GuidesBuying & Value

GaN vs LDMOS Amplifiers: The New Technology Reshaping Ham Radio Power

August 15, 2026

Last updated: August 15, 2026

For the past 20 years, LDMOS (Laterally Diffused Metal Oxide Semiconductor) transistors have dominated solid-state ham radio amplifier design. Every modern solid-state kilowatt amp — from the Expert/SPE to the Elecraft KPA-1500 to homebrew pallet amplifiers — uses LDMOS devices.

Now GaN (Gallium Nitride) is arriving in ham radio. It promises higher efficiency, smaller size, and better high-frequency performance. But the hype needs context. Here's what GaN actually offers and what the practical implications are for ham radio operators.

What Makes GaN Different

GaN is a wide-bandgap semiconductor. Compared to silicon LDMOS:

Higher breakdown voltage per unit area: GaN devices handle higher voltages in smaller die sizes. A GaN transistor can operate at 50V or even 65V drain voltage while being physically smaller than an equivalent LDMOS device at 50V.

Higher power density: More watts per square millimeter of silicon. This means smaller packages for the same power output.

Better high-frequency performance: GaN has higher electron mobility and lower parasitic capacitance. It works efficiently at frequencies where LDMOS devices struggle. This is why GaN dominates the 5G base station market (frequencies up to 6 GHz).

Higher efficiency: GaN amplifiers achieve 70-80% efficiency in some operating classes, compared to 50-65% for typical LDMOS designs. Less wasted heat means smaller heat sinks and less cooling infrastructure.

Faster switching: Lower capacitance means GaN devices switch faster, which matters for broadband designs and higher-order operating modes (Class F, Class J) that boost efficiency.

LDMOS: The Incumbent Technology

LDMOS devices from NXP (BLF188XR, MRF1K50H) and others have been refined over decades for exactly the ham radio use case: 1.8-54 MHz, 50V operation, kilowatt power levels.

Strengths of LDMOS for ham radio:

  • Extensively characterized and documented for HF use
  • Massive application note library (NXP/Freescale has published hundreds of HF amplifier designs)
  • Well-understood failure modes and protection techniques
  • Available in packages specifically designed for HF push-pull amplifiers
  • Lower cost per watt at HF frequencies (mature manufacturing)
  • Rugged — handles moderate SWR mismatch without instant failure
  • Proven in thousands of ham amplifier designs

Weaknesses:

  • Efficiency typically 50-60% at HF in Class AB (40-50% of input power becomes heat)
  • Large die size means significant output capacitance (limits bandwidth and high-frequency performance)
  • Package parasitics limit performance above 200-300 MHz
  • Requires substantial heat sinking for kilowatt operation

GaN in Ham Radio: Where We Are Now

As of 2026, GaN is appearing in ham radio applications:

Commercial HF amplifiers:

  • A few manufacturers are releasing GaN-based HF amplifiers. The efficiency advantage means smaller, lighter amplifiers that run cooler.
  • Expert/SPE has GaN models in development
  • Several Chinese manufacturers offer GaN amplifier modules for the homebrew market

VHF/UHF amplifiers:

  • GaN's greatest ham radio advantage is above 50 MHz. On 2m and 70cm, GaN devices significantly outperform LDMOS in efficiency and gain.
  • Several 2m amplifiers using GaN have appeared (300-1000W class)

QRP and portable:

  • GaN's efficiency means less battery drain for the same output power. A GaN-based QRP radio could potentially run 20% longer on the same battery compared to LDMOS/silicon.
  • The size reduction allows smaller amplifier modules

Microwave:

  • For 23cm (1296 MHz) and above, GaN is already the clear technology choice. LDMOS barely functions at these frequencies. GaN thrives.

Practical Comparison: 1kW HF Amplifier

LDMOS Design (BLF188XR or MRF1K50H)

  • Efficiency: ~55% at 1kW output = 1.8kW DC input, 800W dissipated as heat
  • Power supply: 50V at 36A peak = substantial transformer or switching supply
  • Heat sink: Large (800W to dissipate). Forced-air cooling with significant fan(s)
  • Size: Typically 19" rack width, 5-6" tall, 12-16" deep
  • Weight: 20-40 lbs depending on power supply type
  • Cost of devices: ~$100-200 for the transistor pair
  • Bandwidth: 1.8-54 MHz typically with 4:1 output transformer

GaN Design (Equivalent Power)

  • Efficiency: ~70% at 1kW output = 1.4kW DC input, 400W dissipated as heat
  • Power supply: 50V at 28A peak (smaller supply needed)
  • Heat sink: Half the size (400W vs 800W dissipation)
  • Size: Potentially 30-40% smaller overall
  • Weight: 30-50% lighter
  • Cost of devices: ~$200-500 (GaN devices cost more currently)
  • Bandwidth: Potentially wider (lower output capacitance)

The efficiency difference is dramatic: 400W less heat means smaller heat sinks, smaller fans (quieter!), smaller power supply, and smaller enclosure. The entire amplifier shrinks.

Why GaN Hasn't Fully Taken Over Yet

Cost: GaN devices cost 2-5x more than equivalent LDMOS devices. For a 1kW amp, the transistor cost goes from $150 to $400-600. This premium will decrease over time as manufacturing scales, but today it's real.

