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 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.

Cones taper from magnetron antenna stubs to coaxial cores (75 ohm)

Cones mounted on magnetron antenna stubs

Rubber plugs seal 1/2 holes, 10 ga. copper wire for antenna

Collection of 20 antennas

Copper helical antenna coated with ceramic

5 polar antennas without baffles

5 polar antennas, no baffles

5 polar antennas with their baffles

5 polar antennas

Can't resist such a cool shot

First coax attached

Coax through home-made connectors to antennas

Coax from lower 5 magnetrons to antennas, South hemisphere

Coax attached, midsection South hemisphere

Overview South hemisphere with coax

Coax with plastic magnet spindle in place

Upper 5 coax cables South hemisphere

Borescope (ProVision 300) enters N pole

Hey! I see something!

Fitting borescope through polar pipe

Making right-angle coax connector from plumbing parts

Soldering small tube onto coax core, covering with copper plumbing part

Filling connector with hot glue

Coaxial sparker goes through this guide tube along polar axis through baffles

Centering guide with PVC spacers

Valve, hose clamps, center guide for sparker (S hemisphere)

Valve with coax sparker tube (stainless steel, 1/4 inch); allows removal while keeping vacuum or gas mixture

Sparker valve

Magnet spindle construction; PVC hemisphere sliced and I added this flange

Magnet spindle construction

Mounting spindle

Mounted spindle

Mounted spindle; won't need the magnet for ball lightining experiments

This spherical magnet provides the cylindrical cusp field required for SMC