Re: [TenTec] SLIGHTLY OT

["Dr. Gerald N. Johnson" <geraldj@weather.net>]

Yah it gives consistent numbers but NEC2 and NEC4 are distinctly bad 
about working with close spacings. MININEC does better but its not 
perfect either. In the May/June 1989 issue of QEX Cebik compared the 
three to real world measurements in an article titled: "NEC-4.1: 
Limitations of importance to hams." Subtitled: "Antenna modeling seems 
deceptively easy. Come tour the pitfalls of the latest software." Pages 
3 to 16. The next article in the same issue was by the authors of 
MININEC titled; "Wire Modeling Limitations of NEC and MININEC for 
Windows." Subtitled: "Come listen to the authors of MININEC as they 
describe its workings and compare it to NEC-4." Pages 17 to 21.
On 4/8/2011 4:27 PM, Steve Hunt wrote:
> On 08/04/2011 22:01, Dr. Gerald N. Johnson wrote:
> > Careful about the EZNEC model for impedance. None of the NEC models do
> > coupling to the earth. Their earth model is only used for reflections
> > for the elevation pattern.
> No - it also uses the earth model to generate feedpoint impedances. You
> only have to see the effect that soil type has on the impedance to
> realise that is true. You need to be careful with verticals and
> associated ground systems where it has some shortcomings, but it's
> pretty good on the impedance of horizontal dipoles and derivitives down
> to heights of 0.005 lambda.
>
> > On my Smith chart slide rule, that 11-j930 a long ways from the low R
> > side of the chart spins over a quarter wave to about 4 + j2.75 though
> > the resistive component is practically on the outside edge of the chart.
> I don't trust my Smith chart for those extreme impedances; I suggest you
> don't either because it's giving you the wrong answer - your reactive
> component is grossly in error. Much easier to use the simple maths for a
> quarter-wave transformation:
>
> Zin = Zo*Zo/Zload = 600*600/(11 - j930) = 4.6 + j386
>
>
> 73,
> Steve G3TXQ
73, Jerry, K0CQ