Cube fusor build

For posts specifically relating to fusor design, construction, and operation.
User avatar
Richard Hull
Moderator
Posts: 14991
Joined: Fri Jun 15, 2001 9:44 am
Real name: Richard Hull

Re: Cube fusor build

Post by Richard Hull »

I agree 100% with Jim. Deposition related to high field points on the cylinder. You can the it on the stalk in the photo. There is a rather even deposit due to uniform field about the smooth high field stalk

Richard Hull
Progress may have been a good thing once, but it just went on too long. - Yogi Berra
Fusion is the energy of the future....and it always will be
The more complex the idea put forward by the poor amateur, the more likely it will never see embodiment
Jon Rosenstiel
Posts: 1494
Joined: Thu Jun 28, 2001 1:30 am
Real name: Jon Rosenstiel
Location: Southern California

Re: Cube fusor build

Post by Jon Rosenstiel »

I think Jim may have hit upon what's going on here.

Below images are of a brand new, freshly machined aluminum (6061-T6) cathode. I believe the greenish plasma is related to the “burning off” of aluminum oxide and/or other contaminates.

As a side note, neutron production rate of this cathode was about one-fourth that of its similarly dimensioned stainless-steel counterpart.

Jon Rosenstiel
Attachments
9-minutes into initial conditioning run. (20 kV, 15 mA, 14.3 mTorr)
9-minutes into initial conditioning run. (20 kV, 15 mA, 14.3 mTorr)
16-minutes into initial conditioning run. (20 kV, 15 mA, 24.9 mTorr)
16-minutes into initial conditioning run. (20 kV, 15 mA, 24.9 mTorr)
2 ~ 3-hours of runtime and several runs later. (20 kV, 15 mA)
2 ~ 3-hours of runtime and several runs later. (20 kV, 15 mA)
User avatar
Jim Kovalchick
Posts: 717
Joined: Wed Apr 13, 2011 8:00 pm
Real name:

Re: Cube fusor build

Post by Jim Kovalchick »

I actually posted an inquiry about the color bands before and posted some of my own pictures. viewtopic.php?f=18&t=13077&p=87033#p87033

I didn't understand them either. I have since come to the conclusion that the only things that could be making uniform and distinct color transitions are fields.
User avatar
Richard Hull
Moderator
Posts: 14991
Joined: Fri Jun 15, 2001 9:44 am
Real name: Richard Hull

Re: Cube fusor build

Post by Richard Hull »

I have yapped about fields for years here and a full understanding of them in the assembly of our super high voltage systems is key to avoid arcing internal or external to the system and naturally to deposition. Most of my knowledge and respect for high field conditions grew from the 12 years spent in Tesla coiling. The beautiful diamond lozenge images in the inner spherical shell of fusor III and IV due to the geodesic grid photographed 15 years ago, spoke to the field causal distribution of material via the multi-beaming ports. This is a form of incidental electrostatic focusing, deposition and heating.

For most every person in electronics 100 DC volts is considered high voltage for we fusion folks, 10,000 volts DC is considered far too low a voltage of any genuine value. Tesla coilers work in the million plus volt range albeit at RF frequencies. Field control is far more important in preventing arcing and huge electrical losses due to corona, (which can foster arcing). However it can also affect and control deposition in high voltage components in a vacuum system. I have grown so use to such depositions over these many years, I just do not give such things a second thought beyond being an indicator of high field regions, which are to be avoided or looked at as a possible danger point in the system. It also indicates a point of lost energy in the system or, conversely, a point of successful deposition where desired.

