"A pair of vertical half-wave dipoles spaced a half-wave would be touching each other. That puts the feedpoints 180 degrees apart."
Jeff, the real antenna is physically shorter than the free-space wavelength, so they don't touch. My dipoles on 29.6 were 185 inches long, versus a free-space half wavelength (used for spacing) of a little over 199 inches. The ends were close enough to produce minor issues in balancing currents, but they don't touch. The feedpoints are 180 degrees apart in space, but are fed in phase. It results in gain toward the horizon, but the goal in this case was to create the deepest possible null along the axis (for vertical separation of antennas on a 10M single-site repeater) using the least amount of tower, and nulling the currents flowing in the tower legs. As admitted elsewhere, I was confusing these dipole arrays with some others I tried using elements fed at +/- 45 degrees. I concede that for any frequncy at which the lobe toward the horizon can be considered the makor lobe, it will be perpendicular to the antanna's axis if the elements are in phase. 73, Paul, AE4KR PS - With work, I was able to model tower-mounted antennas for 10M, using two-element colinears for both transmit and receive, that achieved 70 dB isolation at 400' vertical spacing. When I improved the model of the tower to account for other hardware, guy wires, etc., not so much. It might be a useful application on 6M. ----- Original Message ----- From: Jeff DePolo To: [email protected] Sent: Sunday, June 08, 2008 9:14 AM Subject: RE: [Repeater-Builder] Re: antenna suggestions for 440mhz > Jeff, the pattern depends on both phasing and spacing. As > frequency drops, the interelement phasing, expressed in > degrees, remains the same, but the spacing, expressed in > degrees or wavelengths, drops. This is true. However, the antenna was originally designed for no downtilt, and all of the elements are in phase. Typical peak gain on the horizon usually ends up occuring somewhere around 0.9 wavelength spacing (varies depending a lot of other things, but consider this a nominal value), with little change in peak gain when varied up or down a tenth of a wavelength or so. Minor variations between inter-bay spacings will also not shift the major lobe off the horizon, but it will affect the elevation angles of the minor lobes (think "null fill"). > If you model a colinear array > of parallel-fed dipoles in an antenna software program, and > don't scale the dimensions as you scale the frequency, you'll > see the main lobe shift up or down, and "butterfly" lobes > appear, as you get a few per cent off-frequency. At some point you've gone so far away from the design frequency that what had been the "major lobe" no longer is the "major lobe". Until you get to that point, the major lobe will remain on the horizon as long as the elements are fed in phase. It won't shift up or down. At least it never has in any array I've ever modeled for two-way, broadcast, or otherwise. > In an extreme case, a pair of vertical colinear dipoles fed > in phase with half-wave spacing has the familiar big lobe > toward the horizon. A pair of vertical half-wave dipoles spaced a half-wave would be touching each other. That puts the feedpoints 180 degrees apart. Or am I not understanding what you're saying...? > As frequency rises, the pattern degrades > until, at a frequency of 2X, it becomes an end-fire array, > with most energy directed straight up and down. This happens > with no change in phasing or spacing. Yes, but the point is that what had started out as the "major lobe" was on the horizon, but you've gone so far away from the design frequency to the point where what started out as the major lobe no longer is the major lobe. You're talking about a 2X change in frequency - the antenna is no longer the antenna it was designed to be. We're not trying to use a 220 dipole array on 440 here, we're talking about relatively minor excursions from the design frequency when using "commercial" antennas out of band by a few percent or so. --- Jeff WN3A
