So, I wanted to build a Tesla Coil


This post is two parts in one: The Build, and Deep Dive. The build walks through my own experience building the coil, and deep dive goes into nitty, gritty details on how it works.

This is a very dangerous device. A lapse in focus, or judgement while working on, or maintaining it could result in serious injury or death.

The Build

I've been fascinated with electricity for as long as I can remember. The first class project I ever had to do was in first grade and I chose to write a report on electricity. My Dad had made his own high voltage transformer that I'd constantly sneak off and play with. It was always fun watching the arcs jump across different surfaces. In high school we put together a Jacob's Ladder as part of the haunted house we'd build every Halloween. For the rest of the year, I'd stick things between the wires and try controlling the arc. The one thing I hadn't yet been able to play with was a Tesla Coil. Now I can.

First thing was to learn how and why it worked. What was the deal with the coils and that crazy thing on top? I found a couple YouTube videos and blog posts that explained it using jargon I didn't understand. Got frustrated. Kept looking. Then, I finally found a video that stepped through the process in a way that I understood. I'd had a vague idea of how transformers and capacitors worked, but this required a much deeper knowledge of both. I also learned that a spark gap is effectively a switch that's activated by sufficiently high voltage. And I learned what a tank circuit is. See the bottom of this post for more details.

Being the person I am, I didn't want to start small. It had to be significant without being ridiculous. I scoured the Internet for plans or suggestions and came across an article on Make Magazine's site. Getting capacitors rated high enough for something like this isn't easy (at least not for me). One of the most appealing aspects of the article is that it teaches you how to make your own capacitors with salt water, aluminum tape and beer bottles. It also walks you through the build process step by step.

I built everything out of order starting with the simplest or most accessible bits first. In this case, the capacitors were first. All they require is a super saturated saline solution, aluminum tape, copper wire, beer bottles, and mineral oil. I filled a pot with some water, heated it to just under boiling, then dumped in as much salt as would dissolve. I used root beer bottles - my favorite being Red Arrow, hence the name of the coil. The bottles were individually wrapped with aluminum tape. The saline solution was added. Then I checked the capacitance with my multimeter. Given the value of each bottle I taped together enough bottles to get the overall value I needed. The oil was poured in last to insulate the saline solution. Finally, I soldered up some heavy gauge wire and placed it into the bottles. Capacitors of that size can hold a lot of energy and release it very quickly. They are extremely dangerous. To mitigate the danger, I store them with a wire attached to each pole (the wire coming out and the aluminum tape) ensuring they don't spontaneously charge.

Next was the Terry filter. This filter wasn't originally part of the design of the coil, but present-day coilers have learned a lot since Nikola Tesla's initial work. One of those lessons being how to protect the initial transformer - usually an old school neon sign transformer (or NST) - from current flowing back into it from the tank circuit. This back flow has a tendency to destroy the NST over time. The filter basically works by dumping any charge large enough to harm the transformer directly to ground. The most challenging part of building the filter was drilling all the holes in the lexan. I don't have a drill press so I had to drill every one by hand. Doing so got very tedious because you have to go slowly so as not to crack the lexan. After the holes were drilled, soldering everything went pretty quickly. Crimping and routing the bits of this wire took a little trial and error. Unfortunately, I had already cut the lexan to size before realizing I wouldn't have any way of mounting it without the wire getting in the way. My solution was to chisel out some slots for the wire to fit. The finished product is a pretty cool looking device. The giant resistors, brass screw terminals, large repetitive components, and super heavy gauge wire give it an industrial feel.

Once the filter is built, it needs to be tuned. The spark gaps need to be set so they'll fire if the voltage goes just over the NST's working voltage. Setting the gaps means setting them close enough together that the NST will fire them on its own, then gradually backing them away until they no longer fire. The process is a bit nerve wracking, because you have to fire the sparks, unplug the NST, short the capacitors, then loosen and adjust the points of the spark gap. A misstep here and you're in for a world of hurt. On the other hand, you get to make a lot of small, loud sparks.

Four major components remained: the main spark gap, the top load, and the primary and secondary coils. The top load and spark gaps were trivially simple. The spark gap is simply six pieces of copper pipe with holes drilled into the top, glued to a piece of MDF with an even gap between them. The top load is a small, round board glued to a foam flower ring, then covered over with aluminum tape, and smoothed out.

Making the primary coil meant cutting radial dados in a large, round board, cutting a bunch of slots in small MDF boards, and winding heavy gauge wire around it through the slots. I didn't have a table saw, so I put off cutting the slots in the MDF. There are a lot of them and I wasn't sure how to do it without a table saw and jig. That left the secondary coil.

