Tesla Coil Designer: Resonance, Coupling and Spark Length

Design primary and secondary resonance, find the tuning tap, integrate geometric coupling, and watch energy beat between two coupled circuits. Includes a toroid capacitance and empirical spark-length estimate.

#Freau #Medhurst #Neumann integration #ode45 #resonance #Tesla coil #toroid #Wheeler

This Tesla coil calculator brings together the main choices in a spark-gap Tesla coil design: secondary coil resonant frequency, toroid capacitance, primary tank tuning and magnetic coupling. Engineers, students and experienced makers can compare geometries and see how energy moves between the two resonators after the tank capacitor fires.

Start with the secondary winding diameter, height and turn count. Wheeler's inductance formula and Medhurst's self-capacitance correlation give the bare-coil resonance. Enable a toroid and enter its outside and tube diameters to estimate its isolated capacitance. The total effective capacitance sets the loaded secondary resonance. Add a known capacitance correction when measurements or a separate electrostatic model justify one.

Choose a flat spiral or helical primary, its inner conductor centreline radius, pitch, conductor diameter and elevation relative to the secondary base. Enter the available turn count and tank capacitance. Automatic tuning finds the fractional tap that matches the uncoupled primary and secondary resonances. Manual tuning lets you explore detuning; a handle at the connected tap can be dragged in the drawing. The tuning chart shows the primary frequency across the available winding. The alternative tank-capacitance result answers the reverse question for the currently connected tap.

Coupling is calculated from the coil geometry by numerical integration of the mutual magnetic interaction. Raising the primary changes both the coupling coefficient and the rate of energy transfer. The elevation chart shows this dependence. The two coupled mode frequencies explain why equal uncoupled resonances produce beating rather than one persistent sinusoid.

Set the tank's initial voltage and effective circuit Q values, then play the time control. The shaded perspective drawing shows the coil in proportion. Orange primary glow and violet secondary/toroid glow follow calculated RF envelopes, making energy transfer and decay visible. Three bars show the current energy split and accumulated losses. The waveform chart has separate voltage axes because primary and secondary amplitudes differ greatly. The energy chart includes resistive losses and signed mutual magnetic energy, so their sum can be checked against the initial capacitor energy.

The spark-length result uses John Freau's empirical square-root relationship with real wall-plug watts. It describes an optimistic comparison for a well-adjusted spark-gap coil. It does not follow from the simulated peak voltage. Freau also reports that ordinary arc reach tends to stay within roughly four toroid outside diameters. A warning identifies predictions beyond that empirical applicability range; the estimate is not clamped. The illustrated streamer starts at zero and follows the secondary voltage envelope up to the empirical maximum reach; its growth is illustrative, without simulated discharge-channel dynamics.

Use this Tesla coil design calculator to establish a tuning range and understand coupled-circuit behaviour, then compare the completed coil with measurements. JavaTC provides a more detailed treatment of the electrostatic surroundings and distributed resonator. This tool does not cover nonlinear streamers, quenching, insulation coordination, grounding, protective controls or a DRSSTC bridge. Its spark estimate must not be used to set a safety clearance.

How it is calculated

Secondary and topload. Wheeler's 1928 single-layer formula is LμH=a2N2/(9a+10b)L_{\mu H}=a^2N^2/(9a+10b), with radius aa and winding length bb in inches. Medhurst's 1947 Table V is linearly interpolated for Cself=H(h/D)DcmC_\mathrm{self}=H(h/D)D_\mathrm{cm} pF. The isolated-toroid correlation attributed to Bert Pool is Ct=2.8(1.2781−d/D)0.1217d(D−d)C_t=2.8(1.2781-d/D)\sqrt{0.1217d(D-d)} pF, with outside and tube diameters in cm. This is the form used by Denicolai (2001), printed pp. 25–26. Effective secondary capacitance adds these terms and the user's additional capacitance; fs=1/(2πLsCs)f_s=1/(2\pi\sqrt{L_sC_s}).

