made a new small power capacitor. 4x 1uF 100V wima fkp2, in parallel more than enough , and much smaller

I remember having those ripples with the battery driven gate driver also.

It seemed to be related to a low frequency /high series capacitor. And driving it at higher power.

https://teslauniverse.com/nikola-tesla/articles/roentgen-rays-or-streams

it seems this article is missing pictures.
I am Looking for a specific picture, where tesla describes longitudinal waves and their origin (abrupt discharge of capacitor plates).
the picture looks like a plate in a square box. but drawn in 2d.
so a line in a rectangle

lol
I thought I had something.
the energy suddenly jumped.
but....
Turned out the square wave generator was set to sine instead of pulse...
yep. that happens.

This is the circuit I'm working with now. Almost the same as the april 2019 radiant power circuit.
Except, now L3 is in the middle, and voltage ability is much higher. And note the C2 series resonant tuning capacitor is now on the other side of L2, on the drain of the mosfet. this is the L2 inside rim
the L2 outside rim is connected to the source.
L1 is connected from the inside rim, to source. the L1 outside rim is grounded.
L3 is grounded on the outside rim.

This creates 2 feedback loops.
one magnetic, and the other Dielectric.
Both loops are counter rotating.
L1 into L3 into L2 into L1 = magnetic loop
L1 into L2 into L3 into L1 is dielectric loop

All coils are pancake tesla bifilar. (patent 512340).
no coil is flipped over, all are wound the same direction.

L2 and L3 are tuned to the LMD resonant mode (voltages out of phase).

L2 has a positive Dc offset, created by the C6 cap, and the diodes (mur8100eg)
C1 is left out, it is optional to put it parallel over L1.  (1nF 2000V range) it tunes the impulse duration
C2 is the L2 series resonant tuning capacitor (46nF 2000V range, WIMA fkp1/mkp10)
C3 is the parallel resonant L3 tuning capacitor (61nF 2000V range, WIMA fkp1/mkp10)
C4 is the power supply capacitor (4xparallel 1uF 100V, wima fkp2)
C5 is a DC blocking cap that also provides a stable ground for the resonant L2. (1360uF =2x680nF 2000V in parallel)
C6 is the DC offset cap (1360uF= 2x680nF 2000V in parallel).

IGD= isolated gate driver, for this, the nov 2020 radiant half bridge PCB is used, but both series mosfets, are switched simultaneously by the same 50% duty square wave, that is pulsing at the L2 L3 LMD resonant frequency.

Mosfets, are c2m or c3m 1200V Sic mosfets (depends on isolated gate driver PCB, C3m needs 15-4V at the gate, C2m needs 20-5V and the gate)
tnx for the correction bogirish :thumbsup2:


TEM resonance (lower frequency) should be much higher in current from the power supply, as L1 and L3 are 180 degrees out of phase with the current. feedback is now destructive/hindering
In LMD mode, L3 is in phase with L1, and feedback is assisting/constructive. current from supply is much lower.

L4 is not drawn in. it is the output coil (not L3!) L4 is close coupled to L2 on the outside of the stack. L4 is not grounded, but is rectified to DC, buffered in caps and Resistive loaded for power measurements. L4 should be magnetically couple to L2, and is not resonant. L4 has a larger center hole then L2.

All coils have equal outer diameter.
L2 and L3 have equal copper mass. this is not necessary  for L1 and L4

I have been working on the same feedback circuit that I used with the radiant half bridge circuit.

but now I am using the single switched radiant power circuit of april 2019.

And I encounter the same triangle wave, and the same phase shifts of L2.

To solve this, I will keep L1 and L3 close coupled. L3 parallel tuned for max resonance.

L2 (on the other side of L3) is probably to close coupled to L3.
this makes the L2 phase shift.

probably because L3 and L1 are TEM resonant coupled, while L2 is LMD resonant, which makes L2 out of phase with L1.
So L1's magnetic field pushes the L2 out of phase when the are to close coupled.

To solve this, L2 simply needs more distance from L3.
Move further away, until L2 and L3 are both max resonant at turn off of the mosfet, when the impulse is created.

once this distance is determined, I can make an epoxy cast, and start experimenting with higher voltages.

Tuning then is done in low voltage, with the boards, then the board is replaced by a matched series parallel capacitor block without switches that can withstand the high dc voltages.

The bigger distance would make the L2 L3 coil capacitor weaker. but the impulses should make this coil capacitor behave differently. Tesla talked about this in his first wireless power patent.
He said the impulses made the air conductive at normal pressure and much farther beyond the corona discharge.

