In high-frequency radio engineering and historical telephony, circuit designers adhered strictly to the Maximum Power Transfer Theorem: maximum electrical power is transferred when source impedance perfectly equals load impedance. However, in modern electroacoustic amplification, impedance matching between amplifier and transducer is an obsolete topology that severely degrades acoustic fidelity. Contemporary high-fidelity amplification operates entirely on voltage bridging—an engineering principle where the output stage presents a source impedance ($Z_s$) approaching zero ohms, forcing the amplifier to behave as an ideal, non-sagging voltage source regardless of complex dynamic load variations.
Within high-performance headphone and loudspeaker chains, the ratio between the nominal transducer impedance and the amplifier’s total source impedance defines the damping factor ($DF$). This electrodynamic parameter dictates voice coil transient settling time, back-electromotive force dissipation, and frequency response linearity. Across our audio engineering blog, we examine how discrete Class AB push-pull output stages achieve ultra-low source impedance, analyze the physics of dynamic driver back-EMF braking, and explore why output stage impedance control remains foundational to reference audio reproduction.
The Class AB Output Stage: Establishing Sub-Ohm Source Impedance
A classic discrete solid-state Class AB output stage employs complementary pairs of bipolar junction transistors (BJTs) or power MOSFETs configured in a push-pull emitter follower (common-collector) or source follower (common-drain) topology. The upper NPN/N-channel device sources current during positive signal excursions, while the lower PNP/P-channel device sinks current during negative cycles. A precision $V_{be}$ multiplier circuit maintains a constant quiescent DC bias across the transistor bases, eliminating crossover notch distortion while optimizing thermal stability.
The raw open-loop output impedance ($Z_{out,ol}$) of a bipolar emitter follower is governed by the driver stage source impedance ($R_{driver}$), transistor current gain ($\beta$), the small-signal intrinsic emitter resistance ($r_e$), and the physical emitter degeneration resistors ($R_e$):
$Z_{out,ol} \approx R_e + r_e + \frac{R_{driver}}{\beta + 1}$
Where $r_e = \frac{V_T}{I_C}$ ($V_T \approx 26\text{ mV}$ at room temperature) and $R_e$ typically ranges from $0.1\,\Omega$ to $0.47\,\Omega$ to prevent thermal runaway and ensure current balancing in parallel output pairs. In an open-loop configuration, $Z_{out,ol}$ typically settles between $0.2\,\Omega$ and $1.0\,\Omega$. However, when deep negative feedback (NFB) is wrapped around the amplifier, the closed-loop output impedance ($Z_{out}$) collapses dramatically:
$Z_{out} = \frac{Z_{out,ol}}{1 + A_{ol} \cdot \beta_{fb}}$
With an open-loop voltage gain ($A_{ol}$) often exceeding $80\text{ dB}$ ($10,000\times$) at audio frequencies and a calibrated feedback attenuation factor ($\beta_{fb}$), the closed-loop output impedance of a well-engineered Class AB stage plummets to fractions of a milliohm ($< 0.02\,\Omega$). This ultra-low source impedance forms the foundation for uncompromising transducer damping.
Electrodynamic Damping Mechanics: Taming Voice Coil Back-EMF
A dynamic headphone driver is not a passive resistor; it is an electrodynamic motor and reciprocal electrical generator. When an alternating voltage drives current through the voice coil suspended within a concentrated magnetic field ($B$), the Lorentz force ($F = B \cdot l \cdot I$) accelerates the diaphragm assembly to radiate acoustic pressure waves. However, when the input transient suddenly ends, the mechanical compliance of the suspension ($C_{ms}$) and the moving mass of the cone ($M_{ms}$) cause the assembly to oscillate at its fundamental free-air mechanical resonance frequency ($f_s$).
