Concept below

However this is not an accurate model of a transmitter.

For an example, take an old Heathkit DX-100 generating a measured 180 watts
of CW RF into a matched 50 ohm load. To do this, it does NOT also dissipate
180 watts of RF into some "virtual" internal RF load in the DX-100. In
fact, the PAs and power supply in the DX-100 could not produce a total RF
output power of 360 watts without exceeding their ratings.

The dissipation in the PA is essentially related only the DC to RF
conversion efficiency of the PA, which in this case probably is about 75%,
max (Class C). So a PA input power of about 240 watts DC is required to
produce 180 watts of RF output power. The other 60 watts of plate input
power is converted to heat by the PA tube anodes.

The entire RF output generated by the PA stage is applied virtually 100% to
the output connector. How much of that is absorbed by the load connected
there is a function of load SWR and system losses.

- RF
________________

"alhearn" wrote
If you leave out the complex part of impedences for the moment and
think of 100 volt generator that has a 50 ohm internal impedance
driving a 50 ohm load, current is 1 amp and the power dissipated by
the load is 50 watts. There is also 50 watts dissipated by the
generator's internal impedance, for a total of 100 watts dissipated by
the entire system. Therefore, the "available" power for this generator
is 50 watts.