Kapanadze Cousin - DALLY FREE ENERGY

[Сергей В.]

to Itsu

What's going on with your Nano-Pulser Itsu ?

I think your ferrite rings are ok, transistor ok, driver ok, oscillograph ok. Maybe is time for you to start new blank page

If you can't get proper Nano-Pulse that it mean your 1n5408 diode is not good for that. Try to find old Diffusion Rectifier Diodes (bunch of them) and to make test every single example.

In prinicpe you can get old rectifier diodes from old electronic scratches, old inverters, old accumulator chargers, old TV sets, old AT power supplies or old power electronic pcbs. Put your ferrite core on 90 degrees relative to your ground plane and make a disctance about 2 cm from it. For precise differential measurements you need good and quality wideband Active Differential Probes for your oscillograph.

I have attached these photos from Realstrannik member Strimer. Maybe some details can help you to get NANO-PULSE !! As you can see Nano-Pulse 640V on 56ohms with 48V supply. Work excelent. Oscillograph is Aktakom 2GS/s 100Mhz bandwidth. Primary coil Litz wire 32(litz)х0.05mm (from UTP cable) with tinned ends. Ferrite ring 7x4x2 µ=2000 NM1-36. I think nothing complicated to you!!
You can try other primary-secondary ratio 1:3, 1:4 but not to much secondary inductance.

ps. The goal is to get proper Nano-Pulse not to know every  evrery parameter of your torroid. You don't research but replicate!

Удачи Сергей В.

[TinselKoala]

The variable delay in the transistor turn-on time as the collector voltage is varied, doesn't seem strange to me. You are filling a capacitance of sorts, even in the bipolar transistor, and that takes time, less time with more voltage.

The drop in current just after the turn-on, at the lower collector voltages, could be the power supply not keeping up and sagging, maybe? When the supply is turned up in Itsu's video the pulse squares up nicely. It would be nice to see what happens as the pulse width is decreased.

[TinselKoala]

@Sergey: Nice, thanks! That's got to be the first time I've ever seen a wire soldered directly to the drain tab of a mosfet, though!

[verpies]

You don't research but replicate

Are you advocating blind experimentation without knowing what's going on in the circuit?

With so many variables to coordinate the probability of success of such approach is very low.
Just look at Itsu's previous torroid. The core took so long to saturate that no diode could have been conducting forward for so long and exhibiting the DSR effect.
If Itsu did not know its core saturation time, he would have been experimenting till the cows come home. The same goes for the current gain in his BJT and his insufficent transistor saturation.

Itsu is just systematic. He might be slow by your standards but he's a good experimenter and he is very close to driving driving his DSR diode for optimum forward time, e.g. opening his transistor when the current in the core is saturated and the current in the secondary is crossing zero (or the voltage in the secondary capacitor is at its maximum).

If you are impatient and feel that you'd proceed at a quicker pace than Itsu, you are welcome to make your own experiments faster.

Itsu might not have the best DSR diode (the кд diodes are hard to acquire where he lives) but with the experience he gained during this research, that will soon become very evident.

[verpies]

The variable delay in the transistor turn-on time as the collector voltage is varied, doesn't seem strange to me. You are filling a capacitance of sorts, even in the bipolar transistor, and that takes time, less time with more voltage.

I was thinking about the internal transistor capacitances too, but why would the base-emitter capacitance increase so much when the collector voltage is low?

Also, the base-collector capacitance would pull the base down and delay the transistor turn ON, through the Miller effect, when the dv/dt on the collector is high but it seems that the dv/dt increases with the collector voltage, not the other way around.

Maybe the +V2 is sagging when it is low, but shouldn't that be visible on the yellow Ch2 trace which is referenced to +V2 when capacitively coupled to pin 4/8 ?

I agree with Сергей that researching the cause of these delays when the +V2 voltage is low, is not essential to generating the nanosecond pulses which will be generated at higher +V2 voltages anyway.