feat(plecs): add PLECS circuit simulator processor and sample files by tinix84 · Pull Request #57 · OpenMagnetics/MKF

tinix84 · 2026-04-01T07:28:09Z

Summary

Add CircuitSimulatorPlecs processor for generating PLECS XML simulation files from MKF magnetic component models
Add PLECS sample files (simple inductor, saturable inductor, transformer, E-core transformer) for XML reverse-engineering
Add PLECS to the exporter enum and declare PlecsModel class in CircuitSimulatorInterface
Add Catch2 test suite for the PLECS processor

Test plan

Verify PLECS XML output matches expected structure for inductors and transformers
Run TestCircuitSimulatorPlecs test suite
Validate sample .plecs files can be opened in PLECS simulator

Closes #56

🤖 Generated with Claude Code

…ring Add E42/21/15 two-winding transformer PLECS model with N87 core material, 24/12 turn ratio, saturable permeance, and core loss equivalent. This sample file supports reverse-engineering the PLECS XML schema for the upcoming PLECS exporter (#55). Refs: #56 Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

…ering Add simple_inductor.plecs (E32/16/9, N87, 20 turns, 1mm gap), saturable_inductor.plecs (E32/16/9, N87 with BH curve, 20 turns, 0.5mm gap), and ecore_transformer.plecs (E42/21/15, 3-leg magnetic circuit, 24/12 turns, center-leg 1mm gap). All files use the same PLECS format as transformer.plecs with parameterized init commands and magnetic circuit schematics. Refs: #56 Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

Add PLECS to CircuitSimulatorExporterModels enum, from_json/to_json serialization, factory method, and class declaration with all method signatures in CircuitSimulatorInterface.h/.cpp.

Add CircuitSimulatorPlecs processor for generating PLECS XML simulation files and corresponding Catch2 test. Update simple_inductor sample file with revised XML structure. Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

Copilot

Pull request overview

Adds initial support for exporting OpenMagnetics magnetic models into PLECS .plecs XML schematics, along with sample PLECS files and a Catch2 test suite to validate basic exporter output.

Changes:

Introduces CircuitSimulatorExporterPlecsModel to generate PLECS schematic text for magnetic subcircuits and symbols.
Extends CircuitSimulatorExporterModels with PLECS and wires it into the exporter model factory.
Adds PLECS sample .plecs files and a new Catch2 test suite for the PLECS exporter.

Reviewed changes

Copilot reviewed 8 out of 8 changed files in this pull request and generated 7 comments.

Show a summary per file

File	Description
`src/processors/CircuitSimulatorPlecs.cpp`	Implements PLECS export generation (header/init commands/schematic emitters).
`src/processors/CircuitSimulatorInterface.h`	Adds `PLECS` to the exporter enum and declares `CircuitSimulatorExporterPlecsModel`.
`src/processors/CircuitSimulatorInterface.cpp`	Hooks `PLECS` into the exporter model factory and updates the unsupported-model error message.
`tests/TestCircuitSimulatorPlecs.cpp`	Adds Catch2 tests that generate `.plecs` output and check for expected tokens.
`tests/testData/simple_inductor.plecs`	Adds a sample PLECS model file (currently placed under `tests/testData/`).
`tests/testData/saturable_inductor.plecs`	Adds a sample PLECS model file (currently placed under `tests/testData/`).
`tests/testData/transformer.plecs`	Adds a sample PLECS model file (currently placed under `tests/testData/`).
`tests/testData/ecore_transformer.plecs`	Adds a sample PLECS model file (currently placed under `tests/testData/`).

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

Copilot · 2026-04-01T07:36:42Z

src/processors/CircuitSimulatorPlecs.cpp

+    std::optional<std::string> filePathOrFile,
+    CircuitSimulatorExporterCurveFittingModes mode) {


filePathOrFile and mode are currently unused in this implementation, which can trigger compiler warnings and makes it unclear whether these parameters are intentionally ignored for PLECS. If they’re not needed, consider removing the parameter names or marking them [[maybe_unused]]; if they are needed, wire them into the export logic (e.g., curve-fitting mode for loss models, including external schematic fragments).

