:shocked:
 I just  realized I can put a dc offset on both coils!
L2 positive
L3 negative

this doubles the voltage on the plate capacitor.
and this squares the field charge.

Altough the discharge by the impulse wont be fully (only half the voltage) still the displacement should be stronger, due to the higher tension (voltage) of the dielectric field between L2 and L3 which are the plates of the coil capacitor.

I am thinking about the dc capacitor at the base of L3 to ground. it needs to be high voltage, but I think the capacity can be relatively small, as it won't be i fluenced by the parallel resonance of L3.

I think I will try 2x 10nF 2000V dc in series first and see if the voltage holds.

at the same time the current in L3 is amplified so this might still cause a need for a bigger capacity. not sure yet...

My new video is in production,
showing the differences between positive and negative discharges
In this short Teaser that I recorded today, I show the bright white Negative DC discharge.
Note how the White spark, makes the tungsten red hot.
I love to see what happens in slowmotion.

https://youtu.be/CJwULXC9jqY

I found a joule ringer variant.
it works on a large toroid.
one half is wound with the 2 windings if a joule thief/ringer circuit.

the other half of the toroid also has two equal size windings that are series connected like a Tesla bifilar coil.
it is full bridge rectified to dc and connected to a battery to charge it.

the joule ringer is based on a 2n2222a transistor with 1k ohm on the base, and powered by 3V (2x1.5AA in series).

it's a simple small circuit, but takes some time to build due to the windings.
cut 4 equal size length and diameter electric enameled wire to make it.

Should be a fun result.

as a variant, I would connect the bifilar in parallel to the transistor base and emitter, and put 2 more coils on top of it, also bifilar, and use that as output (rectify to dc in to cap for resistive load).

that way, the impulses at the collector can travel into the resonant first bifilar.

Another great variant would be to use a PNP transistor to get negative impulses

made a 630nF 4kV capacitor for the negative DC offset of L3, including 4 diodes. I could have gone smaller probably, with 2x 10nF in series. but I had these bigger 630nF so why not use them.

Also tested the new L1 L2 (non flipped L1) close coupled again, without L3 or L4 and no DC offset.
Power was lower. 0.39A at 64,4V dc= 25W

impulse was a bit slower 400nS but will most likely  be quicker with L3 tuned and loose coupled to L2.
"only" -2600V now, but with a quicker impulse the voltage gets higher negative.

Tuning with both L2 an L3 having a DC offset, will become very tricky now.
The tuning boards only can withstand so much.
If I really want to push the DC up, Ill have to use parallel-series tuning caps for L2 and L3 without switches again, that can with stand 4kV.

The Idea is still the same. L2 impulse will not cause a L2 current implosion at high enough voltage.
L2 L3 coil cap displacement current then amplifies L3 parallel resonant current.
The L3 current steepness must match the displacement current speed (dI/dt) for maximal amplification.

Another risk is when the displacement current is matched to the L3 current steepness, there will be a huge amplification of the current and thus also the resonant voltage. L4 needs to be in place and loaded to keep the voltage and current down.
OR ELSE...

L3 parallel resonant current steepness must match

yellow, trace shows L2 impulse

edit: L2 was tuned with 65nF and Fres was 69kc/s

New video showing and explaining my new Radiant Energy circuit.

https://youtu.be/qxyf-L48Y8k

I just did the first basic test with the new circuit.
positive and negative DC offsets work.
 I could only tune into TEM resonant mode. around 43kcps
C2 and C3 were equal size around 65nF
 C3 was much bigger in previous tests, but now I need to go smaller I believe, based on previous test with L1 L2 close coupled.

I saw a small change at 90 kc/s but it did got resonant fully.

I had L4 close coupled to L3 but without anything connected to it.

I'll hook up my tuning board again to L3, so I can work more easily. But... I have to be carefull, as the complete board has a negative dc offset. The switches should not notice it, but still... if it sparks Ill be the victim. Very dangerous setup this...
Need to be not grounded when tuning.

Let the testing begin!

I increased the L2 capacity again, doubling up, by the adding  22nF 2000V caps again, that I removed in the previous setup.
So now again L2 has a capacity of around 135nF, with 4000Vdc capability.

I made my last video a bit more "commercial" by promoting it and is has worked.
I gained a lot of new subscribers, which reached 10K

the benefit of this, is that I now can use YouTube for selling merchandise.
So I'm working on a new website and perfecting the series mosfet switch board, so I can sell it.

this test was with C3 (for L3) being 21nF and C2 (for L2) 135nF

The negative impulse on top of V+ max L2 (yellow), isn't really the impulse, but two of the ripples that often unexplainable occur.
L2 and L3 are LMD resonant as can be seen by their signals being out of phase.

The L2 and L3 probes are DC, and show a clear positive DC offset of L2 (yellow)
But the Negative DC offset of L3, is much Lower than expected. Why is that? the L3 DC cap is smaller that the L2 DC cap on their grounds.
Is the negative DC offset of L3 becoming positive charged somehow?