Finally getting grip on the feedback loop system again.
The coil indeed were to much close coupled.
I had L2 on 15mm from L3 and this caused phase shifts, which also happened with the feedback loop of the radiant half bridge setup, as can be seen in the video below.

I distanced the L2 coil further from L3, and now the L2 and L3 reach maximum resonance at the same time again.
Still the phase shifts, when going TEM to LMD resonance, but that is to be expected. The important thing, is now the voltages and current of L2 and L3 are maximum at the same moment in time.

I will decrease the distance again, until I find the optimum setting.

Now with the greater distance between L2 and L3, the power draw from the supply also is increased (doubled) So the feedback does appear to work (as in decreasing the power consumption) but a proper analysis must be done before jumping to conclusions.

Once I have the right distance, I can make a epoxy cast, and use the vacuum chamber, to remove all the air bubbles.

The Idea of feedback, is that the L1 / L3 close coupled coils have a TEM resonant voltage and current.
And L2/L3 has a LMD resonant energy, The TEM and LMD resonant modes, are mixed together (this is something I wanted to do a while back, but could not be done by L2 and L3 alone).
 
Once the epoxy is cast, I can start using high voltage impulses again, to charge up the coil capacitor and use the impulses to amplify the current.

Exciting times!

https://youtu.be/2fLBZu7BJkM

This test was done with L2 L3 distance at 30mm.
L1 L3 close coupled, L3 in the middle of the 3 coil equal bifilar pancake coil stack.
L2 series tuned with 41nF
L3 parallel tuned with 55nF

SCR02:
yellow = L2 voltage
orange = L3 voltage
green is the pulse to the mosfets gate (2 mosfets in series)
Purple is the L1 voltage at the source
Newfile1:
Blue is L1 current at the source.
yellow is a capacitive sensor outside of the system, please ignore

Notice how the L1 current also is slightly amplified by the impulse.
This is tuned slightly above LMD resonance, at max result.
No phase shifts at this distance.
Power consumption:
1,32A and 2x24,0V (series 2 channel power)= 63,36W input
Still high, but already appears to be lower.

No Load present. No L4 present. (L4 to be close coupled to L2, and resistive loaded)

The L2 coil  has its series tuning capacitor  connected at its inside rim.
the capacitor is then connected to the drain.

the outside rim of L2 is connected to the large dc blocking capacitor that is connected to the source.

This means the impulse from L1 enters L2  on the outside rim.

the dc field between L2 and L3, created by the positive dc offset of L2, is thus imploded from the outside rim, inwards.

This means the volume of Aether flow, created by the vacuum flows inwards from the outside rim, between L2 and L3.

once this volume of aether is at the center it can pass out again through the center hole of L1/L3 and L2.
so one hole upwards and one downwards.

the flow outside the coil stack is then becoming a magnetic vortex again.

this also explains, why L2 is opposing im its magnetic field compared to L1/L3.
The flow is outward on both holes.

And as we have learned in university (NOT). the inward or outward volume of aether flow determines the magnetic polarity.
centripetal or centrifugal. Just like high and low pressure wind flow on weather maps

did several distance tests. L2 L3 20mm 22,5mm 25mm 30mm.
25mm seems to work best.
this is the same distance as I determined in another test, when I looked at the best loose coupling distance.
Makes sense they are the same.

pictures:
30mm  scr02 newfile1.jpg  (1,32A input at 48V)
24mm scr03 newfile2.bmp [seems to have no data] (1,3A input at 48V)
20mm scr04.png newfile3.bmp (1,11A input at 48V)
22,5mm  scr05.png newfile1.png (1,17A input at 48V)

yellow= L2 voltage
orange = L3 voltage
green is mosfet gate signal
purple is L1 impulse (incuding ripples I can't seem to get rid off)
blue (rigol)= L1 current

I tried staying away from the coils, and turn them off quickly, but my energy is collapsing rather quickly again.
time to step back and recharge.

If it is correct, that the Aether is sucked in from the side, I might need to change the coil setup, so I am not getting drained[/seems]

there was a minimal phase shift at closer ranges. the impulse was a bit more to the right of the voltage max.

[seems to have no data]
 

the impulse is a fast event.
if it creates a vacuum in the coil cap dielectric field of L2 and L3, the inflow of aether will be also rapid and short.

if this is to assist the magnetic field, it must be at the highest speed of the magnetic field vortex , and this is not at its maximum current. no it is when the currenct is zero and reverses polarity.

then the curve is at its most steep position.
the current is zero, but the volume of aether flow is maximum.

this steepness can be matched by the impulse when tuned with 100pF parallel over L1, or several hundred pf.

but this also decreases the impulse voltage. better not use to much here.

bottom line, impulse should happen at zero current or just after zero current when the flow is high.
. this is at the max voltage point or just after it.

