In the world of high-fidelity personal audio, the quest for the “blackest background”—a state of absolute silence when no music is playing—is the ultimate design goal. However, achieving this noise-free floor is exceptionally challenging in high-gain headphone amplifiers. Because headphones sit directly against the ears and are often highly sensitive (especially modern low-impedance dynamic headphones and multi-driver in-ear monitors), even microvolts of electrical noise are easily audible as an annoying hiss or hum. The primary culprit behind this noise is power supply ripple. To understand how to achieve pristine audio, we must explore the differences between two dominant power supply architectures: Regulated Linear Power Supplies (LPS) and Switched-Mode Power Supplies (SMPS), specifically focusing on their ripple rejection capabilities in high-gain amplification circuits.
Here on Headphone Palace, we explore how power supply quality directly impacts the ultimate listening experience. In this article, we will compare these two supply types and examine how circuit designers mitigate ripple noise to keep audio backgrounds quiet.
Power Supply Specifications: LPS vs. SMPS
Before diving into the technical details of noise coupling, let us examine a high-level comparison of the key parameters that differentiate Regulated Linear Power Supplies from Switched-Mode Power Supplies in audio applications:
| Parameter | Regulated Linear Power Supply (LPS) | Switched-Mode Power Supply (SMPS) |
|---|---|---|
| Primary Ripple Frequency | 100 Hz / 120 Hz (Mains double frequency) | 50 kHz – 500 kHz (Switching frequency) |
| Output Ripple Voltage | < 0.1 mV RMS (Extremely low) | 10 mV – 50 mV RMS (Higher, high-frequency spikes) |
| Efficiency | 30% – 50% (High heat dissipation) | 80% – 95% (Runs cool, highly efficient) |
| High-Frequency Noise (EMI/RFI) | Negligible (No high-frequency switching) | Significant (Requires heavy shielding & filtering) |
| Physical Size & Weight | Large, heavy (Bulky toroidal transformers) | Small, lightweight (High-frequency ferrite cores) |
| Relative Cost | Moderate to High | Low to Moderate |
Understanding Power Supply Ripple and PSRR
Power supply ripple is the residual periodic variation of the DC voltage that has been rectified from an alternating current (AC) source. In a simple unregulated supply, this ripple manifests as a sawtooth waveform at twice the mains frequency (100 Hz in countries with 50 Hz power grids, or 120 Hz in 60 Hz countries). In high-gain headphone amplifiers, any ripple on the power rails can easily couple into the audio signal path through various components, including transistors, operational amplifiers, and passive devices.
The ability of an amplifier circuit to reject noise on its power lines is measured by its Power Supply Rejection Ratio (PSRR). Expressed in decibels (dB), PSRR dictates how much of the power supply noise is attenuated before reaching the audio output. A high PSRR of 80 dB to 100 dB means that only a tiny fraction of the ripple passes through. However, PSRR is not constant across the frequency spectrum. While most amplifiers exhibit excellent PSRR at low frequencies (like 50 Hz or 100 Hz), their ability to reject noise drops precipitously as the frequency increases. This is why high-frequency noise is particularly dangerous in high-gain stages.
Regulated Linear Power Supplies (LPS): The Audiophile Gold Standard?
Regulated Linear Power Supplies have long been considered the gold standard for high-end audio components, and for good reason. The architecture of an LPS is straightforward and inherently quiet. It starts with a large power transformer (often a toroidal transformer to minimize stray magnetic fields) that steps down the mains voltage. This low-voltage AC is then rectified into DC by diode bridges and smoothed using large banks of filter capacitors. Finally, a linear regulator circuit (such as a classic LM317, or a high-performance discrete design) regulates the voltage to the desired output level.
When reading our comparison reviews, you will often find that desktop audio amplifiers boasting external linear power supplies carry a premium price tag. This premium is due to the advantages of an LPS in high-gain audio applications:
- Zero High-Frequency Switching Noise: Because there are no high-speed switching transistors, there is no high-frequency electromagnetic interference (EMI) or radio-frequency interference (RFI) generated.
- Excellent Low-Frequency Ripple Rejection: Regulated linear power supplies can reduce mains-frequency ripple to microvolt levels, ensuring that the 100/120 Hz hum is completely absent from the audio path.
