Vacuum High Voltage Discharge - Experts Only

Keep me informed

Dear Shyam,

this is more a question of geometry and gradients.

Have a look on the "Paschen-curve" - should be in any of the high vacuum books - if not I will search and find.

This curve is plotting the voltage necessary for plasma-generation or discharge over the product of (distance x pressure).

I do not expect that you will have problems with the low pressure you have.

Look at the construction of electron guns used in x-ray tubes.

Pressure is near 1 nbar and voltages well above 100KV.

I am not an expert in this high voltage application, but I have a feeling that cleanliness and avoiding sharp edges and avoiding outgassing and materials that will degrade may be more important than distance.

RHABE

Dear RHABE,

"Parchen curve" has sudden jump in pd parameter in this vacuum zone and no longer obey rule of what we see from atmospheric pressure to 1ubar pressure. 1nbar has fewer atoms and perhaps secondary emission of electrons from electrode may be of greater concern at shorter distances. I am loong for safe distance where pA current may be tolerated but no avalanche should form to cause currents greater than few uA in any case.

Yes geometry makes great difference in cylinrical co-centric electrode designs.

There must be some record of actual data somewhere on low pressure safe spark gap. That is what I am looking for.

Paschen's law is likely what you need to know. Remember though "air" has many different gasses in it that constantly change concentrations from day to day. The gas that is likely skewing your setup is the dipole of water vapour. This life sustaining molecule will dramatically change the the value of pd. The most dramatic demonstration of this change in breakdown voltage occurs in a thunderstorm. There are additional spatial effects that will occur from localized crowding of charge (the sharp edge effect) but this produces locally different voltage gradients than the external source.

Without a full three dimensional representation of your apparatus, predictions of how high of a voltage you should be able to sustain is impossible. Once breakdown occurs though one cannot limit the current exchanged purely by spatial geometries. The circuitry producing your high voltage will briefly "see" the equivalent of a short circuit. Your high voltage circuitry should then be designed to safely respond to an unexpected shorting.

You must do this because of the last item that can and will cause unanticipated shorts, ionizing radiation. Even if your little vacuum chamber sits in the lowest shaft of your nearest coal mine, you will occasionally get some form of ionizing radiation to happen. With a 30kV gradient you will be well into the Geiger-Müller region for an ion chamber. This means that one single electron ionized atom will be so quickly accelerated that all electrons will be likely stripped off. The accelerated heavy ion nucleus will likely collide with another gaseous molecule and compound the amount of charge produced.

(PS Please use the spell checker.)

Dear redfred,

GM Tubes are near atmospheric pressure and just 0.1bar at best and ion chambers are pressurized one. I am talking about 1 nbar vacuum which is also used in vacuum switches for isolation. These switches are not very large in size and safe distance perhaps is only few centimeters in them. How do they manage no-contact and then contact in them?

The Geiger-Müller region for ionization of a gas in an electrostatic field exists at many different pressure regions and for different gasses. Depending on the pressure and temperature of the gas ionized, the electrostatic field thresholds values change. But always the first and last region exists for all pressures and temperatures of gasses. There are four regions, each based on the field strength of the electrostatic field:

1. Recombination region. (I think there's a more formal name but I cannot remember it now) In this region the electrostatic field is insufficient to prevent recombination of ionized gaseous molecules.
2. Ion Chamber region. In this region the electrostatic field is strong enough to likely separate electrons freed from interaction with ionizing radiation. However, insufficient acceleration occurs to cause secondary ionizations. So no more than one electron and ion pair is released per collision. In this mode one can clearly see the difference in the speed of free electron detection versus ion detection. The draw back of this region is that a substantial number of collisions must be expected to produce an acceptable detection. The advantage is that this is a fairly flat plateau region. In other words, the electrostatic field can vary fairly widely while the same one collision one pair production occurs. The bias voltage does not have to be well regulated.
3. Proportional Chamber region. As the name implies in this region as the bias voltage increases, more than one electron gets released per collision. I don't remember if this is from additional collisions of the accelerated electron or ion with non-ionized gas molecules. (I suspect the higher momentum of the ion.) This is sometimes called a gas gain effect. One advantage of this gas gain, less intense radiation can be detected. Also changing the bias voltage can make the detector more or less sensitive to radiation. The drawback is that now the bias voltage must be known and reasonably well regulated to have a calibrated detector.
4. Geiger-Müller region. In this region the electrostatic field is so high that ionization of a particle causes acceleration so rapid that all free valence electrons are stripped from the molecule. * In this region the gas gain is the highest so the least intense radiation will be detected. This is also a plateau region so as long as the bias voltage resides above a critical voltage the detector remains calibrated.

