Question about paschen's law and fusors

It may be difficult to separate "theory" from "application," but let''s see if this helps facilitate the discussion.
User avatar
Jim Kovalchick
Posts: 324
Joined: Thu Apr 14, 2011 12:00 am
Real name:

Re: Question about paschen's law and fusors

Post by Jim Kovalchick » Thu Apr 17, 2014 12:57 pm

here's another
Attachments
non planer paschen.pdf
(827.05 KiB) Downloaded 171 times

User avatar
Richard Hull
Moderator
Posts: 11368
Joined: Fri Jun 15, 2001 1:44 pm
Real name: Richard Hull

Re: Question about paschen's law and fusors

Post by Richard Hull » Thu Apr 17, 2014 3:43 pm

This latter paper was most instructive. This stuff can't be quantified except by experiemnt with each system at varying pressures. Rules of thumb abound in this world of nearly an infinity of geometeries being possible. I have always felt field emission controls it all in most any gas environment at any pressure and that is all about the smallest radius electrode ruling the process. Its placement in relationship to the larger radius opposite polarity electrode is also key to the process. It is a an experimental situation, always.

Doug has experimented and reported his results. Pachen's Law be-damned.... it has rigidly controlled parameters. Anything beyond this fixed rigidity and you go experimental with some new bizarre mathematics controlling the going's on.

It is always nice trying to understand things and reasons for variations left and right of the generalized rules, but acceptance of "flyers" is the norm outside those generalities. Unless those specifics are your "raison d'etre", you can move on after defining your operational regime through real experiment.

Ricahrd Hull
Progress may have been a good thing once, but it just went on too long. - Yogi Berra
Fusion is the energy of the future....and it always will be
Retired now...Doing only what I want and not what I should...every day is a saturday.

User avatar
Doug Coulter
Posts: 1312
Joined: Sun May 27, 2007 7:18 pm
Real name: Doug Coulter
Location: Floyd, VA, USA
Contact:

Re: Question about paschen's law and fusors

Post by Doug Coulter » Thu Apr 17, 2014 6:59 pm

Thanks Jim, I'll see if these fit my analyses and observations when I get a chance to do it. While a fusor isn't parallel plates by any means, and does have more field gradient near the grid than the wall, if you assume PxD - that doesn't matter one whit, it all washes out in the end. But if there are other effects, it's another story. Just trying to work out the wichness of the why here.

I tend to agree with Richard, here and elsewhere. But I've seen this exact (or within measurement errors) with grids of 2x different size wires, ends with various radii, and a host of other changes. Yet, even though my cylinder feeding into a larger tank is vastly different than Richard's setup - and most others - I get the exact same numbers for microns and light-off volts he does, again, within our measurement accuracy. That's what made me think the published "rule" is the thing that's wrong, not some vagarie of a particular fusor or grid configuration. I could, of course, be wrong, which was the point of bringing this up in the first place.
Why guess when you can know? Measure!

User avatar
Jim Kovalchick
Posts: 324
Joined: Thu Apr 14, 2011 12:00 am
Real name:

Re: Question about paschen's law and fusors

Post by Jim Kovalchick » Thu Apr 17, 2014 8:30 pm

I'm not sure that the published PxD curves are necessarily wrong, but rather the empirically designed relationship simply doesn't account very well for other geometries. There is enough other work out there that supports the curves for parallel plates. In his senior year, son Mike tried a bunch of different approaches to simulating electrical interaction of dust collisions in rarified gases like that which could exist in space nebula. His background research looked at a number of papers that discussed the Paschen discharge effect including true parallel plate work. Ultimately, Mike used two, rapidly contacting and separating parallel plates at varied pressures to see what happened. He didn't have a way of measuring plate potential, but he was able to watch relative charging and discharging. He video taped his runs and didn't manage to discern anything other than simple contact electrification until he told me that he suspected that the pattern wasn't simple but the speed of the reciprocator made it difficult to see. I suggested he use a fast frame camera to slow it down. After he did another run, he watched the slowed down video for a pattern, and he found that on average, there was a consistent number of contacts and separations between what appeared to be a discharge without contact. His next step was to repeat at various pressures. Low and behold, plotting the average contacts between discharges against pressure yielded an upside down knee the very shape of Paschen. He surmised that he was seeing support of the idea that objects colliding in a gas will continue to take on tribocharging until Paschen says discharge. It would be interesting to see his experiment repeated with a high impedence electrometer to measure the voltage. I know I'm a little off topic with my ramblings, but I think ultimately a true understanding of Paschen discharges will lead to more optimum manipulation of conditions that effect glow plasma discharges. The impact on processes like fusion could be profound.

