what shoud i make? ideas that need testing?

[UG!]

Hi,
whilst researching my first fusor design have been dismayed to see that most of what i thought of as my 'good ideas' have already been done. i was thinking of a tubular design (i have a nice 4" by 16"ish pyrex tube thats just asking to be used in some fun plasma device) but looking through the pictures someone has beaten me to it are there any novel ideas that need testing? bearing in mind i'm at the design stage and just have a heap of miscellaneous vacuum parts. There doesn't seem much point using all my resources building something that has already been done and done better than i could do atm.
Perhaps there should be a list of things that need researching?


Thanks
Oliver

[Q]

well, to my knowledge a cylindrical fusor (cylindrical grids, not chamber) has never been produced on more than a demo level. however, a pyrex chamber can not be used for fusion safely.
really, i see no reason why you couldn't repeat an expiriment that has already been done. repeatability is an important part of the scientific process. you may produce completely different results from a similar expiriment.

Q

[DaveC]

I will second that suggestion "Q" makes.

Actually it is wisest all around to plan to repeat a successful, significant experiment. The reason is that it is a learning experience for the builder. The expected results/performance are known, which does take away a little of the excitement of being the first to do it... but...there will be many alligators( or crocodiles) in the swamp, and you will have to wrestle a few of them along the way.

Making a clean and tight vacuum system, mastering the art of high voltage power supplies, and insulation, working out the plumbing for the gas delivery and pressure regulation, and setting up the necessary instrumentation to be sure you are safe and able to tell if anything actually is happening... all this is a full meal. And will definitely test and develop your experimental skills. Then.... you are ready to do something original and who knows? You may well find something knew.. and then we will all rush off to repeat your experiment..

And so the Research Process goes in real science.

All the best,

Dave Cooper

[Richard Hull]

To my knowledge no human being has ever built a neutron producing (fusing), IEC, cylinderical fusor! There is no data on the old Farnsworth efforts in this area, which were minimal. Yes there has been a lot of talk, but no action.

It is well worth the experimental effort.

Richard Hull

[Adam Szendrey]

Yea, i wanted to be the first here, but right now i don't have the funds nor the time to build it, though i have a fairly decent vacuum chamber and a diff. pump, but that's all. I think i'll be able to continue in a year or so (at least i hope so), but college takes a lot of time.

Adam

[Hector]

I must agree with the Cylindrical IEC device idea.

Richard, to my knowledge the guys at the University of Illinois have made a cylindrical IEC that was capable of generating neutrons and somewhere in my collection of IEC papers I know I have a report on a neutron producing cylindrical IEC made by some Japanese University team.

One of my friends Dr. Blair Bromley who was part of the UI team wrote his PhD Theses on a neutron producing IEC at UI.

Actually Blair is convinced that the cylindrical arrangement offers great potential for exceeding the neutron count of spherical IEC's because of its axial flow.


Hector

[DaveC]

Rashly assuming the convergence of the deuteron beams can be made a gradual as desired, (arbitrarily small angles) it should become evident with a little thought the gradual sidewise collisions of ions will never produce close enough center to center distances to allow fusion to occur, unless enormous accelerating potentials are used. Fusion requires a certain minimum relative center to center velocity in order for the masses to approach to a "fusing" distance.

The cylindrical devices, while offering NO improvement over the spherical fusor in the above regard, DO make a significant improvement possible: The axial collision distance can be made almost as long as one likes, by having opposing inline beams. Compare this to the miniscule collision volume at the center of the spherical fusor and you can see the potential improvement.

There are numerous caveats here, however. Among the more important are, the spreading from mutual repulsion, that becomes increasingly more difficult to control, as the beam density increases; The increasing difficulty in "synchronizing" the positions of ions, so that they can actually collide and not simply pass by as ships in the night (Theoretically, if all the ions could be made to travel in a single file, collision probability would be as high as possible. This requires sub-pico meter spatial resolution in two axes. Not a small task.)

The "final" major problem that lurks here, is what happens if and when a fusion event occurs. A neutron, proton and other radiative processes occur. Since the neutrons and protons have mass, and somewhat the same kinetic energy, there will be momentum issues, and that means some sort of overall cancellation of momentum vectors....which might.... not sure here... mean the fusion products exit UP the ion stream, thereby colliding with other incoming ions , and producing major scattering effects or... possibly some sort of enhancement effect.

