Smith-Don: Primer

Don (Donald L.) Smith's resonant air-core coil and capacitor "free energy" devices, his "cycles per second squared" energy formula, and Joel Lagace's later capacitor reconstruction: how resonance really stores and moves energy, and how to audit an energy formula with units and a ledger.

Source tags: [read] = from a document in the project folder (or a copy saved by the Investigator in research/files/); [web] = Internet source, URL given; [memory] = recalled from training, with a confidence level; [computed] = arithmetic done for this primer and reproduced by example.py. Verdicts come from arbiter.md; this primer adds none.

Source gap (repeated from claims.md and arbiter.md). The folder holds no measurement, test report, witness statement, instrument reading or patent grant. The only evidence that hardware existed is seven uncaptioned colour photographs in Smith's binder (pp.3-9: coils, high-voltage modules, capacitors, a plasma globe, an oscilloscope and meters) and Smith's own statements [read]. The second document, Understanding The Don Smith Capacitor Energy Method by Joel Lagace (don-smith.pdf), is a secondary source: a later author's reconstruction, with no measurement [read]. Every number in both documents is an illustrative calculation, not data. No conclusion in this primer about a Smith device rests on a measurement of one, because none is on record in the files. The deciding first-hand records were not reviewed by anyone: the recordings of Smith's lectures (1994-2005), the 1996 workshop tape (https://youtu.be/INPIfDGfMGs), the uncut 2001 tape, about 170 pages of Smith originals held by Rick Friedrich, and Lagace's 2026 videos [web: see sites.md]. QE revision (2026-10-08): binder pp.10-35, 73 and 76-79 have now been deep-read (they earlier rested on page-by-page summaries); the full 1996 ExtraOrdinary Science article, the Kelly part of the 119-page archive.org compilation and the Smith sections of Utkin's book have still been read only in part, and no verdict rests on them.

---

1. The idea in one paragraph

Don Smith, a Texas "energy consultant" who lectured at Tesla Society and TeslaTech conventions from 1994 to 2005, claimed that small resonant air-core coils (Tesla-coil style), driven at radio frequency from a battery and a high-voltage module and connected to earth and air grounds, deliver far more electrical energy than they take in. His explanation has three parts. First, he says the energy a resonant system "pumps" grows with the square of the frequency, so a megahertz device moves vastly more energy than a 60 Hz one. Second, he says that at resonance the circuit's impedance (its opposition to alternating current) becomes zero, so the coil behaves like a room-temperature superconductor. Third, he says the extra energy is not created but drawn from the "ambient" electrical background of the Earth and the air (about 200,000 volts, he writes), so the usual conservation bookkeeping does not apply to such "open" systems [read: binder pp.37-49, 54-75]. His one quantitative performance claim, made in his own account about five years later, is a 1996 "suitcase" demonstration with 12 volts at 0.3 ampere in and "a continuous 120 V at 80 Amperes" out [web: energyevo.org compilation, below]. He also described his devices at very large scale: "A megawatt sized unit will sit comfortably on a breakfast table" (binder pp.45, 49), a "Domestic Use Range Module, up to two million volt-amperes-reactive output" (pp.64-65), small units "presently available as replacements for the batteries used in electric automobiles" (p.62), and, in an unsigned passage about Smith (third person) in Patrick Kelly's compilation, "Commercial units of Boulder Dam size ... currently being installed at several major locations throughout the world" [read: binder; research/files/Resonance-Energy-Methods-119p_djvu.txt lines 9239-9241]. No unit, customer or site has been found (section 6). Underneath all of this lies his own physical model: electrons as spinning "doublets", one spin the "source of volts" and the other the "source of amperes", with magnetism an "open system" that "replicates multiple copies of the energy present" [read: binder pp.25, 37-42, via deep read]. Smith's starting points are real physics: resonance really does build up large voltages and currents from a small drive, Tesla coils really do step a few kilovolts up to hundreds, and the Earth really does sit inside an electric field with a few hundred kilovolts between ground and the upper atmosphere. The question this primer teaches you to answer is where the energy in such a circuit comes from and goes to, and how to check that from the formulas alone.

---

2. What you need to know first

Energy and power. Energy (joules, J) is a quantity you can store; power (watts, W = J/s) is a rate at which energy moves. A 1 J store emptied once per second is 1 W; emptied a million times per second it is 1 MW, but only if it is refilled each time, and the refill must come from somewhere. Almost every error in this project is a confusion of the two. Refresher text: any first-year physics book, e.g. Halliday, Resnick and Walker, Fundamentals of Physics, the chapters on work, energy and power.

Capacitors and inductors. A capacitor (capacitance C, in farads, F) stores energy in its electric field: U = (1/2) C V^2, where V is the voltage across it. An inductor (inductance L, in henrys, H) stores energy in its magnetic field: U = (1/2) L I^2, where I is the current through it. Units you will need: 1 F = 1 A s / V (coulomb per volt) and 1 H = 1 V s / A. Text: D. J. Griffiths, Introduction to Electrodynamics, the chapters on electrostatic energy and on electrodynamics (inductance and magnetic energy).

