Fact-checked by Grok 2 weeks ago

Two-port network

A two-port network is an electrical circuit or device characterized by two pairs of terminals, called ports, where each port consists of a pair of nodes allowing for the measurement of voltage and current, facilitating the analysis of energy transfer between an input and an output. These networks model linear, passive systems such as resistors, capacitors, inductors, transformers, and transmission lines, assuming no internal power sources and often exhibiting reciprocity, where the response is symmetric between ports. The behavior of a two-port network is described using sets of parameters that relate the port voltages and currents, enabling simplified analysis of complex circuits. Common parameter sets include impedance (Z) parameters, which express voltages in terms of currents (e.g., V_1 = Z_{11}I_1 + Z_{12}I_2); admittance (Y) parameters, the inverse relating currents to voltages; hybrid (H) parameters, mixing voltage and current for transistor modeling; and transmission (ABCD) parameters, ideal for cascading networks by relating input to output variables. These parameters can be interconverted and are particularly useful for networks that are reciprocal (where Z_{12} = Z_{21}) or symmetrical. Two-port networks are fundamental in for applications including design, synthesis, , and systems, where they allow modular analysis of interconnected components in series, , or configurations. In RF and high-frequency contexts, they extend to (S) parameters to account for wave propagation, aiding in the characterization of antennas, couplers, and transmission lines.

Fundamentals

Definition and Graphical Representation

A two-port network is an electrical circuit or device characterized by four terminals grouped into two pairs, each pair forming a that connects to external circuits. The network is modeled as a , where the internal components—such as resistors, capacitors, inductors, or dependent sources—are encapsulated, and the external behavior is described solely by the voltages and currents at the ports. Each port consists of two terminals, with the voltage defined as the potential difference across them and the current as the net flow into the , ensuring no direct connection between the ports except through the internal network. The variables associated with the ports are the input port voltage V_1 and current I_1 at port 1, and the output port voltage V_2 and current I_2 at port 2. Port 1 is conventionally the input, where I_1 enters the positive and V_1 is measured with the positive at that . For port 2, the output, I_2 is defined as leaving the positive to indicate flow direction, with V_2 across the terminals and positive at the exit point. This directionality aligns with the passive , where positive is absorbed when voltage and current have the same reference. Graphically, a two-port network is often represented in a chain diagram, showing port 1 on the left with its terminals facing right toward port 2 on the right, emphasizing the sequential input-output flow like links in a . In distributed systems, such as transmission lines, it is depicted as a connecting the two ports at opposite ends, modeling wave between them. Another representation is the lattice diagram, which illustrates a symmetrical with series and shunt elements arranged in a crossed (diagonal) and parallel structure, useful for balanced networks. Two-port networks are typically assumed to be linear, meaning responses are proportional to excitations and superposition applies; time-invariant, with parameters unchanging over time; and composed of lumped elements, where component sizes are negligible compared to signal wavelengths unless specified as distributed. They may be passive, containing only energy-dissipating or -storing elements, or active, incorporating sources for .

Applications in Circuit Analysis

Two-port networks originated in the early as a fundamental tool in , particularly for analyzing lines and transmission systems during the expansion of long-distance . Pioneered by researchers at Bell Laboratories, such as George A. Campbell, the theory addressed challenges in signal and over extended cables, with early applications focusing on loading coils and wave filters to improve voice transmission quality. Campbell's seminal work on physical theory of electric wave filter circuits, published in 1922, laid the groundwork for modeling linear networks as interconnected two-ports, enabling systematic design of frequency-selective components for . This historical development, building on theory from the late , transformed by providing a modular framework for cascading network sections, as exemplified in the use of parameters for telephone cable modeling. In practical electronics, two-port networks play a central role in simplifying the of amplifiers, filters, lines, and matching networks by representing complex subsystems as black boxes with defined input and output relationships. For amplifiers, transistors are commonly modeled as two-port devices using hybrid parameters to calculate voltage gain, input/output impedances, and stability factors, facilitating broadband design in communication systems. In , two-port representations allow engineers to predict and , essential for separating signal bands in analog , as seen in early filters and modern RF front-ends. lines, such as cables or lines, are analyzed using or parameters to account for reflections and , ensuring over high-speed links. Matching networks, often composed of lumped elements or distributed lines, employ two-port models to optimize transfer and minimize standing wave ratios in RF systems. The benefits of this approach lie in its modularity, which permits treating subsystems independently while enabling precise calculations of overall gain, impedance matching, and stability in cascaded circuits. By isolating internal complexities, two-port analysis reduces computational demands and supports superposition for linear systems, making it invaluable for predicting performance without full circuit simulation. For instance, in transistor-based amplifiers, the two-port model reveals potential oscillations through stability criteria like the Rollett factor, guiding design iterations. In RF circuits, it ensures signal integrity by quantifying return loss and insertion gain, critical for minimizing distortions in wireless transceivers. Examples abound in electronics, such as modeling a common-emitter amplifier as a two-port to derive h-parameters for small-signal analysis, which directly informs bias and configurations. In RF engineering, two-port networks characterize matching circuits to achieve 50-ohm interfaces, enhancing efficiency in and cellular systems. These techniques extend to modern applications in integrated circuits, where two-port models simulate on-chip s and filters in processes for low-power 5G transceivers. In , scattering parameters describe high-frequency behavior of waveguides and monolithic microwave integrated circuits (MMICs), supporting designs up to millimeter waves. Additionally, in control systems, actuators and sensors are represented as two-ports to analyze loops and dynamic responses, aiding in robotic and automotive applications.

General Properties

Linearity and Superposition

In two-port networks, refers to the property where the output voltages and currents are directly proportional to the input excitations, satisfying both homogeneity and additivity. Homogeneity implies that scaling an input by a constant factor k scales the corresponding output by the same factor, while additivity means that the response to a sum of inputs equals the sum of the individual responses. This linearity enables the application of the , which states that in a linear two-port network with multiple independent sources, the total response at any port is the sum of the responses produced by each source acting alone, with all other sources deactivated (voltage sources shorted and current sources opened). Superposition simplifies analysis by allowing decomposition of complex excitations into simpler components. Mathematically, linearity manifests in the linear algebraic relations between port voltages \mathbf{V} and currents \mathbf{I}, such as \mathbf{V} = \mathbf{Z} \mathbf{I} in impedance form or \mathbf{I} = \mathbf{Y} \mathbf{V} in admittance form, where \mathbf{Z} and \mathbf{Y} are constant matrices independent of the excitation levels. These relations hold for networks composed of linear elements like resistors, inductors, and capacitors. As a consequence, facilitates parameter extraction by applying independent excitations to each sequentially, leveraging superposition to isolate individual effects without from other . It is particularly valid in small-signal analysis, where signals are sufficiently small to avoid nonlinear behavior. However, breaks down in networks containing nonlinear elements such as diodes or transistors, where responses do not scale proportionally; in such cases, small-signal approximates the behavior around an using equivalent linear models.

