    With EFHW antennas using an adjustable matchbox there is a great willingness to use the tuning of the matchbox to “fix” a detuned wire or a wrong length wire.
    IF you do that its not an EFHW as its now a random wire with a random wire tuner
    and one must use a ground or counterpoise to work against. The easiest way to tune the box is a resistive load like a 3300 or 3800 ohm resistor setting it sow the swr is 1:1 or extremely close to that. Then the wire is added and adjusted for the correct length by a low or hopefully 1:1 match. The exact length of the wire Will vary with the height of the wire especially the far end.


  10. Tom WB8COX says:

    I ran across your post the other day while I was considering how to solve an antenna problem when camping.  For VHF contesting, the site has to support mounting the beams on the camper as well as being a high elevation.  For general camping, the quality of the site takes precedence, especially here in the South where shade makes it possible to keep the camper air conditioned and the fridge in the safe temperature range.

    A week ago we were camping at Wind Creek up on Lake Martin, the site was perfect, at least for sitting outside and enjoying the view, but there was no way my 6BTV was going up unless I wanted to rest it against a tree.  So I was beginning to consider an end fed antenna as part of the camping supplies.

    I decided to try building one and checked out just about every website I could find.  The antenna handbook was worthless, as was the regular handbook.  There are many websites that showed the half open transformer like you described in your earlier post and worked very hard to prove it won’t work.  I found one that instead of a fully isolated transformer, it used an auto transformer design so, at least in my mind, there was a complete circuit.  

    The design I started with had 21 turns total, tapped up 2 1/2 turns for the input, it also had a short coax capacitor .  I didn’t have an iron core toroid, but I have a whole bunch of Amidon type 43 toroids.  First I talked to my RF Engineer cohorts at work about what the advantages of the iron versus ferrite were.  They both said the type 43 probably will work better than iron with potentially lower losses.

    The first one I made, I essentially copying the original design.  It didn’t work for squat.  I took it into work and looked at it on a network analyzer which showed that the capacitor was pushing the whole thing in the wrong direction, making it more reactive instead of less.  I took it home and hooked it to a 20M antenna (468/14.2 + about 1 foot for luck).  The MFJ analyzer said that it was seeing almost 1K resistance and a bunch of reactance on the input.  So I started taking winding off and the whole thing started to behave.  I got down to 10 turns total and no capacitance, 50 ohms, very little reactive component.  The antenna was a bit long so I slowly shortened it, about an inch at a time.  At 14.150 it read a resistive 50 ohms and was just a little higher at the band edges.

    So I haven’t tried transmitting with it yet, but the other night I had it on one antenna input and my 5BTV (with a bunch of radials) home antenna going into my TS-480.  On receive the end fed heard everything the 5BTV heard, sometimes at about the same signal strength, other times about 3dB down.  That is significantly better than the Texas Bug Catcher I had on the camper previously and ran a similar test.  I need to rerun the test with my Flex 1500 which has a fairly actually calibrated S Meter but that required taking a lot more stuff out on the porch for the test.

    So that is where I am at the moment on this.

  11. Ray Matteis says:

    I’m working on a multi-band end fed now. My QTH will not allow the install of any kind of tower or pole due to power lines, but I DO have a tall, 120′ tree about 150; from the house. I found a site on the web that had a graph of wire lengths for bands, and found a 135.5′ wire should give me a good multi-band end fed. I made my own 9:1 unun, and hope to shoot a tennis ball over the tree this weekend. Because of the height, the pulley/weight will be on the house, and not the tree.
    I will write back in two weeks and let you know how it works. I have already tested the unun at work, and it does well across the bands. Once I get the wire up, and the ground rods in place, I should be able to try and make some QSO’s. As I just upgraded from a tech, most of this is all “new” to me. I’ve used 70cm and 2M repeaters for almost all of my radio life.

  12. Jay Taft, K1EHZ says:

    This is an interesting set of posts. There haven’t been any in a few months, but I thought I’d add a comment anyway.

    I’ve been messing with EFHWs for a few years, beginning as others have with AA5TB and W8JI. After a number of successful mono-band EFHWs I decided to try a multi-band trap version to use without a tuner. It started with fairly crude traps to cover 10-15-20-40m in a 30′ long vertical wire for 10-15-20m and a horizontal or sloped extension for 40m It is easily supported with a 31′ telescoping fiberglass pole.

    The trap implementation has evolved into a compact version for QRP and a more robust version that I have been using up to 100watts at home and in portable situations. According to EZNEC models the losses aren’t too bad. Depending on frequency (and accounting for trap, matching unit, and coax feedline losses) the trap version performs pretty much the same as a mono-band EFHW on each frequency according to my implementation in EZNEC.