Alternating current, phasors and impedance. For a sine wave at frequency f (hertz, Hz = cycles per second) we use the angular frequency omega = 2 pi f (radians per second). A resistor R opposes current equally at all frequencies; an inductor's opposition (reactance) is X_L = omega L, a capacitor's is X_C = 1/(omega C). Written with complex numbers (j = sqrt(-1), the engineering symbol for i), the series impedance is Z = R + j(omega L - 1/(omega C)), in ohms. The real part dissipates energy as heat; the imaginary (reactive) part stores and returns it. Text: P. Horowitz and W. Hill, The Art of Electronics, chapter 1; or any circuits text (e.g. Nilsson and Riedel, Electric Circuits), the chapters on sinusoidal steady state and AC power.

Real, reactive and apparent power. With rms voltage V and rms current I differing in phase by an angle phi: real power P = V I cos(phi) (watts, actually delivered), reactive power Q_r = V I sin(phi) (volt-amperes reactive, VAR, sloshes back and forth, averages to zero), apparent power S = V I (volt-amperes, VA), with S^2 = P^2 + Q_r^2 (the "power triangle"). An ammeter and a voltmeter multiplied together give S, not P. ("rms", root-mean-square, is the effective value of a wave: for a sine wave, peak divided by sqrt(2).)

Resonance and the quality factor Q. In a series L-C-R circuit, the two reactances cancel at the resonant frequency f0 = 1/(2 pi sqrt(L C)). The circuit then rings with energy passing back and forth between L and C. The quality factor Q = omega0 L / R (omega0 = 2 pi f0) measures how slowly that ringing dies away: roughly, the stored energy is lost in Q/(2 pi) cycles. Driven at resonance, the voltage across C is Q times the drive voltage. (Do not confuse this Q with reactive power Q_r or with charge q.) Text: Horowitz and Hill, chapter 1; Griffiths does not cover resonance in depth; the ODE (ordinary differential equation) side is the damped driven oscillator in any classical mechanics text (e.g. Taylor, Classical Mechanics, the chapter on oscillations).

Transformers and Tesla coils. Faraday's law: a coil of N turns in a changing magnetic flux Phi has EMF (electromotive force, a voltage) V = N dPhi/dt. Two windings sharing the same flux therefore have the same volts per turn, so voltage scales with turns and current scales inversely: power in = power out (minus losses). A Tesla coil is two loosely coupled resonant circuits; energy rings over from the primary to the secondary, and because the secondary has small capacitance its voltage becomes very high. Text: Griffiths, the electrodynamics chapter (Faraday's law, mutual inductance); for Tesla coils, R. Burnett's notes [web: https://richieburnett.co.uk/operation.html].

Poynting's theorem (energy conservation for fields). The energy inside any region can only change by energy flowing across its boundary or by work done on charges inside. An "open system" is simply one where you must count the flows across the boundary; it is not exempt from the law. Text: Griffiths, the chapter "Conservation Laws".

The global atmospheric electric circuit. Thunderstorms worldwide charge the upper atmosphere (ionosphere) to a few hundred kilovolts above the ground. In fair weather a tiny current (picoamperes per square metre) leaks back down through the air, and near the ground the field is about 100-150 volts per metre. Text: R. P. Feynman, The Feynman Lectures on Physics, Vol. II, chapter 9, "Electricity in the Atmosphere" [memory, high for the chapter].

---

3. The textbook view

3.1 Stored energy is not delivered energy

A capacitor C at voltage V stores

U = (1/2) C V^2        (U in J, C in F, V in V)

If you discharge it into a load and recharge it f times per second, the average power through the load is P = U f (watts), but the recharging source must supply that same P. Multiplying a stored energy by a frequency gives a rate, and only if a real dump-and-refill happens.

3.2 A capacitor on AC: reactive power

Put a capacitor across a sine-wave source of rms voltage V. The current is i = C dv/dt; it leads the voltage by 90 degrees, so cos(phi) = 0 and the real power is zero. The rms current is I = omega C V and the reactive power is

Q_r = V I = omega C V^2 = 2 pi f C V^2 = 4 pi x [(1/2) C V^2 f]

(for V in rms; with V as a peak value the factor is 2 pi instead of 4 pi). So the quantity (1/2) C V^2 f, up to a factor 4 pi, is the volt-amperes that flow into the capacitor and straight back out every half cycle. It is real current and can heat wires, but it delivers no net energy [memory, high; checked numerically in example.py step 3].

3.3 Series resonance

For a series R-L-C circuit driven by v_s = V0 sin(omega t):

Z(omega) = R + j(omega L - 1/(omega C)),    |Z| is smallest at omega0 = 1/sqrt(L C), where Z = R.

At resonance the current amplitude is I0 = V0 / R, the capacitor voltage amplitude is Q V0 with Q = omega0 L / R, and the average power from the source is

P = V0^2 / (2R) = omega0 W / Q,

where W = (1/2) C (Q V0)^2 is the peak energy held in the tank. Read the last relation carefully: to keep energy W ringing at frequency f0 you must supply power 2 pi f0 W / Q. Higher frequency at the same stored energy costs more input power, not less. Resonance magnifies voltage and current; the source pays for every joule the resistance turns into heat [memory, high; simulated in example.py step 4].

