Ansoft HFSS 3D Boundary Manager Sources

Transcription

1 Lumped Gap Defining s Voltage and Current When you select Source, you may choose from the following source types: Incident wave Voltage drop Current Magnetic bias These sources are available only for driven solutions. Maxwell Online Help System 265 Copyright Ansoft Corporation

2 Lumped Gap Defining s Voltage and Current By default, the interface between all 3D objects and the background is a perfect E boundary through which no energy may enter or exit. Wave ports are typically placed on this interface to provide a window that couples the model device to the external world. Ansoft HFSS assumes that each port you define is connected to a semi-infinitely long waveguide that has the same cross-section and material properties as the port. In solving for the S-parameters, the software assumes that the structure is excited by the natural field patterns (modes) associated with these cross-sections. The 2D field solutions generated for each port serve as boundary conditions at those ports for the 3D problem. The final field solution computed must match the 2D field pattern at each port. Ansoft HFSS generates a solution by exciting each port individually. Each mode incident on a port contains one watt of time-averaged power. 1 is excited by a signal of one watt, and the other ports are set to zero watts. After a solution is generated, port 2 is set to one watt, and the other ports to zero watts and so forth. Within the 3D model, an internal port can be represented by a lumped gap source. Lumped gap sources compute S-parameters directly at the port. The S-parameters can be renormalized and the Y-matrix and Z-matrix can be computed. Lumped gap sources have a user-defined characteristic impedance. Maxwell Online Help System 266 Copyright Ansoft Corporation

3 Defining Terminals Lumped Gap Defining s Voltage and Current A wave port is a traditional transmission line port. In general, use the following procedure to define a wave port. > To create a wave port: 1. For the selected surface or 2D object, select as the source type. 2. Select Wave from the Type pull-down menu. 3. Enter a name for the port or accept the default name. Do not use an embedded blank in the port name. 4. By default, the system assumes that only the dominant mode associated with the port s cross-section is present at a port. To specify more than one mode to analyze at the port: a. Select View Modes from the View Modes/View Terminals pull-down menu. b. Enter a value in the Num field. c. Choose Enter. Each mode is listed in the Modes list. For terminal solutions, specify one mode per conductor. The maximum number of modes on any port is To set an impedance or calibration line on a port mode: a. Select a port mode from the Modes list. b. Select Use Impedance Line or Use Calibration Line. When you select one, a Y appears under Imped or Calib in the Modes list. c. Define the location of the impedance or calibration line. 6. Optionally, define terminals for the wave port if you wish to compute voltage data for the port that can be used for circuit analysis. 7. Select Polarize E Field to polarize the E-field on the port. 8. Enter the impedance multiplier for the model in the Imped Multiplier field. 9. Choose Assign. Maxwell Online Help System 267 Copyright Ansoft Corporation

4 Defining Terminals Lumped Gap Defining s Voltage and Current Defining Terminals For projects containing multi-conductor transmission line ports, a terminal-based description in terms of voltages and currents may be more useful than the traditional mode-based S-matrix. To facilitate such a description, terminal lines must be created on the ports to define port voltages. Use the following guidelines to set up terminal voltage lines: Solve for all present TEM modes. One terminal voltage line must be created for each port mode in the device. In general, draw a single terminal line from the reference or ground conductor to each port-plane conductor. Be consistent with the setup of terminal voltage lines. For example, the setup on port 1 should usually be the same as that on port 2. Voltage loops are not permitted because voltages are not independent. > To define terminals: 1. Select the Define Terminals check box. Select this check box one time regardless of the number of ports. 2. Select View Terminals from the View Modes/View Terminals pull-down menu. New fields appear in the window. Each terminal is listed in the Terminals list. The number of terminals on a wave port must be equal to the number of modes set for the port. The number of terminals on a lumped gap source should be one. Note that if terminals are defined for one port, they are defined for all ports in the model. 3. Select a terminal from the Terminals list. 4. Enter a name for the terminal or accept the default name in the Terminal Name field. The names of the terminals will affect the order in which they appear in the list. 5. Set a terminal voltage line on the terminal. Each terminal must have a voltage line defined. To set a terminal voltage line: a. Select a port terminal from the Terminals list. b. Define the location of the terminal voltage line. A Y appears under Defined in the Terminals list. 6. Choose Assign. Terminals are now specified for the port. Maxwell Online Help System 268 Copyright Ansoft Corporation

