On Wed, 22 Dec 2004 00:19:12 +0000, Ian Jackson
wrote:
What works better?

Hi Ian,

In following up the links that you provided, there is a superior
reference found at the same site:
http://www.minicircuits.com/appnote/an70001.pdf
which in turn leads to a treasure trove on the topic that supports my
favorite discussion on Mismatch Uncertainty. By following the links,
they offer you articles and software to compute the accumulation of
errors (and loss) found in mismatched sources looking at mismatched
loads.

A notable quote:
"A fixed attenuator can help to lower the VSWR of cascaded
(connected) components by providing isolation between the
impedances, effectively masking the impedance mismatches."

For both mismatched source and load, one handy shortcut offered is
that the system suffers a SWR that is not the aggregation of the two,
but the multiple of the two. This is not particularly significant for
a source SWR of 2 seeing a load SWR of 2 (same result of 4 for either
addition or multiplication), but above this value and loss begins to
climb dramatically.

73's
Richard Clark, KB7QHC