Power From Magnetism -- OverUnity Motor Design

Reading the report

This guide reads Harold Aspden's Power From Magnetism -- OverUnity Motor Design (Energy Science Report No. 9, Sabberton Publications, Southampton, 6 November 1996). The underlying source is a selected-excerpt capture of the report; intervening text is marked as not supplied, and Parts I and III are only partially represented. Nothing below should be extended into a summary of passages that were not provided.

The report is not a laboratory report of a working free-energy machine. It is a design proposal with a supporting theoretical argument, written to interest government and the electrical supply industry in funding a large machine of a particular kind. Aspden presents the multi-megawatt concept first, deliberately, and only then describes the small bench machine he actually built. The single most important thing to carry through the reading is Aspden's own concession: the small machine did not demonstrate the effect he claims is achievable, and the case for over-unity rests on scaling up.

The claim, stated as a conditional

Aspden frames his thesis as a recipe rather than a measurement. Given a magnetizing winding with no resistance loss and no significant ferromagnetic inductive reaction when carrying alternating current, and given a motor running at a synchronous speed set by the pulsation rate of that current, he argues that an over-unity motor can be designed. The energy source is said to be the aether, and that part of the argument is deferred to his book Aether Science Papers rather than developed in the report.

He is explicit that the bench machine does not satisfy the conditions: its windings are resistive and do involve inductance, though he says these are minimized by careful design. The stated purpose of the report is therefore not to report a working over-unity machine but to argue that one could be built at scale, and to interest funders in that scale. The multi-megawatt design is presented first for that reason.

The machine: what "ASPDEN" names

The acronym expands to Asymmetric Shaded-Pole Dynamo-Electric Negentropy machine, and each element carries a specific design meaning in the excerpts.

Negentropy is Aspden's term for over-unity operation. He reasons from the usual expectation that entropy — a notional quantity representing heat Q degraded by temperature T, or Q/T — always increases, so negative entropy implies reversal and a regenerative action.

Asymmetric refers to tilting the stator pole laminations so the machine favors one direction of rotation. Aspden attributes the topic to Energy Science Report No. 7.

Shaded-pole is qualified carefully. The orthogonal disposition of laminar planes between stator and rotor poles accounts for the name, but Aspden says this is not strictly a shaded-pole design in the normal sense. Conventional shaded-pole designs, he argues, rely on flux penetration that generates unwanted inductive heating; the current producing that heat shifts the phase of the emergent flux and creates the drive, which he calls inefficient. His arrangement instead aims to keep flux from leaving the poles in a direction that retards the drive.

The distinctive structural feature is the single helical (solenoidal) magnetizing winding, wrapped around the whole body of the machine rather than fitted to individual pole pieces. This lets the winding be large and lets the magnet system, rather than the winding directly, feed energy to the pole gap. Magnets sit in the rotor so that the soft-iron stator bridging yokes experience the applied field; Aspden gives effective permeabilities of about 100 for the yoke iron and close to unity for the rotor magnets, which he says makes it feasible to run the motor from one all-enveloping winding. He notes that a non-conductive magnet such as ferrite avoids eddy-current losses during flux reduction, whereas an alloy such as Alnico would be a problem.

The a/b/c/d argument: the report's central mechanism

Part III is where the quantitative case is made, and it is a geometric argument on an idealized curve. Aspden states plainly that the B–H curve in Fig. 9 has been idealized into straight-line form to facilitate analysis. On that idealized curve he defines four areas:

• a = HδB over a low flux-density range
• b = BδH over a low flux-density range
• c = HδB over a higher flux-density range
• d = BδH over a higher flux-density range

His stated results are that a and b are equal, because the separating straight line passes through the origin, while c is between 4 and 5 times smaller than d, because the separating line has a much reduced slope and intercepts the B ordinate at a high B level.