Fragility: GaN devices are less forgiving of mismatch than LDMOS. An LDMOS amp can survive brief SWR excursions (antenna failure, accidental disconnect) that would kill a GaN device. Protection circuitry must be faster and more robust.

Design complexity: GaN's higher gain and broader bandwidth make stability more challenging. Parasitic oscillations are easier to trigger. The amplifier designer must be more careful about layout, grounding, and bypass capacitor placement.

Thermal management differences: While GaN produces less total heat, the heat is concentrated in a smaller area (higher power density). Thermal management must efficiently extract heat from a smaller die. Poorly designed thermal interfaces can lead to thermal runaway.

Supply voltage: Some high-power GaN devices prefer 65V operation (versus LDMOS's standard 50V). This requires different power supply design and higher-voltage capacitors throughout.

Lack of application notes for ham use: LDMOS has decades of published HF amplifier designs. GaN application notes focus on cellular/5G frequencies (2-6 GHz). Ham radio designers are adapting these for HF, but it's newer territory.

What This Means for You (The Buyer)

If You're Buying an Amplifier Now

LDMOS amplifiers are proven, available, and well-supported. The KPA-1500, Expert 2K-FA, OM Power 2501, and similar are excellent amplifiers with years of field experience. Buy with confidence.

GaN advantages (size, weight, efficiency) matter most for:

  • Portable/field use where weight and battery matter
  • VHF/UHF where GaN's frequency advantage is most significant
  • Stations where heat/cooling is a constraint

For a home station HF amplifier, the practical difference between 55% and 70% efficiency is meaningful (lower electric bill, less AC load, quieter fans) but not revolutionary.

If You're Building (Homebrew)

Stick with LDMOS for now unless you have RF design experience. The NXP BLF188XR push-pull amplifier circuit is well-documented, forgiving of minor errors, and produces 1kW easily. Dozens of proven designs exist with published schematics, board layouts, and construction guides.

GaN homebrew amplifiers are absolutely possible but require more careful design and faster protection circuitry. Wait until more ham-specific application notes and proven PCB layouts appear.

The Future (2-5 Years)

GaN will become standard in new amplifier designs as:

  • Device costs decrease (manufacturing volume from 5G market drives this)
  • More ham-specific designs and application notes are published
  • Protection circuitry designs mature
  • Commercial ham amps prove reliability in the field

The transition will mirror what happened with LDMOS replacing bipolar transistors in the 2000s — gradual, inevitable, and eventually complete.

For VHF/UHF Operators: GaN Now

If you operate 2m, 70cm, or especially 23cm, GaN is already the superior choice:

  • Higher gain at VHF/UHF means simpler circuits (fewer stages)
  • Better efficiency means less heat (critical for tower-mounted amps)
  • Higher frequency capability extends naturally to 1296 MHz and above
  • Most new commercial VHF/UHF amplifiers will be GaN-based going forward

The first place most hams will encounter GaN in practice is in 2m/70cm amplifiers and in multimode portable radios where efficiency directly extends operating time.

Bottom Line

GaN is better technology. It will eventually replace LDMOS in most ham radio power applications. But "eventually" isn't "today" for most HF operators. Buy what's proven and available now. If you're designing a new VHF/UHF amplifier, strongly consider GaN. For HF, LDMOS remains king in 2026 but the transition is beginning.

Frequently asked questions

Is GaN worth paying extra for in an HF amplifier?
For most home HF stations, not yet. The efficiency gain from 55 to 70 percent means a lower electric bill and quieter fans, but it is not transformative. Proven LDMOS amplifiers such as the KPA-1500 or Expert 2K-FA have years of field experience behind them.
Where does GaN clearly beat LDMOS today?
Above 50 MHz. On 2 metres and 70 centimetres GaN offers better gain and efficiency, which matters especially for tower-mounted amplifiers where heat is hard to remove. At 1296 MHz and above LDMOS barely functions, so GaN is the practical choice.
Are GaN amplifiers more fragile?
They are less forgiving of mismatch. An LDMOS amplifier can often survive a brief SWR excursion from an antenna failure or accidental disconnect that would destroy a GaN device, so protection circuitry has to react faster.
Should I build a homebrew amplifier with GaN?
Not unless you have real RF design experience. GaN has higher gain and wider bandwidth, which makes parasitic oscillation easier to trigger and demands care with layout, grounding and bypassing. The LDMOS route has dozens of proven designs with published schematics and board layouts.
Why does GaN need different thermal design if it produces less heat?
Because the heat is concentrated in a smaller die, so power density is higher. Total dissipation drops but extraction from a small area becomes harder, and a poor thermal interface can still lead to thermal runaway.
Will GaN replace LDMOS eventually?
Most likely yes, as device costs fall on the back of 5G manufacturing volume and as ham-specific designs and protection circuits mature. The transition is expected to mirror LDMOS displacing bipolar transistors in the 2000s: gradual, then complete.

Community field notes

No community notes yet. Be the first to share how this performs in your shack.

Log in to share your field notes and photos.