- Simplicity and Reliability: LPS circuits are highly robust and do not produce complex high-frequency harmonic profiles that can fold back into the audible spectrum.
However, linear power supplies have notable drawbacks. They are highly inefficient (often converting 50% or more of the power into heat), bulky, and expensive due to the massive copper transformers and aluminum heatsinks required.
Switched-Mode Power Supplies (SMPS): Efficiency vs. High-Frequency Noise
In contrast to linear power supplies, Switched-Mode Power Supplies (SMPS) operate on a completely different principle. An SMPS rectifies the mains AC voltage directly to high-voltage DC, which is then sliced into high-frequency pulses (typically between 50 kHz and 500 kHz) by a fast-switching transistor. These high-frequency pulses are stepped down by a tiny ferrite-core transformer and rectified and filtered to produce the final DC output.
The benefits of SMPS are substantial:
- Extremely High Efficiency: SMPS designs typically operate at 80% to 95% efficiency, producing very little heat.
- Compact Size and Low Weight: The high switching frequency allows for a fraction of the transformer size, making SMPS units incredibly compact and lightweight.
- Universal Input Voltage: Most SMPS can handle input voltages from 90V to 264V AC, allowing the same device to be used globally.
Despite these advantages, SMPS presents severe challenges for high-gain headphone amplifiers. The high-speed switching generates massive high-frequency noise and electromagnetic fields. While this switching noise occurs far above the range of human hearing (e.g., at 100 kHz), it can interact with non-linearities in the amplifier circuit. This phenomenon, known as intermodulation distortion (IMD), folds high-frequency noise back into the audible band as a high-pitched whine or a general clouding of the audio detail. Additionally, as frequency increases, the PSRR of the amplifier drops, leaving the circuit vulnerable to these high-frequency components.
Visualizing the Noise Spectrum
To better understand this, observe the following graph showing the noise output profiles of both linear and switched-mode power supplies across the frequency spectrum. Notice the localized peak at 100 Hz for the LPS versus the massive high-frequency switching spikes generated by the SMPS:
Ripple Rejection Techniques in Amplifier Design
To visualize how power supply design integrates with amplifier electronics, let us look at the layout of a modern high-gain headphone amplifier board, where heavy regulation components are critical to isolating clean audio traces:
To combat the noise profiles of both power supplies, modern amplifier designers employ sophisticated filtering and regulation strategies. In an LPS-powered design, the focus is on maximizing low-frequency filter capacitance (often using multiple capacitors in parallel to lower Equivalent Series Resistance, or ESR) and using linear regulators with high rejection at 100/120 Hz.
For SMPS-powered amplifiers, designers must employ multi-stage filtering. This includes:
- Common-Mode Chokes and Pi Filters: Inductor-capacitor (LC) networks that target high-frequency switching noise.
- Local Low-Dropout Regulators (LDOs): Highly specific linear regulators placed immediately next to sensitive audio circuits to clean up the power rail.
- Capacitance Multipliers: Active filter circuits that simulate massive capacitance to smooth out voltage fluctuations.
- Shielding: Enclosing the SMPS in a metal Faraday cage to prevent high-frequency EMI/RFI from radiating into high-gain trace lines.
By combining these methods, a well-engineered SMPS-powered amplifier can achieve noise floors that rival, and in some cases exceed, poorly designed linear supplies. However, this level of engineering requires careful design and expensive filtering components. If you want to research specific headphone matches and gear recommendations, head over to our headphones category or check out our latest blog posts where we break down complex audio concepts for audiophiles.
Conclusion: Which One Wins?
Ultimately, the choice between a Regulated Linear Power Supply and a Switched-Mode Power Supply for a high-gain headphone amplifier depends on a balance of performance, cost, and physical space. A linear power supply remains the absolute gold standard for pure, unadulterated ripple rejection without the risk of high-frequency switching artifacts, making it the choice of purists. However, an SMPS, when combined with excellent filtering, shielding, and local LDO regulators, offers a highly efficient, compact, and cost-effective alternative that can deliver outstanding audio quality. When choosing your next amplifier, remember that it is not just the power supply topology that matters, but the overall engineering implementation that determines whether you will hear the music or the power supply noise.
Discuss more about this, FAQ, Announcements and Miscellaneous, over on our community.
