Quote from warj1990 on September 25th, 2019, 07:14 AM

I haven't read your entire thread...(sorry)...but in the mean time...

Say hello to my little friend: IXFH12N120P :cool:

Rds 1.35 ohm  pretty high...
please dont use my board for auctions

I didn't intend it to look like an advertisment and it  wasn't "my" auction...just a place to find the device I was referencing. 

I really don't have much input on your area of work, just noticed you mentioned the voltage limits - so I posted the IXFH series with higher voltage ability.
(just saying I will keep to myself, lol)

Keep up the progress.

W.

thanks,
I found the SiC mosfets of cree/wolfspeed do  the trick  1200v and even 1700V

since the impulse of L1 is at the resonant speed of L1  (its a half wave, a half period),
and it is created in a resonant condition (L1 is charged up with a magnetic field, then quick the mosfet switch is opened)
the L1 is resonant.
When it is close coupled to L2 the current draw is reduced. but the magnetic field of L2 is amplified by the impulse.

since L1 is resonant, and L2 is resonant, and both are close coupled, it is best, to make L2 resonant at a lower octave of L1's resonant frequency (twice the L1 impulse duration= period of resonant frequency).

L3 is parallel resonant, and thus also a lower octave of the L1 impulse.

I still am not convinced, the L3 should be parallel resonant tuned. Best would be, to not tune it down by adding parallel capacitance at all.
 just let the coil ring by itself ( grounded on the outside rim).
Then L1 impulse is a half period of L3.

but for the L1 impulse to travel this fast (short duration) through L2, it should not exceed L2's resonant frequency.
I do not mean the tuned down L2 series resonant frequency, but, L2's own series resonant frequency (without added capacity), as this limits the maximum speed of the impulse (due to L2 capacity).

Since L2 is cloase coupled to L2, L3 needs to have a lower resonant frequency, else it would be to fast, and demanding a to fast impulse for L2.

Solution is to use smaller diameter wire for L3 (but equal copper mass between L2 and L3).
So a 1,5mm2 L1 and L2,   with a 0,75mm2 L3 (or 4mm2 L1 and L2, with 2,5mm2 L3; or 2,5mm2 L1+L2, with 1,5mm2L3)

I had to repair the diodes of the chair.
I dont slowly turn the power up.
this caused the impulses to overload the dc offset diodes.
they still work, but I sespect the reverse recover isn't fast enough.
it only gives half of the impulse voltage as dc offset. should be almost equal.

I also removed L1 from L2. they were close coupled.
L1 will now rest on the ground (not coupled).

I hotglued L2 to the bottom sit plate of the chair. on top is L3.

the MDF plate is 18mm thick.
L2 and L3 both are clockwise.

Why:
I do not need the power of L2 feedback into L1 to reduce input power.
Instead all energy of L2 is coupled to L3.
L2 has a dc offset and forms a capacitor with L3.
L2 and L3 are the plates of a single capacitor.
L2 is impulsed and discharges the capacitor

L2 is now tuned with 107nF in series (11x 10nF in parallel)

the dc offset cap is only 1 0.68nF (1000v) cap

the large cap in series with L2 (blocking the dc) is 3x0.68nF in parallel giving 2uF (1000v)

L3 is parallel tuned with 7.9nF

Since L2 and L3 are now closer together and L1 isn't close coupled to L2 anymore, I might have to retune for a low enough frequency (42kcps or lower)

and I have 330+100=430pF parallel to L1 to increase impulse duration

the chair now works at 32.67kcps
impulse is - 488V and 840nS duration
power input is 8.7W, 9.5V and 0.92A

Somehow the DC offset still is much lower than the impulse voltage. only around  230V dc offset.

both dc offset diodes are replaced, so I dont understand where the voltage is going. Maybe use a smaller cap, or larger, to buffer the dc voltage into?

While I am typing this, I am sitting on L3 as it is tuned and turned on. I can clearly feel the energy in my body.
I'll measure L3, so I know what the voltage is (must be high with less then 8nF parallel to L3)
If I remember correctly, I used series parallel caps on L3, so they can handle the voltages, and do not over heat.

L3 is not a perfect sine.
only 740Vpp

I see, I didn't use a blocking diode for the positive of the power supply.

when I detune, I can hear the power supply ring... thats bad.

back to the bench

made a simple diode for the positive power supply

now the buck boost module for the gate driver could still get the impulses as it sits after the diode. I don't worry about it.
but would be best if it got power directly feomnthe supply, before the diode

tested it and it works again.
Now I can also tune into the higher LMD resonance at 76.88 kcpc.

L3 now is over 2000V peak to peak. and has a perfect sine. (attached)
a bit high in voltage but the chair only runs for 20 minutes per session. and the caps should be able to handle it.
Maybe I'll add a few more nf in parallel with L3 to get the voltage down a bit.

impulse is still 480V max, but now at 750nS duration. slightly faster.

Im sitting on L3 again, and now it heats up my butt as it always does with LMD (Longitudinal magneto dielectric) resonance

as the LMD frequency is higher, I need more power.

32W  19.8V  1.66A

this is also higher because there is no feedback to L1 any more

On my coil capacitor thread, I will couple L3 back to L1. to do this, L2 will be flipped over, as with LMD resonance the phase between L2 and L3 flips 180 degrees.
then L2 is coupled to L4 which produces power

As someone corrected out in a YouTube comment a while back, in the Buffalo lecture of N Tesla he explained how to use electro therapy, by making use of both ends of the secondary coil.

I have done this long ago, but until now I mostly use the secondary coil to sit upon.

I still want to experiment more into these health effects using a series resonant impulsed primary coil, and a (tuned?) secondary coil.
With both ends used, as isolated plates on the body.

I remember when I used the CD coils I noticed the effects, and mostly in the back of my neck. A friend comfirmed the same sensation in the back of her neck at the time.

I still wonder how the health effects differ between the TEM and LMD in this setup.

With the chair I use LMD. but it is very subtle.

Also when the secondary is not grounded, both ends also show the impulse on the sine waves. both ends then have a 180 degrees out of phase signal.

the coil can be left on itself with no added capacity, but then the frequency will be very high, and only the body will act as a parallel capacitor.

I suspect this double plate will be much more intense as the body is capacitive coupled to both polarities, instead of only one.

careful experiments will reveal more. and then... there will probably be frequencies that work better then other frequencies.

One thing is clear. I will make large diameter (a feet) solenoid coils, not pancakes.
still bifilar to increase the resonant energy and lower the frequency.

As plates I will use copper foil with varnish or latex as I still have that, if it sticks...