Magnetron ion source for Carl's Jr.

For the design and construction details of ion guns, necessary for more advanced designs and lower vacuums.
Post Reply
User avatar
Carl Willis
Posts: 2841
Joined: Thu Jul 26, 2001 11:33 pm
Real name: Carl Willis
Location: Albuquerque, New Mexico, USA
Contact:

Magnetron ion source for Carl's Jr.

Post by Carl Willis » Mon Dec 04, 2006 9:01 pm

Here is a cross-section diagram of the ion source currently in use on the Carl's Jr. fusor (see viewtopic.php?f=6&t=2648#p12158)

The total cost of the brand-new parts needed to make one of these is probably under $100. The magnets are made by stacking grade N42 cube NdFeB magnets into stacks, then aligning and fixing the stacks into a row with N poles all in one direction and S poles all in the other. The diagram depicts how the magnetic poles should be aligned. Note that these magnets will want to fly apart with great force and care must be taken to hold them together. I use aluminum duct tape with good results.

The major strengths of this design are the simplicity (minor machining and brazing is called for) and the versatility (any of the 5/8" feedthroughs sold by CeramTec can be used with the quick connect, and the position of the end of the stalk can be varied by moving it up and down in the quick connect). The major weakness is that we have viton o-rings and NdFeB magnets, neither of which tolerate the high temperatures that this ion source may generate unless precautions are taken. Another weakness is that this source will not work AT ALL iif it is reverse-polarized (e.g. center electrode negative).

Improving this ion source is the subject of my active work now and I will post updates specifically about this source here.

-Carl
Attachments
is2.jpg
is2.jpg (46.65 KiB) Viewed 3020 times
Carl Willis
http://carlwillis.wordpress.com/
TEL: +1-505-412-3277

Starfire
Posts: 1482
Joined: Wed Oct 24, 2001 6:14 pm
Real name:

Re: Magnetron ion source for Carl's Jr.

Post by Starfire » Mon Dec 04, 2006 9:37 pm

Thank you Carl - much clearer now - it should be possible to build this gun external to a fusor and measure the Ion current by setting up a faraday cup.

User avatar
Carl Willis
Posts: 2841
Joined: Thu Jul 26, 2001 11:33 pm
Real name: Carl Willis
Location: Albuquerque, New Mexico, USA
Contact:

Re: Magnetron ion source for Carl's Jr.

Post by Carl Willis » Mon Dec 04, 2006 11:44 pm

Hi John,

This source, by itself, has no extraction field and so it won't make a beam of ions as a stand-alone device. In the fusor environment, there is a strong electric field to take ions out of the source, and this same field is also good at focusing them into a point at the center of the fusor. Most stand-alone ion guns have an extraction electrode and some other beam optics to suit their application.

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

Jon Rosenstiel
Posts: 1369
Joined: Thu Jun 28, 2001 5:30 am
Real name: Jon Rosenstiel
Location: Southern California

Re: Magnetron ion source for Carl's Jr.

Post by Jon Rosenstiel » Thu Dec 07, 2006 4:13 am

Carl-
CeramTec's prices are fairly reasonable, $55 for the 12 kV feedthrough you have, $45 for the same thing in solid copper, and $60 for an 8 kV model that sports a 0.375" solid copper conductor. Too bad they have a $150 minimum order.

Jon Rosenstiel

User avatar
Carl Willis
Posts: 2841
Joined: Thu Jul 26, 2001 11:33 pm
Real name: Carl Willis
Location: Albuquerque, New Mexico, USA
Contact:

Re: Magnetron ion source for Carl's Jr.

Post by Carl Willis » Thu Dec 07, 2006 4:24 am

Hi Jon,

Yes, too bad about that minimum order. I am fairly certain that KJ Lesker gets their parts from CeramTec, though you pay a bit more for them there. I bought $150 worth of different feedthroughs from CeramTec because I'm sure I will use them all at some point. CeramTec also must have redone their website recently. Used to be you could browse the whole catalog and see prices for everything. Now they have moved to a quotation-based scheme that isn't as useful. But their prices are pretty good in my opinion. Oh: and I would be willing to participate in a bulk buy if anyone gets interested in feedthroughs.

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

Jon Rosenstiel
Posts: 1369
Joined: Thu Jun 28, 2001 5:30 am
Real name: Jon Rosenstiel
Location: Southern California

Re: Magnetron ion source for Carl's Jr.

Post by Jon Rosenstiel » Thu Dec 07, 2006 4:46 am

Carl-

Bulk buy... yeah, I'm up for that.

Jon Rosenstiel

Jon Rosenstiel
Posts: 1369
Joined: Thu Jun 28, 2001 5:30 am
Real name: Jon Rosenstiel
Location: Southern California

Re: Magnetron ion source for Carl's Jr.

