Why Class D Amplifiers Are Not Digital Amplifiers
The “digital amplifier” label on Class D gear is a naming shortcut. The real trade-off sits between a compact, efficient switching stage and the analog output filter that turns square waves into music. The filter is where treble reach and speaker damping are won or lost. This design favors listeners who want a small, cool-running amp and can accept a slight high-frequency rolloff.
- The “digital” label describes one part, not the whole signal path. The switching stage chops analog audio into on/off pulses. The input and the smooth output are both analog. The word “digital” sticks to the chopping step only.
- The output filter is the real design choice. Inductors and capacitors clean up the square waves. Their size sets how high the treble reaches and how firmly the amp grips the speaker cone.
- Efficiency and transparency pull in opposite directions. A smaller, cooler-running switcher needs a bigger filter to keep switching edges out of the music. That filter adds ringing and caps the top of the range.
- The label can mislead in specific speaker pairings. A low-impedance woofer or a high-sensitivity tweeter makes the filter’s impedance signature and switching losses audible, and a gentler analog topology handles those loads more cleanly.
The Goal On The Drawing Board
The engineers had a concrete problem to solve. Old analog amplifiers, built on Class A or Class AB topologies, burned a lot of power as heat. The output transistors sat at a fixed voltage most of the time. Even when the music was quiet, energy still flowed through them and turned into waste heat. The amplifier body grew heavy and hot. The goal was to deliver the same musical power with a fraction of the heat, so the whole unit could shrink to shelf size and run cool enough to sit next to your speakers without baking the room.
Think of a bicycle brake. A Class A amplifier is like holding the brake cable tight at all times, even when the bike rolls slowly. The cable is usually under tension, and that tension creates friction, which is heat. A Class D amplifier is like a brake that only squeezes when you need to slow down. Most of the time the cable is slack. The bike coasts. When you do need to stop, the squeeze is quick and decisive. That quick squeeze-and-release is the switching action: the output transistors are either fully on or fully off, with no in-between where energy leaks away as wasted heat.
The Constraints Nobody Could Argue With
No Class D amplifier escapes these limits, no matter how skilled the design team is.
The square-wave problem. The switching stage chops the incoming audio into a rapid stream of on/off pulses. This is called pulse-width modulation (PWM). The width of each pulse encodes how loud the music is at that instant. The raw output is a jagged square wave, not a smooth curve. Something must smooth it before it reaches the speaker. That something is a pair of inductors (coils of wire that resist sudden changes in current) and capacitors (components that store and release electrical charge) sitting at the amplifier’s output terminals. Together they form the output filter, and they are the last thing the signal meets before it drives the speaker.
Filter ringing and bandwidth ceiling. Every inductor-capacitor pair has a natural resonance, much like a wine glass that hums when you tap the rim. When a sharp PWM pulse hits that filter, it rings at its resonant frequency before it settles. This ringing adds a small overshoot to each note. In the bass region, the effect is barely audible. In the upper mids and treble, it shows up as a soft, slightly metallic edge on cymbals and violin harmonics. The filter also sets a hard ceiling on how high the amplifier can pass signal before the level starts to thin out. A smaller, cheaper filter has its corner lower, so the top of the audible range gets quietly clipped.
Source impedance and damping grip. The output filter is not a clean, zero-resistance wire. It carries its own impedance, which means it acts like a small spring between the amplifier and the speaker. Damping factor is how tightly the amplifier can grab the speaker cone and stop it the instant the signal says stop. A Class A or AB output stage drives the speaker with a very low output impedance, so it grips the cone firmly. The Class D filter adds a few ohms of impedance, softening that grip. With a low-impedance woofer, say 4 ohms, the difference becomes noticeable: the bass can sound slightly looser, with a little extra cone travel after the note ends.
