How to optimize spark gap Tesla coils

[Alex Karatzky]

My elementary school had its first annual science fair a few weeks ago, so me and my friend signed up and we decided to do it on wireless electricity. So naturally, we built a tesla coil. It took a couple of tries, but we eventually got it to barely illuminate a portion of a fluorescent light bulb. We somehow managed to make it to the regional competition, but we need to make it at least power the full lightbulb.

The lightbulb only blinks when the spark gap flashes, which is every 1-2 seconds.
We used an HV transformer from a bug zapper which had 2 HV input and 2 HV output wires. The 2 input wires were attached to a 9V battery and the output wires were attached to each end of a capacitor bank comprised of 2 150k capacitors.

We used 18 AWG speaker wire to connect the capacitors to the spark gap and primary coil.
The secondary coil was made of 24 AWG magnetized copper wire wrapped around a 20cm long piece of 1 1/4' PVC pipe, while the primary coil was made from 10 AWG wire wrapped ~6 times around the secondary.

Any suggestions on how we can improve the coil? The regional competition is on March 28th.
Picture ---> https://postimg.cc/sMHVzjG8

Circuit --->https://postimg.cc/mPZMXLfn

 

[jambaugh]

The critical feature of a tesla coil is resonance. Tesla coils are resonant transformers. This means you must drive the circuit at a given frequency and the secondary circuit resonates at or near that frequency.

Firstly the design should take into account getting the frequencies close. The primary having few windings will have fairly low self inductance which makes its resonant frequency quite high (inductance acts like the mass of a mass spring analogy to the resonant electrical circuit which lowers the frequency). Higher capacitance will likewise lower the frequency (1/C is analogous to spring stiffness constant). It may be useful to put a high voltage variable capacitor across the primary winding (with a very well insulated knob) for adjusting input frequencies. If not that then at least some additional capacitance will help lower the frequency to something easier to match with the secondary. You're shooting for something in the 100 kilohertz to 1 Megahertz range I believe.

You'll want to figure out, at least approximately, the resonant frequency of the primary. A good oscilloscope is recommended here since a frequency counter may be counting noise instead of sinusoidal signal. Be careful about isolating the high voltage from the scope of course.

You'll then want to match the secondary. Getting close is a matter of doing some bar napkin calculations on the geometry. You should find several sources to calculate inductance of air core solenoids. And the capacitance of a sphere or near sphere is available here http://www.electrostatics.org/images/ESA_2008_D1.pdf what you'd be looking for is at the bottom in the conclusion.

This should get you to the right order of magnitude which will greatly improve performance, then small adjustment comparing results should get you closer.The spark gap acts as a noisy random signal generator and the resonant circuit will filter out and induce onto the secondary the component in resonance. There's also a feedback effect to the spark gap which tends to make it repeat in phase with the resonant frequency. More sophisticated versions actually use an amplified sinusoidal signal generator to drive the primary.

When I read up on Tesla's experiments back in college I recall one technique to make the spark driver higher frequencies was to have an air jet blowing across the gap so as to shorten the pulse by blowing the plasma away so it can only conduct for an instant. That may be helpful in your project though I have no details on the strength of the air jet or efficacy.

That's all I can tell you off the cuff from my knowledge. Even with a good tuned Tesla coil, to get the florescent bulb to glow more than dimly you'll need some kind of receiver circuit. That may be as simple as a stiff piece of wire with one end soldered to the bulb terminal and wound into a coil with the other end projecting out like an antenna. Same on each end should increase the effective length, i.e. lower the resonant frequency of the bulb. Keep in mind a half meter conductor as a half wavelength dipole will resonate at 1m wave-length = 3x10^8 Hertz or 300MHz. That's getting up to UHF an way beyond what the tesla coil can achieve in my understanding. Another tip is to just increase the capacitance of the ends of the bulb. Rap a few inches of the end in foil attached to the end conductor.

If you want to do both, have a wire coil with one end in contact with the foil cap and the other soldered to the bulb's terminal. make sure the foil is, in this case, Not connected directly to the bulb terminal. That's probably the best setup I can think of. You can play around with number of windings of that wire to seek an optimum.