Richard Clark wrote in
:

I am wondering why you are trying to resurrect this train wreck.

I was seeking comment on the issue of asymetric branch feed topology.

It was put to me that accounting for the phase shift due to the different
branch lengths does not fully account for the time lag. My contention is
that in transforming the problem to the frequency domain, conversion of
time lag to phase lag fully and properly accounts for the different
branch lengths.

Gordon's paper was offered as evidence that my feed was "WRONG!".

Yesterday, I note that Kraus has a clear diagram of branch vs distributed
feed, and the technique of transposition to offset a half wave phase
delay.

I also note the ARRL agrees with me (http://www.vk1od.net/lost/Fig7.png
(c)), but that isn't a bullet proof recommendation!

I am now confident my critic was wrong.

Owen