Grid heating by Current vs Ion impacts

It may be difficult to separate "theory" from "application," but let''s see if this helps facilitate the discussion.
Dan Tibbets
Posts: 578
Joined: Thu Apr 17, 2008 5:29 am
Real name:

Grid heating by Current vs Ion impacts

Post by Dan Tibbets » Tue Dec 13, 2011 8:57 pm

This is a brief argument against against Chris Bradley's claim that the cathode grid becomes incandescent (hot) due to the current passing through it (and through the plasma to the grounded wall to complete the circuit. I have contended that a few milliamps of current will not heat even a thin wire much, and that it is the ion bombardment that heats the cathode grid.

Looking again at this old series of images supports this position. The first image has similar current to a subsequent picture of the grid at lower pressure. In the second picture the grid is red hot. My interpretation is that the difference is purely due to the pressure difference. In the first, the electrons and subsequently accelerated ions have a significantly shorter mean free path (MFP) and thus quickly collide with neutrals. These neutrals are accelerated somewhat, but there are so many of them relative to the charged ions, that the resultant thermal energy is dilluted considerably, and few of the ions survive long enough to impact the grid and heat it with their retained KE that they were initially accelerated to.

In the second picture, with the lower pressure the MFP is greater and thus a significantly larger number of ions (or accelerated/ charge exchanged neutrals) survive long enough to hit the grid directly and heat it, instead of sharing their energy among low energy neutrals.
This is consistent with the voltage differences. The voltage is low in the first picture because of Ohm's law. The Resistance of the higher density gas (mixture of gas and plasma) is lower*, thus the voltage is also less in order to maintain the same total current. This is possible due to the limitations of the power supply and other considerations*

A counter argument that there is increased retained heat in the grid because of the lower vacuum levels- there is less cool gas molecules to carry away the heat by convection. This may apply, but at these pressures the diffrences relative to atmospheric pressure is so small that I assume it can only contribute to a tiny amount to the difference.

Image below was cut and pasted from this original-
viewtopic.php?f=18&t=7815&hilit=Dan+Tib ... low#p55774

*The voltage / current relationship in the chamber is dependent on the capacity of the power supply (voltage droop, etc), and also dependent on the rate of ion production relative to the voltage. The secondary cascading free electron/ ion pair production has the highest cross section at a few hundred eV's. This implies that at these voltage levels ionization rates will max out. Below these voltages the ionization rate will fall below maintainable levels. That is why there is a minimal voltage needed to light a plasma, dependant on the gas and density, and other considerations. At higher densities, the ionization rate will increase, but by proportionately smaller rates. That is why the available energy then goes progressively towards additional KE of the charged particles. There are still neutrals available to absorb the energy and ionize, but it becomes less prevalent. Eventually a balance is reached between the maintained voltage and ionization percentage. At low enough densities, and high enough currents the percentage of ions/ neutrals become very close to 100 %. At that point the rules change somewhat. My impression is that this occurs somewhere around ~ 0.1 to 1 Microns.

A very general appreciation that I have is that at ~ 100 Microns, the ion/ neutral ratio may be ~ 1/100, at 1 Micron ~ 1/1 and at ~ 0.1 Microns ~ 10/1. This has consequences not only for achievable voltages, but also the ratios of beam/ beam versus beam/ low energy neutral or target collisions. Even the low percentage of neutrals that may exist at these low densities have probably had charge exchange collisions, so they are near the same KE as the ions, thus equivalent to beam/beam collisions (though this would effect confluence some).
This is why the Wiffleball effect in the "Polywell" is so important (along with useful fusion rates). The density of the contained ions may be at equivalent ~ 100 Micron levels, while the non contained levels of neutrals remain at sub Micron levels (provided you can pump them out fast enough). This has huge consequences on the scaling of beam/ beam collisions versus beam/ background or beam target rates. This has been hinted at by the fusion rates in some experiments as the pressure drops below ~1 Micron levels.

Sorry for the diversion into Polywell evangelizing.

Dan Tibbets
Attachments
xxxx.jpg
xxxx.jpg (30.87 KiB) Viewed 3063 times

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

Re: Grid heating by Current vs Ion impacts

Post by Doug Coulter » Wed Dec 14, 2011 2:16 pm

Oh, it's ion impact, no question at all about that. Chris doesn't have a fusor...but sometimes likes an argument.