The secondary coil is the second most visible part of the whole project. It's the coil that goes from the base to the top load. Making it requires wrapping very small gauge magnet wire around a PVC pipe a couple thousand turns. The wire has to butt against itself the entire length of the coil. It's not a job to do by hand. The article discusses building a jig that uses a drill to turn the pipe while you hold the wire, and slowly allow it to wrap down the length of the pipe. Starting out, it's a difficult and tedious process. Squeezing the drill's trigger just enough to get the pipe turning at a manageable rate took a lot of patience and focus. Add to that having to hold the wire at just the right angle and tension so it stays taut without wrapping over itself, wrapping too far from itself, or breaking. At a manageable speed, it takes a little over twenty minutes to wrap the full length of pipe. Stopping and starting is tricky, the tension and angle of the wire need to be set just right. Once started, I was in it until either the wire broke or the wrap was complete. All in all, I broke the wire three times, and misaligned it so badly I had to start over four times. From deciding to get it done until it was actually done took about three hours, and a boatload of patience and perseverance. The final wrap isn't perfect. About halfway down, things were going really well so I kept pushing the speed a bit. I slipped and put about a four wire-width gap in the coil. I decided I could live with that and kept going. The final step is to protect it with three coats of varnish that need to dry overnight between each coat. Overall, I'm really happy with it.

At this point, only the primary coil is left. I decided that instead of figuring out how to solve the wire slot problem, I was going to take a break and work on something else for a bit. That something else ended up being the playhouse and the coil pieces sat on a shelf for months.

After the playhouse project started to wind down, I was left with a shop full of tools and precedent for making a lot of dust and noise. It was time to finish the coil.

I cut the dados in the round board with a circular saw and straight-edge. Then, I put together a simple jig with a scrap piece of plywood and a roofing nail to cut the wire slots. Being very careful about where I put my fingers, I cut the two hundred slots in the board and glued them into the dados. After drying for a couple of hours, I wrapped the heavy gauge wire and had my primary coil. All that was left was assembly and testing.

Things almost went wrong. While attaching the secondary coil to the primary coil, I broke the connection plate and a piece of the wire with it. There was very little wire to spare. I had to very carefully re-set the plate and solder the wire back to it. Then, I taped it all to reinforce the connection and attached it to the primary coil.

Electrical ground is important generally, but especially so with a Tesla Coil. The secondary coil is effectively an inductor connected to ground. The primary coil induces a charge, but is not directly connected. All safety features need to connect to ground. There’s a copper ring around the outside and above the primary coil, so sparks off the top load will strike there rather than the primary coil. The Terry filter dumps to ground, as well. Maybe the worst thing you could do is use your house’s ground for this. The coil operates at very high voltages and very high frequencies. If these get into your house's electrical system, it could wreak havoc on your appliances. On the other hand, grounding is crazy simple. It’s literally the ground. You take a three foot, steel dowel and pound it into the earth nice and deep. Then, clamp a very large gauge, stranded cable to it, and connect all the coil’s ground points to that cable. I had done this back when I was tuning the filter.

Finally, I double checked all the connections, ensuring everything was secure and connected in the proper place. Then, I made sure nothing was crossing where it shouldn’t, or touching what it shouldn’t. The moment of truth was approaching.

I unclipped the safety on the capacitors, plugged in the NST, stood back, and hit the switch. Lots of noise, but no spark. At least everything was working and nothing exploded. I got a little closer and hit the switch again. The noise was coming from the main spark gap. That’s good! It means everything’s connected properly. Obviously, I wasn’t getting enough voltage in the top load for it to arc off into the air, though.

The spark gap can be tuned. There are six pipes creating five gaps. The more gaps, the more voltage it takes to arc across them, thus more voltage through the primary coil. Just to start out, I used only the first gap. After the initial try, I unplugged from the wall, used my "capacitor killer" - a wire on a long stick to short the cap - then clipped the safety to the cap. With the coil safe I increased the number of active gaps to two and tried again. Still nothing coming out of the top. Rinse and repeat this process until I had all gaps active. Still nothing. This process took a while. I was extra careful when approaching the coil: checking and double checking that nothing was going to bite me.

Next thing to try was to give the current a point to arc off of. I cut a piece of heavy gauge solid core wire and tied it to the top load using the screw that holds it to the secondary coil. I tried again. Still nothing. Last resort: I connected another piece of the same wire to ground, ran it up along the coil, and pointed it at the piece on the top load. I gave them an initial gap of about an inch. Un-safety the caps, plug everything in, stand back, hit the button. A SPARK! A really nice, dancing spark from the coil to the ground wire. It worked. It actually worked. I let go of the button, pushed the ground wire back with my long stick, and tried again. The longest distance I could get was about three inches, but that was three inches of good solid, dancing sparks.