Reciprocity and Symmetry

In two-port networks, reciprocity refers to the property where the transfer characteristics are identical in both directions, meaning the response at one port due to at the other is the same as the reverse scenario. Mathematically, this is expressed in as z_{12} = z_{21} and in as y_{12} = y_{21}, indicating that the at port 1 due to a current at port 2 equals the at port 2 due to the same current at port 1. This condition holds for passive, linear, time-invariant networks composed of elements like resistors, capacitors, inductors, and transmission lines, provided there are no active devices or materials that introduce directionality, such as gyrotropic media. The reciprocity theorem underlying this property was formalized by in his work on electromagnetic theory, with key extensions applied to network analysis in the early . Symmetry in a two-port network describes a balanced where the input and output ports exhibit equivalent , allowing interchange without altering the overall response. This is characterized by z_{11} = z_{22} in and h_{11} = h_{22} in parameters, implying a mirrored structure that equalizes self-impedances or admittances at both ports. A symmetric network often combines reciprocity with this port equivalence, leading to simplified parameter where the determinant of the matrix equals unity as a condition of . Such is common in balanced transformers and certain topologies, facilitating easier analysis and implementation. To verify reciprocity experimentally, one method involves interchanging the source and load positions between ports and measuring the transfer ratios, such as the ratio of output voltage to input current; equality confirms the property. For instance, applying a at port 1 and measuring at port 2, then swapping and repeating, yields identical ratios in reciprocal networks. This testing approach leverages the invariance of excitation-response ratios under port reversal, as defined in standard . Reciprocity and symmetry have significant implications for network design, particularly in passive components like filters and transformers, where they enable bidirectional signal handling and reduce complexity in modeling symmetric responses. In , reciprocal properties ensure consistent performance regardless of signal direction, aiding in the development of bandpass filters with uniform . Similarly, in transformers, simplifies and core modeling, enhancing efficiency in power and signal applications. Non-reciprocal two-port networks, by contrast, violate these conditions and are realized using active elements like transistors or ferrite materials in devices; examples include circulators, which direct signals unidirectionally via ferrite-based nonreciprocal shifts under magnetic bias.

Parameter Sets

Impedance Parameters (Z-Parameters)

The , or Z-parameters, characterize a two-port network by expressing the port voltages as linear functions of the port s under open-circuit conditions at the respective ports. These parameters are particularly useful for networks analyzed with current excitations and series connections, as they directly yield impedances in ohms. The defining equations are: V_1 = z_{11} I_1 + z_{12} I_2 V_2 = z_{21} I_1 + z_{22} I_2 where V_1 and V_2 are the voltages across ports 1 and 2, and I_1 and I_2 are the currents entering those ports. The individual Z-parameters are obtained by setting one current to zero: z_{11} = V_1 / I_1 with I_2 = 0 (port 2 open-circuited), z_{21} = V_2 / I_1 with I_2 = 0, z_{12} = V_1 / I_2 with I_1 = 0 (port 1 open-circuited), and z_{22} = V_2 / I_2 with I_1 = 0. This measurement approach reflects open-circuit impedance conditions, making Z-parameters ideal for scenarios where ports are not shorted during . Physically, z_{11} represents the driving-point input impedance at port 1 when port 2 is open, indicating how the network loads the source at the input. Similarly, z_{22} is the driving-point output impedance at port 2 when port 1 is open, showing the network's output loading effect. The off-diagonal terms z_{21} and z_{12} quantify the transfer impedances: z_{21} is the ratio of output voltage to input current with output open (forward ), and z_{12} is the ratio of input voltage to output current with input open (reverse ). These interpret as voltage ratios influenced by the network's internal coupling. In matrix notation, the Z-parameters compactly represent the network as: \begin{bmatrix} V_1 \\ V_2 \end{bmatrix} = \begin{bmatrix} z_{11} & z_{12} \\ z_{21} & z_{22} \end{bmatrix} \begin{bmatrix} I_1 \\ I_2 \end{bmatrix} This form facilitates analysis of series combinations, where the total Z-matrix is the sum of individual matrices, and all elements share units of ohms for dimensional consistency. For reciprocal networks, z_{12} = z_{21}, reflecting symmetric energy transfer between ports as discussed in network symmetry properties. The advantages of Z-parameters include their suitability for series-connected networks, where parameters add directly, and their intuitive impedance interpretation for voltage-current analyses in lumped circuits.

Admittance Parameters (Y-Parameters)

Admittance parameters, commonly referred to as Y-parameters or short-circuit , describe the behavior of a linear two-port network by expressing the port currents as linear functions of the port voltages. These parameters are obtained by applying voltages to one port while short-circuiting the other port to measure the resulting currents. The approach is particularly suited to networks where short-circuit conditions are practical for or . The fundamental equations defining the Y-parameters are: I_1 = y_{11} V_1 + y_{12} V_2 I_2 = y_{21} V_1 + y_{22} V_2 Here, I_1 and I_2 represent the currents entering the positive terminals of ports 1 and 2, respectively. In matrix notation, this relationship is: \begin{bmatrix} I_1 \\ I_2 \end{bmatrix} = \begin{bmatrix} y_{11} & y_{12} \\ y_{21} & y_{22} \end{bmatrix} \begin{bmatrix} V_1 \\ V_2 \end{bmatrix} The elements are specifically defined as y_{11} = \left. \frac{I_1}{V_1} \right|_{V_2 = 0}, y_{12} = \left. \frac{I_1}{V_2} \right|_{V_1 = 0}, y_{21} = \left. \frac{I_2}{V_1} \right|_{V_2 = 0}, and y_{22} = \left. \frac{I_2}{V_2} \right|_{V_1 = 0}, with short-circuit conditions applied as indicated. Physically, y_{11} is the input admittance observed at port 1 with port 2 short-circuited, y_{22} is the output admittance at port 2 with port 1 short-circuited, y_{21} is the forward transadmittance measuring the output current response to input voltage under shorted output conditions, and y_{12} is the reverse transadmittance capturing the input current response to output voltage under shorted input conditions. All Y-parameters have units of (S), reflecting their admittance nature. Y-parameters offer advantages in scenarios involving configurations of two-port s, where the overall Y-matrix is simply the of the individual matrices, simplifying analysis of shunt-connected systems. For s, the condition y_{12} = y_{21} holds. Consider a shunt consisting of a RC combination with Y = \frac{[1](/page/1)}{R} + j \omega C connected between the ports (with ). The Y-parameters for this configuration are: y_{11} = Y, \quad y_{12} = -Y, \quad y_{21} = -Y, \quad y_{22} = Y Thus, the matrix is: \begin{bmatrix} Y & -Y \\ -Y & Y \end{bmatrix} At high frequencies, where \omega is large and the capacitive term dominates, Y \approx j \omega C, so y_{21} \approx -j \omega C, demonstrating the network's behavior as primarily capacitive under short-circuit conditions.

