Efficient way to bombard target with electrons in the 50-100 eV range? (not for fusor)

[Arun Luthra]

Here is a first pass at a chamber design:

The dome is a glass bell jar. It's on a silicone rubber gasket ring and a 1/4" thick steel disk baseplate.

I send two high negative voltage lines into the chamber because they differ by the filament voltage. What can I use to supply to filament voltage? I need to add a few volts (preferably an ajustable voltage) to a line that is at -1000V. Maybe a battery and a potentiometer?

What's a good way to shape the field to focus the electrons onto the anode target? The targets will be insulators... could ring magnets beneath the target be useful?

The ground attached to the 0 V line of the high voltage seems slightly odd to me, but I saw that choice elsewhere. Is this totally wrong? In this picture, the entire baseplate is connect to the 0V on the high voltage supply.

[Richard Hull]

Your HV supply must be either totally isolated or either a positive ground supply. If isolated you will need to ground the positive line as you show in the diagram. You cannot use a normal negative grounded supply here. Most normal power supplies are negative ground. You may have to wind up making your own HV supply.

Richard Hull

[Arun Luthra]

My HV supply provides two wires, one is negative high voltage up to -1000V, and the return line. So, this is termed a "positive ground supply"? Good to go then.

For the filament voltage, I have a DC low voltage power supply with negative, ground, and positive terminals. Should I attach the -1000 V cathode wire to the low voltage DC supply ground pin? Then, then + terminal on the low voltage DC supply will be at -997 V? Or will this cause arcing and explosions...

[Richard Hull]

This is the norm. One side of the HV supply always goes to the filament. In this case your filament supply must no have a ground and be capable of handling 1 kilovolt without arcing within itself. A well insulated AC filament transformer driven by a variac is ideal here. There is no need for a DC filament supply at all. Why risk your DC low voltage supply when an AC filament transformer would do fine?

Richard Hull

[John Futter]

Arun
With a 1000 volt supply you will get electrons with 1000eV and maybe a bit less arriving on your target.
You say you want 50 to 100eV electrons arriving then that is what the voltage should be
@ Work I use an electron microscope filament running white hot (as seen through viewport) with a 0 -60 volt supply to bathe a nonconducting sputter target with electrons so that the target charge is taken away so that the positive argon ion beam hitting the target is not repelled off the target

you cannot use vacuum tube filaments or magnetron filaments effectivly for an electron source as the coatings that promote electrons to leave get poisoned once in air and the effieciency / life drops.
SEM filaments use different coatings that are compatible with air but you have to run them hotter than vacuum tube filaments to get the same number of electrons leaving. If you do try a vacuum tube filament you will have to run better than yellow hot to get appreciable electrons

[ian_krase]

Ion beam sputtering? Sounds interesting -- and like something to try if I ever get a bigger chamber.

If I was doing this, I would put the filament at ground potential and drive the target positive, assuming I have three feedthroughs. Makes the filament heater supply easier.


I don't use SEM filaments -- I just weld my own filaments from tungsten wire. Far cheaper. Yes, they need to be white hot. You may be able to coat filaments with yttria or the like, but that's additional complexity.

[Arun Luthra]

Arun
With a 1000 volt supply you will get electrons with 1000eV and maybe a bit less arriving on your target.
You say you want 50 to 100eV electrons arriving then that is what the voltage should be
@ Work I use an electron microscope filament running white hot (as seen through viewport) with a 0 -60 volt supply to bathe a nonconducting sputter target with electrons so that the target charge is taken away so that the positive argon ion beam hitting the target is not repelled off the target

The supply (acquired after starting the thread) is adjustable 0 to -1000V. 50-100 eV is just the main area of interest. I would like the system to be robust enough to withstand -1000V.

you cannot use vacuum tube filaments or magnetron filaments effectivly for an electron source as the coatings that promote electrons to leave get poisoned once in air and the effieciency / life drops.
SEM filaments use different coatings that are compatible with air but you have to run them hotter than vacuum tube filaments to get the same number of electrons leaving. If you do try a vacuum tube filament you will have to run better than yellow hot to get appreciable electrons

I will only put current through the filament under vacuum (e.g. 1x10^-4 Torr). Will the filament be poisoned at atmospheric pressure in air even if there is no current through it?

[Arun Luthra]

This is the norm. One side of the HV supply always goes to the filament. In this case your filament supply must no have a ground and be capable of handling 1 kilovolt without arcing within itself. A well insulated AC filament transformer driven by a variac is ideal here. There is no need for a DC filament supply at all. Why risk your DC low voltage supply when an AC filament transformer would do fine?

