RF ion source sees first light

For the design and construction details of ion guns, necessary for more advanced designs and lower vacuums.
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: RF ion source sees first light

Post by Carl Willis »

Larry,

Thiourea is wonderful stuff. Known to the State of California to cause cancer and reproductive harm, it has unparalleled tenacity on copper and silver oxidation. Your wife will be beaming at the shine on that metal when you're done.

For cleaning up stainless steel after brazing, I recommend vigorous scrubbing with a Scotch Brite pad under concentrated hydrochloric acid with some Whink (2% HF) thrown in. (Those with sensitive skin or paper cuts may enjoy the benefit of some rubber gloves). Then I let the parts sit in a bath of CLR until fully shiny.

-Carl
Carl Willis
http://carlwillis.wordpress.com/
TEL: +1-505-412-3277
User avatar
Steven Sesselmann
Posts: 2127
Joined: Wed Aug 10, 2005 9:50 pm
Real name: Steven Sesselmann
Location: Sydney - Australia
Contact:

Re: RF ion source sees first light

Post by Steven Sesselmann »

Richard,

Thanks for clarifying "differential pumping", I should have known this, as I have built an ion gun with a 1.5 mm nozzle for ion extraction myself

I guess if Carl had said "differential pressure" rather than "differentially-pumped", I would have been less confused, because I imagined two pumps being used.

Steven
http://www.gammaspectacular.com - Gamma Spectrometry Systems
https://www.researchgate.net/profile/Steven_Sesselmann - Various papers and patents on RG
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: RF ion source sees first light (DATA)

Post by Carl Willis »

Some low-power operational data is attached, and a couple more (better) pics. All this operation is still in air rather than deuterium.

I collected some beam current numbers via my Keithley 610 electrometer configured as an ammeter in the target power supply ground return. The beam current as a function of extraction voltage (with RF forward power held constant) is on top, while the beam current as a function of RF forward power (with extraction voltage held constant) is on bottom. (How much of the registered beam current is due to backstreaming electrons? I don't know, but the setup is designed in some simple ways to cut down on the secondary electron current off the target. It is made from graphite, not a very good secondary electron emitter. It has a 30-degree cone and a 1/4" drill hole along its centerline where the beam strikes, geometrically restricting the escape of electrons.) The anode is at 9 kV throughout.

Hard to know what conclusions are really suggested by the measurements here. The current-versus-voltage curve is upwardly monotonic like the Child-Langmuir law predicts it should be, but it either underperforms the V^(3/2) proportionality of the Child-Langmuir law at low voltages or overperforms at high voltages. Whatever the case, this trend clearly suggests that I need to kick the extraction sauce up a notch. The feedthrough for it is rated 5 kV, so that's probably my ceiling without resorting to a potted connection. The RF power curve has a bump in it at about 50W forward power. This probably indicates a new type of discharge mode becoming established that is a better source of ions. When the source is run in deuterium, beam current should go up by at least a factor of three (again a prediction of the Child-Langmuir law). I'm predicting a yield of 2 mA deuterons at 75 W / 5 kV extraction.

The photos both show operation at 75 W RF, 3 kV extraction, and about 300 uA. The beam is easy to see. Base pressure is about 0.1 mtorr and the gas valve is barely cracked. I really cannot do any better than that with the pumping system I have.
Attachments
beam_3.JPG
ion_source_data.jpg
ion_source_5.JPG
Carl Willis
http://carlwillis.wordpress.com/
TEL: +1-505-412-3277
User avatar
Richard Hull
Moderator
Posts: 14991
Joined: Fri Jun 15, 2001 9:44 am
Real name: Richard Hull

Re: RF ion source sees first light (DATA)

Post by Richard Hull »

Carl,

This is really superb work! I am printing all this out for inclusion in my ion gun notebook. Thanks for the detailed and clear data presentation and the pix, which, as always, are worth a thousand words. I will follow this with rapped awe.

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

Re: RF ion source sees first light (DATA)

Post by Chris Bradley »

This is a very high standard of work. The ion beam's edge is so sharp and collimated I would presume there is very little electron backstreaming, at least in the actual beam width itself. I also wonder whether the beam coherence is 'too' good that it might actually over-bias the fusor's operation towards the injection axis. (Not sure there are any particularly bad consequences from that, though.)

Maybe you could also consider driving a ring-cathode with this beam rather than a spherical grid and see if it generates some sort of self-resonant drive along that axis.

Whatever neutrons you get out of a fusor will become the new standard, I am sure.

best regards,

Chris MB.
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: RF ion source sees first light (DATA)

Post by Carl Willis »

Thanks Richard and Chris for the compliments.

This is a work in progress and very much a learning experience. Some things are about to change:

(1) The $100 glass-to-metal adapter will be replaced with a $1.00 standard 19 x 150 mm test tube (I ordered a pack of 72 off eBay). This eliminates some glaring mechanical fragility issues seen in the current version and when sputtering products build up too much, I can just toss the test tube and swap in another. This change also alleviates difficulty changing RF antennae and magnets.