Hybrid Parameters (H-Parameters)

Hybrid parameters, also known as h-parameters, provide a characterization of linear two-port networks by expressing the input voltage and output current in terms of the input current and output voltage. This mixed representation combines impedance-like and admittance-like terms, making it particularly suitable for analyzing active devices such as transistors where the input is driven by current and the output by voltage. The defining equations are: \begin{align*} V_1 &= h_{11} I_1 + h_{12} V_2, \\ I_2 &= h_{21} I_1 + h_{22} V_2. \end{align*} In matrix notation, this is written as: \begin{bmatrix} V_1 \\ I_2 \end{bmatrix} = \begin{bmatrix} h_{11} & h_{12} \\ h_{21} & h_{22} \end{bmatrix} \begin{bmatrix} I_1 \\ V_2 \end{bmatrix}. The individual parameters are determined under specific terminal conditions: h_{11} = \left. \frac{V_1}{I_1} \right|_{V_2 = 0}, the short-circuit input impedance; h_{12} = \left. \frac{V_1}{V_2} \right|_{I_1 = 0}, the open-circuit reverse voltage ratio; h_{21} = \left. \frac{I_2}{I_1} \right|_{V_2 = 0}, the short-circuit forward current gain; and h_{22} = \left. \frac{I_2}{V_2} \right|_{I_1 = 0}, the open-circuit output admittance. Physically, h_{11} represents the with the output port short-circuited, reflecting how the network loads the source; h_{12} quantifies the reverse voltage from output to input with the input open, indicating or ; h_{21} is the forward ratio under shorted output, akin to a factor; and h_{22} denotes the output with the input open, showing the network's output loading effect. These interpretations facilitate practical measurements and , especially at low frequencies where direct voltage and current probes are feasible. The h-parameters originated in the early as part of efforts to standardize transistor characterization amid rapid advancements in technology following the invention of the in 1947. Initially referred to as series-parallel parameters due to their mixed series (impedance) and parallel () nature, the term "hybrid" was coined by D. A. Alsberg in 1953 during discussions on at the IRE Convention, emphasizing their blend of voltage and current variables for analysis. This development occurred during the transition from circuits, where similar parameter sets were explored, to solid-state devices, with efforts by the IRE and AIEE in 1954 promoting their use in data sheets for consistency across manufacturers. A representative application is the common-emitter (CE) configuration of a (BJT) , where h-parameters are derived from the small-signal to predict performance. In this setup, the input port corresponds to the base-emitter junction driven by base current I_1 = I_b, and the output to the collector with voltage V_2 = V_{ce}. The forward current gain h_{21} (denoted h_{fe} in CE notation) approximates the transistor's small-signal current gain \beta, defined as h_{fe} = \left. \frac{I_c}{I_b} \right|_{V_{ce}=0} \approx \beta, where \beta typically ranges from 50 to 300 for BJTs and directly influences the 's voltage and power gains. The other parameters, such as h_{ie} (input resistance, often 1–5 kΩ) and h_{re} (reverse voltage feedback, usually small like 10^{-4}), are obtained by applying test signals to the pi-model equivalents, enabling straightforward calculation of overall circuit metrics like and output resistance without full simulation.

Inverse Hybrid Parameters (G-Parameters)

The inverse hybrid parameters, or g-parameters, characterize a two-port network by expressing the input current I_1 and output voltage V_2 as linear functions of the input voltage V_1 and output current I_2. The defining equations are: I_1 = g_{11} V_1 + g_{12} I_2 V_2 = g_{21} V_1 + g_{22} I_2 In matrix form, this relationship is represented as: \begin{pmatrix} I_1 \\ V_2 \end{pmatrix} = \begin{pmatrix} g_{11} & g_{12} \\ g_{21} & g_{22} \end{pmatrix} \begin{pmatrix} V_1 \\ I_2 \end{pmatrix} The individual parameters are defined under specific termination conditions: g_{11} = \left. \frac{I_1}{V_1} \right|_{I_2 = 0} (input with the output open-circuited), g_{12} = \left. \frac{I_1}{I_2} \right|_{V_1 = 0} (reverse transfer with the input short-circuited), g_{21} = \left. \frac{V_2}{V_1} \right|_{I_2 = 0} (forward voltage transfer with the output open-circuited), and g_{22} = \left. \frac{V_2}{I_2} \right|_{V_1 = 0} ( with the input short-circuited). These parameters have clear physical interpretations in circuit analysis: g_{11} represents the driving-point at the input under open-circuit conditions at the output, g_{12} quantifies the reverse of from the output to the input under short-circuit conditions at the input, g_{21} measures the forward of voltage from the input to the output under open-circuit conditions at the output, and g_{22} denotes the driving-point impedance at the output under short-circuit conditions at the input. The off-diagonal elements g_{12} and g_{21} are dimensionless, while g_{11} has units of (siemens) and g_{22} has units of impedance (ohms). The g-parameters are particularly advantageous for analyzing transistor-based amplifiers where the input is driven by a and the output delivers voltage, such as in configurations of bipolar junction transistors (BJTs), as they naturally align with low and high characteristics. For networks, which contain no dependent sources or non-reciprocal elements like gyrators, the condition g_{12} = -g_{21} holds, ensuring symmetry in the transfer characteristics. As an example, consider a common-base BJT amplifier analyzed using the small-signal hybrid-π model, where the base is grounded, the input is applied to the emitter (port 1), and the output is taken from the collector (port 2). The parameters are derived from the model elements: transconductance g_m = I_C / V_T (where V_T is the thermal voltage), base-emitter resistance r_\pi = \beta / g_m (with \beta the common-emitter current gain), and output resistance r_o. Approximating for high \beta and neglecting base-width modulation initially, g_{11} \approx g_m, g_{12} \approx -\alpha (where \alpha = \beta / (1 + \beta) \approx 1 is the common-base current gain, yielding g_{12} \approx -1), g_{21} \approx g_m r_o, and g_{22} \approx r_\pi / (\alpha r_o (1 + g_m r_\pi)). Including the Early effect via r_o, the forward voltage gain g_{21} is large under conditions where output resistance dominates and current transfer is nearly ideal, illustrating the configuration's utility for high voltage gain in buffered applications.

