To Focus or not to focus - the next generation

[pfostini]

I was going to start working on a focusing system for the gun. Larry's timely post ( beam interactions / theory forum ) has me thinking about this now.

The gun at present is like a flood light , just firing ions down stream. It does allow the fusor to go to lower pressures and can be controlled easily.

So now I have made a second unit the same as the first and I am going to extend the accelerator electrode out a bit from the conflat flange. With the fusor shell connected to ground I tried grounding the conflat on the test setup. What happens is the ion beam becomes very distorted. I then placed an extention on the accelerator electrode which extends it out beyond the conflat and now there is no distortion.

If we make a focusing element for these guns, consider the design we have and how to incorporate focusing into it. I was looking at electrostatic or magnetic type but one that can be adjusted for different fusor sizes. Or is the floodlight effect going to be more fruitful ( neutful ) in its use?

[guest]

This new topic brings to mind something I've been thinking of trying as a gun source in a fusor. Back in the 70's while playing on a small accelerator at the University of Nebraska I made a gun source to calibrate the x-ray detector I was using by taking the neck with deflection plates from a surplus scope tube. I machined a flange to fit a port on the target chamber, epoxied the tube neck to it, changed the filament to I don't remember now what, made a little circuit to apply variable dc voltage to both sets of plates, paralleled another scope tube so I could see where the beam should be and went after it. Crude, but it worked.
I've been pondering whether grafting a couple of these to one of my fusor chambers wouldn't work as gun sources.
At least the plates from a scope tube might help your focus problem.
Just pondering,
mike

[hellblazer]

What's the power difference between using the ion flood and using the straight fusor for comparable neutron counts? Sorry if I missed this...

Certainly the control seems like a big advantage. If the power consumed is less with operating time increased as well, then it seems like a big win all around for such a simple change.

[hellblazer]

I've been playing around with a simple penning type focus element, modeling it in Simion. The results look very encouraging. The only drawback is that one of the electrodes will have to be machined. I can do the machining if you like and ship the part off to you. It's pretty simple, just a cone section out of a disc.

Anyways, I'm still on the learning curve of Simion, so I'll post results when I'm more confident. But it does look very encouraging. The result would be very simple two electrode system instead of just one electrode. In the design, there's no need for the BeO tube, rather the whole thing should be very thin.

Anyways, thought I'd keep y'all updated.

[DaveC]

Glad to see someone is using Simion. There used to be a FREE 6.0 version..but last time I looked it was gone. anyway if you are a student or Teacher, you can get it for less than $50 .

Hal - consider the einzel lens.. In it's simplest form, it is only two cylinders with a different voltage on each.
Used near the cathode or filament, with the accelerating electrode farther out, you should be able to get a very good focus with a couple hundred volts on the cylinders, and about 10 kV on the anode or target.

Havent yet done a hollow anode. this will have some focussing power of its own. but you should be looking to get inside of a 1 micron diameter beam at focus.

The current density in two opposing beams at the focus region will be about 8 orders of magnitude larger than in the standard fusor.

A more interesting approach yet would be to use a cylindrical container with opposing ion guns emitting collimated beams on the order of a few tens of microns. The interaction zone could be quite a few cm long, giving a net interaction region large than in the spherical configuration.
THis last configuration uses a lens system not unlike that of an electron microscope.

Use of some deflection coils or plates to align the beams to the output apertures becomes almost a necessity. Check out the Scanning Electron Microscope sites. Several university sites have some good background material on the electron guns in SEMs.

I have a couple einzel trajectory plots done for electrons, it will be easy to do them for deuterons.

Dave Cooper

[DaveC]

Here's a doc file with a Simion Deuteron plot. I will try to set up a colliding beam one, but the results will be only somewhat suggestive as charge repulsion and space charge effects are not handled too well.

Dave Cooper

[DaveC]

Just checked the download... You probably wont see anything till you print it... at least my system doesn't see the plot, till its printed.

I will try to get it saved in a jpeg format, for on screen viewing.

Dave Cooper

[hellblazer]

Yep, this is what I started to conclude last night. The problem is, the einzel lens is a bit more complicated from a power supply point of view. Seems worth it, though. A penning diode really only gives you a collimated beam - nice input to the lens, though. Mostly I was just looking for something simple to try to correct the flood light dispersion.

1 micron focussing would be spectacularly cool to achieve.

I just received my charged particle optics book I snarfed from Ebay. Lot's of great info in there. Ah, this is such fun.

