Fusor Forums

We're two university students from Denmark, in the process of collecting parts for our fusor, intended as a neutron source. We already have a spherical vacuum chamber (24 cm in diameter), and we're expecting to make the inner grid out of thick wolfram wires. Now, we've gotten access to a large depot of large high-voltage power supplies, so we're wondering if we can go too big?

We were thinking of a 80kV 40mA supply for our fusor, but there is no easy way of regulating the current. How does the current go in a fusor? Does it simply go to the maximum of the supply immediately, or does it reach some kind of equilibrium? Is it possible to calculate the current draw from the voltage, given the geometry and the materials of the fusor?

P.s. We do have professional help from people with experience in high voltages and radiation safety, but they've never worked with a fusor before.

Hi Pia,

The only thing I could add here would be that to make sure your power supplies are current limited. Most fusors that people have reported here on this board operate between 5 - 15mA current range when generating neutrons for a limited time before they become unstable and trip the current limiter or just stop producing neutrons! I watched Richard Hull work his fusor during a HEAS event, and he was all hands adjusting the gas and voltage to keep an even plasma in his unit to generate the neutrons.

The voltage varies widely by user. I've only heard of maybe 2 people on this board reliably operate higher than 50Kv, and that would be Carl and Jon.

If you do go for the 80Kv/40mA supply, it good to know that all that power has to go somewhere, and usually it means your fusor's will run hot and your inner grids will take a pounding!

My 2 cents!

In my operational experience the pressure control system becomes a pretty decent current limiter. You'll find that if you pull the pressure below plasma-level and very slowly increase the pressure, it will suddenly fire at a certain pressure but at that very pressure, the current will be quite low. You can then slowly increase pressure and the current will go up (and the voltage may drop if the supply can't keep up with the new current demand). This could be aided by the use of an appropriate ballast resistor or the use of an internal current limit control if the supply has one.

As for calculating all this, you could roughly based on paschen's curve, but there are so many variables that you can only really master the control art by operating the fusor and learning it's characteristics.

The plasma density acts much like a variable resiseter. At ~ 5 microns, if you can get any current flow at all, it will be small. At 30 microns the resistance might allow 20-50 mA (guess). At 100 microns, the plasma will conduct much better, At multiple hundreds of microns the plasma will come close to conducting whatever you feed it. The relationship is logrhymthic. A current limiter in a fancy power supply or a ballast resister is always a good idea. Also keep in mind that there can be surges in current as contaminants on the electrode are burned off. I think this is what damages a lot of the power supplies without good current limiting.

This begs the question - how does a plasma (at optimal density) compare in conductivity to copper wire or superconductors?

Dan Tibbets

Dan DT wrote:
> The plasma density acts much like a variable resiseter. At ~ 5 microns, if you can get any current flow at all, it will be small. At 30 microns the resistance might allow 20-50 mA (guess). At 100 microns, the plasma will conduct much better, At multiple hundreds of microns the plasma will come close to conducting whatever you feed it. ...
> This begs the question - how does a plasma (at optimal density) compare in conductivity to copper wire or superconductors?

[edit to begin with punch line]
The absolute conductivity of fusor plasma is not far from that of deionized water.
If this observation holds up to review, and others find it intriguing, it's probably been said here before.
-Rich
- - -
That's a fun one, Dan.

IIRC, forward voltages for gaseous conduction bottom out at around 10 volts in arc, as opposed to glow, discharges. In arc lamps and gas tube surge arresters the effective resistivity could be down to 1e-3 ohm-cm (conducting up to 1e5 amperes with voltage drop on the order of 10 or 100 volts).

Back in the world of amateur fusors, let's take a "high conductivity" reference plasma with 100 mA at 10 kV. If that is flowing between spherical grids of 2" and 6" diameter, the effective resistivity is about 5 megohm-cm. Spherical radial resistance is (1/r1-1/r2)/4pi * rho.

Standard resistivity for copper is 1.6e-6 ohm-cm -- it's hard for gassy media to compete with 1e23 carriers per cm^3. Between 2" and 6" diameter, the radial resistance would be 35 nanoohms.

What do you suppose would be the radial resistance if the same space were filled with superconductor?

Pia,

The first thing you have to look for is a power supply with negative output, that will eliminate most of them. Any PSU adjustable obver 50 KV would be perfect.

To protect the PSU and avoid disasters you need a good ballast resistor between the grid and the PSU.

Steven

A mixed bag of answers.....All are helpful

bottom line

1. Only negative voltage output supplies can be used (Ground or case is positive. Hot is negative)

2. The supply must have at least 20 ma capacity over its entire range. 50ma would be more than is normally needed. Working D-D fusors producing 2 million fusions per second ( 1 million neutrons/s) are readily run at 45kv and 12ma with 15 microns of flowing D2. This assumes a well executed fusor and a clean, pure gas environment.

3. The supply must be able to have its voltage smoothly varied over its full range from zero to max voltage.

4. It would be highly desireable to have current selection and or electronic current limiting. If this is not avaialble, you will need to select a suitable resistive ballast to protect the supply.

Richard Hull

Rich Feldman wrote: Back in the world of amateur fusors, let's take a "high conductivity" reference plasma with 100 mA at 10 kV. If that is flowing between spherical grids of 2" and 6" diameter, the effective resistivity is about 5 megohm-cm. Spherical radial resistance is (1/r1-1/r2)/4pi * rho.

Standard resistivity for copper is 1.6e-6 ohm-cm -- it's hard for gassy media to compete with 1e23 carriers per cm^3. Between 2" and 6" diameter, the radial resistance would be 35 nanoohms.


These are very interesting observations, but we're not sure what happens with the last number (the 35 nanoohms). That's for copper, right? With the gaseous configuration, the number should of the order of megohm, right?

As more than one of you said, we will try to look for some kind of current limiter - just to take care of our grid until we get to know the system. And yes, we are aware that we have to get negative polarity

And Richard, thanks for the little list - it's good to have some steady points to go back to. We are considering slicing down the voltage to get to a supply which can actually go all the way from zero and up to the maximum. Earlier, we considered a larger voltage, but then it might not be able to go all the way from zero.

Hello Richard,
Regarding neg. voltage power supplies, can't one just modify a standard pos. supply by reworking the rectfiers and grounds?
Jerry

If you have a positive supply, usually, a reversal of the diodes will turn it into a negative hot supply If you have electrolytic filters in the supply they would have to be reversed as well. As a rule, you should not have to alter the grounding within the supply in any way.

Richard Hull

Thanks Richard-With regard to feedback circuits for metering, are there any appreciable, or output operational differences between a feedback ckt. and a sampling ckt. for meters built into a factory power supply. Can one still have factory metering after reversing the polarity?
Jerry