On 10/21/2011 10:00 PM, Owen Duffy wrote:
wrote in
. fr:

HI,

I picked up two coils that could be a trap for the CWA-840 COMET
antenna. (They have two stickers "CL-840").
The manual of the CWA-840 says : 40/80m dipole.

I found amazing that there isn't any capacitor in // of these coils...
These coils measure 76uH.

That is a rather simplistic characterisation of the trap. If you were to
measure its impedance over a wide range of frequency, the data would
give a better idea of the extent of any parallel capacitance (and there
has to be some, even if only stray capacitance. Somethimes the
capacitance is between some metal tubes that make the end supports for
the trap.


or C from the winding of the coil to the surrounding tubing.

Or the "self C" of the coil (viz Medhurst's formulae for self C of an
inductor)

There's a goodly bit of empiricism in trap design. You can use basic
physics to calculate, but there's enough non-idealness in most
construction techniques that the calculations get you to a starting
point. (e.g. you could model the transition from the antenna element to
the coil in the trap with a lot of detail, or you could just build one
and try it, then adjust)



Note that measuring the impedance of traps is very challenging.

Could someone tell me how they works ?

The trap introduces some reactance and resistance that is intended with
appropriate conductor lengths, to give a modestly low VSWR at the
frequencies of interest.

The traditional explanation is that they are resonant at one of the
frequencies of operation, and act like a switch, cutting off current to
the outboard legs. Of course, current would flow in the outboard legs
eve if you used a physical switch.

That explanation is appealing, but limits your options. There are
advantages is making the resonant frequency of a trap fall outside the
bands of interes



Here's a (not very realistic) example of how it might work

Say you want an antenna that covers both 10m and 15m. You start with a
10 meter resonant antenna. Now hook a trap which has very high
impedance at 10m on the end of it. Since it's high Z, not much current
flows, so it doesn't change the 10m behavior very much. But, you also
want it to work at 15m. So you figure out how long an additional
element you need to put "outboard" of the trap to get the system to
resonate. Typically, the trap is a parallel LC, so below resonance, it
looks somewhat inductive. Since a short element looks capacitive, the
added length for 15m will be shorter than it would be without the trap.


There's a whole raft of design approaches about where you put the
resonance relative to the frequencies of use, and how you want the
current distribution to work, and then there's all the "what's practical
to make and manufacture" and "tolerance to misadjustment/weather/aging".

You can get a pretty good start with a modeling code like NEC, but
eventually, you've got to go out and start fiddling in real life,
because almost all traps are not precisely modeled by NEC for a variety
of reasons.



In the manual of the CWA-840, each leg of the 7MHz section, is
11.1meters in lenght , wich, it seems to me, that it is rather
longer that 1/4 wave. (10.1m).
Why ?

See above.

Thanks a lot for your answers

CWA-840:

-----[ ]-------------ooo--//--
2.3m coil 11.1m balun

The trap is probably self resonant below 40m... but that is just a guess
from experience of having designed such things.

Owen