The conclusion he draws is that at low flux density the equal sharing of energy between core and air gap offers no power gain, but over the higher flux-density range that equal sharing puts more power into the air gap between the motor poles than is supplied to the magnetizing winding. He identifies the relevant range as roughly 1.5 to 2 tesla in the ferromagnetic core, and notes such densities are routine in superconductor magnet technology where no ferromagnetic core assists. He states there is good reason to expect a machine of this design to operate with at least a 5:1 power gain by polarization bias to a basic flux density of 1.5 tesla and operation over a cyclic range between 1.5 and 2 tesla.

Two things are worth holding onto here. First, the argument is made on an explicitly idealized straight-line B–H curve, not on measured magnetization data. Second, the conclusion is a claim about energy accounting derived from that geometry, not a measurement of an energy surplus. The report's own framing — that the only remaining problem is ohmic heat loss in reaching near-saturation — follows from the geometry rather than from a test.

The vacuum reaction and the gyromagnetic ratio

Aspden's supporting mechanism is a vacuum reaction field. He argues the applied field is really 2H, always opposed by a reaction H in the vacuum, so the measured field is H — and that this reaction explains how the vacuum stores energy as a function of H and returns it to the winding by induction when the 2H influence is switched. He says simple analysis told him the maximum energy transfer into the reacting field occurs when the reaction halves the applied field, and that the kinetic energy density from random motion corresponds exactly to what is associated with the magnetic field. He extends this to ferromagnets: the reaction cannot distinguish fields from windings from fields from electrons in atoms, so the reaction field is of strength B, opposing the primary magnetization.

He then appeals to the gyromagnetic ratio. Textbooks, he says, give e/2mc for electron orbital motion while e/mc is observed, and he asks how the anomaly of a ratio of 2 is explained. He attributes electron spin to Paul Dirac and argues Dirac did not explain how magnetic induction energy can return to electrons in spin, because there is no area described by the motion of a centre of charge and so no flux linkages to capture an EMF. He concludes that Dirac was wrong to offer spin as an explanation for the gyromagnetic properties of an iron rod.

This passage needs careful handling. The g = 2 observation is established physics. The claim that Dirac's spin explanation is wrong is Aspden's own interpretation, offered as part of his argument, and is not established by anything in the excerpts. It is also peripheral to the machine design: the a/b/c/d argument does not depend on it, and the report's practical claims do not either. Read it as an illustration of how Aspden motivates his vacuum model, not as a result.

The reported test data, and what it does not show

The test setup: a d.c. motor coupled to the test machine, the latter carrying the all-enveloping helical winding. Aspden reports 50 Hz tests and electronically pulsed tests done at an early stage, and a later vane switch — a single 8-pole electrical steel rotor lamination whose teeth swept a gap in a sensor with a small magnet and a Hall effect switch — used to control commutation. He notes the vane switch was not fitted when the reported data were obtained.

The reported figures:

• Only half the test machine winding carried drive current; the other half supplied the induced EMF for electronic switching.
• The combination was run at 800 rpm and then 1350 rpm, just powered by the d.c. drive motor, so the test machine was a load.
• Electrical power input was 3.331 W and 5.255 W respectively.
• The d.c. motor was rated 68 W at 5100 rpm on a 12 V input, of a model-boat type with permanent magnets, with efficiency about 50% over its main working range. Two such machines coupled back-to-back recovered between 25% and 30% of the d.c. input.

Aspden's stated objective was to see whether the machine derives any drive at all from current pulsing of the helical winding and to get some measure of power gain. He notes that at these speeds the fluctuating pole-gap magnetism would involve hysteresis loss, eddy-current loss, and a stronger drag from the retentive property of the stator bridging yokes during pole separation than the forward drive gained during pole closure.

The critical reading point: 3.331 W and 5.255 W are input measurements with the machine acting as a passive load. They are not output measurements, and no efficiency figure is given. They cannot by themselves support an over-unity claim. Aspden's own concession is stronger than any external criticism: quoting Report No. 7, he writes that one cannot just build a motor such as that illustrated in Fig. 9 and expect it to deliver over-unity power on demand, and that the magnetic flux density across the pole gap was far below the level where over-unity operation can really reveal itself in a dominant way. His position is that the prototype could be scaled up, and that the tests could verify design feasibility. He describes the ability to operate the machine by control of the external winding as the proof he needed to see purpose in advancing the project.

