Output impedance

Output impedance, also known as source impedance, is the impedance of an electrical circuit or device as seen looking into its output terminals when no load is connected and all independent voltage sources are short-circuited while all independent current sources are opened.^[1][2] This parameter is equivalent to the Thevenin equivalent resistance (or impedance) at the output port, representing the internal opposition to current flow from the circuit to an external load.^[2][3] In practical terms, output impedance determines the extent to which the output voltage of a circuit drops when a load is applied, as the load impedance forms a voltage divider with the output impedance.^[3] For effective signal transfer in amplifiers and other active devices, a low output impedance is desirable—ideally approaching zero—to minimize loading effects and ensure that the full intended voltage is delivered to the next stage without significant attenuation.^[4][3] This is particularly critical in applications like audio amplifiers, where high output impedance can lead to reduced power delivery and distorted signals, or in power electronics, where it influences voltage regulation under varying loads.^[4] Output impedance is typically measured using the Thevenin theorem by deactivating sources and calculating the equivalent impedance, or experimentally by applying a test signal at the output and observing the ratio of voltage to current.^[1][3] In frequency-dependent contexts, such as AC circuits, it may include reactive components (inductance or capacitance), but it is often predominantly resistive in the midband frequency range for many amplifiers.^[4] Understanding and controlling output impedance is essential for impedance matching, which maximizes power transfer between stages and prevents reflections in high-frequency systems like RF circuits.^[3]

Fundamentals

Definition

Output impedance is the equivalent impedance presented by a device or circuit at its output terminals, equivalent to the Thevenin impedance looking back into those terminals with all internal independent sources deactivated (voltage sources short-circuited and current sources open-circuited).^[5] This concept, formalized in late 19th-century circuit theory, builds directly on Thévenin's theorem, proposed by Léon Charles Thévenin in 1883 as a method to simplify linear electrical networks into an equivalent voltage source in series with an impedance.^[6] It serves as the dual to input impedance, which is similarly defined looking into the input terminals.^[5] In alternating current (AC) circuits, output impedance is typically frequency-dependent due to reactive components like inductors and capacitors, and it is represented as a complex quantity $Z_\text{out} = R + jX$Zout​=R+jX, where $R$R is the real (resistive) part and $X$X is the imaginary (reactive) part.^[7] For direct current (DC) analysis, it reduces to a purely resistive value. A low output impedance is desirable for devices functioning as voltage sources, as it ensures the output voltage remains stable regardless of load variations by minimizing voltage division effects.^[8] An ideal voltage source has zero output impedance, delivering constant voltage without drop; in real sources with non-zero $Z_\text{out}$Zout​, connecting a load $Z_\text{load}$Zload​ results in a reduced load voltage given by the voltage divider formula $V_\text{load} = V_\text{source} \cdot \frac{Z_\text{load}}{Z_\text{out} + Z_\text{load}}$Vload​=Vsource​⋅Zout​+Zload​Zload​​.^[9]

