[TowerTalk] [Bulk] Re: [Bulk] Re: Radials how deep is too deep?

[Jim Brown]

On Thu,12/10/2015 6:43 AM, Grant Saviers wrote:
> As you well know, buried bare radials are the standard for BC 
> stations, and as are any ground screens therefore "connected" to 
> earth, but that 's not the point.  The CW is bare or insulated makes 
> no difference for buried wire radials. 

An important difference between AM BC stations and typical ham stations 
is budget and FCC Rules, both of which drive the radial system. Yes, 120 
radials and a ground screen have been the gold standard since the 
earliest days of AM Broadcast, and systems like this tend to cause 
current in the radials to be quite well distributed between radials. 
This is VERY different from most ham radial systems, which, limited both 
by budget and real estate, are much fewer in number and much shorter.

The principal reason for having more radials, and for not connecting a 
few, shorter radials to the earth is that making that connection tends 
to increase the imbalance in radial currents, which in turn increases 
ground loss. N6LF has published excellent work on this topic. google to 
find it if you haven't already. Great stuff.

Here's a light bulb that went off for me after studying Rudy's work. The 
loss in a radial is I squared R where I is the current in the radial and 
R is the loss coupled from the earth. As we add radials, I is divided by 
more radials, so the loss in each radial is reduced by the square of the 
number of radials, and the total loss is divided by the number of 
radials (assuming equal current in each radial). This is a great way of 
understanding why loss is reduced as we increase the number of radials! 
Rudy's contribution to this was to observe that if current is unequally 
distributed between radials, those with greater current will have more 
loss (because of I squared), and that increased loss will be much 
greater than the lower loss in radials with less current (again because 
of I squared).

Another lightbulb from Rudy is that the electrical length of radials 
sets their current distribution; a radial that's a quarter wave 
electrically will have a current peak at the feedpoint, while one that's 
longer will have a peak farther out from the feedpoint, and that peak 
will be greater than the current at the feedpoint. Again, since loss is 
I squared R, loss in a radial that's too long will be greater than one 
that's near resonance or shorter than a quarter wave.

It's also well known, and I'll bet that you know it, that variations in 
the soil under a radial system will cause current imbalances, which is 
another great reason for having more radials.

73, Jim K9YC