Primary and tap. The flat spiral uses Wheeler's LμH=a2N2/(8a+11w)L_{\mu H}=a^2N^2/(8a+11w), with mean radius a=rin+Npr/2a=r_\mathrm{in}+Np_r/2 and radial winding span w=Nprw=Np_r. A helical primary uses the single-layer formula with length NpzNp_z. Pitch is measured centre to centre. Brent root finding solves Lp(N)=LsCs/CpL_p(N)=L_sC_s/C_p for the required tap. Automatic tuning connects that tap when available, otherwise the nearest available endpoint. The alternative capacitor result tunes the presently connected tap.

Geometry-derived coupling. No empirical fixed coupling is assumed. Neumann's line integral for two coaxial loops is reduced to

Mloop=μ0ab2∫02πcos⁡ϕ dϕa2+b2−2abcos⁡ϕ+z2.M_\mathrm{loop}=\frac{\mu_0ab}{2}\int_0^{2\pi}\frac{\cos\phi\,d\phi}{\sqrt{a^2+b^2-2ab\cos\phi+z^2}}.

Integrating the uniform secondary sheet height analytically gives an inverse-hyperbolic-sine difference. Midpoint quadrature integrates the primary's radial or axial turn distribution and the azimuth, using ngeomn_\mathrm{geom} and 4ngeom4n_\mathrm{geom} points. Multiplication by both turn counts gives MM, then k=M/LpLsk=M/\sqrt{L_pL_s}. The elevation sweep evaluates 81 positions; axial reflection symmetry supplies mirrored values. Even angular symmetry halves the work while preserving the full-circle angular spacing.

Coupled transient. Starting with Vp(0)=VchargeV_p(0)=V_\mathrm{charge} and all other voltages and currents zero, ode45 (Dormand–Prince) integrates CpV˙p=−IpC_p\dot V_p=-I_p, CsV˙s=−IsC_s\dot V_s=-I_s, and the two coupled KVL equations LpI˙p+MI˙s=Vp−RpIpL_p\dot I_p+M\dot I_s=V_p-R_pI_p, MI˙p+LsI˙s=Vs−RsIsM\dot I_p+L_s\dot I_s=V_s-R_sI_s. Resistances follow R=ωL/QR=\omega L/Q at each uncoupled resonance. Time and states are scaled for conditioning. A fifth state integrates RpIp2+RsIs2R_pI_p^2+R_sI_s^2; the balance includes signed mutual energy MIpIsMI_pI_s. The two lossless eigenfrequencies give the reference beat-envelope time 1/[2(f+−f−)]1/[2(f_+-f_-)]; the chart and peak use the damped ODE solution.

Spark estimate. Freau's empirical rule is lin=1.7PWl_\mathrm{in}=1.7\sqrt{P_\mathrm{W}} for a well-adjusted spark-gap coil. It is a separate power-based comparison, not an ODE discharge prediction.

Playback. The glow uses energy-equivalent RF envelopes: Ip,env=2Ep/LpI_{p,\mathrm{env}}=\sqrt{2E_p/L_p} and Vs,env=2Es/CsV_{s,\mathrm{env}}=\sqrt{2E_s/C_s}. They combine each resonator's capacitive and inductive self energy to make the slow energy exchange visible without RF flicker. The current itself starts at zero; the primary envelope is already large because the tank is charged. At playback time, illustrative streamer reach is l(t)=lFreaumin⁡[1,Vs,env(t)/Vs,peak]l(t)=l_\mathrm{Freau}\min[1,V_{s,\mathrm{env}}(t)/V_{s,\mathrm{peak}}]. The peak in the denominator is the sampled ODE terminal-voltage peak. This envelope mapping is a visualization convention, not a discharge-growth law.

Primary-to-toroid fit. Because the toroid is axisymmetric, minimum separation is found in a radius–height section. Minimize the distance from the toroid tube centre to the full available primary centreline span: a radial segment for the flat spiral, or an axial segment for the helix. Subtract the toroid tube radius and primary conductor radius. Zero or negative separation rejects contact or intersection, including unused turns beyond the tap.