Heat still shows up at the switcher. Even though each transistor is either on or off, every transition (the moment the switch flips from off to on or on to off) wastes a small burst of energy. The more often the switch flips each second, that is, the switching frequency, the more transition losses pile up. Push the switching frequency higher to shrink the filter, and the transistors run hotter. Drop it lower to save heat, and the filter components grow physically larger. The engineer is choosing between a hot, compact amp and a cool, bulky one.
Two Roads, One Chosen
| Design direction | What it gains | What it costs | Where you meet it |
|---|---|---|---|
| High switching frequency, small filter | Compact amp body, tighter treble reach | More heat at the transistors, needs a robust heat sink | Shelf-sized home stereo amplifiers |
| Low switching frequency, large filter | Cooler-running amp, cheaper inductor and capacitor parts | Wider treble rolloff, more ringing on fast notes | Power-amp-only units, car audio, bass-speaker amplifiers |
| Analog Class A or Class AB topology | Lowest output impedance, no PWM ringing, strongest damping grip | Large heat sink, heavy chassis, more wasted heat | Studio monitors, high-end two-way bookshelf systems |
| Hybrid: Class D with a large, high-quality filter | Keeps the compact switcher but recovers much of the analog damping and treble | Bigger chassis than a bare Class D unit, filter still adds some ringing | Premium compact amps aimed at speaker-critical listeners |
How You Hear The Decision
When the high-switching-frequency, small-filter road wins, the bass stays punchy and controlled. Kicks in a drum track have a clean attack and a quick decay because the filter settles fast. Vocals sit in the middle of the mix with a smooth, slightly warm texture. The highs are present and clear, but a careful listener playing a solo violin or a hi-hat will notice the top octave rolling off gently, as if the sound is viewed through a thin sheet of tissue. Cymbals lose a small amount of that bright, sizzling edge.
If the other road had been chosen, a low-frequency switcher with a large, inexpensive filter, the bass would feel softer and slower to settle. The low end would bloom a little before the next note arrives. Vocals would still sound fine, because the filter does its job well in the 500 Hz to 2 kHz range. But the treble would thin out earlier, and fast passages in an orchestra would lose some of their shimmer. The overall impression: the music still sounds good, but the last slice of the frequency range is a shade less detailed.
The analog topology, for comparison, hands the tweeter a clean, near-zero-impedance drive. The cymbal wash extends higher, with more air above the fundamental. The damping grip on a low-impedance woofer is noticeably firmer, and the bass tail snaps shut the instant the signal stops. That grip is what keeps a two-way bookshelf speaker with a small 6-inch woofer tracking a loud, complex bass line without the cone overshooting.
When To Pick The Other Road
A few concrete listening situations make the analog topology the wiser pick, and the “digital amplifier” label should not sway you away from it.
A small cabinet with a high-sensitivity tweeter. Many two-way bookshelf speakers pair a 6-inch woofer with a tweeter rated at 90 dB or above in sensitivity. That tweeter does not need a lot of power. What it needs is a clean, low-impedance drive that lets it ring freely into the room. The Class D output filter, even a well-built one, adds a few ohms of impedance that can color the tweeter’s breakup region, the frequency zone just above its normal working range where the cone starts to flex unevenly. The result is a slight dullness in the top kilohertz that a well-matched analog amp would not introduce.
A low-impedance woofer in a small box. If your speaker uses a 4-ohm or even 2-ohm woofer in a compact sealed cabinet, the damping factor matters a great deal. The small cabinet gives the woofer little room to move. The amplifier’s electrical grip is the only thing keeping the cone from coasting after a note ends. A Class D filter’s source impedance, even at a modest value, loosens that grip. You hear it as bass that drags a beat too long, a slight muddiness when the low end gets busy. An analog topology with its near-zero output impedance stops the cone much more quickly.
A quiet, treated listening room with near-field speakers. In a small studio or a well-damped listening space, you hear every detail, including filter ringing that a large, reverberant living room would swallow. The inductor-capacitor resonance that is inaudible in a big room becomes a faint metallic ring on each note’s decay in a treated, quiet space. Here, the extra damping and the absence of PWM edges in an analog amp make a difference even at conversation-level volume.
