HiIQ-Youtube: Primer

Project: a 143-second narrated animation, "How "Partnered Output Coils" Work Together - Its NOW up to YOU!!!", by Chris Sykes (forum name EMJunkie) of hyiq.org, posted on or before 16 January 2015 [web: https://www.youtube.com/watch?v=Z-V1z2TdQJA; date from the opening post of overunity.com thread 15395, per the Investigator]. The folder name "HiIQ" stands for hyiq.org. The video's written source is Sykes's 49-page Guidelines to Bucking Coils. Lenz's Law Free Power Extraction (version 2.5, February 2015; version 2.7, 26 February 2016), which the Investigator recovered from the Wayback Machine [read: research/files/hyiq-Guidelines-to-Bucking-Coils-2016.pdf].

Source gap (repeated from Step 1, updated by the Investigator, the QE deep reads and the arbiter's QE revision). The project folder holds one file, videoplayback.mp4. Step 1 extracted only its spoken narration (about 1,070 characters). The Investigator then read the on-screen captions from video frames, identified the author, and downloaded the Guidelines, Sykes's 2011 paper Self-Assisted A Vector Potential Oscillation, the 1949 Cohn paper he recommends, and page 1 of the overunity.com forum thread. After the quality check (QE), deep readers read the video frame by frame (1 per second), the Guidelines' figures on pp.13-16, 21-24 and 34-48 as page images, the 2011 paper's oscilloscope screens, the forum page and Cohn's paper [read: deep/]. Those figures contain no input and output power measurement for any partnered-coil device. What is still missing: the figures on Guidelines pages outside those ranges; pages 2 to 350 of the forum thread and the earlier thread (topic 14974) it points to; the third-party paper by Cyril Smith (2015) and the gratisenergi.se build report; and the inventor's other videos (ZhQgch4L5XY, iUy_9jDKDvo, zSadMqEnO7I), which nobody has watched. So "no power measurement exists" below means "none in the documents and figures read". The physics in sections 3 and 5 rests on textbook relations recalled with high confidence [memory: high]; no verdict rests on a medium- or low-confidence recollection.

Two corrections to the transcript, accepted by all reviewers. The narration's "increase the separation of the input coil" is a mis-hearing; the on-screen caption reads "Feed Back and Feed Forward in the Partnered Output Coils can Increase the Seperation of Charge Carriers and Excess Output Energy can be obtained!" "Jaker" is Jaycar, an Australian electronics retailer [read: research/2026-10-06-video-frames.md; deep/videoplayback-3c5c77.md, frames at 1 per second].

Source tags: [read] = from a document in the project or in research/files; [web] = Internet source with URL; [memory] = general knowledge, with confidence; [calc] = computed for this primer (example.py).

---

1. The idea in one paragraph

An ordinary transformer has a price: when you connect a load to its output coil, the current in that coil makes its own magnetic field, which pushes back against the field of the input coil (this push-back is called Lenz's law), and the input coil must draw more current from the supply to make up for it. That extra input current is how the load's energy gets paid for. Chris Sykes proposes splitting each output coil into two "partnered" coils wound in opposite directions, so that their two push-back fields point opposite ways and add up to zero. If the push-back is cancelled, he reasons, the input coil no longer feels the load, so a 1-to-3 step-up should raise the voltage three times without lowering the current, and with the right "feedback and feed forward" between the coils, more energy should come out than goes in. He does not claim to create energy: on the forum he wrote "Energy can not be created or destroyed. You and I agree with Science on this one!", and he says the energy "was already there" outside the circuit [read: deep/overunity-15395-p1.md, replies #7 and #4]. He gives this away freely and asks viewers to build it for "around 20 dollars". The physics he starts from is real: opposed coils really do cancel fields (guitar pickups and non-inductive resistors use this), and returning energy into a transformer's output really does lower its input power. The question this primer works through is where the energy for the load comes from once the fields cancel.

---

2. What you need to know first

Faraday's law of induction. When the magnetic flux Φ (magnetic field times the area it passes through, in webers, Wb) through a coil of N turns changes, a voltage appears around the coil: ε = −N dΦ/dt. This induced voltage is called the electromotive force, EMF (ε, in volts). The minus sign is Lenz's law: any current the EMF drives makes a magnetic field that opposes the change in flux. Faraday's law is the whole basis of transformers and generators. Read: Griffiths, Introduction to Electrodynamics, ch. 7 [memory: high]; Feynman, Lectures on Physics Vol. II, chapters on induced currents and the laws of induction (ch. 16-17) [memory: high]. Sykes himself lists Feynman Vol. II ch. 15-17 among his reference downloads [read: research/2026-10-06-wayback-machine.md].

Inductance and mutual inductance. A coil's self-inductance L (henries, H) says how much EMF it makes in itself per rate of change of its own current: ε = −L dI/dt. Two coils sharing flux have a mutual inductance M (also in henries): a changing current in coil 1 induces ε₂ = −M dI₁/dt in coil 2, and the same M works in reverse (this symmetry, M₁₂ = M₂₁, is a theorem, not a coincidence). The coupling coefficient k = M/√(L₁L₂) runs from 0 (no shared flux) to 1 (all flux shared). Inductance grows as the square of the number of turns, so a coil with 3 times the turns on the same core has 9 times the inductance. A coil with resistance R driven by a step of voltage reaches its final current with time constant τ = L/R. Read: Griffiths ch. 7; any circuits text, e.g. Nilsson and Riedel, Electric Circuits, chapter on magnetically coupled circuits.