    As of this writing (4/5/2014) I am operating portable in Washington state (home QTH is New Hampshire). Last evening I worked VE3, Texas and Louisiana with 2.5watts from Fox Island, WA, with the latest version of the trap vertical.

    • Dan KB6NU says:

      This sounds like an interesting design, Jay. Have you posted the design anywhere? Would you like to post it here?

  13. Ray Matteis says:

    Well, I was not able to get the tennis ball as high as I wanted. The tree bark “grabbed” the line, so I was not able to feed it as I wanted. I ended up with an inverted “L”, and cut the length to about 104′. It seems to work on most bands. I did put some counterpoise wires, and it helped the TX quite a bit. With a friends analyzer I was able to tune most bands. I still need an 8’3″ counterpoise segment for 10 meters, but I was able to make a contact from California to Oklahoma with about 100W on voice.

  14. Bill Collins says:

    I have read much about antenna in the past few days, especially the end fed that utilize a 9:1 unun. I would also like to point out I am a newly licensed ham with no real world experience. So far only books and other references (like this forum).

    My predicament is this…I have no trees in my yard, plus I rent. So everything I do first must be approved. The landlord is a pretty nice guy, but I would like to keeping him out of the loop until absolutely necessary. That said, I’m thinking about an end fed in my attic. What if like to do/try is to loop (aka bend) the antenna wire to follow the contour of the roof, and then it the attic window and down the outside of the brick wall. At a specific support I am thinking of connecting an unun (probably purchased), and then send the coax to the shack.

    I’ve heard much starting that it’s OK to bend antenna into several shapes, but before I follow this plan I decided, can anyone tell me if this is a BAD idea? Would it be better to connect the antenna to the unun in the attic instead? Honestly, I came up with the idea to lower the amount of feedline (to reduce signal loss) as well as to increase the length of the element.

    Any comments or concerns would be greatly appreciated!

    73 de Bill

  15. Hi Dan,

    Nice write up on your experience with an end-fed. I’ve used the PAR EF-20 for quite a while and have experienced some RF in the shack so I’ve limited my trusty FT-840 to about half it’s output rating.

    A good friend of mine feels that the only effective common mode choke is to wind coax on a 4″ PVC pipe. So I’m going to run a test to see if this is any more effective than clamp-on RF chokes. We’ll see.

    In the meanwhile, I’ve gotten really interested in 17 meters and ordered an LNR EF-17. The price is a little high, but I want to help LNR and other Ham businesses stick around.

    I’m setting up an inverted L. Since the band is so narrow, I can expect the SWR to stay fairly constant across the band, and I should get a DX favorable radiation pattern over a strictly horizontal configuration at 18′ high.

    Point is, the end-fed solves a lot of problems for Hams in tight situations space-wise. Even though I remember a PAR FAQ regarding common mode being a non-issue, I can say that I experienced it. Maybe me installation was bad, who knows.

    But with limited power, and a choke that works (do they?), especally above 20, I think they are a good, albeit somewhat controversial choice for many Hams.



  16. Jerry Brown says:

    Hi Dan,

    I’m a new General class Ham (thanks to your study guides) and participated in my first field day this year. I joined up with a good friend, EE with 40 years of Ham experience. I decided to bring an end fed antenna to the party.

    I chose the EARC end fed 6-40. This antenna has a 9 to 1 un/un and requires 30 ft of coax feedline. We strung it in the trees of the campsite, fed-end high with a 60 foot RG58 feed, 45 degree of slope with 35 feet of wire. The RG58 was at a 90 degree angle. This antenna tuned to better than 1.2:1 on 15, 20 and 40 meters. At 100w I was able to make phone contacts from all corners of the country plus the Virgin Islands on each band. It performed almost as well as the 20 halfwave dipole my buddy strung. My LDG Z-100Plus autotuner performed very well with this setup.

    At home in my attic it seems to performs as well as my own 20M half wave dipole, anecdotally. I say that because I have no antenna measurements to back that up, just my own experience. It is my choice for digital modes and does not dissapoimt at the lower power levels used in JT65.


  17. Hi,

    I am very interested is this EFHW antenna stuff also, and have built me quite a few based on this guys design: http://pa3hho.wordpress.com/antennes/multiany-band-end-fed-english/
    What I really want to do is, put up a 75 meter EFHW cut for only that band. I’m thinking a 120 ft wire in a V configuration using this guys transformer. I have used it on other bands and it worked out very well.

    • Dan KB6NU says:

      Joel, thanks for the link to PA3HHO’s site. I need to put up an 80m antenna here before the winter, and his 80m/40m antenna might be just the ticket.

Speak Your Mind