Kapanadze Cousin - DALLY FREE ENERGY

[Dog-One]

has any one used this unit built it and ironed out the bugs in the hard ware ? Also i don't understand why Nick wants to use a pulse width device instead of this fixed resonance capable device, as pulse width is not capable of achieving resonance because it's variable frequency, due to the varying pulse width. The bugs could easily be removed if the circuit is modified correctly.

Speaking from experience, I initially used a SG3524 which gave me full PWM (0 - 50% duty cycle) control and frequency (8 - 40kHz).  Connected to this chip I used a pair of 14 amp IXYS gate drivers and did the switching with a pair of IRFB3077 MOSFETs.  All seemed fine and dandy.  I could control frequency and power level no problem at all.  What I discovered driving the yoke (toroid) transformer with this circuit is that you really need to run the duty cycle pretty close to 50%.  You don't want to be at 50% with no dead time because the core will saturate and begin to draw mega amps.  If you run the duty cycle at say 20%, the wave form on the output side is not even close to a sine wave.  Rather, it's badly distorted.  The catch is when you run near 50% duty cycle you get a good sine wave on the output, but with the grenade and induction heater coils connected, you almost get a runaway condition on current draw.  It's very tricky to find the right value resonant capacitors to tame this current draw down, but if you get them, you'll see a nice sine wave and current draw will sit at about 5 amps powering a 60 watt incandescent bulb.

Looking back at everything, one mistake I see right away is the MOSFETs.  These IRFB3077s will dump a ton of amps, way more than is needed.  Their on-time resistance is next to nothing.  What happens is the filter caps on the power rails get dumped immediately and cause a harsh spike in the yoke, probably near the point of saturation.  With a scope you will see the ringing on the input side that gets absorbed by the snubber network.  The MOSFETs run cold, but the 10 watt resistors get mighty toasty.  What I think is necessary is to find a pair of MOSFETs that have a little higher on-time resistance, but not so much they begin to heat up.  This article explains it better than I can:
http://sound.westhost.com/articles/hexfet.htm

With the right MOSFETs, I'm pretty sure you can drive the grenade and induction heater coils just right and waste very little power in the form of heat.  Using a SG3524 is a piece of cake and does the job just fine for a fixed frequency push-pull signal generator.  Oleg even uses a SG3525 which appears to be nearly identical--haven't looked closely to spot any major differences for this application.

Attached is the board I built for testing.  It's way overkill for this application, but it accomplished what I set out to do.  I'm still looking at driving the toroid transformer with an audio amp, but after blowing two of them now, I need to find a way to incorporate the snubber circuit because in both case of blown amps, it's not the power draw that is wiping them out.  It's the kickback from the grenade and/or induction heater coils pushing through the toroid transformer.  These spikes are simply too much for an audio amp to deal with, since they are made to drive speakers that would never spike back like this Ruslan device does.

I'm convinced there is a solution to be found.  We are all dancing around it.  This device needs a custom driver or at the very least some custom components not found on commercial products.  We get over this hurdle and we can all begin to focus on the real meat that makes this device do what Ruslan claims/demonstrates it does.

[verpies]

In fact, I've demonstrated a system using a PLL chip to generate the correct resonant frequency, and lock to it...

I remember it in a SSTC of yours.  Could you remind me what was the source of the feedback signal in that implementation?

[AlienGrey]

Verpies,

tx = 1.5 mhz   rx is 15khz ,  or tx - 2mhz rx - 20 khz  but do they resonate at the magic frequency multiple ? of the earth or it does not work, you can build drivers till the cows come home but unless your device is stable your pissing in the wind the device i posted is not stable as it stands ! If you inject 1.5mhz into the LM393 and use a scope you will see what happens, also the lower pulse width section is again un-stable and drifts with temperature variations, and the pulsed TX self osc is also unstable and micro phonic a problem radio had in the old Tesla days.

[Hoppy]

Verpies,
Verpies,

tx = 1.5 mhz   rx is 15khz ,  or tx - 2mhz rx - 20 khz  but do they resonate at the magic frequency multiple ? of the earth or it does not work, you can build drivers till the cows come home but unless your device is stable your pissing in the wind the device i posted is not stable as it stands ! If you inject 1.5mhz into the 393 and use a scope you will see what happens, also the lover PW variant circuit lover down also is un stable and drifts with temperature variations, and the pulsed TX self osc is also unstable and micro phonic.

Do you mean as stable as Kapanadze's box devices? 

[AlienGrey]

Do you mean as stable as Kapanadze's box devices? 

Hoppy

I can see your a PW lover, but apart from that, when one delusion falls apart, why jump into the next, why not find out what's wrong and fix it. Is all I'm suggesting.