The impulse is actually still is on the -Vmax of L2, but you cant see it, as it is fully resonant tuned, which makes the c2 cap absorb the impulse

this test was with C3 (for L3) being 21nF and C2 (for L2) 135nF
LMD Fr=102.7 kc/s

The negative impulse on top of V+ max L2 (yellow), isn't really the impulse, but two of the ripples that often unexplainable occur.
this can also be seen by the purple square wave.

So We are using a Ripple, to cause the discharge, while the impulse itself isn't used... That can't be right

L2 and L3 are LMD resonant as can be seen by their signals being out of phase.

The L2 and L3 probes are DC, and show a clear positive DC offset of L2 (yellow)
But the Negative DC offset of L3, is much Lower than expected. Why is that? the L3 DC cap is smaller that the L2 DC cap on their grounds.
Is the negative DC offset of L3 becoming positive charged somehow?

The impulse is actually still is on the -Vmax of L2, but you cant see it, as it is fully resonant tuned, which makes the c2 cap absorb the impulse

if I can't use the ripple, on V+max,
Than I could use the impulse on V-max. But the current of L3 is then opposite, so L3 should be flipped over.
That would also compensate for the minimum voltage of the L2 L3 coil cap, by the AC minimum, which then is avoided.
AC would stay relatively equal.

Ok. L3 will be flipped over.
Impulse placed on -Vmax of L2.

with this setup,
The TEM resonance, has a double wave on L3, so L3 isn't tuned down to L2's TEM resonant mode, but stays at a higher octave.

Also, when tuned above TEM and below LMD, then the impulse is riding nicely on top of the L2 V+max.

So Tuning has begun.

I do want L3 to be also able to tune to the lower TEM resonance. Logic says increase C3 capacity. So I will do the reverse. Start with a small capacity on L3, and See if I can get both resonances. on L2 AND L3

Since the frequency is rather high (101kc\s LMD) I will increase the L2 L3 distance to 25mm instead of 15mm , as this lowers the frequency

I won't flip L3 around yet

I creasing the L2 L3 distance to 25mm did seem to work, the L3 almost shows a sine wave on TEM resonance.

half the inner diameter (radius) of the phi ratio coil would probably be best. which is around 3cm I believe. But might also simply depend on L3 capacity (c3)

L3 having less DC offset than L2,
could be due to the Corona discharge. the L3 DC part if the circuit isnt isolated as it includes the tuning board.
also the diodes of L3 are not isolated.
I used nail polish for isolating L2

hmm Having the impulse on V-max LMD sucks a lot more power from the supply.

I need to be able to get a single discharge on the the V+max, as this is much more efficient.

another insight,
each coil cap discharge would amplify L3 current, resulting in a higher L3 voltage, which adds up the the coil cap voltage and charge
so each next discharge is more energetic (untill it limits itself) this has the possibility of a runaway. thus a need for a load to restrict the  voltages until safe limits.
Im already way over the safe value of the capacitor tuning board.

With my chair I added some capacity (500pf or something like that) to L1 to get the single discharge on top of V+max L2.
but this slows the impulse dowm, which lowers the voltage, and thus it needs a higher input voltage again to get up to the high voltage...

Oh well... I'll just have to play a lot more untill I get it tuned Like I want too

I placed the L2 L3 coils again on 15mm instead of 25mm as I realised, I justed needed more capacity on L3 to get the single sine wave on TEM again.

I compared the ratio's of the LMD and TEM frequency's, the theory says it should be 1/2 pi separated (1,571) but I noticed it wasn't And again I could tune it by changing the L3 capacity.

The nice thing is, I just disconnected the DC wire from the source, so there was no DC on the circuit, only AC. Which is much more safer.

I left the L2 cap alone as it is hard to tune (need to solder), and thus only tuned L3, which is the tuning board (has no DC component).

It came down to this:
with L2 135nF (need to measure it again) and
 L3 51nF (counting the caps, not measured)
the LMD resonance was at, 75,0 kc/s
the TEM resonance was at 47,7 kc/s
So the ratio is 1.572 which is very close to 1/2 pi

With L3 capacity becoming larger, there was also more and more impulse voltage still there at LMD resonance.
With TEM resonance, I could always tune so that the impulse completely vanished (charged up the C2)

The impulse also became faster again, from above 700nS it is now again just above 200nS. This gives a higher voltage, which is good for the DC generation.

I could make the C2 even bigger, but I noticed, with the lower frequencies, the waves are longer, and more ripples appear.