Theoretical Models

The Thevenin equivalent circuit represents the output port of a linear electrical network as an ideal voltage source $V_{th}$Vth​ in series with an output impedance $Z_{out}$Zout​, where $Z_{out}$Zout​ is the equivalent impedance seen at the output terminals when all independent sources are deactivated—voltage sources are short-circuited and current sources are open-circuited.^[8] This model simplifies analysis by reducing complex circuits to a basic form that accurately predicts voltage and current behavior under load. To compute $Z_{out}$Zout​, the deactivated circuit's impedance is calculated looking back into the output port, often involving series-parallel combinations of resistors, capacitors, and inductors.^[8] The Norton equivalent provides a dual representation of the same network, consisting of an ideal current source $I_n$In​ in parallel with the identical output impedance $Z_{out}$Zout​, derived from the same deactivation process as the Thevenin model.^[8][10] The equivalence holds because source transformations between voltage-current pairs preserve the port's $I$I-$V$V characteristics, with $Z_{out}$Zout​ remaining unchanged in both models.^[10] In modeling output impedance, small-signal analysis employs linear approximations around a DC operating point to characterize AC behavior, treating the circuit as a linear network where $Z_{out}$Zout​ is frequency-dependent but constant for small perturbations.^[11] This approach is suitable for audio or RF amplifiers, using hybrid-$\pi$π or T-models to derive impedances like output resistance $r_o$ro​. In contrast, large-signal models account for nonlinear effects in power applications, where $Z_{out}$Zout​ varies with signal amplitude due to saturation or cutoff regions, requiring time-domain simulations rather than linear equivalents.^[11] For a simple voltage divider amplifier formed by two resistors $R_1$R1​ (from input to output) and $R_2$R2​ (from output to ground), the output impedance $Z_{out}$Zout​ is the parallel combination of $R_1$R1​ and $R_2$R2​ when the input voltage source is shorted to ground.^[12] The derivation begins by deactivating the source, leaving $R_1$R1​ connected from the output node to ground and $R_2$R2​ directly to ground, forming a parallel network. The equivalent impedance is then: $$Z_{out} = \frac{R_1 R_2}{R_1 + R_2}$$Zout​=R1​+R2​R1​R2​​ This follows from the parallel resistance formula $1/Z_{out} = 1/R_1 + 1/R_2$1/Zout​=1/R1​+1/R2​, confirming the unloaded Thevenin resistance at the output port.^[12] In reactive circuits, output impedance exhibits frequency dependence, as modeled by $Z_{out}(j\omega)$Zout​(jω), where $\omega$ω is the angular frequency. For instance, in a circuit with resistive $R$R shunted by a capacitor $C$C, the output impedance is the parallel combination: $$Z_{out}(j\omega) = R \parallel \frac{1}{j\omega C} = \frac{R \cdot \frac{1}{j\omega C}}{R + \frac{1}{j\omega C}} = \frac{R}{1 + j\omega R C}$$Zout​(jω)=R∥jωC1​=R+jωC1​R⋅jωC1​​=1+jωRCR​ At low frequencies, the capacitive reactance $1/(j\omega C)$1/(jωC) dominates, making $Z_{out} \approx R$Zout​≈R; at high frequencies, the capacitor shorts, reducing $Z_{out}$Zout​ toward zero. This variation influences bandwidth and stability in AC-coupled systems.^[13]

Measurement

Direct Techniques

Direct techniques for measuring output impedance involve physical electrical probing to directly assess the impedance at the output terminals of a device or circuit, typically modeled using the Thévenin equivalent where the output appears as a voltage source in series with the impedance.^[14] These methods rely on applying known test conditions, such as loads or shorts, and measuring resulting voltages or currents with calibrated instruments to compute the impedance value. The open-circuit voltage method, suitable for voltage sources like amplifiers or signal generators, determines output impedance by observing the voltage drop across a known load resistor. First, the open-circuit voltage $V_{oc}$Voc​ is measured across the unloaded output terminals using a high-impedance voltmeter or oscilloscope. Then, a precision load resistor $Z_{load}$Zload​ (often purely resistive for low-frequency cases) is connected, and the loaded voltage $V_{load}$Vload​ is measured. The output impedance is calculated as $Z_{out} = Z_{load} \times \frac{V_{oc} - V_{load}}{V_{load}}$Zout​=Zload​×Vload​Voc​−Vload​​, derived from the voltage divider principle in the Thévenin model.^[15] This approach assumes a linear response and is effective when $Z_{load}$Zload​ is comparable to the expected $Z_{out}$Zout​ to produce a measurable voltage drop, typically 10-50% of $V_{oc}$Voc​. For current sources, such as certain power supplies or transducers, the short-circuit current method is preferred, where the output impedance is found as $Z_{out} = \frac{V_{oc}}{I_{sc}}$Zout​=Isc​Voc​​, with $I_{sc}$Isc​ being the current measured when the output terminals are shorted using a low-resistance ammeter or shunt.^[14] Safety precautions are essential for high-power devices, including using current-limited supplies, protective fuses rated below the device's maximum output, insulated probes, and personal protective equipment like gloves and safety glasses to prevent arcing or thermal damage during the short. This method avoids high voltages but requires careful handling to limit fault currents. Common equipment includes oscilloscopes for AC voltage measurements, function generators to drive the device under test, precision resistors (e.g., 0.1% tolerance) as loads, and multimeters for DC or low-frequency readings; calibration involves verifying instrument accuracy against known standards, such as shorting inputs for zero offset or using precision voltage sources, to minimize errors from probe capacitance or resistance mismatches.^[16] For DC output impedance, steady-state measurements are used after allowing transients to settle (e.g., seconds to minutes), focusing on resistive components with multimeters. In contrast, AC measurements capture frequency-dependent impedance using network analyzers for RF sweeps (e.g., 1 MHz to 3 GHz), where vector measurements of voltage and current phases enable complex $Z_{out}$Zout​ calculation via the RF I-V method.^[17] An example procedure for low-frequency circuits (e.g., audio amplifiers at 1 kHz) begins by powering the device and connecting its output to an oscilloscope channel set to AC coupling with 1 MΩ input impedance to measure $V_{oc}$Voc​ (ensure no load, verify signal stability). Next, select a precision resistor $R_{load}$Rload​ (e.g., 50 Ω for expected low $Z_{out}$Zout​) and connect it across the output, measuring $V_{load}$Vload​ on the same channel while monitoring for distortion. Compute $Z_{out} = R_{load} \times \left( \frac{V_{oc}}{V_{load}} - 1 \right)$Zout​=Rload​×(Vload​Voc​​−1); repeat with multiple loads (e.g., 10 Ω, 100 Ω) for averaging. Error analysis shows typical accuracy limits of ±5%, arising from voltmeter loading (mitigated by high-impedance probes), resistor tolerance (±0.1%), and signal noise (reduced via averaging over 10 cycles), with systematic errors like cable inductance contributing <1% below 10 kHz.^[15]

