The recent discussion of differences among stations such as N2RM and W3LPL
cries out for some experimental confirmation.
There are three issues relating to stacked antennas at HF. Two are static
(array gain over ground, pattern fill over ground) and one is dynamic
(diversity effect of aperture size on fading statistics). Naturally,
integrity of implimentation (cable loss, aiming errors, etc.) is important,
as are modulation characteristics in the SSB case.
It's been my impression, backed by some informal experience, that not much
array gain is realized over ground, because the requirement for matching
phase and amplitude is not met at all radiation angles. In any event, even
the free-space gain of 3 dB by doubling the number of array antennas may not
account for the apparent perceived difference in signal levels mentioned by
GM/ECO.
Pattern fill over ground, which can be thought of as a form of static space
diversity, raises the mean signal level by eliminating elevation angle nulls.
This couples the antenna system effectively to any active single- or
multi-hop ionspheric modes and reduces fading from shifting levels of the
various active modes (angles). The far-end received signal strength
statistics also depend on the degree of coupling of the far-end receiving
antenna to the same modes, which implies you'd want to illuminate all the
multi-hop angles that might be important to a range of far-end receiving
antenna heights. It also means that comparisons at one station might not
correlate with those at another.
Operating a stack intended to improve stateside 40m coverage, I've noticed
that the phasing is not especially critical even though the upper-lower-both
comparisons always seem to favor the stack. This could imply that the
biggest benefit from a large stack is the reduction of average fading because
of space diversity, which would raise the far-end perceived signal level. As
an example, in a mobile system with 60 dB fades, combining 8 antennas
separated by at least the correlation distance of the fading will result in a
maximum combined fade of only 3 dB. The distance between N2RM and W3SPL
certainly exceeds the correlation distance for HF fading, so some
extended-time measurements would be needed to confirm the anecdotal
impressions (which are usually right if repeated over any reasonable time
period).
An interesting experiment would be to record far-end signal strength of an
unmodulated carrier from a physically large array (say a 4-stack) compared to
the same power level at the same time from a physically small (single)
antenna. The question of terrain topology and ground electrical parameters
can be removed if both transmitting antennas are at the same site, and it's
possible that a rapid-switching measurement would suffice to determine the
source of the difference among small and large stacks. A measureable
difference in fade statistics would confirm the conjecture that diversity is
an important factor. This has been mentioned in connection with, for
example, large wire antennas like rhombics.
Capturing the data is the hard part; you can use the threshold effect of
receiver manual rf gain control to infer the fading statistics. If the
answer is that space diversity provides the major benefit, it's not
necessarily so that vertical stacking is the only way to get it. I'd be
interested to know more.
73, Dave W6QHS
|