There are some interesting things to ponder here. The entire experiment is strongly reminiscent of colliding proton and "other" ion beam experiments, which should be part of paritcle physics antiquity by now.

Dave Cooper

[UG!]

Thank you all for your replys,
I think i have come up with a design that will allow me to investigate both cylindrical and spherical designs, if all the parts i have accumulated fit together. (i will know when i go home in a few weeks.)

viewtopic.php?f=18&t=7731#p55690

was the cylindrical version i had seen.

Oliver

[Adam Szendrey]

We need to be careful here. It seems we are talking about axial flow devices, while i think Oliver wants to build a radial flow cylindrical device, though i might be wrong. We need to be clear about this i think. There has been confusion about this in the past, when i posted a lot about cylindrical arrangements (like the plasma sheet device, or the torus, or the cylindrical hybrid, etc...).

Adam

[Richard Hull]

The cylindrical fusor typically uses a FULLY enclosed inner grid. It is closed with hemispherical wire caged ends. The finished grid looks like a polish sausage or hotdog. A simple helix grid is not quite the same, though close.

Richard Hull

[Richard Hester]

The cylindrical format will be my preferred method of doing the grid when I finally get my fusor up and running. I have the chamber, flanges, pumps, supply pieces, but I'm flat out of time... Cylindrical chambers have the advantage of adaptability and easy access, as you can do what you want to the end caps and replace them fairly cheaply. An alternative configuration to the "Oscar Meyer" (popular brand of "hot dog" sausage in the U.S.) grid for the cylindrical fusor would be a modified Meiro spiral. This could be easily done with pliers and some 1/8" welding stock, using a piece of doweling as a mandrel.

[henryhallam]

Whatever you choose to do, put a lot of effort into the instrumentation. It has been the most difficult part of our project and is worth doing properly if you want to record real data. Interference is a bitch! We are now on our third generation of data collection PCBs, fingers crossed these will finally be able to consistently return valid data without crashing the microcontrollers every ten seconds. Whether or not they are successful I will make a detailed post with the things we have learned, after we get these PCBs fabbed and tested out.

The results from a new type of fusor geometry will have so much more meaning, and be a lot more useful, if the voltage-current-pressure-neutron characteristic data are reliable, precise and detailed.

[UG!]

> Whatever you choose to do, put a lot of effort into the instrumentation. It has been the most difficult part of our project and is worth doing properly if you want to record real data.

Yes, i agree. plan to make a scintillation counter for neutron detection (the problem will be getting it calibrated i expect) i just liberated some NIM modules today but unfortunately have no bin to put them in. I'm going to have everything interfaced to a computer for ease of recording data.

my main worry at the moment is finding a suitable roughing pump, as i suspect my single stage Edwards full of chainsaw oil just isn't going to be good enough for my turbo pump.


>Interference is a bitch! We are now on our third generation of data collection PCBs, fingers crossed these will finally be able to consistently return valid data without crashing the microcontrollers every ten seconds.

Ah yes, interface anoannoyance. I was planning to build an interface based on a sparc station 20 as Sbus is relatively simple but high speed. I was disappointed to discover the nasty IEEE want money to show you the supposedly 'open' standard

Oliver

[UG!]

Adam Szendrey wrote:
> We need to be careful here. It seems we are talking about axial flow devices, while i think Oliver wants to build a radial flow cylindrical device, though i might be wrong.

Yes i was thinking radial, i hadn't even considered axial. My idea was that it should be easyer to make a sausage shaped grid more transparent, as the wires need not touch along the length of the grid as in the normal spherical design and can still be thin unlike in a helical design (weather spherical or er sausagical a helix must support its own weight, but a sausage can be tensioned at each end)

Oliver

[Adam Szendrey]

I had posted my thoughts on the cylindrical design a while back. Somehow i'm a big fan of it.
I think that a cylindrical design MIGHT be more efficient at creating neutrons, even though it doesn't create a spot focus as a sphere , but a string.
Here's something that might interest you (honestly i'm not advertising myself here )
viewtopic.php?f=6&t=2465#p11975

Adam

[Mark Rowley]

Hi all. Here is the link to my demo application of the cylindrical grid system.

viewtopic.php?f=6&t=2467#p11977

Mark Rowley

[Adam Szendrey]

