So the coil cap has low capacity. and it needs high voltage to charge.

But...
What if the close coupled L4 coil,
is grounded?

close coupled means that L2 L4 also works as capacitor plates. But due to the closeness of the (coil) plates the capacity is MUCH higher (than L2 L3).

I always thought that might be a problem, but I "only" have 2000V dc so, that should be fine.

Then... well.... I guess I should try it.
see if I can tune it, make L4 resonant to those implosions.

At the same time L4 should than be a higher octave of L2. So You got double induction.

pretty far out? yes. doable?
yes, but... the impulse now is fast.
and I bet L4 has a lower resonant frequency (longer duration wave) then L1 has (of which the impulse is a half wave)

So I can never match it. Maybe... L4 does not need to be resonant.
I don't know. I'll just have to play with it again. Push it.
it's only 2kV now...

the mosfet is 50% on and 50% off.
when its off the L2 series resonant coil is powered,
and L1 is also powered is series with L2, but... how much?

 so L3 could be loaded (single diode rectified). during that half cycle, without loading the oower supply.

now when the mosfet is on, it powers L1, and not L2.
so L3 cant be loaded, but L2 can (using L4, again single diode rectified)

I still think loading L3 is not the way to go, but it's an interesting thought.

http://www.marksacademyofscience.com/uploads/1/2/8/4/12848543/cold_experiments_tesla_coil.pdf

This pdf is still amazing together with this video:

https://youtu.be/qUvF-ZXejyI

The tech is not disclosed but it makes sense

So the fast discharge of the coil capacitor
by the impulse on L2, produces a longitudinal wave.

I always assumed it would travel away from the coil perpendicular to the plate.

But now I am thinking, since I measure a stronger dielectric field outside the coil at the height of the gap,
that the aether is squished out side ways.

This makes me think of the magnetic field, which is not like we have been tought,like a single bubble around the coils, but more like a double bubble on each side of the bloch wall, Ken wheelers experiments support this.

So the L2 L3 space is is then the bloch wall.
The impulse pushes the aether out, which then is captured in the circular motion of the magnetic field (moving around and back inwards).

then it makes sense that L1/L3 is one part of the whole magnetic field,
en L2/L4 is the other part of the magnetic field.
and these 2 parts are not really sensing each other, if L4 is loaded L1 should not feel it.

another thing I noticed is the big difference in the magnetic discharge of L2. When I used the 5m 2.5mm2 L1 coil the L2 magnetic field was barely discharged by the impulse (250nS)

now with the slower 6m 2.5mm2 L1 coil I suddenly see a much bigger discharge of the L2 magnetic field. I was surprised to see this, as the impulse was just a bit slower 280nS

So, next cap coil will be made with the faster L1. which also draws less current for a faster higher voltage impulse.

I need that high voltage to charge to cap coil with dc. I want to have 3kV

I now have 2kV dc (2k5V impulse) at 72kc/s with 64V input (maxed out).
To increase the voltage further I would need to drop the frequency further, which means even bigger capacity and even higher currents while I already sit at 27 App (making the coils hot)

even more important, I haven't played with L4 enough, not loaded it.

picture: 4 coil stack from the sideview.
green is dielectric field between L2 L3
coils are black
red and bleu are the magnetic field components
so green is discharged and pumped into the magnetic field components

Almost forgot to measure it...
91pF between L2 and L3, with epoxy

Quote

from reply #168:
I stacked the new coils with a 15mm air gap between L2 and L3 and measured the capacity between them: 32.4pF
so 32.4 X 2.72= 88.1 pF with using pure epoxy as a dielectric

so my guess was pretty good, 91 instead of 88 pF

what is interesting to see, is that the impulse of L2 is very short in duration, while the current amplification of L3 seems to take longer.

I wonder what happens when I build the new coil cap with the faster (less windings) L1 coil.
as it showed no implosion of the current of L2, which is strange (but could be due tonthe 500kc/s limit if my current probe)

I want to see how L3 behaves, if it is again amplified, and more or less, with the faster impulse on l2

A lot is coming together. I looked at Peter Lindemann 2006 presentation again, whoch bogirish gave me the pdf with the slides. Thanks again!

