The "wire" ion source

For the design and construction details of ion guns, necessary for more advanced designs and lower vacuums.
User avatar
Richard Hull
Moderator
Posts: 14991
Joined: Fri Jun 15, 2001 9:44 am
Real name: Richard Hull

Re: The "wire" ion source (DATA)

Post by Richard Hull »

Nice work Carl! Thanks for the report

I am just about to dismantle the fusor IV setup and I just might try the long wondered at W needle ionizers I have mused at for so long.

6 to 8 - .020" diameter W needle electrodes of about .5-inch length affixed to and about the interior shell.

I am sure the pressure voltage regime will be bizarre and all the normal operational skills are totally out of the window. Will fusion numbers alter? I guess we will see.

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
The more complex the idea put forward by the poor amateur, the more likely it will never see embodiment
Richard Hester
Posts: 1519
Joined: Thu Jul 05, 2001 12:07 am
Real name:

Re: The "wire" ion source (DATA)

Post by Richard Hester »

Four of these wire sources (long) arranged around a cylindrical inner grid might be interesting. I have appropriately sized tungsten wire. I envision 4 pieces of SS tubing opened out similar to the first pictures. but with a much longer wire. This is far easier than another scheme I was considering using heated tungsten filaments and a big solenoid to persuade the electrons to linger about and make ions.
User avatar
Doug Coulter
Posts: 1312
Joined: Sun May 27, 2007 3:18 pm
Real name: Doug Coulter
Location: Floyd, VA, USA
Contact:

Re: The "wire" ion source - link

Post by Doug Coulter »

Thanks for this, Carl, these look really interesting and worthwhile.

You can get the papers here:
http://www.coultersmithing.com/forums/v ... f=28&t=520
Why guess when you can know? Measure!
DaveC
Posts: 2346
Joined: Sat Jun 30, 2001 1:13 am
Real name:

Re: The "wire" ion source

Post by DaveC »

Just caught up with this thread. Very nice work Carl.

The Wire-Ion Plasma source (WIP gun) is a work horse method to get lots of electrons or ions fairly simply. Some good papers have been published by Hughes in the 60's and 70's.

The ion generating voltage is determined from the Paschen Curve for the Pressure-Distance value of the ionizer geometry and gas pressure. With the grounded screen, the exiting ions are then under the influence of the cathode (negative) field exclusively. And conversely, the ionizer is more or less free of any cathode potential influences.

With a current limited supply feeding the ionizer wire, you can measure ionizing voltage versus pressure and plot the "real" paschen curve for your particular configuration.

A WIP gun built at work a few years back had dual "several" by about 20 inch beams of He ions. It also used a fine tungsten wires such as the one you show. It was a pretty robust source. We used the configuration to make a secondary emission electron gun. Electron operating parameters were generally above 100 kV at around 100 mA.

Pressures were nominally in the -3 to -4 Torr range.

One down side is sputtering of the ground screen by the fraction of ions not exiting the source. Interestingly, cathode sputtering was not an issue. The high field probably causing all sputtered metal atoms to return to home quickly.

Dave Cooper
Dustin
Posts: 105
Joined: Mon Jul 26, 2010 8:40 pm
Real name:

Re: The "wire" ion source - link

Post by Dustin »

Nice work Carl.
Do you think it would be possible to increase efficiency by an irregular,
perhaps octagon outer wall?
This may reduce current consumption/ion current.
Steve.
Attachments
ionsce.jpg
ionsce.jpg (115.26 KiB) Viewed 9204 times
DaveC
Posts: 2346
Joined: Sat Jun 30, 2001 1:13 am
Real name:

Re: The "wire" ion source - link

Post by DaveC »

The field around the ionizing wire controls the emission.

The shell geometry figures only in a minor way, as long as we aren't talking about sharp edges, points and such. So the easiest outer container shape will work.

Ion generation is by avalanche multiplication outward from the wire. The generalities of inception voltage are indicated by the Paschen curve for an electrode system at the pressure "p" being used, with a wire screen distance used for the "d" of the "pd" parameter.

While not exact, since most Paschen data is for either sphere-sphere or planar electrode systems, these data are usually close enough to be "exact" for physics work.

As suggested above in my post yesterday ( which got into the wrong part of the thread, sorry about that) you can plot your paschen curve for your particular electrode system with some V-I measurements of the inception points of discharge.

Dave Cooper
User avatar
Chris Bradley
Posts: 2930
Joined: Fri May 02, 2008 7:05 am
Real name:

Re: The "wire" ion source - link

Post by Chris Bradley »

Positive coronal discharges (that is to say, *if* this device follows the same principles) are much more complicated than just an avalanche discharge. It would be interesting to establish if there is a high frequency cycle with this, as there is with positive corona discharges, in which the region around the positive central electrode repeatedly sheds its ions then the electrons left behind act to re-ionise the gas again and the cycle repeats.

