A dynamic fusor
Posted: Tue Mar 04, 2014 4:42 pm
Hopefully, this "speculation" is not utterly unfounded. Along the way - many hours of runs and fun in the lab, we've noticed a few things.
Some of that is due to our unique lashup, and sometimes, our failings, but we try to keep obvservant no matter what.
We are running our main fusor grid, cylindrical, usually about 1.5" long and almost 1" diameter, in a 6" by 6" sidearm of a much larger tank, about 14" by 26" ID. While we've tried a number of ion sources (mostly documented here) the easiest one is simply another 2 loop grid out in the main tank, off to the side. There are some videos here of us switching the main grid, with 50kv on it - on and off with a mere 10kv or less on that "ion generator" grid. If we push a little into the little guy, the big guy fires off, we turn off the little guy, and the big guy goes off too. In fact, some coarse measurements show that as a "vacuum tube" or more correctly, plasma triode, we've created a PNP tube (polarity speaking) with a power gain of something over 100, and a current gain of at least 5-10. Hmmm.
We also have had a number of faraday probes (really just rods or wires) in the big part of the tank, and when things are going, they always tell us that we have an enormous negative net charge imbalance (and this conflicts with a lot of theory posing as fact in a lot of literature, but there it is).
We've even measured the time it took for whatever negative it was - probably electrons, probably also some D- or D2- - to reach the faraday probe, one time using the secondary ion source grid itself as the probe (important for what follows) and the propagation time appears to be in the 10 us range. That's aweful slow given the voltages involved...either this stuff isn't getting full speed of the voltage, or what we're really seeing is something heavy and negative - as in charge-exchanged D- species.
An apparently unrelated observation is that we see highest Q at the lower gas pressures (which is borne out by the theories of space-charge defocusing we've known since the early days of electron tubes). We also see pulses of the very highest Q - 500-1000 times more than normal, during onsets, not in steady state, and it would be better had I not edited out "outliers" that were 100 times more Q than what I left in the data set, being conservative and all.
Let's put that together for a minute. Highest Q when it's hard to start, and only at the start. Low gas pressure. An incidental "valve" or tube with a crap-ton of power gain. What does that suggest?
Well, to me, it means I should be either driving this, or making an oscillator - we don't know which is best, and the math is kinda like fractal to figure out, so we are more or less reduced to just trying stuff. Could be that the idea driving waveform will be the opposite of what we get as an oscillator if we should make one - but we just don't know. And the idea that we could do this as an oscillator is just to elegant to ignore.
So, I'm making an oscillator. I will report in the proper place here when I run it, but of course I hope to be reporting far higher Q than normal, and various expected and unexpected results - that's the very nature of science. Why am I saying this first, then testing later? Well, good science predicts, and takes its lumps depending on whether the predictions come out true or not.
I therefore predict that an oscillator, or a driven system, will have far higher Q than a steady state fusor, which to me has appeared as what a math guy would call a "strange attractor" for dynamics that are not best, but worst case for actual fusion. We will know soon - why guess?
I just finished winding the secondary coil for the transformer. The primary is in series with the DC feed to the main grid. It has about a 3-4::1 stepdown (turns ratio anyway) and even though fine-pitch, may not go low enough in frequency to "go the right speed" - after all, we DO certainly have more than one species here - and we don't even know which one we want to be driving the oscillation! But we'll know...that's what lab time is for.
Since a picture is worth a lot of words and so on...
Obviously I have a little more to do before this sees power. The bigger pipe is sched 80, for insulation. There is an 8" diameter wire-screened outer pipe not shown - this whole mess is more or less the center conductor of a large piece of coax, to keep EMI out of my other stuff - it can get pretty fierce otherwise. The secondary here will be connected in series with the little supply (and ballast) that normally drives the secondary grid. Both supplies have a lot of series R, for safety. We have so much power gain, I'm not going to worry about the losses there for now, it almost oscillates by itself anyway, from coupling inside the tank, but at some wierd, very high frequency (working more like a regular electron tube it seems). This should put it more on a basis of us choosing the frequency and which particles are sloshing back and forth how far and fast.