A parallel L-C circuit at resonance has the opposite behaviour: its impedance is a maximum (about Q^2 R), not a minimum [memory, high].

Skin effect. At radio frequency the current crowds into a thin surface layer of depth delta = sqrt(2 rho / (omega mu)) (rho = resistivity, mu = permeability), about 65 micrometres in copper at 1 MHz [web: https://en.wikipedia.org/wiki/Skin_effect, as checked by the mathematician]. Resistance therefore rises with frequency, roughly as sqrt(f).

3.4 Transformers and Tesla coils

With tight coupling, V2/V1 = N2/N1 and I2/I1 = N1/N2, so V2 I2 = V1 I1 at best. A Tesla coil is loosely coupled (coupling coefficient k about 0.1-0.2); its voltage gain in the ideal lossless tuned case is about sqrt(L2/L1) (equivalently sqrt(C1/C2)), set by energy conservation: the primary's (1/2) C1 V1^2 reappears as the secondary's (1/2) C2 V2^2, and good coils move over 85% of it [web: https://richieburnett.co.uk/operation.html]. Large voltage, small current, energy conserved.

3.5 Two capacitors sharing charge (needed for Lagace's scheme)

Connect a charged capacitor C at V0 to an identical empty one through any resistance. Charge is conserved, so both end at V0/2. Energy before: (1/2) C V0^2. After: 2 x (1/2) C (V0/2)^2 = (1/4) C V0^2. Exactly half is lost as heat (or radiation), whatever the resistance [memory, high]. Put an inductor in the path (with a diode or switch to stop the swing at the right moment) and the charge oscillates across instead; almost all the energy can then be moved, limited only by resistance. This is the principle of switched-mode converters, routinely 90-95% efficient [memory, high; not web-checked]. The ceiling is 100% of the energy you started with.

3.6 Where energy really can be drawn from the air

The ionosphere-to-ground potential is roughly 150-400 kV, typically 250-300 kV; Wikipedia gives "about 400,000 volts"; the fair-weather current density is about 2 pA/m^2 and the surface field about 120 V/m; the whole fair-weather circuit carries roughly 1-2 kA [web: https://en.wikipedia.org/wiki/Atmospheric_electricity ; https://arxiv.org/pdf/physics/0506077]. That is a planet-wide power of roughly 0.35-0.45 GW, and per square metre between about 2.4x10^-10 W (the local J x E at the surface) and 8x10^-7 W (the whole circuit divided by area, a generous ceiling) [computed in arbiter.md]. Real collectors (Franklin's bells, Kelvin's water dropper, long antennas) work at nanowatts to microwatts. Ambient radio waves can likewise power a crystal radio, at microwatts.

---

4. Where the inventor departs

Smith's text starts from correct ingredients and then changes them at specific points. Each row gives his statement [read, binder page] beside the textbook statement.

The single point of departure. Every one of these rows comes back to one move: treating a quantity that circulates (stored energy counted each cycle, reactive volt-amperes, resonantly magnified voltage) as if it were energy delivered. Resonance is a store, not a source. Smith reads the large numbers that a resonant circuit really does show (high voltage, large circulating current, many joules passing per second through the reactive parts) as net output.

Lagace's departure (claim 6). Lagace correctly states the half-energy loss of charge sharing (section 3.5) and draws a circuit (his Figure D) that routes the discharge through a transformer primary, which is an ordinary inductive recovery path [read: deep/don-smith-43e195.md]. He then departs at one sentence: the primary "does not feel the secondary load because there is no closed current loop". In his own figure the primary is in the series loop C1-T1-C2-SW2, and inside a capacitor the displacement current (the term epsilon0 dE/dt that Maxwell added to Ampere's law) exactly continues the conduction current in the leads, so the loop is closed for AC purposes and a loaded secondary reflects back into the primary. Lagace himself writes elsewhere that "the primary must supply all the energy transferred to the secondary" (pp.17, 38) [read].

---

5. Worked example: auditing "E = (1/2) C V^2 x f^2" with units and an energy ledger

All numbers below are reproduced by example.py:

.venv/bin/python review/Smith-Don/example.py

The example uses Smith's own printed examples A and B and then a real resonant circuit, so you can see both what his formula computes and where the energy of a resonant circuit actually goes.

Step 1. Check the units before any arithmetic

• (1/2) C V^2: F x V^2 = (A s / V) x V^2 = V A s = J. An energy.
• (1/2) C V^2 x f: J x (1/s) = J/s = W. A power.
• (1/2) C V^2 x f^2: J/s^2. Neither.

So the p.47 formula cannot give "Joules [Watt Seconds]", as labelled, whatever numbers go in. (SymPy confirms this in step 1 of the script, and confirms that 1 H = V s/A and 1 F = A s/V, so neither "equals one volt" as p.44 says.)

