Tesla-coil lamp experiment — notes, results, and why it appears like “overunity”

Diagram: Experimental setup overview with power levels

1 — Layman’s introduction: Why Chris Swinson’s experiment matters and what you should know

Chris Swinson’s experiment was an effort to replicate the controversial claims made by Donald L. Smith about devices that apparently produce more energy output than input — sometimes called “overunity” or “free energy.” The experiment involved powering lamps wirelessly using a Tesla-coil style setup with high-voltage bursts and resonant coils.

At first glance, it looked like a single transmitter coil could not drive more than one 100-watt lamp, but three receiver coils each lit their own 100-watt lamp at full brightness simultaneously. This appeared to suggest three times the energy output than input — a classic “free lunch” scenario. However, physical laws such as conservation of energy make this impossible.

The experiment and analysis here explain why such observations can be misleading and how thermal inertia of lamp filaments, pulsed energy bursts, measurement challenges, and hidden power pathways can create illusions of overunity. In short: there is no free lunch. What appears to be “extra” energy is really just the way the energy is delivered and perceived.

Understanding this helps prevent falling for misconceptions or faulty experiments that claim impossible energy gains.

2 — Experiment: hardware & measured behaviour

Transmitter (Swinson’s setup)

• High-voltage transformer: quoted rating ~10 kV at ~50 mA (nameplate ≈ 500 W). Practical input depends on load and spark-gap behaviour.
• Tank capacitance: ~12 nF charged to roughly 10 kV between mains peaks.
• Spark/rotary gap: fires near mains peaks — a few sharp discharges per mains half-cycle, typically ~100 bursts/s on 50 Hz mains (50 Hz × 2 peaks).
• Carrier oscillation in each burst: ≈ 500 kHz (several RF cycles per burst while the tank discharges into the transmitter coil).
• Receivers: three secondary coils placed roughly 1 m away, each connected in series to a standard 230 V, 100 W incandescent lamp on the coil ground/return side.

Observed behaviour

• The transmitter coil could directly drive a single 100 W lamp to normal brightness when coupled closely.
• Surprisingly, when three receiver coils were placed in the field, each lit its 100 W lamp apparently to full brightness simultaneously — even though the transmitter alone did not appear able to drive three lamps at full rated mains power.

Measurement considerations

Chris Swinson did not rely on electrical meters because conventional meters malfunction or give inaccurate readings in strong RF fields. Instead, visual aids such as incandescent lamps and motors were used as qualitative power indicators.

For example, specialist high-frequency diodes were obtained to rectify the RF AC output to DC, enabling Swinson to power a vehicle windscreen wiper motor. The motor ran visibly, proving that real mechanical power was received, although exact wattage was not measurable due to meter unreliability in the RF environment.

3 — Why this looks like “300 W from 100 W” (intuitive explanation)

There are three simple facts that make this observation confusing:

1. Stored energy in the resonator: Swinson’s tank stores a chunk of energy that is dumped into the secondary during each spark burst. That stored energy can be larger than the energy delivered per unit time by the transformer between bursts.
2. Bursts have huge peaks: each burst at 500 kHz contains many RF cycles and very high instantaneous voltages and currents — enough to deliver strong short heating pulses to a filament.
3. Filaments integrate heat: incandescent filaments have thermal inertia; they average the heating over milliseconds, so short strong pulses can make them look continuously bright even if the average electrical power is much lower than mains rating.

4 — Simple, step-by-step energy maths (the heart of the matter)

We’ll show the numbers Swinson used and how they limit the average power available to receivers. This is the practical ceiling for sustained lighting.

Energy stored in the tank (per burst)

  C = 12 nF = 12×10-9 F
  V = 10 kV = 10,000 V
  E_burst = 1/2 × C × V2 = 0.5 × 12×10-9 × (10,000)2 = 0.6 J
    

So each time the spark gap fires (one burst) the tank holds about 0.6 joules of energy to dump into the transmitter coil.

Burst repetition rate and ideal average power

  burst rate ≈ 100 bursts/s (50 Hz mains peaks × 2)
  P_total_ideal = E_burst × burst_rate = 0.6 J × 100 s-1 = 60 W
    

If every burst’s energy were magically delivered perfectly to loads, the maximum steady average power available would be about 60 W. That is the upper bound from the stored tank energy given these numbers.