AC circuits with complex numbers (phasors). For a sine wave at frequency f (hertz), write the angular frequency ω = 2πf (radians per second). A sinusoidal voltage or current can be written as a complex number, a phasor, whose size is the RMS (root-mean-square) value and whose angle is its phase. Then an inductor behaves like an "impedance" Z = jωL, where j = √−1 (engineers use j, not i, to avoid confusion with current), a resistor like Z = R, and Ohm's law becomes V = Z I with complex numbers. Coupled coils give a matrix equation Z I = V, solved with ordinary linear algebra. Read: Nilsson and Riedel, chapters on sinusoidal steady-state analysis.

Real power, apparent power and power factor. With RMS phasors, the real (average) power delivered is P = Re(V I) = V_rms I_rms cos φ, where I is the complex conjugate and φ is the phase angle between voltage and current. The product V_rms × I_rms, read off two meters, is the apparent power S, in volt-amperes (VA); the ratio P/S = cos φ is the power factor. Because cos φ ≤ 1, apparent power is always at least as large as real power. With coils, φ can be close to 90°, so V × I can overstate real power many times. Read: Nilsson and Riedel, chapter on sinusoidal steady-state power.

Magnetomotive force and the ampere-turn balance. A coil of N turns carrying current I pushes flux around an iron core with a "magnetomotive force" MMF = N I (ampere-turns). The core's opposition to flux is its reluctance ℛ (a magnetic analogue of resistance), and the total MMF from all coils on the core equals ℛ Φ. In good iron ℛ is small, so the MMFs of all coils very nearly add up to zero. Read: Chapman, Electric Machinery Fundamentals, the transformer chapter.

Conservation of energy and Poynting's theorem. Energy cannot be created or destroyed. For electromagnetic fields this is Poynting's theorem: the energy flowing out of a region (given by the Poynting vector S = E × B / μ₀, with E the electric field, B the magnetic field and μ₀ = 4π × 10⁻⁷ H/m) equals the decrease of field energy inside plus the work done on charges. A transformer is a device for moving energy from one circuit to another through this field. Read: Griffiths ch. 8.

---

3. The textbook view

3.1 The ideal transformer

Wind N₁ turns (primary) and N₂ turns (secondary) on the same iron core, so both see the same flux Φ. Faraday's law gives each coil an EMF proportional to its turns, so

V₂ / V₁ = N₂ / N₁.

A 1:3 winding gives 3 times the voltage. Sykes states this correctly [read: Guidelines v2.7, p.11].

Now load the secondary with resistor R_L. A current I₂ = V₂ / R_L flows (Ohm's law). Its ampere-turns N₂I₂ push against the core flux (Lenz). The flux is fixed by the primary voltage (V₁ = N₁ dΦ/dt), so the primary must draw extra current to cancel the secondary's ampere-turns:

N₁ I₁ − N₂ I₂ ≈ N₁ I_m ⟹ I₁ ≈ (N₂/N₁) I₂ + I_m,

where I_m is the small "magnetising" current the primary draws with no load. With I_m negligible, V₁I₁ = V₂I₂: power in equals power out, and a 1:3 voltage step-up comes with a 3:1 current step-down. Sykes gives this exact example in the step-down direction: "Input Power = Output Power = (240Volts * 1 Amp) (24Volts * 10Amps)" [read: Guidelines v2.7, p.11].

The extra primary current is felt by the supply as an extra impedance, the reflected impedance: a load R_L on the secondary looks like R_L (N₁/N₂)² on the primary [memory: high]. A 100 Ω load on a 1:3 secondary looks like 100/9 = 11.1 Ω to the primary.

3.2 Lenz's law is the energy bill, not a defect

Here is the one-line identity that this whole primer turns on. For coils that all link the same core flux Φ, coil k has EMF magnitude N_k dΦ/dt, and the power it delivers to its own external circuit is that EMF times its current. Count each current i_k as positive in the direction whose ampere-turns push against the flux direction (the Lenz direction while Φ is growing). Then the power coil k delivers to its external circuit is

P_k = (N_k i_k) × dΦ/dt.

Each coil's power output is its own MMF times the rate of change of the shared flux. Sum over all the output coils:

P_out,total = (Σ N_k i_k) × dΦ/dt.

The ampere-turns that oppose the core flux, the "Lenz field", are exactly the factor that multiplies dΦ/dt to give delivered power [memory: high; this is the standard derivation of transformer energy transfer, written out here for the reader]. No opposing MMF, no delivered power. This is why physicists say Lenz's law is not an obstacle engineers have failed to get around; it is how the load's energy reaches the primary.

3.3 Opposed ("bucking") coils in standard practice

Coils wound or connected to oppose each other are common and well understood [memory: high]:

• Humbucking guitar pickups: two coils, opposite winding and opposite magnet polarity. Hum from stray fields induces opposite EMFs that cancel; the string signal adds. Sykes cites humbuckers [read: Guidelines v2.7, p.42-43].
• Non-inductive (bifilar) winding: a wire folded back on itself so the two halves carry current in opposite directions. Their fields cancel, the inductance nearly vanishes, and the resistor heats just the same.
• Buck-boost transformers: a winding connected to subtract from (buck) or add to (boost) a line voltage.
• Active noise cancellation: a speaker makes an inverted copy of the noise so pressure cancels at the ear. The speaker spends power to do it; the cancelled sound energy is not recovered.

In every one, field cancellation in some region is real, and in none does the cancellation reduce the energy the source must supply to the load.