Step 2. Recompute Smith's examples three ways

Example A (p.48): C = 10 uF = 1x10^-5 F, V = 120 V, f = 60 Hz.

• Stored: (1/2)(1x10^-5)(120)^2 = 0.072 J.
• x f: 0.072 x 60 = 4.32 W.
• x f^2: 0.072 x 3,600 = 259.2 J/s^2.

Smith prints "259.2 Joules", so example A uses f^2. He also says "the Establishment's method only permits less than 10 Watt Seconds": the f form gives 4.32, which is "less than 10", but in watts [read: p.48].

Example B (p.49): C = 10 uF, V = 100,000 V, f = 1 MHz.

• Stored: (1/2)(1x10^-5)(10^10) = 50,000 J. (Smith's own intermediate line agrees.)
• x f: 5x10^10 W.
• x f^2: 5x10^16 J/s^2.

The binder leaves the final line "[50,000 x One Million squared = ]" blank and states "greater than 50 Mega K-Joules" = 50x10^6x10^3 = 5x10^10. That is the f form, not the f^2 formula printed two lines above [read: page image p.49]. The 1996 magazine printing of the same example says "= 500 x One Million squared" and "more than 500 MegaJoules" = 5x10^8; its intermediate "500" should be 50,000, and 500 x (10^6)^2 would be 5x10^14 [read: page image, magazine p.37]. So four different numbers are printed or implied for one example: 5x10^8, 5x10^10, 5x10^14 and 5x10^16. Finally, p.56 says a system "cycling at 60 million times per second produces one million times the energy" of a 60 Hz system: 60 MHz / 60 Hz = 10^6 is the first-power ratio; f^2 would give 10^12 [read: p.56].

What this step shows: the document does not apply its own law consistently. That is the arbiter's "Internally inconsistent".

Step 3. What (1/2) C V^2 f really is

Take example A on a 120 V rms, 60 Hz supply. The capacitor's reactance is X_C = 1/(2 pi x 60 x 1x10^-5) = 265.3 ohm, so the rms current is 120/265.3 = 0.452 A and the reactive power is 120 x 0.452 = 54.3 VAR. Divide by (1/2) C V^2 f = 4.32: the ratio is 12.566 = 4 pi. The same ratio holds for example B (Q_r = 6.3x10^11 VAR). So Smith's "energy pumped" quantity is, to a factor 4 pi, the reactive power. The script then integrates v(t) x i(t) over one cycle for example A: the peak is 54.3 W, the average is zero (about 10^-13 W, rounding error). The energy goes into the capacitor and comes back out each half cycle.

For example B, the reactance at 1 MHz is 0.0159 ohm; at 100 kV that means about 6.3 million amperes, and resonating 10 uF at 1 MHz needs L = 2.53 nH, a few millimetres of wire. No such device is described.

Step 4. A real resonant circuit, with the energy ledger

Now take a circuit you could build: L = 2.53 uH, C = 10 nF, R = 1 ohm (the coil's and wiring's loss), driven by a 10 V (peak) sine wave at resonance. (This is the mathematician's illustration, re-derived here.)

• f0 = 1/(2 pi sqrt(2.53x10^-6 x 1x10^-8)) = 1.0006 MHz.
• Q = omega0 L / R = 2 pi x 1.0006x10^6 x 2.53x10^-6 / 1 = 15.9.
• At resonance Z = R = 1 ohm, so I0 = 10 A peak and the capacitor voltage is Q x 10 = 159 V peak: resonance has magnified the voltage 16 times, which is the real effect Smith built on.
• Average power from the source: V0^2/(2R) = 100/2 = 50 W. All 50 W end up as heat in R.

The script solves the circuit's differential equation, L di/dt = v_s - R i - q/C with dq/dt = i, from rest for 80 cycles (fourth-order Runge-Kutta, 400 steps per cycle) and keeps three running totals:

The ledger closes. In steady state the simulation gives 10.00 A peak, 159.1 V across C (gain 15.9 = Q) and 50.00 W average from the source; the textbook relation P = omega0 W / Q gives 50.00 W with W = 126.5 uJ. A sweep of |Z| from 0.5 f0 to 1.5 f0 has its minimum, 1.0000 ohm, exactly at f0: equal to R, not zero.

Now apply Smith's p.47 formula to this tank: (1/2) C V_C^2 f0^2 = 126.5x10^-6 x (1.0006x10^6)^2 = 1.27x10^8 "joules". The real circuit takes 50 W in and gives 50 W of heat out. The formula has no connection to the circuit's energy flows.

Step 5. What example B would cost to run

To keep example B's 50 kJ ringing at 1 MHz against losses: P = 2 pi f W / Q = 2 pi x 10^6 x 5x10^4 / Q = 3.1x10^11 / Q watts. For Q = 100 that is 3.1 GW of input; for Q = 1,000, 0.31 GW. A higher frequency at the same stored energy makes the device more expensive to run, the opposite of Smith's reading.

