It's difficult to get accurate measurements of trap performance, but a few
general points are:
* High-Q traps with inductors and fixed capacitors are the lowest loss, but
voltage breakdown in the capacitors is a constant problem at high power.
* The next-best performance is the method used by most tribander
manufacturers, using the capacitance between the metal cover and the coil. I
have heard reliable estimates that these traps are roughly 0.1 dB loss per
trap. This is not necessarily total gain loss per trap, but loss in that
element, which may not have a big effect if the current is low (e.g. a
director has lower current than the reflector or driven element in a Yagi).
* Coaxial cable coils plus the inner-to-outer conductor capacitance are
convenient traps, and can have reasonable Q, but will be lower on the
performance list than the others. The dielectric losses in the capacitor are
greater than other types, and the braided shield of the outer conductor can
be a reliability problem as oxidation begins to insulate the strands and it
ceases to function as a single large conductor. At high power, the coaxial
capacitor has had many reported voltage breakdown problems, so if KW power
levels are anticipated, use high quality PTFE dielectric cable and protect
the ends as much as possible, since that where the voltage is usualy
highest, as well as where a path can develop with salt, dirt and water. The
3 dB loss number is way too much -- the trap would melt rather quickly if
dissipating half the power! A well-built coaxial trap is probably some
fraction of a dB loss and well worth a try.
Someone else needs to answer the question about modeling traps, since that
is not an area where I have done much work. My own work has emphasized
> Some questions about traps:
> With regard to coaxial vs. discrete component traps, some have said that
> coaxial traps exhibit about 3 dB loss over discrete types. Where is this
> loss manifested? How does it affect antenna radiation?
> In the case of a trapped inverted L for 80/160, would this loss have
> much affect on 160, or would it tend to reduce the horizontal component?
> Also, how does one determine series R when modeling a trap?
> 73 Dallas W3PP
Antennaware mailing list