Insulating in a vacuum / General Science and Electronics / Forums

Forums

4hv.org :: Forums :: General Science and Electronics

« Previous topic | Next topic »

Insulating in a vacuum

1 2 next

Move Thread

LAN_403

Turkey9

Tue Sept 22 2009, 01:24AM

Registered Member #1451 Joined: Wed Apr 23 2008, 03:48AM
Location: Boulder, Co
Posts: 661

I've been thinking a lot about whether or not it's smart to put 1200v electronics is a vacuum without completely potting them. The voltage breakdown of air decreases with a decrease in pressure... that's one reason that neon lights work.

At first that would make me think that an inch and a half gap would easily be jumped by 1200v. That means that all my expensive power electronics need to be permanently encased in a resin of some sort.

But when I thought more about it, at a vacuum of 10e-5 Torr, the particle density would be low enough that not very much current would flow due to a lack of molecules to provide electrons. I know that the electric fields will rip electrons straight out of the exposed joints, but it shouldn't be too much. If the current is low enough, the losses should be acceptable...

But then again maybe the current is too high. Is there some way that I could keep the electric fields low enough to stop any form of arcing without potting the entire thing? It will all be pretty close to DC but with a huge potential for current flow if shorted.

The system will be in a vacuum chamber pumped by a 5in diffusion pump. I think it will reach at least 10e-5 Torr, am I in the wrong ballpark here? The pump doesn't have any specs. The smallest gap will be around 1in. No solution is too crazy as long as it saves the parts and keeps them available for future projects.

Thanks-
Jesse

Tue Sept 22 2009, 02:39AM

Registered Member #2372 Joined:
Location:
Posts: 62

You are correct in stating that the breakdown voltage decreases with decreasing pressure, but only to a point, then it rises very rapidly as you correctly stated due to the lack of anything to breakdown.

Once you are at a low pressure, 10^-5 is low enough, you just need to keep the electric field down to avoid field emission. You do this by avoiding anything that is sharp as this leads to field enhancement. Or anything that is hot as this leads to an effective lowering of the field necessary to pull electrons out of something.

One thing to mention is that even if you pot something the field enhancement will still be present so electrons can be ripped out of the pointy metal and then cause a secondary avalanche as they move through the dielectric you potted them in exacerbating the problem. Also, putting dielectric in your vacuum can also lead to multipactoring on the dielectric if it becomes charged.

A last thing to avoid is triple points, the interface of vacuum, metal. and dielectric, where the angle between the dielectric and the metal is less then 90 degrees as this can lead to secondary avalanching on the surface of the dielectric.

So, basically, make everything smooth.

Good Luck

Proud Mary

Tue Sept 22 2009, 10:31AM

Registered Member #543 Joined: Tue Feb 20 2007, 04:26PM
Location: UK
Posts: 4992

Hello there Turkey farmer!

It's not really possible to answer your question at all, without some general over-view of the application.

What is this apparatus? What is it's purpose? How long must it work for?"

Why must the 1200V electronics be inside a vacuum or a potting box?

Please do give us a much more detailed picture, including circuit diagrams (US: "schematics") and you'll most likely get an answer of the usefulness and accuracy you are hoping for,

best wishes

Harold Godwinson

Turkey9

Wed Sept 23 2009, 02:35AM

Registered Member #1451 Joined: Wed Apr 23 2008, 03:48AM
Location: Boulder, Co
Posts: 661

I don't really have a schematic per say at this point but it's just a simple coil and capacitor switched with an SCR just like an induction launcher. I'm building a type of pulsed plasma thruster for science fair and the entire thing has to be inside of a vacuum chamber for testing. The full 1200v will only be present for at most twenty seconds after the caps are charged and before the device is fired. The average time for the charge will likely be less than a second. I could have placed the caps outside of the chamber and ran wires in to the thruster, but that would create too much stray inductance for my liking. Below is a picture of the capacitors and terminals along with one of the four SCRs I'm using.

