so the cold current that charges the parallel capacitor very fast is high.
but only from the charging side.
that would mean I could use a resistor to connect it to the dc?
leaving the other cap leg connected to the drain/coil without resistance.

or... hmmm its not clear yet.

if the mosfet grounds the coil
will the voltage on the other side of the coil where it is connected to the  choke be zero/grounded also?

the choke at that high frequency will act as a high ohm resistance.
the coil has low resistance / but some reactance?

reactance of a coil is high at high frequencies so that would be good. but is the coil magnetised? No? then reactance is high and a voltage difference can be developed over the capacitor

I disconnected the push pull coil,
and connected the choke and current coil with parallel caps.
I already tuned to 60.5kc/s and 7.7% duty cycle.
Measured on both sides of the current coil.
Voltages are equal and appear to resonate at 237 kc/s
I used 10V on PSU measured near zero current.

So the current coil appears resonant but not at 60.5 but around the 4th harmonic 237kc/s.

The hv coil is grounded on one end through the push pull drive transformer, so the capacity of the coil capacitor could play a role, and also the inductance of the choke could interact.

not what I expected...
clearly the 5uF cap doesnt play a role, as the voltage is to high for that. 39Vpp
the negative half wave is clipped, due to the mosfet body diode, conducting negative to ground.

ok last test was confusing.
what happens when the HV coil is connected, and made resonant?

at least I can tune the phase shift

HV coil is now driven with 1.7V which already gives 550V pp
The ring of the current coil disappears, when I raise the voltage (moves to the right)

the 10V discharge is now more visible, with the high freqeuncy ring it gives.

I will adjust phase, to the right position (max positive on orange= HV coil)

Very cool to see that high voltage spike appear on the current coil,
the negative hald is rectified through the body diode, but then it rings up, this is from the DC offset that slowly charges up.

BUT I only give 10 volts. the offset appears to be higher!

COOL stuff is going on here

I tuned the discharge to the max voltage point, and gave more power to the current and voltage coils.
push pull is now 0.7A 3.7V dc giving 1200Vpp

Weird is to see the ringing of the current coil is depended on the both voltages. I can tune it so a single wave appears on the current coil (which should be a single wave)

this was set higher to 1.00A 5.2V DC push pull giving 1650V pp

current supply is now 18.1V 0.00A

note how there isnt 18V discharged as the resonant voltage is low
and note the current coil is not resonant as both ends have the same amplitude

lets tune to the discharge to negative voltage maximum and see what happens there

jikes I turned down the PSU voltage back to zero, but the voltage of the PSU stays at 11V.
This means the body diode is delivering voltage back to the PSU from the ringing coil.
Also the ringing is above 100V. This could damage the mosfet which is 100V rated.

So I tried to dial down the voltage, went to zero, and the ringing continues. hmmm
very interesting effects

ther discharge was tuned to the half - voltage point of the hv coil. this gives very high ringing voltages, (also on the+ half voltage point

wow
still zero volts supplied from the curren coil PSU, still it reads 6.4V 0.00 A (voltage dial turn all the way down).
I needed to reduce the voltage of the push pull HV coil to 3.6V 0.67A dc  to avoid burning my current coil mosfet (100V limit)

now a resonant half wave came out of nowhere right before I turned on the mosfet, at the max negative voltage of the HV coil.
I had to delay alot, but cool that this delay works so far. Well designed IVO!!!

so there still appears to be no parallel resonance, but I expect the voltage to remain low, so I should really measure with the current probe, to see if current (hot) is flowing.

So the coil capacitor charges up to a negative voltage, around -500V
and the current coil then out of the blue desides to become resonant,
then the mosfet discharges at that maximum (I could tune it like that) it gives another ring and then dies out again.
All while only driven from the HV coil push pull.
The current psu reads 6.4V dc 0.00 A

MAn this is a weird kind of resonance.
must be parametric or something to behave like this

BTW I shorted out the output coil and it is not grounded.
The first time I saw the current coil ring, it was opened and also not grounded.
So that doesn't look like a problem.
lets ground the out put coil, and see if it makes a difference (while being shorted out)

the irf510 mosfet which has no heatsink feels cold to the touch (from reverse cold current through diode?)
The sudden current coil ringing is to high in voltage, and I don't want to kill my mosfet
so... lets give this a rest, and process what the heaven is going on ???
Very exiting to work with this

also need to hook up current probe.
maybe give a load to the output coil, to eat away some voltage
oh boy.... this is FUN

the choke coil must be resonant.
at 237kc/s and 1440uH it would need 313 pF

it would not suprise me if that is from the output coil, and current coil. lets measure that

378pF @10kc/s measured
265pF @100kc/s measured
Between output coil (shorted out not grounded) and the current coil (not shorted, with 5uF parallel, and choke in in series).
so around 60kc/s yes 300 pf or something like that.
makes sense.

