Dave Lawton WFC circuit, improved

Cycle

Dave Lawton WFC circuit, improved
«  »Last edited by Cycle
Here's Dave Lawton's WFC circuit, with some improvements.

I segregated the power supply to the WFC circuit and the 555 timer circuit, so a voltage drawdown or spike in the WFC part of the circuitry won't affect the 555 timer part of the circuit as much.

The inductor/diode pair allows a 'soft-start' voltage climb (it takes about 7.5 ms to get to operating voltage), then when voltage reaches its operating range, the diode kicks in to keep the inductor 'spinning' without spiking downstream voltage. Because the inductor is always 'spinning' (ie: the downstream circuit is drawing current), a short-term voltage drawdown upstream won't affect voltage supply to the downstream 555 timer circuitry as much, either. The 100 uF capacitor downstream of the inductor/diode pair further acts to smooth voltage fluctuations.

The back-spikes when the MOSFET opens are redirected back to upstream of the bifilar (modeled as a transformer with the WFC connected serially across the primary and secondary). You'll notice that as WFC voltage builds up, the current flow through the large capacitor is more than the total circuit consumes. Once voltage gets high enough, this part no longer contributes to WFC voltage build-up, but it helps to build that WFC voltage more quickly.

I also replaced the resistor on the second 555 timer's output terminal with a diode... you don't want the MOSFET to go into linear mode, you want it to alternate between 'saturation' and 'off'. This gives the biggest voltage spikes. The diode also protects the 555 timer's output terminal from any voltage backspikes should that MOSFET short the drain to the gate.

As for the bifilar itself, I'm experimenting... I've currently got a 1:5 primary:secondary winding ratio, which allows for very high voltage spikes when the MOSFET opens, after WFC voltage has built up sufficiently. This also helps to keep that diode redirecting back-spikes working longer to build WFC voltage, thereby reducing total circuit current draw.

The WFC is modeled as a large capacitor with a spark gap across it, as that's the closest I can come to how an actual WFC would work.

The code:
Code: [Select]
$ 13 1.0000000000000001e-7 4.252108200006279 35 5 82
165 464 240 528 240 2 0
165 768 240 784 240 2 0
v 1056 432 1056 352 0 0 40 12 0 0 0.5
370 1056 288 1056 240 1 0
c 928 320 1008 320 0 0.083 0.001
187 928 336 1008 336 0 1000 10000000000 2500 0.025
w 992 224 1008 224 0
f 976 384 1024 384 0 10 0.02
w 1024 400 1024 432 0
w 1024 208 992 208 0
r 896 384 896 432 0 50000
w 896 352 896 384 0
w 928 432 1024 432 0
c 832 400 832 432 0 1e-8 0.001
w 832 432 896 432 0
w 768 352 768 368 0
w 736 352 672 352 0
w 672 352 656 352 0
d 736 304 736 352 1 0.805904783
d 656 352 656 304 1 0.805904783
174 656 304 720 304 0 47000 0.0347 Mark / Space
w 688 288 688 272 0
w 688 272 736 272 0
w 736 272 736 304 0
174 688 240 752 240 0 10000 0.24260000000000004 Frequency
w 688 240 688 272 0
w 720 224 720 208 0
w 720 208 752 208 0
w 752 208 752 240 0
w 752 208 752 176 0
w 752 176 912 176 0
w 912 176 912 304 0
w 912 304 896 304 0
S 672 432 672 400 0 1 false 0 3
w 672 432 832 432 0
c 736 352 736 400 0 1e-8 0.001
w 656 400 608 400 0
w 656 352 608 352 0
c 672 352 672 400 0 1.0000000000000001e-7 0.001
c 608 352 608 400 0 2.2e-7 0.001
w 736 400 688 400 0
w 832 208 832 192 0
w 832 192 672 192 0
w 672 144 672 192 0
w 528 432 672 432 0
w 768 272 736 272 0
w 896 272 896 160 0
w 896 160 624 160 0
w 624 160 624 288 0
w 592 272 592 208 0
w 592 208 528 208 0
S 624 288 592 288 0 1 false 1 2
w 416 272 368 272 0
w 416 400 384 400 0
c 416 352 416 400 0 5e-7 0.001
c 368 352 368 400 0 0.0000049999999999999996 0.001
w 352 400 304 400 0
c 304 352 304 400 0 0.000009999999999999999 0.001
w 368 432 528 432 0
S 368 432 368 400 0 0 false 0 3
w 368 208 400 208 0
w 368 224 368 208 0
w 336 240 336 272 0
174 336 240 384 240 0 10000 0.9257000000000001 Frequency
w 368 272 368 304 0
w 336 272 368 272 0
w 336 288 336 272 0
174 304 304 352 304 0 47000 0.0545 Mark / Space
d 304 352 304 304 1 0.805904783
d 448 272 448 336 1 0.805904783
w 368 352 304 352 0
w 416 352 368 352 0
w 400 208 400 240 0
w 400 208 400 192 0
w 400 192 640 192 0
w 640 192 640 304 0
w 640 304 592 304 0
w 464 272 448 272 0
w 464 336 464 368 0
c 528 400 528 432 0 1e-8 0.001
T 928 224 992 288 0 0.25 5 0 0 0.999
w 1008 224 1008 320 0
c 672 112 624 112 0 0.00009999999999999999 0.001
g 624 112 624 128 0
w 1024 432 1056 432 0
g 1056 432 1072 432 0
w 672 192 672 208 0
w 672 208 592 208 0
w 1024 368 1024 288 0
g 880 144 864 144 0
c 880 144 928 144 0 0.00009999999999999999 0.001
S 448 336 416 336 0 1 false 1 2
w 416 272 448 272 0
w 464 336 448 336 0
w 720 304 736 304 0
w 768 336 768 352 0
w 768 352 736 352 0
l 688 112 752 112 0 0.25 0
w 672 144 672 112 0
d 688 128 752 128 1 0.805904783
w 752 112 752 128 0
d 1056 144 1056 112 1 0.805904783
s 1056 288 1056 352 0 0 false
w 1056 144 1056 240 0
w 1056 112 752 112 0
w 688 128 688 112 0
w 672 112 688 112 0
c 928 192 992 192 0 1e-11 0.001
d 992 208 928 208 1 0.805904783
w 928 208 928 224 0
w 1024 208 1024 288 0
w 992 288 1024 288 0
w 928 320 928 336 0
w 928 288 928 320 0
d 896 432 928 432 1 0.805904783
d 896 304 896 352 1 0.805904783
l 928 144 992 144 0 0.25 0
w 992 144 1056 144 0
w 928 160 928 144 0
d 928 160 992 160 1 0.805904783
w 992 160 992 144 0
w 1008 336 1008 320 0
w 896 384 976 384 0
w 928 192 928 208 0
d 928 160 928 192 1 0.805904783
w 992 192 992 160 0
o 3 64 0 12553 0.0001 0.0001 0 2 3 3
o 4 64 0 12554 0.0001 0.0001 1 2 4 3
o 4 64 0 12553 0.0001 0.0001 2 2 4 3
o 108 64 0 12553 0.0001 0.0001 3 2 108 3
Re: Dave Lawton WFC circuit, improved
« Reply #1,  »Last edited by Cycle
Mheh. The above circuit doesn't work right. Dependent upon inductance, when WFC voltage gets high enough, the bifilar secondary becomes 'starved', current flow drops to nothing, and the WFC voltage slowly drops as leakage current isn't maintained.

Here's a new circuit. You'll note the WFC 'capacitor' is now more accurately modeled as a 6.3 nF capacitor with a spark gap which easily 'blows out'. I plugged the parameters of Meyer's WFC tubes into a calculator and it gave 6.3 nF. Naudin measured his WFC at 5.9 nF, so it's somewhere in that range.

The new circuit pulls current across the bifilar primary (the old one pulled from the bifilar secondary, which is why it didn't work right). The bifilar secondary acts as an analog of Meyer's "electron extraction circuit". So the side of the WFC connected to the primary of the bifilar remains at or near the 12 volts from the battery, and the secondary side goes hugely negative. Of course, that doesn't take into account the voltage spikes when the MOSFETs open, but you get the point.

I stripped out the second 555 timer (the one that adds in a 'pause' in the voltage stepping). You can easily add it back in if you wish, but it slows down the circuit emulator, so I took it out.

The 10 nF capacitors on the Output terminal of the 555 timer act to completely isolate the 555 timer part of the circuit from the WFC part of the circuit.

You'll note the second MOSFET on the secondary of the bifilar. The combination of this, the 50 mH inductor and the 10 nF capacitor act as a 'voltage pump' helping to pull current through the secondary of the bifilar. When the 555 timer fires, it pushes current into the 10 nF capacitor, which pushes it out the other side into the 50 mH inductor. This builds up the voltage to above the trigger voltage of the MOSFET, which dumps the voltage to the low voltage rail. The MOSFET opens (the differential voltage across the 10 nF cap is essentially zero at this point) when the 555 timer turns off its Output terminal, allowing the voltage on the 555 timer side of the 10 nF cap to drain off. This causes a negative voltage on the WFC side of the 10 nF cap, which pulls current from the secondary of the bifilar, which is trapped by the diode upstream of the 50 mH inductor, ready for the next cycle of the second MOSFET.

If you hover your cursor over the bifilar, you'll note the voltage spikes reach upwards of 16 kV on each side, so this thing should go up beyond the 20 kV breakdown voltage of the spark gap pretty easily (I've only run it up to ~3 kV so far, but it continues pulling just as hard at ~3 kV as it did at 0 volts... I'll let it run until the spark gap triggers, just to make sure it works as advertised). If you slow down the simulation enough, you'll see that there are actually two voltage pulses at the WFC 'cap' for each cycle of the MOSFETs... one from the bifilar primary, one from the secondary. Usually they're so close together they look like one pulse.

Strangely, at some points as the WFC 'cap' builds voltage, if you slow down the simulation, you can see multiple voltage spikes at the WFC... I think it's because the second MOSFET is triggered by the voltage coming off the bifilar secondary, and at lower WFC voltages it sometimes is triggered more than once per cycle of the first MOSFET. The second MOSFET can be triggered by the 555 timer Output terminal or if the voltage coming off the bifilar secondary is high enough.

The 10 uF capacitor across the low voltage rail and the entry diode of the bifilar primary is to help pump voltage into the bifilar primary while smoothing low voltage rail fluctuations. When the MOSFETs are closed, voltage is dumped into the low voltage rail, which raises the voltage on the low voltage rail. This increased voltage pumps into the 10 uF cap, which pushes voltage out the other side of the cap and into the primary of the bifilar through the diode. When the MOSFETs open, low voltage rail voltage drops and that 10 uF cap wants to equalize its voltage, but the diode at the entry to the bifilar primary blocks backflow, so voltage instead flows out of the 100 uF cap. This lowers voltage upstream of the bifilar, but it doesn't matter, since the MOSFETs are open and no current is flowing through the bifilar at that moment. Voltage flows in from the battery to top up the 100 uF cap.