(2) Since the test tube will not have a place to feed in gas, the gas will be routed through the side of the upper CF flange into the annular inter-electrode volume. Should be no differences in behavior caused by this change.

(3) I found that the surface of the positive (grounded) extraction electrode sputters a lot in the RF discharge. To solve that problem, I placed a Teflon washer on top of it. Teflon is a pretty good solution but it also sputters some. So I ordered a few #8 ceramic washers from Small Parts.

(4) The target (or rudimentary Faraday cup) may not be able to control its secondary electrons well enough. A new cup is being designed that will have a shroud around it, made from pipe caps. This will intercept most electrons, and will take a bias negative to the cup if electrically set off from the cup with either resistors or Zener diodes. Equilibrium will be established such that most electrons formed in the cup are thrown back into it by the negative self-bias of the shroud.

>I also wonder whether the beam coherence is 'too' good that it might actually over-bias the fusor's operation towards the injection axis.

The ion-source-enhanced beams in a fusor should be intense and obvious, ruining the normal symmetry of the "star mode" discharge. George Miley's presentation on a fusor with an ion source named ILLIBS is here,

http://fti.neep.wisc.edu/tofeprogram/oral/O-I-6.5.pdf

containing several photos that are a likely indication of what the discharge will look like. My previous magnetron ion source likewise resulted in asymmetric discharges, pictured here in black-and-white:

viewtopic.php?f=6&t=2652#p12162

-Carl
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: RF ion source sees first light (DATA)

Post by DaveC »

Nice work once again, Carl!

I think the Child Langmuir prediction is really only a general relationship. There are a few other factors that figure in, which can modify the functionality. Your data looks quite good to me on that score.

About the beam "coherence" being possibly too good. I'm not sure I understand the use of coherence with regard to the ion beam.... If you mean well focused and well defined ( and it certainly is that), then I would expect this portion of the fusor's ion supply to have a higher fusion yield, at the center.

Thanks for sharing your work.


Dave Cooper
lukeantwis
Posts: 1
Joined: Fri Jan 09, 2009 6:09 am
Real name:

Re: RF ion source sees first light

Post by lukeantwis »

Hi Carl,

Im attempting to construct an RF source very similar to the one you've shown here, however im having a little trouble understanding exactly how to clamp the plasma with the use of the magnet, specifically, what exactly the magnetic field looks like, ie, field lines etc etc. Could you elaborate on the exact configuration of the magnet you've used?

Many thanks

Luke
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: RF ion source sees first light

Post by Carl Willis »

Hi Luke,

Ideally the magnetic field is axial in this design. I have used two sets of magnets, the first being a pair of AlNiCo magnets from a microwave oven magnetron (seen in these pics I think), and the second being a ring of prismatic NdFeB magnets held together with aluminum duct tape that can be wrapped around the discharge tube. The NdFeB magnets result in improved performance. They're also considerably stronger. As to what the actual field configuration looks like, I haven't really worried about this so I don't know. The positions of the magnets and the RF antenna have major effects on the discharge impedance, stability, and ion source beam current. The RF antenna likes to be about 1" above the magnets, and the magnets should be as close to the extraction aperture as possible.

Hope this helps.

-Carl
Carl Willis
http://carlwillis.wordpress.com/
TEL: +1-505-412-3277
Tyler Christensen
Site Admin
Posts: 551
Joined: Mon Feb 23, 2009 9:08 pm
Real name:

Re: RF ion source sees first light

Post by Tyler Christensen »

I'm almost done building one of these, the magnets should be placed with the north pole pointing upwards, towards the coil, correct? Just want to make sure I'm right on that before getting them stuck between fittings.
User avatar
Doug Coulter
Posts: 1312
Joined: Sun May 27, 2007 3:18 pm
Real name: Doug Coulter
Location: Floyd, VA, USA
Contact:

Re: RF ion source sees first light

Post by Doug Coulter »

Truly beautiful Carl!

Finally someone is publishing numbers close to what I've been getting on the microwave version of this design, not needing ridiculous RF power inputs to light it off, and with good output.

So I feel a bit vindicated myself ;~) Thanks!

I've found that the fusor really likes a good source of monatomic ions, which I guess no one yet has measured the ratios of, but the RF sources are known to excel at. I guess one of us is just going to have to put a weak magnet in the ion beam and a ZnS screen back there to see the rough ratios of monatomic vs diatomic ions produced.

I find it especially cool that you can run the fusor at pressures below normal, where it doesn't "take off" and become its own best (or at least most prolific) ion source, and in that mode you can dial the main fusor current up and down easily via changing the ion extraction voltage. My particular version has a too-long extractor electrode, so above a certain extraction potential, the electrostatic lens it makes (like it or not) has too short a focal length and bashes the nice ions into its own walls after they pass the focus. At the peak output I am getting fusor currents of about 2.6 ma at 50kv no matter how low the operating pressure (down to the mentioned 1.6e-6 mbar), which goes to zero at low pressures if I turn the ion extractor off -- not as slick as a faraday probe, but indicative of good operation.

