Perpetual motion machines do not work, and the reason is not that nobody has been clever enough yet. Two laws of thermodynamics rule them out, and in every documented case the device has either concealed an energy source, run down when watched carefully, or relied on a measurement that was never made properly.
Two distinct impossible claims
Machines of the first kind claim to produce more energy than they consume, violating conservation of energy. This is the overbalanced wheel, the magnetic motor, the device that charges its own battery while driving a load. Machines of the second kind claim only to convert heat from a single reservoir entirely into work, which does not create energy but violates the second law by decreasing total entropy. A ship that propels itself by cooling the ocean is the classic example. The distinction matters because inventors often dismiss the first objection and then unknowingly propose the second. Both classes have been attempted for centuries, and no device in either category has survived independent testing.
The losses builders systematically underestimate
Real mechanisms lose energy through routes that are easy to overlook individually. Bearing friction persists even in precision assemblies. Air drag rises steeply with speed. Materials flex and recover imperfectly, dissipating energy as heat in every cycle. Magnets moving near conductors induce eddy currents that oppose motion. Electrical windings have resistance, magnetic cores have hysteresis and eddy losses, and every switching element drops voltage. Sound is energy leaving. A builder who accounts for the largest loss and neglects four smaller ones can convince themselves that a small measured surplus is real when it is actually the sum of measurement errors. This is why efficiency claims require calorimetry or careful input-output measurement rather than an estimate. Magnetic proposals deserve a specific note. A static magnetic field does no net work around a closed loop, so an arrangement of fixed magnets cannot drive continuous rotation, however the geometry is shaped.
The wheel that fooled Europe
The best-documented historical case is the wheel demonstrated in the early eighteenth century by Johann Bessler, who used the name Orffyreus. It was exhibited to serious observers, including a well-attended sealed-room test lasting weeks, and it impressed people whose judgment was not negligible. Bessler destroyed his machines and never disclosed the mechanism, and a former servant later claimed the wheel was turned secretly from an adjoining room, a claim he disputed. The episode is remembered as unresolved in its details rather than as a demonstrated success, and no working reconstruction has ever been produced. It illustrates the pattern that repeats across the history of this field: impressive demonstrations, refused independent examination, and no surviving device. A second recurring pattern is the device that reportedly works until an independent observer arrives, at which point a critical component is said to have failed and the demonstration is postponed indefinitely.
Why demonstrations convince people
A flywheel with good bearings runs for a long time, and long is easily mistaken for indefinite. A device connected to mains power through an unnoticed cable, or running from a hidden battery, will perform beautifully for an evening. Measurements on switching circuits are particularly treacherous, because instruments that average incorrectly on distorted waveforms can report input power well below the true value, producing an apparent efficiency above one hundred percent. Observers also tend to watch the part of the device the demonstrator directs them toward. The defense against all of this is the same: seal the system, measure total energy in and out over a long period with instruments appropriate to the waveform, and let someone else design the test.
The right response to a claim
Skepticism here is not closed-mindedness. Conservation of energy is among the most thoroughly tested principles in physics, underpinning chemistry, engineering and astronomy, and a genuine violation would be enormous news. The appropriate response to a claim is therefore a request for an independent, instrumented, long-duration test with the device sealed and the measurement designed by a third party. Genuine discoveries survive that. Patent offices in several countries decline applications for perpetual motion devices without a working model, precisely because the category has consumed so much examiner time historically. If someone is soliciting investment rather than requesting a test, that itself is diagnostic.
What to check before you act
- Ask whether the claim violates energy conservation or the second law.
- Require a sealed, long-duration test with independent instrumentation.
- Check that power measurements suit the actual waveform involved.
- Treat requests for investment ahead of testing as a warning sign.
Common questions
What about devices that run for years, like some clocks?
Atmospheric clocks draw energy from temperature and pressure changes in the surrounding air. That is a real external energy source, not perpetual motion.
Could new physics allow it?
Conservation of energy is tied to deep symmetries and has survived every test. A violation would require overturning a great deal that is independently confirmed.
The practical takeaway
When you meet one of these devices, look for the missing loss or the hidden input rather than the clever mechanism. Insist on a sealed long-duration test with instruments chosen by someone other than the inventor, and treat a refusal as the answer.
References and further reading
- Department of Energy: measuring real energy use of equipment
- Department of Energy: energy efficiency fundamentals and resources
- ENERGY STAR: how efficiency claims are verified against test procedures
Numerical scenarios are illustrative unless identified otherwise. Follow the exact product instructions; component ratings and local installation requirements can differ.