I woke up from dreaming about impulses. it was explained that they were related to time. can't remember clearly.

But I had insights about time before.
space and time are the dimensions of magnetism and dielectricity.

Eric talks about counter space and counter time, but I think its just time and space, that transform.

the frequency is time based.
the coil impulse is a relatively high frequency event (voltage half wave).

the spacial magnetic field of the coil, is transformed into a counter spacial time field of voltage.

the L1 has relatively large inductance but small capacitance, as it isn't tuned by parallel or series capacity, as L2 and L3 are.

so it takes less Time for L1 to generate the voltage from its large magnetic field.

L2 and L3 move much slower in time, as they have way more capacitance due to the added parallel and series capacitors.

So what happens when this fast L1 impulse, enters the slow L2 L3 system?

Tesla said the impulses change the dielectric properties. They increase the conductivity of the air.
they charge capacitors as if they are smaller.

so the capacity is made smaller, the dielectric is given a smaller K value by the impulses.
This changes the time factor. resulting in taking less time to charge up the magnetic fields. Giving a faster (less time) rise of the magnetic field.

air has a low K factor, so it can't change much. thats why I bought titanium dioxide which has very high K factor.

but I will first use pure epoxy to see how it behaves. once I know the distance between L2 and L3, I can cast the coils into epoxy

its about pumping the volume if aether between L2 and L3, pumping it by using the impulse on L2 to discharge the L2 L3 capacitor. The volume of aether moves longitudinal, which is faster than the transverse magnetic field of L1L3.
Faster, means less time.

I wish the coils didn't make me tired like it did yesterday. But feeling fit again.
let's see if BogIrish has a working stl solution that I can print, so I can put the coils in vertical position.

proceeding the test of close coupling distance L2 L3.
this test was at {correction not 17 but 23mm, other test also +5mm}. still no phase shift between L2 and L3, both maximum voltage at the same time.
Power in is lower, 0.97A 48.0V
res freq=83.7kc/s


notice how the impulse voltage on L1 (purple) is much higher 1698.2V, while less input power is needed
previous testst were  1122.4V  1000.4V and  956.48V with increasing distance, which needed more power.
Feedback working, more impulse voltage, less power needed

but L2 and L3 voltages also become less. I think I need some graphs to make this more visible

another test, again a bit closer coupling, half a turn of the wooden bolts I use. this is 17mm
Power is less: 0,9A at 48.0V input
F res=84.8kc/s

impulse voltage is even higher now, with less power input.
1780V
This might be due to L2 and L1 getting closer. The impulse of L1 and L2 might be working together.
L1 outputs the impulse on the inside rim,
While the impulse enters L2 on the outside rim.
 It makes a loop.

L2 and L3 still stay in phase with their maximums (altough they do phase shift together)

its also wonderfull to see the switch boards holding up with these high voltages.

I noticed the stack wasn't aligned with this test, L2 was off center.
I'll ignore it for now, I did a new test made a new post

reduced the L2 L3 distance with another half turn, distance is hard to measure seems around 15mm.
Still the maximums are phase coupled, not shifted.
F=85.9kc/s
power is again reduced, 0.8A at 48.0V
voltage of the impulse again is increased to -1840V
Notice how the L1 current (blue) sine has less amplitude, and the impulse isn't at the zero anymore, but approaches the negative maximum

I did one more test, with another half turn closer by. This didn't work out. the L2 didn't reach its maximum at L3's maximum, but instead declined. I'll turn it back up one full turn, and compare it with the previous test, that had the L2 coil misaligned.

redid the these whereby L2 mas misaligned
impulse is -1771V
power is 0.89A 48.0V
F=84.9kc/s

I could tune the impulse to the 0 current point of L1 (blue) which gave the most amplification.
Interesting to see the energy rise, when the impulse is at the zero curren point.
Isn't that What Tesla also said. Voltage, without current?

Still feeling good, So I will stop testing for today.
My best measurement guess, says its 20mm distanced

in conclusion for now,
I think I need to tune the impulse at the zero current point, and see when it gives maximum amplification of current in L3

the closer the coils get, the less input ower, and the less resonant energy.
and the higher the frequency.

but also the more negative (saying higher seema wrong) the impulses get.

one explanation is that the impulse duration is getting shorter.

that would be an indication of less impedance (resistance). Which is inline with Tesla's statement of the changing dielectric properties by the impulses.

So if I test again, I also need to look at the impulse duration, next to L3 current

The impulse now enters L2 at the ouside rim.
I was thinking if I could also let it enter in the inside rim of L2.
Yes, but then also L1 would have to be reversed connected.
In previous testing I have seen that when L1 is inside rim grounded, the impulse from the outside rim of L2 is larger.

