Electro-Static Linear Deuteron Accelerator Project Photos

Current images of fusor efforts, components, etc. Try to continuously update from your name, a current photo using edit function. Title post with your name once only. Change image and text as needed. See first posting for details.
Post Reply
User avatar
Dennis P Brown
Posts: 3159
Joined: Sun May 20, 2012 10:46 am
Real name: Dennis Brown

Electro-Static Linear Deuteron Accelerator Project Photos

Post by Dennis P Brown »

Since someone asked, here are a few current photo's - hopefully, I will be able to test this very soon - still an issue with the ultimate vacuum but appears to be out gassing which should be addressable. The only thing missing to "fire" the system is a 12 volt battery to power the HV deuteron gun (located within the closest SS globe.)

Here is the complete accelerator with shielding blocking the view of the accelerator tube.
Attachments
MVC-017L[1].JPG
MVC-017L[1].JPG (45.28 KiB) Viewed 7906 times
User avatar
Dennis P Brown
Posts: 3159
Joined: Sun May 20, 2012 10:46 am
Real name: Dennis Brown

Re: Electro-Static Linear Deuteron Accelerator Project Photo

Post by Dennis P Brown »

Here is a picture of the target chamber. On the left, is the bellows that aligns the electron suppression ring over the target. Center, foreground is the water cooling system for the target. Top is the turning mirror to view the sight glass to focus the beam (when the glass is in place) and right, the target sight glass bellows that places the glass plate into position in the beam.
Attachments
MVC-015L[1].JPG
MVC-015L[1].JPG (40.33 KiB) Viewed 7906 times
User avatar
Dennis P Brown
Posts: 3159
Joined: Sun May 20, 2012 10:46 am
Real name: Dennis Brown

Re: Electro-Static Linear Deuteron Accelerator Project Photo

Post by Dennis P Brown »

Here is a better picture of the complex support systems for the accelerator: the fore pump (center, bottom), turbo pump and controller (top center and bottom right), ion gauge (middle, extreme right), Baratron (middle, center for the gauge; the controller is with the ion display)), deuterium tank/regulator/feed system and leak valve (middle section on the left), high voltage supply to spray the VDG (diode rectified white NST and varic; lower left side), the battery (red, extreme right) for the suppression system and two microamp gauges (0 - 3 microamp and 0 - 50 microamp ranges) for the target (to measure the beam current ... hope there is one soon!)
Attachments
MVC-018L[1].JPG
User avatar
Dennis P Brown
Posts: 3159
Joined: Sun May 20, 2012 10:46 am
Real name: Dennis Brown

Re: Electro-Static Linear Deuteron Accelerator Project Photo

Post by Dennis P Brown »

And, of course, the accelerator tube with field collectors for each Einzel lens system. The VDG can be seen to the extreme right. The closest Globe holds the two HV system/battery supply for the Deuteron gun (a modified PIG ion gun) Each collector plate has its own needle to control the electric field and prevent opposite charge build-up and equalize the charge (very little down stream of the VdG generator.)
Attachments
MVC-019L[1].JPG
MVC-019L[1].JPG (49.06 KiB) Viewed 7906 times
User avatar
Dennis P Brown
Posts: 3159
Joined: Sun May 20, 2012 10:46 am
Real name: Dennis Brown

Re: Electro-Static Linear Deuteron Accelerator Project Photo

Post by Dennis P Brown »

Hope these pictures help for understanding what is involved for building a safe accelerator; a fusor is a lot easier and I'd think more fun to build. The number of support systems is getting ridiculous in number, complexity and level of detail but seriously, if one hopes to get a system that does science (my real objective) and offers versatility in target selection/operation, one does not get away with an easy project. Again, NOT a project to be taken lightly nor one that is done quickly without issues ...still, been interesting so far.
User avatar
Dennis P Brown
Posts: 3159
Joined: Sun May 20, 2012 10:46 am
Real name: Dennis Brown

Re: Electro-Static Linear Deuteron Accelerator Project Photo

Post by Dennis P Brown »

Speaking of subsystems, the globes house various system that are necessary for successful accelerator operation but are both a pain and rather complex. The primary VdG need a high voltage spray system to control the polarity (critical if you want the deuterons to accelerate) and designing the power supply's and spray bars as well as the collectors is not a simple task - not even consistent with what is written on the topic! Even mounting the globes is a pain - the original mfg assumes they fit together by their poorly design seam; that is a joke. Result, have to bend/shape an internal support system that aligns with bolts holes so screws can be used to hold the globes together. The original spray bars and pulleys do not work for a HV spray system so new ones must be made and installed along with many attempts to optimise the best design - not a simple task unless one already is experience with this type of special work.