Accounting for real-world losses (efficiency / coupling)

Real systems lose energy in the spark gap, primary coil, secondary coil resistance, radiation, corona, and imperfect coupling to receivers. Let’s look at a couple of realistic efficiency examples:

• Optimistic case (80% delivered): P_delivered = 60 W × 0.8 = 48 W total → three lamps share ≈ 16 W each (average).
• Conservative case (50% delivered): P_delivered = 60 W × 0.5 = 30 W total → three lamps share ≈ 10 W each (average).

Even in the optimistic case the average power per lamp is far below 100 W. Yet filaments can appear mains-bright because the power is delivered in high-peak bursts (see next section).

Could the transmitter have supplied more than 60 W steady?

Yes — a few ways:

• The transformer nameplate (10 kV @ 50 mA ≈ 500 W) is a rating — actual delivered power depends on duty cycle and spark-gap loading. Under heavy load the transformer may deliver more instantaneous current during peaks than assumed, so average input could be higher than the simple tank-based estimate.
• If the spark gap fires not once per mains peak but in multiple sub-bursts, or if the rotary gap runs faster, the effective burst rate and therefore average power rise.
• Ground and mutual coupling pathways can let receivers tap other energy flows (ground currents, re-radiated near-field), increasing delivered power beyond the naive tank-only calculation.

5 — Thermal filament model — how tens of watts can look like 100 W

To understand why a filament lit by pulsed RF can look as bright as a continuous mains-fed filament, we use a simple thermal argument: the filament mass (m), specific heat (c), temperature swing (ΔT) and pulse repetition (f) set the average heating power required:

  P_avg ≈ m × c × ΔT × f
    

Pick plausible numbers for a typical 100 W lamp filament:

• Filament mass m ≈ 5–20 mg (5×10^-6 to 2×10^-5 kg)
• Specific heat of tungsten c ≈ 134 J·kg^-1·K^-1
• Temperature swing ΔT (approx) ≈ 500–1000 K (depends on base temperature and desired peak)
• Burst repetition f ≈ 100 Hz (mains-peak-driven bursts)

Example calculations (ballpark)

Using ΔT = 800 K and f = 100 Hz:

• m = 5 mg (5×10^-6 kg): P ≈ 54 W
• m = 10 mg (1×10^-5 kg): P ≈ 107 W
• m = 20 mg (2×10^-5 kg): P ≈ 214 W

Interpretation:

• Small filaments (≈5 mg) might reach mains-like brightness with ~50–70 W average if energy is delivered as sharp peaks at 100 Hz.
• Heavier filaments require proportionally more average power to maintain the same temperature swing.

6 — Putting it together: 100 W → apparent 300 W explained clearly

Here are the main ways Swinson’s experiment could produce the appearance of three full-bright lamps while staying within physics:

1. Stored tank energy + bursts: The tank dumps ~0.6 J per burst. At 100 bursts/s that’s 60 W ideal. With losses, delivered average might be 20–50 W. That energy is split among three receivers, giving each ~7–17 W average — but high-peak pulses make each look much brighter than the average.
2. Transmitter actually supplied more than assumed: The transformer rating is a maximum, not a guarantee of lower power. Under the spark-gap duty cycle the transformer may have been delivering hundreds of watts in short peaks or higher average than the tank calculation suggests. If the true supplied average was, for example, 200–300 W, then three lamps at near 100 W each becomes possible.
3. Local ground/near-field coupling: Receivers can tap energy from the environment (ground currents, re-radiated near-field), so not all energy must transfer only through the primary→secondary near-field path. That can make it seem like receivers are getting more than expected from the primary alone.
4. Transient behaviour / stored energy depletion: If the lamps were bright only for a short period (seconds), the system could have been using stored resonant energy until it relaxed to a lower steady value. That would show high short-term output but not sustainable over long times without higher input.

In short: the visual reading (three bright lamps) can be produced by short-high-peak pulses + filament thermal integration, by underestimated input, or by tapping other conductive paths. Proper measurement reveals that total energy out ≤ energy in.

7 — Donald L. Smith and “duplicating fields” — why that idea fails

Donald L. Smith promoted an idea that resonant magnetic fields could somehow “duplicate” energy in receivers without taking it from the transmitter. This is inconsistent with electrodynamics and energy conservation.