But having the center hole grounded doesn't feel right in respect to the aether flow.
So I'll leave it as is.

I did another test with the last setup.
input was 0,84A 44,8V
F=84.9 kc/s

I measured the impulse (-1.65kV) duration (purple) which is around 400nS, that is faster than I normally have with this setup, that explains the higher voltages. But why does it take less time? The impedance must we less. But the power supply is also less current. that would be normally more current with less impedance.

I also measured L3 current, which indeed gave an extra rise. (blue) 17.3A peak to peak...

two more test.
first 1/2 turn closer.
 SCR12 shows this,
 its 0,75A at 2x 22,4V dc input power
F=86.0kc/s L3 shows 16.1A pp

then another 1/2 turn closer
newfile7 in blue shows the L3 current
power in is 0.59A at 2x 21,4V
F=87.0kc/s

Note the input power is way less then before. The impulse is now at the zero volt point of L2 and L3.
SCR is zoomed in on the impulse, it has not become faster, still a bit more than 400nS. But look at the impulse voltage, it is higher again, I could not use 2x24,0V as before, as my cap banks are on the edge of operation voltage. around 1500V dc
The impulse voltage is higher than the DC voltage, might be due to losses, and the high speeds.

L2 (yellow) has less voltage, while this will also be the output power coil, close coupled to L4 that will be loaded.
Maybe I should add L4 to L2, and load it, to see which setup behaves best.
Measure DC output vs DC input.
See which distance works best.

Important note! L2 is not at it's max voltage level when L3 is max. This was the problem before. this also explaines why the impulse now can be tuned to the zero volt point.

This doesn't feel right to me. both L3 and L2 should be at their max resonant voltage point. voltage should be zero.
But hey! again... I might be wrong.  I just want to know which distance I should use, before I make a (messy) epoxy cast of the coils.

with the latest closer coupled setup, the impulse is at the current maximum and voltage minimum.
Less power is needed from the input, but also less power is generated in the coils.

I always asummed, best would be the impulse tuned to the voltage maximum, not the current maximum.
But hey! I might be wrong.

If the current maximum works best with the impulse, than it would make sense to prolong the impulse a little bit by adding a small (300pF?) cap over L1, this would also reduce its voltage somewhat.
But the thing to do is, is it would spend more time one the current maximum.
As then we have voltage and current at the same time... which is power in the classic definition.

And power generation needs a negative current, so a positive current combined with a negative voltage (from the impulse) would do the same.

Let's see what happens, when I go even closer coupled

No I wont go closer coupled. This closer coupling test already can't be tuned properly anymore.
I need to increase distance again, and get the impulse to the zero current/max voltage point again.
Then current should be highest. This might take a bit more input power, but that's not bad.

Im in doubt.
It makes sense to have the impulse on the zero volt, max current point.

I should load L2 with L4 and see what happens.

voltage x current is power.
dielectric x magnetic field= electric power

so  large voltage impulse at maximum current... should give a lot of power

I tried testing with the coils even closer together, end that doesnt work, resonant waves become very small, L3 current is lower.

So I distanced the coil again 2,5 turns and looked at L3 current (blue) while tuning for max current.
I noticed, the max is actually not at the Vmax or Imax. But in between at 3/8ths of the wave form.
At that point V AND I are 1/2 max. If I tune it to that point, the L3 gives the maximum current.

At that point, there always is a little bit of resonant power, as I already showed in a very old video, powering LED's with a single wire (no ground).  As said, this is because at the 3/8th point, between V max, and I max in, both current and voltage are present at 1/2 maximum.

So if the impulse is added at that point, the resonant voltage and current become maximum.
This test was done at around 27x2=54V and around 1,27A input power.
No Load present.
The impulse voltage was around -1300V still relatively low, as I can go up to 3500V max. But that would require much more input voltage.

When L2 is close coupled to L4, and loaded down (inductively would be best probably), the resonant energy of L2 would become less, and this would mean less energy is draw from the supply also.
That is funny, but so true. more load, less input. due to the resonant coupling.

Anyway... Still haven't used a load. I should make L4 close coupled to L2 again.
Also, as I am still in high frequency regions, I need a lot of voltage. I would love to get the frequency lower, so I can input more amps, and a bit less voltage. But that's not really important now. (also better readings from the current probe at lower freq regions, due to less phase shift).

added phase picture for L2 (yellow scope trace is voltage, current not shown, but is 180 out of phase with L3 current blue)
so for L2, the impulse would be at the 7/8th of the wave form