Then the issue of shorting out the high voltage from the VdG occurs through the deuterium gas feed line - a task to fix this problem, as well. The list does on. Every single support system has its own issues and problems that all must be addressed and integrated into the whole - not all fixes turn out to be viable when put into operation due to other constraints by the accelerator. So, again, a rebuild and some are major due to the fact there is no simple work around. The point is this device is far more than what it appears even in the pictures.

I provide a photo of the second globe and what is inside. On the left is the negative 25 kilo-Volt initial accelerator/focus system for the PIG. On the right is the rheostat to adjust both power supplies supply voltage. Center is the support column and magnet (ring) for the PIG. Just to the right, lower side of the metal column is the tiny high voltage supply to ionize the deuterium gas to create the deutrons. The gas supply line and feed connector is also seen (also center of image.) Keeping wires clear and high voltage systems from shorting is a bit of work. Physical support systems are needed in the globe and all these systems and the globe must "hang" onto the end of the accelerator tube - a base support would short out the globe!
Attachments
MVC-020L[1].JPG
MVC-020L[1].JPG (39.89 KiB) Viewed 7906 times
User avatar
Dennis P Brown
Posts: 3159
Joined: Sun May 20, 2012 10:46 am
Real name: Dennis Brown

Re: Electro-Static Linear Deuteron Accelerator Project Photo

Post by Dennis P Brown »

First off, these are excellent questions that are rarely asked/answered in any papers posted on building these types of machines.

So, I will answer each question in order (Aside: I owe a great deal to so many people here who have, over the last few years both flagged design errors/problems/safety issues and offered outstanding solutions to these issues - some I am posting here. Just shows how invaluable this forum is for experimenters!)

" magnetic suppression from permanent magnets clamped around the outside of the tube. "

This accelerator design - for better or worse uses an electro-static power supply (a Van da Graff or VdG) - that means that the electric field produced by the VdG provides all acceleration effects. Any metal not designed into the system will change/deflect the electric field seriously and in an uncontrolled manner - most likely causing the deuteron beam to miss the tube axis completely; hence, placing any large metal mass around the tube would be a disaster. Remember, a simple magnetic field just causes charges to circle around the lines of force and will not keep them focused along the axis alone. Magnets that do that are (I believe quadrupoles) very special and not something I'd ever attempt to build. In any case, then I'd need an alternate way to accelerate the ionized deuterons since I'd be unable to use the VdG since the electric field would be compromised.

"Do you have an idea of what the background pressure is in the beam transport region?"

Yes I do - the ionization gauge is directly attached to the accelerator tube giving a very good reading on that overall pressure - recently, the internal pressure with a controlled leak rate of deuterium was 3.5 x 10 ^-6 torr. If I can re-achieve that pressure I see no issues.

"You should get better stability if each acceleration lens is physically tied to a voltage through a resistor ladder from the HV to ground."

Yes; also, you are not the first person to point this out and yes, I do. Due to the nature of using a VdG as the 'power supply', physical resistors that I can obtain will never work. Instead, I simply use acupuncture needles to both shape the electric field around each collector but more importantly, act as bleed resistors/stabilizers for the voltage potential on each Einzel lens assembly/collector.

"Since it looks like your grading rings are separated from the beam by the glass chamber wall, you may get by with an extremely low drain current."

In reality the collectors are not separated at all from the electro-static lens. These collectors feed directly through the glass tube and power the Einzel lens. I drilled ultra tiny holes and feed a tiny wire through that connects to each lens assembly. These were Ohm'd out to better than 0.2 ohms per lens.

"What is the voltage drop between each plate?"

That is a good question but not important for me. After the first two collectors the acceleration effects are near nil and irrelevant, any way. Most all acceleration occurs by the first real Einzel lens, anyway. The weak field near each succeeding lens can be very small since Einzel lens do work well in weak fields - also, the first set of lens really provide the vast majority of focusing and deuterons too far off axis will just be lost. This is a low current machine anyway. I am only looking for neutron production of 10^5 to 10^8 since the beam will be collimated. So a low count isn't significant for the research I plan to perform.