To get a single ripple, would mean to have a smaller C2 cap, with higher resonances.
I have still some playroom here, as the caps work fine up to 100kc/s but the 500kc/s current probe starts deviating upwards of 100kc/s

Since using the mysterious ripples as discharge, might be working, I prefer, for now, to work with the impulse on the negative V-max of L2.
This also means I will have to flip L3 over, else it's current won't be amplified, due to the wrong polarity.
If I have researched that, I can still tune higher in frequency, and get a single ripple on V+max, which is very efficient

I flipped L3 over, and measured again. looks like its current is now indeed amplified as it should be.
But the impulse is not perfectly on V-max L2, so I'll probably will need to retune again.
Enough for today.
scope shot:
I pushed the psu to the max, (L3 still seems leaky on the DC side)
yellow= L2 voltage
orange =L3 volts
green L3 current
purple mosfet on/off (hi/low)

at 75.1 kc/s LMD
power : see photo , 105.6W input (will decrease with load)

No load on L4 (but L4 is again close coupled to L3)

The voltage is high, but I wanted to get it even higher. The impulse is fast, 400nS right now.
Bringing the frequency down, by increasing the caps would be good, for the DC voltage as the impulse voltage would also become higher.

I tuned to the lower TEM resonance, with L3 still flipped over, and the cap values unchanged. 135nF for c2(L2) and 51nF c3(L3)

Tuned slightly above the resonant frequency, the impulse appeared on top of V+max.
I gave it some power, to get high voltage (DC offset activated).
2.6A x (17.9+17.9)=93W (even less then LMD, which was 105W)

yellow=L2 volt orange=L3 volt green is L3 current

The impulse is now nicely tuned on the max. and the current amplification of L3 is now clearly visible, as the steepness becomes stronger at the zero current crossing. from positive to negative (green)
The energy is transferred pretty good.

The ripples at V-Max of L2 are also disrupting the current, which isn't good. Maybe, at higher frequencies I could get rid of this. but I remember already having tried this.

With Tem resonance, the Idea is the Longitudinal part is only between L2 and L3 as this is the displacement current that is induced by the impulse.
The transverse part is then the amplified current of L3

Most important is, that the LMD option doesn't discharge the L2 primary, with the impulse on V-max.
With this TEM setting, it does discharge the L2 primary which I believe is what we need (also)

ok the line out...

First, reduce the capacity on L2, so the frequency raises, and the ripples (might) disappear.
Then again tune L3, so it is a single sine wave, and the ratio of LMD and TEM is again 1/2 pi (that is a guess, which might be wrong)

Than Tune to TEM, get the impulse on top of V+max L2 (slightly above resonance).

and see if we now get a proper amplification of current in L3 (which is flipped).
if the current transfer to L3 is perfect, a massive amplification should arrise in L3. this is dangerous as the caps could blow. So a load (L4) should then be present.

To tune the perfect transfer of the displacement current, into L3 , the C3 cap should be tuned... until a perfect sine is produced of high amplitude.

Changing C3, will also change the freqeuncies again, and the LMD/TEM ratio, thats why 1/2 pi might be wrong.

Without the dc offset I can tune L2 again, and see if a smaller capacitor works better in TEM mode, aiming for 70kc/s TEM resonance

then the 4kV capacity can be tailored to the right value, and the dc offset can be used again

but can voltage to work if it is not equal and opposite force from pancake coils +/- to open aperture ?

Quote from securesupplies on October 16th, 2021, 10:23 PM

but can voltage to work if it is not equal and opposite force from pancake coils =/- to open aperture ?

I don't understand the question

I tuned the L2 and L3 (without DC offset) to 116.7 kc/s with 11nF on L2 and 11nF on L3
Then indeed as excepted, I only have 1 ripple left.
So tuning higher decreases the amount of ripples,
And tuning lower, increases the amount of (unwanted) ripples.

I suspect, I need different coil to get this to work on a lower frequency (where my caps can work and my current probe also has no strong deviations).

I now use, 2,5mm2 for L1 (for high voltage) and 1,5mm2 for L2 and L3. So to get this right,
I will have to change L2 and L3 to 0,75mm2 which will make more turns, and increase the inductance.
L1 can remain the same. although L1 might be part of the problem, as I have seen the same ripples, with only using L1 without L2 and L3. Thus It might really be a problem that only occurs at lower frequencies.

I suspect the mosfets and their isolated DC DC drivers, to somewhere cause the problem. somehow... ???

L1 needs to be small inductance and capacitance to get the high voltage impulses. but I need high inductance and capacitance to get rid of the ripples at lower frequencies... 

So, If this could all work without high voltage DC offset , then I could use slower impulses, and less ripples.

scope shows TEM resonance at 116kc/s 11nf on l2 yellow voltage. high bump is impulse, low bump is ripple (only one)
 
edit: it seems the supply voltage also has an influence on the ripple duration(frequency). Since my DC/DC converters run from a separate 12V supply, I suspect this has something to do with miller capacitance or, the capacitance of the drain to gate or drain to source capacitance of the mosfet.

F=72,2kc/s
C2= 31nF(L2)
C3=31nF(L3)
tuned to TEM resonance, with L3 flipped
this is with DC offset, and a lot of power:

yellow=L2 voltage
orange=L3 voltage
green= L3 current

Note how there now is only 1,5 ripple on yellow V-max (before mosfet turn on)
Why do I only have 1,5 ripples, and not more at this low frequency?

current amplification of L3 is present but can be better