Suggested change

std::optional<std::string> filePathOrFile,

CircuitSimulatorExporterCurveFittingModes mode) {

[[maybe_unused]] std::optional<std::string> filePathOrFile,

[[maybe_unused]] CircuitSimulatorExporterCurveFittingModes mode) {

Copilot · 2026-04-01T07:36:43Z

tests/TestCircuitSimulatorPlecs.cpp

+    auto plecsFile = outputFilePath;
+    plecsFile.append("Test_Plecs_Simple_Inductor.plecs");
+    std::filesystem::remove(plecsFile);
+
+    CircuitSimulatorExporter exporter(CircuitSimulatorExporterModels::PLECS);
+    exporter.export_magnetic_as_subcircuit(magnetic, 100000, 25, plecsFile.string());
+
+    REQUIRE(std::filesystem::exists(plecsFile));
+


The test writes into .../output/... but never ensures the output directory exists. On a clean checkout this will cause the exporter’s ofstream to fail and the exists() assertion to fail. Consider calling std::filesystem::create_directories(outputFilePath); before exporting in each test case (or once in a shared setup).

Copilot · 2026-04-01T07:36:43Z

tests/testData/plecs/simple_inductor.plecs

+Plecs {
+  Name          "simple_inductor"
+  Version       "4.7"
+  CircuitModel  "ContStateSpace"
+  StartTime     "0.0"


The issue/PR description calls for PLECS reverse‑engineering sample files to be placed under tests/testData/plecs/, but this file is added directly under tests/testData/. If the intended location is the subdirectory, move the sample files accordingly (and update any tests/tools that load them).

Copilot · 2026-04-01T07:36:44Z

tests/testData/plecs/transformer.plecs

+    Component {
+      Type          Reference
+      SrcComponent  "Components/Magnetic/Components/P_sat"
+      Name          "P_satTr"
+      Show          off
+      Position      [405, 240]
+      Direction     up
+      Flipped       off
+      LabelPosition east
+      Frame         [-8, -15; 8, 15]
+      Parameter {
+        Variable      "fitting"
+        Value         "1"
+        Show          off
+      }
+      Parameter {
+        Variable      "A"
+        Value         "A_e"
+        Show          off
+      }
+      Parameter {
+        Variable      "l"
+        Value         "l_e"
+        Show          off
+      }
+      Parameter {
+        Variable      "mu_r_unsat"
+        Value         "mu_r"
+        Show          off
+      }
+      Parameter {
+        Variable      "mu_r_sat"
+        Value         "1"
+        Show          off
+      }
+      Parameter {
+        Variable      "B_sat"
+        Value         "B_sat"
+        Show          off
+      }
+      Parameter {
+        Variable      "F_init"
+        Value         "0"
+        Show          off
+      }
+      Terminal {
+        Type          MagneticPort
+        Position      [0, -20]
+        Direction     up
+      }
+      Terminal {
+        Type          MagneticPort
+        Position      [0, 20]
+        Direction     down
+      }
+    }
+    Component {
+      Type          Reference
+      SrcComponent  "Components/Magnetic/Components/P_air"
+      Name          "P_air2"
+      Show          off
+      Position      [440, 355]
+      Direction     up
+      Flipped       on


Per Issue #56, the transformer.plecs sample is specified as an ungapped 2‑winding transformer with a core-loss Magnetic Resistance (Rm) in series with the core permeance. This file currently includes a P_air element and does not include any magnetic resistance/core-loss element, so it doesn’t match the stated topology/parameters. Either update the sample to match the spec (core-loss element, no air-gap element) or update the issue/PR description so the intended reverse‑engineering target is unambiguous.

Copilot · 2026-04-01T07:36:44Z

tests/testData/plecs/saturable_inductor.plecs

+    Component {
+      Type          MagneticInterface
+      Name          "MagInt1"
+      Show          off
+      Position      [440, 290]
+      Direction     up
+      Flipped       off
+      LabelPosition east
+      Parameter {
+        Variable      "n"
+        Value         "n_turns"
+        Show          off
+      }
+      Parameter {
+        Variable      "Polarity"
+        Value         "1"
+        Show          off
+      }
+    }
+    Component {
+      Type          Reference
+      SrcComponent  "Components/Magnetic/Components/P_sat"
+      Name          "P_satCore"
+      Show          off
+      Position      [405, 240]
+      Direction     up
+      Flipped       off
+      LabelPosition east
+      Frame         [-8, -15; 8, 15]
+      Parameter {
+        Variable      "fitting"
+        Value         "1"
+        Show          off
+      }
+      Parameter {
+        Variable      "A"
+        Value         "A_e"
+        Show          off
+      }
+      Parameter {
+        Variable      "l"
+        Value         "l_e"
+        Show          off
+      }
+      Parameter {
+        Variable      "mu_r_unsat"
+        Value         "mu_r"
+        Show          off
+      }
+      Parameter {
+        Variable      "mu_r_sat"
+        Value         "1"
+        Show          off
+      }
+      Parameter {
+        Variable      "B_sat"
+        Value         "B_sat"
+        Show          off
+      }
+      Parameter {
+        Variable      "F_init"
+        Value         "0"
+        Show          off
+      }
+      Terminal {
+        Type          MagneticPort
+        Position      [0, -20]
+        Direction     up
+      }
+      Terminal {
+        Type          MagneticPort
+        Position      [0, 20]
+        Direction     down
+      }
+    }


Issue #56 specifies that saturable_inductor.plecs should use a saturable core with an explicit B-H curve (5–10 points up to ~0.4 T). This sample appears to use the generic P_sat component without any embedded/defined B-H curve data points, so it may not support the intended XML reverse‑engineering. Consider updating the model to include the B-H table data (or adjusting the issue/PR description if P_sat is the chosen representation).