3.4 Coupled coils with leakage and resistance

Real coils share only part of their flux (k < 1), and their wire has resistance. Then the loop equations for three coils are, in phasor form,

(r₁ + jωL₁) I₁ + jωM₁₂ I₂ + jωM₁₃ I₃ = V₁ jωM₁₂ I₁ + (r₂ + R₂ + jωL₂) I₂ + jωM₂₃ I₃ = V₂ jωM₁₃ I₁ + jωM₂₃ I₂ + (r₃ + R₃ + jωL₃) I₃ = V₃

with r the wire resistances, R the load resistances and V any source in each loop (zero for a plain load). Winding a coil the other way flips the sign of every M that involves it. Because the matrix of L's and M's is symmetric and stores positive energy for any currents, a short calculation shows that the real power going in equals I^H R I, the total heat in all resistances (I^H is the conjugate transpose of the current vector) [memory: high]. For any linear passive set of coils, whatever their placement, coupling or winding sense, output can never exceed input. This is a theorem for that class of circuits; example.py also checks it on 2,000 random cases.

3.5 Two kinds of induction, and "separation of charge carriers"

Induction has two textbook mechanisms [memory: high]. In motional EMF, a conductor moves through a magnetic field B with velocity v, and the magnetic part of the Lorentz force, F = q v × B, pushes positive and negative charges toward opposite ends (EMF = B v l for a rod of length l). In transformer EMF, nothing moves; a changing B creates a circulating electric field (∇ × E = −∂B/∂t), and that electric field pushes the charges. Both end with charges separated and a voltage at the terminals. G. I. Cohn's 1949 AIEE paper, which Sykes recommends, separates these two cases carefully [read: Cohn p.1, page image, per the Investigator; deep/cohn-1949.md]. It does not discuss coupled windings, bucking coils or load current [read: deep/cohn-1949.md].

---

4. Where the inventor departs

Sykes's starting points are mostly standard. He states the ideal transformer rule correctly [read: Guidelines v2.7, p.11]. He states that when a transformer is loaded "there is a NET ZERO Magnetic Field. This means the Ampere Turns on the primary are equal but opposite to the Ampere Turns on the secondary, or close to it", which is the ampere-turn balance of section 3.1 [read: p.21]. He writes that "this reflected power must be compensated for ... and as a result the input Current must increase" [read: p.21]. He describes induction as "the Separation of Charge Carriers" by the Lorentz force [read: p.9], a fair picture of motional EMF. His remark that "Lenz's law is the Result of drawing Current from the already Separated Charge, not the cause!" [read: p.9] also has a true core: an open-circuit coil has an EMF but no opposing field, and the opposing MMF appears only once current is drawn.

The departures come at specific steps:

So the exact point of departure is the second row: treating "the opposing fields sum to zero" as the cure for Lenz's law, when by his own correct statement on p.21 a net-zero field is what an ordinary loaded transformer already has, and when, by the identity in section 3.2, a pair of secondaries whose ampere-turns cancel delivers no net power. The mathematician notes the internal tension: the same page presents net-zero field both as the Lenz effect and as its cure [mathematician.md].

The inventor's circuit. Earlier drafts of this review said the video and the Guidelines describe different circuits. That was wrong, and the arbiter has retracted it. The Guidelines define a "Power Output Coil" as at least two partnered coils, so "One input or Excitation coil per Power Coil" means one input coil per pair, and the input coil should "occupy only one half of the Power Coils Geometry" [read: p.1]. The schematic on p.38 shows it: a function generator feeds a power amplifier, which drives a single input coil L1. L1 shares one core with the partnered coils L2 and L3. L2 and L3 are wired in series through one "Resistive Load 10" (10 Ω, "this can be decreased or increased as needed"). There is no diode, no capacitor and no wire from input to output; no polarity dots are drawn [read: page image p.38, per deep/hyiq-guidelines-2016.md]. An "Output Configuration guide" on the same page shows two wirings: "CW/CW 1 Flipped Coils", a series loop like the main schematic, and "CW/CCW Coils" (CW = clockwise, CCW = counter-clockwise winding), in which the two coils are in parallel across one load [read: same]. The video agrees: one green input coil sits over one of the two gold output coils, which are one wire joined end to end, with their field arrows pointing head to head [read: deep/videoplayback-3c5c77.md; the arrow direction on the input-bearing coil is low confidence at 640×360]. On the forum Sykes wrote: "On the input, really anything that works! I have always used a single coil" [read: deep/overunity-15395-p1.md, #12]. Case B4 of the worked example is this circuit.

One consequence, a deep reader's analysis rather than the inventor's statement, which the arbiter keeps: two coils on a common flux, wired in series so their fields oppose, have induced voltages that subtract. Output then exists only to the extent that the flux through the two coils differs, which is why the input sits over one coil only, and why Sykes writes that "Total 'Cancellation' is also impossible" [read: p.45]. So "the fields sum to zero" and "an output appears" pull in opposite directions [arbiter.md, QE revision §1].

---

5. Worked example: the energy ledger of a 1:3 partnered-coil transformer

All numbers below were computed by hand where shown and checked with example.py [calc]. Run it with:

.venv/bin/python review/HiIQ-Youtube/example.py

Set-up (my illustrative choice, not the inventor's numbers; the documents give no coil values, only the 1:3 ratio and the 10 Ω load on p.38). Primary driven with V₁ = 12 V RMS. Two output coils, each with 3 times the primary's turns (the video's 1:3). Steps 1 to 3 give each output its own 100 Ω resistor, the textbook bookkeeping case; Step 4 then wires the pair as the inventor does.

Step 1. The ideal transformer, by hand

• Output voltage per coil: V₂ = 3 × 12 = 36 V.
• Output current per coil (Ohm's law): I₂ = 36 / 100 = 0.36 A.
• Power per load: 36 × 0.36 = 12.96 W; two loads: 25.92 W.
• Ampere-turn balance: N₁I₁ = 3N₁ (I₂ + I₃), so I₁ = 3 × (0.36 + 0.36) = 2.16 A.
• Input power: 12 × 2.16 = 25.92 W. Equal to the output.
• Each output current is 0.36/2.16 = 1/6 of the primary current (the pair together carry 1/3). Two 3× outputs share the primary's ampere-turns.

