...I know air-core choke is worse then ferrite cores but I am a bit puzzled by the picture of the ARRL Book which shows the spirals which cross one over the other as they used just some tape. I did the same as the ARRL book but should I instead have perfectly wounded the coax... or an easy wound ... has more or less the same effect...
Common mode currents and baluns have not been well understood by many ham authors, and there is a lot of
misinformation available. It is unfortunate that W7EL's classic paper
was tucked away in the
first ARRL Antenna Compendium
rather than published in the front of QST
where it belonged.
(But perhaps it stepped on a few too many influential toes...) And it is rare that anyone actually measures
the performance of the chokes they recommend, but that is even more complicated, because it will depend on
the type of antenna, the cable lengths and how they are grounded, etc.
The primary issue with an air-core choke is the self-capacitance between the turns. On one hand, you need
enough turns of coax for the choke to have a high enough impedance, but at the same time the more turns you
add, the higher the capacitance across the coil, and eventually as you get too many turns (or go higher in
frequency) the capacitive reactance exceeds the inductive reactance, and the coil acts like a capacitor
instead of a coil. As you go above that self-resonant frequency, the choke becomes less effective.
Generally, trying to keep the coil in a neat spiral rather than "scramble-wound" will improve performance
because there are fewer opportunities for capacitive coupling from the first to the last turn, which could
be adjacent in a less tidy winding. A bit of space between the turns will help to reduce the capacitance,
though also reducing the inductance. Generally these differences aren't huge, but, given the relatively
narrow winding of good performance to start with, it could move up or down by a few MHz at the higher
end of the range.
So it's not a binary "this method works and that doesn't" sort of answer. Some approaches may tend
to work better, but might not in a particular circumstance, especially when the choke is marginal in
the first place. The real test is to build a current meter and measure the common mode current on
your feedline, then wind a choke and see how much it has changed.