FAQ - Plasma cleaning & start up of long dormant systems

[Richard Hull]

Long dormant amateur fusor systems and brand new "startup systems" can benefit from the following procedure which I have used for years. Plama cleaning is part of this effort.

When first running after a long period of dormancy or for that important first time evacuuation, I turn on the forepump, open the gas ballast wide open and wait until the foreline is near 15 microns, which is close to my forepumps bottom limit. (This normally takes about 20 minutes). This rids the forepump of most water in the oil. I then close the gas ballast and open all valves right up to and including the fusor. By now the forepump is quite warm if not hot. This is normal.

I then start a hard, hot plasma at about 50 microns. (Note: opening all the other valves causes the foreline TC gauge to rise to high pressures and the process starts anew for the diff pump and fusor) I force the plasma to draw about 20-30ma regardless of voltage. (To plasma clean you must have a "beefy" supply to allow current like this) I up the voltage as needed to maintain at least 20ma until the foreline TC gauge and the fusor chamber Barocell agree in the range of 10-15 microns which is the norm for a long running, bottom to my forepump. By this time, the fusor chamber is almost too hot to touch. (plasma dissapation and electron bombardment of the shell).

Note: In most systems, you rarely ever reach 10kv in plamsa cleaning mode as the pressures are rather high, 15-30 micorns. Often, when first starting to glow clean, 20 ma is drawn at only 2-3kv. You are, in effect, running a demo fusor at this point. Watch out for grid melting. Run at a point where your grid is not glowing more than a dull red, at most!

I next close the fusor chamber valve and start the diff pump heater and cooling fan. I have a temperature gauge on the boiler and as the temperature rises so does the foreline pressure due to water vapor and other volatiles leaving the diff pump oil. (This can be as high as 20+ microns!) As the temperature approaches the full operating temp of the oil to start through the jets, but before any significant pressure drop occurs below the normal foreline "hot" bottoming of about 12 microns, I quickly open the fusor chamber valve. Most of the time I open the valve just before the boiler hits 100 deg C., as my diff pump drops the 12 micron chamber pressure like a stone at about 110 deg. C. As this "jet functioning" temperature is reached, it is natural for all plasma to cease as it can't be maintained at any voltage. IMPORTANT!...Turn all power to the fusor off at this time.

This warrants that most of the water vapor is sent out of the system and that the water vapor and light volatiles don't back flow into the fusor.

There is an art to not fouling a long dormant fusor or new fusor and plasma cleaning is just part of it.

Richard Hull

[Bob Reite]

Interesting. When I first built my system, I was thinking of various ways to bake it out and decided that the easiest way was just to run a plasma at a relatively high pressure and current, get it good and hot, then shut down the HV supply and open the exhaust valve all the way so that the turbo could remove any outgassing from the chamber.

Watch out for grid melting. Run at a point where your grid is not glowing more than a dull red, at most!

This is how I killed my first grid, by getting too 'piggy' even though it was made of tungsten. With a tungsten grid, I can safely go to orange red, but I now limit the grid dissipation to 400 watts and current to 25 mA.

[Richard Hull]

We are all responsible for saving our grids. It is very important to monitor your actual fusor current. In this fashion you know when you are getting close to a meltdown and can avoid it.

Richard Hull

[Frank Sanns]

There are some other questions coming up on the site about starting and operating pressures so I am adding this addendum.

Starting up a chamber for the first time or after opening or dormancy is a task. There are both water and gas molecules weakly bonded and trapped on and within the chamber surfaces.

As stated in the FAQ, pumping for a while and plasma cleaning is certainly a fast way to get started. It is however not the final solution.

No amount of reasonable pumping and plasma cleaning can really purge the chamber sufficiently for a good neutron run on the first day of startup. Even at full vacuum pumps at 10^-9 torr, there is insufficient force to separate most of the air and water molecules from the chamber. Air pressure is 760 torr so vacuums below 1 torr, really do not supply much more "force" to pull off water and air from the chamber walls than even the highest vacuum. With the clarification that there will be equilibriums established that does make lower pressures more advantageous but it is more about the concentrations than the difference between 1 torr and 10^9 torr.

Typically, UHV systems use bake-outs at high temperatures to facilitate the process of removing water and air from chamber surfaces. This works because of the Bolzmann distribution that higher temperatures will have more molecular movements in the upper tails of the distribution. Just like quantum tunneling, room temperature itself is not enough to knock water and air out and off of metal surfaces. It is the higher energy molecules in the more energetic tail of the distribution that is sufficient to get the job done. That is why baking moves the molecular motion into more energetic states that are sufficient to separate the impurities from the chamber surfaces. Even with a high temperature bake out, this is a time intensive process.

Plasma cleaning can do this to an extent. It is great for knocking the water loose but does nothing to remove the nitrogen and oxygen in the air. That molecules get imbedded into the metal chamber surfaces. Once deuterium is introduced, it is a combination of knocking off the nitrogen and oxygen impurities as well as imbedding deuterium into the surface for beam on target fusion.

At some point, a high deuterium flow rate can overwhelm any contamination but patience and operating time wins the race.

[Dennis P Brown]

One thing a plasma is very good at is removing organics - trace oils from the hands and the pumping system. But as Frank says, bake out is best for water vapor and that is the #1 issue for most fusors after being exposed to the air.