[TowerTalk] Coil loss

[Jim Lux]

At 02:17 PM 6/2/2004 -0400, Tom Rauch wrote:
> >          It turns out that the shorter an antenna is
>relative to a
> > wavelength is, the higher the Q of the resonant antenna
>system (antenna,
> > transmission lines, coils, inductors and tuner)  must
>become to effectively
> > radiate RF energy. And in achieving that higher Q using
>coils and/or
> > capacitors there is a good chance of increased losses due
>to the higher AC
> > currents and voltages involved.
>
>The loss isn't that bad. See:
>
>http://www.w8ji.com/mobile_antenna_fs_comparisons.htm
>
>Ground loss dominates the system.
>
>The real issue is all those extra turns hanging dead when
>you go up in frequency. The same thing happens in antenna
>tuners in roller inductors. You can get a series-resonance
>in the coil and then voltages (and circulating currents) are
>incredible. The antenna can also become very narrow banded
>at the point were loss is highest. It really is a complex
>interaction where the unused area of inductor behaves like a
>pair of back-to-back L networks with extremely high values
>of L and very low capacitance.
>
>When an inductor is self-resonant or near self-resonance it
>does not have anywhere near even current throughout. That's
>what happens when all that extra coil isn't used. The same
>thing happens in tank circuits and RF plate chokes.
>
>73 Tom

You're talking about the circuit formed by the unused turns (inside the 
lower tubing) and the C to the surrounding tube? I would think that the C 
is going to be pretty big (the tube is close to the coil).  Say, if you've 
got 30cm inside the pipe, with the coil being 5cm in diameter and the tube 
being 6cm, I get 92 pF. Assuming 100 turns for the coil inside the pipe 
(about 8 TPI), I get 77 uH using Wheeler's formula.  These could be off by 
quite a bit because there is an interaction between the coil and the tubing 
wall. But anyway, the resonance would be around 1.9 MHz, which is close 
enough to be of concern (that is, if the L or C were off by a factor of 4, 
you'd be resonant at 3.8)

And, you're assuming the magnetic field from the part of the coil sticking 
up above the tube (with the antenna current flowing in it) couples into the 
part below.  There's probably also some coupling to the threaded rod that 
actuates things (both Capacitive, and the magnetic field)

I think on the higher bands, though, where more of the coil is inside the 
tube, the problem is less, since the L and C would both be bigger, reducing 
the resonant frequency.  The coupling from the active part of the coil 
would also be smaller.

Hmm.. maybe a reason to use a iron tube (plated with copper)?