Oh haha , I get it now -- The bell/cup thing. I had to tilt my head to the side.
Well depends on where you want me to start. I guess I have a tendency to just assume things without explaining...
----
Back in the 1970s and 1980s, a man from Ohio named Stanley Meyer invented a way to split water using very little input energy. He was able to make ridiculous amounts of hydrogen and oxygen gas using only milliamps of input. To put this in perspective, using conventional theory, what he was doing would have required somewhere upwards of 40,000amps. So to do it with half an amp was nothing short of a breakthrough.
Now Stanley applied for patents on this research, and it was all confiscated by the government and classified for awhile under national security. The rest of the story is predictable. Information suppressed, legal troubles, ending in murder. And of course all his equipment was stolen.
Now in 2008, with oil drifting north of $125/barrel, we are left picking up the pieces trying to figure out how the hell Mr. Stanley Meyer did it.
Conventional electrolysis is based around the Faraday's law. Basically, the voltage you send through the water doesn't matter so much -- what matters is the amps. Most all of the HHO cells on youtube use this method. Lots and lots of amps. Low voltage / high amps is how most people do it -- and its what the textbooks say.
Stanley's method used the opposite -- high voltage/low amperage. It also used high frequency pulses. DC pulses. Sound familiar? The story is almost exactly the same as the TPU. How ironic. In any case, this brings us to the graphs. Stanley's patents included a circuit called the VIC, or Voltage Intensifier Circuit. The VIC part of the design seems somewhat important.
A person from another message board (waterfuelcell.org) has posted his replication of some of the details of the VIC circuit. I reposted his graphs. This shows us two waveforms -- one on top and one on bottom. The two waves are equal and opposite, which is why they make the cool pattern.
(http://img80.imageshack.us/img80/5343/dsc09224885dk3.jpg)
You mentioned that it looks like a cup and a bell going to infinity. This is actually very telling, because it shows us something -- symmetry. The two coils are equal and opposite, down to the smallest detail. They are mirror reflections of one another. The axis of symmetry is the x-axis, running from left to right across the scope traces. What does this tell us? Just that this person's coils are very well made, and they are working almost perfectly symmetrical (from what I can tell).
On that scope trace, the x-axis is 'time' and the y-axis is 'volts'. So basically it is just showing you the two waveforms. Incidently, that spike you mentioned is actually significant, because it shows you how long one cycle of the wave will take. You can actually calculate the wavelength yourself.
Look at the first graph. Count how many boxes go from the top of one peak to the other (left to right). I'm serious, you should try this.
It is approximately 6.5 boxes. We know each box is 50 microseconds (us). How? Look up at the top of the screen, it tells us.
Now since we know each box is 50us, we know the length of the wave... 6.5boxes x 50us = 325us long. So one wave lasts 325us. Not bad, that's a useful piece of information. From that, we can calculate the wave frequency. How? Pretty easy. Just divide into one-million, since there's a million microseconds in a second. Do 1,000,000 / 325us = 3,077. So that means the frequency is 3077hz , or about 3.1khz (kilohertz), rounding up.
If you look at the graph you can see that's correct. The graph says 3.18khz, and we said 3.10khz. You just learned how to read a scope trace..
Can we split water with 0.5 amps yet? Nope, but between this new PDF, and the research that this guy over at waterfuelcell.org is doing, we are on the right track.