Here's the code:
Code: [Select]
$ 13 1.0000000000000001e-7 204.8780465020098 51 5 58
165 768 240 784 240 2 0
v 1216 432 1216 352 0 0 40 12 0 0 0.5
370 1216 288 1216 240 1 0
c 992 320 1056 320 0 6.302e-9 0.001
187 992 336 1056 336 0 1000 10000000000 20000 0.025
f 944 384 992 384 0 10 0.02
r 896 384 896 432 0 10000
w 896 352 896 384 0
c 832 400 832 432 0 1e-8 0.001
w 896 432 992 432 0
w 768 352 768 368 0
w 736 352 672 352 0
w 672 352 656 352 0
d 736 304 736 352 1 0.805904783
d 656 352 656 304 1 0.805904783
174 656 304 720 304 0 47000 0.005 Mark / Space
w 688 288 688 272 0
w 688 272 736 272 0
w 736 272 736 304 0
174 688 240 752 240 0 10000 0.5099 Frequency
w 688 240 688 272 0
w 720 224 720 208 0
w 720 208 752 208 0
w 752 208 752 240 0
w 752 208 752 176 0
w 752 176 912 176 0
w 912 176 912 304 0
w 912 304 896 304 0
S 672 432 672 400 0 1 false 0 3
w 672 432 832 432 0
c 736 352 736 400 0 1e-8 0.001
w 656 400 608 400 0
w 656 352 608 352 0
c 672 352 672 400 0 1.0000000000000001e-7 0.001
c 608 352 608 400 0 2.2e-7 0.001
w 736 400 688 400 0
w 832 208 832 192 0
w 768 272 736 272 0
w 896 272 896 192 0
w 896 192 832 192 0
T 992 256 1056 240 0 4 1 0 0 0.999
c 832 80 800 80 0 0.00009999999999999999 0.001
g 800 80 800 96 0
g 672 432 656 432 0
g 992 112 976 112 0
c 992 112 1024 112 0 0.00009999999999999999 0.001
w 720 304 736 304 0
w 768 336 768 352 0
w 768 352 736 352 0
l 832 80 896 80 0 0.25 0
w 832 128 832 96 0
d 832 96 896 96 1 0.805904783
w 896 80 896 96 0
d 1216 112 1216 80 1 0.805904783
s 1216 288 1216 352 0 0 false
w 1216 112 1216 240 0
w 1216 80 896 80 0
w 832 96 832 80 0
d 1056 320 1056 288 1 0.805904783
w 1024 128 1024 160 0
w 992 320 992 336 0
d 832 432 896 432 1 0.805904783
d 896 304 896 352 1 0.805904783
l 1024 112 1088 112 0 0.25 0
w 1024 128 1024 112 0
d 1024 128 1088 128 1 0.805904783
w 1088 128 1088 112 0
w 1056 336 1056 320 0
d 1216 112 1184 112 1 0.805904783
w 1088 128 1088 160 0
c 1024 160 1088 160 0 1e-11 0.001
c 832 128 896 128 0 1e-11 0.001
w 896 128 896 96 0
w 832 128 832 192 0
w 1184 112 1088 112 0
w 992 432 992 400 0
w 912 384 896 384 0
w 1088 432 992 432 0
d 992 288 992 320 1 0.805904783
w 1024 192 1024 160 0
d 992 192 992 224 1 0.805904783
w 992 192 1024 192 0
w 1056 256 1056 288 0
c 1168 432 1168 400 0 0.000009999999999999999 0.001
w 1168 432 1216 432 0
w 1168 400 1168 192 0
w 992 288 992 256 0
w 992 336 992 368 0
w 1056 320 1088 320 0
c 1088 320 1088 288 0 1e-12 0.001
w 1088 288 1056 288 0
d 1136 256 1136 288 1 0.805904783
w 1136 432 1168 432 0
f 1088 400 1136 400 0 10 0.02
w 1136 416 1136 432 0
w 1088 352 1088 400 0
w 1136 224 1136 256 0
w 1024 192 1168 192 0
c 960 224 960 256 0 1e-12 0.001
c 1056 224 1056 256 0 1e-12 0.001
w 992 224 960 224 0
w 992 256 960 256 0
w 1088 432 1136 432 0
w 1136 352 1136 384 0
w 1088 352 1136 352 0
c 960 352 928 352 0 1e-8 0.001
w 928 352 896 352 0
l 1136 288 1136 352 0 0.05 0
w 1136 224 1056 224 0
c 912 384 944 384 0 1e-8 0.001
w 960 352 1088 352 0
o 2 64 0 12553 0.0001 0.0001 0 2 2 3
o 3 64 0 12554 0.0001 0.0001 1 2 3 3
o 93 64 0 12546 0.0001 0.0001 2 2 93 3
o 93 64 0 12545 0.0001 0.0001 3 2 93 3
Re: Dave Lawton WFC circuit, improved
« Reply #2,  »Last edited by Cycle
Ok, I went back to the start and redid the circuit. I've got it working better. It builds voltage pretty quickly and consistently now. I based it roughly upon the schematics of Meyer, with the WFC parameters a mixture of Meyer and Naudin. I'm still testing it a bit, then I'll post the circuit.


Alberto

Re: Dave Lawton WFC circuit, improved
« Reply #3,  »
Nice topic,
But please, go for the real stuff...
Replace that capacitor/spark arrestor
for one tubular array in water. :cool:

Cycle

Re: Dave Lawton WFC circuit, improved
« Reply #4,  »Last edited by Cycle
Ok, it's working. I set the spark gap to 250,000 volts just to see how high the circuit could go before it stalled. So far, it's up to 40,000 volts in ~26 msec (far above what we need) and still pulling hard... and that's with the leak-down of the spark gap, which simulates the water's natural tendency to leak current.

The trick is to get your inductors 'spinning' before your 555 timer kicks in. So I use a p-MOSFET which holds the n-MOSFET closed as voltage builds in the 555 timer circuit, then the p-MOSFET opens and lets the 555 timer control the n-MOSFET.

Remember that in another post, I described parametric pumping, analogizing it to a kid who stands and sits in a swing to exponentially add energy to the swing, but it requires getting the swing swinging before that works. 'Spinning up' your inductors is analogous to getting the swing swinging. Before, the circuit wouldn't let any current through the inductors until the 555 timer circuit was up to operating voltage, and it just didn't work well.

Here's the circuit, and here's the code for it:

Code: [Select]
$ 13 1.0000000000000001e-7 0.8372897488127266 38 5 58
165 768 240 784 240 2 0
v 1344 432 1344 368 0 0 40 12 0 0 0.5
370 1344 304 1344 256 1 0
c 1200 384 1072 384 0 6.302e-9 0.001
187 1200 400 1072 400 0 78.54 2940000 25000 0.5
r 896 368 896 432 0 10000
w 896 336 896 368 0
c 832 400 832 432 0 1e-8 0.001
w 832 432 896 432 0
w 768 352 768 368 0
w 736 352 672 352 0
w 672 352 656 352 0
d 736 304 736 352 1 0.805904783
d 656 352 656 304 1 0.805904783
174 656 304 704 304 0 10000 0.0347 Mark / Space
w 688 272 736 272 0
w 736 272 736 304 0
174 752 208 704 208 0 10000 0.5396000000000001 Frequency
w 688 208 688 272 0
w 720 240 752 240 0
w 752 208 752 240 0
w 752 208 752 176 0
w 752 176 912 176 0
w 912 176 912 304 0
w 912 304 896 304 0
S 672 432 672 400 0 2 false 0 3
w 672 432 832 432 0
c 736 352 736 400 0 1e-8 0.001
w 656 400 608 400 0
w 656 352 608 352 0
c 672 352 672 400 0 4e-8 0.001
c 608 352 608 400 0 8e-8 0.001
w 736 400 688 400 0
w 832 208 832 192 0
w 768 272 736 272 0
w 896 272 896 192 0
w 896 192 832 192 0
T 1104 304 1168 288 0 2 1 0 0 0.999
c 832 48 800 48 0 0.00009999999999999999 0.001
g 800 48 800 64 0
g 672 432 656 432 0
w 720 304 736 304 0
w 768 336 768 352 0
w 768 352 736 352 0
l 832 48 896 48 0 0.5 0
w 832 96 832 64 0
d 832 64 896 64 1 0.805904783
w 896 48 896 64 0
s 1344 304 1344 368 0 0 false
w 1344 80 1344 256 0
w 832 64 832 48 0
c 832 96 896 96 0 1e-11 0.001
w 896 96 896 64 0
w 832 128 832 192 0
w 912 368 896 368 0
w 1072 384 1072 400 0
w 1200 400 1200 384 0
c 912 368 944 368 0 1e-8 0.001
w 1344 432 1008 432 0
w 896 432 1008 432 0
w 720 240 720 224 0
w 688 272 688 288 0
169 1168 112 1168 176 0 4 3 0 0 0 0.99
f 944 368 992 368 0 10 0.02
w 1008 384 1008 432 0
d 1008 304 1008 352 1 0.805904783
w 1008 80 1008 304 0
w 1008 352 992 352 0
w 1008 384 992 384 0
d 1344 80 1312 80 1 0.805904783
d 1072 304 1104 304 1 0.805904783
l 1248 80 1312 80 0 0.5 0
d 928 48 896 48 1 0.805904783
w 1344 80 1344 48 0
l 1072 272 1072 176 0 2 0
d 1168 80 1168 112 1 0.805904783
w 928 48 1344 48 0
c 1024 352 1024 384 0 1e-12 0.001
w 1024 352 1008 352 0
w 1024 384 1008 384 0
w 1232 80 1168 80 0
w 1008 80 1104 80 0
w 1200 384 1200 304 0
w 1072 384 1072 304 0
w 1200 176 1200 272 0
d 1072 176 1104 176 1 0.805904783
w 1136 176 1136 208 0
w 1184 208 1216 208 0
w 1216 208 1216 112 0
c 1072 272 1072 304 0 1e-12 0.001
c 1200 272 1200 304 0 1e-12 0.001
w 1104 112 1104 80 0
c 1248 160 1248 128 0 0.000009999999999999999 0.001
d 1200 272 1168 272 1 0.805904783
d 1136 208 1184 208 1 0.805904783
w 1232 80 1248 80 0
w 1216 112 1168 112 0
l 1248 80 1248 128 0 0.5 0
g 1248 160 1248 176 0
d 1168 304 1200 304 1 0.805904783
d 1168 176 1200 176 1 0.805904783
d 1232 272 1200 272 1 0.805904783
g 1232 272 1232 288 0
w 1104 272 1072 272 0
f 880 128 928 128 9 1.5 0.02
w 832 128 880 128 0
w 832 128 832 96 0
w 928 48 928 112 0
w 928 144 928 304 0
w 928 304 944 304 0
w 944 304 944 368 0
d 896 304 896 336 1 0.805904783
o 2 64 0 12545 0.0001 0.0001 0 2 2 3
o 3 64 0 12546 0.0001 0.0001 1 2 3 3
o 4 64 0 12545 0.0001 0.0001 2 2 4 3
o 80 64 0 12546 0.0001 0.0001 3 2 80 3

Re: Dave Lawton WFC circuit, improved
« Reply #5,  »Last edited by Cycle
I continue to learn more about how this sort of circuit works... the upstream inductors and capacitors form a resonant circuit which, in this case, has a natural resonance at around 11.7 KHz.

