The most common question that I get when I tell people I am trying to build a nuclear reactor is, "Gee isn't that dangerous?!? What if it blows up?" Well, yeah, it is kind of dangerous, but this reactor will never explode, much less produce a mushroom-cloud-worthy blast. It is definitely possible to electrocute oneself, but the "nuclear" part of the reaction is actually pretty safe.
There are many types of nuclear reactions, the most notable being fission and fusion. When people refer to "nuclear energy," they are most likely talking about a fission reactor. In a fission reaction, a nucleus of a heavy element (such as uranium) absorbs a neutron, becomes unstable, and breaks up into two or more smaller elements in addition to shooting off free neutrons. If there is enough of the heavy element in close proximity, the free neutrons from the first reaction will cause more reactions and so on and so forth. If this chain reaction is controlled and moderated, you have the makings of a power plant.
In fusion, the opposite occurs. Instead of a really big element breaking down, two or more small element nuclei combine (i.e. "fuse") into a nucleus of a larger element, also giving off free neutrons in the process. Unlike fission, it is not the free neutrons that sustain the reaction, but the force that is causing the small nuclei to collide and combine.
In a fission reactor, the nuclear reactions are moderated so that the chain reaction cannot spiral out of control. The Chernobyl disaster was an instance where the moderation failed and there was a catastrophic chain reaction. However, in a fusion reactor, this would never happen because an outside force is required to keep the reaction going, and once this outside force was "shut off" the reactions would cease almost instantly.
As far as radiation goes, fusion deals with light elements, such as hydrogen and helium, which in general are not radioactive. Some reactor designs use a radioactive isotope of hydrogen (called tritium) but my reactor will run on a stable hydrogen isotope (called deuterium). The product of the fusion reactions will be a mixture of helium and the aforementioned tritium. The helium is stable, and while tritium is radioactive, it has a very short half-life and low energy decay. Thus, the reactants and products have low, if any, radiation.
While the reactor is running, radiation in the form of fast neutrons will be produced. Although neutron radiation is dangerous, my reactor will not pose a significant nuclear threat anytime soon. First off, I will be lucky to produce enough neutron radiation to notice with measurement devices, let alone be damaging to a person. Secondly, the danger of the radiation drops off exponentially the further away one is from the reactor. In an article by Tom Ligon on the fusor.net website, he says that if one were to stand one meter away from a typical amateur fusion reactor, it would take 12 days of continuous bombardment before the person would even have to start worrying. Finally, as aforementioned, the reactor only produces the fast neutrons while it is on, and because of the finicky nature of the fusion reactions, I will be lucky to be able to contain my reactions for longer than a minute at a time.
To sum it up, as a fusor utilizes fusion instead of fission, my reactor will not be producing radioactive goo or exploding in a giant mushroom cloud anytime soon.
Here is a picture from a very low vacuum run (just under 2000 microns). We found that the atmospheric plasma would only ignite up to 2000 microns, and that the low vacuum plasma was much redder in color. Also, this plasma would ignite at a much lower voltage (around -400V) as compared to the higher vacuum plasma at -1500V.
Today, February 26, 2010 I turned the completed reactor on for the first time and it performed wonderfully! At a vacuum pressure of 20 microns and a potential of -1500V on the inner grid, a bright purple-colored plasma was observed. On the very first run, I was able to adjust the variac all the way to full power (-10kV on the inner grid) but the high voltage managed to arc from the container to the metal guide wire in the vacuum hose, grounding the whole system and extinguishing the plasma. After that we were not able to get the voltage past -5kV, but we still managed to get plasma, and played around with varying the voltage and pressure in the chamber. Below is a video of one of the runs:
This Friday, I hope to run the fusor for the first time. Today, Anatol and I put the system together and mounted the chamber on some vertical supports that Anatol made. We connected the chamber in series with Dr. Bulman's so we can measure the pressure inside the chamber. After everything was tightened down, I turned on the pump and started evacuating the chamber. Hopefully by Friday the chamber will be down to around 35 microns or so.