[hellblazer]

I pulled it out for ya.

[hellblazer]

Is it possible to upload these Simion files for this? What I'm curious about is how you modeled the planar emitter and the initial ion state. Still learning and this example would be good for me to study.

Simion is much better than I expected, althought the UI is awkward - not because it's poorly designed. Just different than most UI's I've used. I realize this is because it was a DOS program way before windows even existed, so I'm not complaining... It just takes quite a while to get used to.

Anyways, many thanks.

[pfostini]

So if I add another electrode to the cathode end of the laser tube version I should be able to focus the beam? At the moment I added anthr electrode to the ande thinking I can focus after the accelerator but that's not working too well. But then again I have only the quartz target in place.

[Richard Hester]

The quartz target might be collecting surface charge and screwing up the focus. A screen or metallized coating to dissipate the charge may be in order.

[DaveC]

Thanks Hal for posting the Plot. Richard Hester is correct. Without some conductivity in the target, charge will just build up till a leakage current develops, somewhere. Either a thin Aluminum evaporated coating on the quartz, or just a metal target in front of the quartz will work. But dont forget the electrical connection.

The key to a good focus is to do it at the low potential end of the gun, near the cathode. Need the least voltage there.

I have been thinking about the cylindrical design a bit more. Perhaps a ring shaped "anode" (for deuterons , of course it would be a negative polarity "anode" ) would work, with two guns aimed in opposite directions down the axis. The collision area would be in the center of the ring. I have not modelled what effect the ring will have. Will try to do that later this evening and see what happens.

A variant to this would be to have a Penning type reflector ( the same potential as the cathode, located on the beam axis, but spaced an equal distance the other side of the anode ring. This would reflect all but the wildly off axis ions back through the ring and toward the original cathode. The ion beam equivalent of the reflecting ends of a laser tube. It is quite possible that only one einzel lens will be necessary, but a second one symmetrically placed to the right of the ring "anode" probably would improve overall focussing.

If I can do it, will add these last mods to the model, tonight.

One or more pairs of opposing ion guns would also work well in the Spherical fusor geometry, using the regular cage electrode.

There are some other forms for the ion guns that might be even more effective. Should have a bit more to share in a day or so.

Hope this is useful.

Dave Cooper

[hellblazer]

Well, you'll create a lens due to the potential gradients of the gap, so in theory, yes. But as I'm sure you know, just throwing a random lens in front of a beam of light doesn't necessarily do what you want. If it has a very short focal length, it can actually have the opposite effect of what you want it to do.

Part of the problem is that you have to shield the lens element from the fields caused by the tubes. In "Building Scientific Apparatus", they show a standard method for shielding the lens effectively. Basically, the idea is to offset the electrodes so the gap is shielded. Takes a bit of machining, and the gap has to be controlled effectively - too big of a gap and the lens is dominated by undesirable effects. I can't remember what the ratio to diameter to gap is, but I'll investigate.

I do think you need a negative potential target, though. Otherwise, the ions are going to be decellerated back towards to the electrode, and very strange things can happen.

Again, just my beliefs from what I've read - not from any practical experience.

[hellblazer]

BTW, correction. It's the Pierce diode structure, not the Penning diode. My very bad misstatement of the obvious. Sigh. Sorry, y'all.

[DaveC]

Hal - you are "forgiven" (spell check won't catch that one..)

Attached here is another plot of what I was discussion earlier..a reflex ion gun.. for lack of a better name. Perhaps I could prevail on Hal one more time to repost the plot. Not sure yet how to do that without the MS word attachment.


Note how these lens voltages give a reasonably good retention of the beam for at least two passes. The beam waist is apporximately at the center, inside the -10kV "anode" ring.

Shouldn't be too hard to build this.

Dave Cooper
Thanks

[DaveC]

Here is the Reflex Ion Gun w/ Trajectories, in jpeg format.

Dave Cooper

[hellblazer]

Now that is a very cool idea.

[Richard Hull]

Hal is correct and Dave is also modeling it right. You just gotta' have that centralized full, acceleratory potential cathode, semitransparent target to make the deuteron beams slam together for max effect. The normal hollow wire cage in a fusor is fine, but I am opting for the tungsten meshed target ring in the linear version if it gets built. If high power is achieved it will most certainly vaporize. It would be interesting to see about the white hot W creating a virtual cathode of negative electronic emission space charge.

Richard Hull

[grrr6]

<rescinded after doing some reading of principles of charged particle acceleration>

New question:
How did you model that Dave?