So the goal is to get your frequency as close to this as possible (or some multiple of it). When you do so, you'll notice your ammeter 'bouncing' from 0 amps up to whatever the upstream inductors and capacitors are drawing at the moment, and those 'bounces' are successively increasing.

Ideally, you'd figure out the resonant frequency of your WFC, then design the circuit to hit that resonant frequency naturally.

If you get your frequency higher than resonance, you'll notice that the bottoms of the 'bounces' lift off the 0 volt line and start successively increasing. If you get your frequency lower than resonance, you'll notice that the top of the 'bounces' start getting successively lower and your WFC won't maintain voltage.

Then once you're dialed in on the frequency and you're at your operating WFC voltage, your next goal is to get your 'Mark / Space' as wide as possible while maintaining the frequency and the WFC voltage.

Note that by widening your 'Mark / Space', you're lowering your overall frequency, so you need to come into it with your frequency a bit higher than resonance, then widen your 'Mark / Space' by one click and watch your frequency. If it's still higher than resonance, you can widen your 'Mark / Space' by one more click. Let it run for a bit in between each change, as it takes awhile for the changes to propagate.

Increase your 'Frequency' setting as necessary to maintain resonance. You'll note the voltage upstream of the transformer's primary continually increases.

Here's the latest iteration of the circuit, you'll note I added an additional inductor and capacitor upstream of the transformer, as a high-frequency filter. It's not working exactly as I want, but it's getting closer.
Re: Dave Lawton WFC circuit, improved
« Reply #6,  »Last edited by Cycle
Ok, I've been working on this circuit for awhile now, and I think I've hit upon the best I can do.

It builds voltage a little more slowly than the previous circuit (34.4 kV in 26 msec for the new circuit versus ~40 kV in 26 msec for the old circuit), but it does so while charging the battery (after the initial current flow to charge up the 100 uF capacitor which provides voltage fluctuation smoothing for the 555 timer part of the circuit, and acts as a rudimentary timer to allow the current flow through the transformer and inductors to increase sufficiently before the 555 timer kicks in).


Here's a full-sized image.

The charts below the circuit are as follows:
Ammeter (amps)
6.3 nF capacitor on the transformer secondary-side center tap (voltage)
6.3 nF capacitor acting as the WFC (voltage)
Spark gap acting as WFC current leakage (current)
Output of voltage pump (upstream of transformer primary) (voltage)
Voltage pump 1 uF capacitor (voltage)

You'll note in the image above that the ammeter shows negative amperage, indicating that the battery is charging. I did this by "splitting the positive" on a voltage pump circuit (the diodes, capacitor and inductors in the upper-right corner of the circuit), which keeps the 4 Henry inductor "spinning" and thus pumping voltage into the primary of the transformer, despite the fact that it's drawing from the low-voltage rail. This keeps the current flowing through the transformer, without having to draw from the battery... all the battery is used for is a small "kick" on each cycle to keep things moving, which is then returned to the battery (and then some).

As in the older circuit, it uses a p-MOSFET to keep the n-MOSFETs closed until the 555 timer circuit kicks in. This allows the current flow to increase through the inductors and transformer primary sufficiently to get the whole thing rolling. Once 555 timer circuit voltage is high enough that the 555 timer part of the circuit can control the n-MOSFETs, the p-MOSFET doesn't do anything.

You'll note it's now got "electron extraction", as well. It increases the positive voltage on the secondary side of the transformer (remember, a negative voltage denotes a surplus of electrons, whereas a positive voltage denotes a dearth of electrons).

You'll also note I used 4 MOSFETs... this is because the MOSFETs have a low Source-to-Drain current flow-through, so 4 of them allows sufficient current flow.

Here's the circuit.

And here's the code for it:
Code: [Select]
$ 13 1.0000000000000001e-7 0.9487735836358526 34 5 66
165 768 240 784 240 2 0
v 1312 432 1312 368 0 0 40 12 0 0 0.5
370 1312 304 1312 256 1 0
c 1136 272 1072 272 0 6.302e-9 0.001
187 1136 288 1072 288 0 78.54 2940000 250000 0.5
r 896 384 896 432 0 10000
w 896 336 896 384 0
c 832 400 832 432 0 1e-8 0.001
w 832 432 896 432 0
w 768 352 768 368 0
w 736 352 672 352 0
w 672 352 656 352 0
d 736 304 736 352 1 0.805904783
d 656 352 656 304 1 0.805904783
174 656 304 704 304 0 10000 0.0347 Mark / Space
w 688 272 736 272 0
w 736 272 736 304 0
174 752 208 704 208 0 10000 0.5198 Frequency
w 688 208 688 272 0
w 720 240 752 240 0
w 752 208 752 240 0
w 752 208 752 176 0
w 752 176 912 176 0
w 912 176 912 304 0
w 912 304 896 304 0
S 672 432 672 400 0 2 false 0 3
w 672 432 832 432 0
c 736 352 736 400 0 1e-8 0.001
w 656 400 608 400 0
w 656 352 608 352 0
c 672 352 672 400 0 1.5000000000000002e-8 0.001
c 608 352 608 400 0 2e-8 0.001
w 736 400 688 400 0
w 832 208 832 192 0
w 768 272 736 272 0
w 896 272 896 192 0
w 896 192 832 192 0
T 1072 240 1136 224 0 2 1 0 0 0.999
c 832 32 800 32 0 0.00009999999999999999 0.001
g 800 32 800 48 0
g 672 432 656 432 0
w 720 304 736 304 0
w 768 336 768 352 0
w 768 352 736 352 0
l 832 32 896 32 0 0.5 0
w 832 80 832 48 0
d 832 48 896 48 1 0.805904783
w 896 32 896 48 0
s 1312 304 1312 368 0 0 false
w 1312 64 1312 160 0
w 832 48 832 32 0
c 832 80 896 80 0 1e-11 0.001
w 896 80 896 48 0
w 832 128 832 192 0
w 928 384 896 384 0
w 1072 272 1072 288 0
w 1136 288 1136 272 0
w 1200 432 1136 432 0
w 896 432 1136 432 0
w 720 240 720 224 0
w 688 272 688 288 0
169 1136 64 1136 128 0 4 1 0 0 0 0.99
f 960 384 960 336 8 10 0.02
w 944 64 944 144 0
d 1072 272 1072 240 1 0.805904783
d 928 32 896 32 1 0.805904783
l 1040 208 1040 160 0 2 0
w 928 32 1312 32 0
d 1040 128 1072 128 1 0.805904783
d 1136 240 1136 272 1 0.805904783
d 1136 128 1168 128 1 0.805904783
w 1072 208 1040 208 0
d 896 304 896 336 1 0.805904783
f 1008 384 1008 336 8 10 0.02
f 1056 384 1056 336 8 10 0.02
f 1104 384 1104 336 8 10 0.02
w 928 384 960 384 0
w 960 384 1008 384 0
w 1008 384 1056 384 0
w 1056 384 1104 384 0
w 944 336 944 320 0
w 944 320 992 320 0
w 992 320 992 336 0
w 992 320 1040 320 0
w 1040 320 1040 336 0
w 1040 320 1088 320 0
w 1088 320 1088 336 0
w 976 336 976 304 0
w 976 304 1024 304 0
w 1024 304 1024 336 0
w 1024 304 1072 304 0
w 1072 304 1072 336 0
w 1072 304 1120 304 0
w 1120 304 1120 336 0
w 1120 304 1136 304 0
w 1136 304 1136 432 0
w 1136 208 1168 208 0
w 944 64 1072 64 0
d 944 240 976 240 1 0.805904783
c 1104 128 1104 160 0 6.3e-9 0.001
g 1104 160 1104 176 0
d 1312 64 1264 64 1 0.805904783
s 1024 240 1072 240 0 0 false
f 880 128 928 128 9 1.5 0.02
w 928 32 928 112 0
w 928 144 928 384 0
w 880 128 832 128 0
w 832 128 832 80 0
l 1216 64 1264 64 0 2 0
w 1168 128 1168 208 0
l 1264 160 1312 160 0 2 0
w 1200 240 1200 432 0
w 1200 432 1312 432 0
w 1216 64 1200 64 0
c 1264 160 1264 64 0 0.000001 0.001
l 1200 160 1200 208 0 4 0
d 1200 240 1200 208 1 0.805904783
d 1200 160 1264 160 1 0.805904783
d 1200 160 1200 64 1 0.805904783
d 1200 64 1136 64 1 0.805904783
l 944 208 944 144 0 0.25 0
w 1312 64 1312 32 0
w 944 272 944 304 0
l 976 240 1024 240 0 0.1 0
w 1040 160 1040 128 0
w 944 208 944 240 0
w 944 304 944 320 0
d 944 240 944 272 1 0.805904783
c 944 304 976 304 0 1e-12 0.001
w 1312 160 1312 256 0
o 2 16 0 12553 0.0001 0.0001 0 2 2 3
o 99 16 0 12546 0.0001 0.0001 1 2 99 3
o 3 16 0 12546 0.0001 0.0001 2 2 3 3
o 4 16 0 12545 0.0001 0.0001 3 2 4 3
o 113 16 0 12546 0.0001 0.0001 4 2 113 3
o 114 16 0 12546 0.0001 0.0001 5 2 114 3

Now, I'm sure someone will claim "But you're just 'spinning up' the inductors prior to the 555 timer part of the circuit kicking in, then allowing the circuit to 'coast down' as WFC voltage builds! Look at the current flow before the 555 timer circuit kicks in!"

Yeah, but no. Most of the current during circuit initiation goes into the 100 uF capacitor feeding the 555 timer part of the circuit (a kludgy but effective timer which allows some current flow to build in the inductors and transformer primary). The WFC part of the circuit only has ~17.14 mA flowing through the inductors and transformer primary at the time the 555 timer circuit kicks in (at ~10.4 msec), which you can ascertain for yourself by running the circuit and hovering your cursor over the 250 mH inductor downstream of the transformer primary.
Re: Dave Lawton WFC circuit, improved
« Reply #7,  »Last edited by Cycle
Well, never say never... I thought the last circuit was about as good as my meager electronics knowledge could do, but this latest one blows them all away. It builds voltage in the WFC more quickly, while drawing less from the battery (and charging the battery more).


Here's the full-sized image.

As you can see, this circuit builds to 42.97 kV in 25 msec, whereas the last-best circuit built to ~40 kV in 26 msec. It does this because I reworked the voltage pump part of the circuit, using two zener diodes as a crude "voltage regulator" to maintain a voltage differential across the inductors going to and from the battery. This keeps them 'spinning', which helps to keep the 4 Henry inductor 'spinning', which helps to pump current across the transformer primary... apparently the magnetic field in the transformer doesn't rely upon voltage, it relies upon current... we're just drawing that current from the low-voltage rail, rather than the battery, while using the battery to 'kick' the circuit once per cycle to keep things going.

