I will advance the radiant power circuit, sharing in my 2019 video, with the L4 extra coil. in search for LMD resonance amplification

LMD= Longitudinal magneto-dielectric
-lmd, gives 180 degrees out of phase voltages between primary (L2) and secondary (L3)
-secondary (L3) becomes hot in LMD mode
-LMD has its resonant voltage and current IN Phase (producing power!)

TEM= Transverse electro magnetic.
-TEM gives in phase voltages between primary (L2) and secondary (L3)
-Primary (L2) becomes hot in TEM
-TEM has voltage and current out of phase (no power)

Amplification of the L3 LMD resonance, by the L4 Tesla Extra coil.
The extra coil will have no external capcitor tuning. it will have very little capacity and inductance, and therefor it will resonate at a much higher frequency than the LMD L3 coil parallel tuned coil, that resonates at the L2 series TEM resonant frequency.

The L4 will be a higher harmonic of the L3 LMD frequency.
due to L3 having current and voltage in phase, the L4 will also have current and voltage in phase. (assumption needs to be tested).
By having less capacity and inductance, the frequency is higher, but also the LMD current and voltage of L4 will be higher, giving amplification.

Basically this is a nested ring vortex. the shape of these ring vortices might determine the amplification, so coil size (diameter) will also be looked at.

I will step away from the dual primary setup, and only use the negative impulse primary. (maybe compare it with a positive impulse primary).

So it will be a stack of 4 coils.
L1 and L2 close coupled, as before, with high side switched L1
L3 loose coupled to L2.
L4 coupled to L3 (probably tight, need to research).

First I will make a 4mm2 diameter coil, with 6,580m speaker wire, which gives equal mass to L3 L2 and L1.
L1 and L2 are 10m of 2,5mm2 diameter, speaker wire bifilar pancake coils
this coil will first be tested as a L3, but probably will give no results, than It will start as the L4 coil, the  Tesla Extra coil.

the title says it all great 
any picture video of this ?

Quote from securesupplies on September 2nd, 2020, 07:40 PM

the title says it all great 
any picture video of this ?

not yet. first video about current dual primary setup.

It works!
yellow is LMD resonant L3 voltage
orange is L4 extra coil. voltage.

coils from bottom to top
L1 L2 - L3 L4

circuit is april 2019 Radiant Power circuit from video (master Ivo)

input is 0.43A 58.3V dc

L4 current is 6A pp in phase with 2625V pp
at 94.26kC
L3 is around 3A pp and 1688V pp

L4 holds 15kW

Now, need to load it, and rectify to dc to measure real output.

Awesome! I found this work in general through Tesla. I then discovered the Stan Meyer stuff and decided to work on that first because it was more practical and applicable to our modern (current) industrial setup..........however, my mind is always still on the work of Tesla.

Well Done Buddy the 
interesting thing about this arrangement is the LMD measurements

Ok I needed to understand loose coupling.
this link helped:
https://books-guides.com/power-amplifiers/

scroll down tho find "mutual induction"

So now we not only have magnetic induction (which loads down the primary from the secondary)

but we also have dielectric (longitudinal) induction.

We need to find the right distance between L2 and L3.

too close and the L2 will be loaded down from L3

to far, and the L3 will barely feel the dielectric induction.

and, when it is to close, we get 2 peak frequencies, from the increased band width and the loading from L3 on L2.

is it the right distance  or right angle/resistance

I am in search of the right coupling distance between l2 and l3

since all values change when I change the distance, frequency power in power out changes, I need to create a dc output load,
so I can measure input vs output, to determine what has the best ratio.
output comes from L4.

only distance will be changed.

the idea is L2 and L3 have very different aether field forms. L2 is a like a water vortex
L3 is like a water ring vortex.
they need space but also need to interact.

with coupling, I am not in need of power transfer, which happens when impedance is matched.

Instead I need maximum voltage tranfer.
this is known as Impedance bridging.

I already am familiar with this from my sound engineering background.
When a microphone enters a mixing board, the Impedance is not matched, which would load down the power of the mic which is very low.

This voltage bridging is between L2 and L3.
Between L3 and L4 the opposite happens.
L3 is high impedance, and cannot transfer power to the load, for ideal power transfer the source impedance must be equal to the load impedance.

L4 is not parallel resonant. altough it is close coupled to L3 (loading it down)
 L4 is also series connected to the LMD  resonant end of L3.
L4 gives the output on its resonant end, relative to earth ground.
The impedance if L4 should be much lower than the impedance of parallel resonant L3.
So L4 should be able to much better drive a (low impedance) load than the high impedance L3.
Ideally, L4 would be made series resonant, this is what Nikola Tesla did with his top capacity (ball on a stick connected to extra coil).

