Complete graph equivalent circuit of passive devices based on Y parameter matrix and its construction method

Abstract

The present invention relates to a complete graph equivalent circuit of a passive device based on the Y-parameter matrix and a construction method, and relates to the field of electronic device modeling. The present invention is directed to a passive device having n ports P1, P2, …, Pn and a ground port P0, and there is a sub-circuit C determined by the corresponding transfer function between any two different ports Pi, Pj ij , where Pi, Pj ∈ {P0, P1, P2, …, Pn} and Pi ≠ Pj, thus forming a complete graph network; the sub-circuit C ij is composed of a plurality of sub-modules M ijk connected in parallel. Under the condition of knowing the Y-parameter transfer function matrix of the passive device, an equivalent circuit of any multi-port passive device is constructed. This model can ignore the influence of other ports without losing accuracy, and quickly and accurately obtain the topological structure of the sub-circuit and the values of each ideal device, having high efficiency and universality.

Description

Technical Field

[0001] The present invention relates to the field of electronic device modeling, and in particular to a passive device complete graph equivalent circuit based on a Y parameter matrix and a construction method thereof. Background Art

[0002] With the advancement of CMOS technology, the RF performance of CMOS devices has been continuously improved, which has greatly increased the application of silicon-based RF integrated circuits. As the key to RF circuits (such as amplifiers, mixers and voltage-controlled oscillators), passive components usually occupy a large area.

[0003] Passive device simulation mostly uses numerical techniques of electromagnetic (EM) simulation, but it is usually very time-consuming and requires a lot of computer resources, especially memory storage space and CPU time. A simple method is to use tabular data from measurement or EM simulation, that is, table lookup, however, this type of model cannot support time domain analysis. The equivalent circuit model can greatly reduce the calculation time and can be easily integrated into computer-aided design (CAD) tools, which helps to simulate the designed circuit quickly, accurately and flexibly. Therefore, in RF integrated circuit design, accurate modeling of passive devices is very necessary.

[0004] The low-order rational function approximation of the passive device system function matrix can be used to fit a rational expression to the data set obtained by EM simulation using the least squares method. However, for CAD tools, it requires an equivalent circuit model for simulation and optimization. Nowadays, many equivalent circuit models based on the zero-pole method of passive devices have been proposed, but most of them require determining the equivalent circuit topology and then solving a set of linear equations through Kirchhoff's law. Not only does it take a lot of manpower to determine the circuit topology, but the time to solve the equations is greatly increased when the network is complex. Summary of the invention

[0005] Aiming at the deficiencies of the prior art, the present invention quickly constructs a passive device complete graph equivalent circuit based on a Y parameter matrix after determining the rational expression of a system function matrix.

[0006] The equivalent circuit of the passive device complete graph based on the Y parameter matrix. For a passive device with n ports P1, P2, ..., Pn and a ground terminal P0, there is a subcircuit C determined by the corresponding transfer function between any two different ports Pi and Pj. ij , Pi, Pj∈{P0, P1, P2,…, Pn} and Pi≠Pj, thus forming a complete graph network.

[0007] Furthermore, the sub-circuit C ij It consists of several submodules M ijkIn parallel; and for any Pi, Pj, the sub-circuit C ij The number of submodules K ij are the same and are denoted as K.

[0008] Furthermore, the submodule M ijk It is composed of ideal elements such as resistors, inductors, capacitors or controlled current sources, and is determined by the kth term after the corresponding transfer function is converted into a partial fraction, 1≤k≤K.

[0009] The construction method of the passive device complete graph equivalent circuit model is as follows:

[0010] Step 1: Convert each term in the known Y parameter transfer function matrix into partial fraction form and determine the number of terms K.

