You wrote the pll works by locking in...on what??
Any signal that passes through the current sense transformer.
Using this current sense transformer as a mixer has a huge advantage BTW. You can connect the feedback to it and turn things on. What will happen is the PLL will immediately lock at its base frequency of 1kHz thereabout. With the feedback disconnected, the VCO will sit at the base frequency out of lock, but steady at 1kHz. Now what is cool is you can use the feedback and let the PLL lock, then send another signal through the current sense transformer from your signal generator and walk the frequency all around. If you disconnect the signal generator, the PLL will hold that lock at whatever frequency you were at when you disconnected the signal generator. So what we have is a very slick-n-easy way to do frequency scanning while under PLL control. If something in the VIC is resonating, the PLL will drift right to that resonant condition and stay there.
I haven't done this testing with the "real" VIC yet, but I have done it on my mockup VIC and it does just what I'm saying. Another quite interesting thing you can do when your core is gapped is place the pickup coil anywhere near that gap on the outside and the PLL will lock. If you close the gap which lowers the leakage flux, the PLL will come out of lock (no signal). With my big cores pulsing at only two volts and a 2mm gap, I can be an inch out and away from the gap and the PLL with still sense enough signal to lock. It gives you an idea just how much flux is escaping from the core when you have an air gap.
So if this occurs you cant test the pll. You want 50% du in and out...
The PLL will lock to any signal input regardless of duty cycle. It's only looking for the zero cross of the current; that becomes the phase alignment the PLL attempts to sync with.
The output is no less than 50% duty cycle. There is a pot I added that will stretch the range to about 60kHz before 100% duty cycle is reached, depending upon the make of the TIP120 used. You can use the pot to set the cutoff however you like within reason.
So one vic card cant be correctly calibrated for different vics.
True, but it can be referenced to a starting point. Gating, voltage level, center frequency and duty cycle cutoff can be set to known values. I can tell you right now, if I populated one of these boards and sent it to you without this pre-tuning, you'd be pulling your hair out for a while trying to guess what all those pots do and which way to turn them. I've been through it few times myself. You have to know what kind of signals to inject, where to place your probes and what to look for. My thought is to set minimal gating, two volt output, center frequency around 15 kHz with the duty cycle cutoff at 30kHz. From there you could tweak things however they need to be for the VIC you connect it to.
Also all the ic chips must equal type of the same series, change the part and you have different operation of the circuit.
Margin of error. I can tell you right now, I can produce two boards that are so very close in behavior and performance that I could not tell which one was which without marking them. If we need precision greater than this to make the VIC function, we'll be swinging at bats in a dark room for a long time.
Also what still is unanswered, on what frequency the vic had resonance and also the doubling of frequency phenomenon. Right now with the capacitance, inductance and core I have a resonance frequency of 16kHz.
I don't think we can be sure "resonance" is the necessary condition where the splitting of the water molecules happens. The frequency may be close to resonance in one of the coils, or all of the coils or someplace else. I'm pretty sure there is some frequency where the magic happens. We really have no choice but to let the test apparatus talk to us. We have to sense what it wants, what it needs to do the thing we are after.
The circuit can only pulse 50% du, max. Approx. 10khz, after that dutycycle starts to change.
Fifty percent duty cycle minimum. With the pot I added, you can control how this duty cycle cutoff behaves, which I still think is important.
I only can say, Matt try to test it first with a proper vic with fb coil and wfc and experiment with it before you start to sell that card. I want to see some scope measurements of in/out/phase signals and data first btw.
That's the plan. If I had all of Ronnie's stuff sitting on my bench and working, we would be well ahead of the game. I don't and won't, so we need to work a little slower and pull ourselves up inch by inch until we can see where we need to go next. This board is my first crack at trying to elevate just one inch higher; that is all for the moment. I've studied and played with the circuits Stan shows us in the patents and in his estate photos enough to say that this board I produced should be a step in the right direction. We still have a long way to go.