There is a lot of good info in this one

https://youtu.be/mZIOj4BziHQ

before casting the new coil cap (with faster L1, 5m 2.5mm2) into epoxy, I decided to record some data with air as the dielectric.

I pushed the voltage slowly up, watching the impulse voltage climb to 3kV.

BANG

OOPS
I forgot my L2 tuning cap can only withstand 2kV dc...

Making a new 100nF 4kV cap out of 40x 10nF and ordered 24x 22nF 2kv for a smaller solution 110nF 4kV

I made a new capacitor for L2, that can handle 4kV
I made it from 40x 10nF wima fkp1 2000V dc

I ordered 24x 22nF 2kV caps to make it more compact.

This cap is smaller so I had to retune L3 again, I am playing with it while tuning.
Also put a L4 coil (0.75mm2) close coupled on L2.
I loaded l4 with a power resistor of 896 ohm. it gets so hot that it starts smelling.

1.71 A 2x18.1V input= 63 W going in from the supply, so that's no surprise

I have doubt how to tune it.
should I tune the impulse to the current max, where V is minimal (0)

Or tune it to the power maximum of L2 and L3?...

at higher input power the current collapse of L2 disappeared, very strange.

I will just have to play with it some more to get a grip on it.

putting a lamp on L4 could also be "enlightening" to see when output power is max

I think I need another cap coil, based on close coupling L1 and L2 again, so L3 won't be in the middle, but L2 will be.

L3 is then to be on the edge of epoxy, and with the series connection left open, so  a L4 coil can be close coupled to L3.

This setup would (without L4) also be ideal for the health benefits.

I want to do this, because both L2 and L1 are directly loading the power supply, and L3 is loose coupled to L2, while L4 (to be loaded) is only loading L3.

The setup of coils with L3 in the middle of L1 and L2  is interesting, but it pulls a lot of power from the supply, while L1 close coupled to L2 reduced the power input dramatically.

Instead of giving L2 a positive dc offset I could also give L3 a negative dc offset

But this would be dangerous for the health benefits as contact is made with the high voltage coil, this could only be safe if the person is not earthed (and that is possible, and also what Tesla suggested in his Buffalo lecture)

as a variation L3 can be given a negative DC offset.
L2 can then be close coupled to L1, and benefit from the power reduction.

Dc offset module needs to be altered so it stores the negative dc, on L3.
(red in picture)
 for high voltages, multiple diodes in series need to be used.

Quote from evostars on August 22nd, 2021, 12:28 PM

I have doubt how to tune it.
should I tune the impulse to the current max, where V is minimal (0)

Or tune it to the power maximum of L2 and L3?...

at higher input power the current collapse of L2 disappeared, very strange.

So you had established that L2 current sine wave seemed to get the biggest "boost" from the voltage spike, when it hits in BETWEEN the current's peaks (impulse hits when crossing i=0)? And that is a loss which we want to avoid? I'd guessed the boost was a good thing, but I suppose 'extra current' might well be a mark 'against' such timing/placement... Or, if THAT's the 'current collapse' which disappeared at higher input power, that is unexpected.
.
As an aside: you're aware, burning power resistors on internet videos, it's TOTALLY IN these days, might be a wise move.

Quote from coldelectric on August 24th, 2021, 06:28 PM

So you had established that L2 current sine wave seemed to get the biggest "boost" from the voltage spike, when it hits in BETWEEN the current's peaks (impulse hits when crossing i=0)? And that is a loss which we want to avoid? I'd guessed the boost was a good thing, but I suppose 'extra current' might well be a mark 'against' such timing/placement... Or, if THAT's the 'current collapse' which disappeared at higher input power, that is unexpected.
.
As an aside: you're aware, burning power resistors on internet videos, it's TOTALLY IN these days, might be a wise move.

Burning power resistors is IN? haven't seen that

And no, the impulse is tuned to the L2 +current max which then COLLAPSES, while L3 - current Max is amplified

At high voltage The L2 + current collapse dissappears.
Weird yes, but I can explain it:

the voltage impulse from the L1 coil normally transforms into a current impulse which collapses the L2 current.

Now that this does not happen at high voltage, means the energy of the voltage impulse is flowing elsewhere.