[0011] Step 2: For any two different ports Pi, Pj∈{P1, P2, …, Pn} and Pi≠Pj, their transfer functions Y ij The kth term after (s) is converted into a partial fraction determines its kth submodule M ijk ; If the kth fraction is in the form of a / (sp), then consider a submodule in which an inductor and a resistor are connected in series; if the kth fraction is in the form of a / (sp), a≤0, then consider a submodule in which an inductor and a resistor are connected in series and then connected in parallel with a controlled current source, wherein the controlled current source is only used to correct the direction of the current without changing the poles and passivity of the module; if the kth fraction is in the form of (as+b) / (sp)(sp*), p and p* are conjugate complex numbers, then consider a submodule in which an inductor, a capacitor, and a resistor are connected in series and then connected in parallel with several controlled current sources, wherein the controlled current source is only used to correct the direction of the current without changing the poles and passivity of the module; if the kth term is in the form of a constant term, consider a single resistor submodule; if the kth term is in the form of as, consider a single capacitor submodule.

[0012] Step 3: For any port Pi∈{P1, P2, …, Pn}, its transfer function Y ii -∑ i≠j Y ij Determine the subcircuit C with P0 in the same way as in step 2 i0 .

[0013] Compared with other models, the advantages of the present invention are:

[0014] 1. This model can model each item separately through the partial fraction expression of the system function, that is, each item corresponds to a submodule. When the number of passive device ports increases and the complexity increases, the computational complexity increases linearly. Compared with other models, the computational complexity of establishing the equivalent circuit model is reduced, and it is highly efficient.

[0015] 2. For CAD tools, this model can shield the ideal components between the remaining ports on the complete graph when calculating the response between any two ports, and only needs to consider the subcircuit between the two related ports. Therefore, it is also efficient when reversing the system function rational expression from the equivalent circuit.

[0016] 3. This model is applicable to any passive device and does not require the construction of new topological structures for passive devices with different characteristics. It is universal. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 It is a schematic diagram of a passive device complete graph equivalent circuit model based on the Y parameter transfer function matrix;

[0018] Figure 2 is an equivalent circuit diagram of a passive device in an embodiment;

[0019] Figure 3 The figure is a comparison spectrum of S parameters of an equivalent circuit of an embodiment and its original device. DETAILED DESCRIPTION

[0020] The present invention will be further described below in conjunction with the accompanying drawings.

[0021] like Figure 1 As shown, a passive device complete graph equivalent circuit model based on the Y parameter transfer function matrix is ​​characterized in that: for a passive device having n ports P1, P2, ..., Pn and a ground terminal P0, there is a subcircuit C determined by the corresponding transfer function between any two different ports Pi and Pj. ij , Pi, Pj∈{P0, P1, P2,…, Pn} and Pi≠Pj, thus forming a complete graph network.

[0022] The subcircuit C ij It consists of several submodules M ijk In parallel; and for any Pi, Pj, the sub-circuit C ij The number of submodules K ij are the same and are denoted as K.

[0023] The submodule M ijk It is composed of ideal elements such as resistors, inductors, capacitors or controlled current sources, and is determined by the kth term after the corresponding transfer function is converted into a partial fraction, 1≤k≤K.

[0024] The construction method of the equivalent circuit model is as follows:

[0025] Step 1: Convert each element in the known Y parameter transfer function matrix into partial fraction form and determine the number of terms K.

[0026] Step 2: For any two different ports Pi, Pj∈{P1, P2, …, Pn} and Pi≠Pj, their transfer functions Y ij The kth term after (s) is converted into a partial fraction determines its kth submodule M ijk ; If the kth fraction is in the form of a / (sp), then consider a submodule in which an inductor and a resistor are connected in series; if the kth fraction is in the form of a / (sp), a≤0, then consider a submodule in which an inductor and a resistor are connected in series and then connected in parallel with a controlled current source, wherein the controlled current source is only used to correct the direction of the current without changing the poles and passivity of the module; if the kth fraction is in the form of (as+b) / (sp)(sp*), p and p* are conjugate complex numbers, then consider a submodule in which an inductor, a capacitor, and a resistor are connected in series and then connected in parallel with several controlled current sources, wherein the controlled current source is only used to correct the direction of the current without changing the poles and passivity of the module; if the kth term is in the form of a constant term, consider a single resistor submodule; if the kth term is in the form of as, consider a single capacitor submodule.

