Excerpt From
Tesla Coil Experiments
Crossroads, GA, Lab Notes 1978 - 2002
First Tesla Coil, Mk I
Atlanta, Georgia, Georgia Tech dormitory, 1978 In an ancient leather-and-wood bound volume of "Electrical World" from a dusty corner of campus, I read original articles by 'N. Tesla', 'E. Thompson', 'T. Edison', 'G. Westinghouse' and others. The Tesla articles captivated me, and I vowed to build my own TC.
Classmate John Doran and I happened on a demolition of the (at the time) World's Largest Neon Sign in Atlanta. Crushed NSTs were buried in heaps of broken glass tubing. We offered the workmen $1 for every undamaged NST they could produce, and managed to load John's VW bug till the suspension was bottomed out and the wheels were scraping. I kept my share of the booty in my dorm room. From a campus construction site, I fetched a peice of 3 in grey PVC pipe. From Ack Electronics, I bought some #24 copper wire, a can of spray polyurethene, and also a 4 in ribbed porcelain insulator for the top, for no other reasons than it was cheap and looked real cool. These were my only cash expenditures, about $18 total.
Secondary Coil : wound on a form of about 30 in of 3 in grey PVC. The coil was close wound by hand with no jig, and sprayed with polyurethane.
Capacitor Bank : Leyden Jar type, six 32oz Miller beer bottles full of brine, and covered in aluminum foil. The bottles sat on an aluminum pan which was the cover of an old magnetostrictive delay line memory module. The pan sat on a stack of computer printouts about 10 in high. Since it was so much fun to empty the bottles, I did not count the cost of the beer against the construction budget.
Spark Gap : was a coat hanger cut in half and bolted accross the NST's output bushings. For noise abatement, I shielded it with a clear plastic cup liberated from the lunchroom, and poked holes in the side with a hot iron.
Primary Coil : the conductor, all connecting wiring, and the internal electrodes of the SW capacitors was scrap solid copper electricians wire from construction sites. The Primary Coil itself was about 10 turns about 18 in diameter supported with masking tape on paper dowels from clothes hangers.
Results : the TC tuned and operated satisfactorily for a hack effort, and provided many hours of terror for longsuffering roomate Mike Kazmierczak. The TC would produce thin controlled sparks about 10 in long and streamers almost that long. It would also cause the flourescent lights to glow when the switch was off, another source of entertainment and terror for the uninitiated. I allowed the secondary arc to strike my bare skin a few times, but usually clinched a metal fork in my fist to reduce the pain. I also could 'throw' sparks from my body, but the TC was difficult to tune and efficiency was awful. Eventually the dielectric stress was too much for the secondary coil, and surface arcing developed. This rapidly worsened and soon the secondary coil was 'eating itself alive' until it was finally useless.
Continued High Voltage Experiments
Norcross, Georgia, basement of house at Littlejohn Way, 1979 - 1982
Bert Behlow's Oil Filled Tesla Secondary Coil, 1979 - 1980
Bert wound a secondary with thousands of turns of extremely fine wire and sealed it in mineral oil. This was a finely crafted device, a machined housing of grey PVC with turned brass end electrodes. The housing was about 6 in diameter and about 15 in tall, inspired by early Tesla followers Elihu Thompson and Oudin, who made symmetrical balanced secondary coils, usually operated horizontally. Bert and I tried several primary coils with NSTs and SW caps, but the results were unimpressive. By brute force, the coil would produce thin, reedy sparks, but arcing over the outside of the case severely limited the secondary voltage.
High Voltage Power Supplies, 1980 - 1982
NST powered Jacob's Ladder : Just copper tubing bolted right on an NST.
DC power supply from NSTs : Rectified with fullwave bridge of 14kV, 1 amp diode packs and ripple filtered with doorknob type capacitors. MUST use 3 diode packs in each leg to prevent reverse breakdown and destruction. Internal equalization of packs was sufficient and no further compensation was needed.
DC supply for HeNe laser : Centertapped NST, with two arms of two series diode packs for fullwave output. Small doorknob capacitor and series choke filtered DC without defeating NST magnetic current limit.
Running HeNe laser directly from NST : A crude but effective method. Just hook up your HeNe laser like a neon sign.
Solid state "Light Saber" drive : Uses a 2N3055 and a tube type TV audio output transformer in a Hartley oscillator. This incredibly simple circuit works well enough to produce VERY PAINFUL shocks and lights up flouresfcent tubes. Run a thin wire to the other end of the tube and make a "Star Wars" light saber, circa 1979.
Solid state drive for Flyback Xfmr : Used a 555 astable multivibrator to drive a 2N3055 to switch a flyback transformer. Worked but not worth the effort.
