ENDFED/EFHW Antenna & Choke

My 8010 Enfed antenna Configuration

My 1:49 Transformer is located in the Attic , just under a tile and obviously the wire is routed outside and hook on a 12 meters fiberglass pole in the garden.

With this antenna the coverage is 80,40,20,15 and 10 meter band without any antenna tuner and the average SWR is below 1.2 on phone bands.

The total antenna lenght is about 23 meters , with one 20.4 meters long segment from the 1:49 transformer to the 110uh coil and about 2.2 meters long segment from the coil to the insulator.

I am using this antenna for almost 2 years and since i have limited space with my tiny garden. The performance are acceptable below 20m (14Mhz)due to high losses from the transformer during transmit for frequencies above 14Mhz. This antenna using this kind of transformer is a compromise.

DO NOT EXPECT FOR HIGH EFFICIENCY FROM 3Mhz up to 30Mhz ONLY WITH A SINGLE 1:49 or 1:64 TRANSFORMER.

More details here : https://squashpractice.com/2021/06/23/performance-of-491-ferrite-core-transformers/

For better efficiency , the FUCHS Transformer is by far a better choice : https://f5npv.wordpress.com/remote-automatic-atu/

The Transformer is located inside the Attic just underneath a tile and only about 10cm of wire still inside the attic. This configuration is provided a great SWR in all bands and only on 80m the bandwidth is really narrow with about 40Khz.

Since the transformer is inside the Attic , it is well protected against the rain and humidity

WARC BAND ENDFED (30,17 and 12M)

Based on PA3DWC design https://gerritvinkies.wixsite.com/pa3dwc/post/end-fed-antenna-warc-band

The Coil is 10.5uH –>

The SWR is almost perfect on 30/17/ and 12m

Transformer design

Regarding the Transformer build i am using the following Design

I using three 2 x T240-43 + One FT240-52 stacked together in order to provide a better linearity, energy transfer and minimize the loss.

Stacking cores increases inductance and winding capacitance linearly. Stacking core is really a must especially if you are using 80 and 40m bands.

With this configuration , the transformer is a able to cope with 500watts SSB and about 200watts CW/DIGIMODES.

You can also use One FT240-52 and two FT240-43 mix together, and this is what i did.

Transformer Check

In any case you shall control the transformer prior to the final installation with the antenna. Antenna impedance for this antenna is roughly around 2500 Ohms and the transformer is providing the adjustment from the 50Ohm from your TRX to the a ENDFED monopole which is about 2500 Ohm from 3 up to 30Mhz.

If you do not check the transformer there is a high risk your antenna will provide you a good SWR but the antenna will be not resonnant and the Current/Voltage repartion along the antenna will provide a poor efficiency and radiation pattern.

How to check ?

For that you will need an antenna analyzer likely the NANO VNA or the SARK100

You will need also a 2400 up to 2500 ohm non-inductive resistor connected to the output of the transformer and tap to the ground.

The result should be :

If not , you will need to check the transformer and with the analyzer to check the capacitance and inductance provided.

Too much capacitance —> Your transformer primary is not correct and check again. You can also add less of more space between the turns . My experience is adding the minimum space to between the primary turns is providing great results

Too much inductance –> The transformer secondary is not correct and double check again . My experience is to put more space between the turn at the secondary of the transformer.

Just to provide you an example , my 15m long antenna (Please refer to below) , my antenna is providing me on 15m band an almost perfect SWR . I should be happy but unfortunately this band is not resonnant at all and the performance are deeply poor , even catastrophic The antenna analyzer is providing completely upside down R , with a Inductance/capacitance thanks to the mathematics and complex number a great SWR.

As an example this is the good current dispatch on the 15m long with 2 loading coils antenna:

When you manage to acheive a close to perfect SWR on the frequencies you are targetting , the transformer is ready to be installed.

a SWR from perfect to 1:1.3 is acceptable , above you will need to correct the issue upon a great SWR is acheived.

110uh Loading coil

The 80 meters coil is a 110uh with 170 windings of 0,5 mm of enamelled wire on a 19mm PVC pipe.

This Coil is providing the 80m band and i am tuned on 3640Khz with about 40Khz bandwidth.

SWR

Antenna performance

So far so good , since i have limited space available , this antenna is perfect for a tiny gargen or for portable

On 80m dont expect too much therefore this morning i can receive very well WSPR stations from US.

