Electro-Static Linear Deuteron Accelerator

[Dennis P Brown]

Well, the leaks go on - only mid 10^-5 torr; however, the idea by one person here of using a can of duster spray to find leaks worked very well. I located two minor leaks in the accelerator tube/assembly - one I already suspected but the other really caught me by complete surprise. Fixing them may be problematic.
I'll try doing this leak repair project sometime this weekend if I can fit it in between the other minor and major projects that MUST be completed this weekend - the car oil change (I do over 600 miles a week!), and major roof repair (the first of two needed!)

If I get to the accelerator tube project and IF the leaks seal I'll move on. If by some amazing luck the humidity is low, and I reach vacuum then I'll even try the accelerator ... not likely but possible if I can repair the HV gun power supply; otherwise, if the accel tube does not reach the mid range for 10^-6 torr ... I'll simply smash the accelerator tube for parts and a bit of well needed revenge (lol) ... .

Still the #1 priority for the ESLA project is to build the new 400 kilovolt voltage multiplier (VM). The diodes should arrive any day now. The mounting board has been drilled/wired and large caps cut to size and test mounted. Even the NST and varic assembled and tested (by my calculations the NST's current output will be over kill for the VM: I need just 1 ma and the unit is rated up to 60 milli-amps so even with issues, as Richard's work has demonstrated, I'll still have current to spare.)

I just need the diodes to show up and then do a lot of soldering ... . Testing the unit will be an issue (for safety sake I'll only test just two sub-unit VM stages at a time!) Each sub-stage will be rated for 30 - 40 kV and so tested; the entire stack would be impossible to test and dangerous for the diode health. An accidental air flash over from the fully charged stack would instantly fry the diode ladder ... that will be an issue regardless until the entire unit is safely under oil.

Once the completed VM stack is under oil the high ohm/wattage resistor that is the main feed out for the high voltage field supply should then fully protect the diodes in the event of an air break down (yes, the resistor will also be under oil as well.) The final testing of this unit will be interesting; and hopefully, neither shocking nor explosive (diode wise)!

Also, a minor side project to complete is that I need to build a new and very small 25 kilovolt battery powered multiplier for the ion gun/accelerator globe housing. The diode and caps parts are at least a week away. Since I have decided to place this unit (when assembled and tested) under oil and fit it all within a small section of the gun housing globe this project will be both fun and interesting - unlike vacuum testing. Small, high voltage low wattage power supplies can be both thrilling and shocking to test at the same time - if you get my meaning.

I am seriously considering building a new accelerator tube (especially if I get around to smashing my current one) - when George Schmermund mentioned he wanted to build an accelerator tube with only hand tools I really liked this idea - a lot. And while he decided to move on I've been really working hard on designs along these very lines. I believe I have worked out a simple and rather elegant (I hope!) approach that will not just allow the ES linear accelerator tube to be made by only using hand tools but provide a better electric field environment, thus allowing me to create both more and stronger focusing Einzel lens that will provide tighter beam alignment.

When I get to that project I'll try and document its progress. Regardless of the old tube's state - pieces in the trash or working - the new accelerator tube project will be put on the front burner after the two power supplies are completed (I've even obtained a few key components already for the new design accelerator tube ... .)

Once last point - I have decided to be patent and not care about my own self imposed deadlines any more. This ESLA project will get done but done right. All safety and operational as well as instrumentation needed will be installed before I attempt first light. Addressing any x-ray dangers and obtaining quality beam performance are the two critical issues that I will address regardless of the delays this may impose. Doing it right isn't something that is nice but for a project like this - essential.

