First published in WorldRadio, June 2000, and reproduced with permission
1. ~10 dBi gain with a take-off angle of 15 degrees on 10 meters
2. Horizontal Beamwidth of 31 degrees on 10 meters
3. Rotatable using an inexpensive TV rotor for coverage in any direction
4. Low wind resistance - less than three square feet of loading
5. More gain than a 1/2WL dipole on 20m, 17m, 15m, 12m, and 6m
6. Performs close to a resonant dipole on 30m (and is rotatable)
7. Performs well enough on 40m (and is rotatable)
8. Fed with low-loss 300-ohm Ladder-Line or TV Twinlead
9. Requires 1:1 current balun (choke) between the coax and ladder-line
10. Mounted on a 33 foot tall pole or tower (but higher is better)
The Extended Double Zepp (EDZ) has some well-known features - the most well known being the approximately 3-dB gain achieved over a dipole. It accomplishes that gain because it has a narrower horizontal beamwidth than a dipole. For rotatable antennas, the EDZ will outperform the half-wavelength dipole by approximately half an S-unit.
EDZ's have a bad reputation for being hard to match. They are hard to match with coax but not with ladder-line. If the EDZ is fed at a current maximum point on the ladder-line, one doesn't even need an antenna tuner.
For details on this no-tuner matching technique please reference: "W5DXP's No-Tuner All-HF-Band Antenna" at http://www.qsl.net/w5dxp/notuner.htm.
Let's take a look at some "magic" lengths of dipoles.
3/8 wavelength - the minimum effective dipole length recommended by Walter Maxwell in his book, Reflections. Shorter than this length, the resistance decreases rapidly and the capacitive reactance increases rapidly resulting in a very high SWR. A 3/8 wavelength dipole is 48 feet long on 40 meters.
1/2 wavelength - the most popular dipole length - associated with resonance. 1/2 wavelength on 30 meters is 46 feet.
Double Zepp - The Zepp is an end-fed 1/2 wavelength antenna that once was trailed from a Zeppelin airship. A Double Zepp is simply a one-wavelength dipole. A Double Zepp on 15 meters is 44 feet long.
Extended Double Zepp - A 1.25 wavelength dipole giving the maximum broadside gain possible from a dipole, about 3dB gain over a 1/2 wavelength dipole. An EDZ on 10 meters is 41 feet.
If one wants best performance possible on ten meters, then 41 feet is the length of choice. If one wants best performance on 40 meters then 48 feet is the length of choice. This article describes a compromise between 10-meter performance and 40-meter performance. The compromise dipole length is 44 feet.
The antenna is assembled from telescoping aluminum tubing and is a total of 44 feet long center-fed with 300-ohm ladder-line. The aluminum tubing is available from Texas Towers or MFJ. Please reference the The ARRL Handbook and/or The ARRL Antenna Book for the rotatable dipole mechanical and electrical design details.
The length of the feedline is important if one wants to avoid/satisfy an antenna tuner. The 300-ohm ladder-line used for these simulations (Wireman #562) has ~0.8 velocity factor. Other 300-ohm feedline may vary from this value. 450-ohm ladder-line could just as easily have been used and usually has a velocity factor around 0.9.
Let's say we have our 44 foot, 10-meter rotatable EDZ fed with 88 feet of 300-ohm ladder-line mounted and operational on a 33-foot tower. It has essentially the same gain as a two-element Yagi. What else can we do with it? Believe it or not, thereís not much else we canít do with it all the way from 7 MHz to 54 MHz.
The 44-foot dipole is resonant on approximately 10.6 MHz so when it is used on any HF frequency above 10.6 MHz, it exhibits more gain than a dipole. That includes all HF amateur frequencies from 20m through 10m. And 44 feet is long enough to do a good job on 30m. The gain is that of a dipole on 30m and is somewhat lower than a dipole on 40m. Besides, it is rotatable and thus will beat a fixed 40m dipole most of the time.
Figure 1 is what EZNEC tells us about the horizontal radiation pattern of the 44-foot dipole on 10m. Mounted at a height of 33 feet, it has a whopping 10.4 dBi gain over average ground, approximately 3 dB better than a half-wavelength dipole.
Band - Gain - Take-Off-Angle - Feedpoint Impedance
40m: 5.8 dBi @ 65 degrees 27.5 - j422 ohms
30m: 6.0 dBi @ 42 degrees 72 - j70 ohms
20m: 7.5 dBi @ 29 degrees 194 + j379 ohms
17m: 8.8 dBi @ 23 degrees 931 + j1207 ohms
15m: 8.9 dBi @ 20 degrees 2334 - j1141 ohms
12m: 9.4 dBi @ 17 degrees 419 - j959 ohms
10m: 10.4 dBi @ 15 degrees 126 - j471 ohms
6m: 9.7 dBi @ 8 degrees (Cloverleaf) 110 - j241 ohms
The length of the ladder-line is of utmost importance if one wishes to avoid the use of an antenna tuner or make the built-in tuner in a transceiver happy. EZNEC predicts the following necessary lengths of ladder-line for system resonance on the various bands:
7.200 MHz - ~.001uf capacitor in parallel at 86 feet = 50 ohms
10.125 MHz - 81 feet or 120 feet = 68 ohms resistive
14.200 MHz - 75 feet or 102 feet = 68 ohms resistive
18.140 MHz - 77 feet or 99 feet = 35 ohms resistive
21.300 MHz - 83 feet or 101 feet = 31 ohms resistive
24.950 MHz - 85 feet or 101 feet = 32 ohms resistive
28.400 MHz - 88 feet or 102 feet = 35 ohms resistive
52.500 MHz - 84 feet or 99 feet = 65 ohms resistive
These lengths are only approximate and the actual length in practice must be varied to achieve a near-perfect match without an antenna tuner or with a built-in antenna tuner.
However, if one wishes to use a fixed feedline length, it appears that 85 feet or 101 feet would be good fixed feedline lengths to use with this antenna and an antenna tuner. A good 1:1 choke/balun is an absolute necessity. Such a choke is positioned between the coax and the 300-ohm ladder-line.
The 300-ohm SWR is reasonable on all bands except 40m where it is about 30:1. 30:1 is not much to worry about as far as losses are concerned  but it does mean that not as good a match is possible at the low impedance point (approximately ten ohms) on the transmission line as occurs on other bands. The remedy for the 40m matching problem is a capacitive stub or actual discrete capacitor. It is estimated that a ~.001uf capacitor in parallel with the 300-ohm ladder-line about 86 feet from the 44-foot dipole will result in a near perfect match. Of course, an open stub can also be used. An antenna tuner will also work without the capacitor on 40m but at reduced efficiency.
 There is about one half of one S-unit loss in the feedline on 40m. Here is why that's not much to worry about. On 40m, the rotatable dipole beats a fixed 1/2WL dipole in about half the compass directions because it is rotatable. So half an S-unit of losses in the feedline seems a reasonable price to pay for an admittedly compromise design for 40m. Additionaly, the 44 ft. dipole beats a 40m 1/4WL vertical by about 1/2 of an S-unit in ALL compass directions (but not at all elevations).
This 40m-6m rotatable dipole was modeled using EZNEC, an antenna-modeling program available from Roy Lewallen, W7EL.Please feel free to visit his web site and download the free demo version of ELNEC from: http://www.eznec.com