In order to create the high voltages required to energize the inner grid, a special power supply was needed. As commercial kV power supplies are very expensive, I decided to build my own. This supply is comprised of a variac tied to the main transformer which is output through a reverse full-wave bridge rectifier and then through a filter capacitor. Because the voltages were too high to measure with a regular multimeter, I made a voltage divider circuit to cut the voltage down enough to measure it. Unfortunately, the first time I measured the voltage, I failed to take into account the fact that the internal impedance of the meter was the same as the resistor I was measuring over, so my calculated results were wrong. However, once I factored in the internal impedance of the meter, I was able to (correctly) measure a maximum output voltage of about 20kV. I also measured a no-load maximum current of 30 mA. Finally, I measured the output using an oscilloscope. There was an appreciable ripple, so I might build more smoothing circuitry for future generations of the power supply.
The basic operating principle of this fusor is that there will be two "grids" in a vacuum. The outer grid will be a positive ground reference and the inner grid will be a highly negative voltage. When the grids are energized, the free particles in the chamber will ionize (i.e. the electrons will be ripped from the nucleus). The free electrons will be attracted to the positive ground (the metal chamber in my case) and the positively charged nuclei will be attracted to the negative inner grid. As you can see from the picture on the left, the inner grid should be as transparent as possible to allow the nuclei to fly through and collide in the center. These collisions are what should produce fusion reactions once deuterium is introduced into the chamber.
In order to get the high voltage into the vacuum chamber, it is necessary to have a vacuum sealed inlet for the high voltage. In the picture below, my feedthrough is shown in the vice. For the feedthrough I decided to use a spark plug (credit to Andrew Seltzman of RTF technologies for the idea). I chipped away the insulation on the inside of the plug and used a crimp connector to attach a copper wire to the center post of the spark plug. Then, Anatol and I J-B welded a high-alumina ceramic jacket over the wire to stop it from arcing to the casing of the spark plug. We cut and tapped a piece of aluminum to J-B weld to the side of chamber so that the spark plug could simply be screwed in (we used teflon tape around the threads to ensure a vacuum seal).
For the third vacuum test, I tested the system with the addition of the high-voltage port. For the high voltage feedthrough, I will be using a Bosch spark plug threaded through a 1/2'' thick aluminum plate, J-B welded to the side of the chamber. The schematic on the right is of the aluminum plate that was epoxied to the side of the chamber. In order to connect the spark plug, Anatol and I wrapped the threads in Teflon tape and used a wrench to screw the plug into the threaded hole in the aluminum plate. 
Here is a view through the viewport with the vacuum container assembled. The hole in the other side is the 3/4'' fitting which is attached to the vacuum pump. When the fusor is completed, the inner grid should be visible through the viewport.
Here is the lathe that I have been using to machine the metal parts. In this picture, the aluminum round used for the viewport base is being prepared for machining of the hole and the counter-sink.
In order to be able to see into the chamber, Anatol and I designed a viewport to be attached to the side opposite the vacuum fitting. We cut a slice off of the 5'' aluminum round to be the base and then I used the lathe to put a 2'' diameter hole in the center of the round. In order to have a place to put the viewport, I cut an additional 2 3/4'' diameter hole partially into the round to act as a counter-sink for the port itself. The port was made by cutting a 2'' round piece from a 1/2'' thick slab of acrylic. As per usual, we used J-B Weld to attach the base to the bowl and then the acrylic to the base.
After letting the J-B Weld set over the weekend, the system was ready to be tested for the first time. Dr. Bulman had a chamber set up with a thermocouple pressure gauge, so we attached my chamber in series with his, with both chambers being connected to my Alcatel roughing pump. We also included a shutoff valve between the pump and the two chambers in order to test the chamber's ability to hold a vacuum once the pump had been shut off. A sketch of the system is shown below. The middle chamber is Dr. Bulman's (with the gauge) and the one on the right is mine:
After the custom flange was machined on the lathe, Anatol and I used J-B Weld to epoxy the flange to one of the stainless steel bowls. We used a drill press and lots of cutting fluid to punch a 3/4'' hole through the top of the bowl before epoxying the fitting.
The upper, larger disk was epoxied to one of the bowls (after the bowl had been drilled) and the flange on the bottom was custom cut to match the fitting on the pump hose I already had.
