E-H Antenna

Background

Many antennas function because of common mode currents, rather than working in spite of them. Two popular examples are CFA and EH antennas. Another recent example, appearing in Antennex's compact antenna articles, is a thick stub "vertical" with no counterpoise. 

All of these antennas become significantly poorer radiators if common-mode currents on feed lines are eliminated. Why? Because the feed line is the actual radiator, NOT the tiny thing they call the antenna.

Misunderstanding or misapplying Maxwell's equations and the principles behind radiation, in combination with missing some very key points of conventional circuit theory, causes problems. Some of us have unwittingly attributed increased feedpoint resistance and/or seemingly disproportionate amounts of radiation from very small structures to new methods of radiating EM waves. Reviewing these antennas and the theoretical or technical mistakes surrounding them will help us understand how antennas and transmission lines work. With that knowledge, we can build better antenna systems. The fastest and best way to learn is often to look in detail at mistakes!

What They Claim

Articles and user reports of CFA, EH, and thick stub verticals (without groundplanes) appearing in Antennex and other internet publications have one common thread,  the operational descriptions almost always include strong indicators of problems with feed line common-mode current. 

Authors commonly warn users to NEVER choke feed lines with baluns and to "be sure the feed line is straight and in the clear"! Authors lay blame for RF burns from feed lines or shack equipment on the antenna's "high radiation efficiency", claiming these small magical antennas radiate so efficiently they naturally excite the feed line and equipment more than full-size antennas.

For example, the designer of the EH antenna claims the following:

"RF on the Coax

In the above statement, the designer actually acknowledges current on the feed line shield and RF-voltage-on-chassis problems. The problem must be severe when a low-power 100-watt radio causes a burn. Like any good salesman, he turns a design shortfall into a feature! According to the author, unwanted RF on the feed line doesn't come from a feedpoint or antenna design problem like it does on other antennas, in this case the unwanted RF appears because the antenna works so well!

Here's what actually causes RF to appear on a coax shield and radio chassis. RF can only appear on the radio chassis through two methods:

1.) The antenna, from poor feed line or feedpoint design, can couple to the radio chassis through external wiring or cables attached to the radio.

2.) The radio chassis itself, being large in terms of the wavelength, can actually become an antenna and receive energy from actual desired "over the air" signals.   

(Many of us have these problems. Click on this link to see one reason why.) 

In this case, we can probably rule out reason two above. It is unlikely the chassis is a large portion of a wavelength long on 20-meters and that the antenna field is suddenly so strong it is "lighting up the house" with RF. After all, if power is radiating effectively it will all be going out to distant stations and NOT cooking you or the radio gear in your house! That only leaves reason one, poor feedpoint and antenna design, as the cause of  common-mode feed line or wiring currents that excite the coax shield and eventually the radio chassis, as described in method one.

Actually someone has measured this, and posted it to a web page!

http://www.home.earthlink.net/~calvinf15/_technical/ 

When the time-varying current from the transmitter flows in any conductor, we will have charges accelerating in the cable. The outside of a shield is no exception. A feed line's shield will radiate proportionally by the ampere-feet of the cable. just as any other conductor will. Of course the antenna element will radiate also, but there is something else to consider.

A very small current flowing unopposed over a large linear distance will radiate quite a strong signal, because radiation resistance of a long antenna is generally very high compared to very short antennas.  You can find this explained in the Radiation&Fields and Radiation Resistance articles on this site, and in engineering textbooks such as those written by Jasik, Kraus, and Jordan-Balmain. From all of this, we know the shield radiates.

The inventor of the E-H antenna goes on to say:

Not a very knowledgeable statement at all, at least from the standpoint of how shields work! 

Time-varying fields can not pass through a shield that is more than several skin depths thick. The inner part of the shield and outer part are isolated by the skin effect, and nothing passes through. The ARRL Handbook, Maxwell's book "Reflections", Reference Data for Radio Engineers, and dozens of other amateur radio and engineering texts describe this effect correctly. If we bring a time-varying electric field to zero in a system, the time-varying magnetic field is also by definition zero. The shield DOES isolate the center of the cable from time-varying magnetic fields on the outside of the cable!