That is not to say that the 'principles' of the initial Paschen breakdown do not apply, but that the 'd' in the 'pd' term may be defined rather by the geometry of what amounts to a 'virtual plasma electrode' around the wire, rather than the physical wire.
User avatar
Carl Willis
Posts: 2841
Joined: Thu Jul 26, 2001 7:33 pm
Real name: Carl Willis
Location: Albuquerque, New Mexico, USA
Contact:

Re: The "wire" ion source - link

Post by Carl Willis »

In response to Dave and Steven and Chris:

Most of the attested previous uses of these kinds of discharges are in secondary-emission electron guns of the type Dave was describing, high-power pulsed electron sources and lasers. In fact, the pair of Gueroult papers referenced above--and particularly the one from this year in Plasma Sources Sci. Technology--are the first occasion in my perusal of the literature where interest was evoked in this device as a high-quality source of ions specifically. It has a very convenient geometry and simple ruggedness that suggests reliability in sealed neutron generators, hence my attention.

Published Paschen curves aren't predictive in this situation where anode radius is much smaller than electron mean free path. They also don't capture the unique nuance of the normal-to-obstructed mode transition that is critical to the behavior of this device. As seen in my previous post, in the earlier Gueroult paper, and the attached figure below from Druyvesteyn and Penning (in their seminal paper in Rev. Modern Physics 12 (1940), p 87), there is a p-V curve at constant current that does have generally the same shape as a usual Paschen curve, and undoubtedly for the same reasons. But a number of assumptions in Paschen's Law aren't valid here as the Penning paper makes clear. The discharge behaves dramatically differently depending on electrode polarity, and in order to reap the low-V, low-p benefits that interest someone in a neutron generator, the wire MUST be positive. In the obstructed mode, the cathode diameter arguably plays an insignificant role, but as Gueroult shows in his earlier paper, this diameter matters in the diffuse mode where the "Child-Langmuir limiting current is a function of the interelectrode gap dR" and so forth.

-Carl
Attachments
pd_wire.jpg
pd_wire.jpg (56.12 KiB) Viewed 5323 times
Carl Willis
http://carlwillis.wordpress.com/
TEL: +1-505-412-3277
DaveC
Posts: 2346
Joined: Sat Jun 30, 2001 1:13 am
Real name:

Re: The "wire" ion source - link

Post by DaveC »

Chris - no real argument, in general with that. The mechanism IS more complex, if rf is being used.

We used a single 0.25 mm diameter ionizer wire (Tungsten), with a ballasting resistor, and a variable DC supply that had current and voltage control modes.
Changing the ionizer -screen distance affects the combination of voltage and pressure needed to initiate ionization.

At a low enough pressure (gas density) there will be little or no avalanching, hence no need for the ballast resistor, yet ions will still be produced.

Running the ionizer on AC usually doesn't hurt anything, but one half of the cycle may not produce ions, depending on gas pressure and geometry, again.

The obvious AC working example is, of course, the neon sign, or fluorescent lamp.

Dave Cooper
User avatar
Chris Bradley
Posts: 2930
Joined: Fri May 02, 2008 7:05 am
Real name:

Re: The "wire" ion source - link

Post by Chris Bradley »

> The mechanism IS more complex, if rf is being used.
Dave,

An 'RF' (typically reported as units MHz) oscillation is set up in a positive corona discharge when a DC current source is applied. It is a naturally occuring frequency, according to the way the ions are shed from the region around the central electrode. Positive corona discharges are reported as dis-continuous DC, they pulse.

In my experiment, I believe* I have detected such resonant effects but at slower than the MHz range reported in papers on the subject (but that do not have a magnetic field, unlike mine). *[I'll have to re-visit this to rule out other mechanisms or background interferences.]
DaveC
Posts: 2346
Joined: Sat Jun 30, 2001 1:13 am
Real name:

Re: The "wire" ion source - link

Post by DaveC »

Good comments, Carl.

The additional information adds to fill in the picture. Do you have the pressure in the Druyvesteyn-Penning paper? The tests I ran were able to move out to right some distance past the Paschen minimum, until the DC supply ran out of voltage... (around 2kv).

The point regrding curve shapes, that you make is most important. These classic curves come with several important assumptions, that often are not realized in actual systems.

Our ionizer geometry was roughly cylinder-square. The grounded screen had a several mm open mesh, and it and the solid chamber walls were several cm away from the fine tungsten wire. So this was only somewhat like a coaxial cylinder setup, with a D/d ratio of maybe 100.

The ions travelled about 0.25 meter to the main electron beam channel, entering at about 45 deg angle, where the strong local cathode field drew them in on axis.

I was able to model (approximately) the paths of both ions and electrons using Simion 7.xx treating both as non interacting. (I also tried unsuccessfully to use Vector Fields for a more refined analysis, but the mesher choked repeatedly on the relatively complicated geometry of the ionizer screen. Simion's transparent screen is a very, very useful feature. David Dahl knew his stuff, when putting this software together.

