Flynn-C · Source catalogue

A US patent (42 claims, 32 drawing sheets on pp.2-33) describing a permanent magnet (PM) whose two pole faces each carry a pole piece with two path portions, giving two parallel flux paths with a control coil on each. Pulsing the coils in timed sequence, either opposing the magnet's flux in one path (forcing it into the other) or aiding it in one path, switches essentially all of the PM flux between the two paths. Applications: linear reciprocating devices (FIGS. 13, 17), linear motion (FIG. 18), four rotary motor embodiments (500, 580, 650, 740; pp.40-44) driven by LED/phototransistor, Hall switch or brush commutation circuits (p.43), a 'Power Conversion' device 900 in which fixed armatures carry secondary coils (pp.43-44), and two PM-stator/PM-rotor embodiments 1000 and 1030 (pp.44-45). The description argues from the Maxwell pull formula F = B^2 A / (2 mu0) per pole face (F = B^2 A/mu0 for the two faces of an armature): splitting one flux equally between two equal paths halves B in each, so the force falls to one quarter, called 'unexpected' [read: p.36]. Blocking one path doubles the flux and gives about four times the force in the other, with 'one half the power it would require for a coil alone' [read: p.37]. For a reciprocator it says an all-electromagnet device (FIG. 14, U-shaped pole pieces 348, 350 with coils 358, 360 and non-magnetic spacers 352) would need twice the electrical energy for the same force [read: p.39]. FIGS. 3-4 (p.36) state that two PMs in parallel double B at unchanged H, two PMs in series double H but leave B unchanged, and that a coil in series with a PM adds flux only until the PM reaches 'B_max' (the page image shows B_r and B_max; OCR reads 'B_sat') unless a bypass path exists. The bypass and 'excess flux' embodiments (FIGS. 11-12, 15-16; bypass claims 7, 8, 37) and FIG. 16C (flux vs coil current in mA, zero crossing at 'critical point 393', no flux scale or turns count) rest on this picture [read: pp.36-39]. The deep reads found: (1) the 'twice' energy figure is reproduced by the standard linear magnetic-circuit model only if 'electrical energy' means the field-building energy (integral of i dlambda), ratio 1.99 for NdFeB with small gaps, tending to 2 for equal legs; it is stored, not consumed, so it does not support 'more work per unit input energy'; coil MMF and steady I^2R holding power give ratio exactly 1, so p.37's 'one half the power' is not borne out [calc; deep/ read]. (2) The 'saturation at B_max' rule for series coil plus PM is refuted by the linear recoil model (mu_r about 1.05), where the PM acts like an air-like reluctance and B in the magnet can exceed B_r under an aiding coil. The practical advice (use bypasses) is nevertheless right, because the magnet is about 10x the reluctance of a 1 mm gap. FIG. 16C's linear zero crossing and the 'about double' flux on switching (1.98x at the critical current) come out of the standard linear model [calc; deep/ read]. (3) The patent gives no magnet material, no stroke, force, current, torque, speed or efficiency value anywhere. The sole data-like figure is FIG. 16C, and the text does not say it was measured. As read, the patent does not claim output exceeding input energy; it claims better force or efficiency per electrical input relative to electromagnet-only designs. The front page cites six earlier Flynn patents (5,254,925; 5,304,881; 5,455,474; 5,463,263; 5,710,493; 5,753,990) and PM-actuator prior art by Sturman (3,683,239; 3,743,898) and Monroe (3,670,189) [read: p.1]. The claims and motor deep reads found: (4) The 42 claims have 13 independent claims (1, 16, 18, 19, 20, 21, 22, 26, 29, 30, 31, 32, 38) and recite only structure and timing; none states efficiency, gain, COP, output above input, or 'twice' (the only energy word is 'a source of electrical energy') [read: pp.45-49]. Claim groups: bypass = 7, 8, 37; power conversion (fixed armatures with secondary coils) = 12-14 and 38-42; rotary = 20-28, 30-31; linear = 29, 32-36. Three drive modes are claimed separately: opposing only (16), aiding on one coil with opposing on the other in alternation (18), simultaneous push-pull (19, device claim 10). Claims 18 and 19 recite the physical outcome that substantially no PM flux traverses the other path, a testable property but not a performance gain. There are small drafting defects (claim 6 repeats 'first pole piece'; claim 8 depends on 6; claim 26 and 38 antecedent slips). (5) The 'twice' wording is in the description, not the claims: p.39 (FIG. 13A vs 14, energy) and p.44 (FIGS. 45A-45C: flux in the secondary pole-piece portions 'varies by a factor of twice the amount of magnetic flux generated by energizing' the control coils, which is a flux swing, not an energy ratio). The two are different statements and must not be conflated. (6) Pp.40-45 contain no torque, speed, rpm, power, current, voltage, efficiency or loss value and no test of any motor; the only performance statement is the unquantified 'increased efficiency and more desirable torque characteristics' on p.34. The drive circuits (FIGS. 40, 43 LED/phototransistor; FIG. 41 Hall switch; FIG. 44 brush commutator) are textbook rotor-position commutation, described as 'circuitry common in the art' [read: p.41]. By the deep reader's judgement, motor 500 is a PM-biased (flux-switching) switched-reluctance structure, 580 a two-phase PM-biased variable-reluctance motor, 650 and 740 brushless-PM types with stator magnets gated by coils (740 adds a bypass and 'aiding and exceeding' drive that needs coil MMF above the magnet's), 1000 a polarised bistable solenoid actuator (rotary form only mentioned), 1030 a PM-rotor machine described in one paragraph with no timing or geometry [judgement, not the patent's wording]. The text admits start-direction ambiguity (p.42) and the fix is detect-and-reverse circuitry never designed. The described circuits are unipolar low-side drives with no flyback diode shown, yet 1000 and 1030 need reversed coil voltage (p.45), so the drive for those is not shown [read: p.43, p.45; drawings not checked]. Power conversion (device 900) gives no voltages, turns, frequencies or outputs; the patent itself says static flux 'is useless' unless made time varying (p.44), and only the flux change induces voltage, which is bounded by what the control coils can divert. No over-unity statement appears in the claims or in pp.40-45.

