[TowerTalk] re Radials

Dave Tipton dave at lodave.org
Fri Jun 16 13:16:43 EDT 2006

Well.. I have an old 5BTV (Hustler) that I'm going to put up this weekend...  It is spec'd for 1/4 wave radials on each band, telling me that I need to have 3 on each band...
  MY question...  If I were to lay down 8 radials on 80, 80 on 40 and 8 on 20 (The only 3 bands I'm really interested in), do you think this would give me sufficient performance to rely on it for some medium distance work?  I'm not going for big DX here, but I'd take it if I heard it.. I now have an 800 watt amp in the mix too, so I'm going for as near resonant as I can get.

Jim Lux <jimlux at earthlink.net> wrote:
  At 06:52 AM 6/16/2006, hasan schiers wrote:
>Actually, they can be MUCH shorter than .22 wl, as long as you have enough
>of them. Many more shorter is quite a bit better than many fewer longer.
>There are free programs by G4FGQ, RADIAL2 is the name I think, but just
>google G4FGQ and you will get to his site and free downloads. The radial
>program is a good one and answers all the questions being raised. It lets
>you manipulate number/length/freq and ground characteristics to show
>efficiency. VERY HANDY!

Except that Reg's program isn't necessarily a very good model of 
radials. He's modeling the radial field as a bunch of lossy transmission 
lines (which is sort of fine), but doesn't account for the transfer of 
energy into the surrounding soil, some of which reradiates, as opposed to 
being simply loss in an equivalent loss resistance. Reg has some 
explanation of what he's doing in the notes in the program, but doesn't 
publish his algorithms or source code, so it's tough to figure exactly what 
he's doing (and I'm not patient enough to just run lots of cases and 
reverse engineer it). It's also not clear how Reg is modeling wires close 
to the surface, where a significant part of the field crosses the 
ground/air boundary.

That said, it might be a handy way to tinker around and gain some 
understanding of the effect of various radials.

If you really, really want to model radials, you're probably best off to 
fork out the bucks for a copy of NEC4 and do it right. (NEC4 does model 
buried wires, and has been rigorously validated, and all the gory details 
are published) Otherwise, just use the "throw lots of wire out, don't worry 
about the exact length, and more radials is better than fewer" approach.


>The data I got is completely consistent with the ARRL publications that
>describe how to get minimum loss for a given wire length available, as well
>as with G4FGQ's program. I do think Reg's program is a bit optimistic, but
>it points one in the right direction. I have full faith in the studies
>published in the ARRL Antenna Handbook on how to optimize a radial field, as
>my actual measurements agree very strongly with the article.

More information about the TowerTalk mailing list