[Dennis P Brown]

Well, the diodes arrived and I really want to begin soldering (oh, what fun) the massive voltage multiplier stack together; however, I removed the accelerator tube from the vacuum system/shielding cart a few hours ago and discovered that the larger of the two leaks is located in a place that cannot be reached without cutting the accelerator tube's permanently affixed KF fitting off ... that design has massive fail relative to leak repairs. Missed that one.
Since I already have a new glass tube (half the diameter of the old one), most key parts, and the basic idea on how to make the electro-static lens without compromising the vacuum integrity of the new tube in any way, I will move forward (that is, take two steps backward and then three difficult steps forward ... .) Since I will need some parts from the existing tube - and since it can't get below 2*10^-5 torr due to that unreachable leak - it is is useless, anyway. So I guess the wet diamond band-saw will get another work out very soon. This will be the fourth complete build from scratch (not counting one rebuild of an existing tube) - learning through experience leaves a lot to be desired (lol.) I also guess this is the difference between a project collecting dust and one that, no matter how slooooooooowly does in fact move forward (but the snail like aspect is getting old ... or is that just me?)

[prestonbarrows]

Torrseal
http://www.ajvs.com/agilent-technologie ... 0001-15422

or

Epoxy patch
http://accuglassproducts.com/product.ph ... 23&cat=398

work very well for sealing leaks. Apply them while the system is under vacuum to help pull the sealant into the crack. Hitting it with a heat gun during/after application helps reduce the viscosity and get it flowing into every nook.

[George Schmermund]

It's an open secret that Torr Seal is just relabeled Hysol 1C. If you look on eBay or Amazon you can find it for about $10 or so. It is an excellent vacuum epoxy and I did an endorsement post on it way back when: viewtopic.php?f=10&t=3706&p=21114#p21114 . I have since become a convert to JB Weld after Tom Ligon's encouragement to try it out.

Except for its magnetic properties, viewtopic.php?f=15&t=7224&p=49096&hilit ... eds#p49096, I do like it better than the 1C. It's tough as nails and seems to stick to everything that I've tried it on. The 1C epoxy is more brittle than JB weld after curing and has some expansion coefficient issues with some parts that I've joined together. Either one is a good vacuum epoxy for unbaked applications.

[Dennis P Brown]

Thank you both for the suggestion - that is a very good idea.

Still, the allure (or am I just being stupid?) of making a new accelerator tube using no machine work is too interesting to pass up so I did dissembled my existing tube.

I will save the main accelerator section (with the Einzil lens assemblies) and maybe spend some time doing what you both are suggesting - can't hurt and a fall back position is worth the effort. The seals will be easy to test since adding a KF w/o-ring section is trivial.

Besides, I'll be either ordering the Einzil collector rings (again, cheap, off the "shelf" item requiring no machine work) or just bending some copper tubing to form a field collector - that is easy; I did that for my VdG and they work fine. The commercial collectors are 1/4 inch thick plastic that is mirrorized (is that even a word?) Either of these two items can be used to act as the field collector for the lens.

The lens themselves will just be adhesive copper foil wrapped onto the outside of the thick wall glass tube - since the Einzil lens' thickness is irrelevant the Cu foil should work fine (not trying for top quality optics performance here - more like a pin-hole camera optic performance ... ); since its the electric field and abrupt change between metal surfaces that is all that really matters or at least that is what the theory says.

Since the foils will be on the outside of the tube connecting the collector plates and Einzil lens will be easy - a little silver epoxy will do the trick. I'll have to place the foils under oil so the gap between them will not break down under the intense electric field. For that part of the job a clear plexi-glass short piece of tubing place between each collector plate will be filled with mineral oil covering the foils/Einzil lens assembly (and the interface between tube/collector plate will be silicone sealed.)

In this manner, the main accelerator tube containing the vacuum will only require an end gun glued to it (which is easy to seal) so the vacuum system should easily get down to the 10^-6 torr range (the system and target chamber currently does without the old (leaking) tube attached.

This is starting to get fun ... or at least, until someone points out the major mistake I am not seeing ... .

[prestonbarrows]

Having an insulator (glass) between the beam and your conductive lens can lead to problems. There will always be free, relatively slow, ions/electrons streaming outwards from the beam due to charge exchange reactions with background gas, ionization of the background gas from the beam ions, and simple scattering collisions. Also, since this is a basic 'zeroth-order' lens stack without tuned optics, one would expect a fair amount of the beam itself to hit the glass also. Depending on the insulating material, these charges can't really go anywhere when they hit and build up on the surface. The electric fields will always tend to pull in the species which will act to oppose the charge of the lenses and shield the accelerating field the beam sees.