My cylinder grid is composed of 8 .040 tungsten TIG rods, that are used in welders up to 60 amp currents or so - and they can get incandescent in the middle (about orange) at currents over 30 ma or so. 30 ma passing through tungsten rated to carry 480 amps or so lighting it up? I kind of doubt that.
And they get hot in their middles, where the rays are, not at the end where the current is fed in.

While things do get hotter for the same power dissipation in a (near) vacuum...not *that* much! 40 milliamps doesn't make much voltage drop in a tiny fraction of one ohm...
Why guess when you can know? Measure!

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

Re: Grid heating by Current vs Ion impacts

Post by Chris Bradley » Wed Dec 14, 2011 3:42 pm

Doug Coulter wrote:
> Oh, it's ion impact, no question at all about that.
> My cylinder grid is composed of 8 .040 tungsten TIG rods, that are used in welders up to 60 amp currents or so - and they can get incandescent in the middle

Doug, I do concur that the beautiful images you've posted in the past on this, viz. viewtopic.php?f=18&t=7858#p55817 , are convincing of the opposite case.

However, I would hesitate to pass on '..no question at all..' simply from this observation, because this could (in part) be because as those wire elements closest to the source of the current (and sink of heat) begin to glow hot, that the majority of current could then flow from those parts of the wire due to thermionic emission. I envisage this might then lead to some form of 'thermal runaway' with more current dissipating from the hottest parts.

That is to say, if the heating effect is sufficient such that all the 'fed' current is carried away from the grid before it passes into the ends of that grid, then there would be little current heating there.

If this were an effect, you would expect to see this manifest as an 'asymmetry' in a wire [grid]; that is, a heating effect biased towards the current source, and for it to be increasingly asymmetric with higher temperature.

That is to say, at higher currents it would tend to heat up closer to the current source (allowing for a T^4 (?) increase in current through that wire element closest to the source), and at lower currents (where the general temperature of the wire is less) more current would flow to the 'furthest' end of the grid, therefore the heating would reach further away from the current source. This also would then all depend on the conduction of the heat through the wire, versus energy loss rate via EM radiation, so nor would there be direct, linear relationships in all this to judge. So I would look for 'asymmetric' heating at different levels of grid temperature as a flag to whether to look for more going on than just one mechanism, or not.

In any case, I do not at all mind being called 'wrong' on this! In fact I also have experiences to the contrary of my own posit (In one experiment, I had the centre region of 6 tungsten wires glowing with under a mA running through each. fed asymmetrically from one end, which therefore seems highly unlikely to be from 'current'.) However, is 'ion bombardment heating' the whole story? I suspect there are some circumstances where it is... and there again in others where it isn't!

Summary: My posit is wrong..... except for any exceptions....!!

User avatar
Frank Sanns
Site Admin
Posts: 1587
Joined: Fri Jun 14, 2002 6:26 pm
Real name: Frank Sanns
Location: Pittsburgh, PA USA

Re: Grid heating by Current vs Ion impacts

Post by Frank Sanns » Wed Dec 14, 2011 6:45 pm

The entire reason for my developing a gridless electrode system was due to ion heating and at much much higher powers, the potential for fusion heating.

By moving the electrodes out of the ion path, the "Pillar of Fire" plasma electrodes can be run full out without heating as the current through the base electrodes is paultry and that power alone is ineffective as an incandecent light or heater.

Frank Sanns

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

Re: Grid heating by Current vs Ion impacts

Post by Doug Coulter » Wed Dec 14, 2011 6:51 pm

For what it's worth, thermionic emission requires power *input* to work, else over 100 years of vacuum tube tech is wrong, and we didn't need those pesky A batteries. It cools the emitter, since the electrons are given off with net *positive* energy (a few eV, and that spread they have is an issue in electron microscopes that don't use field emission). No cheating on that one. Electrons released as secondary emission due to bombardment always have less energy than the bombarding particle, which leaves some energy behind in the emitter. There are some nice charts on secondary emission in a few of the books I have here - Kohl chief among them. But also some RCA tube design handbooks and other sources.

There of course is one well known (seeming) exception - magnetron filaments are partially heated by energy from the B supply as electrons swing around and impact the filament after being accelerated by the HV input and giving up some of that input energy to RF for a few orbits. But it's still a case of things hitting the filament making it hot by transferring energy to it mechanically, not I^R losses due to the emission current, which is orders of magnitude too small for that to be much of an effect.
Attachments
KohlSE.gif
KohlSE.gif (103.78 KiB) Viewed 3063 times
Why guess when you can know? Measure!

