I played with the coil to see if I can get higher voltage jumps as in previous post.

removed L3 and just worked with L1 and L2 close coupled.

Changing the values of L2 series resonant tuning cap, didn't give an enlarged jump.

It seems, it need the L3 parallel resonant coil, with its load attached, to make a proper jump. But the voltages are also relative low in the previous post. only 110V in L2 max.

Seems L4 also didn't do good. L3 loaded with the lamp seems to work best.

Might this be, due to the L2 voltage jump, having something to work into (high impedance parallel resonance L3, and lamp)?
as if it is providing a ground? A wall for the longitudinal to push off, so the voltage jumps higher in L2?
Hmm.
trouble here is, I want to play with the values of L2, so the tuning capacitor bank, is needed as series cap on L2. But L3 also needs to be parallel resonant, so I need to tune that one also. this asks for another tuning cap bank. or. I could give L3 a fixed value, and tune L2 while changing L2 caps. that should give some playroom. I'll put 30nF (already got it) over L3, and play with L2 values, see what happens.

I have a 2 channel power supply, of 32V 5A maximum.
With the system on, its now reading 2x45V, negative channel is 1A positive channel is almost zero amps. I probably have one diode in the line, need to put the other diode also in the live to prevent feedback into the power supply.
edit: negative line from power supply had diode in it. going to put positive line diode in

L3 parallel tuned and loaded with 28W lamp (bright), 40nF
L2 series tuned 61nF
79kHz

L1 L2 close coupled. L3 1,8cm distanced.
all coils bifilar and equal

both lines of the power supply now have a diode in them to block the feedack.
voltage is back to normal 2x 32V
negative is 0.8A positive is 0.7A
don't know why it's differnt, but the signal also isn't symmetric, the negative maximum shows a distorted jump.

Series parallel resonance with lamp brightly lit.
No Impulses, L1 is feeding straight into L2

nice

meter shows dc voltage over lamp (not dc amps, those are in the white analog meter)

found this video showing the relationship between parallel and series resonance

comparison between series & parallel Resonant circuits

Cant understand what he is saying but the formula's are very clear (subtitles dont work).

This comparison reminds me of CP Steinmetz comparison of the magnetic and dielectric fields (together forming the electric field).
He called the analogous. I came to conclude one is the inverse of the other.

after doing the last experiment, It was very clear we need both series and parallel resonance interlocked, to create electricity (power/pressure)

Its seems series resonance (L2) is related to the dielectric field,
And parallel resonance (L3) is related to the magnetic field.

Now the q factor, provides the magnification. of voltage and current. in case of series resonance we need low resistance for a high Q,
for parallel resonance we need high resistance for a high Q (amplification).

Getting Q high in parallel resonance is now the goal. it would be easy to just use the air as the parallel resistance, but we need to be able to load the parallel resistance coil.

A capacitor is related to a resistor. When it is empy and being charged, it acts as low impedance, giving high current.
When the capacitor is charged, it acts as a high impedance, which gives low current.
When it is discharging it acts as a negative impededance (negative resistance) which gives negative current.

If we put a rectifier over the parallel system, and put a capacitor behind it, we end up with DC. If the capacitor is empty, it will acts as a low resistance load for the parallel resistance. therefore the Q will be low. and the amps are low. the cap will take as much current as possible, but there isn't much to give since q is low and magnification is low.

once the cap is charged it represents a high resistance load. now Q is high. amplification is high.
Therfore it is wise, in the series parallel setup, to have a large bank of caps after the rectifier, that FIRST gets charged up, and AFTER its fully charged, is presented with a LOAD.

So the load should not be on before the cap bank is charged high enough. Else the Q factor will stay low, and amplification will stay low, and not reach its threshold.

now a negative resistance (discharging rectifier cap) with parallel resonance will ....  eh... no, what I meant to say, is I have seen this system create pressure. the power supply was over charged, when the diodes were not in the lines. when I put in the diodes, the pressure found another way to decompress, into the rectifier, into the cap into the lamp that is.