[0027] Step 3: For any port Pi∈{P1, P2, …, Pn}, its transfer function Y ii -∑ i≠j Y ij Determine the subcircuit C with P0 in the same way as in step 2 i0 .

[0028] like Figure 2 As shown, an integrated circuit metal-oxide-metal (MOM) capacitor is taken as an example to describe the present invention.

[0029] In the embodiment, the MOM capacitor has two ports P1 and P2 and a ground terminal P0. Step 1: convert each element in the known Y parameter transfer function matrix into a partial fraction:

[0030]

[0031]

[0032]

[0033]

[0034] The number of items K=3.

[0035] Step 2: Determine the subcircuit C between P1 and P2 12 .Y 12 The first term of is in the form of (as+b) / (sp)(sp*), p and p* are conjugate complex numbers, so the submodule M121 It is a submodule consisting of an inductor L1, a capacitor C1, a resistor R1 in series, and a controlled current source in parallel; Y 12 The second term of is a / (sp), a≥0, so the submodule M 122 It is a submodule consisting of inductor L2 and resistor R2 in series; 12 The third item of the form is as, so the submodule M 123 It is a submodule of a single capacitor C3.

[0036] Step 3, similar to step 2, is done by Y 11 -∑ 1≠j Y 1j and Y 22 -∑ 2≠j Y 2j Determine the subcircuit C between P1, P2 and P0 10 and C 20 .

[0037] It can be seen from this that the calculations between the sub-circuits and sub-modules of the embodiment are independent of each other and highly efficient. Figure 3 As shown, the solid line represents the equivalent circuit, and the triangle represents the electromagnetic simulation result of the original device. It can be seen that the complete graph equivalent circuit model has a good effect.

[0038] The above contents are detailed descriptions of the present invention in combination with specific implementation cases, and it cannot be determined that the specific implementation of the present invention is limited to these descriptions. For those skilled in the art of the present invention, without departing from the concept of the present invention, equivalent transformations or substitutions of the components, positional relationships and connection modes of the present invention without changing their functions also fall within the protection scope of the present invention.

Claims

1. A method for constructing a passive device complete graph equivalent circuit based on a Y parameter matrix, characterized in that: For a passive device with n ports P1, P2, ..., Pn and a ground terminal P0, there is a subcircuit C determined by the transfer function between any two different ports Pi and Pj. ij , the sub-circuit C ij It consists of several submodules M ijk Parallel construction; Pi, Pj∈{P0, P1, P2, …, Pn} and Pi≠Pj, thus forming a complete graph network; the specific construction method includes the following steps: Step 1: Convert each element in the known Y parameter transfer function matrix into partial fraction form and determine the number of terms K; Step 2: For any two different ports Pi, Pj∈{P1, P2, …, Pn} and Pi≠Pj, their transfer functions Y ij The kth term after (s) is converted into a partial fraction determines its kth submodule M ijk ; If the k-th fraction is in the form of a / (sp), a>0, then consider a submodule in which an inductor and a resistor are connected in series; If the k-th fraction is in the form of a / (sp), a≤0, then consider a submodule in which an inductor and a resistor are connected in series and then connected in parallel with a controlled current source, wherein the controlled current source is only used to correct the direction of the current without changing the poles and passivity of the module; If the k-th fraction is in the form of (as+b) / (sp)(sp*), p and p* are conjugate complex numbers, then consider a submodule in which an inductor, a capacitor, and a resistor are connected in series and then connected in parallel with several controlled current sources, wherein the controlled current source is only used to correct the direction of the current without changing the poles and passivity of the module; If the k-th term is in the form of a constant term, then consider a single resistor submodule; If the k-th term is in the form of as, then consider a single capacitor submodule; 1≤k≤K, K represents the subcircuit C ij The number of submodules it has; Step 3: For any port Pi∈{P1, P2, …, Pn}, its transfer function Y ii -∑ i≠j Y ij Determine the subcircuit C with P0 in the same way as in step 2 i0 .

Images