Skip to content

richie-rk/PyTeslaCoil

BranchesTags

Repository files navigation

PyTeslaCoil

PyTeslaCoil

License: MIT Python 3.10+ Tests Hugging Face Spaces

An open-source Tesla coil design calculator — the modern Python replacement for JavaTC, featuring a pure-Python physics engine, Pydantic v2 models, and a NiceGUI web frontend.

PyTeslaCoil Results Dashboard

Table of Contents

Overview

PyTeslaCoil is a complete electromagnetic design tool for spark-gap (SGTC) and dual-resonant solid-state (DRSSTC) Tesla coils. It replicates all core features of JavaTC (classictesla.com) — the standard tool in the Tesla coil community since ~2000 — in a modern, all-Python stack with a browser-based UI.

The application computes secondary/primary inductance, self-capacitance, coupling coefficient, resonant frequencies, spark-gap behavior, transformer sizing, and estimated spark length. It can be used as a standalone library for scripts and notebooks, or as a full desktop app with live-updating calculations and a 2D coil visualizer.

Screenshots

Secondary Coil Design

Enter coil geometry, wire gauge, and winding parameters. Results update live as you type.

Secondary Coil Tab

Primary Coil Design

Configure flat spiral, helical, or conical primaries with auto-tune to match secondary frequency.

Primary Coil Tab

Coupling Analysis

View the computed coupling coefficient with quality indicators. Auto-adjust primary height to hit a target k.

Coupling Tab

Spark Gap Configuration

Static and rotary gap parameters with breakdown voltage, BPS, and energy-per-bang calculations.

Spark Gap Tab

Results Dashboard

Headline metrics (resonant frequency, coupling, spark length), detailed subsection results, coil cross-section visualizer, and one-click export to text, JSON, or PDF.

Results Dashboard

Features

  • Secondary Coil Design: Solenoid, conical, and inverse conical geometries with Wheeler inductance and Medhurst self-capacitance
  • Primary Coil Design: Flat spiral (pancake), helical, and conical primaries with lead inductance correction
  • Topload Capacitance: Toroid (empirical fit) and sphere (exact analytical) with multi-stage stacking support
  • Coupling Coefficient: Neumann mutual inductance via complete elliptic integrals (K, E) with NumPy-vectorized O(N_pri x N_sec) filamentary method
  • Auto-Tune: Automatic primary turns calculation to match secondary resonant frequency via SciPy root-finding (brentq)
  • Static & Rotary Spark Gaps: Paschen breakdown, BPS, dwell time, cap charge fraction, and energy-per-bang
  • Transformer Sizing: NST, OBIT, MOT, and pole-pig support with resonant and LTR capacitor recommendations
  • Spark Length Estimation: Freau empirical formula from input power or BPS x bang energy
  • 2D Coil Visualizer: Proportional SVG cross-section rendering of secondary, primary, topload, and ground plane
  • Export: JavaTC-style consolidated text, round-trippable JSON, and single-page PDF (via reportlab)
  • Demo Presets: Three built-in coil designs (small SGTC, medium SGTC with rotary gap, compact DRSSTC)
  • Library Mode: Zero-UI engine — import and use from scripts, notebooks, or your own applications
  • Dark-Themed Web UI: NiceGUI tabbed interface with live recalculation, debounced at 300ms

Architecture

The system features a three-layer architecture with strict dependency direction — the engine never imports from the UI:

User (browser) <--> NiceGUI Frontend (ui/) <--> Physics Engine (pyteslacoil/engine/)
                                                        |
                                                  Pydantic Models (pyteslacoil/models/)

Physics Engine (pyteslacoil/engine/)

  • Pure Python: No UI dependencies — usable as a standalone library
  • 11 Calculation Modules: secondary, primary, topload, coupling, tuning, spark_gap_static, spark_gap_rotary, transformer, spark_length, environment, medhurst
  • SI Internally: All calculations use meters, henries, farads, hertz — unit conversion happens only at boundaries
  • Pure Functions: No global state, no side effects — Input model in, output model out
  • NumPy + SciPy: Vectorized elliptic integrals for coupling, brentq root-finding for auto-tune, numpy.interp for Medhurst table

Data Models (pyteslacoil/models/)

  • Pydantic v2: 28 classes (6 enums + 22 BaseModel subclasses) with full validation
  • Typed Contracts: Every engine function takes a typed input and returns a typed output
  • Serializable: All models serialize to JSON for export/import round-tripping
  • Reactive: validate_assignment=True enables NiceGUI live binding

NiceGUI Frontend (ui/)