Richard Hull
Progress may have been a good thing once, but it just went on too long. - Yogi Berra
Fusion is the energy of the future....and it always will be
The more complex the idea put forward by the poor amateur, the more likely it will never see embodiment
Jon Rosenstiel
Posts: 1494
Joined: Thu Jun 28, 2001 1:30 am
Real name: Jon Rosenstiel
Location: Southern California

Re: Characterizing the cube fusor’s neutron flux

Post by Jon Rosenstiel »

Previous work with a 2” BC-720 replica fast neutron detector showed that the cube fusor’s neutron emissions were anisotropic in nature. viewtopic.php?f=6&t=12954&hilit=anisotropic&start=30

In order to “see” this in more detail I used a Hornyak button 8mm in diameter by 5/8” thick coupled to a Hamamatsu R6095 28mm PMT. This detector was then swept across the cube’s left end from edge to edge (100mm, 4”) using a linear stage. Data was recorded every 5mm. (One turn of the stage’s crank handle)

As I had no idea of the sensitivity of such a small Hornyak button I decided to cast three buttons of 8, 10, and 12mm in diameter. The buttons consisted of a mixture of ZnS(Ag) and casting resin. Mixing ratio was 5.7% by weight. The buttons, once hardened, were centered in HDPE molds 28mm in diameter and back-filled with clear resin. After hardening, the ends of buttons were machined flat, wet sanded, and then polished on a buffing wheel.

The 28mm Hamamatsu PMT and its housing are SAIC surplus. I have a few of these on hand courtesy of George Schmermund, but they are also often found on eBay. The machined aluminum endcap was my doing.

The SAIC units have a plus/minus 5V powered preamp attached to the PMT’s base, but the output is a 1-micorsecond wide pulse that doesn’t play well with spectroscopy electronics. I ended up taking the output off of the anode’s coupling capacitor and feeding it into an Ortec 113 preamp.

NIM electronics were comprised of a Canberra 3102D hv supply, an Ortec 572 spec amp, Ortec 550 SCA, Ortec 773 timer/counter and an Ortec 778 dual counter.

The “control” fast detector used to monitor NPR was a 2” diameter by 0.45” thick Hornyak button coupled to an EMI 9266 PMT. Electronics consisted of a Ortec 113 preamp, Ortec 571 spec amp, Ortec 550 SCA, and a Canberra 3102D hv supply. SCA output was fed into the 778 dual counter.

As each data run took around 30-minutes to complete, stable operation of the fusor was critical. Some of the steps taken to ensure stability. 1) Using another fast detector to monitor NPR. 2) Running at low power, 50 kV, 6 mA, 300 W, TIER of about 2.4E+06 n/s. 3) Directing the outlet of a portable A/C unit into the cube’s water-cooling radiator. 4) Running early afternoon when my lab’s temperature was most stable. 5) Long warm-up/conditioning period.

Jon Rosenstiel
Attachments
Data
Data
8, 10, 12mm Hornyak buttons with PMT
8, 10, 12mm Hornyak buttons with PMT
The setup
The setup
User avatar
Richard Hull
Moderator
Posts: 14991
Joined: Fri Jun 15, 2001 9:44 am
Real name: Richard Hull

Re: Cube fusor build

Post by Richard Hull »

Great data and presentation! Seems reasonable as the focus is very beam on target. I was rather stunned there was so much off axis neutron detection. Scattering might explain that, but so might some isotropic production in velocity space or maybe simple Maxwellian neutral-fast fusion. As always, follow the beaming where it exists. I am more tempted to go with scattering.

Great work with the homemade hornyak neutron detectors!

Richard Hull
Progress may have been a good thing once, but it just went on too long. - Yogi Berra
Fusion is the energy of the future....and it always will be
The more complex the idea put forward by the poor amateur, the more likely it will never see embodiment
User avatar
Jim Kovalchick
Posts: 717
Joined: Wed Apr 13, 2011 8:00 pm
Real name:

Re: Cube fusor build

Post by Jim Kovalchick »

As always Jon, you work is mastery.
.
I wonder if your work means that everyone who thinks they know their tier numbers needs to rethink it. I have speculated on this in the past when people have parked bubble detectors right next to their fusors and then over simplified an impossibly complex geometry by using point source calculation.

Jim K
User avatar
Jim Kovalchick
Posts: 717
Joined: Wed Apr 13, 2011 8:00 pm
Real name:

Re: Cube fusor build

Post by Jim Kovalchick »

As always Jon, your work is mastery.
.
I wonder if your work means that everyone who thinks they know their tier numbers needs to rethink it. I have speculated on this in the past when people have parked bubble detectors right next to their fusors and then over simplified an impossibly complex geometry by using point source calculation.

