Ion source working, 49 bubs / 3 min!

For posts specifically relating to fusor design, construction, and operation.
Post Reply
User avatar
Carl Willis
Posts: 2841
Joined: Thu Jul 26, 2001 7:33 pm
Real name: Carl Willis
Location: Albuquerque, New Mexico, USA
Contact:

Ion source working, 49 bubs / 3 min!

Post by Carl Willis »

My magnetron ion source is improved and back in business after a hiatus due to a meltdown of the first one. Operation is much smoother and highly controllable with it. I am still using a coaxial configuration, but this time I am simply using the 1/4" SS tube on the CeramTec feedthrough as the magnetron anode. Gas is still delivered through this tube, although I think the pressure difference between gas at the tube outlet and that in the middle of the fusor is insignificant. I do know that at higher currents, the magnetron anode heats up to a reddish color that can be seen reflected in the chamber due to electron bombardment. I will have to have some more experience before I decide if I need a thicker tube or a water-cooling loop on the source, which overall does get quite warm. The discharge produced while the maggie is the primary source of ions lacks the beautiful symmetry of "star mode." Instead, the glow is predominantly along the diameter defined by the maggie anode. Not too surprising. But this discharge is a hot neutron producer, see numbers below.

Furthermore, Jr. has a new grid, this time a 2.25" OD grid, 5 loops of 26-mil tungsten wire. Construction was by the Rosenstiel technique with one modification (a little clip of stainless wire at the "north pole" of the grid to hold it in alignment). The new grid was not necessary at the power levels involved so far, where even the former 1.5" stainless grid did not reach red heat. However, I had the chamber open to clean up slag and chips from the last ion source casualty, and figured I'd go ahead with the mod.

The title gives the neutron yield this evening: 49 bubs / 3 minutes is about 300,000 n/s with Carl's Jr. This was accomplished at 50 kV / 4.5 mA. I don't yet know the magnetron current (have to install a precision dropping resistor in the power supply return lead). But as I said, the ion source is an excellent controlling mechanism. I do know that the anode glows a reddish color due to electron heating, and I can see reflections on the inside of the chamber.

Pics attached; closeup of ion source showing NdFeB magnet "belt" (made on a strip of Al flashing, magnets held down with Al duct tape). This belt wraps around the QF half-nipple to provide an axial field. The feedthrough is held by its long weld-lip in a properly-sized quick-connect. One advantage of this technique is that the position of the anode tip can be varied easily without opening the vacuum. In the last photo is the water-cooling pump thrown in for good measure. I have no specs on it since it is a surplus pump.

-Carl
Attachments
ionsrc.JPG
bigpic.JPG
pump.JPG
Carl Willis
http://carlwillis.wordpress.com/
TEL: +1-505-412-3277
wayne
Posts: 40
Joined: Mon Apr 10, 2006 3:45 pm
Real name:
Contact:

Re: Ion source working, 49 bubs / 3 min!

Post by wayne »

Carl-Great job.

What is the copper tube on the left?
Could I get more information on the magnetron source?
With the cap. manometer at 45 deg. is this a problem?
What is the operating pressure?

wayne
Starfire
Posts: 1482
Joined: Wed Oct 24, 2001 2:14 pm
Real name:

Re: Ion source working, 49 bubs / 3 min!

Post by Starfire »

Carl, be careful with the magnetron - if the load and wave guide are not correctly matched it will radiate and proximity to it can result in severe eye damage and cataracts

- THIS WILL OCCUR AT VERY LOW RF RADIATION LEAKAGE LEVELS -

A general warning to any experimenter with Magnetrons. The eye's are particularlly suseptable to permanent damage by RF at the 2.4Ghz of microwave oven magnetrons and it takes only one mistake. It ain't worth it to go blind if you don't know what you are doing - we need a general FAQ on this before folk start to play.

