Topics

End Fed antennas w/ uBITX #ubitx

 

Not trying to start a "antenna war", but I noticed several people on this forum use "End Fed antennas".(Like Allison KB1GMX).  I have also used them with good success.  I feed them with 33' of coax and a 1:64 unun (PD7MAA design).  I have one with a trap (again PD7MAA blog) doing multiple bands.
On QRZ, everytime a End Fed topic is brought up, there is a bunch of nay sayers that always chime in. What are you experiences with them and QRP?  I'd like to know, being limited by an HOA, for antenna structures.
Thanks,
Jim W0CHL

Nick VK4PP
 

Hi Jim.
I am using a HyEndFed 80-40-20-15-10m at my QTH. I am happy with it, But don't have anything to compare with...

I have built a portable version, same as you (PD7MAA) for use when I go away in 3 weeks time with my uBitx.

Try it, and if the SWR is good and you make contacts, got with it!

Good luck and 73 Nick VK4PLN

Don, ND6T
 

In 1999 I installed an end-fed wire as a temporary antenna to get on the air. An old pine tree stands 120' tall in one corner of the lot. 'Nuthin' more permanent than temporary! Still there, still working the world QRP all bands. 175' of 14 AWG stranded copper fed with a variety of tuners, most often "T" configurations, the poor ground (about 20' of buried wire and two 8' ground rods) is tuned with a series LC network.
Almost as good on individual bands as a high dipole except for noise. Much more noise pickup than my horizontal dipoles at my mountain cabin (over 70' above ground there). The far end of this end-fed is at 100', the near end about 5'. 72, Don

John
 

Jim,
I have an end fed 33' w/9:1 in and I've made contacts all over the US and from Europe to Australia with it from here in central Arkansas.  It is by far my favorite and was also the first one I homebrewed.  It definitely receives better than it transmits, but for me the proof is in the logbook.

73s
John W5JXN 

Doug W
 

I am using a myantennas.com EFHW-8010P (resonant on 80/40/30/20/17/15/12/10m without a tuner) fed with 50 feet of RG8X.  I stick to less than 5w on my v3 µBITX.  Most of my QSO's are digital with some voice mixed in.  I get a chuckle out of people that say end fed antennas do not work because most of them are just repeating something someone told them and have no actual experience.  As for real world usage of operating QRP on an end fed antenna, so far I have confirmed QSO's with 49 US States (still need HI), all continents except Antarctica, and my DX list is growing all the time.
--
www.bitxmap.com

Tom, wb6b
 

I use a 55 foot end fed 80 through 10 meter antenna. It is designed to be non-resonant. The impedance of the antenna is in the 300 to 600 ohm range so I feed it with a 1 to 9 impedance matching unun. The resonate end feds are usually fed with a 1 to 64 impedance unum. It should be noted with end fed antennas there still has to be counterpoise, very often it is the shield of your coax run providing that. So, it is a good idea to add a common mode choke before your coax enters your shack. In my case I wound 10 turns of coax through a large toroid. I, also, provided a good ground at the shield side of the cable for safety and static buildup. I only work 80m for a local net, so the tradeoff in my case is 55 foot of wire at 80m will have reduced efficiency over a full half wave dipole. But, it is invisible running just under the railing of an upper deck. On WSPR my beacon has been picked up as far as Antartica. 

Bo Barry
 

I'm HOA bound too and doing well with the practically invisible MyAntennas EFHW with my ubitx and Icom7300. 
I prefer it for the ubitx since its SWR is flatter than my G5RV (I've hidden in the trees). It needs a tuner for sure.
And actually the EFHW outperforms the G5RV.
If your lot is too small for the 80 meter version, shoot a line over one tree (hoping you have one) and hook each end of the antenna to your two eaves and hung high in the middle by the tree. Inverted slanting V. works great. Good luck Bo w4GHV

Bo Barry
 

John, I had the 9:1 version (but not at a resonant frequency length) and it required a tuner and I wasn't impressed. I've found the 49:1 to work much better. I plan to rewind the 9:1 to make it a true EFHW. Bo w4ghv

John
 

Hey Bo,
Since I'm still laid-up and on sick leave due to knee surgery, I have plenty of time to try out new ideas.  Frankly, I bored out of my mind.  I have extremely good luck with the antenna I've built - had it about a year or so - and used it with good results on SSB, FT8, and WSPR.  That being said, I will look into the 49:1 as mentioned here and maybe it will be the latest and greatest for me as well.  It's the building and trying part of the hobby that I like the most (even though it is sometimes the most frustrating), and I am always looking for another interesting project to try.  