Richard Hull

Thanks for the tip. I have a microwave oven transformer. It has a filament winding (something like 1 or 2 turns) in addition to the HV winding. Maybe I can add some voltage adjustability somehow. I need to be able to limit the current. I can leave the high voltage secondary winding unused.

Then, my cathode would have DC high voltage (tens of volts up to 1000V) with a few volts AC on top of it.

[Arun Luthra]

Still trying to understand Paschen's law.

Suppose I have a hypothetical vacuum chamber with a cathode and an anode. There is no hot filament, just cold flat plates at high voltage. At 10^-4 Torr and 1 kV between the plates, the breakdown gap distance is 2500 meters for Argon. Meanwhile, vendors sell vacuum high voltage feedthroughs that operate at a few thousand volts with gaps that range from a few mm to a few cm. So there is many orders of distance between the practical high voltage situation and Paschen's law. What is going on here? Is the Paschen "breakdown" describing 1 nano amp of current? 1 microamp? Something larger? Are there any tables/charts that answer this?

[Richard Hull]

Read and take in the issues here

http://www-eng.lbl.gov/~shuman/XENON/RE ... report.pdf

There are constants and terms in the law that must be defined in its application. Breakdown is ionization of the contained gas. Plates are the key in the law as originally stated. It is extremely rare to have two identical flat plates in a gas at varying pressures of just air. (what the law was written for). Change just one or multiple conditions and breakdown occurs at a different potential.

It is an equation with unknowns that the user must define.

This is why you can't get a "feel" for this until you do some work with voltages and electrodes in a vacuum. Until that time you, yourself, are in an intellectual vacuum! This is why I spent 1 year of my life with a demo fusor and three books read from cover to cover as I experimented, observed and read. In the book "Ionized Gases", Von Engel is no prima donna, egg head... He gives, dutifully, all the relevant math to handle special cases, but never fails to warn the reader............change one thing outside of the equation's derivation and setup conditions and the equation fails. Fail to define the unknowns in the equation properly and the equation fails.... I learned a lot of things about electrostatics, far beyond my college days teaching.... Not by reading only, but by active hands-on experiment in a demo device.

Richard Hull

[ian_krase]

I will only put current through the filament under vacuum (e.g. 1x10^-4 Torr). Will the filament be poisoned at atmospheric pressure in air even if there is no current through it?

If you're using pure tungsten filaments, they will not be poisoned by air at any time whatsoever, but they will *burn out* if they are heated to red hot or hotter in the presence of air. 1E-4 torr is kind of a minimum vacuum for this kind of thing, they won't last very long at that level but in this kind of experiment you'll probably physically break them first. (Tungsten that has ever been heated to glowing is extremely mechanically fragile.) You can tell that the filament burned out from air if there's yellowish-whitish oxide on the inside of your chamber near the filament. The higher the vacuum, the longer the filament will last -- go to 10E-6 torr if you can!

Pure tungsten filaments work JUST FINE. They just need to be white hot. They are resistant to poisoning unlike other filaments.

If you are using cathodes from vacuum tubes that are coated... who knows? There are different coatings. I am not sure of the details of all of them. Some will hate air, others will not. Definitely don't heat them in air, that will destroy them. I would not bother messing around with this.

The third option is the yttria-coated iridium filaments from ion gauges. These are designed to not be poisoned by air, and can operate at fairly high pressures (even 1E-3 torr!) without rapidly burning out, although they will still have a longer lifetime the better the vacuum. However, they can be poisoned by various chemicals -- this is why tungsten-filament ion gauges are still used, since they are poisoning resistant.


A big part of what is going on regarding feedthroughs is that much work happens at high to ultrahigh vacuum, 1E-5 torr up to "how many zeros do you want". (and how many days do you bake your all-Conflat chamber while ion-pumping it!) I have worked with special motors that use high voltage -- they have a label saying that you can use them at atmospheric, and you can use them at high vacuum, but you cannot use them at between 100 torr and 10 microns because they will have glow discharges inside. Lots of stuff is assumed to just be for high vacuum. You may have seen Doug Coulter's posts on making special feedthroughs for fusion -- in his view, most feedthroughs are not designed to work with medium vacuum and glow discharges in general.

On the other hand, in a complex setup where there are not just simple anode and cathode plates or spikes, but a bunch of weird shapes... things get weird, and often there will indeed not be discharge in places you don't want it.