connected the choke to one of the parallel cap sides.
Frequency is again tuned down to 15.46 kc/s
voltage and current simultaneous max.
but again that phase shifting, from the current coil.
hmmm.
ok.
I need to process this.

I wonder, could I change the phase, if I power the parallel resonance?
connect the mosfets, and see what happens, see if I can shift the phase.

Voltage is low in the parallel resonance, but still, the body diodes of the mosfets will conduct the negative cycle to neutral, so I will need to connect the right mosfet to the right side of the parallel capacitor.

I will proceed without connecting the AC HV coil,
only driving the high current coil ,to see how it works and behaves.
then I can connect the HV coil again, and see if I can phase shift the HC coil relative to the voltage (-90)

I detached the HV push driver,
and attached the pull pull mosfets to the current coil at both sides of the parallel tuning capacitor.
gave it some power, and it works!
loads of amps with just a bit of power from the supply.
I now have 2 different mosfets installed, so I will need to replace one, to make them similar.
didnt expect this to work that well. but it does.

I had the choke connected wrong, to the capacitor side,
so I retested, and its all good.
0.4A 5.0V = 2W of dc input power gives 13A peak to peak

I connected the high voltage coil, and the high current coil to the driver pcb.
and gave both independed power.

the (green) high current was tune to the right phase relationship (-90) to the voltage( yellow)
* voltage is much higher, this was mesured through the isolation mantle

If I give more power to the current coil, the power of the voltage coil automatically reduces.
when I give more power to the voltage coil the power of the current coil automatically reduces.

To me this means, the 2 coils are now feeding into each other. they are combining their fields, to produce real power.

psu:
left channel is the current coil 0.25A 3.5V= 0.88W
right channel is the voltage coil 1.58A 2.0V =3.16W

red is the output coil.
in between the bifilar output coil.
so equal amounts of windings.
thin wire tin plated copper. far from what I need.

so It is made for high currents, low voltage.
will need 12V mist lamps to load it down,
I found 6 12V 20W halogen bulbs that are series parallel connected for 24V 5 A 120W operation, so I will use that as a load, and modify if needed.

The coil seems to be still tuned, but the phase is shifted, so I will need to correct that, which is easy with my board

grounding the output winding doesnt seem to hurt the tuning. just a little change in the high voltages (which I keep low for now)

as load I used a single 12V halogen bulb rated at 20W. it started glowing, so at least some power is generated.
current coil used  4,5V and 1,48A=6.66W
voltage coil used 2.4V and 2.04A =4.901W
total input 11.56W

Since the current coil is out of phase with its current on both ends,
while the voltage stays in phase, due to it being a extra coil,
So I decided to earth ground the load at the side where the current is at the wrong phase.

yellow is the loose capacitive probed HV coil.
the actual voltage is around 3,5 times higher so 1.6*3.5=5.6kV pp

green current is not a perfect sine. the current jumps, I estimate around 8A pp

Orange shows the gate signal of the current coil on the side of the output.
I removed the irfp510 mosfet and replaced it with a c3m0280090d (which is fast and can handle 900V spikes)

I want to increase the voltage of the current coil. so there is a larger change in voltage when the mosfets switch
but the PSU current is very high, it should be much lower. The choke should prevent this.
So the choke should have more inductance.

good info
https://microcontrollerslab.com/tl494-pulse-width-modulation-control-ic/

I use it for both the High voltage coil and the high current coil .

and I am wondering if I should put pin 13 to ground on the current coil.

this makes it operate in single ended mode instead of push pull.

both mosfets are then switched simultaneously.

So the bifilar coil with it center tap to the positive voltage supply, will be grounded to neutral on both ends simultaneously.

this means the capacitor is grounded on both ends to neutral for 50% which is not good?

It will be switched when the cap is empty (max currents) so that is fine.

but... will there be an interaction with the high voltage coil as a capacitor plate?

the displacement current is max. while voltage is zero on the HV coil. so this voltage change at zero reactance point , could induce even higher displacement currents?

amplifying currents to higher max?

hmm

no the parallel capacitor wont work for 50%
that would kill all energy flow

my now common mode choke needs 2x 1.7mH
this would give 166 ohm reactance at 15.6kc/s
the PSU current would then be limited to 0.1A at 16.6V

that is doable. lets make another choke

edit
no... wrong. and I dont know what I did wrong. but, my CMC is now 2x2,9mH at 15,6kc this is 284 ohms.
but my psu gave 1.48A @4.5 V suggesting a load of 3 Ohm

The psu reading might be wrong, from the feedback of the system.
this is so weird. what is going on?