Jim K
Frank Sanns
Site Admin
Posts: 2119
Joined: Fri Jun 14, 2002 2:26 pm
Real name: Frank Sanns

Re: Cube fusor build

Post by Frank Sanns »

I was out your way for a few days last week Jon but could not get the time to get over to see you. Seems traffic has picked back up since the pandemic is winding down and we are becoming immobile again due to antiquated single person transportation systems.

My fusor does not get run for the multiple hours at a time on multiple consecutive days that many of you do. When I am done with a run, I bring the pressure part way back up with a bit of deuterium before letting it sit until I get back to it. It seems to accelerate the restart up to good neutron numbers again.

A 50 KeV deuteron alone is not going to change the direction of a 2.4 MeV neutron just based on collisional vectors. O-P is something else of course.

At first I was wondering why there was a discontinuity of the function (top three points on your graph) as you neared the top of the peak but I guess that is due to the diameter of the detector and the neutron production areas not being point sources. Still, it tells something of the production. Can you do a cure fit and find the equation for the curve for all but the top three points? Does it follow the inverse square law? If you post the data we can do it. Guess I could pick it off the graph but you may have already done it.

Great work! I really wish I would have had more time last week. Alway enjoy the visits.
Achiever's madness; when enough is still not enough. ---FS
We have to stop looking at the world through our physical eyes. The universe is NOT what we see. It is the quantum world that is real. The rest is just an electron illusion. ---FS
User avatar
Mark Rowley
Posts: 909
Joined: Sun Dec 29, 2002 12:20 am
Real name: Mark Rowley
Location: Sacramento California
Contact:

Re: Cube fusor build

Post by Mark Rowley »

DIY efforts like this are the best. Another outstanding presentation Jon.

Aside from the stellar work of casting your own buttons, the efforts you put into establishing a very stable and consistent neutron output are notable. Regarding chamber cooling, I’ve noticed here that 18 Celsius seems to be the magic number for stability. Anything below seems to be inconsequential and above the numbers begin to vary a bit more. What is the optimal temp for your chamber?

Mark Rowley
Jon Rosenstiel
Posts: 1494
Joined: Thu Jun 28, 2001 1:30 am
Real name: Jon Rosenstiel
Location: Southern California

Re: Cube fusor build

Post by Jon Rosenstiel »

Frank, too bad we couldn’t get together, hopefully you’ll have a little more time next time you're out here.

Best fit was with an exponential trendline. I have to wonder if the shape of the curve is affected by the geometry of the Hornyak button vs. the angle in which the neutrons interact with the button. For example, with the button centered on the endcap neutrons are most likely interacting fully with the button’s 8mm diameter and 5/8” depth, but as the button is moved towards the cube’s edge the neutrons are more and more interacting with the button from the side. (Hope this makes sense)

Mark,
For these runs with the A/C blowing through the radiator the chamber temp was around 28~29 C, lowest I’ve ever run. One of these days I need to try adding ice to the cooling H2O, try and get a handle on the "how low can you go" point.

Jon Rosenstiel
Attachments
Chart_02.png
Frank Sanns
Site Admin
Posts: 2119
Joined: Fri Jun 14, 2002 2:26 pm
Real name: Frank Sanns

Re: Cube fusor build

Post by Frank Sanns »

If there is a beam line of anisotropy, then why is it following the inverse square law? Pulling out of a beam (using the term loosely) should give a quick drop when starting to remove the detector. Once out of the beam, the drop would then fall off quickly.