Step 6. Scale check of the 1996 suitcase claim

Smith's account [web: https://energyevo.org/2015/01/18/updated-compilation-of-important-info-don-smith-shared-to-nuenergy-yahoo-group/]: "Input was 12 V DC at 300 milliamperes and the Output measured by an Electrical Engineer, picked by the audience, was a continuous 120 V at 80 Amperes as Documented in the VCR Tape." That is 3.6 W in and 9,600 VA out, a ratio of about 2,700. Three checks, none of which decides what happened on the day:

• VA or W? 120 V x 80 A is apparent power. It equals real power only if the load was resistive and the waveform clean; for a reactive or pulsed load it can be much larger than the real power (section 2). The account does not say which was measured, nor the meters, the load or the duration.
• Battery. The suitcase schematic reproduced in Kelly's compilation names "Gelcell, 12 Volt, 7 Amp Hour" [read: research/files/Resonance-Energy-Methods-119p_djvu.txt, Kelly p.3-11]. That is 12 V x 7 Ah = 84 Wh, which would supply 9.6 kW for about 32 seconds. The same schematic rates the output transformer at "480 V @ 60 Amps" (28.8 kVA) and the capacitor bank at 8,000 µF for 480 V DC, which stores (1/2)(0.008)(480)^2 = 922 J, about 0.1 s at 9.6 kW. It prints no measured input or output. "Continuous" output therefore needs a source other than the battery.
• Atmosphere. At the generous ceiling of 8x10^-7 W/m^2, 9.6 kW needs about 1.2x10^10 m^2, some 12,000 km^2 of the whole global circuit; at the local 2.4x10^-10 W/m^2, about 4x10^13 m^2, around a tenth of the Earth's entire surface (5.1x10^14 m^2). The whole fair-weather circuit carries only about 350 MW.

What the example teaches: the ambient source Smith names cannot supply kilowatts (arbiter: "Refuted by calculation" for that budget). It does not show what was measured in 1996. If a real 9.6 kW was measured, it came from somewhere else, and that is a question for the tape and a bench test (arbiter: "Never properly tested").

---

6. Evidence table

Verdicts are the arbiter's (Step 2 scale). "Tested?" means a neutral test with measured input and output.

The record that goes silent. After the 1996 suitcase claim no neutral instrumented test appears; the one announced for 2002 did not happen; Smith lectured until 2005, and a single forum snippet gives his death as 22 November 2010 (no obituary found) [web: sites.md]. The arbiter sets out competing explanations without deciding between them: illness (a documented 2002 stroke, though he lectured three years after it); no working device (the arithmetic errors, the apparent-power ambiguity, the 1994 partial failure); commercial or contractual secrecy ("Don's personal security and contract obligations", in an unsigned passage about Smith in Kelly's compilation; no contract or licensee identified; no matching Texas "TransWorld Energy" entity found); suppression (Smith's and Patrick Kelly's reading; Kelly says Smith's web site was "taken over by Big Oil", his characterization, with no evidence given, a lapsed domain being the other explanation); absorption into another entity (no evidence either way). The organiser of the 2001 event, Perreault (interested party: organiser and video seller), says he cut about five minutes in which Monroe Fischer questioned Smith, to protect a speaker weakened by a stroke; a critic site called the event report misleading. The uncut tape would decide whether that was kindness or the removal of an unfavourable exchange [read: arbiter.md, absence-of-evidence analysis].

The commercial silence (arbiter, QE revision, addition E). Smith's own statements say units were "presently available" and "being introduced into the World Market", and an unsigned passage about him in Kelly's compilation says units were "currently being installed at several major locations"; the same binder gives R&D "for more than 15 years" (cover letter, 23 September 2002) and "the past seven years" (p.22, undated). Units on that scale would leave records, and none has been found. The arbiter's competing explanations, none decided: promotional overstatement by an interested party (no customer, price, specification sheet or compliance mark in the binder); contractual or security secrecy (that passage's "personal security and contract obligations"; but utility-scale installations are hard to hide from interconnection and permit records); small demonstration units described in grand terms (VAR counted as power, as in the suitcase's VA); sale under another name (no corporate trail found); takeover or suppression (Kelly's "Big Oil", no actor named). Deciding records: sales, invoices, customer testimony, utility interconnection, permit and customs records, assumed-name filings in Houston and Spring, Texas, or a surviving unit.