I used a earth connection on parallel resonant L3 may be it is better to use a full bridge rectifier. altough the earth should be immune for pressure change so thats why I grounded the outside rim of L3 (and L2 and L1) and the amplified inside rim was rectified with only 2 diodes, into the 2 capacitors that also had ground (in the between them).

does this L3 earth ground matter? testing will tell.

best conclusion for now, is keeping a rectifier on L3 and a large cap bank, that is first fully charged, before it is presented with a load (lamp).
I need to make a switch for the lamp.

until now I thought I need the impulses, but it didnt fully sit right with me. For an impulse is a jump up AND Down.

now that the impulse is used to charge up the series tuning capacitor, we see the voltage jump up ONLY. not down.

This also happens with a tsunami (longitudinal wave).
The ocean floor jumps up (not down) and a tsunami is the result. A pressure wave.

In my half bridge setup, L1 and L2 are close coupled, and grounded on the outside rim. The inside rim of both coils is connected in the middel of the halfbridge between the mosfets.

thr proces:
postive switch turned on:
L1 is first charged up with positive current, creating a magnetic field (assisting L2)
Simultaneously L2 is charged up and creating a magnetic field (assisting L1) being resonant, the magnetic current is transformed in to dielectric voltage.

the positive switch is turned off, L1 discharges into the lowest impedance which is the series resonant L2 into ground.
The L2 tuning capacitor is charged up INSTANTLY by the impulse.

after short discharge period (dead time) negative switch is turned on, and the cycle repeats, for the second half period.

This can only happen, by the SUCKING IN action of parallel resonant L3 which draws in current (as it's q is high, current is magnified, and this is felt by L2)(!L3 current should be high!) MEASURE
but when L2 is max voltage current is zero, and current in L3 is also zero. but in fact, the aether is in maximum speed of reversing its rotational direction (current=magnetic field=eather vortex).

it seems L3 provides the sucking action, for the capacitor of L2 to jump charge. Is this Due to the negative resistance of the discharging capacitor? which is parallel with the resistive load, hmm that doesnt make sense. those should add up, and give a positive value. the discharge current is equal to the resistive load current.

the jump of L2 series tuning capacitor is also making L2's dielectric field jump up, representing a true longitudinal wave from L2 to L3. Right at the startup uf L3's current wave, where it assist. It should even put a jump in L3's voltage. which by the diodes, is pushed shraight into the rectiofier cap bank.

the amplification, should work best, if L3 is differnt from L2 in dimension, in voltage and current. I could use thicker wire 4mm2 for L3 which will have fewer turns (equal copper mass of L2 and L3). making the coil tinier, and create a unbalance capacitor of L2 and L3 just like thomas thowsend brown used. or I could go bigger, with smaller diameter, but that doens make sense. as nothing really makes sense.....
I'll stick to equal coils for now. but differnt coils might be needed, to create a lense of the dielectric field.

to get high Q (magnification) of voltage in series resonance (L2) , is done by small capacitance and resistance.
(keeping inductance equal or bigger than L3).

to get high Q of current in parallel resonance (L3) we need large capacity
(inductance equal or smaller than L2).

Since the capacitance is in square root related to the Q factor.

I suggest having a capacitance ratio of 1:4 for L2 and L3 tuning capacitors to get a high Q in both.

I made L3 100nF and tried making L2 smaller in capacitance, but came soon to the conclusion, L2 series capacitance is what sets the frequency. L3 just follows, and by havinf a larger capacitance, the voltage just gets lower.

L2 and L3 are out of phase, and it makes sense, to get them equal voltage. hmm doens't make sense to me yet.
lower voltage higher amps higher amps lower voltage right?. But the goal was to get a High Q in L3 not by increasing resistance, but by increasing capacitance. ah yes, voltage drops when capcitance is increased, current rises, but... it needs more kick to build up its resonant voltage and current in L3.