All these extra charges on surfaces also makes the stack prone to arcing and unstable behavior.

Be careful about localized heating of the glass from the beam striking it also. Could easily lead to it shattering from thermal stress.

Accelerators almost always perform best when all surfaces with direct line of sight to the beam are conductive. This lets all the stray charges flying around flow down to ground when the hit a surface.

Here's a simulation of charge exchange in an ion beam through a background gas with higher pressure to the left. Blue points are uncollided ions and red are those involved with charge-exchange.



That said, I have never seen anyone try what you described. It should be simple/cheap enough to rig up this stack and try it since you already have most of the rest of the system together.

[George Schmermund]

In a recent post about the tube I'm building I mentioned "The photos show what the new parts look like. As can be seen, the electrode extends 0.100" past the Delrin insulator. This should reduce the tendency of scattered beam charging." Maybe I should have used BOLD type.

[Dennis P Brown]

Having an insulator (glass) between the beam and your conductive lens can lead to problems

I wondered about that - I do believe this will be an issue but for a test run, shouldn't cause major issues; still, long term operation may be impossible for that reason. However, could have a simple fix - later - once I have the system finally running and complete my critical neutron test (requires only short time period operation with very long time periods between any runs. Will need only a few short runs to complete the experiment.) I appreciate that warning and I will follow it carefully!

I will use bleed "resistors" to keep the charge buildup under control on the collectors. Simple potential "needles" will provide a continuous cornea discharge to prevent overcharging (they will be symmetric to provide as little disturbance as possible to the Einzil lens.).

[Dennis P Brown]

Started on the new accelerator tube and darn, if I am not now changing horses in the middle of the stream - I may add one drilled hole in the glass tube to access one electrode for deuteron shielding from the VdG field.

Started building my main voltage multiplier and was shocked (almost literally) when something shorted big time (nice explosive sparks) but after four or five repeated tests, no shorts occurred and voltage appeared to be holding ( a 7.5 kV NST AC supply.) Maybe a cap is now fried and is conducting but in a stack, would not be easy to determine - well, I do like de-soldering (lying here) ... not clear what the best test for a cap is - guess I will use a high voltage signal through a diode and resister chain into a micro-amp meter to see if the cap is shorted. Guess I should check the local diodes while I'm shocking myself ... I mean testing the components. I am using 12 ohm resistors to feed each NST lead into the diode/cap stack to provide some ballast resistance but since the unit is air grounded so the load shouldn't be too much. A bit confused since my HV probe is giving misleading values - nothing is simple in the world of high voltage (or just my ignorance ... .)

Glade I use a safe NST supply or the lighting up could be problematic .

I think the caps are low value but the thing is covered with numbers/letters that fit no known standard(!) but what I can find but a site that has similar caps indicate they are 4700 pF and rated 20 kV DC (6 kV rms)

Update: Yes, one cap was fried; used the standard cheat method - tapped the two leads of the cap to a 10,000 volt DC power supply poles and checked all the caps to ground (air gap spark.) All passed but one and that one clearly was "shorting" against the power supply poles as I tried to charge it.

Yes, I know, not good for caps but I was tried of soldering/desoldering a significant number of caps in the VM stack a few times ... .

Have a 25 kV door knob cap that I will use at the end of the voltage multiplier stack to replace this failed cap - that even offers some help since the higher Farad cap will provide a bit more stability for the end of the stack for the high end potential access output.

[Dennis P Brown]

So, in line with what prestonbarrows warned me about, I finished the new accelerator tube using my original design rather than the "simple" external lens I had envisioned. The first vacuum test (today) bottomed out at a very poor 6*10^-4 torr. Some sealant and 1.1*10^-5 torr was rapidly achieved - using both the spray duster can method and alcohol tests shows no effect on the ionization gauge when it is bottomed out @ 1.1*10^-5 torr.

So out gassing is still significant and will be for a few days ( very likely caused by the epoxy used on end seals.)