Smith's place among related work, and other claimed replications. These are kept in the record whatever the verdicts; none gives measured input and output energy. - The Bearden line. In a post placed just before an e-mail of 9 November 2001, Smith wrote of the suitcase device: "This is a close relative of Tom Bearden's Device. However I used a different Coupling System"; in that e-mail he cites the Canadian patent document CIPO 2172240 ("... a generator with no moving parts for direct conversion of magnetostatic energy ..."), which the Investigator identifies as Bearden-related [web: https://energyevo.org/2015/01/18/updated-compilation-of-important-info-don-smith-shared-to-nuenergy-yahoo-group/; research/2026-10-08-energyevo-nuenergy.md]. That places Smith beside the motionless electromagnetic generator (MEG) claims; no reviewer analysed the Canadian document. - "Bird on the wire". Vladimir Utkin (c.2012) writes that "Don Smith named this technology 'Bird on the wire'", but gives no source for it. Utkin also describes an arrangement in which a capacitor is "charged by sparks and powerful displacement current is produced", collected by a ferromagnetic-core transformer. That sentence is Utkin's caption for a 2006 rendering titled "Don Smith Device" (credited to Zero Point Films: spark gap, copper and aluminium plates either side of glass, a "MetGlas Cobalt Transformer"), whose own caption speaks of "radiating electrons", not displacement current [read: research/files/Utkin-Free-Energy-Tesla-Secrets.pdf pp.13, 16, page images]. Smith's binder uses the same image twice: "the Bird on the High Voltage Line" (p.30) and "the bird sitting on the million volt power line and sensing nothing" (p.35, his explanation of why instruments referenced to the ambient show no potential) [read]. - Patrick Kelly (compiler; runs a free-energy site): Smith produced "at least forty eight different devices ... in most cases they do not require any input energy" (Kelly's claim); Smith "never did disclose the full details of his designs" but left enough "for somebody who is experienced in radio-frequency electronics" to deduce the rest; and, against the stated currents, a reader's observation that an output choke in one photograph "appears to be wound with wire that is far too small diameter" [read: research/2026-10-08-kelly-free-energy-info.md]. Partial disclosure is one reason replicators disagree about what Smith built. - Claimed replicators. tswift (row 3e) names Marc Belanger, who "claimed to have replicated it with a 60 Hz NST [neon-sign transformer] and made a video", and "Zilano". Energeticforum thread 13434 (2015) reports weighing-scale experiments on capacitor banks (the poster notes possible interference with a kitchen scale) and "dielectric pulsation" seen on a scope near the spark gap [web: https://www.energeticforum.com/forum/energetic-forum-discussion/renewable-energy/13434-don-smith-device-information-and-tips]. J. E. Dalton II, "Radiant Energy Induction Using RF Pulsed Plasma" (2014), discusses Smith's system [web: http://gratisenergi.se/rfplasma, found by search, not opened]. - Influence. An energeticforum thread on Smith's devices opened in 2009 and runs past page 790; the binder circulates in Russian translations and in Spanish translations of Kelly's chapter [web: research/2026-10-08-energeticforum.md, research/2026-10-08-google-search-records.md]. The cover letter's ">40,000 copies" in seven editions and many languages is Smith's own figure, not independently confirmed. Smith's 2 November 1995 "Tesla Coil Geometry" sheet (binder p.42) is the earliest dated form of his two-ended "dipole" coil in the files.

Patent status, stated carefully. Smith's binder cites "Patent Pending 08/100,074" with a date of 4 February 1992; the serial prefix 08 was assigned from 1 January 1993, so the number indicates a 1993 filing; the date is unresolved without the file wrapper [web: USPTO series-code table, via arbiter]. Perreault (2002) relayed that Smith's Mexican patents "have been approved" and Kelly says "granted May 2004"; the located record is an application, later discontinued. Neither statement is confirmed; neither is shown false beyond the application status.

---

7. What a learner should take away

Solid (and worth learning from Smith's material). - Resonance really does magnify voltage and current: the worked circuit turns a 10 V drive into 159 V across the capacitor. Smith's photographs show he built and ran real high-voltage RF hardware with instruments [read: pp.3-9]. - The Builder's Guide (binder pp.76-79) is a voltage-gain recipe for a resonant air-core step-up coil (quarter-wave wire-length rule, volts per turn, a neon bulb to find the high-voltage node), with no gain claim, but it is not a complete, checked recipe [read: deep read pp.73, 76-79; arbiter, QE revision]: the "Chart A" it relies on is missing; the full-wave constant 998 should be 988 (4 x 247); step 4 computes 26.67 turns and then selects 30 (720 V, unstated); 10 ft of wire on a 2-inch form is about 19 turns, so its turn count and its 24.7 MHz cannot both hold; and it gives no wire gauge, capacitor value, spark-gap width, module current or instruction to measure power. Its p.79 figure is an iron-core, oil-filled impulse transformer, not an air-core coil. It tells the builder to probe the live coil with a hand as ground and gives no safety warning: do not do that. Built with proper values and with input and output power metered, it can show voltage gain with current loss; as written it cannot test over-unity. - Smith's ambient 200 kV is the right size for the real ionosphere-to-ground potential (150-400 kV). - Lagace is right that direct charge sharing wastes half the energy and right that an inductive path can recover most of it. - Smith's "x C.P.S." quantity is a true expression for something real: reactive power, to a factor 4 pi.

Wrong, and why the mistake is instructive. - Multiplying a stored energy by f (or f^2) and calling the result energy delivered. The cure is the units check in step 1, which takes a minute and catches the error before any arithmetic. - Reading reactive volt-amperes, or a meter's V x A on a reactive load, as real power. This is the commonest way an honest experimenter overestimates the output of a resonant or pulsed circuit; true-RMS power meters (which multiply v and i point by point and average) and calorimetry (measuring heat in a load) exist for exactly this reason. - "Z = 0 at resonance". The reactive parts cancel; the resistance stays. Q tells you how much the source must pay to keep the ring going. - "Open systems are exempt from conservation". They are not; they require you to meter what crosses the boundary. The atmospheric budget shows how to do that with real numbers. - "No closed loop, so the primary does not feel the load" (Lagace). Maxwell's displacement current exists precisely so that current is continuous through a capacitor.

