> With this setup, I could measure no significant losses. I am convinced
> (I convinced myself) that the "UHF connector loss" is a myth, right up
> there with that myth about the superiority of N-type at HF.
Probably a more relevant test is not to string a bunch of connectors
together in series. Rather, take the same number of connectors and place
them at random points on say, 100 feet of low-loss line. I think Steve's
description (six ft. line sections connected with short Zo=100 connectors)
accurately reflects what's happening, but it would be even more meaningful
to look at SWR on the individual line sections between connectors as well as
measure power delivered into a 50-ohm load at the end of 100 feet of line --
then mix up the distances and record several different iterations at 28 MHz.
Speaking of non-constant impedance, consider the garden-variety monimatch
SWR meter. Crack one open and you'll see three bus bars, equally-spaced in
a metal enclosure usually about five inches in length. The inner bar is the
transmission line inner conductor. But the outer bars are used only for the
sample; they don't contribute the characteristic impedance (Zo) of the
monimatch. Zo is formed by the inner bus and the metal box which results in
a characteristic Z of probably several hundred ohms, certainly nowhere near
50-ohms. And yet, how many folks have inserted a monimatch SWR device at
random points along the line, expecting near constant-impedance insertion?
The connector quality issue probably boils down to reliability and cost.
However, in instances where the connectors are being used to accurately
measure impedance and return-loss (e.g., VNA loads) then better connectors
are needed as F increases. For HF S11 reflection measurements into most
line and antenna applications, it's probably a non-issue.
Just yesterday, I was bashing the Asian BNCs. Not that I've changed my
position, but even when a connector is so bad that return loss accuracy
shifts from 70 dB to 25 dB, the final results are still fine for most
antenna measurements. But that amount of degradation tells me that if QC
was poor at the time of manufacture, perhaps it's not a connector/adapter I
want to use when I need it most.
TowerTalk mailing list