On 9/16/2012 1:39 PM, Steve Hunt wrote:
The Forward Gain dropped by 0.3dB. The rearward pattern was badly
affected: all the nulls were "filled in" as Jim suggested, and
rearward lobes were typically 8dB higher.
Expanding on this a bit, this result is common to nearly all designs
that produce directivity -- forward gain is produced by in-phase
addition, so small changes in any element have a relatively small
effect, but nulls are produced by the cancellation of two (or more)
signals that are nearly equal and nearly 180 degrees out of phase. The
more precise the equality, the greater the depth of the null, and
relatively small differences in the signals producing the cancellation
can fill in the null.
73, Jim K9YC
TowerTalk mailing list