Dipole symmetry?

In ideal free space, both arms of a tuned wire dipole will be of identical length.

However, the real world is not perfectly symmetric free space. The support heights, ground levels, ground conductivity, parasitic buildings, wiring, feed lines, and metal towers might be close to one end and far from the other end of a dipole.

Given very visible differences (such as the above) between the environments at the two ends of a dipole, should each arm of a "center-fed" wire dipole still be equal in length? Or should the length of each half or arm be separately tuned (for lowest SWR, resonance at the desired frequency, etc.)?

If so, how?

• Good question, Ronald. However, such a question could only be accurately answered by modeling the antenna and the nearby objects. EZNEC (or another antenna modeling program) is your friend. :-) Post a sketch here and perhaps someone with more time than I have will model it for you. – Mike Waters Jun 28 at 21:21
• The "center fed" dipole of your description could in theory be tuned to its environmental imbalances, but a current balun (choke) at the feed point will reduce the risks of imbalanced currents along the feedline by forcing equal currents. – JSH Jun 29 at 4:19
• In your list of things that interact with the antenna, you left out feed line. If the feed line geometry approaching the antenna is asymmetrical, it tends to couple with it and develop common mode currents. As mentioned above, a balun can help with this. – user10489 Jun 29 at 14:14
• How would you determine the magnitude and shape of the distortion of the current distribution? – Brian K1LI Jul 1 at 8:40

Changing the lengths of a dipole on either side of the feedpoint results in an off-center fed dipole. In free space, off-center feeding of a $$\lambda/2$$ dipole raises the feedpoint impedance, which could adversely affect the match to the feedline. The effect in a real-world scenario would be very difficult to predict and would most likely have to addressed through trial and error adjustment.
I added an "extra" current source to a NEC2 model of a $$\lambda/2$$ dipole in free space to examine the practical effect of distorting the current distribution. Even severe distortions like this: and this: produced less than 0.1dB difference in the radiation pattern. It's pretty clear that the minimal possible improvement doesn't warrant the effort that would be required to achieve it.