5 Lumped Gap Defining Terminal Voltage, Impedance, and Calibration Lines Terminal Voltage Lines Impedance Lines Calibration Lines Need for Polarization Impedance Multipliers Defining s Voltage and Current Lumped Gap Lumped gap sources are similar to traditional wave ports, but can be located internally and have a complex user-defined impedance. Lumped gap sources compute S-parameters directly at the port. The S-parameters can be renormalized and the Y-matrix and Z- matrix can be computed. A lumped gap source can be defined as a rectangle from the edge of the trace to the ground (a lumped gap source) or as a traditional port (a wave gap source). The default boundary is perfect H on all edges that do not come in contact with the metal. Use lumped gap sources for microstrip structures and use wave gap sources for strip lines and other waveguide structures. For lumped gap sources, the following restrictions apply: The complex impedance must be non-zero and the resistance must be non-negative. Only one port mode is allowed, or one terminal if terminals are defined. An impedance line and a calibration line must be defined. When a gap source is used as an internal port, the conducting cap required for a traditional port must be removed to prevent short-circuiting the source. More Maxwell Online Help System 269 Copyright Ansoft Corporation

6 Lumped Gap Defining Terminal Voltage, Impedance, and Calibration Lines Terminal Voltage Lines Impedance Lines Calibration Lines Need for Polarization Impedance Multipliers Defining s Voltage and Current In general, use the following procedure to define a lumped gap source. > To create a lumped gap source: 1. For the selected surface or 2D object, select as the source type. 2. Select Lumped Gap Source from the Type pull-down menu. 3. Enter a name for the port or accept the default name. Do not use an embedded blank in the port name. 4. Do the following to define the complex impedance of the gap source: a. Enter the resistance or real part of the impedance in the Resistance field. b. Enter the reactance or imaginary part of the impedance in the Reactance field. The user-defined complex impedance Z s defined for a lumped gap source serves as the reference impedance of the S-matrix on the lumped gap source. The impedance Z s has the characteristics of a wave impedance; it is used to determine the strength of a source, such as the modal voltage V and modal current I, through complex power normalization. (The magnitude of the complex power is normalized to 1.) In either case, you would get an identical S-matrix by solving a problem using a complex impedance for a lumped gap source or renormalizing an existing solution to the same complex impedance. When the reference impedance is a complex value, the magnitude of the S-matrix is not always less than or equal to 1, even for a passive device. 5. To set an impedance or calibration line on the port mode: a. Select the port mode from the Modes list. b. Select Use Impedance Line or Use Calibration Line. When you select one, a Y under Imped or Calib appears in the Modes list. c. Define the location of the impedance or calibration line. 6. Enter the impedance multiplier for the model in the Imped Multiplier field. 7. Choose Assign. The port is assigned to the selected surface and appears in the boundaries list. If terminals are defined, you will also need to define a terminal voltage line for each single terminal. Maxwell Online Help System 270 Copyright Ansoft Corporation