Step 2. What "current is not stepped down" would mean

The output current is set by the load (0.36 A), so in the trivial sense nothing "steps it down"; that is true of every transformer. The strong reading is that the primary draws no more than the output current. If the primary drew only 0.36 A, the input would be 12 × 0.36 = 4.32 W against 25.92 W out: a coefficient of performance (COP, output power over input power) of 6. With one output coil instead of two, the same reading gives COP 3. This matches the reviewers' examples (physicist: 21.6× with 10 Ω loads and a 1 A primary; contrarian: 6× with 1 A everywhere) [arbiter.md, Recomputed numbers].

Step 3. A realistic coil set: winding direction alone changes nothing

Now make the coils real: L₁ = 10 mH, L₂ = L₃ = 90 mH (9× for 3× turns), coupling k = 0.95 between every pair, wire resistance 0.5 Ω on the primary and 1.5 Ω on each secondary (three times the wire), frequency 1 kHz (ω = 6,283 rad/s, so ωL₁ = 62.8 Ω). Solve Z I = V (section 3.4) [calc: example.py Part B]:

What to notice: - Leakage (k = 0.95) and loading pull the output down to 24.97 V instead of the ideal 36 V, as the arbiter's caveat to claim 2 predicts. - Output current over primary current is 0.2497/1.544 = 0.162, close to the ideal 1/6: current is stepped down. - Power in = loads + wire to rounding error (the script prints the difference, about 10⁻¹⁵ W). - Reversing one coil's winding only flips the sign of its current. For coils each driving their own load, winding sense is a relabelling. It matters only when the coils are joined to each other. - B1 and B2 are not the inventor's circuit (he joins the pair into one load); they are here only to show that winding sense alone changes nothing. Step 4 is his circuit.

Step 4. The inventor's circuit: the partners in series, opposing, into one load

Connect the two output coils in series, opposing, into one load. This is the circuit of Guidelines p.38 and of the video (section 4), and the literal "vectors sum to zero" case: the same current flows through both coils in opposite senses. The combined output loop has inductance L₂ + L₃ − 2M₂₃ and couples to the primary through M₁₂ − M₁₃ only.

• B4, asymmetric coupling: the inventor's own circuit, and the headline case. The input coil sits over one of the pair, so it couples strongly to that one and weakly to the other: k₁₂ = 0.8, k₁₃ = 0.3, k₂₃ = 0.5 (illustrative values) [calc: example.py Part B4]:

The cancellation is partial, so some power reaches the load: exactly as much as the primary supplies minus copper loss. Compare B1: the same coils on separate loads take 13.85 W and deliver 12.47 W. Opposing the pair cuts the input power by a factor of about 40, and the load power by the same kind of factor. The primary "feels the load less" because the load gets less. With the 10 Ω load, the 3 Ω of secondary wire takes more power than the load does. - B3, symmetric coupling: the limiting case. If both outputs were equally coupled to the primary (k = 0.95 everywhere), the pair's net coupling is M₁₂ − M₁₃ = 0. Hand check: the primary is then just a lone inductor, I₁ = 12 / |0.5 + j62.83| = 12 / 62.83 = 0.191 A, and its real power is 0.5 × 0.191² = 0.018 W, all heat in its own wire. The load current is exactly zero [calc: Part B3, 200 Ω load]. The primary indeed "sees no load". The reason is that the load receives nothing: the two EMFs cancel around the loop just as the two MMFs cancel. This is section 3.2's identity in action, and it is why the inventor puts the input over only one coil.

The parallel wiring of p.38 ("CW/CCW Coils") is another passive linear connection, so the theorem of section 3.4 covers it too: output cannot exceed input whatever the connection [arbiter.md, QE revision].

This also explains, without new physics, Sykes's rotor test, in which magnets spun past a bobbin with partnered coils and he saw "No or next to no Lenz's Law and at the same time, Electricity is Generated" [read: p.45]. If the partners cancel, little current flows to a load, and little drag appears; the B3 case is the electrical version of that observation. No numbers for load current, shaft torque or speed are given, so the test cannot tell the two readings apart.

Step 5. The inventor's feedback mechanism, modelled fairly

Sykes describes the feedback with an analogy: drive a transformer's primary from the mains, and let the secondary's load "just so happen to be another Power Source" of adjustable power and phase; turn it up and "the input power is reduced" [read: p.46]. That is correct physics. To show it, keep the B1 coils, keep one output on its 100 Ω load, and replace the other output's load with a 36 V RMS AC source whose phase relative to the primary we choose [calc: Part C]:

(Negative primary power means power flows back into the primary's supply.)

What this shows: - With the right phase, the primary's input power falls, and past a point it goes negative: the input coil becomes an output coil. That is the title of Sykes's section, "When does an Input Coil become an Output Coil?" [read: p.45], and the mechanism is the one he describes. On this point the contrarian's reading is right: the Guidelines describe power entering from a second source on the output side [contrarian.md]. - In every row, primary power + second source power = load + wire. The energy that lowers the input bill comes from the second source. - The trap. At 0°, someone who meters only the primary and the resistor sees 9.95 W out for 5.28 W in, an apparent COP of 1.89. The "gain" disappears the moment the second source is counted. In a real bench device the "second source" need not be a labelled generator: a charged capacitor, a battery, a permanent magnet being demagnetised, or a supply that is metered wrongly can play the same role. Sykes's own 2011 paper names a second source in his first device: "we have two sources of input energy ... the Input we put into the device (E = V * I), and two, the Permanent Magnet" [read: Self-Assisted paper]. - The inventor's own feedback experiment says the same. Guidelines p.48: an automotive alternator driven by an electric motor, with its field coil shorted or put in series with the output coils, gets "enough Feed Back ... to get an Electrical Output". He then writes: "Not Free Energy, but this shows that Feed Back Techniques can excite the Coils and Magnetic Fields", and that "ONLY because of Lenz Law, Resistive Mechanical Force is applied on the Shaft in Opposition to the applied Input from the Electric Motor!" [read: p.48]. This is generator self-excitation, a standard effect [memory: high]: the motor pays for the output. No numbers are given [read: deep/hyiq-guidelines-2016.md].