Tesla Coil Mk II
Crossroads, Georgia, basement of house at Oak Hills Drive, 1998 - 1999
Hawg Capacitor A, Aug 1998
8.4 nF, using 2 l plastic soda bottle for dielectric. Contained in stainless steel stock pot. Inner and outer electrolytes are saturated aqueous NaCl, MgSO4 and Ammonioun Nitrate. Inner electrode is 1/2 in copper pipe, outer electrode is containment vessel. Standard test jig is test article in parallel with 2 cm static gap and 15kV, 30 mA NST. Improvements are: strip label, clean outside with alcohol, equalize inner and outer liquid levels, float 1/2 in mineral oil layer on top of inner and outer electrolytes. Longest life 30 minutes in standard test jig. Most punctures randomly distributed, porbably due to imperfections in PET bottles. Failed experiments: outer coating of polyurethane, bubbled under coating.
Flexigap, Static Spark Gap Design, September 1998
Secondary Coil A, Sep 1998
880 turns of #20 copper formvar inslulated closewound on a 4.5 in OD white PVC form. Top Load is 8 x 24 in toroid of black stovepipe.
Primary Coil A, Sep 1998
10 turns of 1/4 in refrigeration tubing 40 in dia in air supported by six upright 1/2 in PVC posts. One inch per turn, overall coil 18 inches high. Conductor attached with ty-raps to PVC set in wooden blocks.
First Light, 14 Sep 98
First light using NST and plastic soda bottle SW Hawg Capacitor. Sparks 3 inches. One NST and many plastic Hawg Caps died here.
PERFOMANCE: Controlled sparks 3 in, no breakout.
CONFIG:
Secondary : 880 t of #20 Cu wire on PVC pipe 4.5 in dia x 30 in long.
Top Load : Toroid, 8 x 24 in stovepipe
Primary A : 10 t of 1/4 in Cu tubing on six PVC posts 24 in high.
Power : 15 kV, 30 mA NST
Capacitor : Hawg Capacitor, 2 l plastic soda bottle in 24 qt stainless steel pot.
Spark Gap : 13 segment Flexigap, set at 5 gaps ~0.05 in ea
Primary Coil B, Sep 98
New tight coupled primary coil. Double spark length tuned at 22 turns. Sec-Pri strike stopped operation until variac increased. NST subsequently degraded over next several runs.
PERFOMANCE: Controlled sparks 5 in, 1 in corona from #20 extended wire.
CONFIG: 14 Sep, plus :
Primary B : 30 t of #6 Cu wire on 10 in green PVC pipe 18 in high.
Primary Coil C, Sep 98
Installed inverted conical primary coil "Primary C"
PERFOMANCE: Controlled sparks 8 in, 3 in streamers #20 extended wire.
CONFIG: 14 Sep, plus
Primary C : Inverted cone, 30 deg wedges, 120 deg included angle, 10 in inner dia x 48 in outer dia, 23 t of #6 Cu wire on pegboard/plywood form. Used pegboard holes as wire slots. Tunes at 16 turns.
HV Filter Chokes, September 1998
Close wound #20 Cu wire 12 in long on 1/2 in PVC pipe with 1/2 in EMT insert.
Boss Hawg Capacitor, Oct 98
Switched to 2.5 nF glass Hawg Capacitor with Primary Coil C, sparks 12 in. Added HV filter coils, PVC over EMT cores. No Hawg Cap deaths in over 20 runs, one more NST died.
PERFOMANCE: Controlled sparks 12 in, 6 in streamers from extended #20 wire.
CONFIG:
Primary C : Inverted cone, 30 deg, 10 in x 48 in, 23 t of #6 Cu wire on pegboard/plywood form.
Capacitor : Hawg Capacitor, gallon cider jug in 24 qt stainless steel pot. Electrolytes is saturated aqueous solution of NaCl, MgSO4 and Amonium Nitrate. 1/2 in. Mineral oil layer. Measures 3.1 nF on RS dvm, probably closer to 2.5 nF
Power Panel, Oct 98
New Power Control Panel, with 240 V, 20 A service, 30 A disconnect, 4 switchable ballasts, variac w/ voltmeter.
Design of tungsten rotary gap, Oct 98
A.J. Smith helping with parts procurement and fabrication of rotor from 24 in disk of 1/2 in phenolic. Drive : 1/3 hp 90 VDC servomotor, powered by 120 VAC variac and bridge rectifier. Some problems with rotor trueing and balance.