7 Lumped Gap Defining Terminal Voltage, Impedance, and Calibration Lines Terminal Voltage Lines Impedance Lines Calibration Lines Need for Polarization Impedance Multipliers Defining s Voltage and Current More Defining Terminal Voltage, Impedance, and Calibration Lines > To define a terminal voltage, impedance, or calibration line: 1. Choose Set from the Edit Line menu for the selected line. 2. Define the line in one of the following two ways: Specify its endpoints: a. Enter the coordinates for the start point in the X, Y, and Z fields or select the new point using the mouse. Choose Enter to accept these coordinates for the start point or Cancel to cancel the operation. When you choose Enter, fields for entering a vector appear below the X, Y, and Z fields. b. Specify the end point in the same manner as the start point. When you specify the end point, a line is drawn from the start point to the end point. This allows you to view the line before accepting it. If necessary, choose Reset Start to set the start point to the coordinates currently entered in the X, Y, and Z fields. Specify a point and a vector: a. Specify the point in the same manner you specify the start point above. b. Specify the vector using the mouse or entering the vector coordinates in the X, Y, and Z fields under Enter vector. The vector specifies the direction from the point you specified earlier. When you enter values for the x, y, and z coordinates, the Vector length field changes to reflect the length of the vector (and the line). You may change all the coordinate fields with nonzero values by specifying a new vector length. If necessary, choose Reset Start to reset the start point. Vector coordinates are entered relative to the start point. Thus, if your start point has coordinates of (0,10,0) and you enter (0,20,0) as the vector coordinates, the absolute coordinates (displayed in the upper left corner) would be (0,30,0). 3. Choose Enter to accept the line or Cancel to cancel the operation. Terminal voltage, impedance, and calibration lines appear in the view window. An arrow indicates the direction of the line and a letter (T for terminal voltage, I for impedance, or C for calibration) specifies the line type. To reverse the direction in which the line points, choose Edit Line/Swap Points to swap the coordinates for the start point and end point. The line s direction will be reversed. Maxwell Online Help System 271 Copyright Ansoft Corporation

8 Lumped Gap Defining Terminal Voltage, Impedance, and Calibration Lines Terminal Voltage Lines Impedance Lines Calibration Lines Need for Polarization Impedance Multipliers Defining s Voltage and Current To copy a previously defined impedance or calibration line s points, choose Edit Line/ Copy Calibration (if you ve defined an impedance line) or Edit Line/Copy Impedance (if you ve defined a calibration line). Terminal Voltage Lines Terminal voltage lines are used to define voltages on port boundaries. In general, draw a single terminal voltage line from the reference or ground conductor to each port-plane conductor. Each terminal voltage line is currently restricted to a single line segment. In certain geometries, this restriction may force you to draw the terminal voltage line through a second conductor. This is permissible; however, you cannot draw more than one terminal voltage line connecting a given reference conductor and port-plane conductor, nor draw a terminal voltage line with its entire length along a perfect conductor. In circuit analysis, the polarity reference for a voltage is designated with + and - symbols. The voltage polarity reference on a terminal voltage line is established by an arrow where the head of the arrow is synonymous with + and the base of the arrow is synonymous with -. Maxwell Online Help System 272 Copyright Ansoft Corporation

9 Lumped Gap Defining Terminal Voltage, Impedance, and Calibration Lines Terminal Voltage Lines Impedance Lines Computing Characteristic Impedance Defining the Impedance Line Impedance Lines and Modes Calibration Lines Need for Polarization Impedance Multipliers Defining s Voltage and Current Impedance Lines The S-matrices initially calculated by the system are generalized S-matrices that have been normalized to the impedances of each port. However, it is often desirable to compute S-matrices that are normalized to specific impedances such as 50 ohms. To convert a generalized modal S-matrix to a renormalized modal S-matrix, the system first needs to compute the characteristic impedance at each port. There are several ways to compute characteristic impedance. Two methods the Zpv and Zvi methods require an impedance line. Computing Characteristic Impedance Ansoft HFSS will always calculate Zpi impedance, the impedance calculation using power and current, which are well-defined for a port because they are computed over the area of the port. Zpv and Zvi are not calculated by default. This is because V is computed by integrating along a user-defined impedance line. To renormalize the solution to a Zpv or Zvi characteristic impedance, you must have defined an impedance line. Defining the Impedance Line In general, select two points at which the voltage differential is expected to be at a maximum. For example, on a microstrip port, place one point in the center of the microstrip, and the other directly underneath it on the ground plane. In a rectangular waveguide, place the two points in the center of the long sides. Impedance Lines and Modes If you are analyzing more than one mode at a port, define a separate set of impedance lines for each mode. The orientation of the electric field differs from mode to mode. Maxwell Online Help System 273 Copyright Ansoft Corporation