Here's the code for it:
Code: [Select]
$ 13 1.0000000000000001e-7 0.9487735836358526 34 5 66
165 768 240 784 240 2 0
v 1392 432 1392 368 0 0 40 12 0 0 0.5
370 1392 304 1392 256 1 0
c 1136 272 1072 272 0 6.302e-9 0.001
187 1136 288 1072 288 0 78.54 2940000 250000 0.5
r 896 384 896 432 0 10000
w 896 336 896 384 0
c 832 400 832 432 0 1e-8 0.001
w 832 432 896 432 0
w 768 352 768 368 0
w 736 352 672 352 0
w 672 352 656 352 0
d 736 304 736 352 1 0.805904783
d 656 352 656 304 1 0.805904783
174 656 304 704 304 0 10000 0.0347 Mark / Space
w 688 272 736 272 0
w 736 272 736 304 0
174 752 208 704 208 0 10000 0.5099 Frequency
w 688 208 688 272 0
w 720 240 752 240 0
w 752 208 752 240 0
w 752 208 752 176 0
w 752 176 912 176 0
w 912 176 912 304 0
w 912 304 896 304 0
S 672 432 672 400 0 2 false 0 3
w 672 432 832 432 0
c 736 352 736 400 0 1e-8 0.001
w 656 400 608 400 0
w 656 352 608 352 0
c 672 352 672 400 0 1.5000000000000002e-8 0.001
c 608 352 608 400 0 2e-8 0.001
w 736 400 688 400 0
w 832 208 832 192 0
w 768 272 736 272 0
w 896 272 896 192 0
w 896 192 832 192 0
T 1072 240 1136 224 0 2 1 0 0 0.999
c 832 32 800 32 0 0.00009999999999999999 0.001
g 800 32 800 48 0
g 672 432 656 432 0
w 720 304 736 304 0
w 768 336 768 352 0
w 768 352 736 352 0
l 832 32 896 32 0 0.5 0
w 832 80 832 48 0
d 832 48 896 48 1 0.805904783
w 896 32 896 48 0
s 1392 304 1392 368 0 0 false
w 1392 128 1392 192 0
w 832 48 832 32 0
c 832 80 896 80 0 1e-11 0.001
w 896 80 896 48 0
w 832 128 832 192 0
w 928 384 896 384 0
w 1072 272 1072 288 0
w 1136 288 1136 272 0
w 1200 432 1136 432 0
w 896 432 1136 432 0
w 720 240 720 224 0
w 688 272 688 288 0
169 1136 64 1136 128 0 4 1 0 0 0 0.99
f 960 384 960 336 8 10 0.02
w 944 64 944 144 0
d 1072 272 1072 240 1 0.805904783
d 928 32 896 32 1 0.805904783
l 1040 208 1040 160 0 2 0
w 928 32 1392 32 0
d 1040 128 1072 128 1 0.805904783
d 1136 240 1136 272 1 0.805904783
d 1136 128 1168 128 1 0.805904783
w 1072 208 1040 208 0
d 896 304 896 336 1 0.805904783
f 1008 384 1008 336 8 10 0.02
f 1056 384 1056 336 8 10 0.02
f 1104 384 1104 336 8 10 0.02
w 928 384 960 384 0
w 960 384 1008 384 0
w 1008 384 1056 384 0
w 1056 384 1104 384 0
w 944 336 944 320 0
w 944 320 992 320 0
w 992 320 992 336 0
w 992 320 1040 320 0
w 1040 320 1040 336 0
w 1040 320 1088 320 0
w 1088 320 1088 336 0
w 976 336 976 304 0
w 976 304 1024 304 0
w 1024 304 1024 336 0
w 1024 304 1072 304 0
w 1072 304 1072 336 0
w 1072 304 1120 304 0
w 1120 304 1120 336 0
w 1120 304 1136 304 0
w 1136 304 1136 432 0
w 1136 208 1168 208 0
w 944 64 1040 64 0
d 944 240 976 240 1 0.805904783
c 1104 128 1104 160 0 6.3e-9 0.001
g 1104 160 1104 176 0
d 1392 64 1264 64 1 0.805904783
s 1024 240 1072 240 0 0 false
f 880 128 928 128 9 1.5 0.02
w 928 32 928 112 0
w 928 144 928 384 0
w 880 128 832 128 0
w 832 128 832 80 0
l 1216 64 1264 64 0 2 0
w 1168 128 1168 208 0
l 1264 192 1392 192 0 2 0
w 1200 272 1200 432 0
w 1200 432 1392 432 0
w 1216 64 1200 64 0
l 1200 192 1200 240 0 4 0
d 1200 272 1200 240 1 0.805904783
d 1200 192 1232 192 1 0.805904783
d 1200 128 1200 64 1 0.805904783
d 1200 64 1136 64 1 0.805904783
l 944 208 944 144 0 1 0
w 1392 64 1392 32 0
w 944 240 944 304 0
l 976 240 1024 240 0 0.1 0
w 1040 160 1040 128 0
w 944 208 944 240 0
w 944 304 944 320 0
d 1072 64 1040 64 1 0.805904783
c 944 304 976 304 0 1e-12 0.001
w 1392 192 1392 256 0
z 1264 64 1264 128 1 0.805904783 5.6
z 1264 192 1264 128 1 0.805904783 5.6
w 1232 192 1264 192 0
c 1264 128 1312 128 0 1e-9 0.001
w 1200 128 1200 192 0
w 1392 128 1312 128 0
w 1392 128 1392 64 0
o 2 16 0 12553 0.0001 0.0001 0 2 2 3
o 99 16 0 12546 0.0001 0.0001 1 2 99 3
o 3 16 0 12546 0.0001 0.0001 2 2 3 3
o 4 16 0 12545 0.0001 0.0001 3 2 4 3
o 113 16 0 12546 0.0001 0.0001 4 2 113 3
o 132 16 0 12546 0.0001 0.0001 5 2 132 3

Matt Watts

Re: Dave Lawton WFC circuit, improved
« Reply #8,  »
Certainly have some step-charging going there Cycle.   Nice work.

Cycle

Re: Dave Lawton WFC circuit, improved
« Reply #9,  »Last edited by Cycle
Resonance is a strange thing. I wanted to see what would happen if I opened the switch on the battery as the circuit was running... the circuit kept going, but the WFC part of the circuit was pumping too much voltage into the 555 timer part of the circuit.

So I redid the circuit a bit, with a separate switch for the 555 timer part of the circuit, and a SPDT switch for the WFC part of the circuit. One side of the SPDT went to the battery, the other went to ground via a diode.

If you run the circuit without first getting some current flowing through the transformer primary, WFC voltage goes to zero, which is to be expected.


Here's the full-sized image.

But if you get current flowing through the transformer primary as the 555 timer circuit voltage builds, then switch the SPDT switch (after the 555 timer kicks in) so it's pulling from the low voltage rail rather than the battery, this happens...

Here's the full-sized image.

You'll note the WFC leakage current (through the spark gap) is nearly as much as the current through the transformer primary during circuit initialization... leave it running, and it goes higher than initialization current (and that leakage current runs for much longer than the ~10.4 msec initialization current). So it can't be just due to the inductors "coasting down"... especially considering that the initialization current is ~16.37 mA (at the time when the 555 timer kicks in), then it builds up from there to hundreds of mA... none of it from the battery. How is it doing this? No idea.

Here's the circuit.

And here's the code:
Code: [Select]
$ 13 1.0000000000000001e-7 0.7087504708082256 36 5 66
165 768 240 784 240 2 0
v 1408 432 1408 400 0 0 40 12 0 0 0.5
370 1408 400 1408 352 1 0
c 1136 240 1072 240 0 6.302e-9 0.001
187 1136 256 1072 256 0 78.54 2940000 250000 0.5
r 896 384 896 432 0 25000
w 896 336 896 384 0
c 832 400 832 432 0 1e-8 0.001
w 832 432 896 432 0
w 768 352 768 368 0
w 736 352 672 352 0
w 672 352 656 352 0
d 736 304 736 352 1 0.805904783
d 656 352 656 304 1 0.805904783
174 656 304 704 304 0 10000 0.0347 Mark / Space
w 688 272 736 272 0
w 736 272 736 304 0
174 752 208 704 208 0 10000 0.5 Frequency
w 688 208 688 272 0
w 720 240 752 240 0
w 752 208 752 240 0
w 752 208 752 192 0
w 752 192 912 192 0
w 912 192 912 304 0
w 912 304 896 304 0
S 672 432 672 400 0 2 false 0 3
w 672 432 832 432 0
c 736 352 736 400 0 1e-8 0.001
w 656 400 608 400 0
w 656 352 608 352 0
c 672 352 672 400 0 1.5000000000000002e-8 0.001
c 608 352 608 400 0 2e-8 0.001
w 736 400 688 400 0
w 768 272 736 272 0
w 896 272 896 208 0
w 896 208 832 208 0
T 1072 208 1136 192 0 2 1 0 0 0.999
c 832 32 800 32 0 0.00009999999999999999 0.001
g 800 32 800 48 0
g 672 432 656 432 0
w 720 304 736 304 0
w 768 336 768 352 0
w 768 352 736 352 0
l 832 32 896 32 0 0.5 0
w 832 80 832 48 0
d 832 48 896 48 1 0.805904783
w 896 32 896 48 0
w 832 48 832 32 0
c 832 80 896 80 0 1e-11 0.001
w 896 80 896 48 0
w 832 128 832 208 0
w 928 384 896 384 0
w 1072 240 1072 256 0
w 1136 256 1136 240 0
w 1200 432 1152 432 0
w 896 432 1152 432 0
w 720 240 720 224 0
w 688 272 688 288 0
169 1136 64 1136 128 0 4 1 0 0 0 0.99
f 976 384 976 336 8 10 0.02
w 944 64 944 208 0
d 1072 240 1072 208 1 0.805904783
d 928 32 896 32 1 0.805904783
l 1040 176 1040 128 0 2 0
w 928 32 1408 32 0
d 1040 128 1072 128 1 0.805904783
d 1136 208 1136 240 1 0.805904783
d 1136 128 1168 128 1 0.805904783
w 1072 176 1040 176 0
d 896 304 896 336 1 0.805904783
f 1024 384 1024 336 8 10 0.02
f 1072 384 1072 336 8 10 0.02
f 1120 384 1120 336 8 10 0.02
w 928 384 976 384 0
w 976 384 1024 384 0
w 1024 384 1072 384 0
w 1072 384 1120 384 0
w 960 336 960 320 0
w 960 320 1008 320 0
w 1008 320 1008 336 0
w 1008 320 1056 320 0
w 1056 320 1056 336 0
w 1056 320 1104 320 0
w 1104 320 1104 336 0
w 992 336 992 304 0
w 992 304 1040 304 0
w 1040 304 1040 336 0
w 1040 304 1088 304 0
w 1088 304 1088 336 0
w 1088 304 1136 304 0
w 1136 304 1136 336 0
w 1136 304 1152 304 0
w 1136 176 1168 176 0
d 944 208 976 208 1 0.805904783
c 1104 128 1104 160 0 6.3e-9 0.001
g 1104 160 1088 160 0
d 1392 64 1296 64 1 0.805904783
s 1024 208 1072 208 0 0 false
f 880 128 928 128 9 1.5 0.02
w 928 32 928 112 0
w 928 144 928 384 0
w 880 128 832 128 0
w 832 128 832 80 0
l 1216 64 1296 64 0 2 0
w 1168 128 1168 176 0
w 1200 432 1408 432 0
w 1216 64 1200 64 0
l 1200 256 1200 320 0 4 0
d 1200 352 1200 320 1 0.805904783
d 1200 256 1200 64 1 0.805904783
l 976 208 1024 208 0 0.1 0
w 960 304 960 320 0
c 960 304 992 304 0 1e-12 0.001
w 1392 160 1392 256 0
z 1296 64 1296 160 1 0.805904783 5.6
z 1296 256 1296 160 1 0.805904783 5.6
w 1392 160 1392 64 0
w 960 304 944 304 0
w 944 304 944 240 0
w 944 64 992 64 0
d 1072 64 1040 64 1 0.805904783
l 992 64 1040 64 0 0.25 0
d 1200 64 1136 64 1 0.805904783
w 1152 304 1152 432 0
c 1072 64 1136 64 0 1e-12 0.001
l 1296 256 1392 256 0 0.25 0
d 1200 256 1296 256 1 0.805904783
d 944 208 944 240 1 0.805904783
c 1296 160 1392 160 0 1e-9 0.001
d 1344 352 1376 352 1 0.805904783
S 1392 320 1392 352 0 1 false 0 2
w 1392 320 1392 256 0
w 1408 304 1408 32 0
w 1344 352 1200 352 0
w 1200 352 1200 432 0
s 1408 304 1408 352 0 0 false
o 2 16 0 12553 0.0001 0.0001 0 2 2 3
o 94 16 0 12546 0.0001 0.0001 1 2 94 3
o 3 16 0 12546 0.0001 0.0001 2 2 3 3
o 4 16 0 12545 0.0001 0.0001 3 2 4 3
o 106 16 0 12546 0.0001 0.0001 4 2 106 3

