W5DXP wrote:
The answer has been found but most people think they already know
everything there is to know. The answer is contained in the following
example. The sources are signal generators with circulator loads (SGCL).
The (50) or (150) subscript is the ohmic value of the circulator load
resistor. The signal generators are phase-locked and can be turned on
and off independently.

100W SGCL(50)--1WL 50 ohm feedline--+--1/2WL 150 ohm feedline--33.33W SGCL(150)
Pfwd1-- Pfwd2--
--Pref1 --Pref2

Step 1: With the 100W SGCL(50) on and the 33.33W SGCL(150) off, the
following conditions exist:

Pfwd1 = 100W, Pref1 = 25W, Pfwd2 = 75W, Pref2 = 0W

Step 2: Now turn on the 33.33W SGCL(150). At the instant the rearward-
traveling signal reaches the impedance discontinuity, the following
conditions exist at the discontinuity:

Pfwd1 = 100W, Pref1 = 0W, Pfwd2 = 133.33W, Pref2 = 33.33W

All anyone has to do is figure out what happened to Pref1 = 25W in
Step 1 the instant the 33.33W arrived at the impedance discontinuity
in Step 2. This is not a transient buildup condition. If we ignore
any distortion in the 33.33W Pref2 wavefront, this is an immediate
event and Pref1 is immediately canceled by an equal magnitude and
opposite phase Pref2(1-|rho|^2) wavefront at the moment it first
arrives.

You're sourcing and sinking an additional 33.33 watts, and yet the
wattmeter can't discern the difference between this scenario and the 100
watt, single source scenario.

The example illustrates perfectly the shortcomings of the idea of power
flow, as well as some of the faulty conclusions that can be drawn from
measurements made by a directional power meter.

73, AC6XG