10 Lumped Gap Defining Terminal Voltage, Impedance, and Calibration Lines Terminal Voltage Lines Impedance Lines Calibration Lines Need for Polarization Impedance Multipliers Defining s Voltage and Current Calibration Lines When Ansoft HFSS computes the excitation field pattern at a port, the direction of the field at ωt=0 is arbitrary: The field can always point in one of at least two directions. In the figure shown below, the mode 1 field at ωt=0 can either point to the left or to the right. Either direction is correct unless a preferred direction is specified. To specify a direction, you must calibrate the port relative to some reference orientation by defining a calibration line. In the case of rectangular waveguides, visualize the difference in terms of a physical connection. If the up side of a port is aligned with the up side of the waveguide carrying the excitation signal, the signal at the port is in phase with what is expected. But if the up side of the port is connected to the down side of the waveguide, the incoming signal will be out of phase with the expected signal. Likewise, it is desirable to define which way is up at all ports on a structure; otherwise, the resulting S-parameters can be shifted from the expected orientation. Calibrate a port to define a preferred direction at each port relative to other ports having identical or similar cross-sections. In this way, the results of laboratory measurements (in which the setup is calibrated by removing the structure and connecting two ports together) can be duplicated. Because calibration lines determine the phase of the excitation signal and traveling wave, they are ignored by the system when a ports-only solution is requested. Maxwell Online Help System 274 Copyright Ansoft Corporation

11 Lumped Gap Defining Terminal Voltage, Impedance, and Calibration Lines Terminal Voltage Lines Impedance Lines Calibration Lines Need for Polarization Impedance Multipliers Defining s Voltage and Current Need for Polarization In some cases, such as when a port is square or circular, not only is the positive and negative direction in question the line with which the E-field is aligned is also arbitrary. For example, in the case of a square waveguide, the E-field of the dominant mode can be aligned with either the horizontal or vertical direction. There is no preferred direction. However, the system aligns the field with the calibration line if you select Polarize E Field. Circular waveguides also require a polarized E-field. The direction of the E-field at ωt=0 can point in any direction. To align the simulated field with a preferred direction, define a calibration line and select Polarize E Field. In this case, the calibration line must lie in the middle of the port that is, in the symmetry plane. Warning: When polarizing the E-fields, observe the following guidelines. Otherwise, the results may not be as expected. Polarize the E-field only on square or circular waveguides. Make sure the port on the waveguide only feeds a single conductor (the waveguide wall). Do not polarize the E-fields if you are using a symmetry boundary. The polarization is automatically enforced by the symmetry boundary condition. Maxwell Online Help System 275 Copyright Ansoft Corporation

12 Lumped Gap Defining Terminal Voltage, Impedance, and Calibration Lines Terminal Voltage Lines Impedance Lines Calibration Lines Need for Polarization Impedance Multipliers Defining s Voltage and Current Impedance Multipliers If you have defined a symmetry plane, the computed impedances will not be for the full structure. Generally, use one of the following values for the impedance multiplier: If the structure has a perfect E plane of symmetry, use 2. Such models have one-half of the voltage differential and one-half of the power flow of the full structure, resulting in impedances that are one-half of those for the full structure. If the structure has a perfect H plane of symmetry, enter 0.5. Such models have the same voltage differential but half the power flow of the full structure, resulting in impedances that are twice those for the full structure. If the structure has a combination of perfect H and perfect E boundaries, adjust accordingly. For example, you do not have to enter an impedance multiplier for a structure with both a perfect E and perfect H boundary since you would be multiplying by 2 and 0.5. Maxwell Online Help System 276 Copyright Ansoft Corporation

13 Lumped Gap Defining s Locations s are Planar Ferrite Materials and s Anisotropic Materials and s Multiple Modes s on Microstrips Length of Uniform Cross- Section Voltage and Current Defining s This section highlights things to keep in mind when defining ports. Locations Only surfaces that are exposed to non-existent objects (such as the background or objects defined as perfect conductors) can be defined as wave ports. Do not define a surface that cuts through an object to be a port unless one of the objects is assigned the material characteristics of a metal. Do not define the interface of two internal objects as a port unless one of the objects is assigned the material characteristics of a metal. Surfaces defined as lumped gap sources cannot be exposed to perfect conductors. s are Planar A port must lie in a single plane. s that bend are not allowed. For example, if a geometric model has a curved surface exposed to the background, that curved surface cannot be defined as a port. Ferrite Materials and s When designing a problem containing ferrite materials and ports, do not arrange the port so that it touches the ferrite material. If you must place a port on a ferrite material, separate the two with a dielectric with a relative permittivity equal to the relative permittivity of the ferrite. If your problem contains a port that touches a ferrite material, the following error message appears: Can not solve portname with ferrite material materialname on the port. Maxwell Online Help System 277 Copyright Ansoft Corporation

