In 1997 I became interested in Tesla coils and decided to build my own. Unfortunately, I lost interest before I could get my first coil to work. Few years later, I dug up my old parts, redid my calculations and continued building. Again, I lost interest before finishing the coil.
In 2007 a friend showed me a coil he had made, reminding me of my unfinished project. Again I dug up my coil parts, recalculated everything, and actually finished the project this time.
A Tesla coil is a resonant transformer. It is basically to LC circuits, tuned to the same resonant frequency. The secondary coil and the topload form one circuit; the coil provides the inductance and the topload most of the capacitance. The second circuit is formed by a high voltage capacitor and a much smaller primary coil.
First, a high voltage transformer is used to charge the capacitor.
Once the capacitor attains a sufficient charge, the spark gap fires. The transformer is shorted out and the primary circuit begins oscillating.
Since the tank capacitor size is fixed, the circuit is tuned by adjusting the impedance of the primary coil by changing the tap point. When properly tuned, a very high voltage will develop at the top of the secondary coil, leading to spectacular air discharges. Unlike in a traditional transformer, the ratio of turns between the primary and secondary coil has practically no influence on the voltage.
Here are links to some sites with information about TC construction and theory of operation:
Designing and building a simple Tesla coil is fairly easy. To a beginner, it seems like a daunting task (I know it seemed like that to me!), but you can get a working coil by following a few general instructions and performing just one calculation. Of course, if you want a coil that performs really well, there is no way around learning a bit more theory and doing more calculations.
Gary Lau has a nice short design guide and DeepFriedNeon is another good place to look. Here are the basic steps to get you started.
Before actually setting out to build a Tesla coil according to the instructions above, I strongly recommend following the links to the other TC sites and reading a bit more. First stop should be the safety sheet at pupman.com.
Also note that I've made no mention of transformer protection circuits above. I haven't used any myself, and have had no problems yet. The keyword here is yet.
The design of my coil was influenced mainly by the parts I had available. Those were:
From a hardware store, I found a ⌀75mm drain pipe and 5 meters of ⌀6mm copper tube.
The secondary coil was the first component I put together. I wound roughly 900 turns of magnet wire around the drain pipe. A height to width ratio of around 1:5 is recommended for small coils, so I made mine 37 centimeters high. The length of wire used was roughly 209 meters.
The inductance and self capacitance of the secondary coil and the capacitance of the topload can be calculated with formulas found on other TC sites. With these, I can calculate the resonant frequency of the secondary circuit:
f = [2π√(L(Ccoil+Ctop))]-1
Using a ⌀14cm sphere as a topload, the resonant frequency of my coil is roughly 452kHz.
My first try was a plastic sphere coated with aluminum foil. I couldn't get the foil smoothed out well enough, so I abandoned the sphere and decided to go with a toroid.
I constructed a small toroid by wrapping a length of corrugated rubber pipe into a donut shape and covering it with aluminum tape. I put a plywood disc in the center for support. I couldn't get the toroid very smooth either, but it works better than the sphere because of its shape and on account of being bigger. The rough surface results in lots of short streamers, rather than one or two long ones.
The primary coil is made of ⌀6mm copper tube, wound spirally around the secondary. The inner diameter is 17cm and the outer 29cm. There are 6 turns total, with a 3mm space between each turn.
Because of the large space between my primary and secondary, the coil might be a little too loosely coupled. The rising conical shape might alleviate that somewhat, but since I don't have the equipment to measure the coupling coefficient, I can't know for sure.
The primary coil forms an LC oscillator together with the capacitor. The desired inductance can be calculated with the following formula:
L = [(2πf)2C]-1
C is the capacitance of the tank capacitor and f is the resonant frequency of the secondary coil.
The javascript flat spiral coil calculator at DeepFriendNeon is useful when designing, but the simple equations used will give only a rough approximation. The proper tuning has to be discovered experimentally, so it's a good idea to make the primary coil big enough rather than find out its too small during testing. My small coil has seven turns, tapped at the fourth, but around 15 turns seems to be a rather common size.
Since I had a large number of small capacitors available, I decided to build a Multi Mini Capacitor. The resonant capacitor value for the power supply can be calculated with the following formula:
C = I ⁄ (2πfU)
The resonant capacitor value for my transformer is 27.8nF. The actual value used should be a little bigger or smaller than this, as rapid voltage rise due to resonant charging may break the transformer and/or the capacitors. A safety gap will afford some protection as well. See Richie Burnett's page for more in-depth information about resonant charging.
I made a javascript MMC design tool to help pick the cheapest set of capacitors.
My MMC is made up of three parallel strings with 24 caps each. The voltage rating of each string is 6600V and the equivalent capacitance of the whole MMC is 41.3nF
Each capacitor has a 10MΩ resistor across its terminals. This is important, as the individual capacitors can retain a charge for a very long time after the power has been shut off.
As evident from the picture below, the voltage rating is a bit too low, even for a 4kV transformer. To be safe, it should be at least 8 or 12kV.
My spark gap is simply two machine screws with a pinball ball in the middle.
The spacing is adjusted so that the gap will just fire when it is the only thing connected to the transformer. Setting it wider increases the firing voltage and can theoretically improve spark length, but risks destroying the transformer. NSTs especially can break easily this way.
For a larger coil I would build an air cooled cylindrical TCBOR/RQ style gap.
Tank circuit | |
---|---|
Transformer | NST 4kV 35mA, 240V 50Hz input |
Capacitor | 3×24 275VAC 0.33µF |
Spark gap | Static gap with two gaps in series |
Primary coil | |
Inner diameter | 17cm |
Tube diameter | 6mm |
Turn spacing | 3mm |
Tube length | 5m |
Turns | 6 |
Secondary coil | |
Diameter | 7.5cm |
Height | 37cm |
Wire gauge | 28AWG (0.3mm) |
Wire length | ∼209m |
Turns | ∼900 |
Short streamers using an energy drink can as a top load. The top load is a bit too small, as there is some corona discharge at the top windings.
Sparks to plastic rod with a grounding wire at the end.
Streamers emanating from a sharp screw placed on the toroid.
A roughly 20cm long spark to a grounding rod. Not visible in the picture is a purple haze between the rod and the toroid which was quite vivid to the naked eye.