Post by Jon Rosenstiel » Thu Jan 04, 2007 10:01 pm

Carl-

I've been fooling around with this also. The NIB magnets I’m using are 0.375” square X 1” long; grade N38SH (max temp = 150C). Initially I tried the NIB magnet "belt". Results were pretty much the same as you reported. Current hungry, and wouldn't strike when biased negative. After several confrontations with the NIB magnet belt (with the magnets usually winning) I gave up on the belt approach and decided to try the negative biasing scheme Takamatsu et al. used.

Basically I copied Takamatsu’s ion source. (But without the water-cooling) Inside diameter of the anode is 35mm and outside diameter of the cathode is 20mm. I machined the cathode out of 6061-T6 aluminum, bored a hole in the center to accept one of my 0.375” square NIB magnets, and machined a cap (press fit) to cover the end. For power I used a NST ballasted microwave oven power supply.

Photo below shows the cathode attached (via not visible set-screw) to an old 30 kV feedthrough I had lying around. Next to the ion source is my D2 gas inlet setup.
Last pic shows the ion source / D2 feed / fusor assy. (Valve to the right is the original D2 feed assy)

To get a “feel” for the ion source I ran 22 tests totaling 41 minutes with the ion source operating at various current levels. The majority of the tests were about 1.3 minutes in duration with the fusor operating at 40 kV and 5 mA. The remaining tests consisted of 3 two-minute and 3 five-minute tests at 45 kV, 20 mA and 1 one-minute test at 50 kV, 30 mA. The fusor was allowed to cool completely between all tests.

Observations: (With fusor operating at negative 45 kV, 20 mA and ion source operating at negative 690 V, 20 mA)
1. Fusor pressure, ion source off = ~12.0 mTorr.
2. Fusor pressure, ion source on = ~11.3 mTorr.
3. Neutron output, ion source off = ~1.4E+06 n/s.
4. Neutron output, ion source on = ~1.8E+06 n/s.
5. Switching on the ion source causes a noticeable increase in fusor power draw.
6. Switching off the fusor has very little effect on ion source power draw.
7. Ion source current draw remains stable down to a fusor pressure of ~ 3.4 mTorr. (But fusor will not operate at that pressure).

After 41 minutes of testing the aluminum cathode and the NIB magnet were in great condition. (Contrary to what I expected).

Added 1/4/07, 17:50:
Link to Takamatsu paper, ($30.00, full version)
http://www.iop.org/EJ/article/0029-5515 ... _1_016.pdf
Link to Takamatsu paper, (free, abbreviated version)
http://www.telegrid.enea.it/Conferenze/ ... /03_02.PDF

Jon Rosenstiel
Attachments
Mag_Ion_Source 018a.jpg
Mag_Ion_Source 018a.jpg (83.81 KiB) Viewed 3020 times
Mag_Ion_Source 006a.jpg
Mag_Ion_Source 006a.jpg (51.49 KiB) Viewed 3020 times

tligon
Posts: 587
Joined: Tue Jan 02, 2007 6:58 pm
Real name:
Contact:

Re: Magnetron ion source for Carl's Jr.

Post by tligon » Thu Jan 04, 2007 11:17 pm

Looks like a reasonable start, similar to some ideas I had at EMC2, but I'm curious about the need at the pressures Joel is working at. Ion sources would seem more useful working below glow discharge.

The end of the D2 tube is shown past the end of the magnets, so that's past the peak magnetic field.

Is the inner tube metal or ceramic? If metal, there is no e-field gradient in side that tube so no reason for ions to light off until they reach the end of the tube. If they bled out the sides, ions would form in that space between the tube and the outer KF tube.

I'm somewhat concerned that generating ions positive with respect to the walls will create losses. Ideally, you want them generated just slightly negative so they never want to return to the walls. Don't let them have enough energy to get out of the potential well.

It might light off at a negative polarity if tickled with a radioactive source, the way cold cathode vacuum gages work. They sell replacement radioactive discs that install on the central electrode of those. That type of gage is called an inverse magnetron.

You could probably water-cool that outer KF tube effectively if needed.

We were working at very low pressures, and what I needed was a region of locally high pressure to create the ions in the presence of a strong field, then some way to let the ions out but not the neutrals. I was looking at making a molecular beam of neutrals passing straight down a tube, ionize what I could, and extract the ions at right angles to the beam. The unionized neutrals would then continue on to a pump port. I don't think we ever tried to build one.

We had a lot of fun using 2.45 GHz power taken from a microwave oven to make ECR along a 875 gauss topology. That creates a zone of frantic electrons that will ionize any neutral that trys to pass it. I suspect a working ion source in one of Dr. Bussard's Polywell (r) machines will eventually need to combine the ECR trick with the molecular beam trick, as the system will work best if there are only fuel ions, not neutrals, in it.

Post Reply