14 Lumped Gap Defining s Locations s are Planar Ferrite Materials and s Anisotropic Materials and s Multiple Modes s on Microstrips Length of Uniform Cross- Section Voltage and Current Anisotropic Materials and s An anisotropic material can be in contact with a port provided that there is no loss on the port, i.e., a lossy material or boundary condition (finite conductivity or impedance) cannot be in contact with the port. Although a radiation boundary is lossy, it can be in contact with a port in this case because it is generally not modeled as lossy where it touches the port. Note that a radiation boundary can be modeled as lossy if the environment variable ZERO_ORDER_ABC_ON_PORT is set. one principal axis of the anisotropic material is aligned normal to the port. Multiple Modes The number of modes should be set high enough to include all propagating modes in the port cross-section over the frequency range of interest. Non-propagating modes may be excluded only if the port plane is far enough away from the 3D model so that such modes have negligible effects. See the proceeding section, Length of Uniform Cross-Section, and the Technical Notes section on mode propagation for more information. Non-propagating modes are those that have an attenuation constant, α, that is greater than their phase constant, β. One way to determine which modes need to be modeled is to set up the problem with multiple modes and generate a solution with no adaptive passes (a ports-only solution). Then, inspect the complex propagation constant, γ = α + jβ, associated with each mode. Each additional mode at a port results in an additional set of S-parameters. For example, if you are analyzing two modes at each port in a three-port structure, the final result is a 6x6 S-matrix. Maxwell Online Help System 278 Copyright Ansoft Corporation

15 Lumped Gap Defining s Locations s are Planar Ferrite Materials and s Anisotropic Materials and s Multiple Modes s on Microstrips Length of Uniform Cross- Section Voltage and Current s on Microstrips When assigning ports to geometric models that represent microstrips, define the ports to include the dielectric below the strip and the air above the strip. Ansoft HFSS then assumes that the port is being excited by a wave traveling down the microstrip inside a package with conductive walls. In the following figure, the shaded faces in the model have been defined as ports. Each port s cross-section extends through the air surrounding the microstrip to the conductive shield surrounding the entire problem. 1 2 More Maxwell Online Help System 279 Copyright Ansoft Corporation

16 Lumped Gap Defining s Locations s are Planar Ferrite Materials and s Anisotropic Materials and s Multiple Modes s on Microstrips Length of Uniform Cross-Section Voltage and Current Length of Uniform Cross-Section The geometry must include a length of uniform cross-section at each port. For example, the waveguide on the left below is not modeled correctly because it does not contain a length of uniform cross-section at either port. The waveguide on the right includes a length of uniform cross-section at each port; it is modeled correctly. No cross-section at ports Uniform cross-section at ports The length of the uniform cross-section should be long enough to allow non-propagating modes to die out. Otherwise, the boundary conditions at the port will prevent the simulated solution from matching the actual solution. For example, if a non-propagating mode takes approximately one-eighth of the dominant mode s wavelength to die out either because of losses or because it is an evanescent mode then you should make the uniform cross-section longer than one-eighth of a wavelength. Otherwise, you must include the effects of that higher order mode in the simulation. e αz Reflected waves attenuate as a function of, assuming that the wave propagates in the z direction. Therefore, the required length of the uniform cross-section depends on the value of the mode s attenuation constant, α. More You must make the port long enough for non-propagating modes to die out because the system forces the field pattern at each port to be a linear combination of the modes you request. For example, if discontinuities in a structure cause higher order modes to be reflected back toward a port face, then the actual field solution near the discontinuity is a linear combination of all relevant modes. If the port length, the length of uniform crosssection leading to the port face, is not long enough for the reflected waves to die out, then the energy in those modes will affect the apparent energy in the dominant mode, resulting in erroneous results. In cases where port lengths are too short for reflected modes to decay, a field solution involving only the dominant mode will not be what you expected Maxwell Online Help System 280 Copyright Ansoft Corporation