On 40 and 20m bands , the antenna is acceptable

On 15 and 10m bands the performance are VERY poor (During transmit , the Transformer is providing significant looses up to 3db)

40m band this morning TX

PSKreporter status after about 12Hours FT8 skimming. Not bad for a simple piece of wire and a transformer

24H FT8 Skimming

20w with EFHW Antenna

49:1 Transformer investigation

After using my 23m long antenna for almost 2 years as mentioned above i notice the following performances during FT8 skimming , WSPR and also with QSO using SSB , CW and Digital (Mostly PSK31)

80m : Performances are average . For QSO withing West europe is just piece of cake , for outside Europe, mostly i am using CW,PSK and CW

40m : Peformance average ++ . i do not have any problem for QSO all Europe and DX .

20m : Compare to my 2el beam obviously this antenna cannot compete , therefore when the propagation is open DX is quit easy.

15m and 10m : perfomance are deeply poor , there is a huge difference in terms of performance with the lower bands. The Transformer is showing significant losses during transmit. For receive , the performance are acceptable.

For my investigation i am using 2 transformers back to back connected to a dummy load in order to measure performances and losses. Results are without any doubt and provide a lot of information regarding the antenna performances i notice.

Prior to the transmit testing , i double check both transformers connected together with the output in a dummy load , the SWR for all bands , which is close to perfect and demonstrate Both transformers are providing a 49:1 ratio without any inductance or capacitance noticed.

Overall SWR Curve with back to back transformers and dummy load

Losses investigation

This test was performed on 80,40,20,15 and 10m bands with 10 up to 30w power input. This test is combining Transmission losses + Insertion losses + Heat losses

Results with 30w maximum

Since i am using strickly identical transformers , the P output is divided by 2.

Needless to say with the low power i was using during the test , the Heat losses were negligeable . When using about 100 up to 200w , Heat losses stated to occur and on from 20 up to 10m band overall losses were average 2dB with a peak at 3dB on 10m band

Results with 300w

i perform some test using FT240-43 and 52 toroids with roughly the same results. Whatever 1:64 ort 1:49 i have pretty much the same results.

Tbhe following results are combining Transmission losses + Insertion losses + Heat losses

When combining Transmission losses + Insertion losses + Heat losses , results are demonstrating one thing : a standard 1/49 or 1/64 toroid transformer is not able to provide a reliable efficiency from 3 up to 30Mhz.

Insertion losses

I did the same testing using a NANO VNA and the results were also significant considering This test is taking i nto account only Insertion losses. The average results was already about 0.5dB from 3 up to 30Mhz. Means if you are QRP with 5w output from your TRX , you loose already 1watt only with the insertion losses . When transmission losses will occur you will have fro sure about at least average 0.8 up to 1dB losses. Means your 5W from your favorite QRP transceiver will be only 3w. With Heat, your losses will be for sure above 1dB average.

The NANO VNA is able to provide losses information with a decent accuracy and whatever the transformer i am using (FT240-52 / FT240-43 , 1/49, 1/64) the measurement are showing significant insertion losses . The main extent is capacitance and inductance variation in the primary or secondary of the transformer will imply losses. The NANOVNA measurement are demonstrating that even without any heat , the losses are significant . Heat and transmitting RF is simply make the losses even worse.

Factually what is happening . FT240-43 core has usually a high resistivity and low core losses at high frequencies (Means the HF band) . Therefore this high resistivity and low core losses are not for the entire HF band means from 1Mhz up to 30Mhz. This kind of core will be the most efficient around the 7Mhz band considering the permeability and phase shifting which is almost null at 7mHz . Above this is not the same story and more you go higher on frequency means above 7Mhz , more inductance and capacitance leakage you will have. I addition a phase shifting will occur between the secondary and the primary of the transformer. This capacitance/inductance strays and phase shifting will impact directly the energy transfer from the primary to the secondary. The winding capacitance is interacting with the leakage inductance –> Losses will happen at this stage and even without any heat due to high power from the transmitter. Adding a 100pf capacitor will assist regarding the stray inductance and will provide a great SWR . Yes a great SWR but with some inductance and capacitance stray and leaks —> This will provide losses for sure .