Indirect Techniques

Indirect techniques for measuring output impedance infer the value from observed responses or models without directly injecting signals at the output terminals, often useful when physical access is limited or non-invasive assessment is preferred. These methods rely on varying external conditions, transient behaviors, or computational models to estimate Z_out, providing insights into circuit performance under operational constraints.^[18] The load variation method involves applying a series of known loads to the output and measuring the resulting changes in output voltage or current. By plotting the output voltage against load resistance (or current against load conductance), the output impedance is extrapolated from the slope of the linear region, typically using a least-squares fit to minimize measurement errors. For instance, with two distinct resistive loads R_L1 and R_L2, the output impedance can be calculated as Z_out = (V1 * R_L2 - V2 * R_L1) / (V2 - V1), where V1 and V2 are the corresponding output voltages for a fixed input. This approach is particularly effective for DC or low-frequency measurements in amplifiers and power circuits, as it leverages the Thevenin equivalent model. Simulation-based inference employs circuit simulation tools like SPICE to model the device and extract Z_out through small-signal analysis. In LTspice or similar software, an AC analysis is performed by injecting a 1 A AC current source at the output node with the input source set to zero, then computing Z_out as the ratio of output voltage to injected current (Z_out = |V_out / I_inj|) across a frequency sweep. Convergence can be improved by adding small resistors (e.g., 1 mΩ) in series with voltage sources, enabling initial conditions, or using .UIC directives for transient-assisted AC simulations; for complex models, the .NET directive computes network parameters including output impedance directly. This method allows exploration of parametric variations and parasitic effects without hardware, though results depend on the accuracy of component models.^[19][20] In RF applications, non-invasive methods use a vector network analyzer (VNA) to measure S-parameters and convert S_{22} to output impedance via Z_out = Z_0 \frac{1 + S_{22}}{1 - S_{22}}, where Z_0 is the reference impedance (typically 50 Ω). The device under test is connected to port 2 of the VNA with port 1 terminated, and S_{22} is measured over frequency; this reflection coefficient-based approach avoids direct probing by treating the output as a one-port network. It excels for high-frequency characterization, providing magnitude and phase data for complex Z_out.^[21] These indirect techniques share limitations tied to underlying assumptions, such as ideal load linearity or accurate model parasitics; errors arise from unmodeled effects like cable inductance, fixture residuals, or non-stationary transients, potentially degrading accuracy by 10-20% at high frequencies. Validation against direct methods is recommended for critical applications.^[18]

