Dave wrote:
a lot of the performance probably depends on the frequency vs length of the
coax. if the frequency is low enough that the delay in the coax is small
then you basically have a capacitor (the inner conductor to the inside of
the shield), in parallel with an inductor (the outside of the shield). it
gets a bit uglier because of capacitance between turns on the outside, but
that is likely much smaller than the internal capacitance. so you end up
with a basic parallel resonant circuit. at high frequencies where the
length of coax is longer than a small fraction of a wavelength that probably
falls apart and would take much more complex analysis to figure out.

This thing is much easier analyzed as a distributed network
transmission line. At the resonant frequency where I2 is 180
degrees out of phase with I1, assuming |I1|=|I2|, common-mode
currents are eliminated so there are none available at 'B'.
On frequencies where I2 is not 180 degrees out of phase with
I1, common-mode currents flow on the outside braid of the trap
and thus continue to flow down the antenna wire at point 'B'.

The key to understanding the operation of this trap is to
realize that current ceases to flow on the outside of the
trap braid at the trap's designed-for frequency.
--
73, Cecil http://www.w5dxp.com