User avatar
Doug Coulter
Posts: 1312
Joined: Sun May 27, 2007 7:18 pm
Real name: Doug Coulter
Location: Floyd, VA, USA
Contact:

Re: Question about paschen's law and fusors

Post by Doug Coulter » Fri Apr 18, 2014 12:11 am

I for one would love to see that, Jim. I'm still confused deeply after reading that paper you put up here, at least in part because even in their simulation, it's still off by factor 10 (actually, more) than what we all see.
I've tried various grid materials, high and low work function - no change. They also have different secondary emissions due to ion impact - again, no change. I have the strangest fusor geometry of all here (as far as I know, anywhere) - no change.
We all see right around 15-16 microns (eg pxd at around .01, not the limit of .1 in the paper where it's already off-scale) almost no matter what, and it scales with size as Paschen says - bigger is lower volts at same pressure, but still off by factor 10 or a bit more. Scales nicely out to my 14" diameter 26" tall tank by the PxD difference, just that the base is off by factor 10+.

For example, Richard uses a sphere. I use this:
TopViewLightsOn.JPG
How much more different can it be? The main tank is 14" diameter, yet the strike volts at 15 microns are the same as in a sphere.
I guess I should stop beating this horse, since no one seems to have a scientifically based answer with experimental backing for why it's off by that much. This isn't in the last decimal place, it's a factor of 10! Reworking Paschen's law for field concentration doesn't get you anywhere, I tried that. The only time I've seen a lower striking voltage is when I had literally atomically sharp points on the ends of tungsten rods to get up into the megavolt/mm range it takes to start getting enough field emission to be more than the natural cosmic ray background. And then it didn't light off, it simply melted the sharp points off the tungsten rods.

Oh, couldn't resist - here it is running through the same window. Note that though there is a line focus in the middle, the rays are centered along the length - which is another "why is that" question.
RunningHard.JPG
running roughly 2m neuts/second, from same viewpoint
Why guess when you can know? Measure!

Andrew Haynes
Posts: 90
Joined: Sat Dec 01, 2012 6:25 am
Real name: Andrew Haynes

Re: Question about paschen's law and fusors

Post by Andrew Haynes » Fri Apr 18, 2014 6:51 am

Could this be something related
http://en.wikipedia.org/wiki/Casimir_effect

EDIT, You are right Richard, the only reason I mentioned it, is the atom might be the plates.
Last edited by Andrew Haynes on Fri Apr 18, 2014 11:07 pm, edited 2 times in total.
Andrew Haynes

User avatar
Richard Hull
Moderator
Posts: 11368
Joined: Fri Jun 15, 2001 1:44 pm
Real name: Richard Hull

Re: Question about paschen's law and fusors

Post by Richard Hull » Fri Apr 18, 2014 4:06 pm

No even close! Totally unrelated. 0.000 quantum effects related to this post in any way. Try to stop grabbing at all the bizarre stuff when this is not even an nuclear based issue. This is a molecular gas ionization issue within an electrodynamically macroscopic system where all distances are stellar in nature in relation to the Casimir effect. read you own Wiki URL about what the Casimir effect demands.

1. uncharged plates (we are talking highly charged non-plate electrodes.)
2. sub-sub-microscopic separations ( we have tens of billions greater separation than the Casimir effect demands.)


Richard Hull
Progress may have been a good thing once, but it just went on too long. - Yogi Berra
Fusion is the energy of the future....and it always will be
Retired now...Doing only what I want and not what I should...every day is a saturday.

Jeroen Vriesman
Posts: 277
Joined: Mon Feb 07, 2011 9:08 pm
Real name: Jeroen Vriesman
Location: Netherlands
Contact:

Re: Question about paschen's law and fusors

Post by Jeroen Vriesman » Fri Apr 18, 2014 6:08 pm

Attached is a graph from a vacuum switch factory, it shows the breakdown voltage at a fixed distance for high vacuum, low pressure and high pressure.
These are simply measured values (flat copper anode/cathode)

They agree with the Paschen curve on wikipedia for the high pressure part, but not for the low pressure part.

Conclusion for now: wikipedia describes the Paschen curve as if vacuum break down doesn't exists, it fits the data to the the theoretical model for low pressures.
In reality vacuum breakdown does exist and the combined Paschen/vacuum breakdown curve does not go to infinity at PxD < 0.2 Torr-cm.