Copilot · 2026-04-01T07:36:44Z

src/processors/CircuitSimulatorPlecs.cpp

+
+        // Emit windings
+        for (size_t i = 0; i < numWindings; ++i) {
+            std::string miName = "MagInt_w" + std::to_string(i);
+            std::string turnsVar = (numWindings == 1) ? "n_turns" : "n_w" + std::to_string(i);
+            bool isPrimary = windings[i].get_isolation_side() == IsolationSide::PRIMARY;
+            int yPos = yBase + static_cast<int>(i) * ySpacing;
+
+            if (isPrimary) {
+                schematic << emit_magnetic_interface(miName, turnsVar, 1, {magX, yPos}, "up", false);
+            } else {
+                schematic << emit_magnetic_interface(miName, turnsVar, 1, {magX, yPos - ySpacing}, "down", true);


In the single-column topology, secondary windings are emitted at yPos - ySpacing. For a typical 2‑winding part where winding[0] is PRIMARY and winding[1] is SECONDARY, this places both MagneticInterface blocks at the same Y coordinate (overlapping symbols) because yPos for i=1 becomes yBase + ySpacing and then subtracts ySpacing. Consider making Y placement purely index-based (e.g., yBase - i*ySpacing) and using direction/flipped only for orientation, or computing separate Y tracks for PRIMARY vs SECONDARY windings.

Suggested change

// Emit windings

for (size_t i = 0; i < numWindings; ++i) {

std::string miName = "MagInt_w" + std::to_string(i);

std::string turnsVar = (numWindings == 1) ? "n_turns" : "n_w" + std::to_string(i);

bool isPrimary = windings[i].get_isolation_side() == IsolationSide::PRIMARY;

int yPos = yBase + static_cast<int>(i) * ySpacing;

if (isPrimary) {

schematic << emit_magnetic_interface(miName, turnsVar, 1, {magX, yPos}, "up", false);

} else {

schematic << emit_magnetic_interface(miName, turnsVar, 1, {magX, yPos - ySpacing}, "down", true);

int primaryIndex = 0;

int secondaryIndex = 0;

// Emit windings

for (size_t i = 0; i < numWindings; ++i) {

std::string miName = "MagInt_w" + std::to_string(i);

std::string turnsVar = (numWindings == 1) ? "n_turns" : "n_w" + std::to_string(i);

bool isPrimary = windings[i].get_isolation_side() == IsolationSide::PRIMARY;

int yPos = yBase;

if (isPrimary) {

yPos = yBase - primaryIndex * ySpacing;

primaryIndex++;

schematic << emit_magnetic_interface(miName, turnsVar, 1, {magX, yPos}, "up", false);

} else {

yPos = yBase + secondaryIndex * ySpacing;

secondaryIndex++;

schematic << emit_magnetic_interface(miName, turnsVar, 1, {magX, yPos}, "down", true);

Copilot · 2026-04-01T07:36:44Z

src/processors/CircuitSimulatorPlecs.cpp

+        for (size_t i = 0; i < numWindings; ++i) {
+            std::string miName = "MagInt_w" + std::to_string(i);
+            std::string turnsVar = "n_w" + std::to_string(i);
+            int yPos = yWindingBase + static_cast<int>(i) * ySpacing;
+            if (windings[i].get_isolation_side() == IsolationSide::PRIMARY) {
+                schematic << emit_magnetic_interface(miName, turnsVar, 1, {magX, yPos}, "up", false);
+            } else {
+                schematic << emit_magnetic_interface(miName, turnsVar, 1, {magX, yPos - ySpacing}, "down", true);


The same Y-placement logic in the multi-column (E-core) branch can overlap PRIMARY/SECONDARY windings for 2‑winding parts (yPos - ySpacing for SECONDARY collapses onto the PRIMARY Y). This makes the generated schematic hard to inspect in PLECS and can hide components behind each other.

