Timeshell,
I am more interested in the potential for building voltage, and how he determined his frequency. I agree that keeping SOME voltage going in is good, this whole GATING (stopping the pulses for some period) SHOULD be secondary to getting HHO production with low current consumption. (What is the max voltage across his water capacitor?)
The circuit below shows that if you TUNE the frequency just right, 12V eventually becomes 1+ MV... This is clearly IDEAL and UNREAL, but increasing voltage on EVERY pulse by more than it dissipates is the goal, is it not? Play with that circuit and change the frequency a bit. Just a few HZ makes a dramatic difference.
The second challenge with analyzing wave forms is that that they are often OUTPUT wave forms. The wave form I get for a Square wave, when it goes through a 1:10 step up transformer is NOT EVEN CLOSE to what I put in. Even a PULSE shows the eddy lines after the step up. Due to inductance, the wave form on the input side is altered! (Of course, if you are getting it right out of the controller. Great. I would ask to connect it to YOUR capacitor, to see if the tuning is a big piece?)
I say until you have amazing tuning and HHO production, then I would not GATE off the voltage pulses. Once you can produce the resonant wave, then gating has a benefit. But I view gating as a late stage optimization that Stan did. Especially when I see his older equipment.
On the circuit below, if you add a ZENER diode before the 330mH inductor. You wont get far on 12V. Change it to 24V, and you are back in the game. But it takes 12 Seconds (with 7,000 pulses per second),
to get up near 1MV... (but hey, 84,000 pulses of 24V is near 2MV).
Put in a regular Diode, and it does not work anymore. I keep reading about people getting 2xVoltage in the capacitor thinking this is the point of the LC circuit. I can't imagine all of this talk of resonance not sinking in.
You want HUGE voltage across the plates, pulsating with low current. (Pushing the swing at the playground). (If a diode goes in here, it better be Zener type, or it should go INTO the primary coil side. Although,
if you are using a MOSFET on the primary side for gating). Then there is no real reason to diode the circuit there. The MOSFET, when OPEN (not pulsing) will have the same effect as the diode. Of course, if you
have really long pulses you might need the diode, but I am thinking Spark Gap style pulses. Fast and Powerful. Long pulses through a step up transformer where rate of change is important, doesnt seem right.
But the pulsing you do, should be "positive" only, not putting negative voltage into the circuit (it seems to me that would wreak havoc on the performance). I just tested it. Sure enough, it has the opposite effect (pulsing -12V, after charging it up). (Notice that once running, if you set the Voltage to 0. It will keep running forever at the current voltage. Highly Ideal, but close to the goal)?
Also, Meyers has diagrams of this type of pattern. Like 1/3rd - 1/5th amplitude pulses in between full amplitude pulses. (Then again, I think Meyers had a diagram for everything EXCEPT the WORKING solution!) :-)
HTH,
Kirk Out!
PS: The values in the circuit represent my current cell. I am waiting for an OptoIsolator (so I dont fry any more equipment). By next week, I should have some solid test results, and hopefully a video. If not, a ton of questions, LOL.
A Well Tuned LC Circuit: Watch Voltage Crank up! (a la Tesla style)
$ 13 5.0E-6 382.76258214399064 57 5.0 50
v 160 272 160 192 0 5 6984.4 12.0 0.0 0.0 0.5
c 512 192 512 272 0 7.864E-10 -64857.00378094884
l 384 272 512 272 0 0.325 2.901408948082747
l 384 192 512 192 0 0.33 -2.901408948082745
w 160 192 192 192 0
w 160 272 384 272 0
w 192 192 384 192 0
o 1 64 0 35 163840.0 6.4 0 -1
o 0 64 0 35 20.0 6.4 1 -1