Let's see what happens when I close couple L2 and L3...

close coupling l2 l3  shifts the resonant frequency higher up to 88khz from 74 khz. and also the current draw is higher.
makes sense to have some distance between L2 and L3 as one is series and the other is parallel resonant.
But what is the real logic behind this distance? what is ideal? 
 lets give them some distance again, and see, if making L3 capacitance smaller works better

OK. L2 series capacitance tunes the frequency. L3 capacitance almost has no influence on the frequency.

But L3 capacitance, when made smaller, drives up the voltage in the output and at the same time, current from the supply drops.
more output less input. thats good thats what we need.

Since the size of L3 parallel capacitance doesnt really detune the resonant frequency of the system, I wonder if L3 really is resonant? Is it only reacting on L2 and its voltage jumps? Should I again tune L3 while keeping L2 fixed?

I feel my energy is going down, back to the ascention seat and recharge. Since I retuned that negative impulse on the negative voltage maximum it works a lot better (the when the impulse is on the positive voltage maximum).

I remeber, L3 had only around 4nF in my april 2019 video.
makes sense to make it much smaller here also.  Could even fully remove the capacitance over L3, if L2 was tuned to the proper frequency

blown up diodes again. they keep working but the reverse recovery is not fast enough anymore to block the fast impulses.
so impulses are gone.

need to watch those impulse voltages. max 1000V so will a for max 750V to have some headroom.

Still.... finding the capacitor charging instantly is amazing

also need proper cooling on the diodes.

Here is the working schematic so far.
D1 and D4 are to prevent the presure generated to push back into the power supply. if left out, current drops on the supply and voltage rises above maximum of supply (!)

D2 and D3 prevent the impulses to exit via the body diodes of the mosfets. mur8100E diodes are used (8A 1000V ultra fast)

IGD is a isolated gate driver, 1EDI60n12AF is used (as before) the 20V is created by series connecting a 12 and 6 volt lead acid battery.

L1 L2 L3 are equal bifilar pancake coils. 3x 10M 2.5mm2 speakerwire. outside rims are grounded

L1 creates impulses
L2 is series resonant
L3 is parallel resonant, rectified to DC and given a 28W 230V halogen lamp resistive load.

L1 and L2 are close coupled, L3 is 3cm distanced from L2  (ideal distance not yet determined)

blue scope line is voltage between C2 and L2 showing series resonant sine, voltage jumps are from C2 charging up by the impulses created by L1. This only happens at series resonance, above and below, the impulses are clearly visible. but at the series resonant frequency, the impulse is absorbed into the capacitor.

yellow is L3 inside rim. Wave form is clipped by load.

Fres L2/C2 is now 67kHz, with C2 being 59nF. This gives a low voltage, better make C2 smaller, for larger voltages.

the half bridge is driven by a out of phase square wave of 48% duty cycle. generated by a dual channel wave generator (Rigol DG1022Z)

L1 can get a parallel capacitor over it, to make the impulse duration (inductive kickback) longer, this might result in a better charging of C2.

2 SiC mosfets are used: C2M0080120D (80 mOhm, 1200V)

L1 magnetic field is build up, magnetic energy after switch off, in transformed into impulse, that charges up C2/L2 series resonance.
When L1 is charged, L2/C2 series resonance is also charged.
system re uses its energy. current draw is low.

adding sunflower oil between L2 and L3 might increase the dielectric constant, and give better results.

L2 and L3 can be me made differnt in size, but equal mass of copper must be used.
L3 having lower inductance and capacitance, will have higher resonant frequency).

Dielectric field between L2 and L3 could be increased by DC bias (of high voltage several kV). L2 and L3 should have a (large) DC blocking cap between coil and ground.

the sudden increase of C2 also increases the voltage in L2 giving a longitudinal wave in its dielectric field. L3 catches this wave.

another set of D2 D3 diodes overheated and died on me. Didn't even had that high voltages. Energy in the impulses simply is to high. diodes can only take 8A. peak current is much higher, and more or less sustained.