Step 6. Meter readings versus real power

Multiplying an RMS voltmeter reading by an RMS ammeter reading gives apparent power, which is never smaller than the real power [calc: Part D]:

So V × I at the input overstates input power (by 9× in the light-load case); that error makes a device look worse, not better. The same mistake on the output side, with a reactive (non-resistive) load, overstates output and can make a device look better. Non-sinusoidal waveforms read on meters calibrated for sine waves, and DC supplies feeding pulsed loads, add further errors; a reply in the overunity.com thread (#3589, 19 June 2015) warns how easy it is to "screw up an input power measurement involving a DC supply" [web, via the Investigator; reply not read]. The reliable method is the time average of v(t) × i(t), sampled together on both sides over whole cycles, or calorimetry of the load. Sykes's 2011 paper says "No to minimum (0 5%) Phase angles have to be taken into account when measuring this power" [read: Self-Assisted paper], a statement made without a phase measurement in the text.

The 2011 paper's own scope screens, as a worked exercise. Fig. 4 shows the lamp "3.6 Watt Load" with 1.977 V RMS across the coils and a "current" channel reading 82.41 mV RMS, with no current-sense (shunt) resistor stated; the Specifications section gives no input voltage, current, power or phase; the 15.2 W DC input appears only in the prose [read: deep/hyiq-self-assisted-2011.md, Fig. 4, pp.5-6, p.9]. The lamp is rated 12 V, 300 mA = 3.6 W, so 3.6 W is its nameplate. At 1.977 V, scaling by V²: (1.977/12)² × 3.6 = 0.098 W, ignoring the filament's change of resistance. With a 1 Ω shunt the output would be at most 1.977 × 0.0824 = 0.163 W, about 1.1% of 15.2 W; with a 0.1 Ω shunt, at most 1.63 W [calc: arbiter.md, QE revision §2]. The unknown shunt sets the answer by a factor of ten, and every reading is below 15.2 W, which fits his own "below unity" without testing it (the DC figure may include amplifier losses). The paper's coil voltages also do not reconcile: 8.56 V in Figs. 2-3, 1.977 V in Fig. 4, "10V RMS" in the text.

Step 7. A units check on the inventor's one quantitative coil experiment

The Guidelines report a bifilar coil, 0.231 H with the two strands in series and 0.24 mH in non-inductive connection, 7.8 Ω either way. Charge times: 75 ms and 650 µs. Sykes concludes "This proves that the Magnetic Field Slows the Charge Flow Rates dramatically" [read: Guidelines v2.7, p.15-16]. His schematics put a 1 Ω current-sense resistor in series with the coil, so the resistance in the time constant is R = 7.8 + 1 = 8.8 Ω [read: page images pp.14-15, per deep/hyiq-guidelines-2016.md; the arbiter's QE revision corrects its earlier 7.8 Ω]. By hand: - Inductive: τ = L/R = 0.231 / 8.8 = 26.3 ms. Current reaches 90% of its final value at τ ln 10 = 2.30 τ = 60 ms, and 95% at 3τ = 79 ms. His 75 ms fits an ordinary inductor. - Non-inductive: τ = 0.24 × 10⁻³ / 8.8 = 27.3 µs, 90% at 63 µs. He gives 650 µs; the deep reader reads about 400 µs from the photographed scope screen. Either way the rise is some 15 to 25 times slower than L/R predicts, and the time ratio (75/0.65 = 115) is much smaller than the inductance ratio (962), so something other than the coil (the MOSFET gate drive, the generator edge, the probe) set the fast case. - The two runs also differ in supply (15.4 V against 7.0 V) and drive frequency (about 6 Hz against about 647 Hz), so "the same resistance" is the only thing held equal [read: deep/hyiq-guidelines-2016.md]. - A check on his inductances: two strands of 55.6 mH giving 0.231 H in series imply M = (0.231 − 2 × 0.0556)/2 = 0.0599 H and k = M/L = 1.08, above the physical limit of 1; read it as near-perfect coupling within measurement error [calc: arbiter.md]. - Conclusion: the slow case is the textbook back-EMF, ε = −L dI/dt. It is a correct observation, explained by standard physics, and it says nothing about net energy [calc: Part F; agrees with the arbiter].

Step 8. The general result

Over 2,000 random passive coil sets (2 to 4 coils, random couplings of either sign, random loads), the largest output-to-input ratio found was 0.993 [calc: Part E]. That is only an illustration; the theorem in section 3.4 is the proof. What the theorem does not cover is a device that is active, nonlinear (saturating cores), parametric (inductance changed in time by something external), or that contains a stored-energy source such as a magnet or battery. For those, the energy question is settled by measurement of every input, not by this calculation (the arbiter's point, taken from the contrarian).

---

6. Evidence table

Verdicts are the arbiter's, including its QE revision (no verdict moved; the wording for claim 4 changed).