OK so the choke coil is resonant. I can prevent that.
I will wind another coil on the choke toroid with equal windings (100) and put a diode to ground, that should kill that resonance.
since the choke is low side switched it gives a positive swing first.
If I reverse the second choke coil polarity, I can give it a negative diode to ground. else it would conduct the DC supply to ground.
back to winding.

the diode could even prevent the back flow into the psu. it should be a fast one. MUR440 comes to mind

the output coil with have earth ground on one side. and only this is earth grounded. no other part of the circuit is earth grounded.

the HV coil highest resonant voltage side is opposite to the output coil earth ground side.

the mosfet switch to neutral of the high current coil is also on the opposite side of the outside coil earth ground side.

so the push pull driver coil is connected to the HV coil at the side of the output coil earth ground

and the choke coil is connected to the high current coil at the output coil earth ground.

Since my choke is resonant, I have the frequency and thus I know the capacity.
so then I can also calculate the ideal quarter wave resonance that I would need for charging up the dc offset to its maximum voltage.

thus then I know what my inductance of the choke should be.

Because right now it is much to fast and to high in power (1W).

so I will use an iron core toroid again to make 2  coils in it and then, it should work.

237kc/s and 1440uH , 313 pF is the resonance of the choke.
But that is induced by the high voltage coil resonance.
I dont want that to be the power source,
I just need the DC offset.
But... that also charges the same capacity.
So RC is the rise time but only up to 65%
so if I want to charge higher in the same time,
than it would not be 4R but 1/4 R
the time contant is thus 1/4th instead of 4

right.

60.5 kc/s=16.53uS period. I have a 8% duty cycle= a total of 15.21 us of time available for charge, so this is the 4T
T=RC but now 1/4T=RC
so what is R?
T/(4C)=15.21us/(4* 313pF)=1215 ohm

And this is what the reactance of the choke should be at 60.5 kc/s
so the inductance needed for the choke then is 3.20 mH

And this is to be wound twice on the toroid.

6 turns gives 433uH
I need 3.2mH
So I would need approximately 45 turns twice.
That is do able

26 cm for 6 turns is 4 +1/3 cm per turn. 45 turns is around 2 meters of wire times 2 is 390 cm plus leads is 4m of 0.44 wire is needed

:D haha turns to inductance is not linear!

I got 103mH and 109mH ... way to much, but lets see what it will produce
39154 and 41434 reactive ohms... lol extra low current.
I bet I will see almost no voltage rise....

32V (max dc PSU) / 40k=0.8mA gives 25.6mW
very very low.... well ok, shall I do a test or... remove some of the windings.

edit: I removed 15 and 14 windings, now I am down to 20 and 21 mH
Still high, but This makes a lot more sense.
7603 and 7983 reactive ohms at 60.5 kc/s
32V/7603=4.2mA 0.135W
32v/7983=4.0mA  0.128W

scr775 shows the current coil on both ends, (voltage) with the hv coil not powered.

scr776 the same, bu now the HV coil is given a very small amount of power (1.1V)

clear is that the choke is still resonant, as it should be, since I still have no diode to ground on the second choke toroid coil.
now lets connect the ultra fast shotky diode to neutral and see if we can kill that resonance, and get a dc offset saw tooth like signal on the coil

inductance is higher, thus resonant freqeuncy of the choke is down, so far that is good

I have put a mur440 to ground allowing negative voltages from the second choke to ground.

I turned off the current PSU and still it read 5.6V when the HV coil PSU was turned on (3.0V giving 1100V)

This time the signal is different, I tuned the discharge of the HC coil to the positive hv coil maximum again.

measured again is the voltage on both ends of the HC coil


scr777 is with only power from the current coil psu (6.7V)
scr779 is with only power from the voltage coil psu (3.0V)

So now the resonant half wave appears to be still present, but the frequency has lowered even further (from the added choke coil to ground?)

The signal doesnot appear to be clipped off on the negative half.

There is high frequency ringing when the HV coil is on. where is this coming from?

voltages are still the same on both ends of the current coil.

When tuning the discharge to the voltage maximum, I cut into the positive voltage, so this time there really is a displacement current generated.

But... the voltages on the current coil rise to fast, together with the voltage coil. I need to protect the mosfet from over voltage.

even higher freqeuncy ringing occurs when the mosfet is discharging the positive voltage. this is on the drain side, so it would be wise, to also put a snubber circuit here, to drain away the noise.

