833-A VACUUM TUBE TESLA COIL
Revised 09 / 17 / 2000
Diagram of Circuit Parts List Circuit Explanation Photos Helpful Links
This is the schematic of my current 833-A Tesla Coil. Heavy wires
are short and carry high current. (Leads
connect if they cross, but not if they are jumped like the filament
leads) Components and sources are given below.
L1.....................................Two stacked flat-spiral coils in series, each has 13 turns of #16 zip cord (both sides of the cord) wound on
4.5" inside diameter. Total inductance is ~120 microhenry (mutual inductance accounted for).
L2.....................................One flat-spiral coil of 20 turns #16 zip cord (both wires) wound on 4.5" I. D. and positioned below L1
L3.....................................One flat-spiral coil of 25 turns #16 zip cord (both wires) wound on 4.5" I. D. and positioned above L1
L4.....................................12" of #27 enameled Cu wire, wound on treated 3" Sch. 40 PVC pipe in the counterclockwise direction and
coated with 3 coats of clear PU varnish
C1.....................................MMC capacitor, two rows in parallel of eight capacitors in series. Caps are part no. P10506-ND from Digikey
and the total C1 has a 0.0021 microfarad capacitance.
C2.....................................Tuning variable capacitor, 100 pF 4 kV air unit immersed in mineral oil for a total capacitance of ~230 pF.
Get these at Surplus Sales of Nebraska or RF Parts Inc.
C3.....................................0.001 microfarad, 10 kVDC (Hosfelt or All Electronics)
C4.....................................Two microwave oven caps in series; any should work. If you can't find enough broken ovens, go to AMI Parts.
C5.....................................Two parallel sets of two microwave oven caps in series (total capacitance ~ 1 microfarad)
C6.....................................same as C4
C7.....................................0.05 microfarad cap, 200 V or up. Anything works.
C8.....................................3" by 8" dryer-duct toroid
R1.....................................5000 Ohm, 100 Watt {continuous oscillator} OR 40 kohm 100 Watt {"sputter" mode}
R2.....................................50 Ohm 100 Watt wirewound protection resistor from Digikey
D1.....................................20 6A, 1000 PIV silicon diodes in a chain, from Hosfelt or All Electronics. Bridge all diodes with a 10 Mohm
resistor and a 0.001 microfarad capacitor.
Q1....................................."Staccato" thyristor, I use a 25-A, 600 VRRM SCR from Digikey (part no. S6025L-ND)
T1......................................Microwave oven transformer (get a big one) with the shunts removed and the primary tapped at ~ 4 / 5 the winding.
T2......................................10 V, 10 A filament transformer with a few hundred V isolationnote: a 20 A variac is highly recommended for turning up the power.
The dots by the coils denote the top turn if it is a helical coil and the outer turn if it is a flat coil. This Tesla coil requires a breakout point on the toroid or you will blow stuff (like your tube, your SCR, or your diodes) from kickback. It will make up to 22 inches of spark.
How this coil works: as you can see from the schematic, this is your standard tuned-plate oscillator with a Class-C bias. When the circuit is first turned on, the grid at ground potential and the tube conducts. This causes an oscillation in the plate tank circuit, L1-C1-C2, and a voltage is induced across L2 that is 180 degrees out of phase with that across L1. L2's voltage makes the grid negative, and the tube cuts off. With no plate current flowing, the
tank circuit continues its oscillation until the voltage induced across L2 goes positive and turns the tube back on, replenishing the tank circuit's energy. As this action continues and a strong sustained oscillation is built up in the tank circuit, the grid current that flows whenever the grid swings positive is charging up C3 so that negative charge begins accumulating on the grid side. Leak resistor R1 determines how fast this charge drains, and ultimately what the steady grid bias potential will be when the oscillator has reached stable operation.If the leak RC constant is very large, a few positive cycles of L2 will allow a large enough negative charge to build up on the grid side of C3 that the tube is cut off until the charge can drain. Then the oscillations commence again for a while, the tube is off for a while, etc. This pulsed operation is known as "sputter" mode and in this oscillator, it occurs when the values shown by "sputter mode" above are used. It is not a healthy choice for your components in the long run, since the oscillator's turning on and off at random times on the plate voltage input cycle can stress the tube and kickback can bake all manner of parts- but it is a way of vastly improving the efficiency of the system at a small loss of spark length.
The recommended mode of operation is "staccato", which like "sputter", decreases the duty cycle of the tube with minimal loss in spark length. However, "staccato" is designed to operate the oscillator by grounding the cathode in sync with the positive cycle of the power supply, rather than at more or less random times in the cycle. Dave Sharpe and John Freau on Chip Atkinson's Tesla List have pioneered this method. It is friendly to the tube and the rest of your circuit. The "staccato" driver is not shown in the above diagram, but I use Dave Sharpe's design.
Plate-circuit impedance matching is a difficult concept to theoretically address with Tesla coil loads. Brush discharge, sparks, plasma globes, etc. all contribute a different load to the machine that change frequently with time. The big idea is to design for the longest sparks, and that is much easier to do experimentally than theoretically- just change the tank L / C ratio and see what happens. I have found for all tube coils that I have built that L / C should be on the order of 50,000 - 100,000. Plate voltage will also have an effect in impedance considerations.
NEW!!!! Recent photos of the 833-A coil in Staccato mode.
Operating at 15 pulses per second in Staccato mode, a thick gnarly arc jumps about 21" to my hand. (Note that you need to be extremely careful in allowing discharges to yourself.)
A similar shot at night.
Here are some photos of previous versions of this coil in sputter mode.
This photo shows the coil at reduced power with a flask of 1 atm. neon nearby. The neon forms a long thin snake of brilliant discharge much more readily than air, making a beautiful effect. The flasks of various gases are made by the EGL company and distributed for the manufacture of neon signs. Helium has a similar effect, but with a whitish-pink arc. Argon forms a spray of steely blue, crackly streamers. Xenon and krypton undoubtedly make some cool effects too, but...I have a college bill to pay!!
This photo shows a long spark obtained with the earlier version of the 833 coil, approximately 19-20" long. It's a bad photo, but I should have some better ones of the new "staccato" version shortly. When photographing sparks, use low light, dark background (if possible) and wide aperture. For pulsed tube coil operation, you want exposure times longer than 1/30 second in order to not miss the spark. Sorry I can't be more specific about these things- but it's because I haven't perfected the technique. I'd like to thank my mom for helping take the photos- she has a good sense for photography and has taught me a few things!
Click to download the RCA 833-A specs (zipped jpegs)
Get an 833-A or 833-C at RF Parts Inc.