---

7. What a learner should take away

What is solid. - Opposed coils cancel fields in the space around them, and this is used every day (humbuckers, non-inductive resistors, buck-boost transformers). Sykes's starting observation is real. - His statements of the ideal transformer and of the loaded transformer's net-zero ampere-turns are correct [read: p.11, p.21]. - His description of "feedback" is correct physics: a source on the output side, with the right phase, lowers the input power and can reverse it (Step 5). His remark that the opposing field appears only when current is drawn is also correct. - The 1:3 voltage step-up is standard. - He keeps self-excitation apart from free energy: his alternator feedback experiment is labelled "Not Free Energy" [read: p.48], and he says energy is conserved [read: deep/overunity-15395-p1.md, #7].

What is wrong, and why the mistake is instructive. - The opposing ampere-turns are not a removable side effect; they are the term that carries power from the core into the load: P = (N i) dΦ/dt. Cancel them and the delivered power cancels with them (Step 4, case B3). The instructive part is seeing that "field cancellation" and "energy cancellation" are tied together by a single identity, and that the identity is just Faraday's law multiplied by current. - "Magnetic must cancel, electric must add" is not available for coils on a common flux joined in one loop: the winding sense that adds the EMFs also adds the MMFs. - "Current is not stepped down" in a 1:3 step-up means 3 to 6 times more power out than in (Step 2), and it contradicts the example Sykes himself gives a few lines earlier. - Lowering input power by feedback is not a gain unless every source is metered (Step 5's 0° row shows an apparent COP of 1.89 that vanishes once the second source is counted). This is the most useful lesson for anyone testing such devices.

What is genuinely open. - No instrumented input/output power measurement of a partnered-coil device is in the documents read. The records most likely to hold one are unread: later pages of overunity.com thread 15395 (replies #3589, msg469812, msg471869, and a Prezi quote of an input:output ratio of "1 to 100" whose context nobody read), the Cyril Smith paper "Bucking Coils produce Energy Gain" (2015), the Energetic Forum "Bucking Coil Inverter" thread (a 6 W LED lit through 30 AWG partnered coils, input power not seen), and the inventor's other videos [web, via the Investigator]. - The theorem in section 3.4 covers passive linear coils. Whether a saturating, resonant or magnet-assisted build shows a real anomaly, or only measurement artefacts, is a bench question. The arbiter's deciding test: same coils, sine drive, resistive loads, simultaneous time-averaged v × i on input and output, partnered versus conventional winding as a control, with calorimetry of the load as a check. - Where does "COP 1.7" come from? A critic attributes it to the inventor's device and the inventor did not dispute it on page 1 of the thread. Its likely origins, the earlier thread (overunity.com topic 14974 "Magnet Myths and Misconceptions", "from about page 76 on") and the opening post's attachment "Net Zero Magnetic Field.png", have not been read [read: deep/overunity-15395-p1.md]. It and the Prezi "1 to 100" differ by a factor of about 60, and neither has a known measurement behind it. - Did the testers MileHigh named ("TK or Itsu ... with some help from MarkE and/or Picowatt", whose replication he said "will be definitive") ever test it? Page 1 shows no test by any of them; the answer, if any, lies in the later pages [read: deep/overunity-15395-p1.md, #8]. MileHigh's 10-point protocol (measure the load resistor, state both test points, give error tolerances, do not trust bench-supply current displays) is a fair statement of measurement standards from a critic, and it fits the arbiter's deciding test, run on the p.38 circuit and then on the parallel wiring.

