thebugger

modified 9 years ago

Small Solid State Tesla Coil

315

03:37:11

Okay, I'm really all in for old methods, way before amplifying devices. The original tesla coil design didn't use any amplifying device, but utilised the harmonics from a spark gap, to excite high frequency oscillations in the primary of a HV transformer, and produce a much larger spark from the torus of the secondary. Now there are sooo many ways to do this but let's stick to the most basic. Typically tesla coils produce voltages from 50kV to several MV, at frequencies from 50kHz to 1MHz. I won't go into how it works, but I'll give you tips for the coils of the transformer. The primary should be wound around the secondary, on the base of the transformer. The secondary should be wound around a cone from the center of the base upwards. The primary should be wound over a 30cm diameter with a height of 55mm with 10turns of 1mm wire. The secondary should have an inductance of 470mH and can be wound over an enormous set of cores and parameters, but i recommend an air core (a plastic cone) with a diameter of 10cm, height of 12cm and 2800 Turns of wire around it. It should be placed at the center of the outer coil (the primary) with one end connected to ground and the other to the torus. You may experiment with the secondary's dimensions and turns, it's highly tolerant. Experimenting with the coil can be beneficial, because if the primary is made from 5 turns it doubles the voltage at the secondary (to up to 150kV)

published 9 years ago

WTFCircuit

9 years ago

If I can say, this is more similar to a flyback driver. Of course there are some tesla coil designs with a 555 timer. Anyway the tesla coil is a big LC oscillator with his specific frequency and you want the oscillator to work as near as possible to that frequency, and to do that is generally used some sort of feedbacks. Also in a spark gap tesla coil the elettromagnetic radiations emitted from the spark aren't enough to exite the secondary coil that much. This is why a cap and an adjustable primary are used, them are to make another LC oscillator which oscillate when the spark is present and tuned to almost the same frequency of the secondary. On the solid state ones PLL ICs are used to lock the frequency into the one in which the coil resonates. I already made a schematic for a sstc which seems to be working fine even if over simplified

thebugger

9 years ago

Actually only the old design tesla coils had to be tuned for them to work. See in the old days, even before vacuum tubes, the only two ways of producing high frequency oscillations were either mechanical (as the Alexanderson alternator) and by spark gaps and then filtration. The tesla coils need a way to transform low frequency oscillations into high frequency oscillations. The spark itself has most energy at the fundamental, but the fundamental was only 50/60Hz, so you had to ,,extract'' the high frequency component from the spark. The only way to do that was to tune the primary to a specific frequency, so that it gets excited by the spark and starts building a charge at the high frequency. This was a very archaic way of transforming low frequency oscillations to high frequency ones. Nowadays a tesla coil resembles more a flyback transformer, because with solid states, you can directly supply it the high frequency oscillations without needing to transform it from mains. You can alternatively make sinewave oscillations, by tuning the two sides, but it's not always needed.

WTFCircuit

9 years ago

Well flyback transformers and tesla coils have quite different from the working mechanisms. For example a flyback can work in flyback mode or resonant mode and in a wide range of frequencies. On the other hand a tc must work only in resonance state and at is resonant frequency because it is a big LC resonator. Moving to the sstc and the most powerful ones, the drsstc (double resonant tesla coil) and the qcw drsstc (quasi continous wave drsstc) has to be tuned properly. Anyway check this if you didn't find it yet :-) : http://everycircuit.com/circuit/5134306383495168

thebugger

9 years ago

Yeah i guess a real tesla coil works with harmonic oscillations such as sinewaves, but a similar effect can be achieved without such painstakingly tweaky tunings of resonances, but just as simple as a pulse driven flyback. I think it'll work very similarly

WTFCircuit

9 years ago

I'm not saying that in the primary you have to use a sinewave, as you say it can be any wave. But it has to be in resonance with the secondary so some waves, as the square waves that has some ringings aren't very good for a tesla coil. Anyway have you checked my design?

thebugger

9 years ago

Yes I've checked it. That's one super powerful mega, tesla coil. But i think it's drawing insane amounts of current from the mains, and it's spiking, which may introduce noise back to the mains. I suggest more EMI filtering. Now to the subject, i thonk the primary should also be tuned, as the secondary.

WTFCircuit

9 years ago

And this is a very small driver :-p also I think it won't that powerfull in real life. The primary, to be tuned with the secondary needs a tank capacitor, called mmc bank, and it is used in high power tesla coils

VivekChoubey

9 years ago

I don't think this will generate a large shockwave. I guess

thebugger

9 years ago

WTF it probably will build up to a few hundred kV but therw are limiting factors. The active resistance of the coils, the parasitic capacitance of the secondary (which will be large) and such

thebugger

9 years ago

I even think that the parasitic capacitance on the secondary is enough to tune it to a given frequency. You just have to tune the primary too.

WTFCircuit

9 years ago

The parasitic capacitance is very important becauseit sets the maximum frequency that the coil can get, in general it is between 1-5pF and has to be considered even with a top load. The primary has to be tuned only if you use a tank capacitor with it

thebugger

9 years ago

Yes, exactly, but i suggest double tuning, both primary and secondary. It'll be harder, but if you can get it to work the mutual resonance will increase the output much more. Maybe even at what EC shows. By the way, how do you plan to rectify the mains and filter it sufficiently to withstand such spikes? The largest one i saw was 20A and that's right on the limit on most household installations. Many can withstand up to 40A, but still that's a hell of a lot of power. I mean peak consumption is 4.4kW

WTFCircuit

9 years ago

To recrify the main most people passas it through the primary of a big transformer with the sencondary shorted out. Then there is the diode bridge rectifier and as more capacitance as possible. An isolation transformer can be used but, obviously it has to be very powerfull. This is what I see people to do

thebugger

9 years ago

I'm gonna try to simulate the power supply and I'll let you know the results, but More like rewinding a semi large transformer with thick wire, so that you'd use the maximum space available. If the primary is rated at 20A, the transformer is more likely to be an AC welder transformer. Can you imagine the size of that thing.

thebugger

9 years ago

Actually i got some results, but you may not like them. First the bridge rectifier and transformer must be quite powerful, at least 20-30A. Second the rectifier must be a Pi configuration with...wait for it... 2x 15mF 400V capacitors and a 16mH coil with 1ohm dampening resistor. The coil is easy, and 1ohm is probably given by the coil itself, such capacitors exist but you won't like the price. Just the power supply will cost you around 1000$ :/ Maybe if you turn to IGBT switching you may find it cheeper, but it still will cost you above 100$ i don't know.

thebugger

9 years ago

Let me precisely calculate the current draw of the tesla coil, it may improve the results.

thebugger

9 years ago

Nope it's still 20A, thia is your power supply, sorry. Using the primary of a large transformer is a lot of inductance, and with that a lot of active resistance. At 10H you'll get around 150ohm resistance from the coil itself which will dampen the voltage too much. At such current draws, you'll need a lot of capacitance and low inductance.

WTFCircuit

9 years ago

There is another think I forgot to mencion: such high power tesla coils are never used in continous wave mode, but in interruped mode. It means that the oscillator gets turned on and of several times a second. This makes it drain less average current, for exaple I have a little tesla coil (with a different driver) which drains 1.5A at 33V in cw mode and only 0.3A in interrupted mode. But the pick current stays the same. I think this should change things a little bit.

thebugger

9 years ago

I know I've averaged it by observation. It has negative peaks too. When you add a diode and filter them, the spikes increase to 40A. If you average them, you still get around 20A

add comment in editor