Pulling energy from the ambient energy field using a coil capacitor

[Belfior]

@itsu:

is there a need for any protective resistor or else for the primary, or is it actually better protected from spikes and other stuff now, that it is an open circuit?

What I mean is that ho is the circuit different from the FG's point of view now that the primary is an open circuit? I don't think FG sees it as an infinite resistance, since the 2 wires in the primary are coupled together and exchange energy

[itsu]

B,

as the diagram in this post shows: http://www.66pacific.com/calculators/capacitive-reactance-calculator.aspx   we have 2.3nF capacitance
between the 2 primary coils.

According to this reactance calculator: http://www.66pacific.com/calculators/capacitive-reactance-calculator.aspx  this means that at 291 Khz,
the FG encounters a reactance of 238 Ohm.

As we are working at near resonance frequency (291 Khz), we see a big influence from the resonante coils onto the FG signal turning it
from a square wave into a sine waved square wave signal (mutual inductance i guess).

This mutual inductance could cause pulses to be fed back to the primary and thus the FG,  i think, so care needs to be taken indeed to protect the FG
when higher voltages are going to be used.

Itsu

[Jack Noskills]

Jack,

i have a hard time following you as you seem to run back and forth between your testcases.

I now reply on your post #219 and specific the below in bold request:


First screenshot below is again the output as show in my post #212 (so NOT #214 as you mentioned) meaning
inputting a 10Vpp square wave at 291Khz @ 50% duty cycle.  (the white trace is the unloaded FG signal used as input)
Yellow is the voltage across the output coil see diagram.
Purple is the CP2 current probe signal
green is the CP1 current probe signal


2th screenshot is same setup, now with 10% duty cycle see white trace again.

3th screenshot is again same setup. now with 0.9% duty cycle, see white trace.  (be aware, the yellow trace amplitude is now set as 2V.  instead of 20V earlier!!)

So with only 10% duty cycle not much signal is left, let alone with 0.9% duty cycle.


Itsu

Thanks for the scope shots again Itsu! Things are clearing bit by bit, slowly but steadily.


Keeping the primary coil capacitor charged seemed to increase the low frequency oscillation and decrease the high frequency oscillation. When lowering the duty cycle to 0.9% the high frequency oscillations became evident and the increment of oscillation amplitude is clearly visible in the green current trace. After the input pulse we can see current oscillations that last about 500ns. The amplitude of this oscillation increase as the duty cycle is decreased. This can be seen also in the current RMS value which grows from 1.743mA (50% pulse) to 1.958mA (0.9% pulse). Note how the low frequency sine wave disappears when going from 50% to 0.9% pulse which should decrease current RMS but the current RMS increases instead! Same increase can be seen in the voltage trace but because of scale difference it is a bit hidden. Voltage oscillations seem to last longer than current oscillations, about 2000ns. Maybe current oscillations get buried in the noise? Based on woopy's scope shot voltage and current waveforms should be in phase.
Short 0.9% pulse results in free oscillation: high frequency and tubular ringdown (with diodes) which means it is not typical LC oscillation. It should have the Coulomb effect: the voltage amplitude of this oscillation should increase four times when current of the input pulse is doubled. Comparing the voltage waveform of the very first short purple pulse from the 3.9 MHz system at 5V which had 200mv peak:

http://overunity.com/17119/pulling-energy-from-the-ambient-energy-field-using-a-coil-capacitor/msg514973/#msg514973

to the most recent 0.9% 10V pulse that had over 2V peak this squared relation (and possibly more) seems to exist. Well, not able to look exactly because of clipping and also diodes were used with 5V pulse so ringdowns are not quite comparable. When doing pulse tests you can confirm if this relation exists. Getting the voltage high with this ringdown means more charge will be present in the output coils and squared relation should exist between this voltage and charged capacitor as explained in the pdf. Resonant rise with this frequency should produce the white spark when the output is shorted. The spark has a distinct look and sound and worth to make a video if you can create it.


Amplitude of the low frequency oscillations depends on the width of the input pulse so this is not interesting. Ringdown is a damped sinusoid which is typical in LC oscillation and it does have the Coulomb effect: 10%/10V voltage ringdown was two times greater compared to 10%/5V ringdown. Possibly this oscillation is caused by the beat frequency between the two coil pairs? I think that getting this voltage high does not mean more charge is present in the output coils. Most likely this voltage cannot charge a capacitor to higher value, i.e. squared relation between this voltage and voltage in a charged capacitor does not exist.


So when doing tests with 10%/1kHz pulse test also short 5-30ns pulses to get the ringdown of the high frequency oscillation. To test for the existence of the Coulomb effect, vary the current of the input pulse by using different voltages: 5V, 10V and even 20V if possible.


One more thing, when testing with ferrite test also how iron core works. Iron is not good with high frequency inductive pulses but maybe it behaves differently with capacitive pulses ?

[Jack Noskills]

Hmmm,

not sure where you want me to add this capacitor and blocking diodes.
Should the cap go as shown in the below diagram?  And what is its value?

What about the diodes, are they suppose to come in the leads next to where i have drawn them?

No diagram in your PDF equals my present setup


Itsu

Use the upper circuit of figure 8 like you once already did

[Jack Noskills]

Use the upper circuit of figure 8 like you once already did

Trying to be more accurate about how to continue with testing..
Get the ringdown of the high frequency oscillation using a short pulse and measure the current from the free coil ends. First use 50% duty cycle pulse and then begin to reduce the pulse length. Note the length of the input pulse when the current is still high. Cutting of the drive pulse while current is high should produce strongest ringdown (current reversal method of capacitive pulsing). With ten meters of wire in the primary, 30ns pulse length should be enough. So the best pulse should be such that current is high after 30ns, over 90% from maximum. With 2 nf capacitance in the primary this might not happen though so adding capacitance should fix this problem (C2 capacitor in figure 9 of the pdf). This behavior of the input pulse can be seen by monitoring the input current. Now get the resonant rise using the pulse that gave the strongest ringdown. Next add FWBR and DC capacitor in the output coil pair. This setup you have already used once. Measure the voltage where the capacitor is charged to. Detach voltage probe from the DC cap, put some load and connect ground to minus terminal of the DC capacitor. Measure the current through load.

Next do the same with blocking diodes. First use the upper circuit of figure 8 without C, C' and FWBR. This setup you have also already used in the system when you took the very first scope shots. Find the optimum pulse length as explained above, measure the voltage ringdown across the safety spark gap and current ringdown on both sides of the spark gap. Then add FWBR and DC capacitor and do the same tests as above.

Test the Coulomb effect by changing the voltage of the drive pulse.