It appears that the Perepiteia Motor is nothing more than a hysteresis brake. Placing the steel rods (wound by coils) near the spinning magnets induces alternating magnetic flux within the rods, the resulting magnetic hysteresis causes drag on the rotating disk, and heat losses within the steel rods. Shorting out the coils effectively shields the steel rods from the disk's magnetic field, eliminating the hysteresis drag. This causes the motor to speed up - but not as much as it would if the steel rods were removed completely.
BUSTED!
Hi:
blindsangamon is correct. This is a common phenomenon regards "generators", but one often not commonly known about if you are not working with AC motors and generators all the time. So the professor at MIT may not have this practical engineering savvy to identify the issues at first glance.
An hysteresis brake is one way to describe the apparently anomalous increase in speed when you short the generator coils. What EVERY ferromagnetic core does when exposed to varying magnetic fields is to have it's domains rock or flip direction in accord with the magnetic field changes impinging on them.
This consumes power in the "friction" between domains as they sort of scrape past each other. It results in the material heating up. In addition to this hystersis "loss" is an eddy current effect within bulk steel from the very same time varying magnetic fields, also making heating of the core. These two effects combined are commonly termed "core loss".
Core loss produces a reaction torque in a generator, in that the domain "friction" resists their aligning with the external field - causing more drag torque. Eddy currents make magnetic fields which oppose the fields making the eddy currents too, making more drag torque.
Now "core loss" in any ferromagnetic core material is directly proportional to the induction, B. Put another way the higher the delta flux density, the more core loss you get. (it is also proportional to the frequency, but let's assume a constant freq here, even though it is not at a constant one - it speeds up and slows down, again a neophyte mistake - you must measure things here at common speeds/freqs to make comparisons accurately)
And the induction, B is then what produces the coil voltage via Faraday/Lenz laws. That is voltage is the time derivative of delta flux. So people, when you short a generator coil and it's voltage drops to near zero, you can be certain that the delta B within the coil's core is also near zero!
So if you started with a delta B of say 1,000 gauss at no load on the coils, and your core material produces say 15 watts of core loss per pound of core (solid steel is in this ballpark, which is why we laminate special steels for transformers which takes the core loss down to about 2 watts per pound) then you'd have some serious drag torque experienced by the drive motor with coils open circuit.
Now if you short the coils and drop the delta B down to say 10 gauss, you have REDUCED the core loss by a factor of 1000/10=100 times less core loss when shorted than when open circuited!
This means 100 times less drag torque felt by the drive motor! (therefore the common shaft speeds up when coils are shorted, duhhhh)
This is amateur hour gone mad - both in the videos and mostly in these lists! Which does nothing but hurt the cause of getting O/U to the masses in my view, as it simply reinforces to the powers that be in the scientific community that it is a bunch of flakes and idiots making these claims!
Now I will also say, that heavily loading certain geometry of generator, can produce some gain. I have several examples on the bench which do. But they are proprietary and I don't care to share this with lists. BUT you have to do proper energy/power balances to measure this gain. And you have to endeavor to reduce core losses to a minimum and account for core loss change when you heavily load the coils too.
I have one which gets a gain in excess of the entire core loss value, both eddy and hysteresis - therefore the gain cannot be from this artifact that plagues all coil/core systems. But it is a modest gain, and yes the rotor does want to speed up. But you have to manage this speed, and measure the loaded and unloaded condition at the same shaft speed, because friction and windage change too when speed changes.
Then you have to measure True power at the shaft input via torque sensing and speed, against True output power, including friction, core loss, coil heating and direct electrical output for a complete energy/power balance. In fact there is an IEEE protocol for doing a complete power balance on motors and generators, which includes all these things.
This person did few if none of these things properly and is delusional about the apparent speed increasing meaning it is O/U. There could be a small amount of gain in his sloppy and amateurish system, but it is completely overriden by mundane, conventional effects as "blindsangamon" correctly points out.
In short his videos are a waste of bandwidth and show nothing conclusive except that author is both ignorant of EM, and of how to perform measurements on said systems.
Sorry for being so terse with you folks, but it is very annoying to watch so many people do harm to the cause by spouting off without really having a grasp of conventional ElectroMagnetics. Both amateur's like in these videos, and indeed a large percentage of the armchair critics populating these lists! Do your homework before putting foot in mouth!
There's a few rational voices out there, blindsangamon being one, and most of you then deride these voices with nonsense and blind faith!
here's a simplified protocol for measuring a generator's complete power balance:
Pick or know the optimal final speed of the system. Use only this shaft speed for all measurements.
1 Measure all parameters in a generator "no load" condition including:
2 Friction alone, meaning with no magnets or mag fields acting on the cores.
3 Then include the mag fields and measure the input drag power (torque times angular velocity).
The difference between 3 minus 2 is the core loss at no load.
4 Measure the DC resistance of all coils as they would be connected in a loaded condition (i.e. series or parallel).
5 Load the generator at the same speed as the no load tests.
6 measure input power via torque times speed. (Newton-meters times RPM times 0.1047 = shaft power in watts)
7 measure True output electrical power. Not with DMM's. but with appropriate True Power meters or analyzers.
8 measure coil current, and calculate coil's "Joule heating" via I^2R.
9 measure and compare coil voltage compared to no load voltage for a ratio with which to discount core loss.
Then take the loaded input shaft power in watts as INPUT to system.
Against this Input, you add the following:
a electrical output in watts
b friction in watts
c core loss via no load core loss times the voltage drop ratio (so if no load core loss were 37 watts, and no load voltage was 125V and loaded voltage is 83V, then the ratio is 0.664. Multiply 0.664 times no load core loss of 37 watts to equal 24.57 watts output core loss)
d coil heating via I^2R
Add up item a through d for the total system OUTPUT.
Now divide Output by Input for your COP. (Coefficient of Performance)
Note friction, core loss and coil heat are legitimate outputs.... they heat the room! Useful output is an arbitrary distinction based on subjective criteria. If you want shaft power then heat is not useful. If however you want a heater, then shaft power is not useful! So to know in the absolute sense if a thing is over unity or not, you have to account for ALL outputs in a balance sheet.
That's another pet peeve of mine - those who dismiss everything they deem as "not useful"! Now suppose you had a system which routinely produces 200% more heat output in coil heating and core heating while it turns a shaft as in some newfangled motor. The shaft power COP is only 35%, but overall the system is 200% gainful. These persons I refer to would dismiss this as not being useful because the shaft power is under unity!
When in fact a home heating system would require a heat exchange mechanism to get heat from your machine to the air, thus it requires a pump - moving air or water or both. So you could make "use" of both the excess heating and the shaft power from said system!
My point is at these early stages it is imperative that you measure all aspects even if you may "think" they aren't useful. For complete energy balances and because overunity may not come in the form you wish it to!
DMBoss