Skip to content

matlab/matlab-agentic-toolkit / skills-catalog/rf-and-mixed-signal/matlab-model-rf

matlab-model-rf

Verified

RF Toolbox and RF Blockset in MATLAB -- S-parameter I/O, network conversions (S/Z/Y/ABCD/T/H/G, mixed-mode), cascade/de-embedding, rfbudget analysis, circuit composition, matching networks, amplifier stability, mixer spurs, rational fitting, SI channels, baseband processing, Circuit Envelope simulation. Trigger: sparameters, Touchstone, .s2p, .s4p, rfplot, smithplot, rfparam, rfwrite, zparameters, yparameters, abcdparameters, s2sdd, cascadesparams, deembedsparams, rfbudget, noise figure, OIP3, IIP3, amplifier, modulator, nport, rffilter, attenuator, seriesRLC, shuntRLC, lcladder, txline, circuit, setports, clone, matchingnetwork, stabilityk, stabilitymu, powergain, gammams, gammaml, mixerIMT, OpenIF, rational, rationalfit, stepresp, txlineWRLGC, rf.Amplifier, rf.Mixer, rf.Filter, rf.Sparameter, rfsystem, RF Blockset.

reference/matching-networks.md

View on GitHub
# Impedance Matching Network Design

Design, evaluate, and optimize **1-port** impedance matching networks. The `matchingnetwork` object and `matchingNetworkDesigner` app match a source impedance to a load impedance — they do not yet design 2-port (e.g., filter-style) matching. The object generates candidate topologies, ranks them by performance, and exports as circuit objects.

## Creating a Matching Network

```matlab
mn = matchingnetwork( ...
    'SourceImpedance', 50, ...
    'LoadImpedance', 75 + 20i, ...
    'CenterFrequency', 2.4e9, ...
    'Bandwidth', 200e6, ...
    'Components', 2);
```

### Properties

| Property | Default | Description |
|----------|---------|-------------|
| `SourceImpedance` | 50 | Source impedance (see formats below) |
| `LoadImpedance` | 50 | Load impedance (see formats below) |
| `CenterFrequency` | 1e9 | Design frequency (Hz) |
| `Bandwidth` | 100e6 | Target bandwidth (Hz) |
| `LoadedQ` | 10 | Loaded Q factor |
| `Components` | 2 | Topology: `2`, `3`, `'L'`, `'Pi'`, `'Tee'` |
| `Circuit` | (auto) | Read-only array of generated circuits |

### Bandwidth and LoadedQ Are Coupled

`LoadedQ = CenterFrequency / Bandwidth`. Setting one updates the other. If both are specified, `Bandwidth` takes precedence.

### Components / Topology Options

| Value | Circuits Generated |
|-------|-------------------|
| `2` or `'L'` | 2 L-section designs (series-L/shunt-C and series-C/shunt-L) |
| `3` | 8 three-element designs (all permutations) |
| `'Pi'` | 4 shunt-series-shunt designs |
| `'Tee'` | 4 series-shunt-series designs |

**Gotcha:** Maximum is 3 components. `Components = 1` or `Components = 4` error.

## Impedance Source Formats

Both `SourceImpedance` and `LoadImpedance` accept:

```matlab
% Scalar (constant impedance)
mn.LoadImpedance = 75 + 20i;

% S-parameter object (frequency-dependent, 1-port)
sAnt = sparameters('antenna.s1p');
mn.LoadImpedance = sAnt;

% Z-parameter or Y-parameter object
mn.LoadImpedance = zparameters(sAnt);

% Touchstone file path
mn.LoadImpedance = 'antenna.s1p';

% 1-port circuit object (must have setports called)
ckt = circuit('Load');
add(ckt, [1 0], resistor(75, 'RL'));
add(ckt, [1 0], capacitor(1e-12, 'CL'));
setports(ckt, [1 0]);
mn.LoadImpedance = ckt;

% Function handle (frequency in Hz)
mn.LoadImpedance = @(f) 50 + 30i*(f/1e9);
```

## Inspecting Candidate Designs

```matlab
[topology, performance] = circuitDescriptions(mn);
disp(topology);       % Component types and values for each design
disp(performance);    % Pass/fail, tests failed, performance score
```

The **topology table** columns: `circuitName`, `component1Type`, `component1Value`, `component2Type`, `component2Value`, etc. Component types are `"Series L"`, `"Series C"`, `"Shunt L"`, `"Shunt C"`.

The **performance table** columns: `circuitName`, `evaluationPassed` (`"Yes"`/`"No"`), `testsFailed`, `performanceScore`. Designs are ranked by score (descending). The `performanceScore` is a weighted composite — higher is better, but the scale depends on your evaluation parameters (it is not a fixed 0–100 range). Use it for relative ranking between candidates, not as an absolute quality metric.

