Archived - Van de Graaff Fusion

[Frank Sanns]

There have been pros and cons of low current fusion but since I had a medium sized Van de Graaff generator I gave it a go.

The first picture is a photo of the VDG and my fusor.

The second picture is the half inch (1.25 cm) spark gap between the VDG and an inner grid electrode.

Last picuture is an electron beam at a pressure just before star mode. These are visible in the fusor albiet faint compared to a multi milliamp run, but still clearly visible. Small amounts of fluorescence of the pyrex bowls and the aluminum grid was faintly present. I will hold back quoting any neutron numbers since there is a chance that some noise from pulsing could affect the already paultry neutron numbers that I appeared to have been detecting.

Frank Sanns

[Jon Rosenstiel]

Way cool experiment Frank, you da’ man!

I have no experience with a Van de Graaff, but I’m guessing that the spark across your half-inch gap was continuous… is that correct?

Jon R

[Mike Beauford]

I've got to second that. This is pretty freaking neat.

[Frank Sanns]

It is high rep rate but not continuous. From the sounds and sights I would have guessed around 10 hz to 15 hz but I figured I had the evidence in the photo so below is photo number two from my first post blown up.

To my eye, I see four sparks. The shutter speed was 1/3 of a second so that is 12 hz. It was just one photo so the uncertainty is high for any one measurment but coupled with my observations I feel that the 10 hz -15 hz is a good number.

Frank Sanns

[John Futter]

Frank
I presume you are spaying electrons onto the belt to get negative charge on the terminal



if so ----Loose the spark gap and rely on the insulator to fusor for the breakdown ---throttle back the D2 gas until the insulator breaks down occasionly and re measure the neut output. Couch measurement indicates that your present voltage is 50 to 75 kV depending on the gap 3kV per mm ( my guess 20mm).

we have been doing a D2 run @ work today @ 1.4MeV for NRA measurement on light element contamination in a sample, rem ball 3.5 metres away from beam tube indicated 0.12 rem /hr @ 25nA beam current.



Volts are your friend here and at very small currents I suspect you will not see anything in the viewport

FWIW !!

[Steven Sesselmann]

Frank,

This is a way to fusion which is more affordable and safer, and puts it within the acceptable limits of high school experiments.

Well done..

Steven

[Frank Sanns]

John,

You of course are correct. Much higher voltages is one of the big advantages of the VDG as well as the constant current when in continuous contact with the fusor terminal. I actually tried to operate that way for the reasons that you stated but my results were ZERO detectable neutrons.

I decided to use my handy dandy built in biological electrostatic volt meter to diagnose the problem. I used my hand to see how far the spark would jump at various spark gaps from zero to beyond an inch. At this point I HAVE to mention for the noob that this technique is NOT to be used with any other source of fusor power or would result in instant death to the one touching the HV terminal. Ideally, I should have used a small grounded sphere to determine the voltage but it was getting late and I needed data.

What I found was the lowest voltage on the terminal was when it was in contact with the VDG sphere. There now are many possibilities here but I think the most likely is leakage current of the entire system. HVDC can induce charges in other elements of the fusor and can distribute and disperse everywhere including ionization of of the air at the sharp radiuses on the terminal of the feedthrough. I think this is the cause for the low voltage when no spark gap is present.

When the spark gap IS present, the charge builds up on the VDG sphere and then discharges in a pulse of consolodated energy. Momentarily, the leakage of the system is overcome with the energy dump and results in momentary plasma and fusion. A larger or more capacitive VDG sphere/shape should help for pulsing. For coninuous HVDC, a higher VDG current (i.e. wider belt, spray more electrons, or run the motor faster) should help things. So would a sphere (as Richard uses on his fusor) on the HV feedthrough terminal.

Frank Sanns

[Frank Sanns]

Steven,

Definately a safer way to go. Neutron counts are very low though and detection for the noob may be an issue.

We do want to keep our noobs so they can become OFs (old farts) one day!

Frank Sanns

[myID]

Hi-

had no time for calculation on this so just a quick thought:
Might be fusion happens- I guess for a very short moment (arc over- if you want to call this an arc)- the Voltage is high enough and you have a "high" current.
(Mostly depending on size of sphere on top of generator)
I think the "constant current" is much to low to sustain a discharge...(and goes well with your outcome of the"low voltage" when directly connected)
I guess what you interpret as "several 100 N/sec" could be noise from the spark?
Did you make a run without D2 to compare the outcome?
As Richard always says: bubbles would be nice to proof but I guess your "flux" is a little low here;-)

Independent of this- cool experiment and a nice fresh approach!

Greets
Roman

[Richard Hull]

To do this thing right, the insulator needs to support the full VDG voltage. This is not a criterion that is met. Ideally, a larger VDG with a larger electrode and a short arc to the larger Fusor electrode could develop the voltage better. The salient point here is the ARC DROP in the fusor. This is an old vacuum tube term applied to thyratrons back in the day. (my day).