If that is true, why then does a shielded cable passed through a current transformer used in a directional coupler appear to pass RF magnetic fields? Why does the RF magnetic field seem to "pass through" the shield of a shielded receiving loop antenna? The answer is quite simple. There is a gap or intentional break in the shield. 

Current on the inside of the shield "spills over" the edge of the shield where the shield is broken, and causes a current on the outside of the shield. There is also a voltage across the gap at each end of the shield. We have both our time-varying voltage and current, via a circuitous path to the shield ends! Our Amateur texts explain that effect, as do all of the engineering texts dealing with shields on transmission lines. If the gap is closed and the shield's ends are shorted together, making voltage across the shield gap zero, the magnetic field no longer "seems to" penetrate the shield! 

If a shield did not behave this way, we'd be in serious trouble in the radio world. Without the shield stopping both magnetic and electric time-varying fields, we could not shield our radios. We also could not shield our microwave ovens, with non-magnetic materials!

If anyone thinks a ferrite core affects the impedance inside a coaxial cable that does NOT have significant common mode current, they only need to slip some beads over a cable on a working normal "cold-for-RF" feed line. You will find absolutely no difference in system performance when beads are added, proving Ted does not "have it right" in the text above that appeared on the E-H antenna web site. 

The author and inventor of the E-H goes on to say:   

Of course adding a ground wire might help! The ground wire becomes the path for common-mode currents, or at least a portion of them. The additional wire to ground becomes part of a long-wire antenna (made by the cable shield) that actually does most of the radiating! 

It is understandable why this is a preferred method! The outside of the coax shield can remain the primary antenna, saving us the bother of installing the additional ground wire that becomes the antenna in the previous suggestion!   

"It may take one or more of the above to solve your problem. Remember that if you have a good ground on the antenna, you have also minimized problems with lightning."

In the above text, we can see every solution carefully avoids installing an effective choke balun on the feed line. A properly designed and effective choke balun has no effect on a coaxial feeder or system SWR, unless the feed line is radiating!

Let's look at another E-H antenna experimenter and former proponent of the E-H antennas test of a 160-meter E-H antenna. You can read the text directly at:  

http://www.qsl.net/iz7ath/web/02_brew/18_eh/english/pag11_eng.htm

Steve writes: 

"With my short (and easy) tests I deduce the EH lose about 10 dB versus my short 10 mt vertical with capacitive hat in the top;
10 db is a big amount of power, but remember we are using 2 small cylinder of 37,5 cm on 160 mt band;
my vertical performed better than a 40 mt long sloper, so, may be, loss versus full size dipole is not too much."

Steve claims the E-H antenna looses about 10dB compared to a 1/4wl sloper, which has an unpredictable efficiency. A typical sloper 1/4wl sloper is likely only around 50% efficient in the best situations, and more likely much worse! The actual range can be from 10 to nearly 100% efficiency. Steve's data repudiates Ted's  claim and the CFA inventors claim that "crossed-field" antennas provide high efficiency.   

"There seems to be something good;
May be optimizing something we can have more gain (or, better, we can lose less dB);
This antenna can be useful on the low band, expecially on 137 KHz (the EH cylinder should be 6 mt high with a 2mt diameter).
On the web I've seen an other similar antenna (ISOTRON), wich seems caming (as EH Antenna) from CFA antennas;
On the italian Radio Rivista (1995-1996) there were same informations about an "Antenna Toroidale" wich in same way remember the EH: will be interesting build and test one."

Steve seems to be saying something we all agree with. A short antenna will radiate, but not nearly as well as a full-size antenna. 

"By the way, I think that my EH is not for Dxing; it will be useful for that radioamateur who have no possibility to install "long" antennas for low band, but want just to have local QSO or few contest-contact."

I agree! 10-20dB loss from a full size antenna would not make a good DX antenna! Now here come the current-on-the-feed line problems: 

"Other point: tuning;
I think it's not as simple as others say; I've used an MFJ259B which tells me all about the antenna, but the tuning was critical. On the other side construction is easy;
I want suggest to you , when tuning, to connect the MFJ259B to the shack ground (if there's one). If not, EH will probably resonate higher (20-30 KHz)."