Instead the input impedance of the mixer is made large relative to mic impedance , so maximum voltage is transferred. This low power voltage is then amplified by the mixer.

same thing here. L3 is parallel resonant high impedance, L2 is series resonant low impedance. We dont want power from L2, we need voltage from L2 (impulse, longitudinal dielectric)

close coupling will load L2 down with the high impedance of L3. this will dampen the voltage amplitude of both sine wave and impulse.

https://en.m.wikipedia.org/wiki/Impedance_bridging

L2 L3 is like the microphone and mixer, connected by impedance bridging

L3 L4 is like the mixing board which amplifies the signal and gives it more power.

This voltage bridging is between L2 and L3.
Between L3 and L4 the opposite happens.
L3 is high impedance, and cannot transfer power to the load, for ideal power transfer the source impedance must be equal to the load impedance.

L4 is not parallel resonant. altough it is close coupled to L3 (loading it down)
 L4 is also series connected to the LMD  resonant end of L3.
L4 gives the output on its resonant end, relative to earth ground.
The impedance if L4 should be much lower than the impedance of parallel resonant L3.
So L4 should be able to much better drive a (low impedance) load than the high impedance L3.
Ideally, L4 would be made series resonant.

This is what Nikola Tesla did with his top capacity (ball on a stick connected to the extra coil).

I put a load on the extra coil.

I dont remember correctly so I have to retest
 (dentist appointment sedation messed me up). :-(

I first put the 15W parallel over the extra coil. That didn't work. as expected.
Then I put the 15W lamp in series with extra coil to ground. that worked.
then I rectified to DC and loaded again with 15W lamp, this time lamp didn't work. But That was a mistake, had had not connected the DC caps to ground.

once I realized my mistake, I tested again. extra coil rectified to DC, into 2 caps (one plus one minus) in series, with center ground.
This time I could bring the lamp to full brightness. (but very low efficiency)
 :cheerleader:
The extra coil signal into the rectifier diodes, was a clipped high frequency wave. much higher frequency then the L3 sine.
This tells me, I need to tune the L4 extra coil.

Ideally I would use a series resonance by adding series capacity, because this will again lower the impedance, which is perfect for power transfer (equal low  impedance source and Load).

there are 2 options for series resonance.
one, is place a capacitor between L3 and L4 (this would mean the loose end of L4 whould behave as ground, not swing... very doubtfull)

The other is put a capacitor in series with the loose end of  L4, like a classic tesla coil. but... what then to to with the capacitors loose end? put it to ground? I guess I have strong doubts here, but testing is always truthful

then there is the option of parallel resonance. this could work, but will give a much higher impedance. not ideal for output power transfer (no impedance matching, source is much to high load to low), This could still work, if the parallel capacitor of L3 is much bigger, than the parallel capacitor of L4 (very small).

The L4 when unloaded now shows a amplified sine voltage, bigger in amplitude than L3 (perfect matched sines). But this is only from close (capacitive)  coupling of L3 and L4.
In reallity, L4 is very high in frequency (as it is untuned, which explains how the DC rectified signal shows the high frequency signal below the clip off threshold).

We are dealing with LMD resonance here, so the regular more common known rules of TEM resonance might not apply , thats why we test...

Another idea:
L3 and L2 can be equal size coils. L4 ideally whould be smaller in diameter than L3.
this is because smaller, means less windings, less inductance, less capacitance, which will amplify current and voltage. but I doubt it is necessary. For now I keep using 4 equal coils, as I still need to tune. So adding capacity, will do a lot. but this means inductance is the same. I should measure again, and see if current is still amplified, with the equal l3 l4 coils.
Earlier I had a big L3 and a smaller L4, which doubled the current. but its a resonant transformer, so I doubt this. still... need to test.

This might be a wild idea, but it also might work.

L4 amplifies the signal of L3 which is parallel resonant.
L4 for maximum output should in theory need to be series resonant. (impedance matching to load).

Adding a series capacitor to L4 requires it to be discharged some how.
If we add 2 series capacitors (in parallel) to L4 we could make a rectifier from them. getting DC out.

Idea, is the amplified LMD resonant L4 voltage, charges the C4 capacitor, by having the cap connected in series to ground by a diode.
When the L4 voltage polarity changes, the capacitor is discharged back into L4 (partly) and overflows into the capacitor bank, where it is DC, and is put under load.

Same process for the other polarity, except now the diodes are placed in reverse, so it works for the opposite polarity.

Goal is to get L4 into series resonance with the cap. So its impedance will be low (wire resistance), so it will be able to power the load (also low impedance, thus impedance matched).