Despite the fact the measurement with antenna analyzer is providing good results in terms of SWR , The capacitance and inductance on 15m and 10m bands are not acceptable (Thats why the antenna is showing a good SWR) , the losses are about 30 up 50% when applying some QRO Power. Those results are confirming the antenna and my transformers are not performing well at all for those bands. On the other hand , the losses for 20m and below are very promising and despite the fact the antenna is not the best performer in particular for 80m band (The antenna is only 23m long) , it is acceptable for this band and more than acceptable for 40m and 20m bands.

Multiband capabilities :

Using a single FT240-52 or 43 49:1 transformer will be difficult to achieve multiband capabilities for transmit. The losses for frequencies above 20m are significant. Since there is no chance at all this kind of transformer will perform well from 80m up to 10m, the antenna works “equally badly” on all bands at once.

For SWL or Receive only , the performance are acceptable

Despite the fact this antenna is providing great SWRs , it does not mean the antenna will perform well . A good energy transfer is not granted even your SWR is perfect when using ferrite core. A SWR meter gives you a reading (2:1) or even a good 1:1 SWR but gives no information as to what you can do to improve your antenna.

How many times i read on different forum : My antenna has a 1:1 SWR and works very well . Are you sure ? Did you check the reactance , capacitance and inductance strays ? The Smith Chart will also provide some relevant information regarding your antenna performance. Transformer with capacitance or inductance stray and leakage will have a lot of losses and needless to say the radiation pattern will or may be also affected. When running 1Kw from your ACOM or OM POWER Amplifier in a poor perfomances antenna is like putting dust under the carpet!

Be aware that if our antenna is providing a great SWR but with significant capacitance/inductance strays –> There is a good chance you will have some RF strays everywhere with RF Feedback in your shack and spurious for your TV Sets .

When using a ferrite ring, the inductance at the secondary of the transformer rises very quickly with the number of turns and to achieve good performance on the all HF bands you will need 2 different transformers. It is impossible to simultaneously obtain a PERFORMING 1: 49-1 and a 64-1 transformer suitable for seven HF bands at once. 

The Fuchs ATU is a great alternative and will provide a really high-efficiency transformer (https://f5npv.wordpress.com/remote-automatic-atu/)

Why using a Common Mode Choke

Common mode chokes are added as series elements to a transmission line to kill
common mode current. The line may be a short one carrying audio or control signals between a computer and a radio, video between a computer and a monitor, noisy power wiring, or feedlines for antennas. This application note focuses on the use of chokes on the feedlines of high power transmitting antennas to suppress received noise, to minimize RF in the shack (and a neighbor’s living room and television sets) and to minimize crosstalk between stations in multi-transmitter environments.

I am located nearby 400.000 Volts power line and the noise on 80m without the CMC is about S9 (-69dBm) and the addition of the CMC with a proper grounding system is decreasing the noise level significantly. Nowadays with the CMC my noise level on 80m band is below S6 (-90dBm))

Enfed or EFHW antenna are usually generated by design some Common Mode Currents on your feeder and using CMC-Common Mode Choke/Line isolator will assist you to reduce and even supress spurious and EMI. It will also help you receive by filtering induced noise from various sources.

Consider a simple dipole fed with coax. In the common mode circuit, the coax shield becomes part of the antenna, acting as a single wire connected between one side of the center of the antenna and ground. As a common mode circuit element, its VF is near 0.98 (depending on the diameter of the shield and the dielectric property of the outer jacket). In the common mode circuit, this wire (the coax) has some impedance, (RS +
jXS), by virtue of its electrical length, which is different at every frequency. At some frequencies, XS will be positive (inductive), at others it will be negative (capacitive).

Without a choke at the feedpoint, the feedline becomes part of the antenna; if the antenna system, including the feedline, is unbalanced, this causes the feedline to radiate part of transmitted power; when receiving, signal and noise picked up by the feedline is
coupled to the antenna. This is most easily understood with coax, where skin effect and proximity effect combine to cause common mode current to flow on the outside of the shield and differential mode current to flow on the outside of the center conductor and return on the inside of the shield.

In oder to achieve a CMC , few solution are commoly used and the one i am using is using Toroids or Fair-Rite type : 43, 31 , 52 and 61.

According the table below , it is noted the toroid you should use in order to achieve an efficient CMC device for your EFHW or Enfed Antennae.

According this table , using only one choke will not cover the entire HF and a combination of Toroids should be used.

