Reflected Power Does Not Automatically Re-Reflect

“Reflected power just bounces between the antenna and tuner until the antenna finally uses it.” That explanation can illustrate one special case, but it is not a general rule.

Re-reflection is possible. It occurs when a reverse-travelling wave reaches another impedance discontinuity and that boundary has a non-zero reflection coefficient. But it is not automatic, it is not always complete, and the same watts are not repeatedly added to the transmitter’s output as though new power has been created.

The popular bouncing-power story mixes a transient picture with continuous-wave steady state. It also adds power numbers where the physics requires complex voltage and current amplitudes to be added first.

Why the Bouncing-Power Story Is So Popular

The story usually goes like this:

1. The transmitter sends 100 W toward the antenna.
2. The antenna accepts 75 W and reflects 25 W.
3. The tuner or transmitter sends that 25 W back toward the antenna.
4. The antenna accepts 75% of it and reflects the remainder again.
5. The process repeats until almost all 100 W has been accepted.

This sounds convincing because a geometric series can be made to add up:

As a simplified energy-accounting exercise involving separated pulses, a lossless line and a particular source boundary, that picture can be useful. The mistake is turning it into a universal description of a continuously driven RF system.

The calculation silently assumes all of the following:

• the returning wave is completely re-reflected at the source end;
• the re-reflected wave has the required phase;
• the line, tuner and source-side network are lossless;
• the load reflection coefficient stays unchanged;
• the wave encounters can be treated as separate packets; and
• the forward-power reading remains fixed while returned energy is added to it.

Those conditions are not automatically true. The last assumption is especially troublesome: once a returning wave interacts with the source or tuner, the total forward wave at that reference plane has changed. The original forward reading cannot simply be retained and then supplemented with another count of the returned watts.

Re-Reflection Depends on the Source-Side Boundary

A load reflection is controlled by the load reflection coefficient:

When that reflected wave reaches the source end, its fate is controlled by a different quantity: the source-side reflection coefficient looking into whatever termination or network exists there.

For a passive linear source termination, three broad cases are easy to see:

Source-side condition	Reflection coefficient	What happens to the returning wave
Matched termination	ΓS = 0	The returning wave is absorbed. It is not re-reflected.
Partial mismatch	0 < |ΓS| < 1	Part is re-reflected and part is absorbed or dissipated.
Ideal fully reflective boundary	|ΓS| = 1	The wave is fully re-reflected, with phase set by the boundary.

Therefore the absolute statement “reflected power re-reflects” is no better than the absolute statement “reflected power is absorbed by the transmitter.” Either can occur. The source-side impedance and network determine which description applies.

A Real Transmitter Is Not Automatically a 50 Ω Termination

A radio specified to deliver power into 50 Ω should not automatically be modelled as a passive 50 Ω resistor looking back into its antenna socket.

“Designed for a 50 Ω load” describes the intended external operating condition. It does not prove that the active PA presents a linear 50 Ω source impedance to a returning wave under all drive levels, phases, frequencies and protection states.

Between the feedline and the active device there may be:

• an output transformer or matching network;
• a low-pass filter;
• a tuner;
• relays and directional couplers;
• protection and foldback circuitry;
• a combiner; or
• a circulator or isolator with a dump load.

A returning wave encounters the complete source-side RF boundary, not an imaginary universal 50 Ω resistor. In a practical driven PA, source reflection behaviour may also be power- and phase-dependent, so a simple small-signal ΓS model has limits.

Multiple Reflections Are Real, but Conditional

If both ends are mismatched, multiple reflections can occur. After one full round trip, a wave contribution is multiplied by the load reflection coefficient, the source reflection coefficient and the propagation factor for travelling down and back along the line:

Here γ includes both attenuation and phase, while l is the line length. This compact expression exposes what the bouncing-ball story hides:

• the load may reflect only part of the wave;
• the source may reflect only part of the returning wave;
• the line attenuates both journeys;
• every round trip adds phase delay; and
• later contributions can reinforce or oppose the existing wave.

If the transmitter launches a short pulse into a line whose delay is long compared with the pulse width, the separate echoes can be observed. In that time-domain case, drawing individual trips is physically intuitive.

With a continuous carrier, however, the later contributions overlap. After the transient has settled, the system is better described by steady forward and reverse wave amplitudes, not by trying to label which cycle of RF is on its third or seventh trip.

Add Complex Amplitudes Before Calculating Power

This is the central mathematical correction.