I will posit the following idea that may make yours work even better -- I found with mine that getting a uniform H field wasn't trivial (and I never quite managed that, in my case about 20 gauss out of ~980 over the length of the glow was best I could do), but that it's moderately critical around the electron cyclotron resonance area. If you really nail that -- you'll be able to run it at lower enough pressures with good performance to not need differential pumping, or at least mine does, and will stay lit down to 1.6 e-6 mbar or so....nicely low gas flow rates at that point and no need for better pumping.

I did that on mine feedforward via a hall effect magnetometer and some shims to space the magnets (or more properly, hold the spacing I set). This was however confirmed to be the best setting by also just empirically tuning that while in operation for best ion current and lowest possible operating pressure -- pretty much got the same answer by theory or experiment, which is always gratifying.

Good show all around! This helped our fusor effort probably more than any other single thing we've tried so far. Hopefully you will find the same on your gear.

I intend to fool with the geometry a bit more on mine to correct that issue with over-focus and see if that helps. In my case I'd used too many thick adapter flanges which is why the puller had to be too long, but that can be fixed up. I'm pretty sure that there's a lot more output possible with that and by better control of what the H field does after the glow area (so it doesn't also bash ions into the tubing walls). In my case, I have the RF normal to the tubing axis, and the H field along it.

Tyler -- the polarity of the H field shouldn't matter -- just the orientation vs the RF, you want E and H fields crossed at about 90 degrees to get the cyclotron resonance, and the field strength correct for the RF frequency in use. I doubt it matters whether the electrons are going clockwise or counter clockwise, as long as they are going around. The fields required shouldn't affect the actual ions too much, due to their much higher weight.


Edit:
On cleaning copper: The firearm business has this as one of its more diverse and profitable lines, as bullet gilding metal (Cu-Ni alloy) ruins accuracy when it gets inside a rifle bore by being rubbed off the bullets. The stronger cleaners tend to have a high component of ammonia (eye watering levels), which used alone works fantastically here. Most of the gun supplies also have some other stuff in there which itself can be hard to get back off -- various oily stuff. But 28% ammonia (diazo machine grade) really whips up on Copper fast and is easy to clean off once you are done -- simple water rinse, which I do in an ultrasonic cleaner, but mainly because I have one and that makes it easy -- plain old water spray would do as well.

In a pinch I've found that plumbing flux (ammonium chloride + zinc chloride + HCl in liquid form) also really gets the crud off quick, and cleans up easily as well. That last would almost have to win the "cheap" award, as it's about a buck for a large bottle. It's also easier on the eyes and throat than the concentrated ammonium hydroxide by quite a large factor, but more difficult to really rinse off fully.

Any of these that etch copper tend to get into the inter-grain space and continue their action for awhile, so no matter what you use, a really thorough rinse is a very good idea. That effect makes the pure ammonia a little better in real life use as it's easier to get rid of.
Why guess when you can know? Measure!
User avatar
John Taylor
Posts: 97
Joined: Wed Jun 22, 2011 1:43 pm
Real name:
Location: Dardanelle, Arkansas

Re: RF ion source sees first light

Post by John Taylor »

Carl,
I know this is an older thread but I wondered how this particular ion gun performed as far as sputtering goes. Has the copper given you any problems?
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: RF ion source sees first light

Post by Carl Willis »

Hi John,

Yeah, copper sputters like crazy. It's also less than ideal for hydrogen plasmas, being established as a particularly good catalyst for molecular ion formation.

Copper eventually (10+ hours of operation or so) coats the downstream interior of the glass discharge tube to the point of opacity. It probably has a lot to do with degradation in performance over time. For this reason, just last month I built a new version of this source that, among other improvements, gets rid of the copper entirely and replaces it with stainless steel. I also eliminated some braze joints. Sometime "soon" (maybe this weekend, maybe next month, who knows) I will post an update here about the new source. I have not operated it yet.

-Carl
Carl Willis
http://carlwillis.wordpress.com/
TEL: +1-505-412-3277
User avatar
Chris Bradley
Posts: 2930
Joined: Fri May 02, 2008 7:05 am
Real name:

Re: RF ion source sees first light

Post by Chris Bradley »

Would it not be better to pick aluminium - the lowest sputtering of all common materials?
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: RF ion source sees first light

Post by Carl Willis »

Chris,

I can't think of how you would practically follow this blueprint with aluminum. It calls for a number of braze or solder joints using off-the-shelf insulators, feedthroughs, etc. and uses standard stainless CF hardware. That doesn't mean aluminum is beyond consideration if a more custom approach is feasible. The original Kiss-Koltay ion source upon which mine is based does in fact use extensive aluminum.

-Carl
Carl Willis
http://carlwillis.wordpress.com/
TEL: +1-505-412-3277
User avatar
John Taylor
Posts: 97
Joined: Wed Jun 22, 2011 1:43 pm
Real name:
Location: Dardanelle, Arkansas

Re: RF ion source sees first light

Post by John Taylor »

I would like to see how a stainless version would perform. will you TIG weld the places that are soldered on the current version?

I would like to replicate your design in stainless as well. Would you be willing to share your dimensions so I could attempt to machine the parts to your already proven design?
Post Reply

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