Common mode current is carried on the outside of the coax shield, or as the difference of unequal currents on the two conductors of 2-wire line. A line carrying common mode current acts as antenna for both transmit and receive. Common mode current that couples to the antenna changes the directional pattern of an antenna by filling in the nulls of it’s directional pattern. In simple terms, the common mode circuit becomes part of the antenna – the part of it that is close to noise sources picks up that noise; and when transmitting, radiates RF that poorly designed equipment will hear as RF interference.

Because a Fair-Rite choke can be inductive or capacitive, and because the common mode circuit will be inductive at some frequencies and capacitive at others, the
choke can cancel part or all of the of the common mode circuit

For an antenna from 160m up to 10m bands , the CMC shall be located about 4 meters from the 1:49 or 1:64 transformer.

DESIGN

In general, any combination of chokes can be used in series to provide the
desired choking impedance over the desired bandwidth. Their combined choking impedance, RS, will be the sum of their RS values on each band.

According the table underneath , the desired and highlighted in red will be a good choice in order to achieve a great CMC for your EFHW antenna.

It is highly recommended to use RG400 for the choke winding , therefore according some experimentation , RG58C is not a bad choice since RG58 has less losses on the HF band. If you are using RG58 , do not tight it too much on the Toroid because the of radius bends limitation for RG58 and to prevent any short circuit with the center conductor and the shield.

RS value of 5,000 Ω is the Minimum you should target

According the Table i am using : 17T on FT240-43, 10T on FT240-31 and 11T on 2xFT240-52 with a PVC pipe.

This Combination is providing me a RS total average value of 6,000 Ω from 160m up to 10m which is excellent.

This design is able to handle up to 1kw with ease and the separation between the coils is not really critical.

The Choke shall be located 4m from the feeding of the transformer. Why 4 meter , simply because the 4 meters braid (shield) is acting as a counterpoise according 1/20 wavelength which is the minimum lenght regarding the lowest frequency band (80 x 0.05 = 4 meters).

Alternative Combination

Another relevant combination : Cores are (3) 31 mix with 15 turns on FT230-31 , 12 turns on FT230-31, another 12 turns on FT230-31 and (1) 43 mix with 5 turns on FT230-43 for 20m.

CMC Choke Summary

I do have any RF feedback issue even using QRO amplifier (Ma maximum power is about 300w) , thererfore since EFHW or ENFED antennae are subject to provide interference , it is mostly a safeguard system .

This CMC did not provide significant losses (not really measurable and insertion losses are negligeable) and the SWR remains very good.

The CMC is providing a high impedance circuit for the coaxial shield and will prevent spurious and RFI to come back inside your receiver and acting as a CMC trap. The coaxial center conductor will not be affected.

Using this kind of CMC is providing the ability to add inline a very efficient trap for Common Current Mode ”AKA CMC” since this kind of stacked CMC is performing using the toroids frequencies where they are the most efficient. Using only one toroid is fine cannot be efficient from 160m up to 10m on its own. Using different toroid mix is more efficient. Bear in mind , a unique toroid mix will be efficient for only one frequency range used as a CMC . This inline CMC stacked mixing toroids is in my opinion also a great choice.

CMC installation

How far from the 1:9 Transformer the CMC should be installed ? Ditto according my knowledge , the CMC acting also as a counterpoise should installed 1/4 wave from the transformer. Thus Choke shall be located 4m from the feeding of the transformer. Why 4 meter , simply because the 4 meters braid (shield) is acting as a counterpoise according 1/20 wavelength which is the minimum length regarding the lowest frequency band (80 x 0.05 = 4 meters).

CMC Efficiency

This is a quick video regarding CMC (Common Modes Choke balun) for assisting regarding spurious during receive.

My neighbourhood has solar panels everywhere and maybe other systems disturbing my receive capabilities on 40m band. The solar panels in my neighbourhood are used for external lighting system .

During day time until sunset , my 40m band is not usable and sometimes 20m band (less spurious but still) are full of spurious. With or without the inline CMC it is simply day and night. Initially i did some tests using only a FT240-31 CMC which was efficient but the inline choke stacking few different core mix is performing simply better especially regarding the serious spurious i have daytime on 40m band.

This choke is simply assisting me to provide me very decent noise floor level in particular on 40m and 20m bands. I did also some investigation regarding the Shack grounding system and whatever the gounding is separated or common it is not assisting a lot regarding spurious.