Permanent damage can occur in less than a second of exposure - if your eyes feel rusty when working with an active Magnetron - it may be too late.
UG!
Posts: 116
Joined: Thu May 19, 2005 4:21 pm
Real name:
Contact:

Re: Ion source working, 49 bubs / 3 min!

Post by UG! »

i second Johns call for a magnetron or general RF FAQ. i have been considering investagating RF ion generation and high power ion resonance in the fusor, but need more information on keeping all that RF where its suppost to be, instead of in my eyes, turbo pump, computer systems and instrumentation etc.

also great work Carl! :)

Oliver
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: Ion source working, 49 bubs / 3 min!

Post by Carl Willis »

I'll try to briefly explain the ion source a bit more.

First of all, this is not a cavity magnetron that generates or uses RF in any way. It is a DC magnetron ion source. That is what this kind of crossed-field device is called. Cavity magnetron oscillators are a subset of this general category of magnetron devices and are not what I am using.

A magnetron ion source works by trapping electrons in E and B fields at right angles to each other, causing them to travel in long helical paths and hence be more effective at ionizing the background gas. It's a very simple concept.

My ion source has only three parts: the inner electrode (the anode); the outer electrode (the cathode, a QF-25 half-nipple that is part of the chamber); and a magnet that produces an axial field in the QF-25 half-nipple. The magnet is made of a bunch of little neodymium magnets I picked up cheap on eBay. They must be held close to one another with their north poles all at one end and their south poles all at the other, so naturally they want to fly apart with great force. Aluminum duct tape is used to ensure that they don't, and the whole set of magnets is wrapped around the QF-25 like a belt. Now the same thing could be accomplished by putting a single ring magnet like those in microwave ovens around the QF-25. However, you'd have to put the magnet on before welding the QF-25 nipple into the chamber, and it would never be able to come off afterward. The belt idea allows the magnet to be moved, changed, etc. So the feedthrough conductor is the anode, the QF-25 is the cathode, and there's a magnetic field acting on the gap between them to trap electrons that otherwise would just sail radially from one to the other.

Did I do a good enough job explaining this?

Oh-Wayne, the "copper tube on the right" is the deuterium feed. It's got a metering valve, then you can see where it enters a tee and meets the valve for the vacuum clean-out line. Above this tee is a glass insulator held in little quick-connects that allows gas to flow through but isolates the ion source HV. After this is the stainless capillary, which is brazed into the ion source feedthrough.

-Carl
Carl Willis
http://carlwillis.wordpress.com/
TEL: +1-505-412-3277
Jon Rosenstiel
Posts: 1494
Joined: Thu Jun 28, 2001 1:30 am
Real name: Jon Rosenstiel
Location: Southern California

Re: Ion source working, 49 bubs / 3 min!

Post by Jon Rosenstiel »

Carl- I’ve been thinking of adding a magnetron ion source to my fusor, but so far have lacked the knowledge on how to proceed. Your explanation helped tremendously, thanks.

I do have a couple of questions though…
1. Does the gas exit through the end of the anode?
2. How far does the anode extend into the fitting in relation to the magnets?
3. Power requirements of the ion source?
4. The N & S poles must be at the ends of the magnets?

Jon Rosenstiel
Starfire
Posts: 1482
Joined: Wed Oct 24, 2001 2:14 pm
Real name:

Re: Ion source working, 49 bubs / 3 min!

Post by Starfire »

Yes - a great job Carl - I stand corrected - I had assumed it was a MO cavity Magnetron - the caution will not be wasted

I like the belt idea - quite inovative - what is the anode voltage?
What beam current are you getting?
***** I should read the threads more or make a bigger effort to understand what I read :)), sorry Carl, I just researched back to your explaination =

viewtopic.php?f=6&t=2620#p12130
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: Ion source working, 49 bubs / 3 min!

Post by Carl Willis »

Hi Jon,

1. Right now the gas does exit through the end of the anode. I've thought of plugging it and putting holes around the circumference so that the gas will be closer to where I think the discharge is, but for all I know the discharge is actually at the end of the anode so I have refrained from this mod.