The current coil is capacitive coupled to the HV coil.

But the PSU of the current coil always has a choke coil which should limit the current. but... it doesnt?

I checked the diode to neutral and the choke value, all is good.
but I do see large inductive spikes from the coils, as they are cut to ground by the mosfet, when the current is max.
So lucky that I used 900V mosfets, which could handle these.
So I need snubber circuits first.

Then look for the current source. The choke should limit it. but doesn't
somehow  there must be a currents source leak

the only path for the positive inductive impulse spikes is back to the positive power supply.
So If I install the snubber circuits this high current might disappear?

still not sure

snubber circuit (in purple circle) should connect to V positive

this also makes me realize, the high current coil is always referenced to neutral

but how can an inductive spike exist if no current is supplied to the coil to build up the magnetic field.
the choke should have worked

 :shocked: :roll:

no snubber.

I removed the diode to neutral from the second CMC winding.
This solved the high current problem.
no more resonance also in the coils. also correct
50mA at 10v = around 200 ohm, which seems to be right.

So what is happening?
the diode measures correctly.
it must be the coil connection, reversing the current flow?
lets reverse the second CMC choke coil , and see what happens

FIXED
I reverse connected the choke winding connected to the diode and neutral, and this fixed it.

the copper connecting bridge, was diagonal, now it is straight

still a bit to much current, so I will need a slightly bigger inductor (more than 2,9mH).
but it will have to do for now.
let's continue

spikes still large, so I do need snubber circuits

installed 2 snubbers, between drain and V+ of the current coil mosfets.
spikes are away.
but now...
the current is up again, and the coil is resonant, so I did something wrong???

maybe the duty cycle has not got enough dead time. it would allow dc current, for which the reactance is zero, so only resistance remains .
so I should check the signals.

this is one of the problems with the tl494.
you can tune everything. but they are all independent.
So if I change the frequency , I change the period , the time length. but
the duty cycle does not relate.

this is because
with the high voltage coil driver it does.
but with the current driver, a pulse signal is used. and this signal is created by the hc123 IC
to get it phase locked with the HV coil.

So I might have set the dead time to low.

Haaaaaa :hillbilly:

I have one positive (choke) but 2 paths to negative (coils and mosfets)
So If I have a push pull with 48% duty cycle
that means 96% is seen by the choke.

So basically the choke has a DC signal so the reactance drops to near zero and only  the wire resistance is liming the magnetic current. that is why the psu current is so high

so. I need a 25% duty cycle for the bifilar current coil.

this makes it switch at zero reactance giving amplification of magnetic current
and it switches at infinite reactance which boosts (or drains?) the voltage of the high voltage coil.

what a hoot to realise this.
it only took me 50 years :cheerleader:

25% duty cycle dialed in.
now why is one signal 16V while the other is 23V?

now the current is way less, although still a bit to high.
but the magnetic current also it acts weird.
so maybe 25% duty is still not the best choice. for this kind of driving.
maybe I really need a bigger choke? yes I do!

psu 10.0V 2,8A

at lower setting the current sine is visible, as a true sine wave, but clearly at higher levels, it is chopped by the 25%duty

basically I only need a switching voltage without current.
the only current I need, Is to counter the losses due to switching, wire resistance, capacitor ESR.

but how does that work?
inductive reactance is calculted for a AC signal with 50% duty cycle.
But I will use more like 98% duty or not?
or should I reduce current flow simply by adding more windings to the choke?

Should I maybe give each winding its own choke?

the current now acts up but, there should be no current.
as said at lower levels, the current looks fine.

is this just a probe glitch?

this is at 6.0V and 0.13A a lower setting with less current distortions
still not a good proper sinewave. but I feel this will all be solved by a bigger choke,
limiting the currents, and only providing the voltage switching.

then all is fine. haha, right. well see what comes up then.

the voltage differences of the square waves do not matter. the gates can handle it so, lets ignore it

it turned out, my orange probe wasnt calibrated.
now voltages are equal. perfect!

looking at those 25% duty signals,
it makes less sense to drive the current coil like that.
or maybe I should step away, look at the bigger picture. I could still work

one of the 2 windings only gets a 25% on time, 75% off.
While the other is 50% phase shifted, and then again gets 25% on and 75% off.....

I am more and more convinced I should use 2 chokes, one for each winding

Ok I will solve this differently
keep the current coil bifilar
with a parallel cap over.

but onw side of the cap is a mosfet
and the other side is the choke. to the dc positive.
lets see if that works or not.

if not then I  need a much bigger choke