I know most of you are convinced that there is more out of the ends but I still am not convinced it is due to neutron direction from fusion itself. I still think it is a geometric artifact on where the fusion is occurring. I know that ALL of the evidence does not exactly support that but some of it does like this most recent experiment. Still pondering.
Achiever's madness; when enough is still not enough. ---FS
We have to stop looking at the world through our physical eyes. The universe is NOT what we see. It is the quantum world that is real. The rest is just an electron illusion. ---FS
User avatar
Jim Kovalchick
Posts: 717
Joined: Wed Apr 13, 2011 8:00 pm
Real name:

Re: Cube fusor build

Post by Jim Kovalchick »

Frank,
I am with you on this one. This is why I have been harping on geometry of detection. You can't put detectors right up against shells when we have said there is a lot of fusion happening on the walls. All this variance proves is that the fusion isn't happening on the walls uniformly. Why would we assume otherwise when Jon's fusor has a single center line beam?

To understand if there is direction to the neutrons, we need to back out our detectors. I don't believe we are making directional neutron beams.

Jim K
JoeBallantyne
Posts: 326
Joined: Tue Jul 20, 2010 4:08 pm
Real name: Joe Ballantyne
Location: Redmond, WA

Re: Cube fusor build

Post by JoeBallantyne »

Jon, very outstanding work!

One thing you might try if you are worried the angular orientation of the beam relative to your detector might be having an impact, is to simply do a run where you leave the detector centered on the end cap where the total measured neutron flux is the highest, and then rotate your detector 90 degrees in the horizontal plane while maintaining it centered on the beamline. Start with the detector exactly in line with the beam, and end up with the detector at exactly perpendicular to the beam. You could rotate it in both directions (right and left), although I suspect there should not be much difference in the response between the two rotations. You will need to shift the detector slightly in order to keep the beamline centered on the horniak button as you rotate it.

If there is some detector sensitivity to angle, you should see a change in measured output.

You could then use that data to factor out any detector angular sensitivity from your measured neutron output based on side to side displacement of the detector.

Joe.
JoeBallantyne
Posts: 326
Joined: Tue Jul 20, 2010 4:08 pm
Real name: Joe Ballantyne
Location: Redmond, WA

Re: Cube fusor build

Post by JoeBallantyne »

Another experiment you could do, would be to figure out the exact location on the endcap that the beam is hitting. Make that the center of a circle of sufficient diameter that you can sweep your detector in a quarter circle around that point, always keeping the detector exactly aligned on a line passing through the center of that point, and so when the detector is at 90 degrees to the beamline, it doesn't hit your chamber. ie: make the radius big enough that your detector can sweep continuously around that point without impacting the chamber at any point.

So the front of the detector will be on the circle, and the angle of the body of the detector with the beamline will be zero degrees initially, and will end up at 90 degrees to the beamline. ie: the detector is always exactly aligned with a ray that goes through the center of the beam impact point on the end cap of your chamber.

If a large percentage of fusions are happening in the wall at that point where the beamline hits, and you maintain a constant distance from that point, and those fusions result in isotropic output, then there should be no difference in detected neutron output from that portion of the neutron signal if you move your detector as described.

Since the distance to the point where the fusions are happening is constant, and the horniak button will always be exactly perpendicular to the line that intersects with that point.

If most of the fusions are happening in the end cap point where the beam hits, then the amount of variation you get from this measurement should be significantly less than the one you get from the side to side sweep.

Joe.
User avatar
Richard Hull
Moderator
Posts: 14991
Joined: Fri Jun 15, 2001 9:44 am
Real name: Richard Hull

Re: Cube fusor build

Post by Richard Hull »

I am beginning to see this device as a BOT system fusor! Backing off my position many posts back in this thread. This thing (cube) as long as the ends are kept cool might just implant D in the end caps and at the same time does a good bit of fusion there as all real evidence so well presented here seems to point out. I am also sure fusion is happening in velocity space as well and I'll bet there are good counts to be had from the tube arm walls, just nothing like Jon is getting on the beam line which is obviously where the action is. Of course, in activation work all this beaming will just slow and scatter in the moderator if it does its true job. However, it is nice to concentrate the neutrons at the moderator. That is where beaming comes in real handy. I like it.

Make no mistake, beaming or not, this is still a fusor!....A simple two electrode amateur fusor that is really cookin'! Jon is a master at setting up experiment and relaying great data and results of his efforts. This may be a long thread but a really valuable one that teaches and inspires thoughtful comment and is another way to skin the cat in amateur fusion.