"It is hard to tell from the picture, but are there metal parts inside the vacuum in the glass tube accelerator section?"

Yes - the metal parts are the Einzel lens. (see attached photo.) They focus and guide the deuteron beam. Only the first two really provide any significant focusing but the others do provide (rapidly decreasing, of course) some focusing effects.

"You should get better performance if you space the accelerating plates as close together as possible without running into arcing issues. "

While not a question it is a statement that I will address.

Most current designs use exactly what you propose. I didn't use that method for a few reasons - first off, when I stated this project I'd never seen any other type of design except this classic, large distant separated, Einzel lens design. Second, and why I didn't change to the current design - I simply did not want to manufacturer that many metal plates and glass plates with holes drilled through them. I just didn't feel that that level of effort was worth it since drilling the glass tube in the manner I have was, and is, extremely easy. That said, I'd really rethink the design of the accelerator tube if I did this over again. Still, electric field breakdown issues between very closely spaced glass/metal plates do require careful design and possibly the very special resistor assemblies often found on that type of design - resistors tolerant of extreme electric fields (read a device for use in a VdG field); however, these types of resistors I simply cannot obtain - the work around for a stacked plate would then require even more effort on my part (more work arounds) and when combined with all the machine work this multi-stack system then entails might then turn out to be more challenging than the simple, and classic arrangement I did use. Also, the cost of all the metal and glass plates isn't something that is minor compared to the costs for this older design. In hindsight, the stack might have been better but that is not something I am going to change this late in development.

Balancing all these considerations to make a better or more optimized accelerator is a bit above my pay grade. I do intend to do some ground breaking research relative to a neutron detector design and a radical (but very modest) new approach to some fusion by an interesting modification to zeta pinch fusion but these two experiments hardly are earth shaking nor ones that really require more than a low total neutron count beam that this machine (hopefully!) can produce.

I am including a close up of an Einzel lens assembly. The large circular collectors have a wire that connects them through the glass tube. This hole/wire is sealed with vacuum grade epoxy. Notice the tiny needles at the top of each collector. These can be moved and even replaced with longer/shorter ones (as needed) to adjust the electric potential for each assembly. These needles will help prevent over charges if some deuteron collisions occur and/or by negative electric charge build up from electron back flow into the tube hitting that given Einzel lens. These needles should act as bleed resistors to equalize charge between the plates/collectors. (Two notes on this picture: First, the plastic tube support (no metal parts at all) is near the third Einzel lens assembly where I feel it will still provide good support (prevent the very serious risk of accelerator tube breakage) but any electric field effects are not as serious here than if mounted closer to the globe. Everything about these devices is a compromise and this design aspect will be revisited if I problems with the beam occur on first light tries. Second, and more seriously, the Einzel lens are made of the wrong material - brass. They should have been a lighter atomic mass material like aluminum so that electron impacts would produce lower x-ray intensities. Too late to go back and fix this error but these lens are heavily shielded by the very thick slate walls both to the side and under the assemblies. The slate wall thickness was calculated to fully defeat 500 KeV X-rays - far beyond anything my weak VdG could ever achieve.)

I hope these answers clarify your rather good questions.
Attachments
MVC-021L[1].JPG
MVC-021L[1].JPG (26.85 KiB) Viewed 7906 times
prestonbarrows
Posts: 211
Joined: Sun Jun 24, 2012 1:27 am
Real name:

Re: Electro-Static Linear Deuteron Accelerator Project Photo

Post by prestonbarrows »

Thanks for answering my questions Dennis.

So are you saying each of your lenses are completely isolated from each other and floating? I have never seen anything like this and feel it would just lead to horrible arcing and stability issues. The bleed current from these is going to be very non-linear when stray beam hits it leading to strange feedback and will change with things like ambient humidity etc. The leaked corona current from each will also be dependent on the lens' absolute voltage meaning each will be different...

You have the right idea, but what you want is to replace those needles with some very high resistance power resistors between each plate/lens. This will draw a continuous current and peg each lens at evenly spaced voltages all the way down the stack. In principle, any resistance will work as long as they are equal (to within tolerances). In practice, you will have to choose between higher resistance giving lower wasted base current from your power supply but lower stability, and lower resistance giving more wasted base current and less stability. The exact value will depend on how much overhead your power supply has (its maximum rated current less what you typically pull through the beam, ion current and backstreaming current combined) and how much stray current typically hits your lenses.