17 Lumped Gap Defining s Locations s are Planar Ferrite Materials and s Anisotropic Materials and s Multiple Modes s on Microstrips Length of Uniform Cross-Section Voltage and Current that is, it will not be for a structure being excited with only the most dominant mode. An incident wave (plane wave) is a wave that propagates in one direction and is uniform in those directions perpendicular to its direction of propagation. The angle at which the incident wave impacts the device is known as the angle of incidence. The equation that the solver uses to calculate the incident wave is E inc = E 0 e jk 0 ( kˆ r) where E inc is the incident wave. E 0 is the E-field polarization vector. k 0 is the free space wave number. It is equal to ω µ 0 ε 0. kˆ is the propagation vector. It is a unit vector. r is the position vector and is equal to xxˆ + yŷ + zẑ. When defining the incident wave, you may enter the propagation vector k and the E-field polarization vector E 0 in either cartesian or spherical coordinates. If you are specifying the incident wave in cartesian coordinates, only one incident wave may be defined. If you are using spherical coordinates, you must specify the number of incident waves present. Plane-wave sources are specified in a peak sense. That is, if the plane wave magnitude is 5 V/m, then the plane wave incident field magnitude is Et () = 5cos( k r + ωt). > To set up an incident wave: 1. For the selected surface, select Incident wave as the source type. 2. Do one of the following to define the propagation and E-field polarization vectors: To define the vectors using Cartesian coordinates: a. Select Cartesian (the default) as your coordinate system. X, Y, and Z fields appear for the k and E 0 vectors. b. Enter the x-, y-, and z-components for k vector in the X, Y, and Z fields. The k vector must be a unit vector. For example, to define a plane wave travelling in the positive z direction, enter (0,0,1) as the coordinates for k. Maxwell Online Help System 281 Copyright Ansoft Corporation

18 Lumped Gap Defining s Voltage and Current c. Enter the coordinates for E 0 vector in the X, Y, and Z fields. The magnitude of the E 0 vector cannot be zero. When entering k and E 0 in cartesian coordinates, the propagation vector k must be orthogonal to E 0. However, when k and E 0 are entered in spherical coordinates they are automatically specified as orthogonal. To define the vectors using Spherical coordinates: a. Select Spherical as the coordinate system. The Phi, Theta, and Eo vector fields appear. b. Under Phi, enter the following: Start Stop Points The point where the rotation of φ begins. The point where the rotation of φ ends. The number of points on the sweep of φ. Keep in mind that the number of points is not the same as the number of increments. For example, to divide a sweep from 0 to 180 into 10 increments, you would enter 19 points for 18 increments. Each point is equidistant from the next point with the first point being the Start point and the last point being the Stop point. At each point selected in the φ direction, the system will sweep through the range of θ points. Use the View button to view the values of φ. c. Under Theta, enter values for Start, Stop, and Points. Because θ is swept through each φ point, a spherical grid is created. At each grid point an incident wave is present traveling towards the origin. The number of incident waves and grid points can be calculated by multiplying the number of φ points by the θ points. d. Enter the φ and θ components of E 0 in the Phi and Theta fields. 3. Choose Refresh Arrow(s) to redraw the arrows representing the vectors. 4. Choose Assign. The incident wave is defined and appears in the boundaries list. Maxwell Online Help System 282 Copyright Ansoft Corporation

19 Lumped Gap Defining s Voltage and Current Voltage Drop Current Creating a Voltage or Current Source More Voltage and Current Voltage and current sources (or gap sources) allow you to define the electric and magnetic field strength on a boundary by specifying the electric potential or current flow on that surface. Circuit gap sources are specified in a peak sense. That is, if a voltage gap source magnitude is 5 volts, then the time domain circuit source behaves as v(t) = 5cosωt. This is also true for a current gap source. Voltage Drop The voltage source boundary condition lets you specify the voltage and direction of the E- field on a surface. This type of boundary is used when the feed structure is very small compared to the wavelength and a constant electric field may be assumed across the feed points. In this case, Ansoft HFSS assigns a constant electric field across the gap on which you specified the voltage. Current Because voltage and current sources allow you to construct a problem without ports (and thereby without S-parameters), you can generate a field solution without calculating S-parameters. As a result, commands dealing with ports and S-parameters will be removed, disabled, or replaced with new commands. The current source boundary condition lets you define the magnitude and direction of the current flow through a surface. This type of boundary is used when the feed structure is very small compared to the wavelength and the electric current on the surface is assumed to be constant across the feed points. Maxwell Online Help System 283 Copyright Ansoft Corporation