I removed all parallel tuning capacitance from L3. as this gives the highest voltages. (higher octave resonant frequency)

Tuned L2 series capacitance again. logic is inverted. I need even more capacitance to get a high resonant voltage.
should have been reversed, but I guess this is due to L1 feeding its inductive spike into L2. and with lower frequencies (more capacitance) the magnetic field induced into L2 is much stronger.
I should put a current clamp on L1 again, and see when it produces the highest current.

Also add some series tuning capacitance to L2, to get even lower in frequency, and see how the voltages and current develop. I need more voltage.

But first.  replace the diodes AGAIN.
THEY WERE BLOODY HOT. NEED ACTIVE COOLING!!!

or maybe go back to the single mosfet circuit again. seems I still need to learn how to tune properly.

the diodes d2 and d3 heating up is bad
it's a huge waste of energy.
the impulse should not pass through any diode.

I'm going back to the single mosfet circuit,
and compare high and low side switching, as I have both those circuits ready.

Most of all I'll keep an eye on the current.

For now I intend to go silent end work until something note worthy is to be mentioned. probably in video style.

talkimg about video, the half bridge circuit is worthy of publishing. so it can be used and seen by all.

peace

edit: at resonance there is no impulse. as it is fully absorbed into the capacitor, and the diodes then stay cool.  amplifing this sudden charge would be a goal, but how?

played alot, got the 5 coil sandwich working, now I will add the dc offset again, on both L2 coils. one positive one negative dc offset.

L3 output in the middle of the dc offset dielectric field.

I finished the Double dc offset module.
used mur1660ct diodes

 the caps are rated 1000V dc. 680nF

giving the possible 2000V dc over the l2 coils if 1000V impulses are used.

I'll probably stick to max 700V impulses. giving 1400V dc.

Still need to connect the caps to the impulse source on the board. tomorrow...

DC offset modules work.
I now have 2 L2 pancake bifilar coils, series resonant on top of a DC offset.

One L2 is positive DC the other is negative DC. I can go up to 1000V positive and negative, giving a 2000V DC field between the L2 coils.

Goal now is to collapse this DC field, by making use of the impulses injected from the two L1 coils into the 2 L2 coils.

prefered would be a setup, where both polarity impulses are created at the same time, and devided over the L2 coils, so together they cancel out the DC offset field.
I read this in the vril compendium book 9 page 19-22  the sudden collapse of the DC field is what works.

finally have the dual setup working.
did a few quick tests. looked good
7th sub harmonic of L3 for L2 seem to work best.
L1 impulse then fits half period of L3.
L3 then is tuned down to L2 frequency, giving 2 impulses per sine wave.

It's very strange to see L2 and L4 (series resonant) being tuned by around 60nF
While L3 (middle parallel resonant) only needs 8nF to be tuned.

Why so much less capacitance is needed for L3?
Is it because the longitudinal induction is that much faster? faster then transverse light waves, as Eric Dollard showed.
making L3 longitudinal resonant. i

Still need to fine tune L2 and L4 series resonance (7th subharmonic of L3)
and the L1/L5 impulses to get the half wave duration of L3
then tune L3 to L2/L4 and retune again.

Rectify L3 to DC, and measure output.

two bifilar pancake coils, can act together as a capacitor.

each pancake coil, is one plate of the same capacitor. you can even put a dielectric medium between them, like place them in a refined sunflower oil bath.

if you give one plate/coil a positive dc offset, and the other plate/coil a negative dc offset, this would charge the capacitor made out of 2 coils that act as plates.