The system rises at 1 micron every tens seconds at pump valve off and slows as times passes (60 minutes the system rose to 167 microns. Eight to ten hours - 720 microns.) I did install internal aluminium Einzil lens(these are now Al to reduce the danger of x-rays since copper is really bad in that respect) that are each fed by a steel needle through a hole in the glass tube body. The design (with sealant) works fairly well but I dare not heat treat the tube - microcracks would than grow rapidly causing a very likely tube failure. So I must wait for out gassing to slow by repeated pumping.

[Dennis P Brown]

On the accelerator front - really just an oversized ion gun - the out gassing has improved further: less than 100 microns/hour for the entire system with the new gun/accelerator tube. Decided to try the turbo and the system pumped down to 5*10^-5 torr in under 5 minutes (from 500 microns existing base pressure. This is similar to my best previous performance so leakage issues are under control - for now. Getting into the mid 10^-6 torr is another matter and I'll let it run for a while to see if/now long to get to the upper 10^-6 torr.

Again, first light is once again doable in the not too distant future if I can resolve the high voltage issue. Also, still have to build the mini-voltage multiplier for the internal, battery operated 25 kV deuterium ionizer. That will be a relatively easy build now I have the parts and know-how. I plan to put that in a small tube and fill that with oil, as well. Will simplify volt arcing issues inside the accelerator globe. Also, have to order the Einzel lens collector plates. These will be plastic with a metal coating. A lot easier both to install and in weight on the accelerator tube.

Been at this point before - that is, before new issues de-railed me and major rebuilds occurred. Again, three steps back, four steps forward. Wish it had just been one back and two forward!

Post edit: the system was hung up on 1.0/1.1 *10^-5 torr so I explored the accelerator tube/fittings etc and no luck. Couldn't find a leak. So, I check other locations (these have been checked before and withstood 5*10^-6 torr) and of course, a new one occurred; a window port for the beam target had its O=ring leaking a little. Some alcohol forced it to 9.5 *10^-6 torr.

So I shut the system down, back filled with dry air, and disassembled the offending KF joint. Carefully examined all surfaces and all seemed good. Then cleaned off all the surfaces and O-ring. Used some high vac grease (put on, then wipe off - while I generally avoid using that stuff, sometimes I just give in. The o-ring/surfaces looked perfect so I decided to try a more drastic step) and re-installed. Will try again.

Nice (not) when a previously sealed connector springs a new leak ... .

Post, post edit: success to start. After less than 45 minutes (from atmosphere, no less) the system (accelerator tube) hit 9*10^-6 torr. I will say, the use of a can of dust off (compress gas for cleaning key boards) really works well for high vac leak detection! That was a great suggestion!.) Very pleased by this result.


Add on post to the post, post edit - the system reached 6.1*10^-6 torr after an hour and half. Sufficient for deuteron work so I'll stop at this point and worry about the high voltage systems (one is a simple build for the deuterium ionizer, the other is more art, apparently, than science. Creating the high voltage for the electric acceleration field has been up/down even within five minutes!) Still, progress is progress and getting the new accelerator tube to sufficiently low vacuum to operate is a nice goal to check off. Still, amazing the number of techniques one has to use to deal with leaks.

Maybe dealing with the Van da Graff will run somewhat more smoothly (hasn't in the past; hence, the VM project-which has become its own disappointment) ... not going to believe that one- I think I will make a new, better belt for the VdG (since the company that made it is out of business I've no choice in the matter. Still, a somewhat wider belt with better material properties might solve most the HV problems - then again, that will not be known until I manage to get to that point; once more, another project, new skills to master and problems to solve - these sub-projects to enhance/improve an existing system is getting old.)

Relative to that program, I've tried installing a number of HV static pick ups/corona generators (pins/needles/screens/mess/high current braided cable (to name the major ones) but really most have failed and few perform as advertised - again, apparently art, not science. Even the 25-30 kV sprays to develop a positive potential on the belt has been a rather depressing no effect on VdG performance - strange.

Well, back to the small VM supply - order the parts I will need for the belt & accelerator tube and do more waiting.

