Ed Cregger wrote:

I am beginning to suspect that traditionally made baluns are not as exact in
practice as they are theoretically. This is not a surprise, really. Few
things in electronics are exact as we humans like to assume, as you well
know.

Good seeing your post, OM.


Ed Cregger, N2ECW former NM2K

The whole problem is "liking to assume" that things are simpler than
they are. When the theory you apply is too simple, guess what -- you'll
find that the real thing doesn't behave as your oversimplified "theory"
predicts. Theory works just fine, and accurately predicts how a real
object will work. Oversimplified "theory" often doesn't work so well.

A well made balun or RF transformer behaves reasonably well like an
ideal transformer, that is, infinite winding inductance, no coupling
capacitance, zero leakage inductance, no loss, and so forth, but only
under quite a narrow range of circumstances. Those circumstances include
being terminated with a fairly narrow range of impedances and over a
limited frequency range. Usually, one side is designed to be terminated
with 50 ohms, purely resistive. That means the other side of a 4:1 balun
has to be terminated with something fairly close to 200 or 12.5 ohms
(depending on how it's designed), also resistive, in order for it to
work as intended. If the impedance differs very much at all from that
value, you'll find that the transformation ratio is no longer 4:1, and
that the balun will add a series and/or shunt impedance to the circuit.
This can be accounted for by theory, but only with great difficulty
since it requires careful characterization of the core and windings.
People tend to design, and often test, a 4:1 balun in a 50 ohm
environment, then attach it to a multiband antenna that has rather
extreme (but entirely predictable) impedance variations. Then they're
surprised because the impedance seen looking into the balun isn't 4
times or 1/4 times the antenna impedance, but is something wildly
different. They shouldn't be. A 4:1 balun or transformer that effects a
nice 4:1 impedance transformation when presented with a very wide range
of termination impedances simply doesn't exist. Any "theory" that
predicts it is oversimplified and invalid.

Roy Lewallen, W7EL