Key claims

• Energizing control coils on two parallel flux paths in a timed sequence can switch substantially all of a permanent magnet's flux between the two paths (claim 1; p.37).
• A given permanent-magnet flux can be controlled by a smaller electromagnetic flux (p.34).
• Dividing a fixed flux equally between two equal paths reduces the holding force on each armature to one quarter, not one half (p.36).
• Blocking one of two equal paths doubles the flux and gives about four times the coupling force on the other armature, with one half the power a coil alone would need to make that flux (p.37).
• Switching permanent-magnet flux between two paths needs half the electrical energy of an all-electromagnet reciprocating device (FIG. 14) for the same coupling force, and so gives more work per unit of input electrical energy (p.39).
• A permanent magnet caps added coil flux at its own B_max (series magnets double H but not B), so excess coil flux needs a bypass path that does not pass through a permanent magnet (p.36, p.39; bypass claims 7, 8, 37).
• Two PMs in parallel double the flux density B through an armature while H is unchanged (p.36).
• As the coil current rises, flux on the coil-opposing side falls through zero at a critical point (393) and then reverses, giving a repelling force on a PM armature (FIG. 16C, p.39).
• Permanent-magnet flux-path switching motors (embodiments 500, 580, 650, 740) have increased efficiency and more desirable torque characteristics than many existing motors (p.34; an unquantified assertion, and pp.40-45 give no torque, speed, power or efficiency number or test).
• With fixed armatures and secondary coils, the same switching converts the magnet's static flux into time-varying flux that induces voltage, like a transformer or power inverter (device 900; claims 12-14, 38-42; pp.43-44); the secondary flux swing is 'twice' the control-coil flux (p.44), not an energy gain.
• A two-magnet bridge (embodiment 1000, linear actuator, FIGS. 47-48, claim 29) and a PM-stator/PM-rotor motor (embodiment 1030, claims 30-31) can be controlled by timing coil voltages so the coils alternately aid and oppose each magnet (pp.44-45).
• Claims 1-42 recite only structure and drive timing (three separate drive modes: opposing-only claim 16, aiding/opposing alternation claim 18, push-pull claim 19); none recites efficiency, energy gain or 'twice' (pp.45-49).
• The 'twice' statements are description only: p.39 (FIG. 14 needs twice the electrical energy) and p.44 (secondary flux varies by twice the control-coil flux); they refer to different quantities.
• The motor drive circuits (LED/phototransistor, Hall switch, brush commutator; FIGS. 40-44) are conventional rotor-position commutation 'common in the art'; the novelty claimed is the flux-steering magnetic circuit (pp.41, 43-44).

Relevance: This is the only document in the project and the whole basis for it. The central practical claim is that PM flux can be switched with small control energy to do motive or conversion work, which alt-science circles often cast as a route to free or over-unity energy (not stated in the patent). The patent is also a good teaching case: its force-splitting arithmetic is standard, its energy comparison is right only under one reading, and its magnet 'saturation' rule is wrong while the bypass design it motivates is sound. The claims themselves are structure-and-timing only and make no efficiency or over-unity statement; performance language lives in the description and carries no data.

Mainstream conflict: Little as a design. Flux switching with coils is established engineering (switched reluctance, flux-switching and brushless PM motors, magnetic latching relays; Sturman and Monroe prior art on the front page). The one-quarter force result is plain F = B^2 A/(2 mu0) arithmetic. Two statements conflict with the standard linear magnetic-circuit model: the PM 'B_max saturation' ceiling for series coil plus magnet (refuted by calculation in the deep read), and 'one half the power' read as I^2R holding power (ratio 1 in the model). The efficiency-per-input-energy argument compares static stage forces, not a full energy cycle; the mainstream view is that a PM is a conservative source and the supply must still provide net work plus losses. No over-unity claim is made in the claims (checked across all 42) or in the description pp.40-45. The motor drive electronics and commutation are conventional; the motor efficiency assertion (p.34) is unsupported rather than in conflict.