2. The end of the anode is flush (+/- 1 mm maybe) with the inner surface of the fusor, just 1/8" further from the bottom of the magnet.

3. Right now I am using a variac-controlled Plastic Capacitors 15 kV / 5 mA power pack. However, I think it has been drawing more than 5 mA and hence operating beyond the design envelope of this unit, so I plan to move up to a 10 kV / 10 mA Universal Voltronics power pack. On the other hand, the highest voltage I have used is probably ballpark 5 kV. I'm sorry I don't have more precision. Good measurements are on their way.

4. The N and S poles are at the ends of the magnets (i.e. the ends of the cylinder they form around the QF).

It turns out I am not the first to try using a magnetron ion source on a fusor. There is an excellent paper by Takamatsu, Masuda, et al., "Intertial electrostatic confinement fusion device with an ion source using a magnetron discharge," Nucl. Fusion 46 142-148 (2002). Very nice writeup and data showing neutron yield with and without the ion source.

-Carl
Carl Willis
http://carlwillis.wordpress.com/
TEL: +1-505-412-3277
Jon Rosenstiel
Posts: 1494
Joined: Thu Jun 28, 2001 1:30 am
Real name: Jon Rosenstiel
Location: Southern California

Re: Ion source working, 49 bubs / 3 min!

Post by Jon Rosenstiel »

Carl-
It was a paper titled "Magnetron Discharge Characteristics for the Improvement of IEC Performance" by Mizutani, Masuda, et al., which got me interested in trying this.

Thanks for the info and the tip on the paper; I'll have to track it down.

Jon Rosenstiel
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: Ion source working, 49 bubs / 3 min!

Post by Carl Willis »

*oops. The Takamatsu citation should be 2006 (the rest is correct).

I posted below a fair-use excerpt from the Takamatsu paper (for those without the luxury of a library account, can be found here: http://www.iop.org/EJ/article/0029-5515 ... _1_016.pdf)
that basically illustrates their magnetron concept and the resulting discharge. In their case they put the magnets in the center electrode, and they prefer to make the center electrode the cathode instead of the anode. Principle is the same as mine. I will have to see what happens when I change polarity.

-Carl
Attachments
takamatsu.jpg
Carl Willis
http://carlwillis.wordpress.com/
TEL: +1-505-412-3277
User avatar
Richard Hull
Moderator
Posts: 14975
Joined: Fri Jun 15, 2001 9:44 am
Real name: Richard Hull

Re: Ion source working, 49 bubs / 3 min!

Post by Richard Hull »

More great work Carl! It is good to see that the numbers are really picking up. I gotta' try such a source.

I see what looks like a venerable old Kiethly 610BR rack mount electrometer on your cart there. Am I right? otherwise it is a Keithley pico or nano ammeter. I have a few of those electrometers.

Your ionizer is a cross between and modification of an ion pump and or a standard discharge gauge tube. One could nearly instantly modify a stock varian cold cathode gauge tube to be the equal of what you made. It is a pigatron without an extractor setup. This ia pretty cool bulk, undirected ionizer.

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
Carl Willis
Posts: 2841
Joined: Thu Jul 26, 2001 7:33 pm
Real name: Carl Willis
Location: Albuquerque, New Mexico, USA
Contact:

Re: Ion source working, 49 bubs / 3 min!

Post by Carl Willis »

Hi Richard,


The Keithley 610BR ordinarily spends most of its life just sitting on the shelf, so I pressed it into service as a readout for my capacitance manometer. Not a job that really taxes its capabilities.

A standard cold-cathode Penning gauge does indeed work by the same E x B trapping mechanism, but they are not meant to handle anywhere near the discharge power that is required to form a plasma from which milliamperes of ion beam can be extracted. If one tried to use a gauge in such a manner, meltdown would occur very quickly.