Assumptions

Coaxial, air-core, single-layer cylindrical secondary; a flat Archimedean spiral or cylindrical helical primary. Each turn is represented magnetically by a circular loop, with continuous uniform turn density and short winding pitch. Primary inductance uses the stated centreline radius and the continuous winding span N × pitch; fractional taps use the same homogenized geometry. Leads and external conductive structures are excluded.

The secondary is one lumped fundamental-mode resonator with uniform current for the magnetic-coupling calculation. The toroid is isolated in free space; its centre height changes the picture, not the capacitance correlation. The additional-capacitance input represents a known positive correction. Q values represent constant effective loss, including the conducting gap in the primary. The gap remains conducting during the modeled ring-down; no charging source drives the transient. Circuit equations and the reduced Neumann integration are derived here from Kirchhoff's laws and magnetostatics.

Where it applies

Use the Medhurst table only for 0.5 ≤ winding height/diameter ≤ 10. Wheeler is most accurate for many closely spaced turns; the helical approximation is about 1% for length > 0.8 × radius, and very short or few-turn primaries need measurement. Current-sheet mutual inductance neglects lead fields, pitch components and the secondary's distributed RF current profile. Large toploads make the latter approximation more reasonable. The published Denicolai comparison also shows that nominal geometry and measured inductance need not coincide.

Adding isolated toroid capacitance to Medhurst capacitance is a first estimate: coil shielding, nearby ground, walls and the toroid's height can shift the effective capacitance and resonance substantially. Denicolai's design calculation predicts 58.2 kHz whereas his completed apparatus measured 65.56 kHz; the calculation does not claim to reproduce that distributed loaded resonance. Measure the completed resonator and tune accordingly. This is not JavaTC's full electrostatic/distributed-coil solver.

The simulated peak is the largest sampled value within the selected time window before nonlinear streamer loading. No corona inception, streamer channel, arc clamp, ground strike, gap quenching, insulation stress distribution, driver switching or higher secondary modes are simulated. The beat time is a lossless envelope reference; it need not equal the time of the sampled damped voltage maximum. Increase the duration for weak coupling and refine the two numerical controls to check the peak and mutual inductance.

Primary glow follows the energy-equivalent RF current envelope; secondary and toroid glow follow the square root of secondary self energy. These envelopes describe the slow beat and ring-down rather than instantaneous RF polarity. The three snapshot bars independently show primary self energy, secondary self energy and dissipated energy as percentages of initial tank energy. Their sum can differ from 100% because signed mutual magnetic energy is shown separately in the full energy chart. The bars share a fixed scale fitted to the whole transient; self energies above initial energy are retained when offset by negative mutual energy. Envelope and bar values are interpolated linearly from the ODE samples. Dense secondary turns are represented by at most 170 winding traces to keep playback responsive; dimensions retain their true proportions. Spark shape is illustrative: its reach starts at zero, follows the secondary voltage envelope, reaches the full Freau estimate near the voltage peak and shrinks during ring-down. Power affects the empirical maximum reach only, not the independent single-firing circuit solution. It is not a clearance, approach distance, air-breakdown voltage law or guaranteed performance. High voltage, stored capacitor energy and RF currents require appropriate engineering controls; this tool does not design those controls.

Empirical spark applicability. Freau's 13 April 2020 reply reports that ordinary arc reach tends to stay within roughly four toroid outside diameters, with exceptions at unusually high power. When the square-root power estimate exceeds that range, a contextual warning marks it as outside the ordinary-coil empirical applicability range. The reported value and maximum illustrated reach retain the unmodified power correlation; four diameters is an approximate observation, not a hard physical cap or a safety clearance. Without a toroid, this diameter-based comparison does not apply. See Freau's reply (https://www.pupman.com/listarchives/2020/Apr/msg00009.php).

History size. The requested history has one initial point plus the ring-down cycles × samples per fastest cycle × highest-mode/secondary frequency ratio, rounded up. Histories above 20,000 points are rejected before ode45 runs; reduce duration or sampling, or use more primary turns. No automatic thinning is applied. Valid histories retain the same adaptive solver and requested sampling; the normalized time and five states are stored as six matrix rows to avoid the library-result display-size limit.

