Brian's Fusor: The Story
←Back to "What it Takes" Fusion Main Home Ask a Question about the Fusor On to Parts List→
This is a running diary of all my fusion work, which is updated regularly. Scroll to the bottom to see my latest entries.
Updated: November 16, 2005
The Beginning: April-May 2003
Many people ask me how I came across this strange area of amateur science, and I point them back to a single event that occurred in late April of 2003. I was surfing the web, looking at a wonderful web page about a man's personal collection of all the elements of the periodic table (link here). Under a listing for Beryllium, the author had a picture of a beryllium oxide high voltage standoff insulator, and linked to an image that showed one "in use."
The image is at left and the caption read: "Want to see [the standoff] in action? Here's a picture of what looks like this exact insulator being used to make a fun little gadget, a tabletop fusion machine. I have it on good authority that this is not like cold fusion, that it really does work to produce fusion."
Fusion on the tabletop? I thought that was impossible. Like so many others, I had been made to believe that fusion could only take place in billion-dollar machines with huge amounts of input energy. Here was a man who had done it "for the price of a set of golf clubs!"
After reading the information at the various links given, I felt that such a project was within my capabilities, even as a high school freshman. My work in amateur rocketry had given me a bit of experience in the machine shop, and I also had some experience with nuclear processes.
By May, I had made my first post to Fusor.net, and had ordered Richard's informational video tapes on the fusor. By that time, I had acquired a 30 year old vacuum pump off of ebay for $75, whose operational condition was unknown to me at the time. I made a list of everything I needed and started gathering parts for a fully operational fusor.
The first piece of equipment I acquired was this 30 year old, CENCO "Hyvac 7" vacuum pump
My friends and family naturally thought I was off my rocker when I said I was going to make a fusion reactor, boy would I show them.
By July, I had decided on a preliminary design for my first fusor. It was to be made of all stainless steel, and be roughly 4" in diameter. It was necessary to have several ports placed around the fusor to allow for the input of electricity, Deuterium, a viewing port and a port for the vacuum pump. I had decided upon using "Conflat" fittings for these ports, which are a special type of vacuum fitting that use copper gaskets to create a perfect seal. Unfortunately, conflat fittings are somewhat expensive due to their high-precision construction. I pressed onward, getting only what I needed and not buying off of impulse. I snagged a conflat high voltage feedthrough (rated for 25,000 volts) and a vacuum pressure gauge early on, both for incredibly low prices on ebay.
Another problem was going to be the power supply. I had no prior experience with high voltages, and I felt building a fusor power supply was out of the question for a beginner. I had looked into the professional power supplies, like those made by Glassman or Spellman, but their prices were well into the kilobuck range. This issue was going to have to be tackled at a later date.
Fusion work slowed to a crawl at this point, since I had all the parts, but no machining capability. I had overestimated my own metalworking skills, and finally realized the amount of precision that would have to go into making this thing was something I could not achieve in a short amount of time.
I settled on a 6" spherical fusor, larger than I had originally planned, but easier to work with. My goal was to make my fusor similar to the Hirsch-Meeks device from years ago. It was to use all conflatted fittings, and two huge Conflat ring flanges would be used to bolt the two hemispheres together. The vacuum port was a QF25 o-ring seal flange, and the Deuterium inlet was a Swagelock "tubing to weld systems adaptor."
By January, everything ground to a halt as there were no machine/welding shops that could do the work with the precision I wanted. I had all the parts I needed to complete the vacuum system, but no way to assemble them at this point. I had also spot welded a few inner grids together from some stainless steel wire, which took quite a few tries to get right.
Click Here for a full parts list with diagrams!
By March, I had acquired a 20,000 volt, 50 milliamp Glassman power supply on ebay. It was the exact model I needed, even though the output voltage was a bit low for my tastes. I inquired with the manufacturer about getting an upgrade to 50,000 volts, but that would cost $2500. Not a trivial cost, but doable in about a year or so.
The best breakthrough came with the chamber construction. I had found a local company in Brockton, MA called "Sharon Vacuum Systems." They regularly build vacuum equipment for MIT, NASA and other high profile organizations, as well as individuals. They could build my chamber for a very reasonable price, and with a fast turnaround time.