As the voice coil moves through the permanent magnetic flux lines during post-signal overhang, Faraday’s Law of Electromagnetic Induction dictates that an electrical potential—known as Back-Electromotive Force (Back-EMF)—is generated across the coil terminals:
$V_{emf} = B \cdot l \cdot v(t)$
Where $v(t)$ represents the instantaneous velocity of the vibrating diaphragm. This generated voltage seeks a complete electrical loop to discharge its stored kinetic energy. If the amplifier presents a negligible output impedance ($Z_s \approx 0\,\Omega$), it acts as a virtual dead short-circuit across the driver’s generator terminals. The resulting induced back-EMF current ($I_{emf} = \frac{V_{emf}}{R_{dc} + Z_s + Z_{cable}}$) produces an opposing Lorentz braking force:
$F_{brake} = -B \cdot l \cdot I_{emf} = -\frac{(B \cdot l)^2 \cdot v(t)}{R_{dc} + Z_s + Z_{cable}}$
This electrodynamic braking force rapidly dissipates stored kinetic energy as heat within the voice coil’s DC resistance ($R_{dc}$), critically damping diaphragm oscillations and terminating mechanical ringing within a single cycle. Conversely, if source impedance $Z_s$ is high, $I_{emf}$ is choked, braking torque vanishes, and the transducer continues to oscillate unrestrained, smearing leading-edge transients and generating severe acoustic overhang.
Transient Decay Analysis: Damping Factor vs. Diaphragm Settling
The Damping Factor is mathematically defined as the ratio of nominal load impedance ($Z_{load}$) to the sum of amplifier output impedance ($Z_s$) and interconnect series resistance ($Z_{cable}$): $DF = \frac{Z_{load}}{Z_s + Z_{cable}}$. The engineering chart below illustrates the time-domain impulse response and displacement decay of a $32\,\Omega$ dynamic driver when paired with varying output stage impedances:
As demonstrated in the decay plot, the discrete Class AB output stage with a high damping factor ($DF = 800$) brings the voice coil to a complete mechanical rest within $2.5\text{ ms}$, eliminating transient overshoot. In contrast, underdamped source stages ($DF = 2.1$) suffer from sustained mechanical ringing that persists past $12\text{ ms}$, introducing muddy, uncontrolled low-frequency response.
The Voltage Divider Mechanism: Frequency Response Warping
Beyond time-domain transient ringing, non-zero output impedance introduces significant frequency-domain coloration via passive voltage division. A dynamic driver exhibits a highly non-linear electrical impedance profile across the audible spectrum. At its fundamental mechanical resonance frequency ($f_s$, typically between $50\text{ Hz}$ and $120\text{ Hz}$), the impedance surges from a nominal $32\,\Omega$ to over $120\,\Omega$ due to peak motional back-EMF generation. Similarly, voice coil voice-gap inductance ($L_e$) causes the impedance to rise steadily across the upper treble octave ($10\text{ kHz}$ to $20\text{ kHz}$).
When driven by an amplifier with non-negligible source impedance ($Z_s$), the instantaneous voltage delivered to the transducer terminals ($V_{transducer}$) fluctuates dynamically as a function of frequency:
$V_{transducer}(f) = V_{in} \cdot \left| \frac{Z_L(f)}{Z_s + Z_L(f)} \right|$
When $Z_s$ is high (such as in transformer-coupled tube amplifiers or resistor-padded headphone jacks), a larger proportion of the amplifier’s internal voltage drops across the transducer at $f_s$, generating an unintended, bloated bass peak ($+2.5\text{ dB}$ to $+5.0\text{ dB}$). Concurrently, across mid-range valleys where driver impedance dips to nominal values, delivered voltage sags. In contrast, a low-impedance discrete Class AB stage holds voltage variation under $0.03\text{ dB}$ across all frequencies, preserving the pristine acoustic profile intended by manufacturers of reference audiophile headphones.