The record, kept whatever the verdict. - The inventor's own history, in his words. In 2011 he wrote, of his first device, "The device I have shown is below unity" and "Please Note: The device I have shown is not an Over-Unity Device", calling the effect "the door to achieving above unity Devices"; the device drew "15.2 watts DC" and lit "an Automotive Globe - 3.6 Watt Load"; "if I short the Coils, the Output Load is increased at no extra cost to the Input" [read: Self-Assisted paper]. By 2015 the caption says "Excess Output Energy can be obtained" and the Guidelines say "this Technology DOES WORK!!!" [read: p.45]. No measurement bridging the two is in the documents read. But the record is not a straight line from "below unity" to "gain". The same Guidelines label his alternator feedback experiment "Not Free Energy" [read: p.48]; on the forum in 2015 he wrote "Energy can not be created or destroyed. You and I agree with Science on this one!", "Lenz's Law is ONLY apparent when Current is drawn ... So Energy is not Generated at all! It was already there!", "Loose Coupling is a requirement!", and "To-date, I have found 3 configurations to work! I have not included one because it over complicates the situation for beginners", without describing them or defining "work" [read: deep/overunity-15395-p1.md, #7, #4, #11, #12]; and the 2011 paper already called the effect "the door to achieving above unity Devices". The arbiter's conclusion: the change between documents is one of emphasis, not a demonstrated reversal [arbiter.md, QE revision §5]. He is also candid: "For every one device that works, expect 10 devices that don't" and "I may be wrong" [read: p.34]; "I believe I still don't quite fully understand this section" [read: p.45]; "Total 'Cancellation' is also impossible" [read: p.45]; and "I have given everything freely, all I ask in return is to give credit to those who deserve it!" [read: video end card, per the Investigator and deep/videoplayback-3c5c77.md]. - "COP 1.7". The only performance figure for the device in the record. The forum critic MileHigh (a self-described "hostile witness") wrote on 16 January 2015 of "your alleged COP 1.7 device" and "how are we even supposed to know that your COP 1.7 claim is even credible? I looked through your pdf doc and I didn't see anything like that". COP, coefficient of performance, is output power over input power, so 1.7 would mean 70% more out than in. Sykes quoted the post, called it "sensible and logical", and replied "I have decided that the only proof is in the eyes of the end user ... It only makes sense to build and test for ones self!", neither confirming nor disputing the figure [read: deep/overunity-15395-p1.md, #8, #11]. It is a figure attributed to the inventor by a critic, documented in one forum post so far, with no circuit or measurement given. It can neither be used as a measurement nor dismissed. - The inventor's other experimental claims and principles, with the physicist's judgement beside them. The CRT-magnet experiment (p.23; see the evidence table). "So Mechanical Energy is NOT the Source of the Energy Generation!" (p.8), drawn from a loaded versus an open generator; the physicist answers that in a generator the shaft work is the source, and an open generator turns almost freely because no current means no Lenz drag, which is what conservation predicts [physicist.md; memory: high]. "A is the Electric Component of the Magnetic Field and, Phi, is the Magnetic Component" (p.6), where A is the magnetic vector potential; the physicist notes these are not the textbook meanings of A or Φ (flux) [physicist.md]. "I seem to get a better result if the Input is resonated, for example, LC Tank" (p.24). "Magnetic Gating" with a diode (pp.40-42), marked "work in progress". His first device dates from 2010-11 (p.24) [read: deep/hyiq-guidelines-2016.md]. - Third-party reports, each documented in one source so far. gratisenergi.se: "I made a Bucking Coils Transformer and was dissapointed when I found no Free Energy" (Google snippet; page now 404; no second source found yet). Cyril Smith (2015): title "Bucking Coils produce Energy Gain", first line "There are many claims of overunity for systems that employ bucking coils"; the text could not be read, so its conclusion is unknown. Energetic Forum: a builder's inverter lights a 6 W LED through partnered coils; input power not seen. aboveunity.com threads from 2025 ("Floyd Sweet VTA Partnered Output Coils", "Fringe Experiments With VTA Style POCs") show builders still at work [web, via the Investigator]. The aboveunity.com forum and hyiq.org are the inventor's own sites; "above unity" labels there are his and his community's characterization. - Prior art he names (his attributions, mostly unverified): Floyd Sweet's Vacuum Triode Amplifier (VTA, 1:5 step-up), Paul Raymond Jensen's UDT transformer ("consider each secondary winding as acting as the primary winding for the other secondary winding when an output current is drawn" [read: p.34], which is the textbook reading of a bucking pair), Earl Koenig US 4,806,834, Gennady Markov, Melnichenko, Frolov, Bill Alek, Daniel McFarland Cook US 119,825 (1871), Tom Bearden [read: Guidelines]. The arbiter checked Koenig's patent: filed 1987, granted 1989, mirror-image coils in parallel; it reports a step-down transformer giving 5.77 V against an expected 3.09 V and "efficiency of almost 100 percent", and makes no explicit over-unity claim; it is the patent holder's own account, unverified, and a different circuit [web: https://patents.google.com/patent/US4806834A/en]. No independent measurement of Sweet's VTA was found; its claims come through Sweet's associates and promoters [web: scienceforums.net and KeelyNet, per the physicist]. - A quoted claim that fails a calculation. The Guidelines quote Sweet: "the same gauge wire would carry 10 times more current at the same temperature" [read: p.17]. With the wire's resistance fixed, heat is P = I²R, so 10 times the current means 100 times the heat [physicist; memory: high]. - Cohn cuts both ways, narrowly. The paper Sykes recommends for its care with induction also says confusion between the two induction laws produced "innumerable d-c generators without moving contacts which have been invented (and do not work)" [read: Cohn p.1, page image, per the Investigator]. That line is about those machines, not bucking coils; beyond p.1 Cohn does not treat coupled windings, load current or Lenz back-reaction, so his paper neither supports nor refutes partnered coils [read: deep/cohn-1949.md; arbiter.md, QE revision §4]. - The record going dark. On 2026-10-06 hyiq.org and aboveunity.com did not respond (both on one server address); overunity.com, which hosted the 2015 test thread, is a domain for sale; the last good Wayback capture of the Guidelines is 2024-10-27 [web, per the Investigator and the arbiter]. Competing explanations: a hosting fault (Google still indexes the pages; aboveunity threads are dated December 2025), a move to another platform (he has moved before), withdrawal, or a payment or domain lapse; he writes "PayPal De-Platformed me" [read: Wayback 2022, per the Investigator], which is his own account. Evidence so far favours a transient or hosting fault; a repeat visit, 2026 Wayback captures and domain records would decide. - How the claim was treated. No journal, institution or controlled test examined it; the contrarian found it was ignored rather than tested or suppressed, and nothing read documents suppression [contrarian.md].

---

8. Check yourself

1. A single-secondary transformer has a 1:3 turns ratio. The primary is driven at 24 V RMS and the secondary feeds a 360 Ω resistor. Assume an ideal transformer. Find the output voltage, the output current, the primary current, and the power.
2. Two identical output coils sit on the same core with identical coupling to the primary. They are joined in series, opposing, across a lamp. Why does the primary current hardly change when the lamp is connected, and what does the lamp do?
3. A builder reads 12 V RMS and 1.5 A RMS at a coil's input on true-RMS meters, and 30 V RMS and 0.4 A RMS on a pure resistor at the output. Can the apparent power reading at the input by itself create a false impression of gain here? What would?
4. Using the identity P_k = (N_k i_k) dΦ/dt, show that if N₂i₂ + N₃i₃ = 0 at every instant, the two output coils together deliver no net power. Which of Sykes's own quoted sources says the same thing in words?
5. The Guidelines' bifilar coil has L = 0.231 H and 7.8 Ω, measured through a 1 Ω sense resistor in series. What is the time constant, and how long does the current take to reach 90% of its final value? Does a 75 ms charge time show anything beyond ordinary inductance?