The voltage impulse collapses the high voltage charged dielectric field between L2 and L3.
My guess is, the voltage impulse energy is used to set up a Longitudinal displacement current, which amplifies L3 current.

I'm making a video showing this

normally with tem resonance the dielectric field moves side ways, transversly

with lmd it moves longitudinal.

but now with L3 close coupled to L1, is it really lmd? no not 100%
 so is it a mix of lmd+tem? yes probably.

as L1 is pulsed and creates the impulses when resonant it is TEm resonant.

So L3 has 2 forms of resonance?
hmm lets take a look

when L2 is lmd we can tune l2 down to the tem frequency (1/2 pi lower)

But if we do so, L3 will not create a single sine, like l2, so it seems L3 is tem resonant when L2 is LMD

Could that be due to tuning? Whould increasing the L3 capacity lead to LMD resonance again? I think not.

L3 is TEM resonant due to the close coupling to L1 which also is TEM

So the L2 is LMD and L3 is TEM.

this means the longitudinal dielectric field of L2 when discharged by the impulse, sets up a Longitudinal discharge towards L3.

the magnetic field current of L3 then is amplified by this longitudinal energy current from the discharge of the coil cap.

since L3 is TEM this means the magnetic field also moves sideways again. transverse.

So we first have a longitudinal energy current from L2 to L3.

then we have a transverse energy flow around L3.

to me that sounds like the start of a ring vortex. to close the loop back to L2 where it started, I would like to have L4 transverse again, to circle the energy current back inwards.
but as L4 is close coupled to L2, will that be possible? L2 is now the primary, so maybe it will.

then will L4 raise the frequency of L2 again? yes but only a little bit. and by tuning L4 to a lower frequency we can keep L2 also low.

Interesting but complex. L3 and L4 should then be the poles of the magnetic field.
centrifugal and centripetal, flowing inwards and flowing outwards.

I measured the blown L2 tuning capacitor board and to my surprise I wasnt 120nf but 60nf

only 1 nf cap was blown, so I thought maybe the added 60nF in parallel was broken.

adding it or removing it by the switch,
didn't do much.

now I looked at the board connection of the added 60nF, and it is not in parallel at all! I hooked it up to the wrong lead.

so the experiment was done with 61nf on L2 not 121nF.

the L3 tuning is connected the right way and measured : 120 nF

so... hmm my new 4kV L2 cap is 134nf. way... to big

my new l2 cap is  134 nF 4kVdc
but now I found out the original experiment had only 61nF.

So I will cut it in half, and make it 61 nF also. so I can replicate the experiment.

on the other hand I am thinking about using the 134 nF first. But that will drive the L2 current way up and it already was 19A pp so... I could try to tune with it.

more importantly I thought I had L2 and L3 at almost the same capacity.

but now it makes more sense as I already found out L3 wasnt LMD resonant but TEM.

then again... I prefer having L3 LMD resonant as it showed amplification.

so... what to do? I guess I'll just have to try again. play with it.
maybe I wont rrpeat the experiment that ended with a bang, but use different cap values. bigger. so both l2 and l3 are lmd resonant

now I think about it L2 was to small,
now it's bigger, I might have more chance in tuning L3 to LMD resonance.

on the other hand. L3 should have already been LMD, having a larger capacity so, why could I tune down the frequency of L2 to Lmd but not for L3?

was L3 Lmd resonant if tem? it was out if phase with L2, but L1 was reversed...
grumble this messes it up

I based the feedback loop on LmD resonance of l2 l3. thats why I phase reversed L1,
but it isn't that does explain why I need to pump a lot of power into it... the idea was to reduce it.

what is going on?

hmm I really need to play more with this

I now have 3kV diodes, but I can get up to 3.5kV so I will add another 1kV diode to the dc offset circuit. The higher the better, as it just started to work at 2900V

I added the extra diodes so have a max of 4kV now.

I burned my dual mosfet driver. Instead of 12V I gave it 31V. Luckily I had a fuse, but it was to big. The parallel diode shorted out, and the circuit was damaged.

So I need t9 make a mew C3M circuit. Luckily I already have the new pcb for it. And the parts. Only need to solder it. And see what fuse would be better.

Meanwhile I have the old c2M as backup.
Now that L1 has less inductance it draws a lot more amps.

