The E-H Antenna Revisited

By Steve Katz, WB2WIK/6

A recent article on eHam.net, which can be found at http://www.eham.net/articles/3586, by KE0VH, discussed the virtues of the `E-H Antenna.'

Since that article can still be accessed on line, there's no reason to rehash what it said. If you visit the E-H Antenna website, http://www.eh-antenna.com, or a related site, http://www.qsl.net/w0kph/, you'll find more information relating to the theory, design and operation of E-H antennas. Unfortunately, for all its virtues, which I'll summarize here, the E-H isn't much of an antenna.

Having said that, I should also say that it's definitely not a dummy load. It makes contacts, solidly and easily. It's reasonably easy, and inexpensive to build. The only drawbacks I found, operationally, are that within reason and based on the designs provided in the articles, it won't handle a lot of power and the design is single-band in nature. It's a high-Q antenna that won't tolerate multiple band operation, at least not in the form discussed.

Virtues

Very small, very lightweight, very inconspicuous and unobtrusive - a good `stealth' antenna that could be employed almost anywhere, despite antenna restrictions.

Evidently quite efficient. Not to say its performance is on par with a ½-wave dipole - it's definitely not. However, the power applied to it is mostly radiated, as evidenced by the fact that I can run 100W CW to it for an hour and observe almost no component heating, which would be obvious if much of the 100W were dissipated, due to the small size of all the components.

Narrowband, but not so much that it won't cover most of an amateur band. The 20 meter model assembled, tested and shown in this article covered about 150 kHz of the 20 meter band between 2:1 VSWR points. Not so narrowband that it's not useful over the entire band; however, it's narrowband enough to reject out-of-band signals well. (It does, and much better than the typical `wire' antenna - which I find useful.)

Reasonably environmentally stable, for a hi-Q antenna. The model shown, using two mica compression trimmer capacitors for tuning, would drift 100 kHz or so down the band if it got wet, in the rain. When it dried off, it would drift right back up to its originally resonant frequency. I operated it in very hot (100 degree+) weather, and also in cooler (55 degree) weather, and the resonant frequency did not notably change. (I did not try sub-freezing weather.)

Best of all, the damned thing made contacts. Lots of them. I'll discuss that more in a moment, but making contacts is what it's all about, and the E-H can do it.

What's not great

Well, it's not a panacea. If I had to choose between having no antenna or having the E-H, the decision's clear. But, if I had to choose between having the E-H or having a ½-wave dipole, or a good vertical, that decision would also be clear, and the E-H would not be the choice.

Also, being a hi-Q, single-band antenna design, for anyone who wants to work a bunch of bands, they'll need a bunch of E-H's. Of course, the Bilal Isotron antenna (http://www.rayfield.net/isotron), which is amazingly similar operationally to the E-H, has the same problem: If you want to work four bands, you need four antennas. People seem to do that.

And, don't be tempted to `QRO' with this antenna. The breakdown voltage rating of even mica compression trimmers isn't sufficient to run much more than 100W of power to this antenna. I suppose the antenna could be redesigned using much larger, heavier components to accommodate higher power, but then the advantages of small, lightweight, easy and inexpensive to build are lost.

The test

The antenna shown was borrowed from Dave, KD3V, in Hollywood, CA. He already had one that he said worked well, so rather than start from scratch building one, I simply borrowed his (thanks, Dave) for a period of weeks. Dave didn't build this antenna, he just loaned it to me; it was actually constructed by Ivo, WA6SUA, evidently for Bob, KL7EAL, who is its rightful owner. This particular unit is beautifully done, obviously by someone technically talented, using premium materials. The cylindrical radiators are copper tubing, the inductors are Teflon-insulated wire, and the trimmer capacitors are high-quality mica compression units, as recommended by W0KPH.

Figure 1: The E-H construction, close up. Ivo did a beautiful job, and I doubt anyone could do much better.

The most important tests were the `on-the-air' ones performed mostly in October and November 2002. Hundreds of contacts were made from my station on 20 meters, CW and SSB, over paths ranging from direct wave (2-3 miles) to long-path F2 propagation at distances of more than 10,000 miles. In all cases, I had other antennas to compare to, instantly. The E-H antenna was installed on a non-conductive PVC pipe, as recommended by the articles, and placed 20 feet above ground on the roof of my home. Its feedline, about 20' of RG58A/U connected to 80' of high-quality RG8X coax (International 9092, about the best RG8X on the market), went to one port of my Alpha-Delta coaxial antenna switch. Two other ports of the same switch are connected to my 6 element LPDA beam up 53' on a tower in the backyard; and to a Hustler 6BTV vertical antenna on a roof tower (about 25' above ground at its base, and about 30' away from the E-H installation), with several radials. It takes less than one second to switch between all three antennas, observing signal strength and listening to strength and clarity of any station worked, or even heard.

I'll tabulate the results with some specifics, just for fun. Possibly some reading this article will be in my list of logged contacts, which is too long to reproduce in its entirety.

Overall, here's what happened:

E-H compared with 6BTV vertical: Typically -20 dB (down) from the vertical, on virtually any path - long, short, or between. Kind of amazing. Confirmed the 20 dB delta using a precision attenuator (HP 355D), it's really 20 dB alright. And quite consistently.

E-H compared with Tennadyne T6 6-element LPDA: Typically -30 dB (down) from the beam, dependant upon path, but still amazingly consistent. On some paths, perhaps only -20 dB; on some, up to about -35 dB, but on average, about -30 dB, again confirmed using the 355D precision attenuator (lab calibrated, traceable, +/- 0.2 dB).

