When Entertainment Outruns Engineering

A Reality Check on Polar Modulation, Old Parts, and End-Fed Antennas

I recently watched an interview on Dave G3LRC’s channel featuring Josef Hoffman, KD7QOW. Dave seems like a likable host, and I understand why these videos get attention. Curiosity is welcome. The problem starts when curiosity gets presented as settled engineering, because that is exactly when newcomers walk away more confused than informed.

This particular discussion mixed together polar modulation, switching PA classes, “old” components in commercial radios, and end-fed return currents as if they all belonged in the same drawer. They do not. Each topic has its own engineering context, its own trade-offs, and its own limits. Once those boundaries are blurred, the conversation may still be entertaining, but it stops being a reliable technical reference.

Video discussed: Dave G3LRC interviewing Josef Hoffman, KD7QOW.

Polar Modulation Is Not a Magic External Box

The first point that needs cleaning up is polar modulation itself. Polar modulation is not a magic external amplifier that you simply bolt onto the back of an ordinary 100 W transceiver. That is a misunderstanding of the architecture. A polar transmitter does not want the same kind of low-level composite RF drive that a conventional external linear amplifier expects. It wants the signal split into phase and amplitude information, handled as part of the transmitter system itself.

That matters because it kills the lazy version of the argument right away. Once the architecture depends on tightly coordinated handling of carrier and envelope information, the idea of a universal “just add this polar amp to your existing rig” box becomes far less realistic. The integration is not an optional detail. It is the design.

QMX and Aurora Belong to the Same Family, Not the Same Drawer

That is also why the QRP Labs QMX example does not contradict the point. Hans Summers’ QMX SSB implementation is indeed part of the broader polar or EER family. The phase and amplitude paths are handled separately, and a non-linear switching PA is used in a way that fits that architecture. But family resemblance is not the same thing as implementation equivalence.

A clever low-power SDR doing this at QRP level is one thing. An integrated 500 W commercial transmitter is another. Same broader family, yes. Same scale, same product constraints, same thermal problem, same integration burden, or same market case? Absolutely not. That is exactly where too many online discussions collapse into slogan thinking.

Why the Payoff Gets More Serious at Higher Power

Polar modulation becomes a far more compelling story when the transmitter power gets high enough that heat, power supply size, cooling, and physical volume become real design problems. At 500 W, efficiency is no longer a nice bullet point for the brochure. It directly changes how large the box needs to be, how much waste heat you must remove, and how practical the complete station becomes.

At 100 W, the complexity does not vanish, but the payoff is smaller. That does not mean a 100 W polar transmitter is impossible. It means the engineering and business case is less dramatic. The architecture still asks for tight integration, but the thermal and packaging rewards are less transformative than they are at 500 W. That is a much better explanation than the throwaway claim that mainstream manufacturers simply “do not understand.”

That is an engineering inference about trade-offs, not a claim that 100 W implementations are impossible.

Using Mature Parts Is Often the Sensible Choice

The complaint that Icom, Yaesu, or Kenwood use “older components” usually comes from people who have never had to ship a product in volume. Once a device is selected, qualified, stocked, serviced, laid out on a board, and bought in real quantities, changing it is not a casual fashion decision. It is a redesign event.

A mature part with known behavior, stable availability, field history, known thermal margins, known EMC behavior, and established service stock is often the rational engineering choice. Replacing it with something newer can ripple through the PCB layout, thermal validation, compliance work, firmware assumptions, repair logistics, and procurement chain. That is not a sign of incompetence. It is what responsible product engineering looks like.

In other words, you do not redesign a shipping platform every time somebody spots a newer transistor on a distributor website. “Newer” and “better product choice” are not the same sentence.

Class D and Class E Are Not Interchangeable Labels

The class-D versus class-E part of the discussion is another example of hand-waving replacing technical clarity. Yes, both are switching topologies. No, that does not make them interchangeable labels for “efficient amplifier.” They do not solve the problem in the same way, and they do not place the same stress on the devices or the surrounding network.

Once you understand that, the lazy habit of throwing class D and class E into one bucket starts to fall apart. In this context, saying “it is a switching PA” is only the beginning of the sentence. You still need to ask which topology, what device voltage and current stress it creates, what matching network it expects, what frequency range it is intended for, and how the modulation method interacts with it.

Efficient does not mean identical. That distinction matters.

End-Fed Antennas: The Return Path Is the Real Story

The EFHW discussion suffered from the same oversimplification. Yes, every antenna system needs a return path. That is not controversial. But the statement that “an antenna needs two wires” is only the teaching version of the story. In real installations, the actual return may be a dedicated counterpoise, the outside of the coax shield, stray capacitance, or some combination of all three.

That is why Werner Schnorrenberg, DC4KU, and many practical measurements keep pointing to the same conclusion: the coax can become part of the counterpoise, and choke placement changes how much of the feedline participates. That is also why end-fed systems can appear more forgiving about choke placement than many operators expect. The system may still work because the feedline is quietly doing part of the job.

Until you deliberately force that outer-surface current to stop, the feedline is not a neutral spectator. It is part of the antenna system. That is the real reason slogans about “two wires” or “just add a counterpoise” are usually too shallow to be useful in actual installations.

Entertainment Is Not Engineering

None of this means the interview was malicious. It was entertaining, and amateur radio needs enthusiastic people. But entertainment is not the same thing as engineering. These videos are best treated as conversation starters, not as technical references.

By all means, enjoy them. Just do not confuse a lively discussion with a validated design explanation. The funny part is that their real output power does not show up on a wattmeter. It shows up in my mailbox.

Interested in more technical content? Subscribe to our updates for deep-dive RF articles and lab notes.

Questions or experiences to share? Feel free to contact RF.Guru.