The finished chamber was beautiful, having a polished chrome appearance to it. It could definitely do the job it was designed to do.
Also acquired during this period was an Eberline PNC-1 Neutron counter. It was old, but in good shape and operational. I sent it off to Ludlum Measurements for a comprehensive calibration, so as to avoid any claims that I had bad instruments.
The fusor as it looked when I first flipped the switch--June 30, 2004
Now that everything had been acquired, it was time to start assembling the system. I cleared out a small corner of my garage and started placing equipment on the table. I had also done some vacuum testing with the chamber to make sure I really could get it don to a low enough pressure.
Everything went together in a weekend, and all that was left by early June was to drive in a new grounding rod outside my house, to provide for electrical safety. By June 30, 2004, the fusor was ready for its first taste of power. There would be no turning back. I flipped the switch on for the first time at roughly 11am, achieving the first plasma in my fusor. I did some dry run tests with the neutron counter, making sure that the plasma processes did not create spurious counts. Fusion was only a bottle of Deuterium away at this point.
Excellent example of "star mode" fusion--Sept. 2004
After returning from a 7 week excursion in D.C. and the Appalachian trail, it was time to finish the work I had started. I ordered some Deuterium gas from Advanced Specialty Gases, which arrived at my front door in a few days. I hooked it up to my gas regulator, replacing the nitrogen bottle that had been there for the non-fusion tests. Upon admitting the Deuterium to the chamber, the neutron counter started to register counts, up to 10 or 20 times greater than the background count. This reduced to a total fusion output of 2500 neutrons per second. This was a low emission rate, but it was detectable with my calibrated instruments. I had joined the neutron club and thus became the 14th amateur to be credited with doing fusion in the home laboratory.
The early fusion setup in my garage--Aug. 2004
As time went on, I started to get better and better fusion results. It wouldn't be long before I reached the limits of my power supply. Better alignment of the inner grid produced a more stable plasma, and thus higher neutron counts.
At this point, word had gotten out that I had actually finished this crazy project. People at school wanted to see demonstrations, so a more portable setup would be required in the future. An article appeared in the school newspaper about my activities, thus adding an even greater level of interest in the project.
The current cart-mounted variant of the fusor--Dec. 2004
When I realized that I had to make the system portable, I purchased a moderately sized utility cart from a foodservice supply house. Amazingly, everything fit on the cart with only a few minor modifications, making a more compact and attractive setup that could be assembled and dismantled in less than 30 minutes.
It was with this system that I finally maxed out the capabilities of the electrical system. 25,000 neutrons per second was my limit with only a 20,000 volt power supply. The time had come for an upgrade to 50,000 volts. At the time of this writing, the supply is currently in the shop, and should be back by February of 2005. Hopefully, It'll allow me to get into the million neutrons per second range eventually, and do some neutron activation experiments.
March 23, 2005
Today was the first "official" documented test of the new 50,000V power supply. Using an input of roughly 25,000 volts and 20 milliamps, I achieved a neutron output of nearly 30,000 neutrons per second. This is actually a bit lower than expected, and I suspect that contamination in the chamber is to blame. Still, it is a small step towards the "millionaires club" (1,000,000 neutrons per second) that only a few amateurs have had the privilege of belonging to.
On a different note, the electrical insulation was tested at the full rated voltage and successfully did its job. I can say with great confidence that there will not be lightning bolts flying around the cart in the event of an accident.
April 21, 2005
I managed to activate some silver foil today using a neutron output of
61,000 n/s!
Here are the specifics:
---------------------------------------------------
Voltage: 39kV
Current: 20mA
Pressure: 12mT
Avg. CPM on PNC1: 95cpm
Avg. isotropic emission rate: 61,000 n/s
Silver Foil dimensions: 5cmX5cmX0.01cm
Irradiation time: 140s
Avg. Background: 27cpm
Max count rate immediately after irradiation: 210cpm
Moderator material: tamped paraffin pellets in plastic bags
Moderator thickness: 4.2cm on either side of foil
-------------------------------------------------
I took readings every 5 seconds using a Ludlum model 3 and a pancake probe.
