Kapanadze Cousin - DALLY FREE ENERGY

[verpies]

I now have a nice sharp 200V pulse of about 70ns which can be varied by a 10Kohm trimpot till 500ns.
Repetition rate of the pulse can be varied between 700Hz till 350Khz (not shown in the video).

This range of frequencies and pulse widths should be sufficient to tune into the DSR effect in wide selection of suitable DSR Diodes.

The gate spikes are due to the drain voltage spikes coming through the drain-gate capacitance of the MOSFET.
Those pulses are dangerous because they can turn-on the MOSFET or keep it conducting longer than you want to (a.k.a. Miller turn-on). 
Try to supress these pesky pulses. A half-solution is to connect a pull-down resistor from gate to ground in the range between 100Ω - 4Ω (experimentally chosen to suppress the gate spikes but not degrade your drain fall-time {offâ†'on transition} too much).
A full-solution to this pesky drain-gate capacitance is a better MOSFET driver, designed to combat the Miller turn-on effect, by pulling the gate hard to ground when the MOSFET is supposed to be off (e.g. the UCC27512 or UCC27511).

Next step will be to hookup the small toroid and the DSRD diodes and hunt for the 1ns/1KV/1Khz pulse

Really protect your scope at this point! The HV spikes can damage its input amplifiers.

The first thing to do is to connect the primary winding of T2 and a series 0.1Ω current sensing resistor (CSR) instead of that 100Ω load resistor, you have now, and scope the current in that CSR (without any load on the secondary) to observe the saturation point in the ferrite core, as shown here.  You must know the saturation point unless you want to proceed blindly.
Begin the saturation measurement with very short pulse widths and 20V HV power supply on the drain and increase them carefully while watching the scope for any signs of saturation, excessive spikes and heating.
If V/R limit manifests itself before saturation, then add more turns.

P.S.
Your "spike remover", made of the 1kΩ & 1nF parts (pt.4a), may be incompatible with the DSRD driver using the saturable transformer (pt.4d), so prepare to remove it when you get to the DSR Diodes, later.

[verpies]

No worries; so I guess the back spike is the field behind collapsing

You have worse then spikes. You have reversal of the collector current due to LC oscillations!.

Of course those aren't right... I couldn't have luck and pick the part that was high enough current  and not be inductive    

Just Murphy's law at work.
However those current sensing resistors do not need to be rated at more than 4 Watts.
A 0.1Ω resistor at 12V will dissipate 14W of heat, however the current in the T2's primary does not have this value 100% of the time because of the duty cycle and current ramping in the inductor, so at 50% Du.C. it will dissipate only 25% of 14W which amounts to 3.6W.

Just to confirm, was looking at dally pics,... and the power wire would be one of the white wires coming off the toroid, but disappearing beneath the tl494 board;  the other two from two different posts go to a green screw down connector on the edge of that board... those are all thinner wires than the wires on the atx power supply... so they don't have to be so thick?

Maybe they don't but for testing they should be unless you want to be blind to what's going on in your circuit

Most of it is rewired, but What I do have is 8 inches of 32 gauge wire which should be .1 ohm... but it's far from high current...

The current in the orange wires can be many times higher than the current draw from your power supply, when the transistors are pulsing. Also don't forget about the mutual inductance between long wires with high di/dt.

Compare with Itsu's setup and notice how powerful and clean his waveforms are at T2.
He even thinks of rewinding T1's secondary because he is getting 800V_p-p there... although I think it is much easier to adjust the trimpot P2 in this circuit than rewind a transformer.

[mihai.isteniuc]

Hello,

My TL494 converter it's done. I still don't have a transformer with 2x150V secondary windings, but with the one I used right now I'm able to light very easy at about 3/4 full brightness a 230V/60W light bulb.

Without any load the converter needs 0,5Amps@12V to run. With the light bulb connected as load the current taken from the source goes up at 4.5Amps. The good news: transistors stay cool. I'm holding back for the moment 'til I will find a suitable transformer. I will start working on the nanopulser - the high power part, because I really have to know how much power it needs to run.

Mihai

[verpies]

My TL494 converter it's done. I still don't have a transformer with 2x150V secondary windings, but with the one I used right now I'm able to light very easy at about 3/4 full brightness a 230V/60W light bulb.

Without any load the converter needs 0,5Amps@12V to run. With the light bulb connected as load the current taken from the source goes up at 4.5Amps. The good news: transistors stay cool.

Three questions:
1) What is your switching frequency?
2) What transistors are you using to drive the primaries of T2 ?
3) What are the inductances of your T2's primary windings without anything connected to the secondaries? (if you don't know the inductance, at least let us know how many turns on what core).

[mihai.isteniuc]

As mentioned before the frequency can be varied between 6khz up to 200khz with a maximum of 40/60 duty cycle. The transformer I'm using it's taken from a computer power supply. I cannot provide any tech specs regarding it.
Variations from Dally's unit:
MJE13009 replaced by some 2SC9245 (I'm not at home and I don't remember exactly). The main difference it's the power dissipared by the 2SC....-80W
2N5401 replaced with BC327
TL494 supply line has a 7812 regulator with some caps for better filtering.
The radiator for the main transistors has a cooler (btw the current at no load posted previously 0.5Amps@12V included also the working cooler).
As expected increasing the frequency and the duty cycle light the bulb brighter and brighter. Maximum was obtain around 150khz and40/60 duty cycle. The bulb wasn't at full brightness only because my secondary wasn't deliver 230V because it didn't has enough turns.

Mihai