This test was performed with my Long Wire antenna which is maybe the worst scenario since this a kind of OCF antenna and was even worst with a G5RV antenna (On my G5RV the noise floor was simply terrible on almost all bands without the choke during daytime)

CMC efficiency measurement

To be honest i was puzzle regarding all method and discussion regarding CMC measurement. I just prefer to let the “Experts” to argue each other and just introduce a method using a NANOVNA . Obiously since i am not a seller or do not have any intention to reach certification level , the following methodology will just provide a good idea with figure regarding the choking impedance and attenuation a choke may provide .

What we are looking for , generally speaking for a ENFED antenna , the good figures are:

-About 25dB CMC attenuation

-About 4K to 5K impedance choking.

In order to do so we can use S21 from the VNA with a little circuit in order to perform some interesting measurements .

S21 indicates how much signal is coming into port 2 (CH1) from port 1 (CH0), hence the name S21. With this we can e.g. measure the transmission characteristic of a filter. Not only the amplitude of the transmitted signal is measured and thus the degree of damping in dB is determined, but also the phase so that capacitive or inductive behavior becomes visible.

To perform measurement you will need to use the shield of the coaxial cable or CMC since this the measure we are looking for.

we need also to ensure a proper load for both VNA channels (about 50 ohms for instance) and the VNA shall be calibrated.

To ensure a proper load in parallel i am using a 47 ohms resistance both side S11 port 1 and S21 port 2 since i do not have 50 ohms one.

Hereafter the pictures of the measurement arrangement using a NANOVNA.

I am using two 47 ohms resistors , 2 SMA chassis plugs and a piece of PCB in order to achieve a tiny interface between the CMC and the VNA Port 1 and 2. i did some test without the 47 Ohm load resistors but the results were almost identical with minor differences.

Considering that the nanoVNA 3.2 doesn’t exactly have a 50 ohm input resistance on port 1

Just connect the two SMA sockets on the Input side of the board labeled “VNA”, to the ports on your VNA using SMA jumper cables. Port 1 to Port 1, Port 2 to Port 2.

Once the Test Rig is connected to the VNA, and the VNA is calibrated, its time to measure some chokes!
-Connect the Common Mode Current Choke to the appropriate output from the interface on the CMC braid/shield.
-Measure “S21 LOGMAG” with your VNA. We want it to be as large of a negative number as possible, across as much of the frequency range of interest as possible.

The VNA is generating a signal on Port 1 of a known amplitude and phase, and measuring the relative amplitude and phase on Port 2 on the center pin of the input to the shield of the output with a signal through the Common Mode Current Choke, then route the result back to the center pin of Port 2 of the VNA. So now the signal being measured by the VNA has been attenuated by the CMC, and will show up as a negative S21 LOGMAG value.

The larger a negative value for S21 LOGMAG, the more the CMCC is attenuating the common mode currents, the better.

For a receiver -25db is about 4 S-units (6dB per S-unit) . According the video from above , you can notice the attenuation is more than 25dB and according the measurement i had performed on the CMC i am using the measured attenuation is almost 40dB means it is about 6 S-Units from the S-Meter . On the video from above without the choke the noise level was about S-units 59 and with the choke the S-units was about S2 / S3. Definitly i am very close in real life with the measurements i had performed.

The most interesting is to use different Core Mix and generally speaking the following mix are providing the best results : 31, 61, 43 and 52.

Hereafter you will find a complete methodology regarding CMC attenuation measurement using a NANOVNA and NANOVNA SAVER (Special thanks to W0LEV for the sharing).

https://drive.google.com/file/d/1EmeiP-ANp4IlJhQgcvFroTwYebpgjY1T/view?usp=sharing

A video regarding the NANOVNA Calibration :

CMC Attenuation results

Hereafter the result for a FT240-31 only with 17 turns RG58 , the Best attenuation is about -37dB at 3Mhz down to -25dB at 30Mhz

Hereafter the attenuation for a FT240-43 only with 12 turns RG58. a bit less efficient below 40m band and the maximum attenuation is about -40dB for 30m band down to 28dB at 30Mhz.

Hereafter the attenuation for a stacked inline choke using FT240-43,31 and 52 mix. The attenuation is reaching -51dB for 80m band , about -40dB for 40m and down to 32dB at 30Mhz. By far this combination is providing the best results . Therefore we can reasonnably assume a inline choke using a FT240-43 + FT240-31 will provide also great result. Obviously this configuration using 3 Toroids mix is overkilling , therefore why not ?