So with a bit more optimization, we could have a circuit which builds voltage in the WFC with extremely low current draw.
Re: Dave Lawton WFC circuit, improved
« Reply #10,  »
I redid the circuit again with a SPDT switch with one lead going to the battery, and one going to the low-voltage rail, feeding into the 4 Henry inductor, with a diode between the low-voltage rail and the SPDT switch so voltage could only go into the WFC part of the circuit. After the 555 timer kicked in, I switched the SPDT so the WFC part of the circuit was only drawing current from the low-voltage rail.

It ran harder than any other circuit (it pulled to 39.4 kV in 16.47 msec, just 6 msec after the 555 timer part of the circuit kicks in... so it built up by 39.4 kV in 6 msec, more than double the rate of the last-best circuit). But it's unstable. I can't seem to get it into resonance.

I think it was doing that because as the "electron extraction" part of the circuit charged up, it was taking current from downstream of the transformer primary. This created a 'vacuum' which sucked more current in via the diode and SPDT switch. Current being taken off downstream of the transformer primary, and being put in upstream made the whole circuit accelerate from the ~16.51 mA at 555 timer circuit kick-in, to hundreds of circulating milliamps (none from the battery). So as the circuit ran, it just kept re-biasing to higher and higher voltages. But like I said, it's unstable and I can't seem to get it to stay in resonance... it'll hit resonance and the voltage will skyrocket, then voltage rise will slow to a crawl, then it'll hit resonance again, etc. It seems the resonant frequency is changing as the voltage changes.

So I'm going back to the last-best circuit, and adding a diode between the battery and the WFC part of the circuit... charging the battery isn't our main goal, building high voltage quickly in the WFC is.

While running the circuit, I found a couple tweaks that help it to build voltage faster. After I test the circuit more, I'll post it.
Re: Dave Lawton WFC circuit, improved
« Reply #11,  »Last edited by Cycle
Ok, it's still not perfect... but this is a big improvement.

The circuit worked better when it was allowed to charge the battery, so I rolled back a couple iterations. It also seems to like a negative voltage downstream of the n-MOSFETs, so I used a zener diode between the WFC part of the circuit and the low voltage rail... this lets the battery push current through the transformer primary during circuit initialization to get everything rolling, but after the circuit is running (and pulling current from the low-voltage rail), it holds a slight negative voltage downstream of the n-MOSFETs.

The 555 timer kicks in at ~10.47 msec. Here's a breakdown of how quickly the WFC voltage builds:
At 555 timer kick-in: 3.76 V
1 msec after 555 timer kick-in: 32.93 kV
2 msec after 555 timer kick-in: 82.96 kV
3 msec after 555 timer kick-in: 125.96 kV
4 msec after 555 timer kick-in: 154.39 kV

Yeah, it runs up pretty fast now. :shocked:

And it does that while charging the battery. :bliss:


Here's the full-sized image.

The circuit emulator is a bit flaky... sometimes it'll crash when the ammeter crosses the zero threshold. If that happens, the circuit will stop. You can get it running again, picking up where it left off, by opening then closing any of the switches in the circuit, then hitting the Run/Stop button. It'll keep running from the point the circuit emulator crashed. I've noticed with other circuits I've experimented with that it does this more and more frequently as a circuit becomes more complex.

Here's the circuit.

And here's the code:
Code: [Select]
$ 13 1.0000000000000001e-7 0.625470095193633 35 5 66
165 768 240 784 240 2 0
v 1344 432 1344 384 0 0 40 12 0 0 0.5
370 1344 384 1344 336 1 0
c 1136 240 1072 240 0 6.302e-9 0.001
187 1136 256 1072 256 0 78.54 2940000 500000 0.5
r 896 384 896 432 0 25000
w 896 336 896 384 0
c 832 400 832 432 0 1e-8 0.001
w 832 432 896 432 0
w 768 352 768 368 0
w 736 352 672 352 0
w 672 352 656 352 0
d 736 304 736 352 1 0.805904783
d 656 352 656 304 1 0.805904783
174 656 304 704 304 0 10000 0.0347 Mark / Space
w 688 272 736 272 0
w 736 272 736 304 0
174 752 208 704 208 0 10000 0.5 Frequency
w 688 208 688 272 0
w 720 240 752 240 0
w 752 208 752 240 0
w 752 208 752 192 0
w 752 192 912 192 0
w 912 192 912 304 0
w 912 304 896 304 0
S 672 432 672 400 0 2 false 0 3
w 672 432 832 432 0
c 736 352 736 400 0 1e-8 0.001
w 656 400 608 400 0
w 656 352 608 352 0
c 672 352 672 400 0 1.5000000000000002e-8 0.001
c 608 352 608 400 0 2e-8 0.001
w 736 400 688 400 0
w 768 272 736 272 0
w 896 272 896 208 0
w 896 208 832 208 0
T 1072 208 1136 192 0 2 1 0 0 0.999
c 832 32 800 32 0 0.00009999999999999999 0.001
g 800 32 800 48 0
g 672 432 656 432 0
w 720 304 736 304 0
w 768 336 768 352 0
w 768 352 736 352 0
l 832 32 896 32 0 0.5 0
w 832 80 832 48 0
d 832 48 896 48 1 0.805904783
w 896 32 896 48 0
w 832 48 832 32 0
c 832 80 896 80 0 1e-11 0.001
w 896 80 896 48 0
w 832 128 832 208 0
w 928 384 896 384 0
w 1072 240 1072 256 0
w 1136 256 1136 240 0
w 1216 304 1136 304 0
w 896 432 1216 432 0
w 720 240 720 224 0
w 688 272 688 288 0
169 1136 64 1136 128 0 4 1 0 0 0 0.99
f 976 384 976 336 8 10 0.02
w 944 64 944 208 0
d 1072 240 1072 208 1 0.805904783
d 928 32 896 32 1 0.805904783
l 1040 176 1040 128 0 2 0
w 928 32 1344 32 0
d 1040 128 1072 128 1 0.805904783
d 1136 208 1136 240 1 0.805904783
d 1136 128 1168 128 1 0.805904783
w 1072 176 1040 176 0
d 896 304 896 336 1 0.805904783
f 1024 384 1024 336 8 10 0.02
f 1072 384 1072 336 8 10 0.02
f 1120 384 1120 336 8 10 0.02
w 928 384 976 384 0
w 976 384 1024 384 0
w 1024 384 1072 384 0
w 1072 384 1120 384 0
w 960 336 960 320 0
w 960 320 1008 320 0
w 1008 320 1008 336 0
w 1008 320 1056 320 0
w 1056 320 1056 336 0
w 1056 320 1104 320 0
w 1104 320 1104 336 0
w 992 336 992 304 0
w 992 304 1040 304 0
w 1040 304 1040 336 0
w 1040 304 1088 304 0
w 1088 304 1088 336 0
w 1088 304 1136 304 0
w 1136 304 1136 336 0
w 1136 176 1168 176 0
d 944 208 1008 208 1 0.805904783
c 1104 128 1104 160 0 6.3e-9 0.001
g 1104 160 1088 160 0
d 1328 64 1264 64 1 0.805904783
f 880 128 928 128 9 1.5 0.02
w 928 32 928 112 0
w 928 144 928 384 0
w 880 128 832 128 0
w 832 128 832 80 0
l 1216 64 1264 64 0 2 0
w 1168 128 1168 176 0
w 1216 432 1344 432 0
w 1216 64 1200 64 0
l 1216 128 1216 176 0 4 0
d 1216 208 1216 176 1 0.805904783
d 1216 128 1216 64 1 0.805904783
l 1008 208 1072 208 0 2 0
w 960 304 960 320 0
c 960 304 992 304 0 1e-12 0.001
w 1328 96 1328 128 0
z 1264 64 1264 96 1 0.805904783 5.6
z 1264 128 1264 96 1 0.805904783 5.6
w 1328 96 1328 64 0
w 960 304 944 304 0
w 944 304 944 240 0
w 944 64 992 64 0
d 1072 64 1040 64 1 0.805904783
l 992 64 1040 64 0 2 0
d 1200 64 1136 64 1 0.805904783
w 1216 336 1216 432 0
l 1264 128 1328 128 0 2 0
d 1216 128 1264 128 1 0.805904783
d 944 208 944 240 1 0.805904783
c 1264 96 1328 96 0 1e-9 0.001
w 1344 272 1344 32 0
s 1344 272 1344 336 0 0 false
w 1344 336 1328 336 0
w 1216 256 1216 208 0
w 1328 224 1328 128 0
s 1328 336 1328 272 0 0 false
l 1216 256 1216 304 0 4 0
z 1328 272 1328 224 1 0.805904783 5.6
z 1216 304 1216 336 1 0.805904783 5.6
o 2 16 0 12553 0.0001 0.0001 0 2 2 3
o 93 16 0 12546 0.0001 0.0001 1 2 93 3
o 3 16 0 12546 0.0001 0.0001 2 2 3 3
o 4 16 0 12545 0.0001 0.0001 3 2 4 3
o 104 16 0 12546 0.0001 0.0001 4 2 104 3
Re: Dave Lawton WFC circuit, improved
« Reply #12,  »
A slight tweak, and slightly better performance. This goes to 164.93 kV in 14.47 msec (4 msec after the 555 timer kicks in), and is more stable. It takes a bit longer for the circuit to settle down and start pushing current back into the battery, but that's an acceptable tradeoff.