73's
John W5JXN

Warren Allgyer
 

The reason an end fed antenna “works” on multiple bands as indicated by low VSWR is because of high losses introduced by the 9:1 or 49:1 transformer. For non-resonant wires these losses are 90% or more.

The reason such antennas “work” from the standpoint of making contacts is that losing 90% of your power only reduces your signal by a little over one S unit.

The transformer in such antennas does not “match”. It masks the true VSWR. It is a lot like putting whiteout on the X-rays to cure a broken leg.

WA8TOD

Thomas Owens
 

A local ham was disposing of a SK estate and had a EFHW8010 for a great price. I purchased it and some other items from him, hung the EFHW8010 in the trees and was so pleased I ordered a EFHW4010 to take with me for travel, or field use. I figured I'd hang it in the attic for bad weather use. Testing it out with my 7300 (antenna is in my attic mind you), I got the following decode results on FT8 a few weeks ago. I think I'll keep it, thanks :)

ajparent1/KB1GMX
 

WA8TOD,

There are two forms of end fed. 

Non-resonant with 1:9 transformer.  The idea there is to keep the SWR on the cable down to a dull roar.
Must use a counterpoise.  Depending on the transformer used the losses can be low or very high.  Also
wire length and height makes a big difference in presented impedance.

Resonant, these use a transformer to match the feedpoint impedance typically in the 1800-4000 ohms range.
Most relay on the coax as a return path.   The transformer matched types can operate on harmonically related bands
as a 40M half wave (about 66ft) is a fullwave on 20 and 2 full waves on 10M. With a strategic located tails or chokes
other bands like 15M and such are possible.   The trick is to make it look like a resonant piece of wire at some
multiple of a halfwave.  If the wire is resonant the transformer is a matching device.  If the devices are well designed
the losses are small.

If I were loosing 90% of the power the matching transformer would get very hot at 100W.

Allison

ajparent1/KB1GMX
 

Bo,

The 1:9 designs (not 9:1) can be very good or terrible.  Depends on the wire length the longer 53ft and up behaves better.
Also you do need a counterpoise or ground radial system.  I've used one of about 89Ft as a inverted L with ground plane
and for 160, 75, and 40M it was a very good antenna save for as a vertical and its ability to hear all noise.

If you go with 1:49, the wire must be resosnant as in a half/full/1.5/2 wave lengths and present Rx of
about 2500ohms with Jx of near zero (resonant).

Think of it as the ultimate off center fed Dipole.

Allison

Jacob Farnes
 

I took an opportunity to install a long piece of random wire on the roof of my QTH. Didn't measure the physical length with the limited time I had but later measurements put it at under 130 feet. (The original speaker wire was cut at 66-feet, the the pairs split and soldered. Not an ideal length but what I had on hand for a one-time opportunity. I may have clipped a few feet from it and forgot while it was in my portable kit.) Tucked it down under the tar shingles at the Ridgeline.

Brought one end to a junction box and used it for listening mostly. Longest wire i ever have had in the air. Used some Iron tripwire for four 40-foot groundplane wires as I needed to be as invisible as possible.

Many months later I got a RigExpert AA30.Zero to analyze my whole system as the T-match tuning wasn't making sense to my brain. Was averaging 200-ohms on each band at 40m and higher. So wound a 4:1 balun on a T152-2 core.  That dropped 40m-10m to 25-75 ohms plus reactance.

So I use a simple VK6 L-match with a T106-6 core on a 14-position switch (military surplus), a 365pF air-variable capacitor, and  some silver mica capacitors on SPST switches to add more capacitance as needed. It's not needed on 40m as that was 50-75 ohms.