---

9. Further reading

Mainstream - D. J. Griffiths, Introduction to Electrodynamics: ch. 7 (Faraday's law, inductance, magnetic energy) and ch. 8 (Poynting's theorem) [memory: high]. - R. P. Feynman, R. B. Leighton, M. Sands, The Feynman Lectures on Physics, Vol. II, the chapters on induced currents and the laws of induction (ch. 16-17), including the "exceptions to the flux rule" that separate motional and transformer EMF [memory: high]. Free online at feynmanlectures.caltech.edu. - J. W. Nilsson and S. A. Riedel, Electric Circuits: chapters on sinusoidal steady-state analysis, AC power, and magnetically coupled circuits (mutual inductance, ideal transformer, reflected impedance). - S. J. Chapman, Electric Machinery Fundamentals: the transformer chapter (ampere-turn balance, magnetising current, leakage, equivalent circuit). - P. Horowitz and W. Hill, The Art of Electronics: practical inductors, transformers and measurement. - G. I. Cohn, "Electromagnetic Induction", Electrical Engineering (AIEE), May 1949, p.441 ff. A copy cleaned up by hyiq.org is in research/files/.

The inventor's own key works - C. Sykes, Guidelines to Bucking Coils. Lenz's Law Free Power Extraction, v2.5 (February 2015) and v2.7 (26 February 2016), 49 pp. Original URL http://www.hyiq.org/Downloads/Guidelines%20to%20Bucking%20Coils.pdf (site not responding on 2026-10-06; Wayback captures 2015-2024). Copies in research/files/. - C. Sykes, Self-Assisted A Vector Potential Oscillation, V1.0, July 2011, 10 pp. Copy in research/files/. - Videos (channel "Aboveunity and hyiq Research Channel", @SweetSQM): the project video https://www.youtube.com/watch?v=Z-V1z2TdQJA; "Explanation of Partnered Output Coils" https://www.youtube.com/watch?v=iUy_9jDKDvo; "Partnered Output Coils - Why it's Important to have Current Flowing!" https://www.youtube.com/watch?v=phVE5V3pCTc; first device https://www.youtube.com/watch?v=iJsVSMQqCOM; second device https://www.youtube.com/watch?v=ZhQgch4L5XY [web, per the Investigator; not watched in this review]; and "Video 3", https://www.youtube.com/watch?v=zSadMqEnO7I, listed in the forum opening post with the others [read: deep/overunity-15395-p1.md, #0; not watched in this review]. - Forum thread "Partnered Output Coils - Free Energy", overunity.com topic 15395, opened 16 January 2015; page 1 on Wayback: https://web.archive.org/web/20150923150619/http://overunity.com/15395/partnered-output-coils-free-energy/ (copy in research/files/). Its opening post points to the earlier thread overunity.com topic 14974, "Magnet Myths and Misconceptions", "from about page 76 on" (not read). - The inventor's forum, https://www.aboveunity.com/ (Replications category; not responding on 2026-10-06; Wayback 2022 capture https://web.archive.org/web/20220122043807/https://aboveunity.com/cat/replications/).

Related, by others - Cyril Smith, "Bucking Coils produce Energy Gain" (2015), https://www.scribd.com/document/572304933/Bucking-Coils-Produce-Energy-Gain2 (not read in this review). - Earl Koenig, US 4,806,834, "Electrical circuit for inductance conductors, transformers and motors" (1989), https://patents.google.com/patent/US4806834A/en.

---

Answers

1. V₂ = 3 × 24 = 72 V. I₂ = 72 / 360 = 0.2 A. Ideal primary current I₁ = 3 × 0.2 = 0.6 A. Power: 24 × 0.6 = 14.4 W in, 72 × 0.2 = 14.4 W out. The load is reflected to the primary as 360 / 9 = 40 Ω, and 24 V / 40 Ω = 0.6 A, the same answer.

2. With identical coupling, the induced EMFs of the two coils are equal and, in a series-opposed connection, cancel around the loop. Almost no current flows, so the coils' ampere-turns are near zero and nothing is reflected to the primary. The lamp stays dark (Step 4, case B3: zero load current, primary at its no-load 0.191 A). The primary does not "see" the load because the load receives no power.

3. No. For any waveform, true-RMS V × I is at least the real power, so the input reading (12 × 1.5 = 18 VA) can only overstate the input; the real input is 18 W or less. The output on a pure resistor is real power, 30 × 0.4 = 12 W. A false gain would need an overstated output (V × I on a reactive load, or meters that are not true-RMS on a distorted waveform), an understated input (a DC supply feeding pulses, read on an averaging meter, or a supply whose voltage sags), or an unmetered second source (battery, charged capacitor, magnet).

4. The two coils together deliver P₂ + P₃ = (N₂i₂ + N₃i₃) dΦ/dt = 0 × dΦ/dt = 0 at every instant. If one coil delivers power, the other absorbs exactly that power: one coil is driving the other. Sykes quotes Paul Raymond Jensen saying it in words: "consider each secondary winding as acting as the primary winding for the other secondary winding when an output current is drawn" [read: p.34]; and Sykes writes "One Partnered Output Coil is a Generator (Charge Separation) for the other Partnered Output Coil" [read: p.39].

5. The resistance in the loop is 7.8 + 1 = 8.8 Ω. τ = L/R = 0.231 / 8.8 = 0.0263 s = 26.3 ms. The current reaches 90% at τ ln 10 ≈ 2.30 × 26.3 = 60 ms, and 95% at 3τ = 79 ms. A 75 ms charge time is what an ordinary 0.231 H inductor does; it is the back-EMF, ε = −L dI/dt, and shows nothing beyond inductance (and nothing about energy gain). (Forgetting the sense resistor gives 29.6 ms and 68 ms: the same conclusion.)