Maybe this is common in very high voltage van de graff systems? I don't have direct experience with those. I work more on systems in the 100's of kV 100's of mA type range.

How much current does your corona needle setup pull without the ion source running, i.e. no beam? A good first guess would be to size the resistors such that you will pull this same 'dark current' as you are with the needles. On all systems I've worked on it will be up in the mega- to giga-ohm type range.

Are your 'lenses' actually shaped in any special way? or are they simple sections of pipe basically?

As for the voltage on each lens, to first order they should be total_voltage/n where n is the number of jumps to ground. That is, if your corona needles are doing their job haha. I don't think I see that info anywhere. What voltage and current does your van de graff provide? Do you have a way of measuring? I am guessing the van de graff does not have a calibrated voltage/current knob you can set?

Knowing how much current is leaving your supply versus how much arrives at your target gives you a diagnostic as to how much you are loosing by striking the lenses for example. If you can accurately measure the temperature of a water cooled solid target, you can do calorimetry and work out how much ion current is hitting the target if you know the power supply voltage; this also gives you an idea of what fraction of the current being sourced by the power supply is useful ion beam versus wasted backstreaming electrons.

10^-6 torr in the column is very respectable, you should not be getting horrible levels of background ionization from that. It sounds like you already have an electrostatic suppression ring upstream of your target? That should be plenty with your low column pressures.

Even though they probably are unnecessary for this system, you are correct about the metal magnets shaping E-fields. However, in this case the cylindrical conductors of your accelerating lenses basically shield the E-field that the beam sees from any changes in shape outside of them. Small magnets which do not approach the lip of each metal tube would have no effect on the beam's E-field. They would narrow the standoff gap between each lens on the air side; but no more than those corona needles do already. If each B-field stage was aligned in opposite direction of its neighbors, the net effect down the beam line would wash out but would tend to make the beam wider (something similar to a wiggler). Again, putting them on each stage of the accelerator would be excessive, but one between your target and accelerator could possibly be helpful. At those low pressures, most of your backstreaming electrons are probably going to come from secondaries off surfaces like the target and lenses. Hopefully you do not have much beam scraping on your lenses along the way.

Again you are correct in principle in regards to the quadrupoles etc. but two dumb permanent magnets spanning the beamline will work fine here. I have used such setups on much more complex systems.

The idea is to design the magnetic field such that the gyroradius of the electrons is small compared to the size of the vacuum vessel so they get bent into the wall while the gyroradius of the ion beam is sufficiently large that the beam is bent only a small angle and still transported. This is another reason that it is good to put a magnetic suppression close towards a solid target so that the slight bend does not really matter since you don't have to transport it much further. Also, near the target, the electrons have not been accelerated significantly yet while the ions have; this means the slow electrons are easier to deflect than the fast ion beam.

Electrons are much much lighter than ions and much much easier to push around. Like blowing a fan on a mix of bowling balls and ping pong balls rolling by.

For the relevant equations, look in the plasma formulary
http://wwwppd.nrl.navy.mil/nrlformulary ... ARY_13.pdf
on page 28. Feel free to ask if your are unfamiliar with that.
User avatar
Dennis P Brown
Posts: 3159
Joined: Sun May 20, 2012 10:46 am
Real name: Dennis Brown

Re: Electro-Static Linear Deuteron Accelerator Project Photo

Post by Dennis P Brown »

Thank you for the responses and the link. A lot to cover ... .

Resistors for a Van da Graff are very special items and not obtainable; regular (even Giga-ohm high voltage resistors) will cause massive corona discharges and shorting between the lens collector. I could design & build such a resistor but the effort would be a waste since needles are used (a common method in the past) and do work. As for the details of the current loss, voltage drop these are issues. For a big machine they are important to understand and adjust for maximum performance, but for my toy VdG, this would be a waste of time - issues (like you mentioned) of humidity causes more serious problems (and it is!) The needles are easy to adjust and are mainly there to provide focusing and, to a lesser extent, prevent opposite charge buildup issues.

By the way, the Einzel lens design like mine were commonly used in the past - they are "simple" tubes but have been rounded and polished for HV use. They use a static field to focus and are rather good even at low voltages but are far from the best design - many other designs work better but require far more work. If the accelerator was my principle project, focusing on better designs would be a good use of energy; however, this isn't my goal.