His scaling argument is that I²R loss can be halved by using both windings for drive and a separate signal source for control, that more turns or thicker wire can be used, and that because loss scales as the square of linear dimension while drive power scales as the cube, larger machines make the loss small. This is a design argument, not a measurement.

Patents as evidence

Aspden states that the chopped a.c. waveform technique with a reluctance machine biased near saturation is already the subject of his U.S. Patent No. 4,975,608 dated 4 December 1990, and that the single-helical-winding features have evolved since and are the subject of independent patent applications — pending, not granted. The patent's abstract title page and claims were in an appendix to the first version of the report but are omitted here.

This is the same pattern documented elsewhere in this wiki: a granted patent and pending applications cited as validation for a performance claim. patent-application-as-evidence sets out why a patent is not proof of effectiveness — a patent establishes that a technique was disclosed and examined for novelty, not that it works as claimed. Whether US 4,975,608's claims actually cover the described technique is asserted by Aspden and is not verified in the excerpts.

On superconductivity there is a visible tension. Aspden states that superconductivity is not essential to over-unity operation, yet repeatedly invokes it as the enabler for eliminating the ohmic loss that he identifies as the only remaining obstacle. He also notes that past superconductive machines used the windings in the rotor of an alternator to produce a d.c. field, and that this has no bearing upon the over-unity machine proposed in the report.

The eddy-current anomaly and CAD

Aspden recounts asking a university lecturer who specialized in electric machine design and wrote CAD programs how he allowed for the eddy-current anomaly, and being told the lecturer had never heard of it. Aspden says he spent three years of Ph.D. research on that subject in the Department of Electrical Engineering at Cambridge University. He argues that the precise variation of B with H in iron is not something one relies on from electrical steel manufacturers' specifications, and that his design would not be something that could perform in the manner predicted by use of those programs for computer-aided design. He asks whether a commercial CAD program could really predict the over-unity performance of an electrodynamic machine.

This is a claim about the limits of standard modeling, offered as a reason the design would not be predicted by conventional tools. It is not itself evidence that the design works.

Where this sits in the collection

The Aspden report belongs with the other free-energy claims documented here, where a named device or technique is asserted to deliver more output than input and the available evidence does not establish it. The closest structural parallels are Bedini Claims vs. Evidence: The Patents, the Inventor's Disclaimers and the Absence of an Energy Balance and Tetrasilver Tetroxide: The Electron-Jumping Claim, the Patent Set and the Bounded Evidence. The two framing concepts that apply most directly are patent-application-as-evidence and reported-vs-measured-performance: Aspden cites a granted patent and pending applications, and reports input measurements rather than measured over-unity output.

The claimed mechanism is distinct from earth-field-energy-extraction-claim (Hendershot), which posits energy drawn from the earth's field rather than an aether/vacuum reaction. The two should not be merged; they are different claims about different sources.

Aspden's claim is framed as testable in principle — scale the machine up and the effect should reveal itself — which is worth noting against the discussion in falsifiability. A claim that can in principle be tested by a specified scaling is not the same as a claim that has been tested, and the report is explicit that the specified test has not been performed.

Open questions

1. No output-power or efficiency figure appears in the excerpts. Is one present in the omitted text?
2. The status of the independent patent applications for the helical winding is unstated.
3. Whether US 4,975,608's claims actually cover the described technique is asserted, not verified here.
4. The multi-megawatt design of Part I is referenced but not reproduced in the excerpts; its specific features cannot be assessed from this material.

References

• Harold Aspden, Power From Magnetism -- OverUnity Motor Design, Energy Science Report No. 9, Sabberton Publications, Southampton, 6 November 1996. ISBN 0 85056 0241.
• US Patent No. 4,975,608, 4 December 1990.
• Harold Aspden, Aether Science Papers (deferred to for the underlying aether argument).
• Energy Science Report No. 7 (quoted on the earlier test system and its control).
• Energy Science Report No. 8 (Correa and Correa).
• Energy Science Report No. 3 (announced refrigeration/air-conditioning technology).