Hi Kai,



Not quite, at least not when it comes to circuit boards.

First, anytime you accelerate an electron, i.e. change either its speed or direction, it produces electromagnetic radiation.

Now consider a trace on a circuit board running over a GND plane, and in particular think about the return current associated with that trace. The return current is not going to take the path of shortest route. It's going to take the path of least impedance. And the impedance of the ground plane, relative to the trace, is always least directly beneath the trace, at the minimum distance from the trace. Thus, the return current associated with a particular trace will always follow a path in the GND plane that is essentially directly under the trace. (Yes, there will be some spreading out of the return current, fringing the edges if you will, depending on stuff like current density, but that isn't really germane to my point.)

Now imagine that the trace crosses a short slot in the GND plane. Suddenly, in that acutely localized region, the return current must pass around the end(s) of the slot. In this very localized region the return current experiences an abrupt, albeit short lived, increase in impedance. This impedance impulse is the cause of the sudden and localized acceleration (slowing and steering followed by speeding back up) of the electrons and is thus the cause of the radiation.

So you can say that the slot is radiating, the GND plane around the slot actually, but it won't do so without the trace crossing the gap. Reroute the trace so that it doesn't pass over the slot and your radiation will disappear. This is a handy thing to understand if you ever have to deal with a split plane.

Joe