Overunity Parametric Oscillator?

[neptune]

@Sprocket . Thanks for trying that experiment . My idea of the 2 microwave transformers is at best a crude lash-up , but it is interesting to note that you got some energy out of it with no input except switching .As nul-points says , the switching frequency needs to be twice the resonant freq of the tank circuit .Ultimately , a better set up would be a custom wound inductor on a core with various taps , so the percentage of the coil that gets shorted can be varied . It seems to me you already have energy from nowhere .Now optimisation is the name of the game .I suggested using the secondaries as a test point , but with a custom inductor , a low value resistor could be used for current measurement.Note aso , on Naudins site , the circuit by Fred Apps sing a bifilar coil . One could also look at varying the capacitance instead of the inductance , by having 2 caps in series , and shorting one of them .

[Sprocket]

Sorry about that, I thought you said half the frequency!  That said, I was checking doubling as well when I tried the 2 (maybe) and 4.7uF (definitely) capacitors.  All produced more or less the same output.  The output itself I had presumed was simply induced from the copious amounts of mains-tied wiring in the average house, though why I see 200Hz rather than 50Hz is a mystery.

I have a feeling that the change in inductance or capacitance would need to be 'dynamic' as opposed to a step-function you'd get from shorting.  Of course I bow to Naudin if he says differently.  Someone posted a pic of that large spinning aluminium capacitor earlier - this I could see working!  I used to have several of those vintage aluminium 500-1000pF tuning capacitors from radios at one time - lost now though - but they would be interesting to play with, even though quite a lot of engineering would probably have to be done to get it to spin freely. 

[neptune]

@Sprocket .I feel that if this is going to work , it is going to be very critical on tuning . Ideally the switching rate needs to be exactly twice the resonant frequency , and correctly phased .So the points of transients of the square wave need to coincide with the peaks and crossover points of the output sine wave . Once the principle is proved to work , some kind of feedback mechanism could do the job .Re your theory of dynamic change as opposed to step change . Your theory is as likely to be valid as Naudins or anyone elses .This could be done by varing the capacitance [see my last post] and substituting a load of varicap diodes for one of the fixed caps .Now apply a driving signal to the varicaps  which is like a sine wave superimposed on a DC voltage so the signal never goes negative . Frequency of said sinewave to be twice resonant frequency of tank .

[nul-points]

hi guys

my brain has obviously been in 'park' whilst you've been discussing using two inductors in series - you're right Sprocket, the L & C have to be the same single component whose parameter gets changed - you can't switch another L or C in or out because the stored field has to exist within the component to be modified

so the sequence is something like:
- apply signal to store field in component value X
- alter value X to value Y
- recover new signal from stored field in component value Y

i think that its going to be easier for you to use L as the parametric component because varicaps have such a low capacitance that your frequency would need to be 10s of MHz, or more likely at least 100MHz

variable rotor capacitors would enable a lower frequency, but still probably in the range units-10s of MHz

as you say, you'll need to tap the coil at a suitable point (or mount two coils on the same core)

hope this helps
np


http://docsfreelunch.blogspot.com

[romerouk]

@giantkiller
Where did you get the picture from Magnacoasters?