Small accelerator

[Rapp Instruments]

Hi,
I just finished a small accelerator. The tube was made of glas rings which are epoxied to aluminium rings holding the copper accelerator tubes smoothed with brass rings. The traget in the last copprtube is made of a titanium disk.



The tube is powered by a homemade supply consisting of flyback transformers and multiplier chains. Accelerator voltage and channelray source current are measured with small meters, target current is shown by the flash rate of a neon bulb. The target is bias positive by some hundred volts to surrounding tube to inhibit secundery electron emission.



with a deuterium pressure of ~ 10-2 mBar in the source and < 10-3 in the column I reach a accelerating voltage of 150 kV. No arcing, but some corona a the high voltage end. Althoug I see a well definded beam (low intensity, photographing with 10s, f 8 and ASA 1000) throughout the tube there is only a very low ( < 5µA) target current. X-rays but no measurable neutrons flux.



I Think I should try to improve the focusing by changing the voltage at the first accelerator electrode. Does anyone here has access to the program
>> Simion << to make some simulations of my ray optics.

Thomas

[JakeJHecla]

Great work! I have access to Simion 7, but I would recommend finding someone with Sim 8. Sim 7 is exceptionally buggy to say the least.

[John Futter]

Thomas
You need to bias the
secondary electron electrode negative to get secondary electron suppression around 400 v is enough if the suppression ring is around 1cm away from the front of the target.

you can do this automatically with a transil diode approx PKE400A or similar
I have posted on how to do this on the before (a few years back)

edit Thomas you have already done this I should have looked more carefully at your schematic

[Bruce Meagher]

You have a very impressive looking accelerator. Congrats!

Can you explain a little bit about your target design? Is it a thick Ti disk or Ti coated on a Cu substrate? Are you loading the target with D2 outside the accelerator, or using the beam to load the D2 into the target? If you're using the beam to load the target how long does it take to load the target to some reasonable level at your measured beam current / beam size?

Bruce

[Rapp Instruments]

Hi Jake,
if I don't hav success in finding >IONSIM 8 < would it be possible for You do do this simulation, if it's not too time consuming?

Hi John,
yes, I think a diode would be better. The bias would not depend on the accelerator voltage like with my voltage divider.

Hi Bruce

the target is a 2 mm thick plate of titanium. I also tried a titanium plated copper disk, but I can only prepare thin titanium layers by vacuum evaporation. This very thin ( ~ 50 nm) layers are quickly destroyed by the ion beam.
In my small neutron source some months ago I used a thick zirconium plate as target. It work very well and immediately as deuterium was feed to the system. I have no zirconium sheet left but I think titanium should work as well.



Thomas

[prestonbarrows]

That is some nice work. You might want to add a smooth cap on the HV end to enclose the pointy terminals and connections you have exposed, this will cut down on your corona losses.

Besides your power supply and unoptimized lenses, the limiting factor in the design may be excessive pressure in the accelerator column. Because of the long/thin geometry of the accelerator tube and having your pumps located pretty far away through bellows etc., you will tend to have a significant pressure in the column given that you have a few Torr in your source.

Do you know what the pressure is at the target end while HV is off?

Ideally, you want to be down more towards the 10^-5 Torr type range in the accelerator. Lower is even better. High pressures will cause your beam to undergo many collisions and down scatter and/or charge exchange with the background neutrals. Both of these will reduce your target current and beam energy while generally making the column more prone to arcing. You can see the characteristic 'halo' surrounding the beam due to interactions with background gasses.

Practically the only way around this is to make wider columns with high conductance pumping paths around the lenses or floating UHV pumps up on the column. Both approaches quickly exceed the scope of a benchtop project like this...

[Rapp Instruments]

Hi,
for reducing corona losses I have bought one of the stainless steel >IKEA salad bowls< which I used already with van de graaf generators and tesla coils. The bowl mounted on the top and some corona rings made from copper tube down to the earth side.
I also think that the pressure in the column is to high. Otherwise one would not be able to see the beam. On the other side the pressure is low enough to prevent sparking and glowing in the column if the ion source is switch off.
The pressure in the ion source is around 5x10-2 mbar, the diffusion pump is connected to the column just at the exit of the canal of the ion source. In the moment I can't measure the pressure in the column because I have no working penning meter, only a pirani meter which lower end is 10-3 mbar.
To reduce the pressure in the column I can increase the length or reduce the diameter of the canal of the ion source, but of course with resulting lower ion output. Or change to a penning or rf-operated ion source which allow substantially lower working pressure.

Thomas

[prestonbarrows]

Im not sure on the standoff limits of your system, but one suggestion is to increase the voltage drop across the first few lenses toward the source. Since you are just using resistor dividers, you should be free to choose any voltage profile within the limitations of arcing between lenses. First, you will ideally be able to extract more current from the plasma (assuming you aren't already limited by plasma density) with a higher electric field from Child's law.

Basically, it is generally good to get the ions moving as fast as possible as soon as possible so that space charge is reduced. Otherwise, you have slow ions hanging around the emission aperture repelling any more from coming through, and 'plugging' the source.

Also, getting the beam up to higher energy sooner will also reduce the angle of the cone which the beam makes as it travels. To first order, it is just two perpendicular vectors with the accelerator-driven velocity down the axis and the space-charge-driven velocity radially outwards. So boosting up the component along the axis sooner tends to shrink the beam diameter and helps get more current to the target.

Of course, messing with the voltages will change the focusing action of the lenses. But since I think all your lenses are just straight 'pie plate and tubes' without any specific design, you will have more leeway with sweeping voltages around. The first electrode by the source will be the most important. You should be able to find a specific voltage which is best matched to your lens geometry.