Applications in Electronics

Amplifiers

In voltage amplifiers, a low output impedance ($Z_{out}$Zout​) is essential to ensure efficient power transfer to the load while minimizing voltage drop and gain reduction. Ideally, $Z_{out}$Zout​ should be much less than 1 Ω to drive varying loads without significant attenuation of the output signal. When a load impedance $Z_{load}$Zload​ is connected, the effective voltage gain becomes $A_{v,loaded} = A_v \frac{Z_{load}}{Z_{load} + Z_{out}}$Av,loaded​=Av​Zload​+Zout​Zload​​, where $A_v$Av​ is the unloaded gain; this voltage divider effect highlights how even modest $Z_{out}$Zout​ can degrade performance if not sufficiently low.^[22] Negative feedback plays a crucial role in optimizing $Z_{out}$Zout​ for amplifiers, particularly operational amplifiers (op-amps) in configurations like noninverting or voltage-follower setups. By sampling the output and feeding a portion back to the input, negative feedback reduces the effective $Z_{out}$Zout​ by the loop gain factor $(1 + A\beta)$(1+Aβ), where $A$A is the open-loop gain and $\beta$β is the feedback fraction; thus, $Z_{out,closed} \approx \frac{Z_{out,open}}{1 + A\beta}$Zout,closed​≈1+AβZout,open​​. This derivation stems from the feedback stabilizing the output against load variations, enabling op-amps to achieve $Z_{out}$Zout​ values as low as milliohms in closed-loop operation.^[23][24] Amplifier classes influence $Z_{out}$Zout​ through their output stage topologies, with variations arising from biasing and conduction patterns. In Class A amplifiers, such as the emitter follower (common-collector) configuration, the output stage provides linear operation but at lower efficiency; here, $Z_{out} \approx r_e + \frac{R_E}{\beta}$Zout​≈re​+βRE​​, where $r_e$re​ is the emitter resistance (approximately $kT/qI_E \approx 26$kT/qIE​≈26 mV$/I_E$/IE​), $R_E$RE​ is the emitter resistor, and $\beta$β is the current gain, yielding low $Z_{out}$Zout​ suitable for buffering. Class B and AB push-pull stages, using complementary pairs, improve efficiency (up to 78.5% theoretically for Class B) but introduce crossover distortion if not biased properly; their $Z_{out}$Zout​ is similarly low due to the follower arrangement, though it rises slightly at low signal levels from reduced transconductance. Class C amplifiers, biased for tuned applications, exhibit higher $Z_{out}$Zout​ variations due to nonlinear conduction, making them less ideal for broadband voltage amplification.^[25][26] High $Z_{out}$Zout​ in amplifiers can lead to distortion and stability issues, particularly in audio applications where load impedance varies with frequency. Frequency-dependent loading from elevated $Z_{out}$Zout​ causes uneven frequency response, manifesting as harmonic distortion (e.g., up to 2 dB deviations at bass frequencies with $Z_{out} \approx 0.8$Zout​≈0.8 Ω for an 8 Ω load). The damping factor (DF), defined as $DF = \frac{Z_{load}}{Z_{out}}$DF=Zout​Zload​​, quantifies control over the load; low DF (<100) reduces damping of speaker resonances, potentially exciting oscillations with reactive loads and compromising stability. Amplifier damping factors greater than 300 are typically targeted to achieve a system-level DF above 150, limiting audible effects.^[27] A typical design example for a power amplifier output stage employs complementary MOSFETs in a source-follower configuration within a Class AB amplifier, biased at 50-100 mA quiescent current to minimize crossover distortion. Global negative feedback from the output reduces $Z_{out}$Zout​ to approximately 0.1 Ω at audio frequencies (20 Hz to 20 kHz), enabling effective drive of 4-8 Ω speakers with minimal gain loss and high damping (DF >80). This setup, common in high-fidelity audio, uses devices like IRFP240/IRFP9240 pairs with emitter degeneration resistors (0.1-0.22 Ω) for thermal stability, achieving total harmonic distortion below 0.1% at 100 W output.^[28]