Here's the full-sized image.

Here's the circuit.

And here's the code:
Code: [Select]
$ 13 1.0000000000000001e-7 0.21170000166126748 35 5 66
165 768 240 784 240 2 0
v 1344 432 1344 384 0 0 40 12 0 0 0.5
370 1344 384 1344 336 1 0
c 1136 240 1072 240 0 6.302e-9 0.001
187 1136 256 1072 256 0 78.54 2940000 500000 0.5
r 896 384 896 432 0 25000
w 896 336 896 384 0
c 832 400 832 432 0 1e-8 0.001
w 832 432 896 432 0
w 768 352 768 368 0
w 736 352 672 352 0
w 672 352 656 352 0
d 736 304 736 352 1 0.805904783
d 656 352 656 304 1 0.805904783
174 656 304 704 304 0 10000 0.0347 Mark / Space
w 688 272 736 272 0
w 736 272 736 304 0
174 752 208 704 208 0 10000 0.5 Frequency
w 688 208 688 272 0
w 720 240 752 240 0
w 752 208 752 240 0
w 752 208 752 192 0
w 752 192 912 192 0
w 912 192 912 304 0
w 912 304 896 304 0
S 672 432 672 400 0 2 false 0 3
w 672 432 832 432 0
c 736 352 736 400 0 1e-8 0.001
w 656 400 608 400 0
w 656 352 608 352 0
c 672 352 672 400 0 1.5000000000000002e-8 0.001
c 608 352 608 400 0 2e-8 0.001
w 736 400 688 400 0
w 768 272 736 272 0
w 896 272 896 208 0
w 896 208 832 208 0
T 1072 208 1136 192 0 2 1 0 0 0.999
c 832 32 800 32 0 0.00009999999999999999 0.001
g 800 32 800 48 0
g 672 432 656 432 0
w 720 304 736 304 0
w 768 336 768 352 0
w 768 352 736 352 0
l 832 32 896 32 0 0.5 0
w 832 80 832 48 0
d 832 48 896 48 1 0.805904783
w 896 32 896 48 0
w 832 48 832 32 0
c 832 80 896 80 0 1e-11 0.001
w 896 80 896 48 0
w 832 128 832 208 0
w 928 384 896 384 0
w 1072 240 1072 256 0
w 1136 256 1136 240 0
w 1216 304 1136 304 0
w 896 432 1216 432 0
w 720 240 720 224 0
w 688 272 688 288 0
169 1136 64 1136 128 0 4 1 0 0 0 0.99
f 976 384 976 336 8 10 0.02
w 944 64 944 208 0
d 1072 240 1072 208 1 0.805904783
d 928 32 896 32 1 0.805904783
l 1040 176 1040 128 0 2 0
w 928 32 1344 32 0
d 1040 128 1072 128 1 0.805904783
d 1136 208 1136 240 1 0.805904783
d 1136 128 1168 128 1 0.805904783
w 1072 176 1040 176 0
d 896 304 896 336 1 0.805904783
f 1024 384 1024 336 8 10 0.02
f 1072 384 1072 336 8 10 0.02
f 1120 384 1120 336 8 10 0.02
w 928 384 976 384 0
w 976 384 1024 384 0
w 1024 384 1072 384 0
w 1072 384 1120 384 0
w 960 336 960 320 0
w 960 320 1008 320 0
w 1008 320 1008 336 0
w 1008 320 1056 320 0
w 1056 320 1056 336 0
w 1056 320 1104 320 0
w 1104 320 1104 336 0
w 992 336 992 304 0
w 992 304 1040 304 0
w 1040 304 1040 336 0
w 1040 304 1088 304 0
w 1088 304 1088 336 0
w 1088 304 1136 304 0
w 1136 304 1136 336 0
w 1136 176 1168 176 0
d 944 208 1008 208 1 0.805904783
c 1104 128 1104 160 0 6.3e-9 0.001
g 1104 160 1088 160 0
d 1328 64 1264 64 1 0.805904783
f 880 128 928 128 9 1.5 0.02
w 928 32 928 112 0
w 928 144 928 384 0
w 880 128 832 128 0
w 832 128 832 80 0
l 1216 64 1264 64 0 2 0
w 1168 128 1168 176 0
w 1216 432 1344 432 0
w 1216 64 1200 64 0
l 1216 128 1216 176 0 4 0
d 1216 208 1216 176 1 0.805904783
d 1216 128 1216 64 1 0.805904783
l 1008 208 1072 208 0 2 0
w 960 304 960 320 0
c 960 304 992 304 0 1e-12 0.001
w 1328 96 1328 128 0
z 1264 64 1264 96 1 0.805904783 5.6
z 1264 128 1264 96 1 0.805904783 5.6
w 1328 96 1328 64 0
w 960 304 944 304 0
w 944 304 944 240 0
w 944 64 992 64 0
d 1072 64 1040 64 1 0.805904783
l 992 64 1040 64 0 2 0
d 1200 64 1136 64 1 0.805904783
w 1216 336 1216 432 0
l 1264 128 1328 128 0 2 0
d 1216 128 1264 128 1 0.805904783
d 944 208 944 240 1 0.805904783
c 1264 96 1328 96 0 1e-8 0.001
w 1344 272 1344 32 0
s 1344 272 1344 336 0 0 false
w 1344 336 1328 336 0
w 1216 256 1216 208 0
w 1328 224 1328 128 0
s 1328 336 1328 272 0 0 false
l 1216 256 1216 304 0 4 0
z 1328 272 1328 224 1 0.805904783 5.6
z 1216 304 1216 336 1 0.805904783 5.6
o 2 16 0 12553 0.0001 0.0001 0 2 2 3
o 93 16 0 12546 0.0001 0.0001 1 2 93 3
o 3 16 0 12546 0.0001 0.0001 2 2 3 3
o 4 16 0 12545 0.0001 0.0001 3 2 4 3
o 104 16 0 12546 0.0001 0.0001 4 2 104 3
Re: Dave Lawton WFC circuit, improved
« Reply #13,  »Last edited by Cycle
Ok, I tweaked the circuit a bit more, improved performance a bit more, and I did some divergent testing... I disconnected the 555 timer part of the circuit and just allowed current to flow through the WFC part of the circuit, with the n-MOSFETs held closed with 12 volts to their gates. Current flow peaks at ~70 mA under this condition.

With the new tweaks, there's more than 2 amps circulating in the WFC primary part of the circuit, and it jumps up to ~165 kV just 3 msec after 555 timer kick-in. I'll test it a bit more, then post it.

I noticed that the n-MOSFETs were firing more often than the 555 timer... the circuit was running at its natural frequency, rather than the frequency set by the 555 timer. This is because of the zener diode downstream of the n-MOSFETs causing a slight positive voltage (when the 4 Henry inductor isn't sucking in current from the low-voltage rail), and a slight negative voltage (when the 4 Henry inductor is sucking in current from the low-voltage rail). When the voltage downstream is negative, the n-MOSFETs close, when it's positive, they open.

So I was hoping to bias the gates of the n-MOSFETs with just enough voltage that the circuit would oscillate without any need of the 555 timer circuit... it'd make for a simpler and more robust circuit if we didn't have to rely upon any chips... but every once in awhile the circuit needs the 555 timer to slam those n-MOSFETs open in order for the whole thing to work.

{EDIT}
Ok, the circuit stalls out at a WFC voltage of ~675 kV, then just maintains that voltage. That's sufficient. :D

I set the WFC spark gap to spark at 50 kV (analogous to the WFC 'dielectric' breaking down), and I set the "electron extraction" spark gap to spark at 250 kV. I added a spark gap that drains voltage off the 6.3 nF "electron extraction" capacitor, because when it gets too high, it starves the primary current flow. That current stored in the cap is largely returned to the battery.

The circuit will take the WFC voltage up to 50 kV about once every 0.4 msec, so you're looking at ~2500 catastrophic breakdowns of the WFC 'dielectric' (the water) per second.

Here's the circuit.
{/EDIT}

{EDIT 2}
I worked out a way to hopefully prevent the circuit emulator from crashing as much. I put a diode and 2 kOhm resistor in parallel between the battery and the WFC part of the circuit. Thus the circuit gets full voltage (minus the drop across the diode) in the forward direction, but it has to go through the resistor in the reverse direction. This biases current flow slightly in the forward direction, and so far there haven't been any crashes.

This also has a beneficial side-effect. It increases the voltage downstream of that diode when the circuit is trying to push current back into the battery, which increases the current circulating in the WFC part of the circuit.

When the "electron extraction" spark gap fires, its pushes current through the transformer primary, then the current is sucked through the already-'spinning' 4 Henry inductors and pushed into the 10 uF capacitor and into the battery. Thus very little of the current goes to the low-voltage rail even when that cap is dumping through the spark gap.

If you watch the little moving dots which represent current flow, right at the battery + terminal, you'll notice that they don't really move all that much, they sort of slosh back and forth while slowly crawling forward (out of the battery). Then the "electron extraction" spark gap fires and current flows into the battery.

I can't think of any way to monitor cumulative power into / out of the battery, though... too bad that they don't have a cumulative power chart.

I'm going to do a long-term test run, letting the circuit run for an hour of emulator time. This'll take quite awhile, since the emulator is stepping at 100 ns, but I want to see if it's truly stable.

Here's the circuit.