Geometric fit. The primary-to-toroid check includes all available turns and conductor thickness. Positive surface separation means the solid geometries do not intersect; it does not establish an electrical insulation or operating clearance.

References

  • H. A. Wheeler, Simple Inductance Formulas for Radio Coils, Proceedings of the IRE 16(10), 1398–1400 (1928). DOI: https://doi.org/10.1109/JRPROC.1928.221309
  • R. G. Medhurst, H.F. Resistance and Self-Capacitance of Single-Layer Solenoids, Wireless Engineer 24, 35–43 and 80–92 (1947), Table V, p. 84. https://hamwaves.com/inductance/doc/medhurst.1947.pdf
  • Marco Denicolai, Tesla Transformer for Experimentation and Research, Helsinki University of Technology (2001), printed pp. 25–28 and 63–66. https://www.eltem.fi/lthesis.pdf
  • Barton B. Anderson, JavaTC version 13.1 documentation: geometry, coupling and energy transfer. https://www.tesla.nu/programs/javatc/script/k.html
  • Jovan M. Cvetic, Development of the Tesla Coil Apparatus, Sixth International Symposium Nikola Tesla (2006), Tables I–II and bare-coil measurements. https://www.researchgate.net/publication/228591367_Development_of_the_Tesla_Coil_Apparatus
  • John Freau, Tesla Coil Mailing List discussion, 13 April 2020: empirical spark rule, topload size and real watts versus VA. https://www.pupman.com/listarchives/2020/Apr/msg00009.php

Inputs

InputSymbolUnitDefault
Secondary winding diameterD_smm150
Secondary winding heightH_smm600
Secondary turnsN_s1000
Toroid toploadtoploadtrue
Toroid outside diameterD_tmm450
Toroid tube diameterd_tmm100
Toroid centre above winding
Measured from the winding top to the toroid centre; used in the drawing.
h_topmm100
Additional effective capacitance
Add measured lead, surroundings or streamer capacitance beyond the bare coil and isolated toroid.
C_extrapF0
Primary shapeprimary_kindFlat spiral
Primary inner centreline radiusr_pmm105
Radial pitchp_rmm12
Axial pitchp_zmm12
Primary conductor diameterd_pmm6
Primary start above secondary base
Negative values place the first primary turn below the winding base.
z_pmm10
Available primary turnsN_max12
Tank capacitanceC_pnF30
Primary tap selectiontuningTune to secondary resonance
Connected primary turnsN_manual8
Initial tank voltageV_chargekV15
Effective primary Q
Includes coil, capacitor and conducting spark-gap loss at the primary resonance.
Q_p20
Effective secondary Q
A lumped constant-loss Q before nonlinear streamer loading.
Q_s150
Wall-plug real power
Freau's empirical rule uses watts, not transformer VA.
P_inW500
Ring-down duration
Duration in uncoupled secondary oscillation cycles.
cycles24
Mutual-inductance quadrature
Primary quadrature points; the angular integration uses four times as many.
n_geom64
Samples per fastest cycle
Controls waveform sampling and peak detection; ode45 also adapts its internal steps.
n_time80
Time after firingtus17.1

Results

With the default inputs above, the calc gives:

ResultValue
Secondary resonance159.28 kHz
Primary resonance159.28 kHz
Required tuning tap9.07
Connected primary turns9.07
Tuning rangeTuning tap available
Secondary inductance33.177 mH
Medhurst self-capacitance10.8 pF
Isolated toroid capacitance19.3 pF
Total secondary capacitance30.095 pF
Primary inductance33.282 uH
Capacitance for this tap30 nF
Coupling coefficient0.1658
Mutual inductance174.2 uH
Lower coupled resonance147.52 kHz
Upper coupled resonance174.39 kHz
First beat-envelope maximum18.6 us
Simulated peak secondary voltage368.8 kV
Simulated peak primary current434.1 A
Energy per firing3.375 J
Freau empirical spark length0.97 m

Checked against sources

Each case below was calculated by this calc and agrees with the value from its source (within the tolerance where one is shown).

