Originally posted by Ted Ewert
View Post
So, in your opinion, using a coil to move a magnet does not consume energy because the magnet can move itself back? Sorry, but you have made a gross error here. Not only does it consume energy, but in your application it consumes even more. Not only must it move the mass, but it also must push against the attraction of the magnet. So 100% of the potential energy stored at the 1" location is supplied by your input power.
The problem is, you think that energy is equal and it is not. That stored energy is only a minuscule fraction of what you are applying to the coil. Why to think it would be equal needs to be explained in full detail.
Passing a square wave through an transformer is a very simple thing when you understand the principles involved. Part of the problem with experimenters is that they tend to think in simple terms with the components they use. Thinking that a transformer is just two inductors wrapped around a core is a very simplistic thing and for many applications this simplistic approach is satisfactory. But when it becomes necessary to understanding the precise dynamics involved in passing energy from one winding to another, that is where experimenters often lose there solid footing on reality. Take for example the Pulse Transfomer. These transformers are made specifically for passing squared waveforms. If you remember from your studies that a square wave is the sum of all odd harmonics, you will understand quite readily why they pass through a transformer. Especially, when you consider that there is much more going on than just inductance. Notice in the link above that there is Cps. This is very high in bifilar wound coils and especially those made of square conductors that fill the voids normally produced by round conductors. This capacitance offsets the inductive delay and increases the rise time of the pulse. The Lp1 and Lp2 in the transformer play an important role in keeping the flat line up because they support the field when the change in current begins to stabilize across the top of the wave form. It is like dropping a magnet down a copper tube. The induction stops the magnet, which stops the induction which then moves again and the process repeats. The same is true with that inductance of Lp1 and Lp2. If you could view it with a fast enough microscopic oscilloscope, you would find small ripples across the top of your flat line where the current changes up and down holding the charge. And of course there are all the other factors there that play a part in the wave shape. Then, not even mentioned in that document, there are factors relating to variable permeability in a core material and effects of saturation, both of which can readily impact the wave shape passed through a transformer. For more on understanding the passing of square waves through a transformer see Pulse Transformers. So what do you think your coils would do with a duty cycle of 75% ON? Would it retain its square waveform?
You stated that there is no 'time issue' in your experiment. So have you completely dismissed the dv/dt equation then? What do you think would happen to your output if the on time of your wave form was increased to 2 hours? Would you still see a square wave on the output? Any one who has experimented with magnetics as much as we have will know right away that time is a BIG factor in induction. Just try moving your magnet across your coil real slow and see what the gives you even at 90°.
I have tried to help you understand why your tests failed to produce the effects you expected, and why you experienced something you did not expect (square wave transfer). It has little to do with the misapplication of Faraday's Law but more has to do with the specific dynamics of the system.
Try repeating the test with small mass magnets that you can oscillate at high frequencies and I guarantee you will be surprised at how different the exact same experiment becomes.
Also do my experiment with the magnet drop and prove to yourself that a change in flux density on the axial line of the core will in fact produce a voltage in your coil.
The reality is that the MEG and the FLYNN device will work as shown when properly configured. But as I have stated before, there are losses in the coils that work against any OU gain. There is absolutely no difference between the flux of a PM and that of an EM in a core. They are both identical at full potential. Where the two differ is in the building of the flux. The EM starts at zero, and builds the flux to full potential. The PM is always at full potential. What the MEG and FLYNN device do is provide a way of using the PM full potential to gradually build a flux change in one of the core legs while reducing the potential in the other thus effecting a flux density change in those respective legs using the 'always on' potential of the magnets. This effect is supported by simulations as well.
If you wish to experiment with this device I would like to make the following suggestion. Instead of magnets, use DC solenoids. You can then vary the PM flux by using a rheostat on the DC coils and tune in the best flux ratio for your core that you are experimenting with. Once you find the best value with your DC tuning, use a Gauss meter to measure the value in the solenoid coil and select magnets that match this value. Also, as I have stated before, seek out high permeability core material of a wide enough path to accommodate the flux of 3 sources; each magnet and the switching flux. I think the most common mistake made in this regard is a a bottleneck in the core material at the output transformers that results in pushing the flux outside of the material where it end-runs the entire winding through an air path.
Comment