Transmission Parameters (ABCD-Parameters)

Transmission parameters, also known as ABCD-parameters, describe a two-port network by expressing the input voltage and current (at port 1) in terms of the output voltage and current (at port 2), which is particularly useful for analyzing networks where signal flow is predominantly from input to output, such as in or configurations. The defining equations are: \begin{align} V_1 &= A V_2 + B I_2, \\ I_1 &= C V_2 + D I_2, \end{align} where V_1 and I_1 are the voltage and current at the input port, V_2 and I_2 are those at the output port (with I_2 directed away from the network), and A, B, C, D are the transmission parameters. These parameters are determined under specific conditions: A is the ratio V_1 / V_2 when the output port is open-circuited (I_2 = 0), B is V_1 / I_2 when the output port is short-circuited (V_2 = 0), C is I_1 / V_2 with I_2 = 0, and D is I_1 / I_2 with V_2 = 0. In matrix form, the relationships are compactly represented as: \begin{bmatrix} V_1 \\ I_1 \end{bmatrix} = \begin{bmatrix} A & B \\ C & D \end{bmatrix} \begin{bmatrix} V_2 \\ I_2 \end{bmatrix}. The physical interpretations of these parameters reflect their roles in network behavior: A represents the voltage ratio or transfer function with an open output, B acts as a transfer impedance (with units of ohms), C serves as a transfer admittance (with units of siemens), and D denotes the current ratio with a shorted output. A and D are dimensionless, while B and C carry impedance and admittance dimensions, respectively. For reciprocal networks, which satisfy the reciprocity theorem (no dependent sources and passive elements), the determinant of the ABCD matrix equals unity: AD - BC = 1. Symmetric networks, where the ports are interchangeable, further require A = D. These properties are summarized in the following table:
PropertyConditionDescription
ReciprocityAD - BC = 1Holds for passive, linear networks without non-reciprocal elements.
A = DApplies when the network is symmetric about its midplane.
The primary advantage of ABCD-parameters lies in their suitability for cascading multiple two-port networks, such as in ladder networks or transmission lines, where the overall matrix is obtained by simple of individual ABCD matrices from right to left in the signal flow direction. This multiplicative property facilitates efficient analysis of complex systems without solving the full circuit equations each time.

Scattering Parameters (S-Parameters)

Scattering parameters, commonly known as S-parameters, characterize the response of linear electrical networks to incident and reflected traveling waves, making them particularly suitable for high-frequency applications where wave propagation effects dominate. Unlike voltage and current-based parameters, S-parameters relate the amplitudes of outgoing waves to incoming waves at the network ports, providing a framework that inherently accounts for mismatches and power transfer. This approach originated in to simplify analysis of distributed systems, such as transmission lines and antennas. For a two-port network, the S-parameters are defined through the relationships between the incident waves a_1 and a_2 (entering ports 1 and 2, respectively) and the reflected waves b_1 and b_2 (exiting the ports): b_1 = S_{11} a_1 + S_{12} a_2 b_2 = S_{21} a_1 + S_{22} a_2 In notation, this is expressed as: \begin{pmatrix} b_1 \\ b_2 \end{pmatrix} = \begin{pmatrix} S_{11} & S_{12} \\ S_{21} & S_{22} \end{pmatrix} \begin{pmatrix} a_1 \\ a_2 \end{pmatrix} Here, S_{11} represents the input at port 1 when port 2 is terminated in a matched load (a_2 = 0), quantifying the fraction of incident power reflected due to impedance mismatch. S_{21} is the forward , measuring the power transmitted from port 1 to port 2 under matched conditions at port 2, often interpreted as the forward gain or . S_{12} denotes the reverse , indicating between ports by showing power transmitted from port 2 to port 1. S_{22} is the output at port 2 when port 1 is matched (a_1 = 0). These parameters are dimensionless and complex-valued, capturing both (related to power ratios) and (related to delay). The incident and reflected waves are normalized to a reference impedance Z_0, typically 50 \Omega in RF and microwave systems to match common transmission lines and measurement equipment, ensuring that |S_{ii}|^2 directly represents the power reflection coefficient. For arbitrary or non-real Z_0 (e.g., in active devices or lossy lines), power waves are employed instead of voltage waves to maintain a physically meaningful interpretation in terms of available and delivered power, avoiding issues with non-positive power quantities. This normalization allows S-parameters to remain bounded by unity in magnitude for passive lossless networks, facilitating stability analysis. S-parameters offer significant advantages in high-frequency networks by naturally handling discontinuities, radiation losses, and distributed effects that challenge lumped-element models like . They enable straightforward cascading of networks through and are directly measurable using vector network analyzers (VNAs), which inject a swept-frequency signal into one port while terminating the other in a matched load to isolate each . This technique provides accurate characterization over broad bandwidths, essential for designing amplifiers, filters, and antennas in circuits.

Scattering Transfer Parameters (T-Parameters)

Scattering transfer parameters, commonly referred to as T-parameters, provide a of a two-port network by expressing the incident and reflected waves at the input port (port 1) in terms of those at the output port (port 2). This approach is especially valuable in high-frequency applications, such as circuits, where networks are often connected in and wave effects dominate. Unlike S-parameters, which relate outgoing waves to incoming waves at all ports simultaneously, T-parameters facilitate sequential analysis from input to output, making them ideal for systems with directed signal flow. The T-parameters are defined by the relations: a_1 = T_{11} a_2 + T_{12} b_2 b_1 = T_{21} a_2 + T_{22} b_2 where a_1 and b_1 are the incident and reflected waves at port 1, and a_2 and b_2 are those at port 2. In matrix notation, this is written as: \begin{bmatrix} a_1 \\ b_1 \end{bmatrix} = \begin{bmatrix} T_{11} & T_{12} \\ T_{21} & T_{22} \end{bmatrix} \begin{bmatrix} a_2 \\ b_2 \end{bmatrix} The matrix \mathbf{T} is unimodular for networks, satisfying \det \mathbf{T} = T_{11} T_{22} - T_{12} T_{21} = 1. To relate T-parameters to S-parameters, start with the S-parameter definitions: b_1 = S_{11} a_1 + S_{12} a_2 b_2 = S_{21} a_1 + S_{22} a_2 Solve the second equation for a_1: a_1 = \frac{b_2 - S_{22} a_2}{S_{21}} Substitute into the first equation: b_1 = S_{11} \left( \frac{b_2 - S_{22} a_2}{S_{21}} \right) + S_{12} a_2 = \frac{S_{11}}{S_{21}} b_2 + \left( S_{12} - \frac{S_{11} S_{22}}{S_{21}} \right) a_2 Thus, the T-parameter elements are: T_{11} = -\frac{S_{22}}{S_{21}}, \quad T_{12} = \frac{1}{S_{21}} T_{21} = S_{12} - \frac{S_{11} S_{22}}{S_{21}} = -\frac{\Delta_S}{S_{21}}, \quad T_{22} = \frac{S_{11}}{S_{21}} where \Delta_S = S_{11} S_{22} - S_{12} S_{21} is the determinant of the S-matrix. This conversion assumes S_{21} \neq 0, which holds for networks allowing transmission. The derivation highlights how T-parameters rearrange the S-matrix relations to prioritize chain-like propagation. A key advantage of T-parameters lies in their compatibility with cascade connections. For two two-port networks in series, the overall T-matrix is the product \mathbf{T} = \mathbf{T}_1 \mathbf{T}_2, enabling straightforward computation of the composite network without resolving intermediate waves. This mirrors the utility of ABCD-parameters but employs normalized power waves, preserving energy conservation and matching conditions in transmission lines. T-parameters thus excel in scenarios involving non-reciprocal components, such as amplifiers or circulators, where reflections and transmissions are asymmetric. In practice, T-parameters find extensive use in microwave filter design, where cascaded resonators and sections require iterative matrix multiplications for optimization, and in phased antenna arrays, facilitating analysis of signal distribution across elements. Their wave-based formulation inherently accounts for mismatches better than voltage-current parameters at high frequencies. In the low-frequency limit, where phase shifts across the network are negligible and wave effects diminish, T-parameters approximate the (ABCD) parameters. Here, the incident and reflected relate to voltage V and I via a = \frac{V + Z_0 I}{2 \sqrt{Z_0}} and b = \frac{V - Z_0 I}{2 \sqrt{Z_0}}, with Z_0 the reference impedance. The resulting T-matrix elements align with ABCD as T_{11} \approx A / \sqrt{Z_0}, T_{12} \approx B / Z_0, T_{21} \approx C \sqrt{Z_0}, and T_{22} \approx D, bridging classical circuit analysis with RF behavior.