Denicolai worked secondary and toroid

Source: Marco Denicolai, Tesla Transformer for Experimentation and Research (2001), printed pp. 25–26, PDF pp. 39–40, equations (4-2)–(4-8). Published worked values: 88.2 mH, 27.8 pF, 57 pF, 101.6 kHz bare, 58.2 kHz loaded. https://www.eltem.fi/lthesis.pdf

Inputs: Secondary winding diameter = 400 mm; Secondary winding height = 1,500 mm; Secondary turns = 970; Toroid outside diameter = 1,500 mm; Toroid tube diameter = 200 mm; Toroid centre above winding = 180 mm; Primary inner centreline radius = 279 mm; Radial pitch = 30 mm; Primary conductor diameter = 8 mm; Tank capacitance = 100 nF; the others as above.

ResultSourceThis calc
Secondary inductance88.2 ± 0.1 mH88.198 mH
Medhurst self-capacitance27.8 ± 0.05 pF27.8 pF
Isolated toroid capacitance57 ± 0.2 pF57.017 pF
Bare-coil resonant frequency101.6 ± 0.2 kHz101.6 kHz
Secondary resonance58.2 ± 0.2 kHz58.19 kHz

Cvetic measured bare secondary

Source: J. M. Cvetic, Development of the Tesla Coil Apparatus, Sixth International Symposium Nikola Tesla (2006), p. 2, Tables I–II and measured bare resonance. D = 25 cm, H = 96.5 cm, N = 887; measured L = 44.5 mH, f = 177 kHz. Wheeler predicts 44.918 mH and linearly interpolated Medhurst gives 17.65 pF, rather than the paper's rounded 17.8 pF; 1–2% frequency tolerance reflects that and measurement. Author manuscript: https://www.researchgate.net/publication/228591367_Development_of_the_Tesla_Coil_Apparatus

Inputs: Secondary winding diameter = 250 mm; Secondary winding height = 965 mm; Secondary turns = 887; Toroid topload = no; Primary inner centreline radius = 160 mm; Primary start above secondary base = 30 mm; Tank capacitance = 93.7 nF; the others as above.

ResultSourceThis calc
Secondary inductance44.5 ± 1 mH44.918 mH
Bare-coil resonant frequency177 ± 3 kHz178.7 kHz

Denicolai measured mutual coupling

Source: Denicolai (2001), printed p. 65, PDF p. 79, Table 7: 9.5-turn flat primary, elevation 10 mm, M = 477.55 µH and k = 0.1694. Geometry from chapter 4: secondary 400 × 1500 mm, nominal 970 turns, primary inner clear diameter 550 mm, 8 mm pipe, 22 mm edge spacing (30 mm pitch). The first conductor centreline radius is (550 + 8)/2 = 279 mm, from printed pp. 27–28 (PDF pp. 41–42). The measured secondary inductance is 80.22 mH versus Wheeler's nominal 88.20 mH; this comparison allows 5% on M and 0.012 on k for the current-sheet and nominal-geometry approximations; it is not a fitted calibration. https://www.eltem.fi/lthesis.pdf

Inputs: Secondary winding diameter = 400 mm; Secondary winding height = 1,500 mm; Secondary turns = 970; Toroid outside diameter = 1,500 mm; Toroid tube diameter = 200 mm; Toroid centre above winding = 180 mm; Primary inner centreline radius = 279 mm; Radial pitch = 30 mm; Primary conductor diameter = 8 mm; Available primary turns = 9.5; Tank capacitance = 100 nF; Primary tap selection = Use manual tap; Connected primary turns = 9.5; the others as above.

ResultSourceThis calc
Mutual inductance477.5 ± 25 uH480 uH
Coupling coefficient0.1694 ± 0.0120.1641

Related Resources

Responses

Your email address will not be published. Required fields are marked *

L ading...