I agree with your information on magnets but my system is so complex I cannot justify further adding more secondary systems (LOL!!!) Even now, I am dealing with two new problems (one minor, one major) - I am installing a by-pass for a high vacuum purging capability of my deuterium feed line (will test later today); more of a concern is that humidity is killing my VdG performance (its spring and will only get worse! Also, my VdG needs a new belt!)

These issues have dramatic effects on my system - they will impact it working at all. Now, once I finally get/have a deuteron beam (any!!!) I could address issues to optimize this a bit (with a beam to study, it is far, far easier to adjust parameters and see the effects. Without a beam, it is all just theory and more work delaying me getting a beam!

Since my real focus is not the accelerator but rather obtaining a neutron beam (from the deuteron beam - I will have a deuterated target) in order to do some science experiments - the neutron beam need not be very intense but mostly just collimated and available.
prestonbarrows
Posts: 211
Joined: Sun Jun 24, 2012 1:27 am
Real name:

Re: Electro-Static Linear Deuteron Accelerator Project Photo

Post by prestonbarrows »

Don't know what your budget is, but keep an eye out on eBay for a nice glassman/spellman HV supply. This will give you a rock solid acceleration voltage and beam current of known values.

I'll add a bit more in regards to the grading resistors, even if you are not planning on using them (though I still think you should if your supply allows). You mention issues with corona or arcing on the resistors. This occurs at locations with high electric field which is subtlety different than electric potential (voltage). Mathematically, the E-field is given by the negative gradient of the potential. In more straight forward terms, the E-field is how quickly the potential is changing over a distance.

Static electric fields must always be perpendicular to a conducting surface. This means sharp points on conductors tend to create high E-fields while smooth round surfaces give much lower fields. This is why a Van de graff typically is topped by a large smooth sphere; it creates a relatively lower E-field on its surface allowing for a higher potential to build up before arcing/corona. One classic thought experiment is to imagine stretching a big rubber sheet over the device towards ground. Spots where the sheet is likely to break, like corners and pointy bits, will have the highest fields and greatest issues with arcing or corona.

Pull the imaginary sheet over a sharp needle and it rips right through. But, put the smae needle inside a hollow sphere, and things get spread out and the sheet holds. Even if something is at a very high absolute potential with respect to ground, you can make it have zero E-field on its surface by "propping up" the field with other conductors as in the example above. In terms of arcing, it does not matter if something is at 10,000,000 V if it is surrounded by things also at that potential. It is all about the change in potential through space. Looks like you already know this as your pointy source hardware is housed in a nice sphere.

In your case, you will have extremely high fields on the needle tips, quite high fields on the rims of the plates, and very little field on the faces of the plates (even less closer to the corners where they join the lens cylinders at the same potential). Attaching your HV resistors down in this region should give you no problems with corona/arcing. Just need a small hole through the plates to mount the lead through and solder on the back side leaving a nice round solder blob. You need nice long resistors rated for high voltage like these.

Image

If you wanted to get really fancy and were starting from scratch, you could mount them through the plates directly into the ends of the resistors, like shown below. By using counterbored bolts and spacing the plates the same distance as the length of the resistors you will get very good surfaces for holding HV and can shrink down the length of your accelerator to combat space charge effects.

Image

Also, you mention a 'neutron beam'. Just so you know, a solid target will still be pretty close to isotropic and neutrons will be produced in all directions. There is a slight peak in intensity/energy along the direction of the beam but it is typically only a fraction higher than any other direction. This effect depends on the energy of the beam.

What material is your target?
User avatar
Dennis P Brown
Posts: 3159
Joined: Sun May 20, 2012 10:46 am
Real name: Dennis Brown

Re: Electro-Static Linear Deuteron Accelerator Project Photo

Post by Dennis P Brown »

Excellent post and I realize that this issue should be revisited. Without a doubt, the last four stages of my Einzel lens system should be high meg or even giga resisters. This, as I just now realized after thinking over your post, is the location my lens' most likely will overcharge due to electron interactions caused by the target emission. Due to the very low electric field my existing resisters, using the ideas you have provided, should far better perform than simple needle point corona leakage devices (and be far less disturbing to the Einzel lens.)
Post Reply

Return to “Images du Jour”