Just for fun, I tried using a random-length wire, #18 gauge insulated copper hookup wire, just cut to 65' long and tossed out the ground-floor window, with no tuner or any matching devices of any kind, plugged into the fourth port of the switch using a `banana plug' to terminate the wire and make it fit the center post of an SO-239 connector. The wire and the E-H antenna were frequently neck-in-neck on received signals; however, I had no way to match the wire to make it work for transmitting, so this was a `received signal' test only.

The lab

I brought the antenna to the EMC lab at JMR Electronics Inc. in Chatsworth, CA (http://www.jmr.com) and placed it on the non-conductive rotating test table in the RF anechoic chamber to measure the E-H's radiation pattern and gain referenced to a NIST-traceable standard antenna. A snapshot of the test process and chamber used can be found at http://www.jmr.com/custom/compliance_1.html. This chamber is quite large and is virtually reflection-free from about 100 kHz through 10 GHz, and performs extremely well (like `free space') in the 14 MHz region, so measurements made therein are considered quite accurate by the FCC, CISPR and other agencies. In fact, this particular chamber and test site is NVLAP-accredited and on the FCC list of approved test facilities, of which there are fewer than seventy in all the U.S., last I checked.

Results: As well predicted by actual on-the-air use, the E-H measures between 20 and 22 dB below a 0.5WL dipole* at 14.150 MHz, its resonant frequency, depending upon planar orientation. I rotated the antenna in every possible manner (remotely, using robotic equipment in the chamber) and could not come up with any `amazing' results, no matter what I did. It measures about the way it works, no better, but no worse.

What does this mean?

I'm not sure. I guess it means that if you have a 100W transmitter connected to an E-H antenna in free space, that will perform about a well as a 1W transmitter connected to a ½-wavelength dipole in free space. Sounds like quite a compromise, but then, most people using an E-H probably don't have room or permission to install a ½-wavelength dipole, which, on 20 meters, would be more than 33' long and need to be 33' above ground to work like a dipole.

When one considers the 20 meter E-H is about 0.025 wavelengths long (!), can be made to look like a ventpipe extension, bird feeder or many other acceptable accessories, might cost $10 to $20 to build, and actually makes contacts on the air with reasonable ease, it's a hell of a deal.

In my operating experience, I often started a QSO using one of my regular, larger antennas, then switched to the E-H, logging the difference in signal reports (both ways). I sometimes told my contact what I was doing, and sometimes not. Sometimes I told my contact I was `switching to QRP,' and requested a new signal report, as accurately as they could provide. That 20 dB figure just kept recurring. If I was S9 with the 6BTV vertical, I'd drop to S5-S6 (18 to 24 dB change) with the E-H, if I could trust people's S-meters. I don't trust S-meters in general, so that's why I kept reaffirming with my own S-meter, calibrated by a lab standard attenuator.

In most cases - probably 90% -- the station I contacted with the large vertical or beam continued to hear me with the E-H, although weaker. And I could hear them, too. In about 10% of the cases, the contact would be lost by making the switch to the E-H. Those were the cases where the initial signal just wasn't that strong, and switching to the E-H would drop it into the noise level.

During the ARRL November Sweepstakes (CW), I initiated and completed several contacts on 20m using only the E-H. I think every single station I called came back to me, and we completed the contact - all five seconds of it!

I worked a `big gun' in India, Sabu, VU2ELJ, long-path (over the south pole) on 20m SSB using the E-H and 100W PEP output. Sabu gave me a `56' report. Normally, I work him using a kW and a beam, and I'm S9+. But, you know, 56 is a workable contact, and this guy is on the other side of the world.

Summary

Would I willingly sacrifice 20 dB of signal strength, both transmitted and received? Nope. But would I rather have an E-H antenna than be off the air? Yep, yep, yep!

I subscribe to the “if it was that great, everybody would be using them” theory of product development. And perhaps even more so to the “if it was that great, contesters would be using them” theory, because amateur radio contesters will try anything that provides the slightest operational edge. In the case of the E-H antenna, everybody isn't using them, and I don't know even one contester who is. Clearly, this design is not replacing Hertzian antennas any time soon.

If you want to read about the operational theory, go to the websites referenced earlier. But all that aside, if you want a miniscule antenna that can make global contacts with a bit of operator skill and propagation, try building an E-H. Especially if you can't install a `regular' antenna, due to restrictions. You'll be surprised.

WB2WIK/6

*Note: A 1/2WL 20m dipole isn't required to make this measurement. We use a NIST-traceable calibrated antenna reference standard and correlate data from that.

Contact tabulation

I promised an abbreviated contact tabulation, so here it is! All contacts are on 20m, either CW or SSB as indicated by RST reports:

STN DATE/TIME RS(T) RS(T) NOTES

w/6BTV w/E-H

VE3XAP 10/3 0031 59 55 Sam-Toronto

WO3Z 10/6 0203 599 559 NR 179-PA

KY5R 10/6 0210 59 55 NR 171-AL

AL7ES/7 10/7 0014 59 56 Terry-mobile in WA

W4USR 10/7 2353 589 539 Dennis-NC

N2UI 10/8 0054 599 549 Tom-NM

K1EKF 10/10 0049 59 56 Rich-GA

CO8ZZ 10/18 0121 59 55 Raoul-Cuba

These are very representative of results over a 6-week period. A total of 296 contacts were made with the E-H during that time.

-30-