How to Design an HF Linear Amplifier for 400 W PEP Output

| September 25, 2026 | Amateur Radio (Ham Radio) | No Comments

The legal limit for amateur radio operators with General Class license in this region is 400 W PEP on SSB Mode for HF bands. Let us go through the various technical aspects and design concerns.

Designing a 400W PEP HF linear amplifier requires a fundamental shift in architecture compared to building in the 100W class. At this power level, inexpensive switching MOSFETs like the IRFP150 hit their frequency and thermal limits, and relying on RF sensing (carrier-operated relays) for transmit/receive switching becomes highly risky due to relay arcing. That is in comparison with the EVQ linear amplifier which I had homebrewed earlier. To achieve clean, linear SSB amplification across 1.8 to 30 MHz, a modern solid-state design utilizing LDMOS technology is the most reliable approach. Here is the architectural breakdown for a 400W HF linear amplifier.

1. The Active Device: Stepping Up to LDMOS

For 400W, a 50-volt LDMOS transistor is the standard. Operating at 50V keeps the DC current manageable (~14-16 Amps) and makes output impedance matching significantly easier than 12V or 24V designs.

  • Top choices: A single ruggedized LDMOS like the BLF188XR (capable of 1kW+ but extremely clean and virtually indestructible when under-driven at 400W), or a pair of MRF300AN/BN transistors.
  • Biasing (Class AB): To achieve linearity for SSB, the gates require a highly regulated, temperature-compensated bias voltage (usually between 1.5V and 2.5V for LDMOS). The bias circuit must include a thermistor mounted directly on the transistor flange to reduce the gate voltage as the device heats up, preventing thermal runaway.

2. Input Matching and Attenuation

High-gain LDMOS devices require very little drive power—often just 2 to 5 watts to yield 400W out. If you are driving this with a standard transceiver or a digital setup capable of higher output, you must build in protection.

  • Input Pad: Use a massive, non-inductive resistor network (e.g., a 10dB or 20dB pi-attenuator bolted to the heatsink) to absorb excess drive power. This also provides a perfect 50 Ω load to your exciter, stabilizing the system.
  • Input Transformer: A 9:1 or 4:1 impedance transformer (using BN-43-202 or similar binocular cores) steps down the 50 Ω input to match the very low gate impedance of the LDMOS.

3. Output Network and Impedance Matching

The output impedance of the transistors must be matched to your 50 Ω antenna system. You can determine the required drain-to-drain load impedance (Zload) using the formula:

Zload = (Vdd – Vsat)2/(2 x Pout)

Vdd: The positive power supply voltage provided to the circuit.

Vsat (or 𝑉𝑑𝑠,𝑠𝑎𝑡): The saturation voltage. This is the minimum drain-to-source voltage a MOSFET needs to stay in the saturation region (where it acts as a stable current source or amplifier). Below that there will be severe distortion.

For a 50V supply generating 400W, the output impedance is roughly 3 Ω.

  • The Balun / Transformer: To transform this up to 50 Ω, you need a 1:16 impedance transformer (1:4 voltage ratio). This is typically constructed using heavy-duty coaxial cable (like RG-316 or RG-400 with Teflon insulation) wound through large Type 43 ferrite tubes.
  • DC Blocking & Chokes: The 50V DC supply is fed to the drains through a heavy bifilar RF choke to prevent RF from flowing back into the power supply. Large, high-voltage blocking capacitors (ATC or heavy silver mica) keep the 50V DC out of the output transformer.

4. Heavy-Duty Low Pass Filtering (LPF)

A broadband solid-state amp generates massive harmonic content. At 400W, a 3rd harmonic only 15dB down is pushing out 12W of illegal interference.

  • Filter Banks: You need a switched bank of 5- or 7-pole Chebyshev Low Pass Filters. Typical groupings are 160m, 80m, 40m, 20/30m, 15/17m, and 10/12m.
  • Components: Use large powdered iron toroids (T-130 or T-200 sizes, usually Type 2 or Type 6 material). Capacitors must be high-voltage, high-current NP0/C0G ceramics or high-quality silver micas rated for at least 1kV. Standard capacitors will melt under the RF circulating currents.

5. Control, Switching, and Thermal Management

At 400W, the tolerances for error are nearly zero. A millisecond of high SWR or hot-switching can destroy the device.

  • Mandatory PTT: Do not use RF sensing. You must use a hard PTT line from the radio to a sequencer. The sequencer must actuate the output relays, wait a few milliseconds for the contacts to settle, and only then apply the bias voltage to the LDMOS to allow RF amplification.
  • Thermal Dissipation: A standard aluminum heatsink is no longer sufficient on its own. Because an LDMOS die is tiny, it creates a massive thermal bottleneck. You must mount the transistor to a copper heat spreader (at least 10mm thick), which is then bolted to a large, fan-cooled aluminum heatsink.
  • Protection Circuits: Incorporate an SWR bridge at the output. If reflected power spikes (e.g., SWR > 2.5:1), a fast-acting comparator circuit must immediately cut the gate bias, shutting down the amplifier before the transistor fails.

These are few of the theoretical aspects of design of a solid state HF linear amplifier. While practically implementing it, there are many more factors to be considered like RF PCB design, which is quite different from PCB design for audio frequencies or other simple electronic circuits. Just learning a bit hoping that in the long run I will be able to implement a 400 W PEP linear amplifier which has been my dream ever since getting Advanced Grade Amateur radio callsign way back in 1988! In those days, the legal limit for Grade I was was 150 W and that for Advanced Grade 400 W. Maximum which I have been able to achieve in those days was 120 W DC input into my homebrew 3 x 807 vacuum tube set up, running at about 850V!