COUNTERPOISE

I Cannot elaborate a lot regarding counterpoise for this kind of antenna . You will find a lot of information on internet saying a counterpoise is necessary or not. My antenna knowledge is driving me at the early stage when testing the antenna : i need a counterpoise in order to achieve a good radiation pattern and efficiency .

I made extensive tests with and without counterpoise. From a single counterpoise wire to multiple wires using also different length i did not noticed any difference or any increase of the efficiency. All counterpoise are now removed and i use simply the coaxial shield with the CMC as a counterpoise and the antenna is performing decently with this configuration.

As stated above , all tests i performed with a counterpoise are conducted regarding my environment and the antenna configuration . The results may differ according different environment and antenna configuration . it must be tested in due course according yours.

Regarding the noise floor level or CMC , with or without a counterpoise wire does not make any difference.

Installation pictures

Since the mast is flexible , the wire is not tight too much.

For the guying i am using chopping boards. Efficiency and tough

The CMC nest is located 4m from the transformer

For the guys i am using chopping board

100uh coil

Alternative Design: 15m long antenna

Since i like to experiment , i just acheive the following ENFED antenna. This antenna is only 15m long using 2 loading coils.

This antenna is working for 80,40,20,10 and 6m bands

The total lenght is about 15mtr only and using 2 coils and a little stub(s) on the 80m coil or/and the 40m coil

The 2 coils are 35uh and 90uh

The overall lenght is only 15m

Using 2 coils in order to reduce the size of the antenna is great thereofre the bandwidth is really narrow on the 80m band (about 60khz)

35uh Loading Coil

Using a 19mm pvc tube you will need 80 turns of 0,75 mm winding wire

90uh Loading Coil

Using a 19mm pvc tube you will need 160 turns of 0,5 mm winding wire

Alternative configuration

Hereafter two configuration

One is using only one Stub at the output of the 90uh coil

One is using 2 stubs : one at the output of the 35uh coil and at the output of the 90uh coil, in this one you will notice the wire lenght between the 2 coils is only 2.7m . Both configurations are working very well

Stub

Adding a stub on the output of the 90uh coil is providing fine tuning features which is very convinient.

The stub lenght is about 10 up to 20cm long using the same wire you are using for your antenna. you just need to cut to adjust the frequency you are targetting

Performance

Compare to my 23m ENFED antenna i cannot see any difference for the receive signal and regarding the TX , it is also pretty much the same performance. The only difference i notice is regarding the bandwidth which is narrow compare to the 23m endfed antenna.

TX Performances: with this antenna lenght , the losses are significant compare to the 23m long one and was confirmed according some WSPR testing.

I will this antenna lenght remains a great choice for SWL purpose , for transmit keep focus on the 41m or 23m long using a coil to access the 80m band.

Check

Performance on 20, 40 and 80m

On 80m , 20w, FT8 : about -15dB USA East Coast

Receive 80m

On 40m , 20w, FT8: about -9dB in US East Coast

-12db in the East Coast USA

-3db in New Zealand

Transmit on 20m band

Receive on 20m band

Considering this antenna with only 15m long , the results are not so bad .

Antenna Installation (15mtr long antenna)

This Antenna is dedicated to my parent place in the south of France and i am planning to test it during 1 or 2 months before to put back the 23m one.

Stub Way-Ahead testing

For testing and experimentation purpose i will test some different shape Decoupling stubs on the Coils.

The Stubs will be like in the following picture below:

Currently for testing pupose both stub lenght is 15cm folded (10 + 5cm). Results are promising and make the antenna easier to tune

Decoupling stubs

– Current block role all along the antenna . It makes the loading coils more efficient. The wire leg tailored for your frequency will radiate more efficiently. The antenna radiation pattern will have a better shape . Your antenna is more efficient. The end wires (So the output of the Coil) presenting a high impedance due to the coil (your coil has a very high impedance according the frequency is tailored for) connected to a lower impedance. Regarding this mistmatch means a high impedance connected to a lower impedance , the Stub will act as a choke and will assist the coil regarding the current dispatch at the end of the wire and coil regarding the frequency they suppose to work. For multiband antenna using diffent frequency , the antenna will need to harmonise the current all along the antenna taking into account different frequencies and obviously antenna lenght. The different antenna lenght used by your antenna will be used more efficiently.

