On Mon, 03 Oct 2005 03:14:52 GMT, David wrote:

Doing an A/B comparison between the commercial antenna and my version.
The range achieved by the commercial antenna was approximately 4 times
the range of my version.
The range was determined by loss of using one end as a beacon and moving
the test unit further away until it could no longer receive valid
packets, then move in and out of range several times to confirm
consistency. The same was done with my version.

Hi David,

This suggests your problems are indeed very profound. A 4 times range
difference would suggest a massive mismatch.

Before the test, I monitored the return port on the directional coupler
on a Spectrum analyzer while seeping frequency from the sig. gen and
noted the SWR on the commercial unit was lower than my version. (Though
the measurement was not calibrated I could see the SWR was better but
did not determine how much).

Then I would bet that your system topology is to blame. That is, your
line from the coupler and test equipment to the antenna is not
normalized out of the measurement as a network analyzer would do. And
when you shift from measurement to application, your line length
shifts such that your test match goes to hell. This implicates the
isolation of the sleeve - which is already in serious jeopardy by many
variables anyway. Your test scenario does not reveal the efficacy of
this isolation, it simply affirms that using the spectrum analyzer
establishes a suitable, but serendipitous arrangement.

You might add ferrite beads to the lead going to the antenna/sleeve
and move them along its length to test this hypothesis. This
prophylactic measure is also commonly employed as a necessary adjunct
to design.

73's
Richard Clark, KB7QHC