Alternating Current Demo Fusor

[Mark Rowley]

Last week I spent an afternoon building a tiny demo fusor (2" diameter by 3" long). It is powered by a 9kV NST with full wave rectification. After pumping it down to about 15 microns, the usual star mode became apparent. At that point, I deceided to take out the rectifier and run it with AC. Suprisingly the poisser and star mode remained ! Thinking that my NST may have some unexpected internal rectification, I switched the input wires and achieved the same results.

When the rectifier is in place, it displays the expected plasma phenomena when the polarities are reversed.

Whats happening here...self rectification ?

I will post a photo soon.

Mark Rowley

[Wilfried Heil]

>Whats happening here...self rectification ?
Yes. The hollow cathode discharge acts as a rectifier. Good observation!

The Paschen product p*d is highest when the grid is the cathode, because the ions can oscillate and keep the discharge alive. This is not the case when the voltage is reversed and almost no current will flow then.

Because of this, the fusor has self-rectifying properties, within a certain voltage range.
It works well with AC up to a few kHz, but not with RF of several 100 kHz or more.

[Mark Rowley]

Thanks Wilfried,
I had an idea thremionic emission may have something to do with it. After reading your post I revisited Paschens Law (http://home.earthlink.net/~jimlux/hv/paschen.htm)and read the following:

Temperature dependence
Paschen's law ( V = f(pd)) should really be stated as V = f( Nd) where N is the density of gas molecules, which is, of course, affected by the temperature as well as the pressure of the gas ( n/V = p/RT). An empirical formula for air (considering it as an ideal gas) is:

x = 293 * p * d / (760 * T)
Vbreakdown = 24.22 * x + 6.08 * SQRT(x)

p = pressure in Torr (mm Hg),
d = distance in cm,
T = Temperature in Kelvins
V = breakdown in kV



In a neutron producing fusor, would operation under these conditions effect the output in any way ?

Mark Rowley

[Steven Sesselmann]

Mark,

Well done, you might have come across some significant effect here.

A fusor is really a simple diode, and when the grid is hot, current flows much easier from grid to shell than from shell to grid.

Someone out there should try running a standard fusor from a raw unrectified x-ray transformer, and report back what happens. Needless to say that the earth connection from the shell should best not go to your domestic water tap.

Steven

[Chris Bradley]

I think you've hit the basic issue there, Steven. There may also be an issue of asymmetric loading on the transformer, was the NST a kHz type or an ironed core mains type? I imagine those buck-boost inverter types will suffer considerable asymmetric compression effects - but there again, only once the grid has warmed up to thermionic-emitting levels, and if below that with an ironed core type then what dangerous, potential lethal, effects may there be with floating loads with respect to earth.

This was a good experiment with Mark's diminuitive fusor, with such a small grid and a floating inverter NST (?), but other set-ups will not be so forgiving. Connecting an x-ray, or ironed core type, may bring parts of the internals of the transformer to too high a level with respect to the other internal parts which normally do not see those potentials. I do not recommend such experiments, excepting where the experimenter is fully knowledgeable over all these things. (Gee - I do sound the nanny these days, but there's a number of new folks that may need to hear the warning shots.)

[Richard Hull]

A quick check will show that the hollow cathode diode rectifier was looked at back in the 1920's by Langmuir and Blodgett.

The fusor is a rectifier, though not a very effiicient one. There is no normal source of classic diode electron emission that would support high current in the device until the inner grid gets orange hot. This is why it works best with the large pressure gas load that supplies the ions and electrons, much in the sense of the mercury vapor rectifier.

Very old news, indeed, for us vacuum tube guys........Not all vacuum tubes have a deep vacuum. Gases and metal vapors have been pumped into cold cathode glass tubes for years to get all sorts of electonic jobs done. The GM tube is just one example we use all the time.

Richard Hull

[Jon Rosenstiel]

Brings to mind the 0Z4 full-wave gas rectifier tubes that were commonly used in automobile radio vibrator power supplies.

Jon R

[Doug Coulter]

60 Hz is so slow compared to the transit times of either electrons or ions that what's happening here doesn't need much explanation at all. It has plenty of time to get stable poisser on the right half cycle, and is a glow discharge on the tank walls the other half (which you didn't mention -- is that what you saw?).

Though some rectification does take place due to different transit times, no need for that here.