Network Connections

Series and Parallel Configurations

In two-port networks, series and parallel configurations allow for the combination of multiple networks while preserving the overall port structure. The series-series connection involves connecting the input ports of two networks in series and their output ports in series, resulting in the voltages adding at each port while currents remain equal. This configuration is particularly suited to (z-parameters), where the total impedance matrix is the sum of the individual matrices: \mathbf{Z}_{\text{total}} = \mathbf{Z}_1 + \mathbf{Z}_2. Thus, z_{11,\text{total}} = z_{11,1} + z_{11,2}, z_{12,\text{total}} = z_{12,1} + z_{12,2}, z_{21,\text{total}} = z_{21,1} + z_{21,2}, and z_{22,\text{total}} = z_{22,1} + z_{22,2}. The parallel-parallel connection places the input ports in parallel and the output ports in parallel, so currents add at each port while voltages are equal. Admittance parameters (y-parameters) are ideal here, with the total admittance matrix given by the sum: \mathbf{Y}_{\text{total}} = \mathbf{Y}_1 + \mathbf{Y}_2. Specifically, y_{11,\text{total}} = y_{11,1} + y_{11,2}, y_{12,\text{total}} = y_{12,1} + y_{12,2}, y_{21,\text{total}} = y_{21,1} + y_{21,2}, and y_{22,\text{total}} = y_{22,1} + y_{22,2}. This addition simplifies analysis for networks like parallel amplifiers or filters. For the series-parallel configuration, the input ports are connected in series (voltages add, currents equal) while the output ports are in parallel (currents add, voltages equal). parameters (h-parameters) are appropriate, as they model input voltage in terms of input and output voltage, and output in terms of input and output voltage. The total h-parameter matrix is the sum of the individual matrices: \mathbf{H}_{\text{total}} = \mathbf{H}_1 + \mathbf{H}_2, yielding h_{11,\text{total}} = h_{11,1} + h_{11,2}, h_{12,\text{total}} = h_{12,1} + h_{12,2}, h_{21,\text{total}} = h_{21,1} + h_{21,2}, and h_{22,\text{total}} = h_{22,1} + h_{22,2}. This setup is common in amplifier stages where input series elements combine with parallel output loads. The parallel-series configuration reverses this, with input ports in (currents add, voltages equal) and output ports in series (voltages add, currents equal). Inverse hybrid parameters (g-parameters) suit this arrangement, relating input current to input voltage and output current, and output voltage to input voltage and output current. The total g-parameter matrix adds directly: \mathbf{G}_{\text{total}} = \mathbf{G}_1 + \mathbf{G}_2, so g_{11,\text{total}} = g_{11,1} + g_{11,2}, g_{12,\text{total}} = g_{12,1} + g_{12,2}, g_{21,\text{total}} = g_{21,1} + g_{21,2}, and g_{22,\text{total}} = g_{22,1} + g_{22,2}. Such connections appear in current-driven circuits with series output impedances.

Cascade Configurations

In cascade configurations, two or more two-port networks are interconnected end-to-end such that the output port of the first network connects directly to the input port of the second, allowing signal flow from the input of the first to the output of the last. This arrangement is common in systems like multi-stage amplifiers or sections, where the overall behavior is determined by multiplying the parameter matrices of the individual networks. Transmission parameters, such as ABCD-parameters and T-parameters ( transfer parameters), are particularly suited for this due to their chain-like representation that facilitates straightforward without needing to account for intermediate terminations explicitly. The ABCD-parameters define the relationship between the input voltage V_1 and current I_1 at port 1 and the output voltage V_2 and current I_2 at port 2 (with I_2 directed out of the network) as: \begin{bmatrix} V_1 \\ I_1 \end{bmatrix} = \begin{bmatrix} A & B \\ C & D \end{bmatrix} \begin{bmatrix} V_2 \\ -I_2 \end{bmatrix}. For two networks in cascade, the overall ABCD matrix is the product [ABCD]_{\text{total}} = [ABCD]_1 [ABCD]_2, where the matrix for the input-side network (network 1) is multiplied by that of the output-side network (network 2). This multiplication yields the total A, B, C, and D elements that describe the combined network. Similarly, for T-parameters, which relate the incident and reflected waves from left to right, the overall T-matrix is [T]_{\text{total}} = [T]_1 [T]_2, enabling efficient analysis in high-frequency applications like microwave chains. This approach applies directly to a two-stage , where the ABCD-parameters of each common-emitter stage (derived from hybrid h-parameters and external components) are multiplied to obtain the total, revealing the cumulative through the A and D elements under open-circuit conditions at the output. When cascading networks, between stages is essential to maximize power transfer and minimize reflections, particularly at high frequencies. Non-ideal effects, such as loading where the of the second network alters the output conditions of the first, are inherently captured in the matrix product assuming direct connection; however, mismatches can degrade performance, necessitating buffers or matching networks in practical designs.