[DaveC]

Greg -

Here's a little bit about how the electron optics works. You should probably think about getting hold of some text on electrostatics for a more thorough treatment.

Simply put, whenever charges are placed on an electrode, a voltage or potential is present. The potential is actually a measure of the energy a charge particle would either gain or lose as it leaves the electrode. At points away from the electrode, the particle would have different energy and thus a different potential. Potentials around the electrode can be plotted as lines of equal potential known as equi-potentials.

The rate at which the voltage changes ( gradient of potential) is expressed in volts per unit of length, and is what we call the "Electric Field". The term "electric field" is simply a way to describe the force experienced by a charged particle at that particular point in space.

So, when a voltage is applied to an electrode, it creates around the electrode an electric field which is capable of exerting a force on a charged particle.

Now depending on the shape of the electrodes we can get differently shaped electric fields, which can bend the paths of moving charged particles - electrons or ions.

The electrodes with voltages on them, act to the ions like a lens acts to light waves or photons. The two cylinder electrodes comprise what is called an "Einzel" lens, which works very much like a simple optical lens. If an ion is going through the cylinders, say from left to right, and if the first cylinder has a lower voltage than the second, the electric field in the gap between the cylinders, acts to bend the flight path of the ion towards the axis of the cylinders. If the first cylinder has a higher voltage than the second one, the fields bend the ion outward. This gives either a converging lens in the first case, or a diverging lens in the second case.

In the case of the two cylinders, lines of equal potential curve into the bore of the cylinder. making what actually closely resembles the outline of a convex or concave lens. Since electric field lines intersect the equipotential lines at right angles, and since at low speeds, ions and electrons more or less follow the electric field lines, we can actually draw the paths of the ions as they go through the lens.

The plots I posted above could be displayed with the equipotentials and the ion trajectories. I'll rerun one and post it for you.

That's the broad generalities of how bits of metal with voltages on them can bend the paths of moving ions. Hope the explanation was somewhat useful.

Dave Cooper

[DaveC]

Here's a much simpler version of the reflex ion gun. It can hardly be much simpler than this.

The ion source is on the left emitting positive ions of 2 AMU. It has a potential of +1V.

At center is a ring electrode at - 10kV.

On the right side is the reflector at same potential as the source, +1 V.

I put a grounded (0V ) metal can around everything along the beam axis.. and... dodged the issues of how to get the -10 kV to the ring in the center, for now.

Notice the beam path resembles a bow tie, somewhat. The return paths of some ions can clearly be seen. What is interesting is that this dense pattern is from only 30 ions!. The plotting took a number of seconds to complete, as each ion flies separately, until it collides with an electrode. Most of the ions made many passes before finally colliding with the reflector or source.

The second picture is an enlarged view of a portion of the source end, showing how the returning ions are reflected again and again. The squares are 1mm in both pictures.

Looks like the reflex ion gun is a pretty simply thing. We should note too, that this model does not account for residual gas molecules. So the pressure assumed is such that the mean free path is a few tens of meters, perhaps around 10^-6 Torr.

I have done some models of scattering in air at these pressures. There is not much. So the trajectories should be more or less representative of the real thing.

Keep in mind too, that the earth's magnetic field will have an effect if the tube is not enclosed in a magnetic shield.


Hope this helps stir the imagination and guide the construction a little.

Dave Cooper

[DaveC]

Here's the enlargement mentioned in the post above. Tried to send both together, but seemed to print one over the other.

Hope this is not too confusing..

Dave Cooper

[hellblazer]

One thing I'm wondering about, though, is the scattering at the focus. The cross section of scattering is much larger than the cross section of fusing. So, what I'm thinking is that this will be similar to a magnetic mirror in that there will be a lot of losses in the equitorial plane at the focus. Also, those ions that aren't lost through scattering, most will still have their trajectories significantly altered through the scattering. This will change their energies, and thus their ultimate focus. It seems like this would have the effect of diffusing the fine point focus at the center.

Anyways, just some completely random thoughts. I still think this is a pretty cool idea, and I think it will be extremely interesting to see how it turns out.

[pfostini]

This thread has been very informative. I am going to try some of the ideas on an actual gun. I will also rig some type of target to prevent the scattering and such in order to truly see where the beam hits. On the last gun I will use a much smaller extractor and seal it in such a way as to have a higher pressure behind it and use the pinhole as an orfice for slowing the gas flow and have a higher pressure gradient.