Fusor Forums

In a preliminary test, my electron cyclotron injector sucessfully ionized under RF ionization for a sustained period. Since the 900MHz amplifier to generate 900MHz for the ECRF resonance is not complete yet, the system was tested at 150MHz, ionizing as an inductivly coupled plasma instead of by ECRF excitation.

File manifest:

PIC1: 20W in at 150MHz, 5W reflected.
PIC2: DC extractor field femlab sim
PIC3,4: magnetic field femlab sim
ECRFmacro.m : nec file coil generator macro
ECRFcoil.m : nec file laser feedthrough coil generator for matlab
EyscanGIF2.gif: y component of rf field at 900MHZ as simulated in nec
Current-pan.gif: antenna currents and structure
ecrf-coil.JPG: molybdenum antenna coil

get 4nec2 and 4nec2x to view simulations from:
http://users.ece.gatech.edu/wrscott/

for other details on the injector
http://www.rtftechnologies.org/


Andrew Seltzman
www.rtftechnologies.org

Andrew
lovely work both physical and at the design level

What power at 900 MHz will it need
and what sort of ion current do you expect after extraction
thirdly is this going to be a negative source??

I am aiming at driving each injector at between 20W to 50W RF input at 900MHz, however based on my experimentation using it as an inductivly coupled plasma it may need less then 20W using ECRF ionization since it is more efficient.

I am aiming at about 10mA of current after extraction based on previously published designs of simular size.

The ion source will be a postive source. The extractor cone will be grounded and the antenna coil will be biased at +15kV and surounded by a ceramic ionization cup.

Andrew Seltzman
www.rtftechnologies.org

Andrew - 20W in 5W reflected, swr indicates bad impedance mismatch - a pi tank will help to insure a better power transfer and better matching. It will also help the driver stay cool.

Boy, I'm always amazed at the quality and level of engineering that you guys do here. I'm impressed.

I have a pi match already, however both the PI network and the antenna were designed to operate at 900MHz, I was running it at 150MHz just for preliminary testing, hence the very bad mismatch.

Andrew Seltzman
www.rtftechnologies.org

Ahh! that explains it - you are ahead of me Andrew

I congratulate on very good work and watch with interest.

Hi Andrew,

Good work and nice "eye candy" to share with the forum.

I'm watching your ECR source project with a lot of interest. Right now I am planning an upgrade to my fusor that will give it six ion sources--the ability to get high ion currents at high voltages seems limited by the availability of ions, and my little magnetron source is not good enough. I am planning on inductively-coupled discharge sources with B > B[ecr] and f probably around 50 MHz.

I am not expecting much additional plasma density or electron energy at the ECR condition, however. It seems that for low frequencies and small source geometries, one really can't get high energy electrons in circulation per the theory. The electrons orbit at low energies, but as soon as they get hot they have spun into the source wall on account of the small dimensions. The only way to improve the electron temperature for the given geometry is to raise the frequency and the magnetic field.

I have run an ECR proton injector as part of the work related to my dissertation. It uses 2.54 GHz from an oven maggie delivered to an alumina-lined discharge chamber that is surrounded by water-cooled solenoids in a magnetic-bottle configuration eating 75 A at 24 V. This source can kick ~25 mA of protons on a good day. I've wondered why it couldn't be designed to use lower frequencies and smaller magnets, and I think the answer is that the plasma density will drop per my observation above.

So, I just wonder how well one can do with ECR vs. ICP if the B, f, and dimensions are all small. Your project should shed some light on that.

-Carl

10 ma of ion current is a lot of current! Do you really think the device you have there Andrew will do that? I hope it gets you where you need to be once you get the match at 900 mhz.

Richard Hull

Hirsch-Meeks ion source was running at equivilent extractor potentials and produced 10mA od beam current at 12kV extractor potential.

Attached is the Hirsch-Meeks paper.

Andrew Seltzman
www.rtftechnologies.org

Hi Andrew!

I know how to calculate the Ion Cyclotron-Frequency.
But how did you choose the diameter and the number of windings of the RF-coil?
Or did you just take the available space according to the magnetic field and distance to magnet to achieve the right B-field?

I talked to several people. Some told me to calculate a kind of resonance frequency (coil + wires as capacity) others told me to just take a "few" windings.
I have about 1 kW @ 13,56 MHz available using a MFJ-962-D tuner as matching network ...

BR,
Werner

The diameter was chosen based on space available(the spacing between the feedthroughs). The number of turns was simulated in NEC, and chosen to produce the highest transverse(perpendicular to B) electric field.