From here, I’ll try to tune the coil to get more out of it, but it’ll probably be some time before I get a chance to dig in and do that. I’ll definitely be posting more as it happens.

I’m still a little shocked (if you’ll pardon the pun) that I managed to build a Tesla Coil from scratch.

Deep Dive

From here, I’ll go into a deeper explanation of how the coil actually works. While researching the principles behind Tesla Coils, I kept wanting a single resource that would walk through all the components, and break down what they did, and how they did it. Most resources assume some amount of knowledge of electrical components, and the principles behind them. I’m hoping the following section will be a clear guide for anyone, regardless of current understanding.

Here we go.

A Tesla Coil is an air-core, dual-resonant transformer.

I didn’t really know what all that meant initially, either. So, I’m going to talk through this from the very beginning, starting with what happens when electrical current moves through a wire.

When we talk about electricity, we’re really discussing the electromagnetic force. That is, electricity and magnetism are interconnected. Energy can move between magnetic flux and electrical current similar to how energy can move between heat and mechanical force (e.g. in an engine). As electrical current flows through a wire - any conductor really - a magnetic field flows around the wire. The magnetic flow depends on the flow of current both in magnitude and direction. That is, if current is flowing in one direction, a magnetic field will also circle the wire in one direction. When the current is reversed, so is the magnetic field. The magnitude of the magnetic field is proportional the electrical current flowing through the wire.

Here's a quick animation showing this phenomenon:

That’s how electrical energy becomes magnetic energy, but the reverse is also possible. Moving - it must be moving - a magnetic field near a wire will induce an electric current in that wire. The faster the movement, the larger the current.

If you take a piece of wire and coil it up like a spring, you’ll amplify the magnetic field that is produced by running a current through it, and you’ll have a weak electromagnet. Put something ferrous in the middle of that coil - like a steel nail - and you’ll give the magnetic flux something it likes to flow through, and a stronger electromagnet. This is a fun science experiment, and can be applicable when trying to lift heavy ferrous things, but doesn’t help our cause much.

What does help is that you can make another coil of wire and put it next to the first one. As the amount of current increases (usually due to increasing the voltage) in the first wire, the magnetic field will move. That movement will induce a current in the second coil of wire. We’ve just transferred electrical energy through the air. Neat. Except, it only works when the amount of charge in the first wire is changing. As luck would have it, the electricity coming out of the wall socket does just that: it alternates in a sine wave at 60Hz. Meaning, every 1/60th of a second the voltage will go from 0 up to about 170 volts, back down to 0, then down to about negative 170 volts (thus reversing direction), then back up to 0. Hooking that up to a coil will cause the magnetic flux to grow, then shrink, then grow going the opposite direction, then shrink. It’ll move.

So, now we take our two coils and put them next to one another. Then, plug one coil into the outlet. If we measured the charge on the second coil, we’d see it move in the same sine wave. Well, it’d be shifted: when the first coil was positive, the second would be negative, but it’d still be moving. We could plug a light bulb into that and it’d glow.

We’ve just created a transformer. In fact, we created an air-core transformer: so named because the space between the coils is air. Usually, transformers have a ferrous ring that goes into one coil, around and through the other coil, and connects back to itself. That allows for a much more efficient transfer of energy - because the magnetic flux prefers traveling through ferrite rather than air.

This video goes into further detail (and includes some descriptive animations):

One interesting - and extremely useful - property of transformers is that the ratio of the number of wraps in each coil is directly proportional to the ratio of voltages on either side. So, if you were to wrap the first side ten times and the second side twenty times, then run a current at a specific voltage through the first side, you’d see twice that voltage on the second side. You’d step up the voltage. That works the opposite way as well. If you ran a current at some voltage through the second side, you’d see half that voltage on the first side. This is how electricity is transferred from the power station to your house. It’s stepped up at the power plant to make it efficient at traveling over large distances. Then, it’s stepped down before going into your house.

Now that we understand (hopefully I explained in a way that was clear) transformers, we can see how the NST steps up the house voltage from 120 volts to 12,000 volts. That’s the first step.