I've done tests on electron emission of W and Ti (which I use here for grids) and at "orange" I can't easily measure ANY above the noise. For W, think yellow to white, Ti melts before any real emission.
This was for an electron source for an ion gage tube I was making. I wound up with a coated W filament from a 6v halogen bulb I cut the top off of on my lathe with a diamond grinder.
I don't have a source of tantalum to try with just now.

I use Y2O3 for coating things I want to have emit, it withstands atmosphere exposure far better than Ba/Sa which I've also tried -- with the latter you get about 2-3 times before it won't reactivate. At first you put BaCO3 on there (and the other earths), then activate to Ba and BaO (and the other compounds from Sr and Ca). But exposure to air makes them into hydroxides, which won't reactivate like the carbonate would -- you get 2-3 times at most. No such problem with Y2O3.

For what it's worth, I've measured transit times for ions at the speeds and feeds we use. It's in the 10's of mHz even at this large scale due to the slow ions and high voltages. Electrons would be a lot faster at the tens of kV input range -- roughly the sq root of the mass ratio is the scale factor due to E = mv^2. ( I say large scale because fusor spacings are much larger than in electron tubes).

The neon sign guys did try flyback solid state power supplies for awhile. They ditched them fast, as all the tubes burned out at one end in early life. The very slow Hg ions didn't have time to traverse the long tube during the short flyback pulse, so all piled up at one end. Most modern neon is just Hg (plus Ar) and phosphors, sadly. They went right back to good old 60 Hz iron, at a cost increase, because the asymmetric waveform just didn't work out at those frequencies, in the 10's of kHz, needed to make the replacement ferrite work out and be cheaper than the old iron.

[Wilfried Heil]

Let's not miss the important consequence of what Mark has observed and brought up:

>>> There is no need for a HV rectifier with a fusor. The fusor already is a rectifier!
This happens because of the hollow cathode gas discharge mode only. We do not run the grid so hot that it emits electrons thermally, which would be very undesirable as this current is a total loss for fusion. Nor is there any significant amount of field emission (Fowler-Nordheim tunneling) from the rather thick grid wires.

The shell of the fusor does not light up during the positive half cycle (on the grid) with low frequency AC and at the low pressure of "star mode". It will light up with RF however, where this method no longer works.

[Chris Bradley]

Wilfried Heil wrote:
> This happens because of the hollow cathode gas discharge mode only. We do not run the grid so hot that it emits electrons thermally, which would be very undesirable as this current is a total loss for fusion.
Sorry, I don't follow you. A gas discharge *itself* cannot rectify, it needs a preferential direction for one of the charge carriers to travel. The electron path *is [almost all] the* total loss for the fusor. I would not exclude other mechanisms that promote the ions to stay in the centre and the electrons to preferentially go, but electrons *must* come out of the grid one way or the other whilst there are ions around. In a DC mode, for each ion plodding its way to the cathode, 60-or-so electrons will have made it from the cathode to the anode, but will the reverse not also occur if the gas discharge is reversed and there are no other effects to create asymmetry of current?

[Wilfried Heil]

The gas in the hollow cathode does the rectification.

When the ions can recirculate a number of times, this increases the length of their path through the gas and the amount of secondary ionisation. Given enough voltage and gas pressure, this can sustain a discharge. When the hollow part becomes the anode, during the other half of the AC cycle, the ions traverse it only once and the discharge will extinguish. If the pressure is too high, then the gas will conduct in both directions.

The Paschen product is p*d*n in the conducting direction and p*d in the other.

[Richard Hull]

Field emission electrons will preferentially occur off the sharp radius grid wires where the field is extremely high, but not off the anode shell where the large radius smooth surface is located. Simple electrostatics.

Richard Hull

[John Futter]

Wilfried is right
The rectifying effect of high voltage AC in the presence of ions is how many cheaper burner controllers work (natural gas or fuel oil) to detect the presence of flame. Better controllers use a UV cell to detect flame. A flame will give some ions, evidence of which can be seen by putting a strontium rich source into a candle flame, some ions are produced making the pronounced red colour
Most expensive controllers for boilers give both options of flame rods or UV cell.

the fancier types of controllers use a DC source for the flame ionisation.

FWIW

[Todd Massure]

I agree with Richard. Not only are the grid wires a much tighter radius, but the inside of the shell is convex too, and any excess charge will reside on the outside of the shell, not the inside. There would be a much greater electric field on the wire causing much more emission.
Not to get off topic, but I believe this is also more or less the process responsible for the directional ion wind in "lifters".

Todd