Some of that is due to our unique lashup, and sometimes, our failings, but we try to keep obvservant no matter what.
We are running our main fusor grid, cylindrical, usually about 1.5" long and almost 1" diameter, in a 6" by 6" sidearm of a much larger tank, about 14" by 26" ID. While we've tried a number of ion sources (mostly documented here) the easiest one is simply another 2 loop grid out in the main tank, off to the side. There are some videos here of us switching the main grid, with 50kv on it - on and off with a mere 10kv or less on that "ion generator" grid. If we push a little into the little guy, the big guy fires off, we turn off the little guy, and the big guy goes off too. In fact, some coarse measurements show that as a "vacuum tube" or more correctly, plasma triode, we've created a PNP tube (polarity speaking) with a power gain of something over 100, and a current gain of at least 5-10. Hmmm.
We also have had a number of faraday probes (really just rods or wires) in the big part of the tank, and when things are going, they always tell us that we have an enormous negative net charge imbalance (and this conflicts with a lot of theory posing as fact in a lot of literature, but there it is).
We've even measured the time it took for whatever negative it was - probably electrons, probably also some D- or D2- - to reach the faraday probe, one time using the secondary ion source grid itself as the probe (important for what follows) and the propagation time appears to be in the 10 us range. That's aweful slow given the voltages involved...either this stuff isn't getting full speed of the voltage, or what we're really seeing is something heavy and negative - as in charge-exchanged D- species.
An apparently unrelated observation is that we see highest Q at the lower gas pressures (which is borne out by the theories of space-charge defocusing we've known since the early days of electron tubes). We also see pulses of the very highest Q - 500-1000 times more than normal, during onsets, not in steady state, and it would be better had I not edited out "outliers" that were 100 times more Q than what I left in the data set, being conservative and all.
Let's put that together for a minute. Highest Q when it's hard to start, and only at the start. Low gas pressure. An incidental "valve" or tube with a crap-ton of power gain. What does that suggest?
Well, to me, it means I should be either driving this, or making an oscillator - we don't know which is best, and the math is kinda like fractal to figure out, so we are more or less reduced to just trying stuff. Could be that the idea driving waveform will be the opposite of what we get as an oscillator if we should make one - but we just don't know. And the idea that we could do this as an oscillator is just to elegant to ignore.
So, I'm making an oscillator. I will report in the proper place here when I run it, but of course I hope to be reporting far higher Q than normal, and various expected and unexpected results - that's the very nature of science. Why am I saying this first, then testing later? Well, good science predicts, and takes its lumps depending on whether the predictions come out true or not.
I therefore predict that an oscillator, or a driven system, will have far higher Q than a steady state fusor, which to me has appeared as what a math guy would call a "strange attractor" for dynamics that are not best, but worst case for actual fusion. We will know soon - why guess?
I just finished winding the secondary coil for the transformer. The primary is in series with the DC feed to the main grid. It has about a 3-4::1 stepdown (turns ratio anyway) and even though fine-pitch, may not go low enough in frequency to "go the right speed" - after all, we DO certainly have more than one species here - and we don't even know which one we want to be driving the oscillation! But we'll know...that's what lab time is for.
Since a picture is worth a lot of words and so on...
Obviously I have a little more to do before this sees power. The bigger pipe is sched 80, for insulation. There is an 8" diameter wire-screened outer pipe not shown - this whole mess is more or less the center conductor of a large piece of coax, to keep EMI out of my other stuff - it can get pretty fierce otherwise. The secondary here will be connected in series with the little supply (and ballast) that normally drives the secondary grid. Both supplies have a lot of series R, for safety. We have so much power gain, I'm not going to worry about the losses there for now, it almost oscillates by itself anyway, from coupling inside the tank, but at some wierd, very high frequency (working more like a regular electron tube it seems). This should put it more on a basis of us choosing the frequency and which particles are sloshing back and forth how far and fast.