[Dennis P Brown]

Well, a few up dates on trying to finish my deuteron accelerator project (or at least, just get to first light ...):

First off, the project was delayed for good reasons. I was rather busy last week since my daughter (Jeannette, who some of you meet at Richard's place during the previous year's "forum") was home from MIT for a few days. This was the only time she could visit me this year since she is working full time for a Professor @ MIT all of this summer. Very glad she is working in her chosen field (theoretical Physics) but very sorry I cannot get more time with her this summer ... .

On the deuteron accelerator front (or also called an ion gun on steroids) I just received the new low voltage (3 kV) battery operated power supply (boy, are those things delicate - one mistake and the power transistor fries instantly - very bad design. I've order some of these transistors so hopefully, I can than fix these darn units when they do fail - these things border on being junk.)

The voltage multiplier for the accelerator's own ion gun is now complete, sealed in oil and works great (easily achieves 25 kV using the a fore mentioned HV supply.) The cap/diode stack of the multiplier system fits nicely in a small plastic bellows tube (just the end piece of a plastic drain pipe tube - what's nice is that a KF-25 vac plastic end cap covers fit's perfectly on one end of the plastic pipe!) The complete unit works far better than I had hoped (at least this voltage multiplier system does ... one nice thing about building voltage multiplier units is that I now know a bit more about their problems/limitations and issues. I guess that really is a good way to learn about real world issues, sometimes.

While this unit really does not need oil coverage to operate well, by placing the HV multiplier stack in oil anyway I can better enable all the system's pieces - the 12 volt lead-acid battery, HV power supply, voltage multiplier, rheostat, low and high voltage wiring - to fit nicely within the stainless steel globe while significantly lessening the danger of any flash over by the really high voltage output; the issue is that all these pieces in the globe are rather a tight fit; besides the metal surfaces of the globe itself, all the wires are just too close for comfort with this HV unit.

This SS globe also fits on the end of the deuteron accelerator tube and contains the ion gun/deuterium gas line and the rather large ring magnet for the gun. Since the globe will be charged to 150 - 250 kV (if I can get the Van da Graaff working again like it use to do!), this globe must float (i.e. be electrically isolated) in the "air" so all this weight must be carried by the end of the accelerator tube.

Speaking of the VdG, I have made a series of replacement belts for the unit and look forward to testing them this weekend. This so-called "minor" mini-project has been more of a pain than I thought (or the article prepared me for.) Learning to properly cut and then piece together/glue these belts is a bit more complex than I ever imagined - getting the length correct is really an issue. The new belt material does not stretch at all so getting the exact length is very critical and, of course, more difficult to determine than I realized. Glade I had a lot of the material ... .

If this new belt does not fix the problem with the VdG performance issues than I guess I will be forced to return to my alternate power supply approach and restart backup plan #1 again. One good thing that came out of my failed attempt to build a 400 kV voltage multiplier stack is the unit I did build easily reaches + 60 kV and works well as an ionizer for the VdG belt spray system (this controls VdG polarity - rather critical if one wants to accelerate the deuterons rather than electrons ... .)

The "sub" project for the new plastic collector plates - i.e. to charge the accelerator Einzil lens - has developed into a major issue of its own (what else is new) since these "mirrored" disks are insulated by "paint". Did not see that problem/issue coming - so, I will now attempt to vapor coat the plastic disks (paint side) with some copper in a few days - really need that to work since my regular copper plates are a bit heavy and weight on the accelerator tube is getting serious.

So far, not a single subsystem with this accelerator project has run smoothly - each and every one of these support systems has become its own mini-project and major headache ... . I guess that is SOP for any system of this complexity but I think an accelerator just seems to attracts far more serous problems than most other projects I have read about here. While I am learning a lot as a result, I am starting to realize that assuming people's designs as posted in articles, just will work as well as they claim. I do believe this is for a number of reasons - most, honest errors but still I feel careless. For instance, I now realize that for some of the posted accelerator projects that one finds online, that their HV power sources just didn't work any were near as well as they think. Also, for the few projects that did work, they left out a host of problems/issues that they did end up facing in achieving their working equipment but just wanted the project to appear faultless.

I think I'm starting to understand ITER's real issues ... lol.