Because the plates are coils, we can make the coils series resonant (low impedance).
Assume both coil/plates, are series resonant at the same frequency and in phase.
 then the voltage difference between them would stay the same.

due to their low impedance we can feed the coils impulses.
a negative impulse for the positive DC plate/coil at the first half of the cycle
a positive impulse for the negative DC plate/coil at the second half of the cycle (period)
see picture attached

What these impulses do, is collapse the DC dielectric field generated between the 2 coils.
Creating an implosion. a vacuum. to be filled...

This resembles the discharge of the capacitor.
but only very shortly during the time of the (alternating ) impulses.

then we can put a third coil in the middle of this DC field. inside the capacitor.

one half of this circuit is already out there. the other half of this circuit is a mirror of this circuit, producing the negative dc offset, and the positive impulses.
working together, each half of the total circuit is driven by an out of phase square wave from a TTL A and B output.

L1/L2 close coupled
L4/L5 close coupled.
L3 loose coupled to L2 and L4 which are series resonant.
L3 is parallel resonant, picture doesnt show parallel capacitor.
A and V are amp and voltage meter
IC is isolated gated driver.

all coils are pancake bifilar tesla coils

both circuits work together as one. meaning in phase.
isolated gate driver is driven by TTL (transistor transistor logic) A and B, of a single square wave generator.
.thus the 2 circuits recieve an out of phase square wave, making L2 and L4 in phase series resonant.

Tuning is crucial.

very curious.

I had problems tuning (for a long time now)
and I decided to take the circuit apart and start over with just a single low side switched coil.

simply pulsed way below it's resonant frequency.

I expect it to ring. but it doesn't.

Instead I get a single voltae impulse, and a current impulse. afterwards there is only a negative current.

Why does it not ring?

with a high voltage probe, I can measure 300V outside the center of the coil while the probe is not connected!

it looks like the coil is transmitting the energy. I remember seeing the coil ring when I used negative impulses.
need to test this again.

I thought the impulse might charge the caps of the power but... that doesnt make sense

32kC  resonant around 500kC
green is current 500ma per div
blue is not connected probe 100V/div
orange is impulse from coil at mosfet switch 100v/div (800v)

power also is very low. 30mA at 32.2V dc

I repaired the diode on the negative power supply and it blew again.

its a mur8100 8A 1000V

I kept the impulse at 800V

its a low side switch, so why is that diode blowing?

coil is grounded on positive. negative is switched.

there is a 3 uf 100v cap between positive and negative, that power cap is fed through 2 diodes from the power supply.

if I probe  positive ground I see no impulse, only between coil and mosfet

strange. I have to think this over

the power supplied is much to low. 30ma 32V is 0.96W. for 800V spikes 2.4us at 32 kC...  seems very little energy.
current probe also looks like more.

but dc current and voltage are measured from supply. but probe seems to so no ac stuff...

hmmm...

is positive pulse transmitted? though counterspace/ aether?

my gate drive is grounded, so the negative diode is bypassed through earth ground (wire).

so diode isn't broken.
need to remove earth from gate drive circuit

still strange... why does it not ring and have a slow negative current after the spike?
is the 3uF power cap playing tricks?

v+ of low and v- of high should both be tied to earth ground.
this is for the impulses to work into earth ground. it seems they really need earth ground and not a large capacitor as ground.

this means the low side switch circuit cant be driven with a buck boost gate driver, as that ties earth ground to v-

both gate drivers need to be isolated. easiest, is by 2x  battery (12v+6v series).

or 2 x isolated dc-dc converter. 1w per switcg is enough. it seems. not sure if gate power supply is influenced by impulses.

I use a separate ps for gate 12V. it uses less then 1W for both

I again made the symetrical power system for the 2 circuits to work.
and it works!
weeks of previous testing didn't work due to not using symetrical power.
did a quick test and it looks good.
very low power consumption.
and a pure sine with spike.
all good.

I used 4x 680nF caps 2000V dc rated. absolutely overkill, but if the circuit is untuned the caps charge up to high levels of voltage, so 1000V dc would be the right rating.

tuning will have to wait a bit. heatwave is to intense for concentrated work (no airco YET)