One interesting result I remember getting was from an attempt to measure incoming ion energy. By adjusting the bias on a collector plate in the electron channel (cathode off) until current went to below 1 nA, a rough ion energy spectrum was developed. I'd have to go dig up the test results, now, but I recall ion energies (He) were pretty large, perhaps up to half the voltages in the ionizer.

This suggested the ions at 10-4 Torr pressures were fairly long lived, and could travel significant distances ( and around corners). This probably is not what happens at the higher pressures in the fusor. Again, as you say, it's mean free path.

We always ran with a positive wire ionizer, because the ions were for the SEE (secondary electron emission) gun application.

Our pressures had to be quite low, in order that main e gun did not itself go into glow discharge... (which did happen from time to time).


So much for the anecdotes....

Dave Cooper
User avatar
Chris Bradley
Posts: 2930
Joined: Fri May 02, 2008 7:05 am
Real name:

Re: The "wire" ion source - link

Post by Chris Bradley »

In the limited sample diameters of electrodes I have tried out, there also seems to be a 'cut-off' point when the electrode diameter gets bigger than some value between 1 to 2.5 mm. Up to then, electrode diameter appears to play only a very small part in the voltage (600 to 900V - with respect to an intermediate electrode positioned at the ends of the volume) at which the thing lights up. I moved to 0.2mm wire (from the picture I show above, with 1mm diameter electrodes) and was surprised to find that it made very little difference to the voltage at which a corona was initiated. At ~2.5mm diameter, it just wouldn't light up at all, at seemingly any voltage (at least, below 10kV).

My preliminary conclusion is that you need a certain voltage gradient to get the process underway, but if the diameter gets too large then the voltage gradient does not drop off (with radius) quick enough, such that ions are repelled from the region before they have a chance to sustain a discharge through further ionisations. This is just a 'working hypothesis'.
DaveC
Posts: 2346
Joined: Sat Jun 30, 2001 1:13 am
Real name:

Re: The "wire" ion source - link

Post by DaveC »

Chris-

Quite likely you moved to a value on the Paschen curve whose breakdown voltage (or field) exceeded yoursupply capability for the pd product of the gas you were using.

Away from the Paschen minimum, the voltages climb with either a higher or lower pressure for a given electrode geometry.

The natural oscillations that I've measured in long ( 1meter long) plasma discharges were in the 60 kHz range. They occurred towards the low end of an unassisted mechanical pump's pressure, (guessing now at perhaps a few tens of microns maximum).

They coincided with the appearance of plasma bunchings that were approximately spherical. I've called them "ion acoustic" oscillations in past references, simply because the plasma bunches coincided with the nodes of the measured oscillation acoustic frequencies. I used a shielded piezo-electric membrane to detect the acoustical) signal.

I need to set up this apparatus once again, and this time take and post photos along with scope traces of the acoustical signal. It was an interesting experiment.... sometime soon, I hope, when the shop-rebuild project plateaus.

Dave Cooper
Dustin
Posts: 105
Joined: Mon Jul 26, 2010 8:40 pm
Real name:

Re: The "wire" ion source - link

Post by Dustin »

Sorry Dave, I think I have been unclear,
In very low pressures ie:the mean free path being greater than the device diameter,
I was hoping to illustrate that an irregular outer shell may extend electron oscillation lifetime.
As a circular outer shell, all fields focus the electrons on the wire, so an electron has a relatively high probability of hitting the wire.
By changing the shell shape, you can reduce this probability by defocussing and promoting eliptical orbits.
Steve.
ab0032
Posts: 86
Joined: Sun Oct 23, 2011 2:50 am
Real name: Alexander Biersack

Re: The "wire" ion source

Post by ab0032 »

Would one also get a positive corona discharge if the center wire was at zero potential or perhaps even negative and the outside more negative, if one wanted ions that start their existence at a certain negative potential?

Could one use a grid or mesh on the outside if one is not interested in sending the ions only in one mayor direction?
User avatar
Carl Willis
Posts: 2841
Joined: Thu Jul 26, 2001 7:33 pm
Real name: Carl Willis
Location: Albuquerque, New Mexico, USA
Contact:

Re: The "wire" ion source

Post by Carl Willis »

For the purpose of establishing the wire discharge, all that matters is that the wire is positively charged relative to the outer cylindrical surface. The potential relative to any external point of reference is irrelevant.

You can certainly extract ions over the entire 360-degree azimuthal range, provided you provide the necessary mechanical support for the extraction grid (or slits) and you keep any external fields from penetrating too far into the trapping region.

-Carl
Carl Willis
http://carlwillis.wordpress.com/
TEL: +1-505-412-3277
Post Reply

Return to “Ion Gun Design and Construction (& FAQs)”