Richard Hull
Progress may have been a good thing once, but it just went on too long. - Yogi Berra
Fusion is the energy of the future....and it always will be
The more complex the idea put forward by the poor amateur, the more likely it will never see embodiment
Jon Rosenstiel
Posts: 1494
Joined: Thu Jun 28, 2001 1:30 am
Real name: Jon Rosenstiel
Location: Southern California

Re: Cube fusor build

Post by Jon Rosenstiel »

Frank,
I’ve found that the inverse square law does not hold up well at small source to detector distances. I ran some tests using a 2” x 2” NaI(Tl) detector and a 10 uCi Cs-137 source and found that in order to get good data the source to detector distance should be no less than 4”. What I’m trying to say is, we probably should not put too much faith in close-in measurements.

As to the anisotropy, take a look at pages 18 thru 20 in the attached paper. (Characterization of Neutron Fields Around an Intense Neutron Generator) I believe this applies to our case, but not really sure. (Dammit Frank, I’m a mechanic, not a physicist) Anyway, I would like to hear your thoughts on this.

Joe,
Nice to hear from you! Thanks for the good ideas, I’ll keep them in mind for future reference.

Richard,
I’ve finally come to the same conclusion; the cube is a BOT device.

Pic of a Monte Carlo neutron flux simulation for a BOT device. Looks quite familiar, huh? Link to complete paper below. (Didn’t attach the paper as it’s nearly 18 MB is size)
https://escholarship.org/content/qt6z74 ... ee762b.pdf

Jon Rosenstiel
Attachments
Characterization of Neutron Fields Around an Intense Neutron Generator.pdf
(1.99 MiB) Downloaded 413 times
Image1.png
Frank Sanns
Site Admin
Posts: 2119
Joined: Fri Jun 14, 2002 2:26 pm
Real name: Frank Sanns

Re: Cube fusor build

Post by Frank Sanns »

Jon,

Good paper. Nice compilation of lots of information that we have been discussing.

Some conclusions.

1. We agree that close measurements are not good because neither the neutron producing area nor the detector are points. That is why I asked you to throw the top 3 points away on your original curve. I could see they were corrupted because of the near distance. The minimum distance for a good measurement will depend upon the size of the detector. Smaller detectors look more like point detectors at closer distances.

2. Deuterium is a unique beast in that it is such a simple atom, that minor shift between the center of mass of the atom and its charge location (a neutron and a proton), have meaning in the real world. The nucleus is neutral at one end and positively charged at the other. The inverse square at play so and impact from the neutron side is slightly more desirable since there is just a little less repelling charge at that end. By tritium, this effect is much less and the heavier atoms is probably non existent.

3. There is no neutron beam coming out of the ends of fusor. There is a higher probability of of productive collisions when the deuteron is aligned in an electric field that in a random field.

4. The paper also points out that the polarization of the neutrons are also anisotropic. Good luck with that portion of the experiment. Although, it is interesting to think about that as keeping your detector in place and rotating it 90 degrees should also show a difference. Want to handle some cadmium with slots cut in it for the front of your detectors. lol. I AM KIDDING. DON'T EVEN THINK ABOUT IT. Gadolinium is probably not a whole lot better. Maybe Boron Carbide pieces arranged into slats. Just thinking.

5. Your results are most excellent Jon. The effect is not a huge one but you are detecting it. It is important to point out that the confidence in your numbers is in your technique and setup but the repetition of measurements along multiple changes gives the confidence in the numbers. I single reading close and at another distance for a total of two points cannot tell the story. Multiple data points in good agreement gives a good confidence interval to evaluate the data.
Achiever's madness; when enough is still not enough. ---FS
We have to stop looking at the world through our physical eyes. The universe is NOT what we see. It is the quantum world that is real. The rest is just an electron illusion. ---FS
Jon Rosenstiel
Posts: 1494
Joined: Thu Jun 28, 2001 1:30 am
Real name: Jon Rosenstiel
Location: Southern California

Re: Cube fusor build

Post by Jon Rosenstiel »

Frank, you were correct in your skepticism, allow me to explain.