We see that in a working neutron producing fusor, this arc drop can be very high voltage indeed, but requires a delicately adjusted gas pressure of flowing D2.

With a VDG you can't find this point easily and would have to "poke around" in the dark on the optimum pressure.

I would have to count the neutrons myself before I would believe a VDG would be a good device for the fusor. The volts sound nice, but if there is any fusion only plus ultra (read expensive gear) would detect it. What you would save on the power supply you would have to spend X5 on the detector to affirm absolutely to a hyper critical audience that fusion has occurred. Even then, there would be little of it.

Richard Hull

[Doug Coulter]

Frank,

You know my theory about space charge repulsion preventing fusion at the supposed focus, something I'm working on right now -- we built a pinhole camera for X rays and charged particles (the x ray one without the Be foil in front of the pinhole to stop the charged particles). In your experiment, you possibly win in two ways here -- one is the fast risetime, getting a bunch of stuff going before there's a lot of ions in the grid center, and the other is just generally running lower currents, because in my theory, most of the higher currents are wasted.

I'll be posting my pinhole camera pix real soon, it's been a challenge to get pictures digital anywhere near as well as you can see with the naked eye, but what I'm seeing is beginning to bear my theory out. We're seeing most charged particles from the exact center of the cylindrical grid I'm using (looking end on at a hole there), but most of the really hot X rays from a ring around that (still inside the grid), which is kind of interesting. We can add a little E field to displace one image from the other and see both at the same time, which is cool.

BTW, this camera was really easy to build, and we put it on the end of a wiggle stick to look around which is nice too. It's just a piece of 2" copper pipe with a piece of lead soldered over the input side, a 40 mil pinhole in the lead, and a 2" screen of ZnS:Ag on plexiglass (don't use this, use real glass! Ours got a little melted).

Note -- you can build a VDG for much higher currents, it's just cubic charge carrying belt/pellets per second that determines that, so a faster belt or wider, or both, and pretty soon you've got real power going on. If you can find belt material that doesn't have static charge control conductive stuff added these days, that can be a little bit of a challenge, I'm still casting about for the good stuff here (I have the same VDG you do, or its twin -- not much current at all). I'd bet you do better if you have ions in there already from an ion source, btw. Once going, the fusor is a pretty good ion source of its own, but...for pulse operation that doesn't work out very well at all, by the time it becomes a good ionizer, the pulse should have been over for quite awhile. Running at a better ion/neutral ratio should boost results in this setup.

While bigger can be better, I'd like to see more things like this that potentially make more fusion per joule of input -- then (and only then) worry about scaling things up.

[Frank Sanns]

I have removed all claims of neutrons from my post pending other routes of verification. My setup can detect down to single neutron/sec levels and I doubt the pulsing at such small energies could be affecting the detector when it is well behaved at high power contiuous running but still, my numbers are retracted for now.

Frank Sanns

[Chris Bradley]

Good work, Frank. It is the do-ing of it here that counts, I think, and seeing what is what.

Just to add on the neutron-counting front, though, the simple thin-target yield calculation might be usefully kept in mind;

viewtopic.php?f=11&t=4657#p30233

Going from, say, 50kV drive voltage to 200kV drive voltage 'only' gives you a single order of magnitude improvement. That is, just for example, if you're down to single uA at 200kV compared with multi mA at 50kV then you will be 3 orders of mag lower in neutron count in the high voltage/low current stakes, according to the basic maths. This graph suggests once you're above around the 70kV mark, you're just as well off going for current rather than voltage, for a given power.

None of the maths takes anything away from your experiment, though.

[Doug Coulter]

Chris,
It's kind of hard to tell from your link what you're using for assumptions on your maths.
Since you are talking center mass, it would seem you're calculating fast on stopped (probably a good approximation for some cases, perhaps the more common ones, as my measurements seem to indicate), but it also seems to assume that the applied field is the NET field the ions see, which is definitely wrong almost no matter what, in a fusor, with no target connected to a power supply to allow all ions to become neutral as soon as they hit, and thus not repel further incoming ones via their own field. That is, if you don't consider the grid wires the target, which certainly is part of what's going on, but hopefully not the main show. If we thought that, we'd go away from fusor design and be trying to improve borehole tubes where at least nothing misses the target.

Looking at DD cross section on a log-log plot for fast on (stopped) target, indeed you're most of the way there at 70-125kv or so, the curve begins to bend over in that region and higher volts no longer gives you more fusions per joule input. Which is why the borehole guys run right in that range. Obviously, they can read. And they are doing this on battery power deep in the ground...

