Every now and then, a putative inventor showcases his perpetual motion machine which he/she claims is truly perpetual, only to find-out that the invention is a do-nothing machine that relies on a well-hidden outside-source of power, and even then, the machine is not able to do any useful amount of work.
The second type, like this Finsrud machine, mentioned in the last part of this video, may be such machine, with an elaborate system of magnets, pendulums and a ball rolling on a circular track. Looking at the way it worked and why it seemed to work for a longer-than-expected period of time (Few days in a row), made me think how this could have been done, all without violating the first law of thermodynamics. To explain my theory, I would use a simpler model:
A ball rolling on a track, passing-by a magnet located near or underneath the track. The ball approaches the magnet with an initial speed V1. The magnet accelerates the ball towards it. At a point closest to the magnet, the magnet allows the ball to attain a higher speed V2, so that V2>V1. The ball passes by, leaving the magnet. The magnet keeps pulling the ball back towards it, so the ball decelerates. When far enough, the added speed is lost and the final speed is, again, V1. This is the case for ideal magnets, no friction and no air resistance. There is no account for gained or lost energy. "Before" and "After" are the same.
The magnets, are reported to lose magnetic field over time. According to Lenz law, a magnet induces an opposite magnetic field at a magnetized object. An opposite field may have a cumulative demagnetization effect, so that the magnet loses a fraction of its magnetic field after each interaction.
The result of a demagnetization, is that the magnet loses a fraction of its magnetic field after each interaction. As it pulls the ball to a speed V2, it becomes slightly weaker. The ball is pulled back at a weaker force when it leaves the magnet, so is not fully decelerated to V1. The ball leaves the magnetic field at a speed V3, where V2>V3>V1. Conclusion: Some of the "lost" magnetic energy of the magnet is converted into a kinetic energy of the ball. The machine is therefore said to draw its energy from the magnets and over time the magnets are demagnetized. In an electric machine, this energy conversion is constantly replenished by electric power. In case of putative perpetual motion machines, it draws the magnetic energy from the magnets until depleted and reaching a halt. The energy gained by the ball, equals the energy of the lost magnetic flux minus the friction losses, so that it's too small to be useful.
Given that the energy needed to roll a ball on a track is very low, does that explain why Finsrud's machine "seems to work" for such a long time as claimed by the inventor? Does this magnetic field suffice to create this illusion? If my explanation is correct, then you could observe the following:
1- If magnets are of poor quality and demagnetized easily, the ball may attain higher speeds but the machine will work for a shorter period
2- If the magnets are of high quality and demagnetized very slowly, then the ball may barely overcome the friction on the track
Edit: here is picture