20 Lumped Gap Defining s Voltage and Current Voltage Drop Current Creating a Voltage or Current Source More Creating a Voltage or Current Source > To set up a voltage or current source: 1. For the selected surface, select Voltage drop or Current as the source type. 2. Enter the value of the source in volts or amps in the Value field. 3. Choose Set Vector to select the direction of the source. Define the vector in one of the following two ways: To define the vector by specifying its end points: a. Enter the coordinates for the start point in the X, Y, and Z fields or select the new point using the mouse. Choose Enter to accept these coordinates for the start point or Cancel to cancel the operation. When you choose Enter, fields for entering a vector appear below the X, Y, and Z fields. b. Specify the end point in the same manner as the start point. When you specify the end point, a line is drawn from the start point to the end point. This allows you to view the vector before accepting it. If necessary, choose Reset Start to set the start point to the coordinates currently entered in the X, Y, and Z fields. To define the vector by specifying a point and a direction: a. Specify the point in the same manner you specify the start point above. b. Specify the direction using the mouse or by entering the vector coordinates in the X, Y, and Z fields under Enter vector. The vector specifies the direction from the point you specified earlier. When you enter values for the x-, y-, and z-coordinates, the Vector length field changes to reflect the length of the vector (and the line). You may change all the coordinate fields with nonzero values by specifying a new vector length. If necessary, choose Reset Start to reset the start point. When you enter coordinate values under Enter vector, remember that the vector coordinates are relative to the start point. Thus, if your start point has coordinates of (0,10,0) and you enter (0,20,0) as the vector coordinates, the absolute coordinates (displayed in the upper left) would be (0,30,0). 4. Choose Enter to accept the vector or Cancel to cancel the vector. When you have finished, a line appears on the model. An arrow indicates the direction and a letter (v or i) indicates the type of source. 5. Choose Assign. The source is assigned to the selected surface and appears in the boundaries list. Maxwell Online Help System 284 Copyright Ansoft Corporation

21 Lumped Gap Defining s Voltage and Current Uniform Internal Bias Non-Uniform More When you create a ferrite material, you must also define the net internal field that biases the ferrite. The bias field aligns the magnetic dipoles in the ferrite producing a nonzero magnetic moment. When the applied bias field is assumed to be uniform, the tensor coordinate system is user specified through a rotation from the global coordinate system. When the applied bias field is non-uniform, the user-specified coordinate system rotations are not allowed. The permeability tensor s local coordinate system is calculated on a tetrahedron by tetrahedron basis, with the direction determined by the field directions calculated in the static solution. Uniform The applied DC bias that causes ferrite saturation is always in the positive z direction of the tensor coordinate system. Initially the tensor coordinate system is assumed to be aligned with the fixed coordinate system the tensor s z-axis is the same as the model s z-axis. To model other directions of applied bias, the permeability tensor must be rotated so that its z-axis lies in another direction on the fixed coordinate system. This is accomplished by specifying the rotation angles about the axes. > To specify a uniform applied bias field: 1. For the selected object, select Magnetic bias as the source type. 2. Select Uniform as the applied bias field type. 3. Enter the internal bias of the ferrite in the Internal Bias field. 4. Enter the rotation of the permeability tensor with respect to the xyz-coordinate system in the X, Y, and Z fields. 5. Choose Assign. The angles should be defined in such a way that the tensor coordinate system is obtained in the following manner: 1. Rotating the tensor coordinate system by α degrees (from the X Angle field) around the fixed x-axis. 2. Rotating the resulting tensor coordinate system by β degrees (from the Y Angle field) around the new y-axis. 3. Rotating the new tensor coordinate system by γ degrees (from the Z Angle field) around the new z-axis. Maxwell Online Help System 285 Copyright Ansoft Corporation