How The Decision Evolved
When Class D amplifiers first appeared in the 1970s and 1980s, they were almost exclusively for bass speakers. A subwoofer only needs to reproduce 20 Hz to 200 Hz. The switching frequency could stay low, in the tens of kilohertz. The filter could be a simple, cheap pair of inductors and capacitors. Nobody cared whether the treble rolled off at 10 kHz because there was no treble in the signal. The “digital amplifier” label stuck because the switching stage looked digital on a block diagram, and marketing ran with the word.
By the late 1990s and into the 2000s, engineers pushed the switching frequency higher, into the hundreds of kilohertz and beyond. A faster switcher needs a smaller filter to do the same smoothing job. A smaller filter means a lighter, shelf-sized amplifier. But a smaller filter also moves its ringing frequency up and drops the treble ceiling. Designers started pairing the high-speed switcher with carefully matched, low-loss inductors and film capacitors to keep the ringing in check. The trade-off shifted: the heat load on the switching transistors grew, and the filter, though physically smaller, became more expensive to build well.
Today, the balance has moved further. Some Class D designs switch at 1 MHz or above, which lets the filter be tiny and keeps the treble extension close to that of a good analog amp. The cost is that the transistors must be very fast, and the cooling still has to handle the transition losses. At the same time, analog Class A and Class AB topologies have improved their own efficiency through better thermal management and lower-voltage designs, so the gap in heat output is narrower than it was twenty years ago. The crossover point between “Class D is the practical choice” and “an analog amp sounds cleaner in my specific system” keeps shifting, and it depends on your speaker’s impedance, your room’s size, and how far you sit from the drivers.
Questions About The Design Choice
Does “Digital Amplifier” Mean The Whole Signal Path Is Digital?
No. The word “digital” describes only the switching stage, which turns the audio signal into a stream of on/off pulses using PWM. The signal going into the amplifier is analog. The smooth audio coming out of the output filter is analog. Only the middle chunk, the rapid switching, uses digital-like logic. If the input were truly digital, you would see words like “digital-to-analog converter” on the spec sheet. The misnomer persists because it sounds more precise and modern than “switching amplifier.”
Why Does The Output Filter Limit Treble Extension?
The inductor and capacitor in the filter form a low-pass network. They are tuned to pass the music, roughly 20 Hz to 20 kHz, and reject the switching frequency, often 200 kHz to 1 MHz. The corner where the filter starts to roll off is set by those component values. A smaller, cheaper filter has its corner set lower, so the very top of the audible range gets attenuated before it reaches the tweeter. A larger, more carefully matched filter pushes that corner above 20 kHz, and the treble comes through unharmed.
Will A Class D Amp Hurt My Small Bookshelf Speakers?
In most cases, no. If the amplifier is well-designed and your speakers sit in the 4- to 8-ohm range, the output filter’s impedance is low enough that damping remains acceptable, and the ringing is typically inaudible at normal listening levels. The “digital” label should not scare you away. But if your speakers are a very low-impedance design, below 4 ohms, or use a particularly sensitive tweeter, compare the Class D amp against an analog unit in the same room before you decide. Your ears will tell you if the damping grip is too loose.
Is The Heat Really That Different Between Class D And Class AB?
In most modern designs, the gap has narrowed. A well-built Class AB amplifier running at moderate levels wastes roughly 30 to 40 percent of its input power as heat. A Class D unit wastes far less, because the transistors are either fully on or fully off. But if you run a Class D amp at very high output into a low-impedance load, the transition losses pile up, and the heat sink still gets warm. For a shelf-sized amp driving a typical 8-ohm speaker at conversation levels, the difference in back-of-chassis temperature is the clearest sign of the efficiency gain.
(This piece is an original hifireport.com publication. Reuse without consent is prohibited.)