Standard Delay Line Systems
Many phasing systems use standard 180-Spacing=Delay line phasing. In other words, the end-fire distance in degrees is subtracted from 180-degrees, and the result is the delay line length. This is an acceptable method in single frequency or narrow-band systems.
Conventional delay line systems have the following problems when used in endfire unidirectional arrays:
Some systems use 90-degree Hybrids or L/C phasing systems. Hybrids offer ideal distribution of power and provide the expected phase-shift only when perfectly constructed and terminated. While there is some tendency to self-compensate phase, they still suffer bandwidth limitations. Hybrids are very useful in electronic systems, amplifiers are one example. In an amplifier system, we might want a constant 90-degree phase shift despite slight frequency shifts.
Unidirectional antenna systems are never optimized when phase shift is fixed at some arbitrary value that remains constant as frequency is varied. ALL unidirectional endfire arrays require phase to track with element spacing change in degrees, as frequency is varied. Lumped component systems and Hybrids might save space, but they do not enhance array performance over the proper choice and design of transmission line phasing systems. This is true in both transmitting and receiving antenna systems!
If hybrids are so poor, why do we see so many of them in antennas? There are a few reasons authors and manufacturers use hybrids. They often think:
In reality, none of the above are true in a broad sense. Hybrids have their place, but it certainly is not in wide-bandwidth phased arrays.
Transmission-line phasing systems are a bit more tolerant than Hybrids. For example, a 90-degree long transmission line has zero degree phase error, even when grossly misterminated. A 90-degree phase delay transmission line has less phase error than a quadrature (90-degree) lumped component Hybrid when mismatched. Phase error peaks in misterminated transmission lines when the line is any odd-multiple of 1/8 wavelength, and is minimum with lines any multiple (even or odd) of 1/4-wl. For example, a 45-degree long transmission line provides 27 degrees phase lag when terminated in 25 ohms...not the 45 degrees people often expect! (Remember this when you see phasing designs that just throw a certain length of cable in series with a mismatched impedance!)
With all systems, amplitude errors are a problem. There isn't any passive system that provides correct phase and amplitude when load impedance changes, especially over a wide frequency range.
I prefer cross-fire phasing, rather than the conventional narrow-band phasing methods discussed above. Cross-fire phasing, when designed properly, ensures phasing is always correct regardless of frequency. When elements (in this article Beverages) offer a near-constant impedance that is almost entirely resistive over a wide frequency range, cross-fire phasing can function perfectly from VLF to LF all in one antenna. Phase and amplitude can be designed for a back-fire null, with the upper limit in frequency set by element spacing and the lower limit set by array sensitivity. It is possible to design cross-fire receiving arrays that maintain the same basic directional response over several octaves of bandwidth.