Looking at E-2 mbar in the graph, the breakdown voltage is about 5kV at 10mm, which translates to about 32kV at 63.6 mm (it is incorrect to assume a linear relation between distance and breakdown voltage, but I do it anyway).

My previous post about microplasma's...would still be interesting to see if the current spikes are there.
Attachments
holec_vacuum_paschen.jpg

User avatar
Chris Bradley
Posts: 2931
Joined: Fri May 02, 2008 11:05 am
Real name:

Re: Question about paschen's law and fusors

Post by Chris Bradley » Fri Apr 18, 2014 7:24 pm

Jeroen Vriesman wrote:Looking at E-2 mbar in the graph, the breakdown voltage is about 5kV at 10mm, which translates to about 32kV at 63.6 mm (it is incorrect to assume a linear relation between distance and breakdown voltage, but I do it anyway).
It is not simply incorrect, it is highly incorrect. This is the pressure around which the relationship reverses, where you can get less insulation against breakdown for further distances. This is because as you approach the molecular flow regime, the distance over which a seed ionisation can happen, leading to a cascading breakdown, increases with increasing distance rather than decreases.

After that, whether it cascades or not rather depends on the gas type, the higher the nuclear number the more electrons may get cascaded if struck by a dislodged electron/ion.

User avatar
Doug Coulter
Posts: 1312
Joined: Sun May 27, 2007 7:18 pm
Real name: Doug Coulter
Location: Floyd, VA, USA
Contact:

Re: Question about paschen's law and fusors

Post by Doug Coulter » Fri Apr 18, 2014 8:00 pm

Richard is, of course, correct. The issue here is the very long mean free path - at least before "lighting off" which we here have proved (should be obvious, but always nice to measure) that we have enough "focus" or "compression" to go from molecular flow to viscous flow - at which point Paschen kicks in and works more or less correctly - the pressure in the "rays" is higher than the surrounding, for example, so the P in PxD becomes tenable. We've even gotten flowing glowing ions to flow around corners in (short) pipes once it was "compressed" like that - at least it proved we have actual pressure (not a lot, but more than in the bulk gas).

It took the best scientists at NBS and then NIST to even measure the Casimir force, it's so tiny - it's at Planck scale, 10^-39 class stuff. This ain't it. This is every time and for everybody. It's something at a much larger basic scale.

It's what's happening at the onset that intrigues me, and I admit I have an ulterior motive. When we make the thing oscillate, and pay close attention to our neutron data (some of which is from an unmoderated hornyak and therefore fast) - we find most Q right as the thing starts up. It drops off by factors of thousands thereafter, very quickly as we reach that "dynamic equilibrium", or so we measure here. Could be all the spins tend to align - (D has spin=1), and that's bad for fusion...we simply do not know at this point what happens once things settle down. Could be random spins are better than "all wrong except for random collisions" - the oscillations give time to re-randomize? Again, don't know. What I want to know is how this thing starts up from 1/10 the PxD that any interpretation I'm aware of allows for even at infinite volts.

Here's a video demo of us getting around 2800 times the normal "Q" we get in static mode. Skip out to 2:00 if you're easily bored, or 2:18 or so. When we get it to oscillate (and we don't even know what octave we should be in, we could be way on the tail of a resonance and missing a huge peak effect at this point - this means a lot of hours in the lab trying stuff), we get this high fusion Q. We'd like to do this on purpose and the right way, of course, rather than having it happen almost by accident (not quite accidental, there's a transformer, untuned, between the main and ion source grids with a very low electrical Q due to the ballast resistors involved at the moment). https://www.youtube.com/watch?v=ZGiU_Ck ... Ij&index=2 Yeah, it's a problem that Q is an "overloaded" symbol here that can mean fusion out vs power in - or electrical resonance (kind of means Quality either way). You have to judge from context or word very carefully.

I'm not the first guy to ask this question, though I forget which scientist did (WAY back at the beginning, maybe Rutherford?) - he noted that while theoretically impossible, vacuum arcs do indeed occur. The "plasma at the surface" theory is probably closest, as everything has little atomic bumps that do have some finite (but unmeasurable here) field emission, and therefore heat up and release any trapped gas which might just do it. Doesn't take too many femto-amps of FE to heat an atomic sized point - and even though those low currents don't ionize anything much - there just aren't enough electrons or D2 atoms in their path - they may release surface gas along with the vaporized metal and that does the trick.

My point being we just don't know, and we like to think of ourselves as scientists, albeit amateur ones. That holds us to certain standards, one of which is if we don't know something, we admit it and try to find out what we don't know. It's how we move forward, which is the glory of science.
Why guess when you can know? Measure!

Post Reply