How to Build a Vacuum Tube Tesla Coil (VTTC)

Can't find a neon sign transformer? Want to build your first Tesla Coil without facing the complexities of going solid state? Here's some good news: Vacuum tube Tesla Coils, which have been making a comeback in recent years, can be just as rewarding as any other type of coil without breaking the bank. This is one such coil that I built during the spring of my 8th grade year.

You can even use it to wirelessly transmit electricity to a lightbulb! (12/3/12)



Noteworthy:

While this project does work in its current form, I have detected some problems and and working to fix them. You would best be advised to postpone your building until then - it seems that theses tubes could operate more efficiently at higher frequencies and my primary RLC tank circuit's natural frequency is much higher than my secondary side RLC circuit's natural frequency; a new secondary coil with a frequency of approximately 1.5MHz is being designed and the primary circuit will be retuned. I expect a great leap in performance, with sparks possibly as long as 7" to 9".

3/10/10: I decided to try to estimate the resonant frequencies of my primary and secondary circuits using deepfriedneon's formulas, and I found that my coil is oscillating about 100kHz above my primary circuit. I don't have any parts to fix this now, but will add a capacitor or two to the primary circuit to lower its frequency soon. IMPORTANT: I found a 6.3V at 12A Hammond power transformer and replaced my 5V computer power supply - the results were truly impressive; I am getting better performance with one tube than I ever got with two, filament voltage really matters! Here is a quick video:



4/16/10: The MOT (plate transformer) burnt out because the secondary windings were damaged by previous experiments (SGTC,s, Jacob's Ladders, etc.). It was replaced with a larger one and the sparks are now almost 7" long - this coil performs as well as Steve's did, but with only one tube and with a poorly tuned primary oscillator!

Thanks!

In the early 21st century, many of us have never even heard of vacuum tubes, and the few who have only know that they were used in old electronics. Therefore, before I begin this project, I feel the need to explain how they work.

Lets take a look at picture 1. This is the standard symbol for a vacuum tube diode. A diode only conducts electricity in one direction. They can be used to turn an alternating current into a direct current. In the diagram, the bottom half hexagon is the filament. It is just like the filament inside of an incandescent lightbulb. The line above it is called the plate. The circle around the filament and plate represents the (usually glass, sometimes metal) envelope of the tube. Almost all of the air inside of this envelope has been evacuated, there is a vacuum. This will become important later.

Now, lets take a look at picture 2. Here we have applied a voltage between the filament and the plate. The filament is negatively charged, and the plate is positively charged. While the electrons in the filament are attracted to the plate, there is not enough voltage for them to do so on their own. So how can we get them to jump? Take a look at picture 3.

In picture 3, a few new things have appeared.First, we have a 10 volt power supply connected to each side of the filament. Just as in an incandescent lightbulb, this heats the filament up. The negative side of the power supply is still connected to the filament, but the positive side is not. Notice that now, the negatively charged electrons are stil flowing into the filament from the 100 volt power supply, but something is different. Why are they floating around the filament? As the filament heats up, thermionic emissions occur. Essentially, the electrons are shaken off of the filament by its thermal energy. This can happen because there is a vacuum. So now, the question is: What happens when we connect the positive side of the 100 volt power supply to the plate? Take a look at picture 4 to find out.

In picture 4, the positive side of the 100 volt power supply is connected to the plate. We have zoomed back towards the tube. In the picture, the electrons floating around the filament are moving towards the plate! There are no air particles to hinder their passage, so after the thermionic emission occurs, the positively charged plate attracts them, and they accelerate towards it, hit it, and move along the wire back into the power supply. That's how a vacuum tube diode works.

The principle of operation is relatively simple, but a Tesla Coil such as the one that we are building is an oscillator. That means that there is a feedback system that turns the diode on and off, to accomplish this, we use a triode. Read on the find out how it works.

The first true electronic amplifier was the vacuum tube triode. It works because like charges (in this case, electrons) repel each other. Take a look at picture 1. It looks similar to the symbol for a vacuum tube diode, but it has an extra part that looks like a grid between the plate and filament.

This grid normally allows electrons to pass through itself for diode operation, but as it grows more and more negatively charged, it allows less and less electrons to travel from the filament to the plate due to electrostatic repulsion. In this way, you can regulate the flow of a relatively large current by using a relatively small one.