Because of this, I could only get count rates to the nearest 5 CPM. When I
get my black cat detector set up, I should be able to make more accurate
decay curves.
Attached is an excel file with the data
entered into the table, and a graph of the count rate over time. While
the decay curve is a bit rough, you can see the exponential component to it.
I kept my Geiger counter well away from the fusor during operation to avoid
having it become activated. After irradiation, I quickly removed the Silver
foil from the moderator and put the pancake probe hard up against it. The
audible difference in count rate was noticed immediately.
This was a fairly crude setup with a crude moderator and crude data
collecting techniques. The only dead giveaway that activation occurred is
the exponential decay curve in the excel file. I'm also a bit suspicious
about my neutron numbers. 61,000 neutrons/sec seems a bit low for nearly 40
kilovolts on the inner grid.
SEE THE DECAY CURVE (MS Excel File)
July 12, 2005
After countless hours and lots of frustration, I have managed to cobble together a crude replica of the device portrayed in the original Hirsch/Meeks patent. This is how a fusor was meant to be made, and should put out many more neutrons than my previous setups. Simply stated, this design uses 3 grids instead of 2; the inner grid is still the same, and the outer grid is still the spherical vacuum chamber. However, a third grid has been added between the two, and it carries a positive charge of about 300-500 volts. Its purpose is to attract electrons that have been emitted from a heated (thermionic) Tungsten filament, which, in turn, produce ions.
This design allows most of the ions to be generated at the outer rim of the fusor, ensuring that they are at fusion energy upon arriving at the inner grid. In the simple, two grid fusor, ions could be created anywhere in the fusor volume, including inside of the inner grid; this prevents them from being accelerated to fusion energies. The 3 grid system additionally enables one to work at much lower pressures than before, since ions are continuously being created via thermionic emission. Previously, I had to rely on field emission in order to strike a discharge, but that process ceases at about 2 microns due to the highly rarified nature of the gas.
This link
describes the design in a bit more detail.
Picture 1:
This picture is of the interior of the fusor, showing the inner grid and the
"middle" or ionizer grid. The filament is too thin to be seen, but it is
supported by a series of ceramic tubes that are attached to a grounded
stainless ring and held in place by friction. The ionizer grid is supported
by 5 alumina "feet" and is a bit larger than 4" in diameter.
Picture 2:
The filament current (and thus temperature) is controlled by a standard 20V,
5A DC power supply, visible next to the TV monitor. Those two objects rest
on top of an old Gelman electrophoresis power supply, rated at 500 volts and
125 milliamps DC, which provides the positive bias for the ionizer grid.
Everything else from the previous variant of my fusor remains the same.
I'm in the process of getting this setup to go online. I'm going to set
aside tomorrow for a bakeout of the Micromaze and should hopefully start
fusing by Thursday.
As far as I know, the only amateurs who have tried this type of setup are Phil Fostini and Joe Zambelli, so this is fairly unexplored territory. Neither of them reported a great deal on the subject, so I had very little empirical information to work from when actually designing and building this thing. The H-M patent and Joe's old website were probably the best resources.
November 16, 2005
Today I finally got to test the B10 neutron counter tube that Carl Willis
gave me at the HEAS meeting. It worked extremely well, much better than I had
expected!
The tube was encased in a homemade wax moderator (6" PVC with a hollow 1.5" PVC
tube running down the center) and connected to a Ludlum 2241-2 digital scaler/ratemeter,
which I found to be much more reliable and insensitive against EMI than my
original setup. I did a 5 minute background count, over which 6 counts were
logged by the Ludlum.
After the background count, I set my fusor's voltage to 18kV, and admitted a bit
of deuterium to the chamber. Instantly, the counter started chirping away at 50
cpm! A bit of fiddling with the valves allowed me to get a final average count
rate of 150 cpm, and a maximum count rate of over 200 cpm! All of this was done
at a current of between 4 and 12 mA.
The B10 tube is very sensitive, almost approaching what I saw with Richard's
He-3 counter. I'm still a little hazy on how to extrapolate isotropic emission
rate from this counter though. A picture is attached above.
←Back to "What it Takes" Fusion Main Home On to Parts List→