The plasma sheet system is one further step from the plasma string setup (which is the simple cylindrical setup), further reducing compression, and further increasing ion collision surface area, and it can also be made spherical.
I'm re-posting the relevant discussion here, so it's in one place:
viewtopic.php?f=6&t=2465#p11975

Mark, have you been able to get a deeper vacuum since then?
Thanks,

Adam

[AnGuy]

I am actively working on constructing a cylinderical fusor. I've been working on it for about two and half years (the building process has been very slow due to lack of time and time building up my metal fabercation skills). I have almost finished all of the machining work on the vacuum chamber parts, and I have about half of the parts welded up. This will be an all metal fusor, using nothing but conflat flanges, and its a monster, measuring 6in diameter by 22in end to end. However I still have to complete the PS output transformer, the radiation detection equipment, aquire a turbo pump, and build a frame to mount it. It might take me another 12 months to complete.

Now, I am not claiming I will have a working cylinderical fusor, just a built cylinderical fusor!

>i was thinking of a tubular design (i have a nice 4" by 16"ish pyrex tube thats just asking to be used in some fun plasma device) but looking through the pictures someone has beaten me to it

FWIW: Even getting to this point has been expensive. I probably have about $10K invested in parts and equipment, most of it in machine tools (tig Welder, Vertical mill, etc). I expect the total cost to be about $15K.The cost of the Pyrex tube is going to be insignificant to the other costs to construct a sound vacuum chamber, and aquired vacuum pumps, radiation detection equiment HV power supply. I am not trying to discourage you from building one, but to just point out the cost factor that you may not have considered.

Finally I would also recommend that you build it out of stainless tubing instead of pyrex. If you do build a working fusor (non-demo mode), you have to deal with x-rays. The stainless will block the majority of x-rays produced by the fusor, and keep you from glowing at night

[AnGuy]

>Ah yes, interface anoannoyance. I was planning to build an interface based on a sparc station 20 as Sbus is relatively simple but high speed. I was disappointed to discover the nasty IEEE want money to show you the supposedly 'open' standard

One alternate you might wisht consider is to use USB instead. While it not super fast, it should be fast enough for your needs. You can purchase off the shelf USB interface controllers to pass both char and block devices. It should reasonable easy to write code for either windows or linux. I was thinking of using USB for my detector data collection needs. Although your probably a lot further than me in detector development.

[Adam Szendrey]

Actually USB 2.x is rather fast, at a whopping 480 MBit/sec (which ofcourse is theoretical). I would say, that 10 Mbit/s should be enough for accurate, real time data collection for a fusor. A 1 Msps A/D converter is good up to about 100 kHz (or less, for increased resolution). For a simple fusor there is probably not more than ten sources of data (like a neutron counter, pressure gauge, voltage, current, etc...).
I have a four channel 200 ksps A/D (single chip) from analog devices. It has a serial output (which should be possible to be connected to an USB interface). Though i have never tested it. Will do as my time, but more importantly as my budget allows.

Oh, btw, this might help. Philips manufactures a series of USB controllers.
Here is their product line (semiconductors):
http://www.semiconductors.philips.com/m ... .html#hubc

Adam

[UG!]

> I have a four channel 200 ksps A/D (single chip) from analog devices. It has a serial output (which should be possible to be connected to an USB interface). Though i have never tested it. Will do as my time, but more importantly as my budget allows.

do you have a part number for that AD chip? i was looking for a serial ADC some time ago but couldn't find one.

Oliver

[Adam Szendrey]

AD974 AN is the part number.
And here is the info site on the analog devices website:
http://www.analog.com/en/prod/0,2877,AD974,00.html

It costs $ 31, but you can order a sample (that's what i did).
Analog devices has a very wide range of A/D converters, though not the cheapest, but their products are really good quality (nope i don't recieve any funds for this lol ).

Adam

[UG!]

> It costs $ 31, but you can order a sample (that's what i did).
> Analof devices has a very wide range of A/D converters, though not the cheapest, but their products are really good quality (nope i don't recieve any funds for this lol ).

yes i think free samples are invaluable to people like us maxim do them too

i do wonder if i'll get an angery e-mail one day asking why i order so many free samples but never actuialy purchase anything hehe

when/if i get a data acuisition system to work, i shall post the pcb designs here

Oliver

[Adam Szendrey]

Yea, i have ordered a lot of samples aswell. On one occasion i've accidentally ordered two of the same chip. I almost felt guilty .
Problem is that i never get to the phase when i build something from these samples lol.

Adam