A Tesla Coil is essentially a very large oscillator. When the primary side of the coil oscillates at the natural frequency of the secondary side, resonance is achieved. This is a fundamental concept that is used in all Tesla Coils and other resonant transformers (such as the ones found in many switch-mode power supplies, and CRT television sets). The Vacuum Tube Tesla Coil that I detail here uses a configuration known as an Armstrong Oscillator.

In the standard model of a transformer, there are two coils, a primary and a secondary coil. Currents are usually induced from the primary coil to the secondary coil (although the opposite sometimes happens, usually with destructive results), this is a concept that we will not go over now, if you are unfamiliar with it, then this is a good place to become acquainted: http://en.wikipedia.org/wiki/Transformer. However, an Armstrong Oscillator works by introducing a third coil, called the feedback, or sometimes "tickler" coil.

Currents are not only induced from the primary coil into the secondary coil, but also into the feedback coil. This feedback is then used to turn off the oscillator by blocking current from flowing into the primary coil. However, when the primary coil is turned off, current is no longer induced into the feedback coil, and it no longer blocks current from flowing through the primary coil. In this way, the cycle repeats indefinitely, until it is interrupted, or the power is switched off.

The basic schematic for an Armstrong Oscillator using a vacuum tube is given in the first picture. (This picture is from Steve Ward's site: http://www.stevehv.4hv.org/VTTCfaq.htm, you can read more about VTTC operation there)

Here is the particular schematic for the Tesla Coil that we will be building. I do not take credit for its creation - it was made by Steve Ward and you can find the full - sized image on his site here: www.stevehv.4hv.org/VTTC1/dual811Aschematic.JPG. A few things that you should note are that I have found that you should make the primary coil (L1) slightly larger but allow for it to be tapped every other turn. Also, I've noticed that a slightly larger (~2nF) tank capacitor (C1) works better for my coil, but this could vary. Also, if you are adventurous enough, you might consider using a level shifter to double the voltage to the 811A tubes to 4000VAC RMS and then use a staccato circuit (something you should consider even without the level shifter) to keep the tubes running cool. However, since this is a slightly more advanced project I will not cover it here (yet!).

Here are the parts that I used, and the approximately how much each one cost:

~$30 (2) 811A Vacuum Tube Triodes
~$0   (1) Microwave Oven Transformer
~$10 (1) 30kV 1.0nF Polystyrene Capacitor (2 or 3nF will also work here, I found that larger capacitance increased the performance slightly)
~$15 (1) Bundle of 1000 ft. of 28 AWG Magnet Wire
~$10 (1) Bundle of 100ft. of 16 AWG Insulated Audio Wire (Can be purchased from Radio Shack)
~$3   (2) Small Circuit Boards
~$10 (1) 1' x 1' x 1" Wooden Board (This price is for about 10 of these boards)
~$5   (1) Box of Nails
~$0 (2) Ferrite Toroid Cores (Should be at least about 1/2" in inner diameter, these are not critical parts)
~$2   (4) 3kV 1.8nF Ceramic Disk Capacitors
~$10 (1) 50W 5k� Resistor (�  = Ohm Symbol)
~$7   (2) 20W 30� Resistors
~$0   (1) 6.3 VAC 10A Filament Transformer (You can use 5 volts from a modified computer power supply instead if you want to save some money, but the performance will decrease)
~$5   (1) Small Container of Epoxy Glue (This is for gluing the secondary coil down, if you want to experiment (like me) then you shouldn't glue the coil down)
~$5  (1) 1' of 4" Diameter PVC Pipe (Primary Former)
~$3  (1) 1' of 2" Diameter PVC Pipe (Secondary Former)

Total: ~$100

This is actually a very crude estimate, and shipping costs will differ depending upon where you buy from and where you live. The parts marked $0 were either salvaged (like the Microwave Oven Transformer) or they were free samples (like the Ferrite Toroid Cores) or were already owned (I used a computer power supply instead of a costly filament transformer).

While there are no general guidelines on how to do this step properly, you should strive to fit everything on one board and keep connections as short and simple as possible with as few overlapping and/or twisted wires as possible in order to minimize stray capacitance and inductance. Here, you can see my main board with all of the major parts.

One thing you may have noticed is that some of my photos appear to show a coil with a different base - this was the original variant of the coil which did not work as well because of the long hookup wires.

NOTE (4.5.11): This circuit layout is now obsolete - arcing between the top leg of the primary capacitor and the microwave oven transformer's iron case has become problematic to the point where I completely reassembled the circuit and put the transformer on its own board. Hopefully, I'll get some pictures of this soon.