Topology Benchmark: Comparing Output Stages Across Architectures
Different amplifier architectures present vastly different output impedance characteristics. On our amplifier topology comparison archive, we evaluate how various designs handle reactive headphone loads. The comparative engineering matrix below details key output parameters across primary amplifier categories:
| Amplifier Output Architecture | Source Impedance ($Z_s$) | Damping Factor ($32\,\Omega$ Load) | Transient Settling Time ($t_s$) | Resonance Frequency Deviation ($\Delta \text{dB}$) | Intermodulation Under Back-EMF |
|---|---|---|---|---|---|
| Discrete Class AB Push-Pull (BJT) | < 0.04 Ω | > 800 | < 2.5 ms (Critically Damped) | < 0.03 dB (Ruler Flat) | Extremely Low (< -110 dB) |
| High-Feedback Op-Amp Buffer Stage | 0.10 – 0.50 Ω | 64 – 320 | 3.0 – 4.5 ms (Fast Control) | 0.08 – 0.35 dB (Minimal) | Very Low (< -98 dB) |
| Transformer-Coupled Tube Output | 2.0 – 6.0 Ω | 5.3 – 16 | 6.0 – 9.0 ms (Moderate Ringing) | +1.2 to +2.8 dB (Warm Bass Bump) | Moderate (< -65 dB) |
| Output Transformerless (OTL) Tube | 25.0 – 80.0 Ω | 0.4 – 1.3 | > 14.0 ms (Severe Overhang) | +4.5 to +8.5 dB (Heavy Bloat) | High (< -50 dB) |
| Resistor-Padded AV Receiver Jack | 47.0 – 150.0 Ω | 0.2 – 0.7 | > 15.0 ms (Sluggish Transients) | +6.0 to +11.0 dB (Severe Warping) | Substantial (< -45 dB) |
Transducer Motor Sensitivity: Dynamic, Planar, and Multi-BA Loads
The acoustic consequence of amplifier damping factor depends fundamentally on the transducer’s internal electromechanical motor structure:
- Dynamic Moving-Coil Drivers: Possess high mechanical compliance and significant motional impedance spikes at $f_s$. High damping factor ($DF > 50$) is essential to suppress mechanical overshoot, eliminate loose bass decay, and preserve midrange clarity.
- Planar Magnetic Drivers: Feature an ultra-thin conductive trace distributed across an entire flat membrane immersed in an isodynamic magnetic array. Because the voice coil trace has negligible parasitic inductance and minimal motional impedance peaking, planar magnetic headphones present a virtually flat, purely resistive impedance curve. While planar drivers do not suffer from frequency response warping, they demand substantial linear current delivery that robust Class AB output stages provide effortlessly.
- Multi-Driver Balanced Armature IEMs: Incorporate passive R-L-C crossover networks that cause extreme impedance swings (e.g., dipping from $40\,\Omega$ at $1\text{ kHz}$ down to $4\,\Omega$ at $8\text{ kHz}$). When driven by high source impedance, crossover slopes warp severely, causing harsh treble spikes and hollow midrange notches.
Design Standards: The 1/8th Rule vs. Modern Ultra-Low Impedance Benchmarks
In classical electroacoustics, the recognized engineering rule of thumb is the Rule of Eighths: an amplifier’s source impedance should never exceed one-eighth ($12.5\%$) of the transducer’s rated impedance ($Z_s \le \frac{Z_{load}}{8}$). Adhering to this limit restricts frequency response errors to within $0.8\text{ dB}$, which is generally acceptable for casual listening.
However, for critical studio mastering and high-resolution audiophile listening on low-impedance multi-driver earphones ($12\,\Omega$ to $16\,\Omega$), maintaining response variations below $0.1\text{ dB}$ requires an output impedance of $Z_s < 0.1\,\Omega$. Discrete Class AB amplifiers—configured with low emitter ballast resistance, heavy negative feedback loop gain, and high-transconductance output transistors—effortlessly satisfy this requirement, guaranteeing absolute transient control, zero audible coloration, and pitch-perfect bass reproduction. Explore our full library of technical hardware analyses on HeadphonePalace to optimize every link in your reference audio chain.
Discuss more about this, FAQ, Announcements and Miscellaneous, over on our community.