And here's the code:
Code: [Select]
$ 13 1.0000000000000001e-7 0.529449005047003 32 10 66
165 768 240 784 240 2 0
v 1344 432 1344 384 0 0 40 12 0 0 0.5
370 1344 384 1344 336 1 0
c 1136 272 1072 272 0 6.302e-9 0.001
187 1136 288 1072 288 0 78.54 2940000 50000 1
r 896 384 896 432 0 25000
w 896 336 896 384 0
c 832 400 832 432 0 1e-8 0.001
w 832 432 896 432 0
w 768 352 768 368 0
w 736 352 672 352 0
w 672 352 656 352 0
d 736 304 736 352 1 0.805904783
d 656 352 656 304 1 0.805904783
174 656 304 704 304 0 10000 0.0347 Mark / Space
w 688 272 736 272 0
w 736 272 736 304 0
174 752 208 704 208 0 10000 0.5 Frequency
w 688 208 688 272 0
w 720 240 752 240 0
w 752 208 752 240 0
w 752 208 752 192 0
w 752 192 912 192 0
w 912 192 912 304 0
w 912 304 896 304 0
S 672 432 672 400 0 2 false 0 3
w 672 432 832 432 0
c 736 352 736 400 0 1e-8 0.001
w 656 400 608 400 0
w 656 352 608 352 0
c 672 352 672 400 0 1.5000000000000002e-8 0.001
c 608 352 608 400 0 2e-8 0.001
w 736 400 688 400 0
w 768 272 736 272 0
w 896 272 896 208 0
w 896 208 832 208 0
T 1072 240 1136 224 0 2 1 0 0 0.999
c 832 32 800 32 0 0.00009999999999999999 0.001
g 800 32 800 48 0
g 672 432 656 432 0
w 720 304 736 304 0
w 768 336 768 352 0
w 768 352 736 352 0
l 832 32 896 32 0 0.5 0
w 832 80 832 48 0
d 832 48 896 48 1 0.805904783
w 896 32 896 48 0
w 832 48 832 32 0
c 832 80 896 80 0 1e-11 0.001
w 896 80 896 48 0
w 832 128 832 208 0
w 928 384 896 384 0
w 1072 272 1072 288 0
w 1136 288 1136 272 0
w 1216 304 1136 304 0
w 896 432 1216 432 0
w 720 240 720 224 0
w 688 272 688 288 0
169 1136 64 1136 128 0 4 1 0 0 0 0.99
f 976 384 976 336 8 10 0.02
w 944 64 944 240 0
d 1072 272 1072 240 1 0.805904783
d 928 32 896 32 1 0.805904783
l 1040 208 1040 128 0 2 0
w 928 32 1344 32 0
d 1040 128 1072 128 1 0.805904783
d 1136 240 1136 272 1 0.805904783
d 1136 128 1168 128 1 0.805904783
w 1072 208 1040 208 0
d 896 304 896 336 1 0.805904783
f 1024 384 1024 336 8 10 0.02
f 1072 384 1072 336 8 10 0.02
f 1120 384 1120 336 8 10 0.02
w 928 384 976 384 0
w 976 384 1024 384 0
w 1024 384 1072 384 0
w 1072 384 1120 384 0
w 960 336 960 320 0
w 960 320 1008 320 0
w 1008 320 1008 336 0
w 1008 320 1056 320 0
w 1056 320 1056 336 0
w 1056 320 1104 320 0
w 1104 320 1104 336 0
w 992 336 992 304 0
w 992 304 1040 304 0
w 1040 304 1040 336 0
w 1040 304 1088 304 0
w 1088 304 1088 336 0
w 1088 304 1136 304 0
w 1136 304 1136 336 0
w 1136 208 1168 208 0
d 944 240 1008 240 1 0.805904783
c 1104 160 1104 192 0 6.3e-9 0.001
g 1104 192 1088 192 0
d 1328 64 1264 64 1 0.805904783
f 880 128 928 128 9 1.5 0.02
w 928 32 928 112 0
w 928 144 928 384 0
w 880 128 832 128 0
w 832 128 832 80 0
l 1216 64 1264 64 0 2 0
w 1168 128 1168 208 0
w 1216 432 1344 432 0
w 1216 64 1200 64 0
l 1216 128 1216 192 0 4 0
d 1216 224 1216 192 1 0.805904783
d 1216 128 1216 64 1 0.805904783
l 1008 240 1072 240 0 2 0
w 960 304 960 320 0
c 960 304 992 304 0 1e-12 0.001
w 1328 96 1328 128 0
z 1264 64 1264 96 1 0.805904783 5.6
z 1264 128 1264 96 1 0.805904783 5.6
w 1328 96 1328 64 0
w 960 304 944 304 0
w 944 304 944 272 0
w 944 64 992 64 0
d 1072 64 1040 64 1 0.805904783
l 992 64 1040 64 0 2 0
d 1200 64 1136 64 1 0.805904783
l 1264 128 1328 128 0 2 0
d 1216 128 1264 128 1 0.805904783
d 944 240 944 272 1 0.805904783
c 1264 96 1328 96 0 0.000009999999999999999 0.001
w 1344 272 1344 32 0
s 1344 272 1344 336 0 0 false
w 1344 336 1328 336 0
w 1328 224 1328 128 0
s 1328 336 1328 272 0 0 false
l 1216 224 1216 304 0 4 0
z 1216 304 1216 432 1 0.805904783 5.6
w 1104 160 1104 128 0
d 1104 160 1184 160 1 0.805904783
187 1184 160 1184 112 0 1 10000000000 250000 1
w 1136 64 1184 112 0
w 1328 224 1296 224 0
w 1328 272 1296 272 0
r 1296 272 1296 224 0 2000
d 1328 272 1328 224 1 0.805904783
o 2 16 0 12553 0.0001 0.0001 0 2 2 3
o 3 16 0 12546 0.0001 0.0001 1 2 3 3
o 4 16 0 12545 0.0001 0.0001 2 2 4 3
o 93 16 0 12546 0.0001 0.0001 3 2 93 3
o 134 16 0 12545 0.0001 0.0001 4 2 134 3
{/EDIT 2}

Gunther Rattay

Re: Dave Lawton WFC circuit, improved
« Reply #14,  »
Cycle,
do you run a real circuit plus simulation or is it simulation only?

Cycle

Re: Dave Lawton WFC circuit, improved
« Reply #15,  »Last edited by Cycle
Simulation only on this so far. If it seems to work in the simulation, I'll consider building it.

I figured out a way to determine whether the circuit was charging the battery or not. I put a 1 nF capacitor in parallel with the battery, with a switch on the positive side and a 1 mH inductor on the negative, then set the circuit run-speed to its lowest setting.

Then I opened the switches to the 555 timer and WFC parts of the circuit, and closed the switch between the battery and the 1 nF capacitor.

Then I ran the circuit and alternated between opening the switch to monitor switch differential voltage, and closing the switch to let the inductor 'spin up' and charge the cap. I got it to exactly 12 volts, accurate to the picovolt (0.00 pV differential across the switch when it's open).

Then I paused the circuit, and changed the capacitor size to 1 Farad. The capacitor voltage doesn't change when you do this.

The end result is that I now had a 1 F capacitor charged to exactly 12 volts, with only 59.1 usec of circuit run-time. It took a few tries to be able to do it that fast.

If I had tried to charge up a 1 F capacitor from 0 volts, it would have taken a large amount of circuit run-time, but since I want to test whether / how much the circuit is charging the battery, I had to minimize the amount of time taken getting that 1 F cap up to voltage. Hence this shortcut to getting the 1 F cap charged.

I then opened the switch between the battery and the (now) 1 F cap, closed the switches to the 555 timer and the WFC parts of the circuit, ran the circuit and set it to run faster.

The voltage drawdown from circuit initialization was 1.7 mV. Then the circuit settled down and started charging the capacitor at ~12.7 msec.

After 21.76 msec of run time, the 1F capacitor was back up to exactly 12 volts.
After 50 msec of run time, the 1F capacitor was at 12.00472 V.
After 100 msec of run time, the 1F capacitor was at 12.01323 V.
After 1 second of run time, the 1 F capacitor was at 12.17652 V (charge rate of 2.1340008 Wh).
After 2 seconds of run time, the 1 F capacitor was at 12.35501 V (charge rate of 2.1614994 Wh)

Here's the full-sized image.

{EDIT}
SqueezingSparks pointed out that the falstad circuit emulator doesn't take into account the resistance of inductors... I'll go through and alter the circuit, adding in resistors to emulate the resistance of each inductor.
{/EDIT]
Re: Dave Lawton WFC circuit, improved
« Reply #16,  »Last edited by Cycle
I'm using DigiKey to find inductors which are close to the current carrying capacity of what this circuit does, then I'll put a resistor on each side of each inductor, each resistor equal to 1/2 the measured resistance of the DigiKey inductor.

I used the following:
4 H inductor - 66 Ohms
2 H inductor - 58 Ohms
500 mH inductor - 30 Ohms
Re: Dave Lawton WFC circuit, improved
« Reply #17,  »
Huh, that didn't seem to affect it much at all. It's running on the battery right now, and it's doing pretty much what it did before the resistance was added to each inductor.

I'm going to set it up with the 1 F cap again and see how it runs.
Re: Dave Lawton WFC circuit, improved
« Reply #18,  »Last edited by Cycle
Ok, it's set up the same as before with a 1 F cap charged to exactly 12 volts.

The 555 timer kicked in at 11.96 msec.
Circuit initialization voltage drawdown was 1.69 mV
The circuit started charging the cap at 14.24 msec.
At 22.29 msec, the 1F cap was back to exactly 12 volts.
At 50 msec, the 1F cap was at 12.00903 V.
At 100 msec, the 1F cap was at 12.0332 V.
At 1 sec, 1F the cap was at 12.42026 V (5.1310008 Wh charge rate).

At 2 sec, the 1F cap was at 13.09187 V (6.8495004 Wh charge rate).

At 3 sec, the 1F cap was at 14.26095 V (9.8956668 Wh charge rate).

At 4 sec, the 1F cap was at 15.15531 V (10.7105004 Wh charge rate).

At 5 sec, the 1F cap was at 15.63223 V (10.03659984 Wh charge rate).

At 6 sec, the 1F cap was at 15.90103 V (9.0701664 Wh charge rate).


It looks as though the 1F cap voltage is starting to top out. It'll be interesting to see how the circuit behaves when all the energy must go into the WFC, rather than some of it going into the 1F cap.

Now, some of you may be asking, "But Cycle, those MOSFETs are going to blow with the voltages this thing is making! It's spiking up into the megavolt range when the MOSFETs open!"

True... and I've thought of a solution. A solenoid-driven variable-spacing spark gap.

Envision a quartz tube with spark gap electrodes on each end. The tube is evacuated to a high vacuum. One of the electrodes can slide upon a polished metal rod. Outside the tube is a ring magnet connected via connecting rod to a solenoid.

When the solenoid is energized, it slides the magnet back. The magnetic attraction to the sliding electrode causes the electrode to slide back, increasing the gap between the two electrodes. And as we know, a strong magnetic field in the vicinity of the spark gap will tend to quench the spark quickly, so the ring magnet serves a dual purpose.

When the solenoid is deenergized, a spring pushes the whole assembly forward, closing the spark gap distance.

The solenoid on and off times can be controlled by the same 555 timer circuit as is used now, through the same MOSFETs that are used now.

The way I envision it working is that the sliding electrode will be touching the other electrode most of the time, and will slide back to open the circuit momentarily, just as the MOSFETs are closed most of the time, and only open momentarily to create the pulse needed to make this system operate. Thus there shouldn't be much heat buildup.

It's the caveman version of a MOSFET. :D

Matt Watts

Re: Dave Lawton WFC circuit, improved
« Reply #19,  »
Hey Cycle, for grins try something...