I discovered that L3 is not in the lmd resonant mode, but TEm. Meaning if I tune L2 down in frequency L2 will go into Tem mode (from lmd mode) but L3 wont.

L3 is already at the lowest resonant frequency.

This makes makes me doubt if I should have flipped L1 over. So to check, I will reverse connect L1 and compare the current draw.

I also again hooked up L4 close coupled to L2 and loaded with 896 ohm (4x17W) which again became smokin hot. But the input was also high.

I need to rectify L4 to dc to see the real power

So again a lot of confusion. I need to get things straight and compare them.

Also the L2 tuning cap is now 12x 22nF in series parallel making areound 65nF (I cut the nee tuning cap in half)

I compared L1 flipped, with L1 flipped and reverse connected.

reverse connected and flipped L1 pulled much more power from the supply.
There was no phase shift of L2, So I could not tune the impulse to the power max.
I could tune BOTH L2 and L3 to TEM and LMD.

With L1 flipped only, I can not tune L3 to TEM.

That leaves the question, If I flipped L2 instead of L1, could I then tune L3 to LMD and TEM mode?

I flipped L2, and compered it to flipped L1.
It gave the same results on lower voltages.
But when L2 is flipped, then L3 and L2 are in phase at LMD, this makes the current of L3 go in the wrong direction to be amplified by the displacement current of the L2 L3 coil cap.

So L1 is flipped, is the right way to go (for now).

question stays, that when L1 is not flipped, L3 can go to the lower TEM res freq, together with L2.

Why cant L3 flip phase to TEM resonance when L1 is flipped over? it stays at the higher LMD resonance, when L2 is dropped to TEM resonance mode...

OK, need to rectify L4 to DC, and load it and measure it. "hot" is fun but not accurate.
And it would be good, to put it on video, so it is clear what the in- output ratio of power is.
Then I can compare it with and with epoxy.

once its in epoxy. I would like to be able to play with L4, so I guess I will need to keep it out of the cast.
Top L2 off to the right level, so L4 can still close couple.

After filming an publishing this feedback setup with yhe L3 parallel resonant secondary in the middel, close coupled to L1, I will continue again with a more conventional setup stack.

The power consumption is very high now, and I bet the output does not show an efficient system.

The feedback loop is a nice idea.

The power peak of series resonant L2 is not amplified but reversed in polarity during the peak.

L3 is amplified in current, but cant properly couple it back to L1 (I think it might need some more tinkering maybe).

So the new idea will be to close couple L1 L2 again. and place L3 loose coupled on L2.
(as this reduces power draw from supply)

But this time, L3 will be given a dc offset.
So now L2 will be the positive plate. And L3 the negative plate.

L2 negative impulses than still discharge the l2 l3 coil cap.
L3 still gets the current amplification but now on the best point. Where current is zero.

L3 and l2 will both be lmd.
L4 output coil to be close coupled to L3.

Back to filming and soldering a new gate drive pcb

Made a new gate drive circuit as I blew up my previous one by putting 31V on the 12V port
 :roll:

Then I started testing it. Thing wasn't working as supposed.
Blew fuse. High amps at low voltage.

Took my infra red camera to see what got hot.
Saw the reverse diode burning.

That's where I should have thought hey... That means....

But no. I yanked it out.

Tested again. Still trouble (duh)
Now the 5v regulators got hot.
Checked polarity, was fine.
Still didn't think, hey that means...

Yanked those out also.

Checked again.
Still trouble...
 :shocked:
Dc dc converters got hot.

Hmm. What is going on.
You know what. I'm going to take a break.
Walk down stairs.
Then it dawned in me.

Did I connect the 12V wrong way round at the psu?

...
Yep :huh:

Worked to long without a break. Lost concentration.... Yes it happens.

One of the mosfets is dead

I replaced the dead mosfet, and the circuit is running again.
But the voltage at the gate of the other mosfet doesn't reach +15V and the circuit draws more amps than usual , also the impulse is slower and thus lower in voltage,
So I conclude the mosfet does switch but is also damaged, causing higher resistance.
I will replace that mosfet also.

replaced second mosfet, my component tested read it as a resistor, so that makes sense. threw it away...
Now both gate voltages are proper again.
Circuit is working.
back to testing and filming