@@@@@@@@@@@@@@@@
Also, L1 and L4 do not need to be bifilar anymore. the L1 impulse is faster when L1 is not bifilar.
L4 output is more amplified, when capacity and inductance is low. low resistance also gives the best results for both coils, so If L1 and L4 were bifilar, disconnect their series connections, and instead put the 2 windings in parallel.
Sadly this also reduces L1 inductance. giving a less energetic impulse, requiring more voltage on the input. but the impulse would still be faster. than with a bifilar coil.
The question still remains if this is what we want? As the lower resonant freqeuncy of L2 (i called it TEM mode  but still need to test if L3 is also TEM), if L3 is also LMD in phase current and voltage I could still use that one. the only trouble than is the impulse needs to be much slower), does need the least amount of power to create resonance.

Also, since L2 and L3 are transferring voltage, a high K dielectric would be best here (to fill the distance of the loose coupling between L2 and L3

Keep in mind
L3 and L4 both voltage and current are ONLY IN PHASE when a (resistive) load is present.

when pure resonant, with no  load, current and voltage is out of phase 90 degrees as is "normal"

l3 and l4 need to differ in size.
L4 needs to be physically smaller, with less turns.

I noticed current wasn't amplified with equal coils

I think I need to do dual test.
one with high side switching, negative impulses

And One with low side switching positive impulses.

See if there are differences.

What I noticed, is the positive impulses only see the mosfet drain, are faster, higher voltage, and can become fully absorbed into series resonant L2.
this does not happen with the negative impulse.

It seems positive and negative, is determined by the faraday tube end point on the conductor (dielectric line of force)

one end of the faraday tube is on the material (magnetic) part of the conductor, the other is in the "non material" (dielectric/counterspace?) part of the conductor.

one end it the start the other is the end of the faraday tube.

And I still wonder if the DC offset really is needed? L3  and L2 are distances so far that a dc voltage is not helping.
but L1 and L2 are close coupled, so hear it should work.

is the power supply current still reduced if L2 has no dc offset? I believe I already showed this in my april 2019 Radiant Power video. No dc does not effect it.

The impulse is a higher current event that helps build up (and amplifies) the magnetic field (current) of L2 (series resonant) which is close coupled to L1 that in turn is also assisted in producing the impulse from its magnetic field.

L1 shows the same current sine wave as L2 (with less amplitude)

I used a low side switch setup and compared it with and without DC offset.
both showed the SAME results.
stepping away from DC offset. it has no benefits at all

As positive impulses also work, i won't need an isolated gate driver...
IXDN14PI is the IC that I plan to use in the future cheap and powerfull.
it has a wide range of voltage use, delivers up to 14A.

to get a stable independent voltage from the supply, a buck boost converter will be used set, to the voltage the gate of the mosfet needs (20V)

I intend to design a simple drive board, for the IC and the mosfet, gate resistor, capacitors(to deliver 14A pulses), power connections, and the coax for the logic input.

This will make everything SO MUCH EASIER!

Hello evostars.

As I already wrote in the additional field of the PayPal donation, your work is remarkable. I am currently in the process of rebuilding your tests.

I have a few questions, after I read your posts.

I cannot find this "IXDN14PI", do you have a link to the component?

Have you ever tried to change the square wave signal so that the pulse width is above or below 50%? I would be interested in what happens then.

Why did you go away from your first attempts with the enamelled wire on the empty CD cases?

A test with different dielectrics would be very interesting!

I hope we can exchange more information about this topic in the future.

Bye Jan

driver IC
https://nl.mouser.com/ProductDetail/IXYS-Integrated-Circuits/IXDN614PI?qs=8uBHJDVwVqxiweeEKpTHmg%3D%3D

yes I tried different duty cycle but doesn't work, impulse is created when signal is turned off.
impulse needs to be on top of series resonant voltage sine  wave.

CD coils had to much resistance.

yes different dielectrics would be nice, but difficult. I tried refined sunflower oil, which is a good dielectric but it messed up my coils. the hotglue detached.

SiC mosfet:
 c2m0080120D or C2M0160120D

gate driver IXDN614PI  driven by logic level square wave 50% duty at resonant frequency L2/C2

XL6009 or XL6019 based step up/down dc dc converter to power the gate driver.

L1, L2 and L3, L4 are both close coupled
L2 and L3 is loose coupled.

L1 produces the impulses
L2 is series resonant recieving imlulses, and is the primary
L3 is parallel resonant and is the secondary
L4 is the extra coil. (this is new and unclear)

I left out the power capacitor over V+ and V- use a 1uf or bigger polypropyleen.

switch is for high voltage use. Dc dc converter (xl6009 based step up/down) can only take 32V. If V+ is higher then use a separate powersupply for the dc dc converter.