  • 8 Tabbed Panels: Secondary, Primary, Topload, Coupling, Spark Gap, Transformer, Environment, Results
  • Observer Pattern: AppState owns one CoilDesign (inputs) and one FullSystemOutput (outputs); widgets subscribe to change notifications
  • SVG Visualizer: Proportional cross-section view generated as inline SVG with color-coded components
  • Debounced Recalculation: 300ms throttle on input changes to prevent UI lag

Presets & Export (pyteslacoil/presets/, pyteslacoil/export/)

  • 3 Demo Coils: Small SGTC (Hackaday Mini), Medium SGTC (TCML reference), Compact DRSSTC
  • 3 Export Formats: Consolidated text (JavaTC-style), JSON (round-trippable), PDF (single-page via reportlab)

Installation

Prerequisites

  • Python 3.10+ (tested on 3.10, 3.11, 3.12, 3.13)
  • pip or uv package manager
  • 4GB+ RAM (coupling calculation is vectorized and memory-efficient)
  • Optional: reportlab for PDF export

Setup

  1. Clone the repository:

    git clone https://github.com/richie-rk/PyTeslaCoil.git
cd PyTeslaCoil

  2. Create a virtual environment:

    python -m venv venv
source venv/bin/activate  # On Windows: venv\Scripts\activate

  3. Install dependencies (choose one method):

    Option A: Editable install with dev tools (Recommended)

    pip install -e ".[dev]"

    Option B: With PDF export support

    pip install -e ".[dev,pdf]"

    Option C: Runtime only (no test/lint tools)

    pip install -e .

    Which method to choose?

    • Option A: Best for development — includes pytest, ruff, mypy
    • Option B: Full install with PDF export via reportlab
    • Option C: Minimal install for end users

  4. Verify installation:

    python -c "from pyteslacoil import calculate_secondary; print('Engine OK')"
pytest tests/ -q

Usage

Quick Start — Web UI

  1. Launch the application:

    pyteslacoil
# or
python -m ui.main

  2. Access the application:

  3. Load a demo coil: Click Load Demo in the header to pick a preset (Small SGTC, Medium SGTC, or DRSSTC)

Library Mode — Use as a Python Package

from pyteslacoil import calculate_secondary, calculate_primary, calculate_topload
from pyteslacoil.models import SecondaryInput, PrimaryInput, ToploadInput, ToploadType
from pyteslacoil.units import inches_to_meters
from pyteslacoil.engine.tuning import auto_tune

# Design a secondary coil
secondary = SecondaryInput(
    radius_1=inches_to_meters(2.125),
    radius_2=inches_to_meters(2.125),
    height_1=inches_to_meters(3.0),
    height_2=inches_to_meters(19.0),
    turns=711,
    wire_awg=24,
)
sec_out = calculate_secondary(secondary)

print(f"Inductance:    {sec_out.inductance_mh:.3f} mH")
print(f"Self-cap:      {sec_out.self_capacitance_pf:.2f} pF")
print(f"Resonant freq: {sec_out.resonant_frequency_khz:.2f} kHz")
print(f"Q factor:      {sec_out.q_factor:.0f}")
print(f"Wire length:   {sec_out.wire_length_ft:.1f} ft")

Preset Loading

from pyteslacoil.presets import load_preset
from ui.state import AppState

# Load a complete coil design and calculate everything
state = AppState(load_preset("small_sgtc"))
state.recalculate()

print(f"System freq:   {state.outputs.system_resonant_frequency_khz:.2f} kHz")
print(f"Coupling k:    {state.outputs.coupling.coupling_coefficient:.4f}")
print(f"Spark length:  {state.outputs.estimated_spark_length_in:.1f} in")

Export

from pyteslacoil.export import to_text, to_json

# Generate a JavaTC-style text report
report = to_text(state.design, state.outputs)
print(report)

# Export as JSON (round-trippable)
json_str = to_json(state.design, state.outputs)
with open("my_coil.json", "w") as f:
    f.write(json_str)

Configuration

Unit System

The UI supports switching between inches and centimeters via the header dropdown. Internally, all calculations use SI (meters, henries, farads, hertz). Unit conversion is handled by pyteslacoil/units.py at the input/output boundary.