Parameter Relationships

Conversions Between Parameter Sets

Conversions between different parameter sets for two-port networks are essential for analysis and design, allowing engineers to switch representations based on the application's requirements, such as or cascading networks. These transformations are derived from the fundamental voltage and current relationships defining each set and assume the network is linear and time-invariant. All major conversions can be expressed using matrix algebra, with specific element-wise formulas for the 2×2 matrices involved. The \mathbf{Y} are directly related to the \mathbf{Z} by matrix inversion: \mathbf{Y} = \mathbf{Z}^{-1} The follows as \det(\mathbf{Y}) = 1 / \det(\mathbf{Z}). This is particularly useful when connections are involved, as \mathbf{Y} parameters add directly. The parameters \mathbf{H} (h-parameters) can be obtained from \mathbf{Z} parameters as follows: h_{11} = \frac{\det(\mathbf{Z})}{z_{22}}, \quad h_{12} = \frac{z_{12}}{z_{22}}, \quad h_{21} = -\frac{z_{21}}{z_{22}}, \quad h_{22} = \frac{1}{z_{22}} where \det(\mathbf{Z}) = z_{11} z_{22} - z_{12} z_{21}. This transformation facilitates transistor modeling, where input impedance and output admittance are mixed. The inverse hybrid parameters \mathbf{G} (g-parameters) are simply the matrix inverse of the hybrid parameters: \mathbf{G} = \mathbf{H}^{-1} with elements g_{11} = h_{22} / \det(\mathbf{H}), g_{12} = -h_{12} / \det(\mathbf{H}), g_{21} = -h_{21} / \det(\mathbf{H}), and g_{22} = h_{11} / \det(\mathbf{H}), where \det(\mathbf{H}) = h_{11} h_{22} - h_{12} h_{21}. This set is advantageous for certain amplifier analyses involving output short-circuit conditions. For scattering parameters \mathbf{S}, conversions from transmission (ABCD) parameters assume a common reference impedance Z_0 at both ports and are given by: Let \Delta = A + \frac{B}{Z_0} + C Z_0 + D. Then, S_{11} = \frac{A + \frac{B}{Z_0} - C Z_0 - D}{\Delta}, \quad S_{21} = \frac{2}{\Delta}, S_{12} = \frac{2 (A D - B C)}{\Delta}, \quad S_{22} = \frac{-A + \frac{B}{Z_0} + C Z_0 - D}{\Delta}. These formulas hold for reciprocal networks where \det(\mathbf{ABCD}) = A D - B C = 1, simplifying S_{12} = S_{21}. The reference Z_0 (typically 50 \Omega) normalizes the waves for power-based analysis. The scattering transfer parameters \mathbf{T} (also known as T-parameters) relate to \mathbf{S} through the chain scattering matrix formulation. Assuming the convention where \begin{pmatrix} a_1 \\ b_1 \end{pmatrix} = \mathbf{T} \begin{pmatrix} a_2 \\ b_2 \end{pmatrix}, the elements convert as: S_{11} = \frac{T_{22}}{T_{11}}, \quad S_{12} = -\frac{T_{12}}{T_{11}}, \quad S_{21} = \frac{1}{T_{11}}, \quad S_{22} = -\frac{T_{21}}{T_{11}}. This is particularly useful for cascading multiple two-port networks, as \mathbf{T} matrices multiply directly. These conversions are valid for linear networks and preserve properties like reciprocity (symmetric matrices for \mathbf{Z}, \mathbf{Y}, \mathbf{H}, \mathbf{G}) when applicable. However, at high frequencies, direct computation in \mathbf{Z} or \mathbf{Y} domains can encounter numerical instability due to large or near-zero elements, making \mathbf{S}-parameters preferable for microwave applications where measurements are normalized to Z_0.

Conditions for Special Networks

Two-port networks exhibit special properties based on their parameter sets, allowing classification into categories such as reciprocal, symmetric, passive, or lossless. These properties are determined by specific conditions on the , , or parameters, which can be tested to verify the network's behavior without requiring full circuit analysis. A reciprocal two-port network is one where the transmission characteristics are identical in both directions, meaning the response from port 1 to port 2 equals that from port 2 to port 1. In terms of Z-parameters, reciprocity holds if z_{12} = z_{21}. For ABCD-parameters, the condition is AD - BC = 1. With S-parameters, reciprocity implies a where S_{12} = S_{21}, and for lossless reciprocal cases, the determinant satisfies |S_{11}S_{22} - S_{12}S_{21}| = 1, aligning with the unitary property. Symmetry in a two-port network occurs when the ports are interchangeable, yielding identical input and output impedances or characteristics. For Z-parameters, this requires z_{11} = z_{22}. In ABCD-parameters, symmetry is indicated by A = D. These conditions ensure the network behaves equivalently regardless of port orientation, common in balanced transmission lines or filters. Passivity defines networks that do not generate power, absorbing or storing it instead. The Z-matrix must have a real part that is positive semi-definite, meaning \mathbf{Re}(\mathbf{Z}) \succeq 0 for all frequencies, ensuring no negative resistances. For S-parameters, passivity requires |S_{ii}| \leq 1 for i = 1, 2, limiting reflected or transmitted power to at most the incident power. These conditions prevent and maintain energy conservation in the network. Lossless networks conserve all power, with no dissipation. The S-matrix is unitary, satisfying \mathbf{S} \mathbf{S}^\dagger = \mathbf{I}, where \dagger denotes the , ensuring total power reflection and transmission sum to unity at each port. In the Z-domain, all elements are purely imaginary, representing only with \mathbf{Re}(\mathbf{Z}) = 0. For reciprocal lossless networks, the ABCD-parameters further specify A and D as real, while B and C are imaginary. Non-reciprocal networks violate directional symmetry, often due to active elements or magnetic biasing. Examples include amplifiers with , where forward exceeds reverse transmission, failing conditions like z_{12} = z_{21} or AD - BC = 1; parameter conversions reveal asymmetries such as S_{12} \neq S_{21}. amplifiers also exemplify non-reciprocity, enabling unidirectional signal flow in applications like isolators.

Advanced Topics

Multi-Port Networks

A multi-port generalizes the two-port framework to an arbitrary number of ports N, where the port voltages \mathbf{V} and currents \mathbf{I} are related by the impedance matrix \mathbf{V} = \mathbf{Z} \mathbf{I}, with \mathbf{Z} being an N \times N matrix whose elements Z_{ij} represent the voltage at port i due to current at port j (with other currents zero). The is reciprocal if \mathbf{Z} is symmetric, satisfying Z_{ij} = Z_{ji} for all i, j, a arising from the linearity and symmetry of the underlying electromagnetic fields. This extension allows modeling of complex systems where interactions occur among multiple access points, building on the voltage-current relations familiar from two-port parameters. Three-port networks, a common case of multi-port configurations, are essential in microwave applications such as power dividers, which split input signals into multiple outputs, and circulators, which direct signals unidirectionally between ports. These devices cannot simultaneously be lossless, , and matched at all ports, often requiring trade-offs like added resistance for matching in power dividers. (Y-parameters) are particularly suited for analyzing three-port networks in nodal formulations, where currents relate to voltages via \mathbf{I} = \mathbf{Y} \mathbf{V}, facilitating simulations by focusing on potentials. Scattering parameters extend naturally to multi-port networks as an N \times N , crucial for characterizing systems where incident and reflected waves at each port define transmission and reflection behaviors. For lossless multi-port networks, the is unitary, satisfying \mathbf{S}^\dagger \mathbf{S} = \mathbf{I}, which enforces power conservation across all ports. This property is vital in design, treating as generalized one-ports within larger multi-port assemblies. Analyzing multi-port networks presents challenges due to the need for N^2 independent measurements to fully determine parameter matrices like or , complicating and hardware requirements in vector network analyzers. Reduction techniques, such as , address this by synthesizing equivalent two-port models from the full multi-port description, simplifying analysis for specific excitation scenarios without losing essential coupling information. Multi-port networks are applied in radio-frequency integrated circuits (RFICs) for components like multi-port hybrids and dividers that enable and signal in mm-wave systems. In multi-antenna systems, they support multiple-input multiple-output () configurations, optimizing phase and amplitude across ports for enhanced data rates in and applications.