Unfortunately, I did not take pictures of this part of my construction - both of my hands were too busy winding and holding!

The ideas involved in winding an effective secondary coil are very simple, but some things are easier said than done. All that you are really doing is winding wire around your 2" PVC pipe former. Here are a few general guidelines for winding your own coil:

DO NOT:

1) Break the wire - if the wire snaps half way through, it is better to buy a new roll (or buy are bigger one in the first place) than to solder the broken wires back together. This is not a good idea because you will risk serious damage to your coil - there will be problems with unwanted discharges (the secondary coil could potentially arc to the primary coil and ruin the entire primary circuit; very bad) and the coil can also destroy itself by burning through the plastic form or by melting the solder you used to hold the rip together, thus unwinding the secondary coil.

2) Drill Holes in the Secondary Coil Former - this is the most common mistake, and you will pay the price for it with this coil. If you drill holes in the secondary former, there will be a huge risk of the coil discharging through the inside of the pipe or discharging upwards though the secondary former (at the top) and damaging itself (not to mention that the impressive sparks won't be flying into the air, but rather melting through the secondary former).

3) Wind the Secondary Coil Haphazardly - If you cross windings or wind them on top of each other, the performance of your coil will suffer greatly and the secondary coil will be at risk of damaging itself. While its okay to make one or two small mistakes (with emphasis on small!) near the bottom of the secondary coil, you will regret it if you do not wind well.

DO:

1) Use relatively thick wire for winding the secondary coil - increasing the thickness of the wire you are using will make it easier for you to wind the secondary coil and will decrease the chances of the wire snapping.

2) Wear gloves while working with your coil and/or wash your hands very well - some of the various molecules in your sweat and on your hands, if caught on the secondary coil, can decrease performance. While you can wind the coil with sweaty hands, you will notice that the sparks will be shorter than if you had used dry, washed hands or worn gloves.

3) Work slowly and deliberately - its not a race, you will make fewer mistakes if you are willing to commit a few hours to winding the coil. Sometimes, if you make a mistake earlier on, you might want to unwind and then rewind the coil entirely. For this particular coil, I wound the secondary half way through before I noticed some crossed windings and rewound the whole thing.

4) Use shellac (I used the spray-on type) or polyurethane to cover your secondary coil - this will help to prevent the coil from unwinding and will hold everything together well, it also looks and feels very nice. Give it a good day to dry off after the shellac (even if the can only says 15 minutes) as the secondary coil might erupt in flames if the coating has not dried thoroughly.

When you're finally ready to turn your coil on, be sure that you have a large open area to work in where there is no danger of sparks from the coil setting anything on fire. I would recommend testing the coil at a lower voltage first (using a Variac - I would start by testing it around 30VAC input first and then working up to full power) instead of plugging it in to see what happens. Also, a 10 ampere FAST safety fuse is REQUIRED in series with the mains electricity you are using in order to prevent electrical fires and other nasty scenarios in the event that your coil does not work properly.

Before you plug anything in, however, you should use a multimeter to make sure that your connections are all correct and that your vacuum tubes are not damaged (ie. burnt our filament, shorted grid and filament, etc. - all of these have happened to me when working with vacuum tubes)

On the first attempt, do not expect to immediately be rewarded with roaring sparks - be glad if your coil works at all. Once you have established that it does, then attach a small topload to the top of the secondary coil (I like to use a filment lightbulb wrapped tightly in aluminum foil) and use the taps on the primary coil to tune your coil for maximal spark length. Note: you will need some sort of breakout point like a sharp nail if you use a topload of any sort).

Here are some pictures of my Tesla Coil's first light:

If you're satisfied with the way your tests are looking, then you can plug everything in and enjoy the plasma! Interesting experiments you can try include observing discharges inside of an argon filled lightbulb, inside of a vacuum tube (be careful some tubes might produce small amounts of X-Ray radiation if you do this by way of Bremsstrahlung), and you can light up fluorescent tubes at a distance. Also, if you remove the breakout point and tune a nearby radio (and sometimes a faraway radio too) to the resonant frequency of your coil (Usually somewhere on the AM band) you will be able to hear the 60Hz buzz of the coil.

This is a page where I will answer general (and sometimes specific) questions that you have and will try to help to explain some of the deeper operating principles principles of the Tesla Coil.

On a different note, the 811A also makes a great audio output tube!