Q: why are impedances strange with type 5 sources?

  I'm hoping that someone with more experience with NEC-2 can clue
me in. I'm probably missing something very obvious..

  I did some modeling of some collinear arrays, and got some
suspicious looking feedpoint impedances. To try to understand
what was going on, I backed away from the problem and modeled
something real simple: a resonant half-wave dipole in free
space: (at 299.8 MHz).

  My problem is that I get bizarre feedpoint impedance
(and a slightly different resonance) when I use a type
5 ("current slope discontinuity") voltage source, instead
of a type 0 (applied E field) source. I would expect the
results to be the almost the same. What am I missing?

  Here's a chart of feedpoint impedance vs number of
segments when using a type 5 source:

tot Source Impedance
segs segment
10 6 88.6 -j0.55
12 7 90.6 -j0.34
14 8 91.9 -j0.37
16 9 92.4 -j0.53
18 10 91.9 -j0.76
20 11 90.2 -j1.11
22 12 86.8 -j1.45
24 13 81.4 -j1.77
26 14 73.4 -j2.05
28 15 62.0 -j2.29
30 16 46.7 -j1.09
32 17 26.2 -j1.58

  I can perhaps understand why it goes berserk above 24 segments, as the
segment length is approaching the diameter (.01) of the wire. Oddly,
I don't see this problem until many more segments when the type 0 source
is used, and even then not as dramatically:

tot Source Impedance
segs segment
9 5 71.3 -j.87
11 6 71.6 -j.55
13 7 71.8 -j.38
15 8 72.0 -j.30
17 9 72.1 -j.20
21 11 72.3 -j.37
23 12 72.4 -j.34
41 22 72.7 -j.93
61 31 72.5 -j3.5

   More importantly, why does it seem to indicate around 90 ohms?

Here's my input deck..

CM half wavelength dipole
CM 20 segments type 5 Voltage source at seg 11
CE
GW 1 20 0.00000 0.00000 0.00000 0.00000 0.00000 0.46200 0.00500
GE
FR 0 0 0 0 299.80000 0.00000
EX 5 1 11 10 1.00000 0.00000
XQ 0
EN

  Many thanks if you can help..

--
Charlie Panek KX7L              Hewlett Packard Company
charlier_at_lsid.hp.com            Lake Stevens Instrument Division
                                 Everett, Washington