Bear in mind that sharing different impedance at different frequency (High and low) connected together may cause power loss and deteriorate the radiation pattern. A decoupling system will miminize such power loss especially to assist the coil which is connected to a different impedance . We can consider as a matching system

The 15m lenght antenna is using 2 coils and resonnated on 4 frequencies.

The element length is relative to the frequency of appearing resonance each other.
The shorter element length is, the higher resonance frequency is. There are few current paths: monopole and another monopole especially in our case for the 40 and 80m bands which are using the 2 coils using 2 elements or leg of the antenna. one more resonance appears at the lower frequency (40m and the first coil )than monopoles. Then, both real and imaginary parts for the 40m band will need to handle two resonances. Like this, you can consider you are connecting 2 antenna together and the decoupling stub will assist to provide a clean current distribution between the two antenna section (40m and the 20/10m section). The connection between the antenna sections resonnating on different frequencies may involve power loss due to resistance and radiation efficiency decrease. Adding a stub especially at the output of each coil will ”trick” the antenna and the current distribution in order to consider the end of the stub is the end of the antenna section for the frequency in use. The end of the path for each section will have to output : the Stub and the other section of the antenna . Thanks to Kirchoff Law , the current will be driven to the stub and will provide a clean cut or choke.

In order to verify my assumptions , i install the stubs at the input of the coil and it was simply a disaster and the antenna was not tune anymore.

Obviously this is therory and testing and make measurement with a VNA , Grip Dip and a hertz loop will provide maybe some evidence of the efficiency of this system. This is oviously experimentation but so far the results are not so bad and i will continue this way in order to make this stub system efficient if i can (Nothing is granted when experimenting)

Way ahead experimentation

According this website is will test during the next few weeks a random wire configuration.

You can find is the following website some interesting information : http://www.k7mem.com/Ant_End_Fed.html#Not_Random_Multi

Final Summary

-23m antenna using 1 loading Coil (110uh):

-The antenna performance are average to good on frequencies below the 20m band. Above the performance are deeply poor, On 10m band , the SWR is low but transmitting losses are amost 50% (the 1/64 transformer is providing sligthly better results in terms of losses for frequencies above 20Mhz)

-SWR magnitude is excellent using a stacked Transformer especially with the FT240-52 Coils

-Transformer : according the max power i am using means 300w , the transformer is very linear and without significant losses on frequencies below 20m, above losses are significant . The transformer heat during long QSO likely PSK31 is ok and i did not notice any overheat issue

-Radiation patterns according the installation configuration are acceptable and i did many tests on WSPR and FT8 with acceptable results on 40m

-Antenna tuning is not tedious and i manage to tune the antenna without any issue

15m antenna using 2 loading Coils (35uh & 90uh):

-The antenna performance are average on frequencies below the 20m band. Above the performance are deeply poor : just forget it.

-SWR is OK on 40 and 20m but very tedious to achieve on 80M

– Radiation patterns according not as good compare to the 23m one. Losses can be about 3 up to 6 or 9dB during transmit above 20m. For receive the performance are almost the same compare to the 23m long antenna.

-Therefore according a limited space or HOA restriction , this antenna remains a good choice for 80m, 40m and 20m. For 10m and 15m the perfomances are not good at all.

It is just a question of compromise but limited space or HOA restrictions are something we need to deal with . The 15m long antenna is better than nothing.

Conclusion

Great Antenna for frequencies below 14Mhz , For above just forget it .

In order to achieve better performance i will recomment to use alternative Transformer . The Fuchs ATU is a great alternative and will provide a really high-efficiency transformer (https://f5npv.wordpress.com/remote-automatic-atu/), using the Fuchs transformer with my Endfed wire is simply day and night regarding the performances compare to my previous 1/49 or 1/64 transformers using ferrite core.

In the case of the EFHW 49:1 transformer, the major shortcomings of the most popular current designs are excessive leakage inductance and poor heat transfer from the ferrite material to the environment. Tackling both of these issues at once led with an efficient transformer (The FUCHS Transformer )with better loss characteristics than previous designs.

A great SWR does not mean at all an antenna is working well and this the main extent for EFHW using standard ferrite core 1/49 or 1/64 transformer and leakage inductance may transform your favorite transformer in a kind of Dummy load.