The following 0-degree/90-degree fast neutron-rate comparisons are based on the assumption that the neutron formation area (NFA) of the cube fusor is centered on (or very close to) the endcap surface. Detector for this test was a 2” diameter by 0.6” thick Hornyak button coupled to a Thorn/EMI 9258KB03 PMT. NIM electronics consisted of the following: Ortec 4001M minibin, Canberra 3102D hv supply, Ortec 113 preamp, Ortec 572 amp, Ortec 550 SCA, Tennelec 534 counter/timer. Both amp and SCA outputs were monitored on an oscilloscope.

My initial plan was to run a 0/90-degree comparison at 3” and 13” to the NFA. (13” is the maximum my setup would allow) At 3” to the NFA the 0-degree position count-rate was about 22% higher than that at 90 degrees. At 13” I expected to find a smaller difference in count-rate between the 0/90-degree positions, maybe somewhere around 15%. I was quite surprised when the 0/90 difference increased to 37%. Additional data points at 5.5”, 8”, and 10.5” did nothing to change things.

At this point I was somewhat baffled so I made a simple drawing of the cube. First thing I noticed was that at 0-degrees there is 0.375” of aluminum between the NFA and the detector vs. 2” of aluminum at 90-degrees. Seems so simple, but something that I had, up to this point, failed to recognize.

Ok, with the above in mind I placed a block of aluminum 1.625” in thickness between the detector and the endcap at the 0-degree position to mimic the 2” of aluminum at the 90-degree position. (Endcap + Al block = 2.0”)

With the 1.625” Al block in place a rerun of the 0/90-degree comparison at 3” and 13” resulted in the following.
At 3” to NFA: 0-deg measurement was 2.3% higher than at 90-deg.
At 13” to NFA: 90-deg measurement was 2.7% higher than at 0-deg.

So, what this is telling me is that the previously measured anisotropy was nothing more than neutron absorption/moderation/scattering in the cube’s aluminum walls. The following Richard Feynman quote comes to mind: “The first principle is that you must not fool yourself – and you are the easiest person to fool”.

Jon Rosenstiel
Attachments
Detector at 90-degree position, 3" from NFA. Shaded areas are 6061 aluminum.
Detector at 90-degree position, 3" from NFA. Shaded areas are 6061 aluminum.
Long-tailed pulse from the ZnS(Ag) loaded scintillator & data pulse from the SCA.
Long-tailed pulse from the ZnS(Ag) loaded scintillator & data pulse from the SCA.
Setup
Setup
User avatar
Richard Hull
Moderator
Posts: 14991
Joined: Fri Jun 15, 2001 9:44 am
Real name: Richard Hull

Re: Cube fusor build

Post by Richard Hull »

Great report and observation, Jon. You are giving a lot of good information based on interesting experiment in detection and sorting out anisotropy. Sure, fooling oneself is easy.

Richard Hull
Progress may have been a good thing once, but it just went on too long. - Yogi Berra
Fusion is the energy of the future....and it always will be
The more complex the idea put forward by the poor amateur, the more likely it will never see embodiment
Frank Sanns
Site Admin
Posts: 2119
Joined: Fri Jun 14, 2002 2:26 pm
Real name: Frank Sanns

Re: Cube fusor build

Post by Frank Sanns »

More great work Jon,

It may just be the physical distance measurement that is giving the effect. If you measured from the outer surface of your Fusor, then you started closer and that would explain the higher numbers that almost met the perpendicular readings as the distance increased. At greater distance, the measurement error on starting distance would be less of contribution.

I am not sure attenuation of neutrons by aluminum is much of a factor. It has a very low cross section so it should not be intercepting many neutrons.

Thanks again for posting the data. Good stuff!
Achiever's madness; when enough is still not enough. ---FS
We have to stop looking at the world through our physical eyes. The universe is NOT what we see. It is the quantum world that is real. The rest is just an electron illusion. ---FS
Post Reply

Return to “Fusor Construction & Operation (& FAQs)”