Suggested change

for (size_t i = 0; i < numWindings; ++i) {

std::string miName = "MagInt_w" + std::to_string(i);

std::string turnsVar = "n_w" + std::to_string(i);

int yPos = yWindingBase + static_cast<int>(i) * ySpacing;

if (windings[i].get_isolation_side() == IsolationSide::PRIMARY) {

schematic << emit_magnetic_interface(miName, turnsVar, 1, {magX, yPos}, "up", false);

} else {

schematic << emit_magnetic_interface(miName, turnsVar, 1, {magX, yPos - ySpacing}, "down", true);

int primaryIndex = 0;

int secondaryIndex = 0;

for (size_t i = 0; i < numWindings; ++i) {

std::string miName = "MagInt_w" + std::to_string(i);

std::string turnsVar = "n_w" + std::to_string(i);

if (windings[i].get_isolation_side() == IsolationSide::PRIMARY) {

int yPos = yWindingBase + primaryIndex * ySpacing;

schematic << emit_magnetic_interface(miName, turnsVar, 1, {magX, yPos}, "up", false);

++primaryIndex;

} else {

int yPos = yWindingBase - secondaryIndex * ySpacing;

schematic << emit_magnetic_interface(miName, turnsVar, 1, {magX, yPos}, "down", true);

++secondaryIndex;

…nto shared helpers - Replace hardcoded B_sat=0.49 with core.get_magnetic_flux_density_saturation() - Fix multi-column mode only using first center gap (removed break, added gap chaining with proper vertical spacing) - Fix electrical connections hardcoded to 2 windings — now loops by isolation_side so each secondary gets its own R_load - Extract build_physical_init, build_magnetic_schematic, and build_electrical_schematic helpers to eliminate ~200 lines of duplication between export_magnetic_as_subcircuit and export_magnetic_as_symbol - Add 3 new tests: B_sat from material, 3-winding transformer, distributed gaps Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

…ample-files

- Merge emit_magnetic_connection, emit_wire_connection, emit_signal_connection into single emit_connection(type, ...) to eliminate code duplication - Extract assemble_plecs_file() helper from both export functions - Add EnsureOutputDir for test output directory creation - transformer.plecs: replace P_air with MagneticResistance (Rm) per spec - saturable_inductor.plecs: add B-H curve data (7-point N87), set fitting=2 Refs: #56 Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

sonarqubecloud · 2026-04-03T15:20:43Z

Quality Gate failed

Failed conditions
22.6% Duplication on New Code (required ≤ 3%)

See analysis details on SonarQube Cloud

tinix84 and others added 4 commits March 31, 2026 19:50

feat(plecs): add PLECS to exporter enum, declare PlecsModel class

29d6bce

Add PLECS to CircuitSimulatorExporterModels enum, from_json/to_json serialization, factory method, and class declaration with all method signatures in CircuitSimulatorInterface.h/.cpp.

Copilot AI review requested due to automatic review settings April 1, 2026 07:28

Copilot AI reviewed Apr 1, 2026

View reviewed changes

AlfVII and others added 3 commits April 2, 2026 19:51

Merge remote-tracking branch 'origin/main' into feat/issue-56-plecs-s…

fba31f9

…ample-files

tinix84 merged commit 652148e into main Apr 3, 2026
1 of 3 checks passed

tinix84 deleted the feat/issue-56-plecs-sample-files branch April 3, 2026 15:16

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

feat(plecs): add PLECS circuit simulator processor and sample files#57

feat(plecs): add PLECS circuit simulator processor and sample files#57
tinix84 merged 7 commits intomainfrom
feat/issue-56-plecs-sample-files

tinix84 commented Apr 1, 2026

Uh oh!

Copilot AI left a comment

Uh oh!

Copilot AI Apr 1, 2026

Uh oh!

Copilot AI Apr 1, 2026

Uh oh!

Copilot AI Apr 1, 2026

Uh oh!

Copilot AI Apr 1, 2026

Uh oh!

Copilot AI Apr 1, 2026

Uh oh!

Copilot AI Apr 1, 2026

Uh oh!

Copilot AI Apr 1, 2026

Uh oh!

Uh oh!

sonarqubecloud bot commented Apr 3, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

		std::optional<std::string> filePathOrFile,
		CircuitSimulatorExporterCurveFittingModes mode) {

Conversation

tinix84 commented Apr 1, 2026

Summary

Test plan

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Reviewed changes

Uh oh!

Copilot AI Apr 1, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Apr 1, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Apr 1, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Apr 1, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Apr 1, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Apr 1, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Apr 1, 2026

Choose a reason for hiding this comment

Uh oh!

Uh oh!

sonarqubecloud bot commented Apr 3, 2026

Quality Gate failed

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants