Kapanadze Cousin - DALLY FREE ENERGY

[Hoppy]

  I guess that this might is possible, as my SG is only a 2Mhz model. But as my 3 turn is now running at 25Mhz down from the 28 to 29.6 Mhz that it was at before I removed the two additional tuning caps that were on the 0.47uf cap.  The problem now is that I don't have any more of the HV wima caps,  and placing other lower voltage caps along with the 0.47uf cap drops the voltage to their lower voltage rating.
  Any way, thanks for the reply.

Nick,

If you removed caps that were in parallel with the 0.47uF cap, then I would have expected your KHz frequency (not Mhz) to increase, not decrease as you have written.

[itsu]

    Itsu:
  You mentioned that you are not that optimistic about the self running aspect.  But, have you tried it yet, or not?
   Maybe try it with no load at first, to see if it will self sustain itself without any loading from the  bulbs, or not. 

   All three guys using the longer Kacher secondary coils, Geofusion, Jpok, and Ursa, are not mentioning anything about self running.
   And, it's been a while now...

Nick,

no i did not tried it, the setup is not complete or correctly tuned.


Ok, i added (to the ALLEGGA snubber setup) a 10 Ohm / 6W (with a parallel UF4007 diode) in series with a 47nF mica 600V  capacitor across the source drain of each MOSFET.
The ringing is now much less and i think to an acceptable level, see screenshot below.

Yellow is voltage across the wima caps (resonance).
Blue is voltage on the drain of MOSFET 1
purple is voltage on drain of MOSFET 2


Regards itsu

[verpies]

I would have to redesign the secondaries to match this same 220V DC output as the primary would be much more turns (whole circumference, and that severall times).

It is possible to connect e.g. a primary of 240 turns in parallel in 12 groups (20 turns each) and then to connect each group in series, in order to emulate a heavy 12 turn winding that will span the entire circumference of the core.

Narrow winding and odd number of layers indeed cause flux leakage and resulting spikes.
Canceling the undesirable circumferential current is something that most people forget about.  Having only one layer almost guarantees it (except for CW+CCW center-fed windings).

P.S.
The groups can be interlaced.

[Hoppy]

Hoppy,  thanks, good to know.

I was (am) thinking on using that verpies circuit, but the present setup with the 2x 12 turns primary, 3-turn and 28-turn secondarys is somehow greatly adjust to deliver 220V DC at the output caps
which we can use for connecting up a 220V/24V ps, bulbs etc.

I would have to redesign the secondaries to match this same 220V DC output as the primary would be much more turns (whole circumference, and that severall times).
How does that pan out at your "verpies" setup?

Regards Itsu

Itsu,

As you may recall, I dismantled my original setup and re-wound my yoke primary to Verpies 'clamp' design. I have only tested the clamp using a PWM generator to observe the waveform under various loads but I did not monitor and note the secondary output voltage level. IMO, its not necessary to use the thick 2.5mm sq conductors, as clearly this would add too much extra bulk to the primary, which is why I used 2 x 4 strand litzed 1mm ECW, which took up very little extra space around the circumference of the yoke. The output voltage should be much the same as my original, given that the turns ratio is the same.

Regards
Hoppy

[Jeg]

Narrow winding and odd number of layers indeed cause flux leakage and resulting spikes.
Canceling the undesirable circumferential current is something that most people forget about.  Having only one layer almost guarantees it (except for CW+CCW center-fed windings).

Hi Verpies
Can you please say some more on this? Why CW-CCW center fed windings don't result spikes?

Tnks

ps. Does anyone know anything about the type of diodes that Ruslan uses at his output stages?