The capacitors are the next step. A capacitor is crazy simple. Take a conductor, place an insulator on it, then place another conductor on that. Done. When a charge is applied to one side, the opposite charge builds up on the other side. To take advantage of this effect, use it like a battery: place something across both conductors and the charge will flow (or discharge) through it. Unlike a battery, however, no chemical process needs to take place in order for the current to flow out of - or into - the capacitor, so current can flow extremely quickly. This makes them very useful, but very dangerous. The amount of charge a capacitor can hold is a function of the size of the conductors, and the thickness of the insulator, as well as its dielectric constant - how well it resists the flow of current. Larger conductors and thinner insulators allow for larger amounts of charge. Some capacitors only work one way. Reversing the flow of current through them will destroy them (spectacularly). However, the caps for a coil must be bipolar and allow current to flow in both directions. More on that when we discuss the tank circuit.

Here's a nice little animation showing how a capacitor works:

I made the capacitors for the Tesla Coil using super saturated saline and aluminum tape for the conductors, and a glass bottle for the insulator. Everyone says to beware of electricity and water because water is conductive. While this is true, pure water is an extremely poor conductor - it’s the minerals in water that are conductive. Salt is particularly good at conducting electricity when dissolved in water. The reason has to do with the molecular makeup of salt, or sodium chloride (NaCl). When dissolved in water these elements separate and become ions (Na+ and Cl-). These ions can carry a current.

These beer bottle capacitors aren’t as efficient as they could be. Most larger coils use a collection of small capacitors connected together to form what effectively becomes one large capacitor. This method is cheaper and allows for easy replacement in the case of a cap blowing out.

Inductors are basically just the first half of a transformer. They’re typically a coil of wire wrapped around some kind of ferrous material. What they do - at least for the purposes of the Tesla coil - is store electrical energy as a magnetic field and convert that field back to electrical energy. Remember that current flowing through a conductor will create a magnetic field around that conductor, and that a moving magnetic field will induce a current in the conductor. Well, take a coil of wire, run a current through it to create a magnetic field, then stop the current. The magnetic field will collapse (move) back down onto the coil and induce a current. Only, the current will flow in the opposite direction it had been flowing before. Neat!

Another detailed video, this time of an inductor:

Now we can finally talk about tank circuits, or LC circuits. The L in LC stands for inductor because inductance is represented by an L in honor of Heinrich Lenz, who formulated a law that describes the behavior of an inductor. C stands for capacitor (that’s pretty straightforward).

I digress. A tank circuit is what you get when you connect one side of an inductor to one side of a capacitor, and the other side of the inductor to the other side of the capacitor. In this configuration an outside source of electrical energy initially charges the capacitor. When the source is removed, the current flows in some direction through the inductor where a magnetic field forms. When the capacitor has fully discharged, current stops and the magnetic field in the inductor collapses, inducing a current in the conductor (moving the opposite direction) and flows back into the capacitor, charging it back up again. When the inductor’s magnetic field finishes collapsing, current stops flowing into the capacitor and it’s permitted to discharge through the inductor again. Given a resistanceless conductor, this would go on forever. We live in the real world and so eventually the circuit needs to be recharged from an outside source. It’s called a "tank circuit" because the current sloshes through the circuit like water in a tank.

The following - very dry - video goes into a good amount of detail on how a tank circuit works:

The really interesting part of this circuit is that it can operate at a very high frequency. While the frequency coming from the house and through the initial transformer is operating at 60Hz, the tank circuit can oscillate in the MHz range. More on that later.

A Tesla Coil is made up of two of these tank circuits. The first being made up of the beer bottle capacitors, the spark gap, and the primary coil. Its source of energy is the NST. The spark gap is there to act as a voltage regulated switch. The NST charges the capacitors until the voltage causes the air in the spark gap to breakdown and allow current to flow through it. That breakdown removes the NST from the circuit and allows the capacitor to discharge and the tank circuit to resonate.

The second tank circuit is formed by the secondary coil and the topload. The capacitor in this case is created with the topload as one conductor, ground as the other, and the air as the insulator. This second tank circuit is magnetically coupled to the first. Its inductor is placed inside the magnetic field created by the primary coil - making a transformer. The primary circuit is then tuned to the second circuit’s resonant frequency. The resonance efficiently transfers energy and amplifies the voltage in the secondary coil.

Here's a quick animation that shows all of this. The capacitor charges until the spark gap fires, enabling the tank circuit to ring, and resonate with the secondary coil.

To understand resonance a bit better, find a tube that will just fit over your lips. Then, buzz your lips in the tube. Try buzzing at different tones. When the sound out of the tube is loudest, you’ve hit its resonant frequency.

My thanks to the people that helped me edit this: @theatrus, @bitbckt, and @gwennie.