1253672964 1451 FT1630 0922092017a 327569

1253672964 1451 FT1630 0922092018a 259641

1253672964 1451 FT1630 0922092018b 309643

Wed Sept 23 2009, 03:02AM

Registered Member #2372 Joined:
Location:
Posts: 62

Can that type of capacitor be pumped down, it might leak all inside the vacuum once you turn the pump on, not only will it be gross but you wont be able to get to a very low pressure if it is leaking. You could build a low impedance coaxial feed for getting the power into the chamber, it is kind of a standard way for transporting pulses into a vacuum system, also it is pretty easy.

wylie

Wed Sept 23 2009, 03:24AM

Registered Member #882 Joined: Sat Jul 07 2007, 04:32AM
Location:
Posts: 103

I Think the part that was missing was how this is a matter of space-worthiness testing. Its not like Turk just wanted to box a TC in a vacuum chamber. (OT, anyone thought about running a VDG in a low-pressure chamber?)

Most systems bound for space have to undergo vacuum testing, even when they're intended for use in pressurized environments, cause you never know when you'll lose pressure.

Cant have materials outgassing and changing values unexpectedly. Dielectrics, i'm looking in your direction

I don't have any expertise on the actual issue, but i'm inclined to agree with dugg. Field and thermionic emmission are the main culprits of loss in a vacuum. Plus i think arcs in a vacuum need a decent glow-discharge (IE field- or thermionic emmission) to even get started.

I say just think in terms of vacuum-tube design. They certainly don't epoxy-pot the elements and support rods inside tubes/valves. Its all about geometry.

Potting seems like its mainly useful for terrestrial@1atm applications. I spose also down to low enough pressure until the mean-free-path starts increasing breakdown voltage. But you better outgass your materials and put them in sealed enclosures ASAP.

(Sorry to chime in my very-uninformed 2cents but i cant resist when vacuum and electrons get involved. time to head over to the "DIY Particle Physics" thread

)

Turkey9

Wed Sept 23 2009, 05:34AM

Registered Member #1451 Joined: Wed Apr 23 2008, 03:48AM
Location: Boulder, Co
Posts: 661

Actually there are a few papers out there dealing with potting of high voltage power supplies for satellite that were written by the folks at JPL; I only assume that they were exposed to the vacuum of space while in operation.

Also those caps are polypropylene not the normal electrolytic and are sealed in an epoxy that is pretty low outgassing. I asked the guys at CDE and they said the caps should be fine in a moderate to high vacuum. The black plastic jacket on them is pvc and will be removed before going into the vacuum chamber. I've done a lot of research on what materials can be used and i'm not worried about outgassing at this point.

As for the feed through, one reason I don't want to do that it that it is another place for a leak to form. In my experience those are always the trouble areas; homemade feed throughs.

I know that in large voltage multipliers rings are used to disperse the electric field in a more even geometry to help cut down on losses from corona, would something similar work here?

Also, as much of the wire as possible will be teflon insulated.

Proud Mary

Wed Sept 23 2009, 10:44AM

Registered Member #543 Joined: Tue Feb 20 2007, 04:26PM
Location: UK
Posts: 4992

Hiya Turk

I'm not at all clear where this vacuum comes in, or what its purpose is, or how anti-corona rings should be neded at 1200V, so I'll leave off further comment until you've stuck up a circuit diagram, and some sort of engineering drawing of the device, cos I'm only a beginner, and if anyone can get the wrong end of a stick, it's me!

all the best

Harold Godwinson

Wed Sept 23 2009, 12:53PM

Registered Member #2372 Joined:
Location:
Posts: 62

The reason it could matter with only 1200V is because there is no air to stop electrons from crossing a gap. At 1200 volts you arent going to be pulling electrons off of a flat surface, but small pointy things like threads on bolts might have huge electric fields on them due to field enhancement and any current can flow right across a gap leading to large losses.

Sulaiman

Wed Sept 23 2009, 08:35PM

Registered Member #162 Joined: Mon Feb 13 2006, 10:25AM
Location: United Kingdom
Posts: 3141

10E-4 Torr is low enough to be well below the paschen minimum.

Here is a great reference

1 2 next

Moderator(s): Chris Russell, Noelle, Alex, Tesladownunder, Dave Marshall, Dave Billington, Bjørn, Steve Conner, Wolfram, Kizmo, Mads Barnkob