So while you do not have a Geiger-Müller tube and your pressure reside well below nominal values used by Geiger-Müller tubes for optimal gains, the voltages you have will make you the most sensitive to radiation effects. Now with a centimeter of clearance (a large distance in my opinion) and 10^-9 bar of pressure you will not be capable of sustaining an arc but any radiation will surely increase your leakage current.

* Above the Geiger-Müller region binding electrons get removed breaking chemical bonds. This makes for a very unpredictable soup of historic significance, far beyond this discussion.

Dear Shyam,

the sudden jump in Paschen-curve at 1 µbar and below is resulting from free mean path length and ionisation probability.

At your pressure of below 10^-9 mbar the free mean path length is above 70 m, so there is no longer the possibility of avalanching and related current conduction.

But there is the possibility of:

a. radioactivity (not likely at pA level, but critical at single electron counting)

b. insufficient insulation quality of material resulting in breakdown

c. insufficient cleanliness: same as b. but restricted to immediate surface layer

d. gas sources that generate localised pressure well above system pressure

(we saw big problems from entrapped -by machining- small amounts -10^-9g- of oil)

e. field emission at sharp edges or needles

f. high field recrystallisation of non-suited materials (whisker formation)

So I am convinced that you have to reformulate your question because it cannot be answered in the classical sense as in much higher pressure regions:

- which materials are suitable,

- how to ensure a surface of sufficient quality with nearly no gas sources

- which materials to use

- how to clean after assembly

- to burn down or not any problematic contaminations with high voltage (slowly rising) with limited current and fast shutdown

If you look to typical constructions, not only the above mentioned x-ray-tubes but also the CRTs for oscilloscopes and TV, these often use acceleration voltages up to 30 KV - so any construction details will be easy to access.

Have success! Good luck

RHABE

RAHABE,

Assuming that electrodes are polished, clean and has almost no gas trapped in them and there are only molecules in the chamber at that pressure whatever can fit into the pressure volume curve at room temperature of 25C.

Do you have a clear idea for example in ion source or ion focusing tube, what is the safe distance for generating field from 30kV electrodes that will not spark at 1nbar pressure? What is the maximum field strength I can have without breakdown?

R.V. Latham (editor), High-Voltage Vacuum Insulation: Basic Concepts and Technological Practice (Academic, London, 1995).

www.geocities.com/afendi63/HV1_v2.pdf

High voltage engineering by J. Kuffel, W. S. Zaengl

High voltage insulation engineering by R. Arora, W. Mosch

High Voltage Engineering by M.S. Naidu

above is useful literature.

Dear Shyam,

"polished" electrodes are often not smooth if electrical characteristics are measured, any content

of oxides or sulfides will play havoc.

"no gas" condition will not exist, as any ion or particle bombardment will free entrapped gas,

if prior "baking 450°C, 1h, while pumping" removes most of the water, any new surface will have a lot of gas freed.

Think about 100 ppm gas inside a metal (quite a usual condition) then 1 m³ of metal of density 10g/cm³ will have dissolved 1kg of gas!

We did measure absorbed water on aluminum (2017), measurement started within 10 s of pump start,

in the first 10 seconds 3 complete layers of waters came off, the same for the next 100s and so on until the measurement was stopped

after the last period of 10⁷ s. This was from a carefully prepared and with a thin (10 to 100nm, nonporous) oxide-layer protected surface.

Safe distance: 5mm is sufficient at 100KV if the system is clean and stays clean and any surface and insulator is ok.

If this cannot be guaranteed then stay with 1 KV/mm. These figures are from the above cited literature.

No own experience beyond 1.5 KV at 0.5mm distance - this causes arcing after pump-down by particles nobody can avoid.

With a fast shutdown power-supply this arcing is not dangerous to the surfaces (but was at arcing below a VITON seal).

If you like I can ask a friend who has experience with synchrotron operation or a nearby high-voltage lab.

RHABE

Dear RHABE,

That is a lot of information.

400C baking 1h-24h and pumping to degas is usually done when chamber is open for some reason. Ceramic seal electrodes are often used at the entry point having spacing of about 10mm-20mm from each other. We also use 10kV and 20kV and 30kV SHV connectors of standard types.

There are electrodes that form focus points, are in the form of discs and tubes and those for accelerating the ions are in the form of mesh or grid.

Not an expert. Did you look here to see if what you need already exists?

http://www.rossengineeringcorp.com/detailed_information_singlepole.htm

milo

Dear Sir,

One of my friend is setting up a metallisation plant for plastic in which two electrodes are discharged in near vacume situation.Is this the similar thing?

anyway

does this add some value for for you?

Dear rakesh_semwal,

First thing I will like to tell you that no one is with title "Sir" in this forum as far as I know, so you can call people by their name or token name they have placed here. All are equal here and this applies for you too.

Vacuum arc deposition processes are different from what I am talking here. I really do not want anything to get ionized and then conduct due to electric field and looking for safe distance where it is rather avoidable.