On the other hand, you may need more applied volts to get a certain net particle energy, due to the other ions repelling the incoming ones. I'd have to say that at this point, in the absence of actual measurements of speeds, or even indirect ones that could be done, which no one here has that I'm aware of, we can't go by this -- the math may be correct in some situation, and probably is. The thing is, is it correct in *this* situation, which I frankly doubt, and am currently taking measurements to find out one way or the other. Nothing beats a guess like a real measurement, and barring high water and other emergencies, I'll have something to report that I can back up in a month or two.

Preliminary results *suggest* that indeed, the space charge gives us net energies far below the applied field would suggest we'd like to assume we have. So at fairly high volts input, we're still not anywhere near the particle energies implied by those applied voltages, and still working on the straight part of the cross section curve at a fraction of the applied voltage.

Which may explain why I can get fusion beam on target at much lower voltages (roughly factor of 4) than in any fusor design I've yet tried (about 5-6 of them). No space charge repulsion, other than beam spreading in the beam on target situation, as every ion that hits the target becomes a neutral right off, where in a fusor, they tend to collect in the center and repel incoming ions.

So, Frank may be trying something smarter than we all know!

[Richard Hull]

I have no doubt that some fusion is probably occuring in a van de graff experiment. To measure it and even quantify it is a job for top flight instrumentation at the lowest levels. I fear a bubble detector would take hours to accumulate one bubble over background with a simple van de graff setup.

Barring a suitable data collection by a large He3 of BF3 detector, I would look twice at anything but a bubble detector for claiming any neutron production.

I have always felt that pulsed fusion in a fusor might be very interesting and that a pre-ionized system that might be fusing at a low level might be able to be "pulled up" to super pulsed fusion levels with a Hydrogen thyratron of proper scale and proportion dumping stored inductive or capacitive energy into the fusor.

Frank's initial VDG image and setup is definitely an attempt at low level, pulsed fusion, albeit at virtually undetectable levels. Fusion, as we know, is a bulk, probablistic thing.

Probabalistic through voltage and bulk through current and fuel pressure levels.

Richard Hull

[Chris Bradley]

Doug Coulter wrote:
> Chris,
> It's kind of hard to tell from your link what you're using for assumptions on your maths.
The calc is a standard thin target yield, based on; [the target density]x[velocity of projectile]x[reaction cross-section]. I do not need to include the projectile effective density as it is just a relative calculation assuming that effective density is the same between the projectile energies.

> Since you are talking center mass, it would seem you're calculating fast on stopped (probably a good approximation for some cases, perhaps the more common ones, as my measurements seem to indicate), but it also seems to assume that the applied field is the NET field the ions see, which is definitely wrong almost no matter what, in a fusor,
I'm sorry, Doug, you've lost me on this bit. The 'voltage' is simply twice the particle energy. Sure, if a drive voltage of 50kV only gives you 20kV particles, then you'd need to look at '20kV' on the plot. I'm just assuming reactions are at the full drive voltage. If you're saying that there are also reactions when the particle is running at less than 50kV, sure, there is a smurge of possible reactions, I'm just comparing reaction rates at the particles' peak velocities. I think that's quite reasonable for *comparative* and normalised purposes (not for an absolute measurement of reaction rates, though) and would only be in error by significantly less than one order on account of the way the reaction rate is exponential to particle energy.


> On the other hand, you may need more applied volts to get a certain net particle energy, due to the other ions repelling the incoming ones.
Depends on the specifics. If it is a reaction with a 'general background medium' then it doesn't matter if they're deflected, so long as they don't loose too much energy in the deflection. (Is that what you were meaning?)

> Preliminary results *suggest* that indeed, the space charge gives us net energies far below the applied field would suggest we'd like to assume we have. So at fairly high volts input, we're still not anywhere near the particle energies implied by those applied voltages, and still working on the straight part of the cross section curve at a fraction of the applied voltage.
I'm very much with Frank on the issue of current. One reason for wanting volts over amps is because as you crank in more amps, reaction rate starts tailing off. On this matter, of a limiting current to operate up to then you should look to volts to go further, I'm in agreement.

[Frank Sanns]

I now have scaled up my VDG operation to a much larger scale. The system that I now have uses a 41 cm (16 inch) sphere with a 7.6 cm (3 inch) wide belt. The arc will easily jump 35 cm (14 inches) in air which puts it in excess of 400 KV. Current flow is measured to be 0.08 ma for a grand maximum power output of 32 watts. In reality it will be somewhat less than that with losses that will come with the higher voltages bleading off energy to the environment.

Using the same setup as my smaller VDG (i.e 50 kv -75 kv) there is a nice glow in the chamber and pressure and D2 flow can be easily adjusted even by eye. No comments on neutrons right yet as I am working on ways to be sure my electronic counting is not picking up RF from the pulses.

For now, I am working on a way to push my HV feedthrough capacity up to the hundreds of KV to take full advantage of this current VDG. A dielectic oil emersion system, encapsulating of my current feedthrough, or a home brew insulator seems to be my best options.

I will post more as this iteration continues.

Frank Sanns