Where his idea intersects reality

• Weak coupling + high stored field energy can make the transmitter barely notice one or more receivers tapping small amounts of power. That looks like duplication because the transmitter field is large compared to the small load.
• High apparent (reactive) power can be present in the resonator; if measured poorly it can be mistaken for real delivered power.

Why it fails physically

1. Energy accounting: if a receiver delivers useful work (resistive heating, light), the energy must be supplied from the source; in steady state the transmitter must supply that power.
2. Back-action: any load on the receiver changes the resonant circuit and imposes a reflected impedance on the transmitter. That change requires additional energy from the source to maintain amplitude.
3. No verified replications: claimed duplicating systems have not been independently demonstrated to produce net gain once correct measurements (true-RMS, good instrumentation) are used.

Chris Swinson’s experiments were motivated by trying to verify Smith’s claims, but his detailed investigations showed no violation of energy conservation.

8 — Practical measurement checklist (how to prove what is really happening)

1. Measure true input power at the mains feeding the whole transmitter using a true-RMS wattmeter (able to handle non-sinusoidal and pulsed loads) or using a current probe + oscilloscope measuring real power over time.
2. Measure the burst repetition rate and burst duration (use a pickup probe on the secondary to see envelope).
3. Estimate energy per burst from the tank formula and compare E_burst × burst_rate to mains input measured power.
4. Measure receiver actual dissipated power where possible (use resistive dummy loads or calibrated RF power meters). Lamps are poor RF power meters because filament behaviour is non-linear.
5. Check for ground/return currents — measure currents on the ground/earth stake to see if energy is flowing through unexpected paths.
6. Observe steady-state vs transient response — record how long lamps stay bright and whether input power rises when receivers are connected. If input rises proportionally, energy is being supplied continuously.

Note: Due to the strong RF fields, conventional meters may malfunction or give incorrect readings, so visual indicators such as lamps and motor operation remain valuable qualitative tools.

9 — Final practical notes & design reminders

• Small changes in coil geometry, burst timing, and lamp filament characteristics make large differences in perceived brightness.
• High-Q resonators store energy and can deliver impressive short bursts, but sustained power transfer to loads requires either higher continuous input or careful matched coupling.
• Always distrust visual brightness as a power measurement in RF/pulsed systems.

10 — Common mistakes made by amateurs replicating “free energy” experiments

Chris Swinson observed many common and fundamental mistakes made by amateurs trying to replicate experiments like his or Don Smith’s, which lead to wildly misleading claims:

• Incorrect power calculations: Many measure short-circuit currents using meters in an RF environment, obtaining thousands of amps. Then they measure open-circuit voltages, multiply these two numbers, and claim insanely high wattage outputs and overunity. This is physically invalid because short-circuit currents occur at very low voltages, and open-circuit voltages occur at zero current; multiplying these together overestimates real power by many orders.
• Ignoring voltage under load: Proper power measurement requires simultaneous voltage and current under load. Short-circuit current times open-circuit voltage is meaningless.
• Claims of megawatts of output with no practical load: Swinson saw people claim mega-watts of power output but could not even light a simple LED with their “energy.” This clearly shows their wattage calculations were incorrect or their devices produced no usable energy.
• Misunderstanding RF measurement techniques: Conventional meters and probes often give wrong readings in high-frequency or pulsed environments, leading to false conclusions.
• Social resistance in free energy groups: Swinson was involved in many free energy and alternative energy groups but eventually left due to repeated dismissals and insults when pointing out these fundamental measurement errors. Many people in those groups refused to accept evidence that their measurements and assumptions were flawed.

11 — Dedicated research and publication history

Chris Swinson invested significant time, effort, and expense into researching Tesla coil wireless power transfer experiments, initially attempting to replicate Don Smith’s work to verify its claims. His research spanned many detailed pages and was originally published online in the early 2000s. However, to avoid confusion and prevent misleading less experienced readers, much of this detailed work was later removed from public forums.

Despite the complexity and initial excitement surrounding apparent overunity results, Swinson ultimately acknowledged how easy it is to be misled by visual effects, measurement difficulties, and misunderstood physics. His honest and thorough exploration contributed valuable clarity to this niche field.