Kapanadze Cousin - DALLY FREE ENERGY

[itsu]

hi itsu,

May i suggest the igbt's which was listed for you in the circuit diagram by Akula as also found before i even started editing the circuit diagram to fix the bug.
These are the ones-
But i would recommend to select the one eg: FGH50N6S2D..........

Ok,  i have ordered some FGH50N6S2D, lets see if that will work better.

Regards   Itsu

[itsu]

I was thinking in terms of loading the 28 turns with a FWBR and 30W of lamps as per circuit schematic. I assume that you have nice short gate leads.

Ok,  i could try that (the original Ruslan diagram does not has a FWBR), i will use fast UF4007's, the 550Vpp would light up the 30W nicely i guess.

The gate leads are the 11 Ohm (in my case 4.7 Ohm) resistors with some short wire length, but if it will not improve by any
snubber methode i use, i will have to change the circuit as to have the MOSFET's separated (now on one PCB / heatsink) each on its own PCB/heatsink and the MOSFET drivers also on these PCB's (drivers now still on the breadboard with the TL494).

Regards Itsu

[magpwr]

Ok,  i have ordered some FGH50N6S2D, lets see if that will work better.

Regards   Itsu

hi itsu,

I had purchased this igbt as well for a another project but did not open tube containing 10 pieces yet.

I have attached the unofficial version of the nanopulser verion 7.0 .Because it is now a considered a prototype using only one I/C:74HC132 and it produce a very clean 20ns pulse(Professional version with no ripple) using 120pf at C3 instead of 470pf.

The nanopulse is clean to the point with only one probe connected to the scope to show nano-pulse and my scope still able to properly detect the primary signal at 27khz.
(Less gates involved so less noise)
But i find that the TC4422 or TC4421 is too slow to be switch at 20ns on so my guess it's somewhere around 50ns...300ns for the actual nanosecond pulse to be used on the actual device.

The IGBT driver needs to be the inverted version TC4421 or TC4451(faster).At the moment i do not have this inverted driver yet.

5kohms variable was connected at the non variable ends so it's 2.5kohms for actual.

------------------------------------------
This is my current findings on Akula/Ruslan device.

I am strongly suspecting the tesla coil is likely powered off from the 12volts or 24supply rather than high voltage,why because taking a look at the old KTxxx transistor rated 150volts and new transistor spec 2SC5200 (250volts).
And then take a look at all the tesla coil circuit found in the internet which operated from mains either 110volts or 220volts.The igbt used is either 600volts or 1200volts at least.
So it's unlikely this transistor is connected to around charged 350volts capacitor to work with tesla coil. (The circuit is intentionally made wrong except for some part of nanosecond pulser section which i have already extracted and modified)

The high voltage capacitor used which is around 350volts....450volts was being charged by the yoke core and rectified by ultra fast high voltage diode.(Please do not think of using bridge rectifier as these are meant for low frequency 50/60hz only)

Now this charged capacitor play a very important role in this device.
Upon researching all the "HV nanosecond or nanopulser technology" in the internet which i can get  my hands on.
It seems one of the main component is the "high voltage capacitor" which was charged up then discharged in nanosecond around 20ns <300ns.The resulting pulse would be in kv.

I ask myself where the pulse is likely to be discharged.The only thing keeps surfacing in my head is coil with high Q (Low inductance which results in low ripples produced).
In Dally version i recall someone mentioned it was cable which was shorted at the end and wind around the device.This is good example of a high Q coil .
If you take a L/C meter and do this experiment like what Dally has done short at the end of cable tv cable and create a winding over pvc pipe just by itself.
The uH reading would be very low to qualify for a high Q coil.

Another thing i obverse there is one unboxed video of Ruslan where i can spot additional outer coil over the multilayer coil.This is a possible area to inject nanosecond pulse which is in sync with PWM generator around 27khz.But can be around 18khz ...30khz.
Air coil eg:B&W air coil is a high Q,Simlar to the outer large coil as seen in the device.

[NickZ]

 
   I'm bit excited, as I got my new 12, 7ah battery fully charged now.
The battery would not charge up when using a 12v, 2 amp wall adapter, but it does charge when I used my car battery charger on it.
NOW, the 100w bulbs are brighter, and the Mazilli/yoke core cranks...
  So, now I can continue with connecting the air coils, the Kacher crt, and further testing and tuning.

  My idea was to compare my yoke's output to what your device and yoke's output can do by itself, while lighting some of the same type voltage/wattage bulbs. But, as some of you are using the 220 volt bulbs, and I use the 110v ones, that may not be a proper comparison.
  I still have a heat issue with the fets (one of the two IRFP260N), which I'm hoping to resolve to some degree by the addition of the air coils, as they are would cw, ccw, and should be self cancelling.
We'll see...

[Hoppy]

@ Mag,

I'm not sure that you have the yoke output connections to the Kacher and Grenade coils correct. I think the original circuit looks more viable as the main lamp load is driven in series with the grenade coil. Your altered configuration has the lamp load relying on air cored induced power from the 3-turn winding on the yoke and Kacher. It also makes more sense to me that the kacher primary is powered from the main supply rail and pulsed by the nano-pulser.

What is the NTC5 doing??