I didnt chop the extra wire of the choke off, which came from the removed 15 windings on both ends.

I also now grounded the output coil on the choke side. leaving the other end open. maybe I should use a 4 ohm power resistor here would be fun if it gets hot

another strange phenomena is that the current coil discharge at start up, doesnt appear straight away. it then starts to flicker, and then it stabelises, but now... on the negative maximum of the High voltage sine wave.
Weird, as I just tuned it to the positve maximum. I have seen it jump like this before.
It might have to do with the not ideal signal from the delay IC.
I do not yet fully understand what is going on.
but again definitely have an effect on me. I feel tired again...
I have not been grounded, and have not been using my shield.
time to do so...
first a break to recover and think these results over.

one thing: I might need to reverse connect the second choke coil

does a displacement current induce magnetic current?
if so, does the displacement current imduce a magnetic current in the current coil,
which then becomes parallel resonant with the parallel large capacitor?

So the displacement current is not directly charging the parallel capacitor,
but first charges the current coil with a magnetic field.

And, what if the choke coil is meant to be resonant?
then a combined series parallel resonance would exist between the current coil and parallel capacitor, and the hv coil capacitor, (and output coil?)
and the choke coil.
wow. that is pretty complex :huh:

The choke causes unnecessary problems that are easily solved.

instead of grounding the resonant secondary of the push pull driver, which provides voltage to the high voltage coil,
I will connect the high current coil to the other side, of the secondary, which was grounded before.
The high voltage will then be present on both coils, but out of phase.

Then the discharge mosfet is used to ultra fast discharge voltage on the current coil.
while also connected to the secondary of the push pull coil. (where before it was grounded).

This brings a slight question, of the body diode passing the negative voltage again, but this can easily be solved with a series diode to ground from the mosfet source (or even at the drain).

the induced displacement current will again make the coil parallel resonant, providing the magnetic current for the output coils voltage.
But this parallel resonance, how does it relate to the series resonance?
hmmm...

And the big question is, how do I keep my voltage low of the mosfet that is connected to the drive push pull secondary coil.

Nikola Tesla in his Colorado springs notes already mention this, whereby an extra coil can be connected in series with each side of a transformer.

Very intersting, and easier to work with/design

I will need to lower the voltage again of the resonance, since the secondary will become half wave resonant, instead of the the grounded quarter wave resonant. So I will need more turns. More inductance to get the frequency down.

and the freqeuncy is now known, as it should match the parallel resonance, 60.5kc/s
So If I connect all if it now, it will function on 121kc/s
The caps dont like this, and will run hot.
but at least, I can already test this, and see if there are unforseen complications

I connected and tested, but without the discharge mosfet connected (expected to see high voltage and I am right)

around 470Vpp for the high voltage coil
and 349V pp for the high current coil. voltage difference might be from the different probes.
Also hooked up the other side of the current coil which gave the same frequency as expected.

but not as expected, the frequency is 64.7114 kc/s  much lower why is that?
I now have the half resonant mode, but I guess it has all to do that the 2 extra coils are capacitively coupled.
The voltages are nicely out of phase which is good.

BUT
the amplitude is way to high! I only had 1.8V on the PSU.
So a irf510 mosfet with its 100V limit would be broken.
Also... one half of the current sine would be rectified due to the body diode. but as said before, a series diode can fix that

So what to do now?
first I can get rid of the series resonant primary,
then if needed I can add more windings to the secondary.
this will give a lower resonant frequency.
but then I would also need to further increase the parallel capacity of the current coil to tune it to the lower frequency.
That would mean even more current, and lower parallel resonant voltages... hmmm
 lets just start with the easy bit.

Remove the capacitor from the primary, and tune again the the secondary frequency and see what is left of the frequency and voltage.

0.4A 1.15V PSU DC
removed the 1uF series capacitor of the primary push pull coil.

frequency dropped to 51.8 kc/s which is a bit weird but good

I removed the hv probe (ornage) and used the green normal probe for the HV coil. which is now 364Vpp
while the current probe is 334Vpp so just a small difference. just a bit less on the current coil. (more series capacity?)

so again, very high voltages for low levels of PSU.
I like the resonant frequency where it now sits,
so since the current is already relativly high on the primary, I will give the primary more windings.
this will get the resonant voltage down, the mutual inductance will change so the resonant freqeuncy should als again change.
With series resonant primary the impedance was low, so now the impedance goes up, and will again drop.
lets see if that fits reality

edit if freqeuncy goes way down, I can even remove windings from the secodnary to get the freqeuncy back up, and lower the voltage