GaN Polar Modulation: Why the 100 W Amp Pitch Falls Short

Updated: March 29, 2026

I watched the YouTube video of Dave G3LRC, but the pitch mixes one very real technology trend with a comparison that does not really stand up technically. GaN polar modulation is absolutely real. It is not hype, and it is not a gimmick. As an integrated transmitter architecture, it is one of the most interesting developments in amateur-radio PA design right now. But comparing a finished multiband 100 W HF amplifier to a bench evaluation board is not a fair buyer’s-guide comparison.

The Xiegu XPA125B, whatever its limitations, is still a complete HF + 6 m product with broadband amateur coverage and an internal tuner. The EPC Class-E board often cited in these discussions is a development platform intended to explore switching PA behavior under tuned-load conditions. Those are not the same category of product, and they do not solve the same problem.

The apples-to-oranges comparison

If you want to argue that a modern GaN switching architecture is superior to a traditional HF linear, then the comparison has to be between finished products that solve the same job. That is where the sales pitch starts to wobble. A lab board may be interesting, elegant, and efficient, but that does not make it a drop-in replacement for a shipping HF amplifier with band coverage, protection, user interface, filtering, and real-world operating convenience.

That distinction matters. A development board can demonstrate what is possible under controlled conditions. A shipping amateur-radio amplifier has to survive mistuning, band changes, reactive loads, user abuse, thermal cycling, and the endless creativity of real stations. “Old tech” is not the same thing as “easy tech.”

Polar modulation, EER, and envelope tracking are not interchangeable

A second problem is that these discussions often blur together polar modulation, envelope elimination and restoration, and envelope tracking as though they were the same thing. They are related, but they are not interchangeable. In polar modulation, the RF path carries phase and the supply path carries amplitude. In envelope tracking, the RF chain still carries the modulated RF signal while the supply path is varied to improve efficiency.

That sounds like a subtle distinction, but it becomes very important once people start talking about a supposed “external amp” that can take arbitrary low-power RF and somehow turn it into a perfect high-efficiency transmitter. In a true polar architecture, the system wants access to envelope and phase information, plus tight timing alignment between those paths. That is not a trivial bolt-on feature.

A true external polar box is not really just an amplifier

This is the biggest conceptual flaw in the whole story. A clean external polar transmitter is generally not just a smarter post-driver box. Once implemented properly, it starts becoming a transmitter subsystem in its own right. It needs coordination, calibration, timing alignment, shaping, protection, and usually some knowledge of what the exciter is doing internally.

That is why the cleanest counterexample is the Polar Explorer concept. It makes the point very clearly: once you do an external polar solution properly, it stops behaving like a generic “5 W RF in, 100 W RF out” black box and starts behaving more like a companion transmitter that follows the host radio. That is possible, but it is not the same thing as replacing a traditional linear with a magical GaN brick.

Why integrated 500 W makes sense sooner than generic external 100 W boxes

GaN polar modulation does not hit a magical wall at 500 W. The real challenge is that broadband, multiband, all-mode, compact amateur products become much harder as power rises. That is why integrated solutions make sense first. When the exciter, modulation path, calibration, filtering, supply design, and protection are all engineered together inside one product, the architecture can shine.

That is also why the first compelling commercial examples appear as integrated transceivers rather than universal external boxes. At moderate power, the efficiency benefit is meaningful and the total system can still be kept manageable. Push much higher, and the entire product problem gets bigger very quickly: supply sizing, protection, filtering, tuner design, thermal margins, connectors, and safety all scale in ways that are far less glamorous than the headline efficiency number.

Efficiency is real, but it is not the whole argument

Efficiency matters. For battery work, field operations, DXpeditions, portable deployments, or generator-powered stations, it matters a lot. A more efficient transmitter means less heat, less current draw, smaller batteries, and more practical operating time. That part of the pitch is absolutely valid.

But the buying argument is still more nuanced than “GaN good, old linear bad.” Device price is only one part of the story. The moment you move into real polar or ET-style transmit architectures, you also need fast supply modulation, calibration, control, protection, and a lot more engineering around the active device. A cheap transistor does not automatically produce a cheap finished product.

The other nuance is that blanket efficiency numbers are often presented as though they apply equally across every band and every operating condition. Real amplifiers rarely behave that way. Band, load, drive level, filtering, and protection margins all influence what the user actually sees in practice.

Signal purity is a system result, not a GaN magic trick

This is where a lot of marketing language falls apart. Signal purity is not something you get automatically because a design uses GaN. It is not something you get automatically because the architecture is polar either. Clean transmit performance comes from the whole transmitter system: the device, the matching network, the modulation method, timing alignment, predistortion, filtering, protection, and how hard the system is being driven.

That is why “switching equals efficient” should never be confused with “switching equals automatically clean.” A badly aligned or poorly filtered high-efficiency architecture can be ugly. A properly engineered linear can be very clean. A properly engineered polar solution can also be very clean. The result comes from execution, not from buzzwords.

The real takeaway

GaN polar modulation is the right direction for integrated HF transmitters, and the technology deserves serious attention. It is one of the few genuinely exciting architecture shifts in amateur transmit design. But that does not mean every traditional HF linear is suddenly obsolete, and it certainly does not mean a generic external “5 W in, 100 W out” polar amp is the obvious next product waiting to happen.

Mature LDMOS linear technology still scales well into higher-power amateur products, and true external polar solutions tend to become much more than simple amplifiers. So yes, the future is real. The simplified buyer’s-guide story, however, is not.

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.