Or, rather, something amusing from someone at VOTB. See attached image.
Full size version is here (it's too big to attach and send to Vortex): http://i47.tinypic.com/j0fl7s.gifThat's a perpetual motion machine. The weights are arranged so that they exercise more leverage on the shaft as they fall than when they're raised, which should result in a net clockwise torque.
The thing which is interesting about this is that it is an *EXACT* analogy to a magnetic motor (such as what Steorn originally claimed to have), as we will see.
The obvious question is "Does this thing really work??" There are two ways to analyze it.
I) Examine the torque exerted by each weight as it falls, and the torque exerted by each weight as it rises. Make suitable assumptions about the friction and/or gearing of the weights as they "flip" while going over the top. Count the number of weights in each state at each moment ... and so on and so forth, and when you've analyzed *all* the details, you'll have your answer. But it won't be easy, and it certainly will be easy to get a wrong answer!
II) Look at ONE weight, and observe that the (gravitational) energy it takes to lift it to the top of its arc is EXACTLY equal to the energy obtained as it is lowered to the bottom of its path. So, as one weight goes up and comes back down, there's no net energy gain. And that's all we need to know! That's because we know that torque times net rotation equals energy, and if we get an average positive torque out of one weight we'd get energy out too, and we can't, because gravity is conservative. Then, just observe that the forces and torques on the weights and shaft all sum linearly, so if we can't get energy out of one weight, we can't get it out of the system, either. QED.
The situation with magmos is, I said, *exactly* analogous. Analyzing any claimed magmo, we can use exactly the same two approaches outlined above: Either we look at all the details (which tend to be hideously complex) in an effort to determine if the sum of the energy involved in all processes comes out to zero or not -- OR, instead, we just observe that we KNOW (from many experiments) how a simple dipole and a wire interact, we know the individual most basic interactions are conservative, and we know (from a very large number of experiments) that the interactions sum *linearly* ... and so we know that the whole system must be conservative.
The key here is linear superposition. If there's any evidence that superposition isn't linear, the game's off. But AFAIK there isn't.
<<attachment: perpetual-motion-1-tiny.gif>>