Quote

driver IC
https://nl.mouser.com/ProductDetail/IXYS-Integrated-Circuits/IXDN614PI?qs=8uBHJDVwVqxiweeEKpTHmg%3D%3D

ok I was looking for a mosfet, not a mosfet driver :lol:

Quote

yes I tried different duty cycle but doesn't work, impulse is created when signal is turned off.
impulse needs to be on top of series resonant voltage sine  wave.

if you have already tried this, then this is an incentive to try it yourself. Basically it is the mosfet no matter how long it is switched on or off. But if the pulse width was 25% I could theoretically increase the frequency by a quarter compared to 50 pulse width.

Quote

CD coils had to much resistance.

Do you still remember the diameter you took then?
i will now wind a test coil with 1 mm². The advantage of the flatter wire coils is that the wires are close together and the space in between could be filled with dielecrum.

Quote

yes different dielectrics would be nice, but difficult. I tried refined sunflower oil, which is a good dielectric but it messed up my coils. the hotglue detached.

I will also test this

Quote

SiC mosfet: (...) C2M0160120D

The "Turn-Off Delay Time of 14ns" is unbelievably fast :blink: :blink:

ok my test setup is now based on your videos from 2019.

in time i will either let it run parallel or test the basics first.

good work, continue ! :thumbsup:

bye Jan

I just ran 2 equal tests. just L1 and L2 close coupled, No DC offset (done with that!)
L2 tuned series resonant with a impulse on its voltage maximum.
changed the L2 series capacitor to get differnte values.
kept the impulse voltage at 1200V (!) with all tests.
Measured, input power, L2 voltage and current.

2 sets of equal tests were made and compared.
one with high side switch,m negative impulses
other with low side switch positive impulses.

WOW. yes. there is definitely a differnce.
negative impulse, needs more power, to get to 1200V,
but also, series resonance is increased in voltage and current. in comparison with low side switch.

Now why is that? Is there a Loss of the impulse energy? does it leak away via the isolated gate driver? which is connected to the source of the mosfet, where the impulse also resides with the high side switch?

With the low side switch, the impulse does not see the gate driver circuit, as the impulse is now on the drain side.

The impulse, if it looses energy into the gate driver circuit, should show up in the gate driver voltages. BUT IT DOESNT.

Where is the impulse energy going then? Is the impulse energy helping the series resonant L2 coil to build up its magnetic field? and thus loosing energy? requiring more from the supply to reach 1200V, and thus ALSO amplyfying L2 more, as it also gets more voltage?.

to be clear some data:
Low side switch:
L2 cap    F res     L2       DC POWER              L2          impulse
61nF   73.65kC   5A   0.15Adc   2x11.8Vdc   160Vpp    918nS 1200V (used DC offset which slows the impulse a bit down)

high side switch:
60nF   72.95kC   6.2A   0.22A dc  2x18.7 Vdc 320Vpp  870nS   1200V (impulse faster, no DC offset circuit)

with the low side switch, the higher in frequency, the smaller the cap, the LOWER the amps.
"     "       high  side  switch,"   "       "     "               "       "         "     "       "    HIGHER the amps
That is NOT supposed to happen.  The smaller the cap gets, from 60nF to 31nF the HIgher the amps go, 6.2A to 9,4A
thus the higher the frequency, the higher the amps go.

That is not explained away with the isolated gate driver influencing the impulse.

SO ...
Finally
the conclusion.
NO DC offset is needed. which is great, as it is very dangerous to have 1000V dc on your caps.
the high side switch is needed, and thus also the isolated gate driver.

L2 in the attached picture, is series resonant and recieves impulses (inductive spikes from L1)
this L2 coil, is how nikola tesla had his primary. So for those who like to build tesla coils... build the primary like this.
 
MAybe I should give it a go one time. And see the difference in quality of the streamers.

Keep in mind, the gate driver, should be ISOLATED, since the source voltage changes.
SiC mosfet:
 c2m0080120D or C2M0160120D

Im going to do the test again, without changing the voltage supply, and see if the current really is amplified by the negative impulse. and not by the positive impulse

Compared impulse polarity differnce again,
now with equal switch, and tuning board. All was the same items.
made 2 screen shots, but... failed menu's are in the way LOL
 :rofl:

Low side switch:
L2 cap   Fr           L2 sine     impulse voltage and duration  L2 amps   DC power
41nF    86.85kC   440Vpp      1200V         832nS                 8.4A pp    0.29Adc 44.0Vdc

high side switch:
41nF   86.75kC    440Vpp      1180V          850nS                8.1A pp     0.29Adc 44.0Vdc

results are almost the same. high side impulse (negative V) is a little slower 18nS, and thus a little less high in voltage 20V.
current is 0.3A lower, this is result of impulse and measurement mistake (current probe varies in position).

CONCLUSION
with the same frequency there is no difference between positive and negative impulses.
Test was done with L1 L2 close coupled, no L3 or L4 coil coupled or present.

Now I need to test again, with different frequencies, and compare them again, see if there is a difference in response.