Kapanadze Cousin - DALLY FREE ENERGY

[Black_Bird]

@itsu

Congratulations. Isn't 800V enough?

[verpies]

despite some problems with noise and ringing in my nano-pulser circuit, even after re-arranging some parts, i managed to squeeze a 800V / 19nS pulse out of the 1n5408 dsr diode.

The MOSFET did get hot, but it was operating without a heatsink.

The good news is that your 1N5408 diode is showing some DSR effect.
The ringing is bad news - this is what makes your MOSFET hot and wastes energy and can produce excessive gate (>20V) and drain voltage (>250V) that will damage the MOSFET.

I don't have the result of measurement from Fig.20 but it is possible for the filters composed of L1+C1 as well as L2+C2, to ring.
If you take a look at the attached schematic it becomes apparent that L2+C2 form a series LC circuit, that will oscillate at its natural resonant frequency.

Indeed, this filter strongly attenuates:
1) Any voltage variations appearing at point +V_2A acting as input and trying to get to point +V_2B acting as output of the filter.
2) Any voltage variations appearing at point +V_2B acting as input and trying to get to point +V_2A acting as output of the filter.

However, this filter allows the point +V_2B to move up and down in reference to the ground, at the frequency of LC resonance, which is  f = 1/( 6.28*(LC)^0.5 ).  This calculates to 7.2kHz if L1= 220μH and C2=2.2μF.

So, are the values of L1 and C1 really 220μH and 2.2μF, respectively?
Note that the inductance of L1 can decrease to nH if its core becomes saturated (e.g. with too much current through its winding).

If "yes" then do you observe any ringing at 7.2kHz?
...or do you see the beat frequency of two LC tanks L1+C1 and L2+C2 (both oscillating close to 7.2kHz) ?

P.S.
The drain ringing frequency seems to be ≈90MHz after the rising edge and ≈30MHz after the falling edge of the signal on the gate of the MOSFET. This frequency is very far away from the resonant frequency of the power supply filter L2+C2 which would ring around 7.2kHz with 220μH and 2.2μF.  For comparison: 1nF and 3nH would ring at 92MHz.
Does the ringing frequency change at all when you short L2 or L1 or alter C2 or C1 ?

[itsu]

The good news is that your 1N5408 diode is showing some DSR effect.

Right,  so wondering what the ordered Russian diodes will do.

So, are the values of L1 and C1 really 220μH and 2.2μF, respectively?
If "yes" then do you observe any ringing at 7.2kHz?
...or do you see the beat frequency of two LC tanks L1+C1 and L2+C2 (both oscillating close to 7.2kHz) ?

Yes, those are the value's as mentioned in the original Dally diagram.
I will try to find out the ringing frequency.

Does the ringing frequency change at all when you short L2 or L1 or alter C2 or C1 ?

I did short out both L1 and L2 one by one with a short (3 cm) wire. but nothing changed (was measuring gate and drain signals)

Regards Itsu

[itsu]

@itsu

Congratulations. Isn't 800V enough?

Well,  probably, but i do not have the feeling this is the right pulse, as it is still fairly wide (20ns).
I expect to see a pulse in the 1-5ns range.

Regards Itsu

[Black_Bird]

@itsu
It can be that your high voltage probe is widening the pulse. Anyway dv/dt seems to be the important parameter. What is the rise time of your pulse? (not the width)
In my experiments I had a bigger voltage by adding a second 1N5408 in parallel. Also, your output capacitor is 9nF, per your video. Try to reduce it.