Available Presets

Preset ID Description Secondary Primary Gap Type
small_sgtc Hackaday Mini SGTC 1.66" OD, 570T AWG32 Helical, 11T Static
medium_sgtc TCML Reference Coil 4.25" OD, 711T AWG24 Flat spiral, 10.8T Rotary
drsstc Compact DRSSTC 3" OD, 1000T AWG30 Conical, 6T N/A (solid-state)

Engine Module Reference

Module Function Physics
secondary.py calculate_secondary() Wheeler inductance, Medhurst C, skin effect, Q factor
primary.py calculate_primary() Pancake/helical/conical inductance, lead inductance
topload.py calculate_topload() Toroid empirical fit, sphere exact (4piER)
coupling.py calculate_coupling() Neumann mutual inductance via elliptic integrals
tuning.py auto_tune() Root-finding for primary turns to match secondary f
spark_gap_static.py calculate_static_gap() Paschen breakdown, BPS, bang energy
spark_gap_rotary.py calculate_rotary_gap() Rotary BPS, dwell time, charge fraction
transformer.py calculate_transformer() NST/OBIT/MOT sizing, resonant & LTR cap
spark_length.py estimate_spark_length() Freau empirical formula
environment.py calculate_environment() Proximity correction for ground/walls/ceiling
medhurst.py medhurst_coefficient() Medhurst H table interpolation via numpy.interp

Testing

# Run the full test suite (73 tests)
pytest tests/ -v

# Run with coverage
pytest tests/ -v --cov=pyteslacoil

# Lint check
ruff check pyteslacoil/

# Run a specific test module
pytest tests/test_secondary.py -v

Test Coverage

Module Tests Key Validations
test_secondary.py 7 Wheeler proportional to N-squared, wire length exact, skin depth ~65um at 1MHz
test_primary.py 7 Pancake/helical/conical geometry detection, frequency decreases with more turns
test_topload.py 7 Sphere exact vs analytical 4piER, toroid in reasonable pF range
test_coupling.py 5 M > 0 for concentric coils, k in typical range, auto-adjust converges
test_tuning.py 5 Required L formula, helical/spiral auto-tune hit target within 5%
test_medhurst.py 6 Table endpoints, clamping, monotonicity, C = 3.9pF for known geometry
test_spark_gap.py 4 Breakdown monotonic, ~3kV at 1mm, BPS formulas correct
test_transformer.py 2 NST 15kV/60mA: VA=900, resonant cap formula exact
test_spark_length.py 3 500W produces ~38in, energy form preferred over power form
test_environment.py 3 Free space factor = 1.0, walls/ceiling increase capacitance
test_units.py 12 Round-trip conversions for all unit types
test_wire_data.py 5 Diameter decreases with AWG, resistance increases, SI cache consistent
test_presets_export.py 6 All 3 presets load and calculate, text export contains name, JSON round-trips
test_integration.py 1 Full end-to-end SGTC: secondary through spark length