Use `component1Type`/`component2Type` columns in the topology table to identify which topology each auto-generated design represents (e.g., `"Series L"` + `"Shunt C"` vs `"Series C"` + `"Shunt L"`).

## Evaluation Parameters

Add performance goals to rank and filter designs. Only two parameter types: `'gammain'` (input reflection) and `'Gt'` (transducer gain).

```matlab
% Require return loss below -15 dB in-band
addEvaluationParameter(mn, 'gammain', '<', -15, [2.3e9 2.5e9], 2);

% Require insertion loss above -1 dB
addEvaluationParameter(mn, 'Gt', '>', -1, [2.3e9 2.5e9], 1);

% View current evaluations
params = getEvaluationParameters(mn);
disp(params);

% Remove an evaluation by row index
clearEvaluationParameter(mn, 1);
```

Arguments: `(mn, parameter, comparison, targetdB, [fLow fHigh], weight)` — all 6 are **positional and required** (weight is not optional).

A **default automatic evaluation** (`Gt > -3 dB` over `[CenterFrequency - Bandwidth/2, CenterFrequency + Bandwidth/2]`) is always present. Clear it with `clearEvaluationParameter(mn, 1)` if the automatic entry is at row 1.

**Gotcha:** Only `'gammain'` and `'Gt'` are supported — no S11, S21, VSWR, or other parameters.

## Exporting Circuits

```matlab
ckt = exportCircuits(mn);            % Best-ranked circuit only
ckts = exportCircuits(mn, [1 3]);    % Specific designs by index
```

Returns `circuit` objects containing `inductor` and `capacitor` elements. These can be used directly with `sparameters(ckt, freq)`, embedded in larger circuits via `add()`, or cascaded.

## Visualization

### Frequency Response

```matlab
rfplot(mn);                          % All designs, auto frequency range
rfplot(mn, freq);                    % Custom frequencies
rfplot(mn, freq, [1 3]);             % Specific designs
```

Plots two panels: input reflection coefficient (gammain) and transducer gain (Gt) in dB. Overlays evaluation goal lines.

### Smith Chart

```matlab
smithplot(mn);                                        % Best design
smithplot(mn, 'CircuitIndex', 2, 'Z0', 50);          % Specific design
```

Shows the impedance transformation path from source to load, tracing through each matching element in sequence (one arc per L/C component). This is the standard way to visualize each element's individual contribution to the match — no need to build intermediate circuits manually.

**Gotcha:** `smithplot(mn)` works directly on the `matchingnetwork` object. To plot an exported circuit's S-parameters on a Smith chart, use the S-parameter object form:

```matlab
ckt = exportCircuits(mn);
s = sparameters(ckt, linspace(1e9, 3e9, 201));
smithplot(s, 1, 1);                                   % Plot S11
```

Do not pass the circuit object directly to `smithplot` — it only accepts `matchingnetwork`, `sparameters`, or numeric data.

### S-Parameters

```matlab
s = sparameters(mn);                       % Best design, auto frequency
s = sparameters(mn, freq);                 % Custom frequencies
s = sparameters(mn, freq, Z0);             % Custom reference impedance
sArr = sparameters(mn, freq, Z0, [1 3]);   % Multiple designs -> array
```

The 4-argument form `sparameters(mn, freq, Z0, circuitIndex)` extracts S-parameters directly without needing `exportCircuits` first. Use `gammain(s)` or `powergain(s, 50, 50, 'Gt')` on the result for analysis.

## Richards Transformation (Lumped -> Stubs)

Convert lumped L/C matching elements to transmission line stubs. See `reference/lumped-to-distributed.md` for full Kuroda identity and realization details.

```matlab
txCkt = richards(mn, 2.4e9);              % Best design at 2.4 GHz
txCkts = richards(mn, 2.4e9, [1 2]);      % Specific designs
```

Returns `circuit` objects with `txlineElectricalLength` stub elements. Each stub has:
- `Z0` — characteristic impedance (from original element)
- `LineLength` — electrical length in radians (pi/4 for quarter-wave)
- `ReferenceFrequency` — the operating frequency argument
- `Termination` — `'Short'` for inductors, `'Open'` for capacitors
- `StubMode` — `'Series'` or `'Shunt'` matching the original topology

Stub element names get a `_tx` suffix: `C` becomes `C_tx`, `L` becomes `L_tx`.

## Tolerance Analysis

Evaluate matching network sensitivity to component value variations:

```matlab
% Default 5% tolerance on all components
toleranceAnalysis(mn);

% Custom tolerances per component
toleranceAnalysis(mn, [0.05 0.02 0.10]);    % 5%, 2%, 10% for each component

% Custom frequency range and specific circuit
toleranceAnalysis(mn, 0.05, linspace(0.5e9, 2.5e9, 500), 1);
```

Arguments: `(mn, tolerances, freqList, circuitIndices)` — all optional after `mn`. Plots the nominal and perturbed responses showing the spread due to component variations.

## Custom Matching Networks

Import user-designed circuits for comparison alongside auto-generated designs:

```matlab
% Build a custom matching circuit
custom = circuit('MyMatch');
add(custom, [1 2], inductor(3.3e-9, 'Ls'));
add(custom, [2 0], capacitor(0.8e-12, 'Cp'));
setports(custom, [1 0], [2 0]);

% Add to matchingnetwork for evaluation
addNetwork(mn, custom);    % Appears as "user_1" in circuitDescriptions

% Remove a custom network by name
deleteNetwork(mn, 'user_1');

% Suppress auto-generated designs to evaluate only custom circuits
disableAutomaticNetworks(mn);

% Re-enable auto-generated designs
enableAutomaticNetworks(mn);
```

Custom networks appear as `"user_1"`, `"user_2"`, etc. in `circuitDescriptions`.

**Gotcha:** Custom circuits must be 2-port with `setports` called, and contain only R/L/C elements. Other element types produce `-Inf` performance scores.

## Matching Network Designer App (R2021a+)

Interactive GUI for designing and comparing matching networks:

```matlab
matchingNetworkDesigner              % Launch empty
matchingNetworkDesigner(mn)          % Pre-loaded with existing design
```

The app provides visual comparison of topologies, interactive evaluation parameter setup, and Smith chart visualization. Configure programmatically with `matchingnetwork`, then open in the designer for interactive exploration.

## Handle Class Behavior

`matchingnetwork` is a **handle class**. These methods mutate in-place (no output argument):
- `addNetwork`, `deleteNetwork`
- `disableAutomaticNetworks`, `enableAutomaticNetworks`

Use `copy(mn)` for independent deep copies.

## Gotchas

1. **Property is `Bandwidth` (lowercase 'w')** — not `BandWidth` as some help text suggests.
2. **Components range is 2–3 only** — `Components = 1` or `Components = 4` error.
3. **Only `'gammain'` and `'Gt'` evaluation parameters** — no S11, S21, VSWR. All 6 positional arguments are required: `(mn, parameter, comparison, targetdB, [fLow fHigh], weight)`.
4. **Handle class** — `addNetwork`/`deleteNetwork` mutate in-place. Assigning their output errors ("too many output arguments").
5. **Custom circuits must be 2-port RLC only** — other element types produce warnings and `-Inf` scores.
6. **Default automatic evaluation always present** — `Gt > -3 dB` over bandwidth. Clear explicitly if unwanted.
7. **`Bandwidth` takes precedence** over `LoadedQ` when both are specified.
8. **`smithplot` only accepts `matchingnetwork` or `sparameters` objects** — do not pass a `circuit` object directly. For exported circuits, call `sparameters(ckt, freq)` first, then `smithplot(s, 1, 1)`.
9. **`performanceScore` is relative, not absolute** — use it to rank candidates against each other, not as a fixed quality metric. The scale depends on your evaluation parameters and weights.

## Conventions

- Use `tiledlayout`/`nexttile` for multi-panel matching analysis plots
- Show `smithplot` alongside `rfplot` for matching network analysis
- Always inspect `circuitDescriptions` to compare candidates before selecting
- Label axes with units (GHz, dB) and include figure titles

----

Copyright 2026 The MathWorks, Inc.

----
slack-gif-creatoranthropics/skillsKnowledge and utilities for creating animated GIFs optimized for Slack. Provides constraints, validation tools, and animation concepts. Use when users request animated GIFs for Slack like "make me a GIF of X doing Y for Slack."Stars: 197kUpdated Claude CodeCodexVerifiedInstall
ponytailDietrichGebert/ponytailLazy senior dev mode for any coding task: the smallest change that fully solves it, and a reply a busy human understands in one read.Stars: 194kUpdated Claude CodeCodexVerifiedInstall
ponytail-debtDietrichGebert/ponytailHarvest every shortcut: comment into one debt ledger, so deferrals get tracked instead of forgotten. One-shot report.Stars: 194kUpdated Claude CodeCodexVerifiedInstall
ponytail-debtDietrichGebert/ponytailList every `shortcut:` comment (and older `ponytail:` ones) as a debt ledger. One-shot report, changes nothing. Use for "ponytail debt", "what did ponytail defer", "list the shortcuts", /ponytail-debt.Stars: 194kUpdated Claude CodeCodexVerifiedInstall