[TowerTalk] Grey PVC Loading Coil

[Jim Lux]

At 10:02 PM 12/4/2006, K4SAV wrote:
>You are going to need something more than just a loading coil to match
>it.  A 40 ft vertical with an 80 uh coil at the base, using #14 wire,
>should be resonant on 160 meters but the feedpoint impedance will be
>around 2.2 ohms plus coil loss and whatever ground resistance you have
>from your radial system. I didn't calculating a matching network for
>you, sorry.
>
>By using a coil program I calculated some coil data. With #10 wire, 5
>inches in diameter, and 5 inches long, you would need about 30 turns to
>get 83 uH at 1.85 MHz. The resistive losses of this coil would be about
>1.2 ohms (2 ohms for aluminum) for an air core. This coil should be self
>resonant at about 7.4 MHz. For a PVC form, you would need about 29 turns
>to get back to 83 uH and the resistive loss would be about 2.2 ohms (3
>ohms for aluminum). This coil should be self resonant at about 5.5 MHz.


How does the program calculate the loss and change in resonance with the form?

I assume that the loss would be from the parasitic C, since there's 
low loss from the magnetic fields?  The change in turns, presumably 
due to the change in parasitic C that "cancels" the L, although I 
would think that you'd need "more" turns (since the C is bigger with 
the form), is only a few percent.

The resonant frequency change (smaller by a factor of .74) implies 
that LC is 1.8 times bigger with the form than without.  The 
reduction in turn count (29 vs 30) actually makes the L smaller by a 
factor of 0.93, so the C must be bigger by a factor of 1.94, that is, 
nearly doubled.  That's a pretty big change, for a form material that 
probably has an epsilon of something like 2-3, and with the physical 
size of the coil slightly smaller.

Does the program assume the coil is "immersed" in the dielectric 
(that is, the dielectric is "between" the turns) or that the coil is 
wound on the surface (or in a shallow groove), in which case only 
half the flux is in the dielectric.

What does the program assume for dielectric properties of PVC?


{I'm curious because this is an interesting problem that would make a 
good homework problem for a class in electro/magneto statics, and I 
know someone who's teaching such a class and is always looking for 
good practical problems.}



>This should give you a hint of the amount of loss contributed by the
>coil with the two different forms.  This doesn't give you any
>information on whether white PVC is better than gray PVC, sorry, I don't
>know.

Jim, W6RMK