On Aug 11, 1:37 am, Owen Duffy wrote:
K7ITM wrote roups.com:

...

Tom, for avoidance of doubt, I am not talking about the type of directional
coupler that uses a couple line and that you would terminate with matching
load. I am talking about the cheap VSWR meters that have about 100mm long
coupled line, that is quite tightly coupled, and the resistor at one end of
the line is adjusted to balance the electric field sample with the magnetic
field sample for a null reading with V/I=Zn.

Owen

Hi Owen,

I'm not sure I see the difference. The load on the cheapie you
describe is just the load required to terminate that line. I have a
freely redistributable field solver program that will calculate the
even and odd mode impedances for you from the geometry and the
dielectric's permittivity, and from those impedances and the length
you can predict the proper termination impedance of both the "through"
and the "coupled" lines, and the coupling at any particular
frequency. It IS a problem if you try to do it in microstrip because
the propagation velocity for the even and odd modes is different, but
in true TEM configurations, I believe the directionality is fine if
you maintain uniform cross-section. Actually, a way that they make
broadband coupled lines is to have a central section tightly coupled,
and another section on each end of that which is less tightly
coupled. You can extend it to 5 sections or more, to get even broader
bandwidth.

Info about them is out there, but it wasn't as easy for me to find as
I figured it would be. ;-)

Cheers,
Tom