Signal Generators

In signal generators, particularly function generators used for RF applications, the ideal output impedance is standardized at 50 Ω to ensure maximum power transfer and minimize signal reflections in transmission lines. This value aligns with the characteristic impedance of common coaxial cables, preventing standing waves that could distort the output waveform. For instance, devices like the Keysight 33600A series function generators feature an optional fixed output impedance of 50 Ω on their primary output terminal, designed for direct connection to 50 Ω loads.^[29] When the load impedance mismatches this 50 Ω source, reflections occur, quantified by the voltage standing wave ratio (VSWR), which measures the amplitude ratio of the forward to reflected voltage wave along the line. A VSWR of 1 indicates perfect matching, while values greater than 1.5 can lead to significant signal attenuation or distortion in RF testing setups.^[30] For oscillators within signal generators, such as those based on parallel LC tanks, the output impedance exhibits a pronounced peak at the resonance frequency $f_0$f0​, where the inductive and capacitive reactances cancel, leaving a predominantly resistive response amplified by the circuit's quality factor. The peak impedance magnitude is given by $Z_{\text{out}} = Q \cdot R$Zout​=Q⋅R at $f_0$f0​, with $Q$Q representing the Q-factor (typically 50–200 for high-performance RF oscillators) and $R$R the effective series resistance of the tank components. Higher Q-factors sharpen this peak, enhancing frequency selectivity but also increasing sensitivity to loading effects that can pull the oscillation frequency. This behavior is critical in voltage-controlled oscillators (VCOs) used for sweeping signals, as it determines the drive capability into subsequent stages without excessive damping.^[31] Calibration of signal generator output impedance relies on NIST-traceable 50 Ω terminations to verify performance across operational bandwidths, such as from 1 Hz to 1 GHz in versatile function generators. These terminations, often precision coaxial loads with tolerances below 0.5%, serve as reference standards during vector network analyzer (VNA) measurements, ensuring the generator's source impedance remains within specifications like ±5% variation. Measurements reveal that output impedance can deviate slightly with frequency due to parasitic elements in the output network, requiring periodic recalibration to maintain traceability for metrology applications. For example, Keysight's N4690D electronic calibration modules provide automated NIST-traceable verification for 50 Ω ports up to 18 GHz.^[32][33] High output impedance in signal generators can compromise signal integrity when driving capacitive loads, such as oscilloscope probes or long cables, leading to amplitude droop at higher frequencies. The capacitive reactance forms a voltage divider with the source impedance, attenuating the signal as $|Z_C| = 1/(2\pi f C)$∣ZC​∣=1/(2πfC) decreases, resulting in up to 50% amplitude loss above 1 MHz for unbuffered outputs into 100 pF loads. To mitigate this, output buffers—typically unity-gain op-amps like those in Analog Devices' high-speed amplifiers—are employed to provide low-impedance driving (e.g., <1 Ω), isolating the generator's core from load variations while preserving waveform fidelity.^[34] A representative example is a sine wave generator configured for RF output, where $Z_{\text{out}} = 50 + j0 \, \Omega$Zout​=50+j0Ω ensures a purely resistive match into coaxial systems. In such setups, impedance transformation occurs via the cable's characteristic impedance, maintaining signal amplitude if terminated properly; for instance, Tektronix AFG series generators deliver flat response up to 100 MHz into 50 Ω coax when buffered against reactive mismatches. This configuration is standard for precision testing, emphasizing the need for matching to avoid phase shifts or harmonic distortion.^[35]