Reduction to One-Port Equivalents

A two-port network can be reduced to an equivalent one-port network by terminating one of its ports with a specified impedance, allowing analysis of the behavior at the remaining port as if it were a single-port device. This reduction is fundamental in circuit design, enabling the calculation of input or output characteristics under practical loading conditions. The resulting one-port equivalent is characterized by its impedance or admittance, which depends on the termination and the two-port parameters such as Z-parameters or Y-parameters. When the output port is terminated with load impedance Z_L, the input impedance Z_{in} seen at the input port is derived from the Z-parameters as Z_{in} = z_{11} - \frac{z_{12} z_{21}}{z_{22} + Z_L}. This expression accounts for the interaction between ports through the off-diagonal terms z_{12} and z_{21}. Equivalently, using Y-parameters, the input admittance Y_{in} is Y_{in} = y_{11} - \frac{y_{12} y_{21}}{y_{22} + Y_L}, where Y_L = 1/Z_L, and Z_{in} = 1/Y_{in}. These formulas facilitate the design of amplifiers and filters by predicting how the load affects the input match. Symmetrically, the output impedance Z_{out} is obtained by terminating the input port with source impedance Z_S: Z_{out} = z_{22} - \frac{z_{12} z_{21}}{z_{11} + Z_S}. This measures the impedance looking into the output port, essential for maximum power transfer from the network. Special cases arise with open-circuit or short-circuit terminations. For an open-circuit output (Z_L \to \infty), Z_{in} = z_{11}, the open-circuit input impedance. For a short-circuit output (Z_L = 0), Z_{in} = (z_{11} z_{22} - z_{12} z_{21}) / z_{22}, which is the determinant of the Z-matrix divided by z_{22}. These extremes are used to extract individual Z-parameters experimentally. Furthermore, the terminated two-port can be represented by Thévenin or Norton equivalents at the active port. The Thévenin equivalent consists of an open-circuit voltage and the computed impedance (e.g., Z_{out} with input terminated), while the Norton equivalent uses a short-circuit current and the admittance dual. Such equivalents simplify cascading with other circuits. In the scattering parameter framework, the input reflection coefficient \Gamma_{in} for a load reflection coefficient \Gamma_L is \Gamma_{in} = S_{11} + \frac{S_{12} S_{21} \Gamma_L}{1 - S_{22} \Gamma_L}. This expression is particularly useful in for analyzing reflections under arbitrary terminations. These one-port equivalents find applications in stability assessment, impedance matching, and noise analysis. For stability in active networks like amplifiers, the real part of Z_{in} must be positive (\Re(Z_{in}) > 0) to prevent oscillations, ensuring |\Gamma_{in}| < 1 for all passive loads. Impedance matching sets Z_L such that Z_{in} = Z_S^* or \Gamma_{in} = 0, maximizing power transfer. In noise figure calculations, the one-port reduction with output terminated allows evaluation of the minimum noise figure by optimizing source impedance, as the noise performance depends on the effective input admittance.

