[TowerTalk] mismatch at feed, ok at transmiter

[Jim Lux]

There are antennas which will appear to be a poor match at the 
feedpoint.  For instance, a helical antenna might have a 100 ohm feedpoint 
Z.  A common strategy is to use a 1/4 wave transformer of 75 ohm coax to 
get the impedance closer to 50 ohms. (56 ohms in this case).  The loss from 
the relatively short section isn't all that high.

Likewise, in phased arrays, there are all kinds of schemes around where the 
elements won't be anywhere near 50 ohms at the feedpoint.

In a single feedpoint antenna, it might just be a design decision to let 
the Z at the feedpoint be non-50 ohms, and tolerate the extra loss in the 
feedline.  After all, how many people string up dipoles, which have a 
feedpoint Z of around 70 ohms, and just accept the 1.4:1 SWR  (with 200 ft 
of RG-213, it drops to around 1.35:1 on 40m)

For small mismatches (2:1 or less), the extra loss from the circulating 
power just isn't all that huge.  It may well be that trying to put a 
matching network at the feedpoint might make the antenna system have more 
loss or (more likely) be narrower band.  The antenna design typically 
assumes a broadband resistive 50 ohm termination when calculating all the 
mutual interactions.  If you put some sort of reactive matching network, 
the performance may be improved at one frequency at the expense of many others.

(The extreme in this might be the Terminated Folded Dipole.. a deliberately 
lossy antenna, that presents a reasonably good match over almost the whole 
HF spectrum)