Lab Pics 2 - Physics - Bill Robinson's Website

Lab Pictures 2

October and November 2006; Room 102-A, Research II

NCSU, Raleigh NC

For economic reasons, and also since several fittings are custom and not commercially available, I have made coax connectors and adapters by hand for each of the 20 antennas and magnetrons. There are inevitable reflections and losses, not to mention leaks, in the circuit that would require several thousand dollars of equipment to avoid and correct. The reactor works reasonably well at the intended pulse duration, but it cannot be considered anywhere near to optimal in safety and efficiency, and cannot be run CW. Each magnetron produces 1000 W; this is within the power rating of the coax cable, but only standard connectors would be durable and reliable even with pulsed operation (at the usual 0.2 s).

The first connector required for the 20 magnetrons lead from the antenna stub usually used to feed microwaves into the oven waveguide, into the coax cable instead. This required an inner and an outer cone, proportioned for 75 Ohm impedance. The cable connected by sliding the inner connector into the inner cone and squeezing the outer shield onto the outer cone with a hose clamp and conical aluminum pieces. This technique worked very well and proved durable and with no evidence of leakage or damage over three years of operation.

For the first several months of construction, I anticipated winding magnets and fitting them around the pressure chamber as in the original proposal. However it became obvious, after I had constructed spindles for the magnets, that this would be expensive, elaborate, and probably impossible for me to accomplish alone in the time required and with only meagre student loans for financing. After operating the reactor it became clear that my original proposal was entirely too tidy and optimistic, and that the idealistic simplifications would not apply in the real world. The reactor design and hardware configurations continued to evolve dramatically all the way to the fall of 2009, when active experimentation halted in order to write my thesis.

For initial plasma formation and introduction of experimental aerosols (vaporized organic powders), I used a miniature coaxial plasma jet (the sparker) at the end of ¼ inch outer diameter stainless tubing about 65 cm long. The inner electrode was a 5 inch long, 1/16 inch diameter tungsten welding rod, with the cavity at the last half inch, the rest sealed with a low-temperature glass frit backed up with porcelain. The current pulse (negative from the capacitors) went along the central tungsten electrode, through the target material placed in the end (organic powder with carbon dust), and back through the stainless tube to ground outside the sphere. It ran in the south polar pipe and just past the inner baffles, carefully insulated from all metal contact, since any path to ground besides the intended one eroded a good deal of metal by intense sparking.