Potential Transformer in Service, Oct 98
Control panel preliminary wiring and PT in service, for 120 VAC only using 7.5 A variac and two shorted NSTs for ballast. Nice Jacob's Ladder. Hooked up to previous TC circuit, controlled sparks 8 in.
PERFOMANCE: Controlled sparks 8 in, streamers 3 in from #20 extended wire.
CONFIG:
Secondary : 880 t of #20 Cu wire on PVC pipe 4.5 in dia x 30 in long.
Top Load : Toroid, 8 x 24 in stovepipe
Primary C : Inverted cone, 30 deg, 10 in x 48 in, 23 t of #6 Cu wire on pegboard/plywood form.
Power : 70:1 PT, ballast = 2||NSTs 450 VA ea, 120 VAC in.
Capacitor : Hawg Capacitor, gallon cider jug in 24 qt stainless steel pot.
Spark Gap : 13 segment Flexigap.
New Static Gap, Bigger Hawg Capacitor, 15 Oct 98
New aspirated spark gap, 1/2 in Cu pipe on PVC manifold, shop vac.
New Hawg Capacitor made from 3 Glass Gallon Cider Juggs, ? 5 nF.
Inner electrolyte, 1/2 can Red Devil Lye, electrodes 1/2 in Cu pipe.
Outer electrolyte, saturated brine, electrode/continment vessel #3 galvanized washtub.
PERFOMANCE: Controlled sparks 12 in, streamers 6 in from #20 extended wire.
Glass Hawg Capacitor Improvements, 18 Oct 98
New Hawg Capacitor made from 7 Glass Gallon Cider Juggs, 17 nF.
Inner electrolyte, 1/2 can Red Devil Lye. Outer electrolyte, saturated brine.
Disassembled, transported to Southern Tech for demo. Good demo until HV filter choke flameout. PERFOMANCE: Controlled sparks 20 in, streamers 12 in from #20 extended wire.
CONFIG:
Secondary : 880 t of #20 Cu wire on PVC pipe 4.5 in dia x 30 in long.
Top Load : Toroid, 8 x 24 in stovepipe
Primary C : Inverted cone, 30 deg, 10 in x 48 in, 23 t of #6 Cu, wood form, tunes at 6 turns.
Power : 70:1 PT, ballast = 2||NSTs 450 VA ea, 120 VAC in.
Capacitor : Hawg Capacitor, 7 gal cider juggs in #3 washtub.
Spark Gap : Aspirated, hollow electrode static gap.
Increase Input Power, Elevate Coils, 19 Oct 98
Converted Power panel to 240v, ballast of 1 to 4 parallel 1500 W, 240 V HW elements. Elevated primary and secondary coils on 24 in PVC legs. PERFOMANCE: Strikes 24 in, streamers 12 in from #20 extended wire.
Adjust Coupling, 22 Oct 98
Adjusted elevation of secondary coil by substituting different lengths of 4 in PVC pedestal. PERFOMANCE: Strikes 24 in, streamers 20 in from #20 extended wire.
Added Strike Rail, Elevated Toroid, Improve Capacitor Efficiency, 24 Oct 98
Adjusted elevation of toroid 6 in above secondary. Added NaOH to inner and outer electrolyte, 2 cans to outer, 1 can dissolves and divided among inners. Initial arcing on glass untill rinsed and dried above electrolyte. Much improved streamers. Reducing gap setting from max breakdown, improves streamer length, gives smoother, higher break rate. Streamers now prefer horizontal, can breakout toroid alone, no extended top wire needed.
PERFOMANCE: Strikes and streamers 30 in from toroid.
CONFIG:
Secondary : 880 t of #20 Cu wire on PVC pipe 4.5 in dia x 30 in long.
Top Load : Toroid, 8 x 24 in stovepipe
Primary C : Inverted cone, 30 deg, 10 in to 48 in dia, 23 t of #6 Cu, wood form, tunes at 6 turns.
Power : 70:1 PT, ballast panel 240 vac 1.5 to 6 kW short circuit limit
Capacitor : Hawg Capacitor, 7 gal cider juggs in #3 washtub.
Spark Gap : Aspirated, hollow electrode static gap.
Further experiments with electrolytic efficiency, Nov 98
Based on conductivity of ionic solutions, the Hydronium ion is by far the most mobile, followed by Hydroxide. Using NaOH (Lye) in the outer electrolyte of the Boss Hawg Capacitor gave notably longer and brighter arcs than NaCl (common table salt). Using HCl (Hydrochloric Acid) for the inner further increased the apparent capacitor efficiency. Within two days, before controlled measurements of resistance and efficiency could be arranged, the inner electrolyte turned blue presumably due to copper leaching out of the inner electrode tree. The outer tub began to leak from corrosion due to the more active electrolyte. The capacitor was disassembled, the electrolytes were chemically neutralized and then discarded. The leaky tub was retired, the phenolic insulating platform was also damaged.Upgraded Power Ballast panel to 1 to 8 parallel 1500 W, 240 V HW elements. Approximately 50A short circuit limit.
