Lab Pictures 1

August and September 2006; Room 102-A, Research II

                           NCSU, Raleigh NC

            My faculty advisor, Dr. David Aspnes, had some spare space in his lab in Research Building II in the Centennial Campus here at NCSU. In August 2006, I started construction. Two hemispheres arrived from Vancouver, and the physics instrument shop welded plate onto them and drilled holes for the antenna feedthroughs. Magnetrons, capacitors, and other gear came in largely from eBay, and the reactor took form—although a very early form, to be repeatedly torn apart and reconstructed over the next three years.
            Cross-talk between antennas has a dire effect; microwaves from one antenna go into another, up the coax cable, into another magnetron, and causes an arc that can destroy the magnetron. Initially I tried isolating each antenna by means of thin sheet aluminum, bolted onto the walls via flanges using conductive silver epoxy. In practice, the baffles were not needed when plasma formed and thus shielded the antennas from each other. When the pressure was too high for plasma, cross-talk was severe even with baffles and the reactor could not operate without magnetron damage. Baffles also severely limited the camera’s ability to observe inside the sphere, and sometimes suffered arc damage when using the sparker.

New Hemispheres

Fresh out of the box, two aluminum hemispheres, 22" diameter, 3/8" thick, spun in Vancouver

With Gasket

Trying out the seal, made of 1/16" rubber and hose clamps

First Capacitors

44 caps, 5100 microfarads 350 V; I now have 16 more 5000 MF at 450 V. Lethal! Fun!

End Shot

Mounted on framework made from two shelves that roll on casters; see polar pipe and struts

Equatorial Magnet Spindle

This plastic spindle will hold the equatorial parts of the spherical magnet that surrounds the aluminum sphere.

Early inside

Shows the inner surface finish before baffles and antennas clutter it up.


1100 W oven magnetrons in 4 banks of 5, each bank using a single squirrel cage fan for cooling.

First Joined

Hemispheres joined together before bells and whistles complicate matters.

First Joined

Another view of joined hemispheres.

Gluing slots

To avoid eddy currents when the magnetic field pulses, there's a longitudinal slot I filled with epoxy. Let's hope it holds under vacuum!

Polar baffles

Each of the 20 icosahedral helical antennas is separated by aluminum sheet metal baffles to avoid cross-talk and magnetron destruction. Both sets of 5 polar antennas have this baffle. Sparker and borescope enter through the poles.

Equatorial Magnet

I inspect the equatorial magnet spindle.

Hemisphere's baffles

This shows the equatorial baffles attached to the polar baffles; each polar baffle has half the equatorial baffles which must fasten on after the antennas are inserted.