Same-frequency waves travelling in the same direction combine coherently. Their voltages have magnitude and phase. If a source-launched contribution and a re-reflected contribution travel forward together, the total forward voltage is:

Power is then calculated from the magnitude squared of the total amplitude. In general:

The missing cross-term contains the phase relationship. Two contributions can reinforce, partially cancel or, at a particular location, nearly cancel. This is why adding a list of wattages without their phases is not a general continuous-wave solution.

The “same power keeps coming back and gets counted again” story treats coherent waves like independent invoices. RF does not work that way.

The 100 W and 25 W Trap

Suppose a directional meter at one reference plane reads:

The 25 W reverse reading is already part of the electromagnetic state at that plane. It cannot be sent forward again while the 100 W forward reading is held fixed unless an additional boundary condition or source contribution changes the wave solution.

If a lossless tuner establishes a matched transmitter input while the antenna-side line still has a reflected-power fraction of 25%, a valid steady-state result for 100 W net transfer could instead be:

This does not mean the transmitter generates 133.3 W and receives a free extra 33.3 W. It means the travelling-wave decomposition on the mismatched line contains larger forward and reverse components whose difference is the 100 W of net average power crossing the plane.

The directional meter counts the total forward component once and the total reverse component once. It does not label individual watts by trip number.

Re-Reflection Does Not Make Loss Disappear

Reflection itself is not dissipation, but real systems are not lossless.

Every extra trip through real coax adds conductor and dielectric loss. Tuners have finite inductor Q, capacitor loss and contact resistance. Transformers, ferrites, relays, connectors, traps, loading coils and antenna conductors dissipate power. The antenna may also accept power that becomes ground or structural loss instead of radiation.

Therefore “it is not lost because it will be re-reflected” is incomplete for two reasons:

1. The wave may not be re-reflected at the source boundary.
2. Any energy that does make additional trips encounters real loss each time.

In a strictly ideal lossless model, reflection does not dissipate energy. That is a useful limit for understanding the fields. It is not evidence that high SWR causes zero additional loss in real coax.

What a Tuner Actually Does

A tuner transforms the impedance presented to the transmitter. When installed at the shack, it can give the PA a suitable input match while the line between the tuner and antenna still carries substantial forward and reflected waves.

Describing the ideal tuner as re-reflecting the returning wave can be a valid travelling-wave interpretation. But the complete statement is that the tuner changes the source-side boundary conditions so the combined voltages and currents present the required impedance at its input.

The tuner does not:

• create power from the returning wave;
• allow the same watts to be added repeatedly as new output;
• remove line loss;
• guarantee that all accepted antenna power is radiated; or
• eliminate antenna-side SWR unless it is located at the antenna feedpoint.

Real tuner losses also remain real, regardless of how elegantly the ideal network is described.

When the Bounce Picture Is Useful

The bounce diagram remains useful in the right context:

• time-domain reflectometry;
• short pulses on electrically long lines;
• explaining propagation delay;
• building the transient solution before steady state; and
• showing why source and load reflection coefficients both matter.

It becomes unreliable when individual bounce powers are added without phase, when source absorption is ignored, when loss is assumed away without saying so, or when the final steady-state meter readings are mixed with the first launched wave.

The Better Mental Model

Do not imagine one packet of watts carrying a passport that is stamped on every journey.

Instead, define the reference plane and ask:

• What are the total forward and reverse wave amplitudes here?
• What are the load and source reflection coefficients?
• What phase and attenuation does the line add?
• Where is real power dissipated?
• What net average power crosses this plane?

For a lossless line with a real characteristic impedance, the last question is answered by:

That accounting is complete at the stated plane. No trip counter is required.

In Summary

Reflected power does not automatically re-reflect. It is re-reflected only when the returning wave encounters a source-side mismatch, tuner boundary or other discontinuity with a non-zero reflection coefficient.

Multiple reflections are real and are especially easy to observe with pulses. But in continuous-wave steady state, all same-frequency contributions overlap and combine coherently. Voltage and current amplitudes must be added with phase before power is calculated.

The familiar geometric story in which the same 25 W repeatedly returns and is added back to an unchanged 100 W forward reading is not valid general power accounting. Once the returned wave interacts with the source boundary, the total forward and reverse wave solution changes.

Reflection is not automatically loss. Re-reflection is not automatically recovery. And neither process creates free power. The correct result always comes from the complete source, line, load, phase and loss conditions at a defined reference plane.

Want more technical RF content? Subscribe for new deep dives and lab notes.

Have a question or field observation? Contact RF.Guru.