Applications in Power Systems

Batteries

In electrochemical batteries, the output impedance, often referred to as internal impedance, represents the opposition to current flow within the cell and manifests as a combination of ohmic resistance ($R_\Omega$RΩ​), charge transfer resistance ($R_{ct}$Rct​), and diffusion-related components. The ohmic resistance arises from the electrolyte, separators, and current collectors, typically measured at high frequencies (>10 kHz) in electrochemical impedance spectroscopy (EIS). Charge transfer resistance, associated with electrochemical reactions at the electrode-electrolyte interface, dominates around 10 Hz, while diffusion impedance accounts for ion transport limitations in the electrodes at low frequencies (<1 Hz). For lithium-ion (Li-ion) batteries, these components contribute to typical internal impedance values ranging from 10 to 100 mΩ for fresh cells, depending on chemistry, size, and state of charge.^[36][37] The effective output impedance increases with discharge rate, influencing battery performance through Peukert's law, which describes the reduction in available capacity at higher currents. This law is expressed as $C_\text{eff} = C \left( \frac{I}{I_0} \right)^{1-n}$Ceff​=C(I0​I​)1−n, where $C$C is the nominal capacity at reference current $I_0$I0​, $I$I is the discharge current, and $n$n (Peukert exponent, typically 1.1–1.3 for Li-ion) reflects losses from elevated internal resistance and incomplete ion recovery between reactions. Higher discharge rates amplify ohmic and diffusion losses, leading to greater effective impedance and reduced usable capacity, particularly in high-power applications.^[38] Battery aging elevates output impedance over charge-discharge cycles, primarily due to solid electrolyte interphase (SEI) layer growth on the anode, which consumes lithium and electrolyte while increasing resistance to ion transport. This growth, driven by side reactions and mechanical cracking, raises charge transfer and diffusion impedances, contributing to power fade and capacity loss. AC impedance spectroscopy at 1 kHz is commonly used to quantify this, as it approximates the ohmic resistance and provides a standard metric for state-of-health assessment, with values rising non-monotonically—often decreasing slightly early in life before increasing toward end-of-life.^[39][36] Different battery types exhibit distinct output impedance profiles. Alkaline batteries, such as zinc-manganese dioxide cells, typically have higher internal resistance (around 200–500 mΩ per cell) due to their non-rechargeable chemistry and thicker separators, leading to more pronounced voltage drops under load compared to rechargeable alternatives. Lithium-polymer (LiPo) batteries, a variant of Li-ion with polymer electrolytes, maintain lower impedances (5–50 mΩ fresh) for better high-rate performance but are susceptible to similar aging mechanisms. For lead-acid batteries, fresh cells show low impedance (~5 mΩ), but degradation from sulfation and electrolyte stratification can elevate it to >50 mΩ, significantly impairing cranking power.^[40][41] Practically, elevated output impedance causes voltage sag under load, where the terminal voltage drops as $V = E - I Z_\text{out}$V=E−IZout​ (with $E$E as open-circuit voltage and $I$I as current), reducing efficiency and triggering low-voltage cutoffs prematurely. In electric vehicle (EV) battery packs, for instance, if each Li-ion cell has ~1 mΩ internal resistance and the pack comprises 100 cells in series (with parallel strings scaling inversely), the total series impedance could reach 100 mΩ, resulting in a 100 V sag at 1000 A draw—highlighting the need for low-impedance cells to minimize range loss and heat generation.^[42][40]

Power Supplies

In electronic power supplies, output impedance plays a critical role in maintaining stable voltage delivery to loads, particularly under varying current demands. Linear power supplies achieve very low output impedance, typically on the order of 1–10 mΩ, through the use of pass transistors that provide negative feedback to minimize voltage variations.^[43][44] This low impedance ensures excellent load regulation, where the change in output voltage per unit change in load current, ΔV/ΔI, approximates the output impedance Z_out. Switching-mode power supplies (SMPS) also achieve low output impedance at low frequencies, typically on the order of 10–50 mΩ, though their output impedance increases at higher frequencies due to the switching operation and output filter characteristics.^[45] For capacitive output filters common in both linear and switching designs, the impedance is dominated by the capacitive reactance, given by Z_out ≈ 1/(2πfC), where f is the frequency of interest and C is the capacitance.^[46] This formula highlights how larger capacitors or lower frequencies reduce Z_out, improving regulation but potentially increasing size and cost. Load regulation specifications directly tie to this, as poor regulation manifests as voltage droop proportional to load current and Z_out.^[47] Transient response to load steps is heavily influenced by output impedance, with lower Z_out enabling faster recovery times from sudden current changes. For instance, supplies with Z_out below 0.1 Ω can recover from a full-load step in under 1 μs, minimizing voltage undershoot and supporting dynamic loads like processors.^[48] In AC generators such as alternators, output impedance is characterized by the synchronous reactance X_s, typically 0.5–5 Ω per phase for medium-sized machines, which causes voltage drops under varying loads and impacts overall system stability.^[49] To mitigate voltage drops due to output impedance in distribution lines, remote sensing techniques are employed, where sense wires connect directly to the load to adjust the supply's output voltage accordingly. For example, in a 5 V USB power supply, maintaining Z_out below 0.5 Ω—often achieved through low-resistance pass elements and filtering—ensures compliance with USB standards for minimal voltage deviation at the load, even with cable lengths up to 3 m.^[50]