Genuinely open (records, not physics). - What was measured at the 18 July 1996 workshop, with which meters, into which load, for how long. The tape exists and has not been reviewed. - What Smith's roughly 170 further pages and the "CAPACITOR FUNCTIONS" paper (14 February 2004) contain, and whether any of it is measured data. - The fate of US serial 08/100,074 and why the Mexican application was discontinued. - Whether the "self-charging battery" seen by Smith, tswift and others is the module's own rectified output or a meter artefact (testable on the bench by logging battery charge against module input power). - What Lagace's 2026 "modified replication" measures. - What Smith meant by the "Heaviside current component": used three times in the pp.11-19 specification, drawn as dashed lines crossing the plates, defined nowhere; the Mexican translation renders it "Campo pesado" ("heavy field") [read: deep read pp.10-19]. - Whether any of the units Smith said were available or being installed (section 6, row 10) ever existed; the records that would show it (sales, customs, utility interconnection, assumed-name filings) have not been searched. - How Smith's device relates to the Bearden-line Canadian patent document CIPO 2172240 he cited, which no reviewer has read.

---

8. Check yourself

1. A 2 uF capacitor is charged to 500 V. (a) How much energy does it store? (b) Smith's p.47 formula is applied at 100 kHz. What number results, and what are its units? (c) If the capacitor is discharged into a load and recharged 100,000 times per second, what average power flows, and what must the charger supply?
2. A series R-L-C circuit has R = 2 ohm, L = 10 uH, C = 1 nF and is driven at resonance with 5 V peak. Find f0, Q, the peak voltage across C, and the average power drawn from the source.
3. A meter on a resonant output reads 120 V rms and a clamp ammeter reads 80 A rms. The load is a capacitor bank in parallel with a 10 ohm resistor. How much real power is delivered?
4. Two 100 uF capacitors, one at 200 V and one at 0 V, are joined by a wire. What is the final voltage, and how much energy is lost? What changes if an inductor and a diode are put in the path?
5. Why does a higher resonant frequency make it more expensive, not less, to keep the same stored energy ringing?

---

9. Further reading

Mainstream. - D. J. Griffiths, Introduction to Electrodynamics (Pearson / Cambridge University Press): the chapters on electrodynamics (Faraday's law, inductance, displacement current) and "Conservation Laws" (Poynting's theorem). - P. Horowitz and W. Hill, The Art of Electronics (Cambridge University Press), chapter 1: reactance, resonance, power in reactive circuits. - R. P. Feynman, R. B. Leighton and M. Sands, The Feynman Lectures on Physics, Vol. II, chapter 9, "Electricity in the Atmosphere" (free online at https://www.feynmanlectures.caltech.edu/). - Atmospheric electricity summary [web: https://en.wikipedia.org/wiki/Atmospheric_electricity]; global-circuit potential and current estimates [web: https://arxiv.org/pdf/physics/0506077]. - R. Burnett, Tesla coil operation notes [web: https://richieburnett.co.uk/operation.html]. - A. Kurs et al., "Wireless power transfer via strongly coupled magnetic resonances", Science 317 (2007) 83-86 [web: https://pubmed.ncbi.nlm.nih.gov/17556549/]: what resonant coupling really achieves.

Smith's own works and the primary record. - Don L. Smith, Resonance Energy Methods binder (96 pp., material dated 1992-2003, cover letter 23 September 2002): the "Electrical Energy System" essay, "Induced Electrical Energy System", the 1994 International Tesla Society speech, the Builder's Guide [read: project folder]. - Don L. Smith, "Electrical Energy System", ExtraOrdinary Science, Apr/May/Jun 1996 (Vol. VIII no. 2) [web: https://archive.org/details/electrical-energy-system-extra-ordinary-science-magazine; copy in research/files]. - Mexican patent application MX NL02000035 A, "Transformer generator of magnetic resonance into electric energy", filed 15 Nov 2002, published 20 May 2004 [web: https://patents.google.com/patent/MXNL02000035A/en; copy in research/files]. - Smith's NuEnergy group posts, compiled [web: https://energyevo.org/2015/01/18/updated-compilation-of-important-info-don-smith-shared-to-nuenergy-yahoo-group/]. - Lecture recordings: 1994 [web: https://www.youtube.com/watch?v=BHr3eDELyHk], 1996 workshop [web: https://www.youtube.com/watch?v=INPIfDGfMGs], 2005 TeslaTech [web: https://www.youtube.com/watch?v=duR1aYqiLzc]. These are the deciding records and have not been reviewed by this project.