Calorimetry to measure input power
Digital meters were unable to measure the power, due to extreme waveform. Input current was estimated by ballast resistance and circuit breaker tripping. Calorimetry offers a great improvement in being able to accurately measure the ballast current.
A Calorimeter reads temperature rise in a known thermal mass. A sealed liquid-glass candy thermometer and stopwatch were used to instrument the ballast bank. Themal mass was approximated to be 19 litres of water plus 1.5 kg of iron. Temperature was measured with a gentle stir immediately before and after a timed run.
Power is the rate of Energy. Energy is Temperature rise times thermal mass. Power is Energy divided by time:
P = ( T2 - T1 ) / time
Real Power is Current squared times Real resistance. Current is square root of Power divided by resistance:
I = SQRT ( P / R )
This gives the current in a resistor in an AC series circuit, and therefore the current in the whole circuit. It is in phase with the voltage drop across the ballast, but not necessarily in phase with the mains voltage. The phase of the voltage across reactive elements, such as the mains transformer, will generally be complex.
Primary D
Inverted cone, 30 deg, 10 in to 48 in dia, 18 T of 3/4 in copper tubing, Nov 98Bigger toroid, 12 in stovepipe
Four elbows of 12 in stovepipe. Development put aside because it makes the coil too big to fit in the lab!
Tesla Coil Mk III
Crossroads, Georgia, basement of house at Oak Hills Drive, 2000 - 2002 Secondary Coil B, Feb 2000
880 turns of #20 copper formvar inslulated closewound on a 8 in ID white PVC form.PERFOMANCE: Strikes and streamers 54 in, controlled strikes 14 in from toroid.
CONFIG:
Secondary : 880 t of #20 Cu wire on PVC pipe 8.5 in dia x 30 in long.
Top Load : Toroid, 8 x 24 in stovepipe
Primary D : Inverted cone, 30 deg, 10 in to 48 in dia, 18 t of 3/4" Cu, wood form, tunes at 6.5 turns.
Power : 70:1 PT, ballast panel 240 vac 1.5 to 6 kW short circuit limit
Capacitor : Hawg Capacitor, 7 gal cider juggs in #3 washtub.
Spark Gap : Aspirated, hollow electrode static gap.
Surplus Capacitor A, Feb 2000
GE capacitor, 50 kV, 0.022 uF performance comparable to 7 gallon Boss Hawg Cap D.PERFOMANCE: Strikes and streamers 54 in, controlled strikes 14 in from toroid.
CONFIG:
Secondary : 880 t of #20 Cu wire on PVC pipe 4.5 in dia x 30 in long.
Top Load : Toroid, 8 x 24 in stovepipe
Primary D : Inverted cone, 30 deg, 10 in to 48 in dia, 18 t of 3/4" Cu, wood form, tunes at 6.5 turns.
Power : 70:1 PT, ballast panel 240 vac 1.5 to 6 kW short circuit limit
Capacitor : Hawg Capacitor, 7 gal cider juggs in #3 washtub.
Spark Gap : Aspirated, hollow electrode static gap.
Also: Tungsten rotary gap rotor unuseable, warped and unbalanced.
Tesla Coil Mk IV
Crossroads, Georgia, basement of house at Oak Hills Drive, 2002 - 2003 Secondary Coil D, Feb 2002
1040 turns of #17 copper formvar inslulated closewound on a 12.5 in OD green fiberglass form.Primary E, Unwound 2 inner turns of Pri D:
Inverted cone, 30 deg, 16 in to 48 in dia, 16 T OF 3/4 in copper tubing, Nov 98PERFOMANCE: Strikes and streamers 50 in, controlled strikes 14 in from toroid.
CONFIG:
Secondary : 1040 t of #17 Cu wire on fiberglass tube 12.5 in dia x 46 in long.
Top Load : Toroid, 8 x 24 in stovepipe
Primary D : Inverted cone, 30 deg, 16 in to 48 in dia, 16 t of 3/4 in Cu, wood form, tunes at 6 turns.
Power : 70:1 PT, ballast panel 240 VAC with up to 8 ea 1500W heater elements.
Capacitor : Hawg Capacitor, 7 gal cider juggs in #3 washtub.
Spark Gap : Aspirated, hollow electrode static gap.