References

  1. [1]
    [PDF] appendix w2 - two-port networks - UCSB ECE
    A port is a terminal pair where energy can be supplied or extracted. A two-port. network is a four-terminal circuit in which the terminals are paired to form ...
  2. [2]
  3. [3]
    [PDF] Two-Port Network Parameters
    A two-port network (Fig. B.I) has four port variables: V₁, I₁, V2, and 12. If the two-port net- work is linear, we can use two of the variables as ...
  4. [4]
    Two-Port Networks
    Two-Port Networks. Models of two-port networks. Many complex, such as amplification circuits and filters, can be modeled by a two-port network model as ...Missing: graphical chain diagram line lattice
  5. [5]
    [PDF] CHAPTER 17 - UCSB ECE
    A two-port network is a four-terminal circuit in which the terminals are paired to form an input port and an output port. Figure 17–1 shows the customary way of ...<|control11|><|separator|>
  6. [6]
    Lattice Network | Series Arms | Diagonal Arms - eeeguide.com
    The lattice network is one of the common four terminal networks. These networks are generally used in filter sections and also used as symmetrical attenuators.
  7. [7]
    [PDF] A Short History of Circuits and Systems - IEEE CAS
    two port networks” IEEE ISCAS, Vol. 4 pp. 1911 - 1913 vol.4, 1992. [32] ... of modern network theory,” History of Telecommunications Conference, IEEE.
  8. [8]
    Cascade ABCD two-port networks - EDN
    Dec 23, 1999 · The telephone industry has always used two-port networks in the form of ABCD matrices to cascade sections of telephone cable.
  9. [9]
    ABCD Parameters of Transmission Lines - Cadence System Analysis
    The ABCD parameters of a transmission line give the relationship of the input voltage and currents to the output voltage and currents.
  10. [10]
  11. [11]
    [PDF] Chapter 6 Two-Port Network Model - VTechWorks
    As a transducer converts energy from electrical to mechanical forms, and vice- versa, it can be modelled as a Two-Port Network that relates the electrical ...
  12. [12]
  13. [13]
    [PDF] 6.061 Class Notes, Chapter 1: Review of Network Theory
    Electric network theory deals with two primitive quantities, which we will refer to as: 1. Potential (or voltage), and. 2. Current.
  14. [14]
    2.2: Two-Port Networks - Engineering LibreTexts
    May 22, 2022 · A port voltage is defined as the voltage difference between a pair of terminals with one of the terminals in the pair becoming the reference terminal.
  15. [15]
    [PDF] AN215A: RF Small Signal Design Using Two-Port Parameters
    INTRODUCTION. Design of the solid-state, small-signal RF amplifier using two-port parameters is a systematic, mathematical procedure, with an exact solution ...
  16. [16]
    [PDF] On Reciprocity in Linear Time-Invariant Networks
    —A general time-domain definition of reciprocity is given in terms of network-port variables, and this definition is applied to conclude the symmetry of network ...<|separator|>
  17. [17]
    11.2 Z, Y, h, and ABCD parameters - Fiveable
    Symmetry in two-port networks implies identical behavior from either port. Occurs when network structure is mirrored around its center; Indicated by Z11 = Z22 ...
  18. [18]
    10.8: CONDITION FOR RECIPROCITY - GlobalSpec
    A two-port network is said to be reciprocal if the ratio of the excitation to the response is invariant to an interchange of the positions of the excitation ...
  19. [19]
    How To Find Z Parameters of Two Port Network (Examples)
    Apr 27, 2018 · Applying voltage V1 at the output port and observing current I1 indicates that the network exhibits mirror-like symmetry between the input and ...
  20. [20]
    Z-parameters definition - Newark Electronics
    Z-parameters are also known as open-circuit impedance parameters as they are calculated under open circuit conditions. i.e., Ix=0, where x=1,2 refer to input ...
  21. [21]
    Two-Port Impedance Model and Z-Parameters
    The two-port network model is a popular modeling technique used to characterize the electrical and electronic circuits.
  22. [22]
    Network Theory - Two-Port Networks - Tutorials Point
    Two port network is a pair of two terminal electrical network in which, current enters through one terminal and leaves through another terminal of each port.
  23. [23]
    Improved Bipolar Current Mirrors | Electronics Textbook
    Figure 4-8. Emitter resistors reduce the current mirror voltage dependence to 0.7%. A measure of the quality of a current source is its output impedance, which ...Missing: parameters | Show results with:parameters
  24. [24]
    [PDF] TWO-PORT NETWORK PARAMETERS - Oxford University Press
    The definition and the method of measuring each of the four z parameters are given in Fig. C.3. C.1.3 h. The hybrid (or h-parameter) characterization of two- ...
  25. [25]
    How To Find Y Parameters of Two Port Network (Examples)
    Jun 24, 2017 · Y parameters (also known as admittance parameters or short-circuit parameters) are properties used in electrical engineering to describe the electrical ...
  26. [26]
    [PDF] Hybrid Transistor Model for small AC
    Typical values for the h parameters for a 2N3904 transistor in the common emitter configuration: h fe. = 120, h oe. = 8.7x10-6. Ω. -1, h ie = 3700 Ω, h re. = ...
  27. [27]
    10.6: INVERSE HYBRID (OR g) PARAMETERS - GlobalSpec
    The constants g 11, g 12, g 21, and g 22 are known as inverse hybrid parameters or g-parameters. The g-parameters are defined as follows by using Equations (10 ...
  28. [28]
    None
    ### Summary of Transmission Matrix (ABCD Parameters)
  29. [29]
    None
    ### Summary of Transmission Parameters Section from https://web.engr.oregonstate.edu/~webbky/ESE470_files/Section%204%20Transmission%20Lines.pdf
  30. [30]
    [PDF] MICROWAVE AND RF DESIGN NETWORKS - NC State Repository
    The book series Microwave and RF Design is a comprehensive treatment of radio frequency (RF) and microwave design with a modern “systems- first” approach.
  31. [31]
    A Primer on Scattering Parameters, Part I: Definitions and Properties
    Abstract—This primer offers a simple and comprehensive overview of the properties and usage of the scattering parameters of linear n-port elements.
  32. [32]
    [PDF] RF engineering basic concepts: S-parameters
    The physical meaning of S11 is the input reflection coefficient with the output of the network terminated by a matched load (a2 = 0). S21 is the forward ...
  33. [33]
    Understanding S-parameters | Rohde & Schwarz
    S parameters are the standard way of quantifying network characteristics. S parameters are measured by injecting power into one port and measuring the power at ...
  34. [34]
    4.2: Measurement of Scattering Parameters - Engineering LibreTexts
    Mar 21, 2021 · S parameters are measured using a network analyzer called an automatic network analyzer (ANA), or more commonly a vector network analyzer (VNA).
  35. [35]
    [PDF] Conversions between S, Z, Y, H, ABCD, and T parameters which are ...
    T parameters to the other 2-port parameters includes complex source and load impedances. The equations developed in this work are valid with port 1 and port ...
  36. [36]
    Microwaves101 | Basic network theory - Microwave Encyclopedia
    ... T-parameters. ... For those of you who enjoy some good matrix algebra derivations, we refer you to Pozar's book Microwave Engineering, which you can find on our ...
  37. [37]
    None
    ### Summary of Interconnection of Two-Port Networks (Series and Parallel)
  38. [38]
    None
    ### Summary of Interconnection of Two-Port Networks from the Document
  39. [39]
    Interconnection of Two Port Network - eeeguide.com
    Let the y-parameters of the network N” be y”11, y”12, y”21, y”22. Let the overall y-parameters of parallel connection be y11, y12, y21, y22.
  40. [40]
    [PDF] Two port Networks - parallel and series connections. - EE@IITM
    Similarly h22 = h22a +h22b and h12 = h12a −h12b . Parallel-series connections can be done similarly with g-parameters as the preferred parameters.
  41. [41]
    Interconnection of Two Port Network: Know Series, Parallel ...
    When two identical two port networks are connected in parallel, then Y- parameters are two be added. In a parallel connection, the input voltage V1 is the same ...
  42. [42]
    [PDF] Two-ports
    Figure: A Two-port may be described in one of the six common ways. Impedance (Z-parameters) description: V1(s). V2(s).
  43. [43]
    [PDF] A Guaranteed Passive Model for Multi-port Frequency Dependent ...
    Poles on the jω axis are simple and their matrix of residues is a positive semi-definite matrix. •. The real part of Z(jω), ℜ{Z(jω)}, is positive semi- definite ...
  44. [44]
    N-port network - UC Homepages - University of Cincinnati
    A typical Two-Port network is shown in the figure below. ... Again two reference planes are chosen; one for the input and one for the output. The E and H fields ...
  45. [45]
    [PDF] THEORY OF ACTIVE NONRECIPROCAL NETWORKS - DTIC
    Two simple devices exhibiting nonreciprocity are discussed in detail: a differential amplifier, and a differential attenuator; and a procedure is given for ...
  46. [46]
    [PDF] Module 13: Network Analysis and Directional Couplers
    Two-port parameters describe a system in terms of the voltage and current that may be measured at each port.Missing: definition | Show results with:definition
  47. [47]
    [PDF] RF Engineering - Basic Concepts: S-Parameters
    An N-port is passive and lossless if its S matrix is unitary, i.e.. S†S = 1 ... 1-ports (Note: an antenna is basically a generalized transformer which.
  48. [48]
    Multiport and differential S-parameter measurements (Chapter 9)
    Microwave designers and engineers are thus facing new challenges in multiport measurement hardware and calibrations. ... multiport when measured with a two-port ...
  49. [49]
    (PDF) Synthesis of Multiport Networks Using Port Decomposition ...
    Aug 5, 2025 · The heart of the algorithm lies with a novel port decomposition technique introduced here, wherein a reduced two-port network (2PN) is obtained ...
  50. [50]
    Deep-learning enabled generalized inverse design of multi-port ...
    Dec 30, 2024 · We demonstrate a universal inverse design approach for arbitrary-shaped complex multi-port electromagnetic structures with designer radiative and scattering ...
  51. [51]
    [PDF] Lecture 16: Properties of Two-Ports - University of California, Berkeley
    The stability and power gain of the two-port is determined by these quantities. • In terms of y-parameters. Yin = I1. V1. = Y11V1 + ...
  52. [52]
    None
    ### Summary of Input and Output Impedances for Two-Port Networks Using Z-Parameters
  53. [53]
    [PDF] Two-Port Gain and Stability - University of California, Berkeley
    N-Port Passivity. • We would like to find if an N-port is active or passive. By definition, an N-port is passive if it can only absorb net power. The total ...
  54. [54]
    2.3: Scattering Parameters - Engineering LibreTexts
    May 22, 2022 · The essence of scattering parameters (or S parameters 1 ) is that they relate forward- and backward-traveling waves on a transmission line, thus ...
  55. [55]
    [PDF] MT-052 TUTORIAL - Analog Devices
    The gain and the noise of a two-port network can now be defined in terms of the available power gain, G, and the noise factor, F, as shown below in Figure 4.