Caveat: My reply, interspersed with the message, is too long. Oh, well. On 02/13/2010 04:35 PM, Harry Veeder wrote: > > Stephen A. Lawrence wrote: > >> >> Lovely page! Thanks, Harry! > > > You are welcome. > >> JLN has done a really clear job of describing the effect, well >> enough that it can be reproduced and fully analyzed, with, as far >> as I can see, no hidden tricks. >> >> Now, what can we say about his page? >> >> First, he measures the inductance of the coil, and observes that >> it's lower when the magnets are nearby. OK, well and good; as I >> understand it that's because the core is saturated and its >> permeability has hit the skids. >> >> Next, he shows the voltage, current, and power curves to >> energize/deenergize the coil with and without the magnets present. >> He asserts, several times in a couple different ways: >> >>> KEY 3 : The electrical power (Current * Voltage ) needed to >>> energize the toroidal stator coil at the TDC position is EQUAL to >>> the electrical power for the REF position and this is fully >>> independant of the position of the magnet of the rotor Vs the >>> toroidal stator coil. The electrical input power is fully >>> decoupled from the output mechanical power. >> >> Now it may not matter with regard to the final analysis of this >> motor (which probably must depend on calorimetry), but it's >> interesting to note that this several times repeated assertion is >> FALSE. This can be seen by simple reasoning, and by looking at his >> curves. >> >> First, simple reasoning: When the magnets are present, the >> inductance of the coil is lower. So, by definition of inductance, >> when the voltage is turned on, the current is going to rise and >> fall *faster* with the magnets present than with them absent. That >> means total power going in during turn-on is going to be higher >> with the magnets present than with them absent, and total power >> going in during turn-off is going to be lower with the magnets >> present than with them absent. Consequently, power consumed is >> going to be larger if the magnets are brought to the coil, the >> power is turned on, the magnets are removed, and power is turned >> off, than it would be if the magnets were either left far away >> throughout the cycle, or were left adjacent to the coil throughout >> the cycle. >> >> Second, look at the curves: The power curve, shown most of the way >> down the web page, is clearest on this point.
The graph in question appears to be a hand-overlaid plot of two scope traces. It was replaced with an updated version after I wrote the above statement. The original version appeared as I described; the later version does not. Assuming the new version is a correction, presumably because the two graphs were slightly off register in the original version, the earlier one shows an artifact of the overlay process, rather than an actual effect, and that *artifact* is what I was describing. Sigh... I haven't been sufficiently motivated to write to JLN about this. If I suddenly find time hanging heavy on my hands and decide to spend some of it winding coils and gluing up magnets I'll definitely give him a shout. >> >> As with all magnetic shields, the only place where you can see any >> power being consumed is as the shield is switched on and off. Look >> at the rise and fall -- don't look at the flat peak, it's just a >> red herring. > > > The claim of no BackEMF means the coils are unaffected by the > magnets. Assuming this claim true, Which it surely is. The cores are affected, but the coils are not; that's an easy proof, based on well tested aspects of Maxwell's equations. > wouldn't the power required for > saturation of the core decrease when the magnet is closer to the > coil? Um ... depends. If you mean "power pumped into the system during the time it takes to saturate the cores" the answer is probably yes, because the saturation will probably take less time when they're already in a B field. But if you mean "power pumped into the system during the 0.5 mS following coil turnon" the answer is NO -- the greater inductance when the magnets are far away means more current flows when the magnets are nearby during turnon, and so more power (I*V) will actually flow into the coil when the magnets are nearby. This is *exactly* what JLN's inductance meter measures, though of course its display is calibrated in henrys rather than amp-seconds or watt-seconds. > If this is also correct then it is inaccurate to equate the > saturation of the core with a form of shielding. Oh, I don't know. I'd call it a "shield" in that it shuts off the effect of the outside field. You could call it a cloak, if you prefer, but I don't see the significance in the distinction. I mean, you turn the coil on and the box with the magnets drops to the floor -- the coil's core is totally dead as far as the magnets are concerned! How much more "shielded" can you get? > > Shielding implies resistance to