Keep a backup of your circuit and build a new one where electrically it's identical but the components have been moved around.  I'm curious if falstad takes into any consideration the physical placement of the components.  In the real world I found certain situations where things on a schematic should be identical but on the bench, not so much.  The easiest thing that can be done is to take a small and larger value capacitor connected in parallel, try them soldered right together then connected to the circuit.  Then take them apart and separate them--put them on opposite ends of the circuit where other circuitry is now in between these two parallel capacitors.  The results you may get on the bench is much different.  I think it is attributable to non-conservative fields, but I'm not completely certain.  Could just be the longer wire lengths now begin to have inductance values seemly too small to have any effect, but in actuality they do.

Cycle

Re: Dave Lawton WFC circuit, improved
« Reply #20,  »
I noticed an interesting thing when playing with the circuit... changing the value of the resistance upstream of the 'electron extraction' spark gap also determines the minimum voltage of the 'electron extraction' part of the circuitry... I've changed it from 5K to 50K to see how the circuit acts. I'll post results here.

Apoc4lypse

Re: Dave Lawton WFC circuit, improved
« Reply #21,  »
This site is awesome for looking at Magnetic fields...

Cycle

Re: Dave Lawton WFC circuit, improved
« Reply #22,  »Last edited by Cycle
Ok, it still operates as expected, except the 6.3 nF 'electron extraction' capacitor bottoms out at 50 kV when that spark gap fires... so I'm trying it at 100 kV now.

{EDIT: It did not like the 'electron extraction' spark gap set with its minimum voltage at 100 kV via the upstream resistor... the circuit simulator crashed repeatedly. It does not like it when the ammeter hovers too long at zero or crosses it too often in a short period of time.}

Here is the circuit, and here is the code:
Code: [Select]
$ 13 1.0000000000000001e-7 13.654669808981877 28 10 66
165 672 256 688 256 2 12.074352393706535
v 1712 480 1712 384 0 0 40 12 0 0 0.5
370 1616 384 1616 336 1 0
c 1232 272 1072 272 0 6.302e-9 33222.0737592401
187 1232 288 1072 288 0 78.54 2940000 50000 1
r 800 432 800 480 0 15000
w 800 352 800 432 0
c 736 416 736 480 0 1e-8 8.045007272428377
w 736 480 800 480 0
w 672 368 672 384 0
w 640 368 576 368 0
w 576 368 560 368 0
d 640 320 640 368 1 0.805904783
d 560 368 560 320 1 0.805904783
174 560 320 608 320 0 10000 0.0347 Mark / Space
w 592 288 640 288 0
w 640 288 640 320 0
174 656 224 608 224 0 10000 0.5 Frequency
w 592 224 592 288 0
w 624 256 656 256 0
w 656 224 656 256 0
w 656 224 656 208 0
w 656 208 816 208 0
w 816 208 816 320 0
w 816 320 800 320 0
S 576 448 576 416 0 2 false 0 3
w 576 480 736 480 0
c 640 368 640 416 0 1e-8 6.617716657468988
w 560 416 512 416 0
w 560 368 512 368 0
c 576 368 576 416 0 1.5000000000000002e-8 0.0010000000138612464
c 512 368 512 416 0 2e-8 0.0009999999965435435
w 640 416 592 416 0
w 672 288 640 288 0
w 800 288 800 224 0
w 800 224 736 224 0
T 1120 240 1184 224 0 2 1 0.6328999303384721 -0.28298766831361155 0.999
c 656 -64 624 -64 0 0.00009999999999999999 12.074352393706535
g 624 -64 624 -48 0
g 576 448 560 448 0
w 624 320 640 320 0
w 672 352 672 368 0
w 672 368 640 368 0
l 704 -64 752 -64 0 0.5 -0.10174633392456178
w 656 -16 656 -48 0
d 656 -48 800 -48 1 0.805904783
w 800 -64 800 -48 0
w 656 -48 656 -64 0
c 736 -16 800 -16 0 1e-11 0.6769767572362149
w 800 -16 800 -48 0
w 736 48 736 224 0
w 832 432 800 432 0
w 1072 272 1072 288 0
w 1232 288 1232 272 0
w 1280 352 1136 352 0
w 800 480 1280 480 0
w 624 256 624 240 0
w 592 288 592 304 0
169 1136 -32 1136 32 0 4 1 0.9503349140927072 -0.28298766835817935 -0.6328999304832706 0.99
f 976 432 976 384 8 10 0.02
w 848 -32 848 240 0
d 1072 272 1072 240 1 0.805904783
d 832 -64 800 -64 1 0.805904783
l 1024 208 1024 160 0 2 0.6328999303380395
w 832 -64 1616 -64 0
d 1024 112 1024 80 1 0.805904783
d 1232 240 1232 272 1 0.805904783
d 1232 80 1232 112 1 0.805904783
d 800 320 800 352 1 0.805904783
f 1024 432 1024 384 8 10 0.02
f 1072 432 1072 384 8 10 0.02
f 1120 432 1120 384 8 10 0.02
w 832 432 976 432 0
w 976 432 1024 432 0
w 1024 432 1072 432 0
w 1072 432 1120 432 0
w 960 384 960 368 0
w 960 368 1008 368 0
w 1008 368 1008 384 0
w 1008 368 1056 368 0
w 1056 368 1056 384 0
w 1056 368 1104 368 0
w 1104 368 1104 384 0
w 992 384 992 352 0
w 992 352 1040 352 0
w 1040 352 1040 384 0
w 1040 352 1088 352 0
w 1088 352 1088 384 0
w 1088 352 1136 352 0
w 1136 352 1136 384 0
d 848 240 896 240 1 0.805904783
c 1104 80 1104 112 0 6.3e-9 41471.79489063148
g 1104 112 1104 128 0
d 1568 -32 1424 -32 1 0.805904783
f 784 48 832 48 9 1.5 0.02
w 832 -64 832 32 0
w 832 64 832 432 0
w 784 48 736 48 0
w 736 48 736 -16 0
l 1328 -32 1376 -32 0 2 -0.6040444173330151
w 1280 480 1616 480 0
l 1280 80 1280 128 0 4 -0.9203480939961453
d 1280 208 1280 176 1 0.805904783
d 1280 32 1280 -32 1 0.805904783
l 944 240 1008 240 0 2 0.3499122620178203
w 960 352 960 368 0
c 960 352 992 352 0 1e-12 -632390.23600912
w 1568 0 1568 32 0
z 1424 -32 1424 0 1 0.805904783 5.6
z 1424 32 1424 0 1 0.805904783 5.6
w 1568 0 1568 -32 0
w 960 352 848 352 0
w 848 352 848 272 0
d 1024 -32 992 -32 1 0.805904783
l 896 -32 944 -32 0 2 -0.9503349125633065
d 1280 -32 1216 -32 1 0.805904783
l 1472 32 1520 32 0 2 0.004200420223828433
d 1280 32 1424 32 1 0.805904783
d 848 240 848 272 1 0.805904783
c 1424 0 1568 0 0 0.000009999999999999999 4.98754333409263
w 1616 272 1616 -64 0
s 1616 272 1616 336 0 0 false
w 1616 336 1568 336 0
s 1568 336 1568 272 0 0 false
l 1280 256 1280 304 0 4 -0.9203480939961685
z 1280 352 1280 480 1 0.805904783 5.6
d 1104 80 1152 80 1 0.805904783
187 1216 16 1216 -32 0 1 10000000000 150000 1
r 1568 32 1520 32 0 29
r 1472 32 1424 32 0 29
r 1424 -32 1376 -32 0 29
r 1328 -32 1280 -32 0 29
w 576 448 576 480 0
r 1280 352 1280 304 0 33
r 1280 256 1280 208 0 33
r 1280 176 1280 128 0 33
r 1280 80 1280 32 0 33
r 1216 -32 1136 -32 0 33
r 1072 -32 1024 -32 0 33
r 992 -32 944 -32 0 29
r 896 -32 848 -32 0 29
r 1072 208 1024 208 0 29
r 896 240 944 240 0 29
r 1072 240 1008 240 0 29
r 1024 160 1024 112 0 29
r 1120 240 1072 240 0 29
r 1120 208 1072 208 0 29
r 1184 240 1232 240 0 29
r 1024 80 1024 32 0 33
w 1024 32 1072 32 0
w 1216 80 1216 16 0
w 1136 32 1232 32 0
r 1184 208 1232 208 0 29
r 1232 32 1232 80 0 33
w 1232 112 1232 208 0
r 800 -64 752 -64 0 15
r 704 -64 656 -64 0 15
w 656 -16 736 -16 0
z 1568 144 1568 32 1 0.805904783 5.6
r 1104 32 1104 80 0 17
s 1616 384 1712 384 0 1 false
l 1616 480 1712 480 0 0.01 0
c 1616 384 1616 480 0 1 11.998337832763589
p 1616 336 1712 336 1 0
w 1712 336 1712 384 0
z 1568 272 1568 144 1 0.805904783 5.6
r 1152 80 1216 80 0 100000
o 2 16 0 12545 0.0001 0.6453839464246314 0 2 2 3
o 3 16 0 12554 33214.11961424729 0.0001 1 2 3 3
o 4 16 1 12545 0.0001 0.0001 2 1 375.2305244046009
o 91 16 0 12546 41460.21038208088 0.0001 3 2 91 3
o 127 16 1 12546 0.0001 0.0001 4 1 149.02529806060724
o 163 16 0 12546 0.001662160939602586 0.0001 5 1

To Set Up Before Running:
1} When you click the link to the circuit above, it'll automatically start running. Click the 'Start / Stop' button, then click the 'Reset' button.

2} Open both of the switches going to the WFC part of the circuit and the 555 timer part of the circuit... in my browser, things slow way down at this point for some reason... let it sit for awhile. Once you notice that the colors change quickly when you hover over a part of the circuit, you're good to go.

3} Edit the 1F capacitor and set it to 1 pF.

4} Slide the 'Simulation Speed' slider all the way to the left, this slows the simulation down while you're getting the 1F (now 1 pF) cap up to 12 volts.

5} Click the 'Reset' button one more time. Since the circuit simulator is already stopped, this will reset it and start it running.

6} Close the switch between the battery and the 1F (now 1 pF) cap. You don't have to keep it closed long... just close it then open it, close it then open it, etc. Note the voltage on the voltmeter above the switch. Keep closing and opening the switch until that voltmeter gets down to 0.00 pV. Now the cap is at exactly 12 volts.

7} Click the 'Run / Stop' button to stop the circuit simulator.

8} Edit the 1F (now 1 pF) cap and set it to 1F again.

{This is the 'shortcut' to get the 1F cap charged to exactly 12 volts without taking a lot of circuit simulator run-time... this lets you gauge more accurately the cap charge rate. Taking a long time to charge that cap would skew the charge rate to the low side.}

9} Slide the 'Simulation Speed' slider up to about 3/4ths of the way to the right.

10} Ensure the switch between the battery and the 1F cap is open, and close the other two switches going to the WFC and 555 timer parts of the circuit.

11} Click the 'Run / Stop' button one more time to start it running.

As for the resistors on each side of each inductor, that's to simulate the resistance of the inductors... if you actually build this in meatspace, you can leave them out.