Skip to content
Article 02 — Station Power

DC Power Supplies for Amateur Radio:
Linear vs. Switch-Mode

A rigorous, myth-free comparison of the two dominant power supply architectures — what they are, how they differ, and how to make an informed choice for your station.


Power demands of the modern HF/VHF/UHF transceiver

Amateur radio transceivers place demanding requirements on their DC power sources. A typical 100-watt transceiver draws in excess of 20 amperes during transmit — a load that the power supply must sustain reliably, with minimal noise contribution to the receive and transmit signal chain.

Two fundamentally different supply architectures are available to the amateur operator: the linear regulated supply and the switch-mode power supply (SMPS). Each carries a distinct set of engineering trade-offs, and neither deserves blanket praise or condemnation without context.

How they work

Linear power supplies

A linear supply begins with a large laminated-iron transformer, which steps the mains AC voltage down to a level close to the desired output. That low-voltage AC is then rectified to pulsating DC, filtered with bulk capacitance to reduce ripple, and passed through a series-pass voltage regulator — typically a bank of power transistors operating in their linear (non-saturating) region — to produce a stable, tightly regulated DC output.

The operative word here is linear: the series-pass element dissipates the difference between input and output as heat. This is inherently inefficient, but it produces exceptionally clean DC — typically less than 5 mV of output ripple under full load. The penalty is size and mass. That large 50/60 Hz transformer is heavy, and there is no engineering shortcut around that physics.

Switch-mode power supplies

An SMPS takes a fundamentally different approach. The incoming mains AC is rectified and filtered directly to high-voltage DC — typically 160–340 VDC depending on the mains standard. That high-voltage DC is then switched on and off at high frequency — commonly 50 kHz to several hundred kilohertz — through the primary of a much smaller ferrite-core transformer.

The key principle is this: transformer size scales inversely with operating frequency. A transformer designed for 100 kHz operation can be orders of magnitude smaller and lighter than one designed for 50/60 Hz, while handling comparable power. The result is a power supply that can be 50–80% smaller and lighter than a linear supply of equivalent output rating — often at significantly lower cost.

Size, weight, and cost: a practical comparison

SpecificationAstron RM-60M (Linear)TekPower TP50SW (SMPS)
Output current50 A continuous / 55 A peak50 A maximum
Dimensions5¼ × 19 × 12½ inches14 × 10.1 × 5.8 inches
Weight60 lbs (27.2 kg)6 lbs (2.7 kg)
Retail price~$589.95~$159.99
Output ripple< 5 mV< 100 mV

Switching at kilohertz frequencies rather than 50/60 Hz yields a roughly ten-to-one reduction in weight and approaching a four-to-one reduction in price for comparable output current. For portable operations, expeditions, or installations where rack space and load-bearing capacity are concerns, this is not a trivial difference.

The noise question

Output ripple

The Astron RM-60M specifies less than 5 mV of output ripple at rated current. The TekPower TP50SW specifies less than 100 mV. Modern amateur transceivers incorporate substantial onboard filtering — series chokes, wide-frequency-range bypass capacitors, and regulated sub-supplies for sensitive receive stages — and a well-designed radio is expected to reject a reasonable level of supply-borne ripple without measurable degradation.

Conducted and radiated switching noise

The more significant concern with SMPS designs is not output ripple per se, but the generation of high-frequency switching noise. The primary switching transistors handle large currents that transition very rapidly — high di/dt events — producing rich harmonic content and potential for ringing and parasitic oscillation. These transients can appear as both conducted noise (propagating back onto the DC output and the AC mains) and radiated interference (directly off PCB traces, transformer windings, and interconnecting wiring).

Regulatory limits in context

Switching power supplies must comply with conducted emissions limits defined by the FCC (Part 15) and CISPR 22 / EN 55022. However, there are no radiated emission limits below 30 MHz in either framework — meaning the entire HF amateur spectrum is covered only by conducted emissions testing.

Key insight

Converting the CISPR Class B limit of 316 µV to RF signal levels in a 50-ohm system yields −57 dBm — approximately S9+10 to S9+20 dB on a calibrated S-meter. A fully compliant, legal product can produce noise exceeding S9 on the HF bands.

StandardClass A limitClass B limit
FCC Part 15 (~7 MHz)3,000 µV250 µV
CISPR / EN 550221,000 µV316 µV (avg)

The role of build quality

The noise performance of any given SMPS is almost entirely a function of design investment. A well-engineered supply will incorporate aggressive line input filtering, snubber networks across switching transistors, output filtering optimized across a wide frequency range, careful PCB layout, and shielding to contain radiated emissions.

It is worth noting that linear supplies are not immune from the noise environment either. A linear supply typically presents the mains directly to the transformer primary through little more than a fuse and possibly a MOV. Mains-borne interference generated by other devices in the building may conduct through to the DC output relatively unimpeded. A high-quality SMPS with a well-designed input filter may actually provide better rejection of mains-borne noise than a linear supply in a noisy AC environment.

Conclusion

The binary framing of "linear supplies are quiet; switch-mode supplies are noisy" is technically imprecise and operationally unhelpful. Both architectures can produce excellent results; both can produce poor results. The determining factors are design quality, manufacturing standards, and the specific application.

Bottom line for the amateur operator
  • A quality SMPS from a reputable manufacturer with proper input filtering, snubber networks, and shielded construction is afully viable choice for station use.
  • A low-cost, unbranded, or poorly specified SMPS is a legitimate noise risk and should be avoided at the operating position.
  • Do not dismiss an entire supply architecture based on the worst examples of its implementation.
  • In this domain, as in most — you reliably receive what you pay for.