Secondary and related (favourable to Smith; not primary evidence). - Joel Lagace, Understanding The Don Smith Capacitor Energy Method (undated) [read: project folder], and his 2026 videos [web: https://www.youtube.com/watch?v=Tiji0JqY90o]. - Patrick Kelly, A Practical Guide to Free-Energy Devices, chapter 31 [web: http://www.free-energy-info.tuks.nl/SChapt31.html]. - Vladimir Utkin, Free Energy: Nikola Tesla Secrets for Everybody (c.2012) [web: http://www.free-energy-info.tuks.nl/VladimirUtkin.pdf]. - Resonance Energy Methods, 119-page archive.org compilation including Kelly's Smith material [copy in research/files/Resonance-Energy-Methods-119p_djvu.txt]. - Energeticforum threads 5012 (from 2009), 13434 (2015) and 13932 (tswift, 2016) [web: https://www.energeticforum.com/forum/energetic-forum-discussion/renewable-energy/]. - J. E. Dalton II, "Radiant Energy Induction Using RF Pulsed Plasma" (2014) [web: http://gratisenergi.se/rfplasma, not opened]. - Related work Smith named: Tom Bearden's motionless electromagnetic generator line; Canadian patent document CIPO 2172240, as cited in Smith's 9 November 2001 e-mail [web: energyevo.org compilation above].

---

Answers

1. (a) U = (1/2)(2x10^-6)(500)^2 = 0.25 J. (b) 0.25 x (10^5)^2 = 2.5x10^9, in J/s^2: not an energy or a power, so the number has no physical meaning. (c) 0.25 J x 10^5 /s = 25 kW flows through the load, and the charger must supply the same 25 kW (plus losses). Frequency multiplies the throughput only if the source pays for every refill.
2. f0 = 1/(2 pi sqrt(10^-5 x 10^-9)) = 1/(2 pi x 10^-7) = 1.59 MHz. Q = omega0 L / R = (10^7 rad/s)(10^-5 H)/2 = 50. V_C = Q x 5 V = 250 V peak. P = V0^2/(2R) = 25/4 = 6.25 W, all dissipated in R. (Check: stored W = (1/2)(10^-9)(250)^2 = 31.25 uJ; omega0 W / Q = 10^7 x 31.25x10^-6 / 50 = 6.25 W.)
3. Only the resistor takes real power: (120)^2/10 = 1,440 W. The remaining current (sqrt(80^2 - 12^2) = 79.1 A) is reactive current into the capacitor bank, delivering nothing on average. The meters' 120 x 80 = 9,600 VA is apparent power.
4. Charge 100x10^-6 x 200 = 0.02 C shared over 200 uF gives 100 V. Energy before: (1/2)(10^-4)(200)^2 = 2 J; after: 2 x (1/2)(10^-4)(100)^2 = 1 J; 1 J lost whatever the wire's resistance. With an inductor and a diode the charge swings across; stopped at the peak, nearly all 2 J ends in the second capacitor (at nearly 200 V), less resistive loss. Never more than 2 J.
5. The power needed is P = omega W / Q. At fixed stored energy W and fixed Q, the energy lost per cycle is 2 pi W / Q, and there are f cycles per second, so the loss rate grows in proportion to f. (In practice Q also changes with f, and the skin effect raises R.)

---

Corrections to earlier versions and documents (QE round, 2026-10-08)

1. The earlier primer (section 4, magnetic-units row) called the pp.91-94 conversion table "Smith's own appendix". That was wrong: it is a textbook table bound into Smith's binder (its footnotes refer to "this text" and "footnote 6, Par. 176"), so the conflict with "1 gauss = 100,000 volts" is within the binder, not Smith contradicting himself (qe.md correction 2; arbiter QE revision).
2. The earlier primer (section 7) called the Builder's Guide "an ordinary, usable recipe". That overstated it; the deep read of pp.73 and 76-79 found a missing chart, three numeric defects and no power measurement (arbiter QE revision, claim 4).
3. The earlier evidence table gave "conservation excluded" the verdict "Refuted", which is not on the scale; it is "Refuted by calculation" (arbiter QE revision).
4. arbiter.md (QE revision, Retention) says that in his 9 November 2001 e-mail Smith both cited CIPO 2172240 and called his device "a close relative of Tom Bearden's Device". The Investigator's note (research/2026-10-08-energyevo-nuenergy.md) places the "close relative" sentence in a separate, undated post about the suitcase device that sits just before the 9 November 2001 e-mail, and the CIPO citation in the e-mail itself. This primer follows the note; the orchestrator should correct arbiter.md.
5. Orchestrator fix (deep reads of the three Investigator files, 2026-10-08). Several statements in earlier versions were wrong. Section 1, row 10 and "The commercial silence" put the "Boulder Dam" and "contract obligations" passage in Smith's voice: it is an unsigned paragraph in Kelly's compilation that refers to "Don" in the third person. The Utkin quotation "displacement current ... around" reads "is produced" in the source, and it is Utkin's caption for a 2006 rendering, not Smith's words. Step 6 gave the 7 Ah battery as [memory]: the suitcase schematic names it. No verdict changed (arbiter.md, "QE revision (orchestrator fix, 2026-10-08)"; deep/further-investigator-files.md).