Some pictures of Bert Hickman's 10" coil This group of pictures show stages of assembly for homemade flat-plate LDPE caps, 10" diameter coil, details of primary construction, and some shots of the coil in operation. The coil was designed and constructed by Bert Hickman of Woodridge, Illinois (a suburb near Chicago, in the USA). Thanks and credit also go to the pioneering work done by Richard Hull, Richard Quick, Duane Bylund, and many others in the Tesla Coil Builders Association, and the master pioneer and designer himself, Nikola Tesla. Safe coilin' to ya! -- Bert -- E-mail: bert.hickman@aquila.com. CAP1.JPG: Two identical flat-plate capacitors were fabricated. Each is designed to operate at 20 KV RMS in Tesla Coil (TC) service. Dielectric consists of multiple thicknesses of 0.004" Low Density Polyethylene (LDPE) sheet. Heavy duty aluminum foil plates are "sandwiched" along the entire edge via 1/8" thick extruded aluminum angle-iron and flat stock via sheet metal screws. Eight layers of LDPE are used between plates, and 37 of these capacitor elements are paralleled in each module. Dielectric size is 1 4 x 9.5 inches (as it comes folded on a 50' x 3' roll from the hardware store). A one inch margin of LDPE surrounds each aluminum foil plate, and the corners of the plates are rounded. Each module has about 0.044 uF of capacitance, and can operate at 5 KV RMS in TC service. Each module has 1/8" thick PVC end-plates (dark gray), and nylon separators (blue). The modules are "loosely" clamped by screws through the PVC end plates into the separators. CAP2.JPG: Here is a bottom view of four identical capacitor modules showing how they will be positioned together. Heavy tinned copper braid interconnects the modules. The dielectric is oriented so that transformer oil (Shell Diala-X) can "wick up" between the sheets of the LDPE, eventually entirely filling all spaces. With careful break-in under TC service at gradually increasing voltage levels and operation times, virtually all trapped air will be displaced by the oil without any vacuum pump-down being required. The series-connected modules result in a robust 0.011 uF 20 KV RMS capacitor. CAP3.JPG: All four modules are electrically and mechanically assembled in this view. 3/8" PVC is used for the bottom support plate and the top module tie-bars. Tinned copper braid permits a degree of movement to occur (i.e., physical contraction and vibration under high voltage AC and pulse discharge operation). The total effective dielectric thickness is 128 mils. Because 32 layers of 4 mil dielectric are used, the design is very resistant to defects and fault propagation. Once all air has been excluded, the capacitor can withstand peak voltage spikes in excess of 70 KV and tank currents of 200 - 300 Amps. a There is probably no application tougher on a capacitor than in a disruptively excited Tesla Coil! CAP4.JPG: The final stages of assembly. The completed capacitors are terminated with 1" copper braid to 3/8" brass threaded rods, which pass through 1/2" Nylon details. Sealing is via buna-n sheet material. The containers are Rubbermaid "Roughneck" trashcans (LDPE), filled with about 5 gallons of Shell Diala-X transformer oil. The tops are then sealed via buna-n gaskets and bolted on. The matching capacitor can be seen to the right. Each finished cap weighs about 45 pounds. These were constructed before I knew of a reliable commercial source for Tesla Coil capacitors such as Condenser Products. Mine were made to be tough, not small... PRIMARY.JPG: The primary consists of about 110 feet of 3/8" soft copper tubing ("refrigerator tubing"), mounted on 2" high Lexan (polycarbonate) standoffs, drilled for 1/4" diameter Delrin (polyacetal) rods. coil spacing is 5/8" Turn-Turn. The standoffs are slightly offset to provide a spiral winding pattern for the primary. The 3/8" copper tubing fits snugly between the Delrin rods, requiring no hold-downs. Successive 50' rolls of 3/8" tubing are joined via solder, with a short piece of 1/4" tubing inside. Wood base is about 34" square. The outer strike-ring is about 2" above the primary. Secondary nominal tune is 91 kHz, with primary tapped at about 15.25 turns via a heavy-duty alligator clamp. The lighter segments seen on the primary are small pieces of HDPE tubing to insulate tapping points from arc-overs to the clamp. RF ground is provided by three 8' ground rods along the outside of the porch. GAP1.JPG: Side view of a low power static gap. This is similar to the design developed by Richard Quick, but uses a taller 6" diameter PVC housing. This design has 2 layers of Copper Pipe so as to provide up to twelve 0.030" gaps in one package. A high-velocity fan pulls in cooling air. At left is the air "choke" which forces air through the outer portions of the sparkgap assembly. Spacing between the two rows of pipes needs to be increased for better performance under high-humidity conditions, since I get periodic flashovers along the inside surface of the PVC housing (about 2.5" long). GAP2.JPG: Inside view of static gap showing the 1.5" diameter x 2" copper pipes. COIL1.JPG: A front view of the "business end" of the coil. A 32" x 8" toroid caps the coil, which has about 1000 close-wound turns of #21 AWG Double Formvar magnet wire (31" long winding length). The PVC has been pre-treated by drying and coating with 3 layers of polyurethane. After winding, the coil was coated with a single thick layer of Behr Build50 clear epoxy. Coil ends are capped with 1/4" Plexiglas. The top has a 3/8" thick plate of Lexan glued on. A 1/4x20 Nylon bolt holds a 2" PVC endcap to the to p. This permits easy toroid height adjustment via changes to the length of a connecting 2" PVC pipe. The toroid is made from two 15" pizza pans and two pieces of 8" x 8 ft flexible aluminum ducting. A small "bump" of aluminum tape insures breakout in one area when desired. Coupling coefficient is about 0.21. COIL2.JPG: The author beside the coil. The coil is housed in a screened-in porch, and runs off two 15 KV 60 MA luminous sign transformers. However, once powered from a 19.9 KV 10 KVA Pole-pig, it will need to be fired outdoors. The coil should really be at a higher elevation. It was lowered to control arcing to the rafters. [When streamers hit the rafters, they follow the sap channels in the wood , causing any trapped moisture to explode. This causes small splinters of wood to rain down on the coil, and the resulting "raggedy" appearance of the rafters also makes for an unhappy spouse...] COIL3.JPG: The coil in operation. The point-to point distance is 63" from the "bump" on the toroid to the grounded conduit above. The discharges are white-hot, and very loud. A 6-gap vacuum gap is used in series with the static gap to provide a total gap of about 0.54". Most quenching is done by the vacuum gap. The static gap alone becomes overloaded at about 2 KVA. The system is running with about 25 - 27A of primary current drawn off the 120 Volt mains, and uses about 200 uF of Power Factor Correction (PF C) capacitance. COIL4.JPG: A 60 Watt clear glass light bulb makes a high-power plasma globe. The current is flowing through the ionized gases in partially evacuated interior of the bulb, capacitively coupling through the glass of the bulb, and terminating in a 62" long streamer to ground. Amazingly, the glass is not punctured (as long as run times are kept rather short). The glass does heat up significantly, and the current flow will also sometimes light portions of the filament. A corona discharge which is forced to firs t flow through a lamp will light a 40 Watt bulb, even if the streamers are not arcing to ground. More to come later...