22 Lumped Gap Defining s Voltage and Current Uniform Internal Bias Non-Uniform More This concept is illustrated in the following graphic. In the first panel, the permeability tensor is rotated α degrees about the x-axis. In the second panel, the tensor is rotated β degrees about the y'-axis (the new y-axis). In the third panel, the tensor is rotated γ degrees about the z''-axis (the new z-axis). The resulting tensor has the coordinate system (x''y''z'') relative to the fixed coordinate system. z x y z y x z y x α x y As an example, to model the DC bias in the x direction you would rotate the tensor coordinate system such that its z-axis lay along the x-axis of the fixed coordinate system. To do this you would enter 0 for the X Rotation, 90 for the Y Rotation, and 0 for the Z Rotation. Internal Bias This is the DC magnetic field within the ferrite that results from the applied bias field. The applied field aligns the magnetic dipoles in the material producing a nonzero magnetic moment commonly known as the demagnetization field. The demagnetization field opposes the applied bias and results in an internal field, which can be much smaller than the applied field. The internal bias field is the net field within the ferrite after accounting for the demagnetization field. The internal bias field is assumed to be uniform in magnitude and direction. The units of magnetic bias field are selectable, but are amperes/meters in the MKS system. Non-Uniform y β To accurately model a ferrite in an applied static magnetic bias field, the non-uniform magnetic bias fields must also be calculated. In Ansoft HFSS, a ferrite s permeability tensor is a direct result of an applied static magnetic bias field. The static field causes the tensor to assume an hermitian form, with cross coupling terms between field components perpendicular to the bias. However, a uniform bias field is difficult to achieve in practice. Even if the bias field is nearly uniform, a non-ellipsoidal shaped ferrite material will have non-uniform demagnetization with resulting non-uniform fields in the ferrite. z z γ x Maxwell Online Help System 286 Copyright Ansoft Corporation

23 Lumped Gap Defining s Voltage and Current Uniform Non-Uniform Use the magnetostatic solver provided in Maxwell 3D to generate a solution for non-uniform magnetostatic fields. Once a solution is generated it may be imported into Ansoft HFSS. > To specify a non-uniform applied bias field: 1. For the selected object, select Magnetic bias as the source type. 2. Select Non-uniform as the applied bias field type. 3. Enter the name of the Maxwell 3D Field Simulator project in the Magnetostatic project name field. You do not need to enter the.pjt extension. 4. Choose Assign. > To generate a solution for non-uniform static magnetic bias fields within ferrites: 1. Start the Maxwell Control Panel and create a new project for the Maxwell 3D Field Simulator with the Projects command. 2. Create and open a new Ansoft HFSS project. 3. Draw and save the geometry for the Ansoft HFSS problem. 4. Open the Maxwell 3D Field Simulator project and import the Ansoft HFSS model the.sld file in the Ansoft HFSS project s directory. You only need to import the objects that correspond to the magnetic materials. Warning: To specify the non-uniform bias field you must have purchased the Maxwell 3D Field Simulator. Refer to the Maxwell 3D Field Simulator documentation for instructions on solving for non-uniform magnetostatic fields. You must import the geometry. When Ansoft HFSS imports the solution information it does so on a tetrahedra by tetrahedra basis. The position of every object must correspond exactly; if it does not, the solution will not be accurate. You may delete unnecessary objects or add virtual objects, but do not change the position of any existing objects. 5. Add any magnetic bias circuitry to the Maxwell 3D Field Simulator model. 6. Set up the materials, boundary conditions, and the solution parameters for the Maxwell 3D Field Simulator problem and generate a solution. 7. Exit the Maxwell 3D Field Simulator project and return to the Ansoft HFSS project. 8. Set up and solve the Ansoft HFSS problem. Ansoft HFSS uses the Maxwell 3D Field Simulator project you specified in the Magnetostatic project name field as the source of the non-uniform magnetostatic field information. Maxwell Online Help System 287 Copyright Ansoft Corporation