Project Structure

PyTeslaCoil/
├── pyproject.toml                   # PEP 621 packaging, ruff, pytest config
├── requirements.txt                 # Runtime dependencies
├── README.md                        # This file
├── LICENSE                          # MIT
├── .gitignore
├── CLAUDE.md                        # Build conventions for Claude Code
├── DESIGN.md                        # UI design system tokens (from Stitch)
│
├── pyteslacoil/                     # Main package
│   ├── __init__.py                  # Public API (10 re-exported functions)
│   ├── constants.py                 # Physical constants (MU_0, EPSILON_0, copper)
│   ├── units.py                     # SI <-> inches/cm/pF/uH converters (22 functions)
│   ├── wire_data.py                 # AWG 4-44 wire table (21 gauges)
│   │
│   ├── engine/                      # Pure-Python physics engine
│   │   ├── medhurst.py              # Self-capacitance table + numpy.interp
│   │   ├── secondary.py             # Wheeler, Medhurst, skin effect, Q
│   │   ├── primary.py               # Pancake/helical/conical + lead inductance
│   │   ├── topload.py               # Toroid empirical + sphere exact
│   │   ├── coupling.py              # Neumann/elliptic-integral mutual inductance
│   │   ├── tuning.py                # Auto-tune via scipy.optimize.brentq
│   │   ├── spark_gap_static.py      # Paschen, BPS, bang energy
│   │   ├── spark_gap_rotary.py      # Rotary BPS, dwell, charge fraction
│   │   ├── transformer.py           # NST/OBIT/MOT, resonant & LTR cap
│   │   ├── spark_length.py          # Freau empirical formula
│   │   └── environment.py           # Proximity correction multiplier
│   │
│   ├── models/                      # Pydantic v2 models (28 classes)
│   │   ├── coil_design.py           # CoilDesign, FullSystemOutput, enums
│   │   ├── secondary_model.py       # SecondaryInput, SecondaryOutput
│   │   ├── primary_model.py         # PrimaryInput, PrimaryOutput
│   │   ├── topload_model.py         # ToploadInput, ToploadOutput
│   │   ├── coupling_model.py        # CouplingInput, CouplingOutput
│   │   ├── spark_gap_model.py       # Static/Rotary Gap Input/Output
│   │   ├── transformer_model.py     # TransformerInput, TransformerOutput
│   │   └── environment_model.py     # EnvironmentInput, EnvironmentOutput
│   │
│   ├── presets/                     # Demo coil definitions
│   │   ├── demo_small_sgtc.py       # Hackaday Mini SGTC
│   │   ├── demo_medium_sgtc.py      # TCML reference (rotary gap)
│   │   └── demo_drsstc.py           # Compact DRSSTC
│   │
│   └── export/                      # Output exporters
│       ├── consolidated.py          # JavaTC-style text report
│       ├── json_export.py           # Round-trippable JSON
│       └── pdf_export.py            # Single-page PDF (reportlab)
│
├── ui/                              # NiceGUI frontend
│   ├── main.py                      # App entry point, tab layout
│   ├── state.py                     # Reactive AppState + recalculate pipeline
│   ├── theme.py                     # Design system tokens + global CSS
│   └── components/                  # 11 UI modules
│       ├── cards.py                 # Shared card & metric components
│       ├── header.py                # Title bar, units, Load Demo
│       ├── secondary_tab.py         # Secondary inputs + live outputs
│       ├── primary_tab.py           # Primary inputs + auto-tune toggle
│       ├── topload_tab.py           # Toroid/sphere selector
│       ├── coupling_tab.py          # k display + auto-adjust
│       ├── spark_gap_tab.py         # Static/rotary sub-tabs
│       ├── transformer_tab.py       # Transformer inputs + cap sizing
│       ├── environment_tab.py       # Ground plane, walls, ceiling
│       ├── results_tab.py           # Summary + export buttons
│       ├── coil_visualizer.py       # SVG cross-section renderer
│       └── presets_dialog.py        # Demo coil loader modal
│
├── tests/                           # 73 tests across 14 files
│   ├── test_secondary.py
│   ├── test_primary.py
│   ├── test_topload.py
│   ├── test_coupling.py
│   ├── test_tuning.py
│   ├── test_medhurst.py
│   ├── test_spark_gap.py
│   ├── test_transformer.py
│   ├── test_spark_length.py
│   ├── test_environment.py
│   ├── test_units.py
│   ├── test_wire_data.py
│   ├── test_presets_export.py
│   ├── test_integration.py
│   └── fixtures/
│       └── javatc_reference.json
│
└── docs/
    ├── PyTeslaCoil_Implementation_Guide.md
    ├── TECHNICAL_SUMMARY.md
    └── screenshots/                 # UI screenshots for README

Physics Reference

Key Formulas

Formula Equation Source
Solenoid inductance L = (u0 * N^2 * pi * r^2) / (l + 0.9r) Wheeler (1928)
Flat spiral inductance L = (u0 * N^2 * r_avg^2) / (8*r_avg + 11*w) Wheeler (1928)
Self-capacitance C [pF] = H * D [cm] Medhurst (1947)
Resonant frequency f = 1 / (2*pi*sqrt(L*C)) LC resonance
Sphere capacitance C = 4*pi*e0*r Gauss's law (exact)
Mutual inductance M = u0*sqrt(ab)*((2/k-k)*K(k^2) - (2/k)*E(k^2)) Maxwell (1873)
Coupling coefficient k = M / sqrt(L1*L2) Definition
Skin depth d = sqrt(rho / (pi*f*u0)) EM theory
Spark length L [in] = 1.7 * sqrt(P [W]) Freau (empirical)
Q factor Q = 2*pi*f*L / R_ac Definition

Data Sources

  • Medhurst table: 20-point L/D-ratio lookup, linearly interpolated via numpy.interp
  • AWG wire table: 21 gauges (AWG 4–44) from Phelps-Dodge / NEC standard tables
  • Toroid capacitance: Empirical fit calibrated against JavaTC and experimental data

Acknowledgements

  • Inspired by JavaTC by Bart Anderson — the standard Tesla coil calculator since ~2000
  • Physics formulas based on the work of Paul Nicholson (GEOTC/TSSP), R.G. Medhurst, H.A. Wheeler, F.W. Grover, and the global Tesla coil community
  • Built with NiceGUI, Pydantic, NumPy, SciPy, and Plotly

Built with ⚡ for the Tesla coil community

About

A pure Python implementation of Java TC or Java TeslaCoil.

Topics

tesla-coil tesla-coil-calculator java-tc

Resources

Readme

License

MIT license
Activity

Stars

2 stars

Watchers

0 watching

Forks

1 fork
Report repository

Releases

1 tags

Packages

 
 
 

Contributors

Languages