As for actual antenna performance, I seem to do very well with NVIS on 80m & 40m. On days 40m goes long, I get into Texas just fine.

On the local HF net (from Boise to Pocatello to Idaho Falls & Rexburg, roughly 325 miles) I'm heard very well at 10W on the Bitx40 & uBitx and at 100W with the IC-718, even if local noise sources make it very hard to hear net control.

That's my end-fed I'm using. Kicks the pants off the tripwire dipole that was 18:1 SWR I was using before that as a cloud burner at this QTH

73 de Jacob AG7CT

ajparent1/KB1GMX
 

Jim,

The vertical for 2M on the truck is a halfwave end fed.  There are marine units for fiberglass boats.

I use the part EF40/02/10 and EF-QUAD (40/20/15/10) here.  I keep a EF40/20/10 in the antenna
bag in the truck a long with a vertical kit (poor girls pac12) and a assortment of cordage for
suspending things.  

A 1:49 transformer is likely adequate.  The end impedance of a resonant half wave follow the
same pattern as a center fed half wave dipole.  It varies with height and proximity to objects.

I also have a few monoband end feds that use a L-network for matching and are also very good.
I use them for 40, 20 and 10M.

Also I tend to keep clear of harmonically related coax lengths like 33ft/66ft  (I use 25, or 40ft) for
portable work however FD setups lengths like 75 and 100FT are used.


The big argument is and has never been do they work.  Its been about the return currents
on the shield of the coax.  That varies with the length and position where its measured.
Generally its low assuming the antenna is actually resonant.  If its not then it behave more
like a transformer fed random wire with much (significantly) higher shield currents.

My experience is every time I encountered a case locally of someone having RF issues
with an end fed.  The SWR was terrible and they were trying to fix it with the tuner
(rigs or external) and the wire was just plain wrong length.  Worst case was a 40M
monoband resonant at 6.1mhz.  The wire was low, inverted V with the middle maybe
25ft with the ends not more than 6ft up and the wire measured 68ft.  Another was
actually the EARCHI 1:9 53ft a very different beast (but its endfed!)...  Usually the
prior install was a dipole or G5RV with the same issues.

Given a bit of effort and care many antenna work well or are at least serviceable.
Things like RFI and hot mics are usually the result of installation related errors
or the entenna being just too close.

Allison

Warren Allgyer
 

Allison

i agree with most of what you said. 

I have made extensive tests of the most commonly used 49:1 transformer as well as VNA analyses of both random and resonant end fed wires. I have not tested 9:1 transformers. 

90% and higher losses occur using the 49:1 transformer on non-resonant wires. If the wire is resonant the losses drop to on the order of 20%. In both cases the measured VSWR is severely and favorably distorted by the transformer losses.

What you you say about harmonically related bands is true in theory but not useful for the directly related HF bands because of the shortening effect of multiple half waves in series. This causes the second harmonic resonant point of, for example, an EFHW wire resonant at 7.1 MHz, to be well
above the 20 meter band. The wire therefore when used on 20 meters will be operating in the non-resonant mode with resulting 90% or more losses.

The EF wire itself, both in resonant and non-resonant configurations is an excellent radiator. The challenge is to deliver power to it effectively. This can be done very well if the wire is resonant with a transformer. It can also be done very well across a range of bands if a tuner is used rather than a transformer.

If  you put 100 watts into a non-resonant antenna fed by the popular  49:1 FT-250-43 toroid transformer, key down for one minute, the transformer will be too hot to hold. If you do it at 500 watts the core will explode. On the other hand, used with a resonant wire and on bands 40 meters and above, the transformer will be Weston but not alarmingly so. 

Rob Snow
 

I have nothing to compare it to, so take this for what it's worth...nothing :-)

My uBITX is tied to a MyAntennas EFHW8010 setup in an inverted L, 50' up and 70+ across.  I've got ~1200 (1050 confirmed) QSOs with it since March on all continents sans Antarctica (did get a WSPR from there, though) and my log says 49 states (RI) and plenty of islands.

Is it any good?  IDK, I don't have another rig or antenna to compare it to.  Am I pleased with it?  Very much so, it's my first and it hasn't disappointed.

I think I'll build my next one, seems it's just a 49:1 and a cap in the matchbox, seems like something I can knock out and learn even more with.

73 Rob de AG5OV

Kunjani Ol
 

I have tried many wire antennas, limited by the small space available.
About five years ago I tried an inverted L with a 9:1 unun and never looked back.
I use it for 160-17, then a modified CB vertical for 15-10.

I have the vertical section as far from the house as possible, which reduces noise appreciably.

The vertical section is supported be a 25 ft fibre glass pole, then it goes over the roof (touching the roof) and tied to a shrub in front of the house.

I have never used a counterpoise/radials, instead I purposely use the coax. I have 6 clip on ferrites which I move along the RG213 until I get the best SWR on all bands. It is easily tuned by the internal tuners on all rigs I have used on it (Yaesu and Icom)

I have recently made a 9:1 balun, and installed an earth rod at the feed point, and taken the ferrites off the coax (they no longer have any effect)
I can't detect any difference.

I don't think it can be bettered considering ease of building, simple supports, small size, ease of use and performance for price.

The transformer will get warm on high power on some bands, if you see the SWR drifting, you know things are getting warm.

Jerry Gaffke
 

Warren,

You are correct, a poorly designed EFHW can have 90% losses.  Or worse.
But properly built, it can work well across multiple bands with very little loss.
Check out this QST review of the myantennas.com EFHW8010, it is resonant on 80,40,20,15, and 10m. 
    http://www.arrl.org/files/file/ProductReviewsForDeb/2016/pr032016.pdf

At the end of the review, they mention hooking two of these matchboxes back-to-back,
and finding there was virtually zero loss.

Here's a facebook page of experimenters playing with EFHW's:
    https://www.facebook.com/groups/EndFedHalfWaveAntennas/
Getting one to work well across multiple bands is not quite as trivial as one might think.
Just building a high powered transformer that works well from 3.5mhz to 30mhz is tough.

Jerry



On Tue, Jul 24, 2018 at 08:07 PM, Warren Allgyer wrote:
Allison

i agree with most of what you said. 

I have made extensive tests of the most commonly used 49:1 transformer as well as VNA analyses of both random and resonant end fed wires. I have not tested 9:1 transformers. 

90% and higher losses occur using the 49:1 transformer on non-resonant wires. If the wire is resonant the losses drop to on the order of 20%. In both cases the measured VSWR is severely and favorably distorted by the transformer losses.

What you you say about harmonically related bands is true in theory but not useful for the directly related HF bands because of the shortening effect of multiple half waves in series. This causes the second harmonic resonant point of, for example, an EFHW wire resonant at 7.1 MHz, to be well
above the 20 meter band. The wire therefore when used on 20 meters will be operating in the non-resonant mode with resulting 90% or more losses.

The EF wire itself, both in resonant and non-resonant configurations is an excellent radiator. The challenge is to deliver power to it effectively. This can be done very well if the wire is resonant with a transformer. It can also be done very well across a range of bands if a tuner is used rather than a transformer.

If  you put 100 watts into a non-resonant antenna fed by the popular  49:1 FT-250-43 toroid transformer, key down for one minute, the transformer will be too hot to hold. If you do it at 500 watts the core will explode. On the other hand, used with a resonant wire and on bands 40 meters and above, the transformer will be Weston but not alarmingly so. 

Warren Allgyer
 

Jerry

i am aware of that product review and I have have discussed the methodology with the author. Their methodology was deeply flawed and cannot be duplicated in a properly equipped test laboratory. 

The antenna is not the issue. The transformer is. When feeding a resistive resonant load the losses in the transformer are under 2 dB for bands 40 - 17, rising above 3 dB or 50% on the rest of the HF bands. 

When feedingING a non-resonant reactive load the transformer loss rises to 10 dB and more on all bands. I am happy to supply test data and methodology, which has been independently corroborated, to any who wish to see it. I am good in QRZ. 

WA8TOD