If grounding and ungrounding an antenna analyzer or any other piece of equipment connected to a coaxial cable causes resonant frequency or SWR of an antenna to change, the system has severe feed line radiation problems. See my article on testing baluns. 

"Coaxial cable influence: inserting more coax cable, the resonance seems to vary a little; I suppose that's normal;
I don't know for certain if the line radiates or not; I've added two iron-powder coil forms back my TX and EH performance was unchanged. Then later I've added 2 big ferrite beads with a lot of turns on the roof (at the feed point of the EH) and all has changed. My signal was 3 S point weaker than before and S.W.R was very high.
I don't know if that loss is due to the changed resonance in EH (as I said before, outside the bandwidth performance goes down quickly)."

We see again that any attempt at reducing feed line radiation results in an antenna that does not radiate very well. With the feed line choked, Steve's EH antenna dropped 3 S-units in addition to the original 10dB from his sloper, or maybe 25dB or more!

There surely is a hidden message in all of the above contradictions!

How the E-H Antenna Really Works

There are many examples where designers intentionally use common-mode currents. Examples are found in textbooks, such as the "Antenna Engineering Handbook" by Jasik on and around page 22-6.

The antennas at the right, copied from Jasik's textbook, outline the derivation of a skirt collinear antenna from a simple feed line with the open end terminated into a conductor. (It could be a ground rod or an antenna, like a Beverage or large loop, the antenna does not have to be an "open circuit".

Looking at (a), we find by hanging any conductor from the end of a coaxial cable the shield is excited (on the outside) with common-mode current. The electrical equivalent of the OUTSIDE of the shield is just as if a generator located at the end of the shield was driving the outside of the shield as longwire antenna. This goes along with Kirchoff's Laws, that tell us the sum of currents entering a point must equal the sum of currents leaving that point. For any current to flow up into the antenna, and equal current must flow back down over the outside of the shield.   

With one ampere flowing up the center conductor into the "stinger" at the coax's end-point, the  same level of current flows back over the outside of the shield. (The shield's inside and outside are isolated by the skin-depth of the current at the operating frequency, an can be treated as two independent conductors that are connected over the open edge of the shield.) We MUST have this current simply because this is how coaxial cables work, the current on the inside of the shield is ALWAYS equal and opposite to current in the center conductor. There has to be some place for that shield current to flow, so it makes the bend over the end of the cable and flows back down the outside.

This is also why, when we use a cable's shield as a ground lead the center conductor and inside of the shield do nothing to reduce resistance. Any current that flows down the center conductor is cancelled by current flowing on the inner wall of the shield, the result is no current at all flows down the center conductor as long as the shield is several skin depths thick.   

Many antennas intentionally and unintentionally use this principle, two examples are shown in (b) and (c).

A recent Antennex Article on a "magical" ultra-compact antennas claims an identical system, using a loaded fat cylinder, has an extremely high radiation resistance and excellent performance because some magical field-trickery increases the radiation resistance of a thick  cylinder at the antenna end. Certainly radiation resistance is somewhat high...but not for the reason something magical or special is happening! 

The small coil-loaded cylinder is actually only a fraction of the antenna length, and being so short has a very small radiation resistance. The point missed is the shield of the cable is very long, and is in series with that short section. Since the shield is long, it also has a reasonably high impedance both from radiation resistance and loss resistances. Shortening the length of the end-stub results in an insignificant reduction in radiation resistance, because the overall length of the radiating system is very long! We have a simple off-center fed dipole, with one very long leg and one very short leg! 

The main radiator is the outside of the feed line shield, not the tiny thing being called an antenna! 

Unless we make the coaxial shield an infinite length or pass it through what amounts to an infinite groundplane with zero resistance, current continues on down the cable shield. Looking at (c), we find even multiple sleeves appearing as parallel tuned high-impedance circuits do not fully decouple the shield. Many collinear antennas work on this principle, yet E-H antennas and others attribute it to some form of electro-voodoo.