Microstrip
Physical microstrip constructors use substrate, conductor, and geometry inputs to produce frequency-dependent n-ports. Use lowercase names such as nP.mlin() for these models. For a lossless line with no physical substrate, use nP.Tlin().
Units and common defaults
Call each constructor with no arguments for its default geometry, or pass one options object. All lengths, widths, gaps, and thicknesses are in meters. A mil is 25.4e-6 m; an inch is 0.0254 m. Positional values are not supported.
The line, coupled line, tee, step, bend, and cross default to 25 mil substrate height, 0.0000125 in metal thickness, relative permittivity 10, copper resistivity 1.72e-8 Ω·m, loss tangent 0.001, and zero RMS surface roughness. Use resistivity for an absolute material value in Ω·m. These models use the selected circuit frequency list; the line, coupled line, tee, and film resistor also use the selected circuit temperature.
nP.global.fList = nP.global.fGen(1e9, 3e9, 21);
var line = nP.mlin({
width: 30 * 25.4e-6, // 30 mil
length: 0.0254, // 1 in
height: 25 * 25.4e-6,
relativePermittivity: 10
});
var response = line.out('s11dB', 's21dB');
nP.lineChart({
inputTable: [response],
title: 'Physical microstrip line',
metricPrefix: 'giga'
});Lines and junctions
| Constructor | Ports | Default geometry | Main options |
|---|---|---|---|
nP.mlin() | 2 | 23 mil width, 0.5 in length | width, length |
nP.mclin() | 4 | 19.1155 mil width, 5.82185 mil gap, 719.794 mil length | width, spacing, length |
nP.mtee() | 3 | Three 23 mil arms | commonWidth, branch1Width, branch2Width |
nP.mstep() | 2 | 46 mil input, 23 mil output | inputWidth, outputWidth |
nP.mbend() | 2 | 23 mil width, no miter | width, miterLength |
nP.mcross() | 4 | Four 23 mil arms | leftWidth, topWidth, rightWidth, bottomWidth |
All six accept the common substrate and material options above. mclin() uses the clockwise coupled-line order: port 1 upper left, port 2 upper right, port 3 lower right, port 4 lower left. mtee() uses port 1 as the common arm and ports 2 and 3 as branches, matching ideal Tee().
var split = nP.mtee({
commonWidth: 30 * 25.4e-6,
branch1Width: 23 * 25.4e-6,
branch2Width: 23 * 25.4e-6
});
var coupled = nP.mclin({spacing: 10 * 25.4e-6});Film resistor and vias
| Constructor | Ports | Default geometry | Main options |
|---|---|---|---|
nP.mtfr() | 2 | 50 Ω/square, 10 × 10 mil film | ohmsPerSquare, width, length, temperatureCoefficient |
nP.mvgnd() | 1 | 100 µm diameter, 25 mil height | diameter, height, resistivity |
nP.mvia() | 2 | 100 µm diameter, 25 mil connection height, no pads or stubs | diameter, connectionHeight, padDiameter, antipadDiameter |
mtfr() also uses a 25 mil substrate, relative permittivity 10, and loss tangent 0.001 by default. Its temperature coefficient defaults to zero. The vias default to copper resistivity; mvia() also defaults to relative permittivity 10.
var film = nP.mtfr({ohmsPerSquare: 100});
var groundedVia = nP.mvgnd({diameter: 200e-6});
var transition = nP.mvia({diameter: 200e-6});Inspect a physical model
These constructors return ordinary n-ports with .out(...). They also expose .physicalModel metadata describing the model family, geometry, material, and analysis settings:
var line = nP.mlin();
var geometry = line.physicalModel.geometry;
var material = line.physicalModel.material;
nP.log(geometry);
nP.log(material);Use named properties and the same dimensions for components you intend to connect. An optional fList property overrides the selected frequency list for one component, but all n-ports in the same analysis must have matching frequency rows. For the full option lists and model assumptions, see RF Constructors.