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

Re: Grid heating by Current vs Ion impacts

Post by Chris Bradley » Wed Dec 14, 2011 7:11 pm

Conceded!

Dan Tibbets
Posts: 578
Joined: Thu Apr 17, 2008 5:29 am
Real name:

Re: Grid heating by Current vs Ion impacts

Post by Dan Tibbets » Fri Dec 16, 2011 12:29 am

That was easy. I didn't expect Chris to give up so quickly. It can't be because of relative argument quality/ credibility of the contributors (read D. Coulter) can it?

In any case, I had wondered (a little bit) if the hot cathode thermally emitted electrons, were at low energy, at the applied voltage , or if there was any practical separation of the electron condition between the point of emission, and subsequent acceleration. Thanks to the graph provided by D. Coulter's link, apparently there is a distinct and measurable difference. That these emitted electrons have a peak at a few eV (~ 20-30,000 degrees C equivalent) and that the glowing electrode may be at a temperature of up to several thousand degrees C, suggests that it is the high energy thermalized tail electrons that escape. Despite being a small portion of the thermalized distribution on the surface of the wire, the height of the peak implies that they make up the large majority of the emitted electrons. Another example of an energy dependant cross section (or rate determining) relationship.

Dan Tibbets

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

Re: Grid heating by Current vs Ion impacts

Post by Chris Bradley » Fri Dec 16, 2011 1:06 am

Dan DT wrote:
> I didn't expect Chris to give up so quickly.

I'm always persuaded when I recognise a better argument than my own! (I don't view the point of a debate to 'win', but that human knowledge would not develop if no-one ever suggested and/or discussed a different point of view.)

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

Re: Grid heating by Current vs Ion impacts

Post by Chris Bradley » Sat Dec 24, 2011 4:46 pm

I'm still not sure I have seen a 'clincher' argument on this yet. I was doing some plasma cleaning on a new central electrode set-up in my magnetic experiment, in which the central electrode is an anode for the purposes of coronal ion generation. It used to look like this:

viewtopic.php?f=12&t=5045#p33338

The insulators were teflon and had carbonised over time and tracked under load, so could no longer hold voltage. So I've replaced them in the last few days with parts I have made using fire cement, and am now pumping it down (it takes days, due to all the stuff in the chamber).

Less than 2 mA is enough to get the tungsten wires glowing, in this set up, and they are at a positive potential, so why are they red hot if not for current alone, albeit only 2 mA? Electrons can't build up enough straight-line speed, on account of the magnetic fields, so unless it's due to negative ion impacts I do not see any other explanation other than 'current', in this particular case.

(note: still plenty of 'contaminants' in the chamber at this point)

...
Attachments
glowing_W_at_2mA.jpg
glowing_W_at_2mA.jpg (166.84 KiB) Viewed 3063 times

User avatar
Frank Sanns
Site Admin
Posts: 1587
Joined: Fri Jun 14, 2002 6:26 pm
Real name: Frank Sanns
Location: Pittsburgh, PA USA

Re: Grid heating by Current vs Ion impacts

Post by Frank Sanns » Sat Dec 24, 2011 6:23 pm

The current is equal everywhere in a circuit. If there is a loop of wire or a long stretch of wire, then the entire length carries the same current. In your device and in a fusor the circuit is not from one end of the wire to the other but is the summation of an infinite number of circuits along the wire.

For simplicity, consider a long wire with no terminiation at one end and HV fed into the other end; a wire sticking straight out from the feedthrough and nothing at the far end. When the power is switched on, a voltage will appear on the wire and current will flow through the plasma that is created around the wire and will flow to the shell in a fusor. All power must come from the feedthrough so the current will be at a maximum AT THE FEEDTHROUGH and fall till the far end of the wire. This ensures that if glow is due to current in the wire, then it must be most incandescent at the feed point since current is at a maximum there. There is also the heat sink of the feedthough so it would be likely that maximum incandescence would be just away from the feedthrough and drop along the length of the wire as fewer and fewer circuits are present up to the open end of the wire where there is only one cirucuit and current would be the lowest at the far end of the wire (or loop).

If the incandescence is due to current flowing through the wire then how would you explain maximum incandescence that far from the feedpoint and far from where currrent and power would be at a maximum?

Frank Sanns

Post Reply