>But this applies to a well behaved plane wave or TEM propagation, I th=
ink.
>When you break up the ground plane, your wave is no longer strictly TE=
M.
>It may be close enough to assume TEM, though, like we often do. If yo=
u
>look at velocity as 1/sqrt(L*C) then gridding a ground plane should in=
crease
>L and decrease C. Which one changes more with the gridding effect wou=
ld
>dictate the change in velocity. Again, at the dimension that we work =
with,
>this should be a very slight change. I'm still working on the bigger =
problems.
Some work I did long ago with some crude dc simulations on 50/50 grids
suggests that the dc resistence at 0, 45, and 90 is the same. The path=
length is longer at 45, but the copper is used more effectively. I'd
expect the same thing on GHz type signals.
=