penetrative forces, so a shield is > only as effective as the force it can block, consequently the power > requirements of a shield are proportional to the external penetrative > forces. I disagree. The power requirements of a shield are, a priori, unknown. It is an *assumption* that the power requirements must be proportional the force against which it's shielding, and in fact I don't see the reason for that assumption. > In the case of a permanent magnet motor, a stronger magnet > implies more magnetic force aimed at the shield, so the shield will > require more power to block the magnetic force. Consequently, a > shield's power requirements are potentially unlimited. That is, again, an assumption, and whether it's correct for a given shield depends on how the shield is implemented. In general, shields -- the physical sort -- require no power at all once they're in place. It's only moving them into place and moving them away again that takes the power. > > Many materials are inherently invisible to magnets, since magnets are > not drawn to them. However, the core in the coil is not inherently > invisible, so it requires a cloak instead of shield. Unlike a shield, > the power requirements of a cloak are limited since the intent is to > only make the core appear "invisible" or "insensible" to the magnets. > > > Or is this wrong? The distinction you're drawing between a "cloak" and a "shield" is interesting but I think it hides the fundamental similarities between this design and all magnetic shield perpmos. They *all* share the very important trait of drawing no power (to operate the motor) while the shield is stationary; only the effects during shield motion really matter. Furthermore, I'm not so sure that power requirements of this "shield" are independent of the applied B field. The B field penetrates the coil (as you pointed out) and certainly affects the core. Thus, in order for the coil to block the action of the B field, it must totally overwhelm the effect of the external field within the core. Otherwise, the field of the core would be somewhat aligned with the field of the external magnet, and they would continue interacting. Consequently, a stronger external magnet will necessarily require a stronger internal field to overwhelm it. In fact, this may be part of what makes this thing so hard to analyze from the scope traces -- the current put through the coil has been set high enough to totally swamp the effect of the external field, not just "barely suppress" it, and in consequence we've got a *very* large "background signal" from which we're trying to pick out a very small "shield shifting" signal. > > >> Now, the other issue is warming of the core. As I understand it, >> when the core is saturating, things are not behaving "elastically" >> and some energy is being lost to heat. I *think* that amount is >> different depending on whether the core starts out saturated (by >> the external magnets) or doesn't. That heat must be measured to >> get a full energy balance of the motor, and of course JLN hasn't >> done that in this series of experiments. > > Any heat produced is part of the output energy so knowing this heat > loss should support the claim of overunity rather than detract from > it. Not true. Steorn's claim is that heat loss = electrical power in. That is true if the motor is not turning. The thing which I *think* may be true -- though the longer this discussion goes on the less sure I am -- is that *LESS* heat is lost to the cores during motor operation than when the motor is stopped. That is to say, it is waste heat which is being diverted to power generation. The motor gets less hot while running than it would otherwise. This is what Bill Beaty was saying, also, if I understood him correctly. But after realizing that the reduced induction means there *must* be more power going in when the coil is energized while the magnets are close, and after realizing that I don't understand the saturation phenomenon very well at all, I decided I'm not so sure. I sent part of this off to a serious EE I know but I haven't heard back (and haven't phoned him recently). If he has an interesting opinion I'll relay it. If I can figure out how to calculate the expected additional power cost to energize the coils when the magnets are nearby I'll post it (but don't hold your breath, I have no idea how to get a handle on that). > >> But once again, I'd like to say I think this is a great page; by >> putting everything down, in detail, with measurements and >> specifications, JLN has made it possible to fully analyze exactly >> what is going on, and determine once and for all where the energy >> is going and where it is coming from. Excellent! >> >> > > Harry > > > __________________________________________________________________ > Yahoo! Canada Toolbar: Search from anywhere on the web, and bookmark > your favourite sites. Download it now http://ca.toolbar.yahoo.com. >
