top of page

Building a Fusion Reactor: The Beginning

Updated: Dec 21, 2023


I’ve wanted to create a desktop nuclear fusion reactor, specifically the Farnsworth Fusor, ever since Freshman year when I saw a video online of someone creating a “star in a jar”. This purple dot seemed at once so simple yet so fantastically confusing. I knew I had to make one. Initially, after some preliminary research though, I thought it would be utterly impractical. Living in a high-rise in downtown Chicago, even if we ignore the tremendous cost, there would be no space for me to manufacture such a device. Fast forward 3 years and here I am writing this document, as my team and I are currently working through our vacuum chamber iteration 1 in our school lab. Our baby fusor is being born!


First, I had to do plenty of research. This stage was probably one of the more fun parts so far, and I watched no less than 10 YouTube videos on fusors, read probably a gazillion essays and procedures by others who had created such devices, and found the love-of-my-life website,, which would be the starter and eliminator of many headaches.

One thing that was immediately apparent was that there was no one way to create such a device. People were fusing atmosphere particles to deuterium, constructing chambers out of lead and stainless steel encasing to pickle jars, and used everything from welding, iron cutting, 3D printing and epoxying. The end goal is to create a high pressure vacuum with the least amount of particles present as possible, then create a large potential difference between the walls and the inside, via use of a negatively charged anode. Atoms near the walls are stripped of their electrons due to the positive potential difference, and the positive ion is sent hurtling towards the negatively charged center where they collide with enough force to convert some of their mass into pure energy via the infamous E=MC^2.

My Initial Design

In a fusor, there are 3 main components: the vacuum chamber where fusion takes place, the high voltage system, and finally the vacuum system. Usually, there is also a deuterium intake, as deuterium gas allows for greater fusion than atmospheric particles, but acquiring this hydrogen isotope is expensive, dangerous, and complicated. As of now, I decided to do without that fuel, and instead just try to make it work with air.

My goal was to create the most feasible and simple design possible, so I finally settled upon, after numerous iterations, using a 316 inline sight glass (to the right) as my chamber. This tubing is normally used in liquid/gaseous piping systems, so provides a steady and mostly sealed way to construct a chamber, and already has a good shape. Additionally, the top and bottom lid could be already outfitted with holes, leading to high-voltage feedthroughs. These holes can be better sealed as parts are custom fit, and use NPT (National Pipe Taper), thus, less custom manufacturing is needed, and thus less human error.

I planned to use a stainless steel wire to create a inertial confinement chamber for the particles, with the idea that shaping it into a hollow spherical shape would cause the ions to get trapped in the center and thus increase the chance of a fusible collision. I find stainless steel to be most ideal due to its high melting point, conductivity, and resistance to things such as rust. This would be attacked to a high voltage system, allowing it to reach our goal level of 30kV. My initial voltage system design did have some faults and unknowns, such as the need for a capacitor, the lack of grounding, and lack of research or documentation for current. As of the writing of this paper, our high voltage system looks like the image below. Essentially, we use a Variable Transformer or Variac, which allows the user to control the voltage output, takes an outlet voltage of 120VAC at an extremely deadly 15A of current, and it supplies it to Neon-sign Transformer (NST) at 120-130 VAC at 60Hz and with a lower but still very lethal current of 5A. The NST steps up voltage to 12,000 VAC but drops current to a still dangerous but safer 30mA. Positive out is connected to earth ground while negative is fed through 4 (to prevent overheating) parallel diodes which act as a half-wave rectifier, converting standard AC to DC, but also only outputting 6kV DC due to half of the wave being blocked. Thus, the output voltage in the inertial confinement chamber (anode) in the fusor body will have a final voltage of 6kV so far, and a dangerous but lower 30mA current. 6kV is a safe starting voltage, and if tests go as planned, I might exchange the transformer for an X-Ray transformer to get a higher voltage set-up. Chamber walls will be grounded. There are still many aspects of the circuit that require clarification, such as the need for a possible current transformer and a low-pass filter, and as well as the possible emittance of X-Rays.

Finally, I decided to use a dual stage rotary vane vacuum pump, whose high-pressure and low-pressure system allow for higher quality, yet still affordable, vacuums. The pump I'm using is graded for 15 microns of pressure, which is a decent level of pressure for the low-quality fusion that we are attempting. All parts are fitted together through 1/4" diameter NPT (National Pipe Tapered Thread) except for the Vacuum pump (to the left) whcih uses inlet port fittings, so a inlet fitted barbed hose fitting has to be used instead of NPT for pipe connections.


Related Posts

See All
bottom of page