OT: Ground Loss (long)

[alorona]

I wasn't clear enough in my last post. Let me try to put this idea out there one more time. 
If you have an antenna modeling program like EZNEC you can perform a simple experiment to illustrate a strange behavior of ground loss. 
The EZNEC example "Elevrad1" is a 1/4-wave vertical with 4 radials about 2 cm above ground. I re-scaled the antenna to 40 meters while keeping the radial height the same. 
The "common" or "average" ground that everybody talks about is conductivity = 5 milliSiemens per meter (mS/m) and relative permittivity = 13. With this ground EZNEC gives an antenna gain of -2.46 dB at a takeoff angle of 26 degrees. 
If conductivity = 30 mS ("very good ground") the gain improves to -0.47 dB while the elevation angle lowers to 22 degrees. This is what we expect. 
Now, we want to see what happens as the conductivity gets *worse*. Here are the EZNEC results, where the 1st column is conductivity in mS and 2nd column antenna gain in dB. 
30        -0.4715        -1.6810        -2.25 8        -2.44 5        -2.46 4        -2.30 2        -1.52 1        -0.78 0.1      +0.12
 0.01     +0.20
 
See what's happening? There's a point where the loss hits maximum, but then the system gain actually improves as the ground worsens more.
 In my experience poorer ground also has lower permittivity. Setting permittivity = 4, the losses do increase (so that there are no positive values) *but the trend remains*. 
It's also true that the elevation angle increases as the ground gets worse, but it stays in the 20 - 30 degree range and doesn't get crazy or anything.
 
This suggests that ground loss is sort of an impedance-matching problem. Clearly, as your ground gets really bad your antenna can still work just fine, even better in some cases!
 
Totally unintuitive. But upon further reflection it makes sense that an "insulator" as ground is okay, otherwise an antenna in free space wouldn't work. I never meant to suggest something crazy like that one not use radials; I only wanted to point out a peculiar characteristic of verticals above real ground.

I will re-state my original thesis this way: If you have perfectly conducting ground, there is no ground loss. If you have perfectly insulating ground, there is no ground loss. There's always some ground conductivity in between those extremes at which the loss is maximum. This value depends on the frequency.
 
Al W6LX

______________________________________________________________
Elecraft mailing list
Home: http://mailman.qth.net/mailman/listinfo/elecraft
Help: http://mailman.qth.net/mmfaq.htm
Post: mailto:[hidden email]

This list hosted by: http://www.qsl.net
Please help support this email list: http://www.qsl.net/donate.html
Message delivered to [hidden email]

[wayne burdick]

Matches my experience, Al. This humble arrangement of wires radiates somewhere along a continuum between "vertical with one or more radials" (over perfect ground) and "inverted V mounted at a weird angle" (in free space). I think that explains the variance in loss, takeoff angle, etc.

Most portable antennas live between these two extremes. Your analysis proves the operator is better off enjoying the effect of radiation than obsessing over its shape.

Wayne
N6KR

On Jul 22, 2015, at 6:05 PM, Al Lorona <[hidden email]> wrote:

> I wasn't clear enough in my last post. Let me try to put this idea out there one more time.
> If you have an antenna modeling program like EZNEC you can perform a simple experiment to illustrate a strange behavior of ground loss.

> ….

> I will re-state my original thesis this way: If you have perfectly conducting ground, there is no ground loss. If you have perfectly insulating ground, there is no ground loss. There's always some ground conductivity in between those extremes at which the loss is maximum. This value depends on the frequency.
>  
> Al W6LX


______________________________________________________________
Elecraft mailing list
Home: http://mailman.qth.net/mailman/listinfo/elecraft
Help: http://mailman.qth.net/mmfaq.htm
Post: mailto:[hidden email]

This list hosted by: http://www.qsl.net
Please help support this email list: http://www.qsl.net/donate.html
Message delivered to [hidden email]