-Carl
Carl Willis
http://carlwillis.wordpress.com/
TEL: +1-505-412-3277
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: Ion source working, 49 bubs / 3 min!

Post by Carl Willis »

I made some measurements on the ion source. It is current-thirsty but not very demanding on voltage. Also, it is utterly incapable of operation with the center electrode negatively biased.

For typical usage at about 7 mtorr, the supply voltage is about 1 kV and current draw ranges from 10 to 30 mA depending on fusor discharge current. With the 10 kV, 10 mA power pack installed, I can draw up to about 40 mA (I'm sure the power pack is none too happy with that). It seems that 100 mA could be drawn if desired. In light of this result I am probably going to go for the microwave-oven-transformer route and get rid of the power pack. I will also look into the prospect of water cooling, and I will probably aim to replace the stainless tubing anode with a solid copper anode to facilitate heat removal. I hope the vapor pressure of copper will not cause this to be a failure. One final thought is that in order to get a discharge to strike with the center electrode biased negative, I will probably have to increase its diameter so that the surface is in a higher B field.

-Carl
Carl Willis
http://carlwillis.wordpress.com/
TEL: +1-505-412-3277
Wilfried Heil
Posts: 590
Joined: Sat Dec 24, 2005 7:31 pm
Real name:

Re: Ion source working, 49 bubs / 3 min!

Post by Wilfried Heil »

Carl -
what kind of detector are you using on the Ludlum meter? Is this a neutron detector with a small moderator (42-9 BF3 counter) or an ion chamber? How high is it's sensitivity to DD neutrons?

Nice work with the ion trap/generator. I think Takamatsu put the center electrode of the PIG at negative bias to prevent it from discharging to the main grid.
WildBill
Posts: 19
Joined: Fri Nov 24, 2006 11:49 am
Real name:

Re: Ion source working, 49 bubs / 3 min!

Post by WildBill »

1) A more precise descriptor is "Sputtering Magnetron", unlike the RF device.

2) These discussions on RF dangers are exceeding important.

3) This hardware looks niiiiiiice...

Fine piece of work...

Harringbone!!!
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: Ion source working, 49 bubs / 3 min!

Post by Carl Willis »

Hi Wilfried,

The Ludlum detector is a 42-9, a cylinder of cadmium-covered polyethylene around a N. Wood G-10-2 BF3 tube. I just use it as a small general-purpose neutron detector. It sure beats the heft and bulk of a Bonner ball, although I don't know what its sensitivity to 2.5 MeV neutrons is supposed to be. In the same picture the lower portion of the moderator for a big He-3 tube can be seen.

On another metrology issue, I did put an ion chamber up over the viewport in a recent run and found about 30 R / hr there of very soft x-rays.

-Carl
Carl Willis
http://carlwillis.wordpress.com/
TEL: +1-505-412-3277
User avatar
Richard Hull
Moderator
Posts: 14975
Joined: Fri Jun 15, 2001 9:44 am
Real name: Richard Hull

Re: Ion source working, 49 bubs / 3 min!

Post by Richard Hull »

I got to thinking and the discharge gauge has an alnico circular magnetic with internal north-south poles at right angles to the tube. Your device, like a pigatron, has an axial field associated with it and not a cross field. The circular magnet on the gauge tube is oddly magnetized to accomplish the cross field magnetization. My bad.

I do have some potent alnico's with 1" holes in them that are magnetized along the axis. (I think from a couple of dismantled TWTs or noise sources. They would work OK and be able to handle really high surface temps that Nd:Fe:B magnets could not. Based on your 1kv 100ma discussion, a 100 watts disappation is not too bad.

The MOT based supply is a good choice for those current levels. I have a couple of salvaged ion pump supply transformers rated up to 800ma @ 4-5kv. Wearing a truss is needed to move them about, unfortunately.

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
Post Reply

Return to “Fusor Construction & Operation (& FAQs)”