(* this page is a bit out of date, I finished it and made some improvements to the design and chose _much_ better power transistors to use. I will update these details soon*)

I am currently building my first Tesla coil. I started out after being inspired by Mehdi Sadaghdar's design. I tried copying his design but, wasn't able to get the driving circuit to operate at a high enough frequency. Not sure why, I did build this on a bread board so maybe that is the issue(?) I have done high frequency projects on these before and didn't have issues until I got to much higher frequencies (100s of MHz), but I am not ruling this out.  Either way, I decided to just design the thing myself. There is still some work that is necessary to make a sparking, solid state driven, pulse width modulated, low power Tesla coil. But, it is well on its way -- I can currently illuminate light bulbs in my hand. I think the reason I can't get more power out of it is that I am not able to operate the power transistors in the switching regime and am basically just pumping all the energy into heating those bad boys up.

### A Tesla what-now?

A Tesla coil is actually a special type of transformer: a resonant transformer. Tesla coil transformers can provide voltage increases by amounts impossible with normal transformers. Unlike with most transformers you hear about in a physics class, with a Tesla coil you want a very weak mutual inductance between the primary and secondary winds (to mitigated the effect of back emf caused by the discharging torus on the driving circiut). The relationship between voltage increase and number of turns is not as simple as the ratio of turns like it is for a typical transformer. To elucidate this, I worked out a simple model of the Tesla coil to see what the effect of the coupling is.

A simple model for a Tesla coil is an LRC circuit that is being driven by a black box oscillator and coupled to the secondary wind through Faraday's law (i.e. )  (click here for the details).

### The design

I designed the circiut in modules on different days to try to make the design more manageable. As a result, my design is not optimum. I struggled for a while trying to get each chunk of circuit to work properly with the other chunks. Here is an interactive version of the circuit diagram.

The general idea is that the resonant frequency of my secondary coil and cap is very high (in my case about 222 kHz). The sparks that come off of the coil after just a few oscillations (I think). So, if I were to turn the circuit on and off at a frequency in the audible range, say 10 kHz, the sparks will be emitted then not emitted then emitted and so on at that frequency. The noise generated from the sparks will also be generated at that  lower frequency allowing the human ear to perceive the frequency that the circuit was turning on and off at. Therefore, if I turned the circuit on and off according to an audio wave form, the Tesla coil would become a speaker.

To achieve this modulation I designed an oscillator with pulse width modulation. By varying the pulse width I will be varying power delivered to the coil.

The oscillator consists of a comparator oscillator  that produces a triangle like wave that is then fed into another comparator. The second comparator turns on when the voltage coming into it is above the threshold value set by the other leg. This other leg is where the audio signal will be fed in. Finally, the signal is sent to the primary drive transistors, but first through an op-amp. This op-amp increases the voltage in the signal so that the voltage is enough to operate the transistors in the saturated regime.

The brunt of the power being delivered to the primary coil will be coming from a home made power supply. The design for the power supply consists of a full wave rectifier and some smoothing capacitors. In addition to the bridge and capacitors, I put in a fuse and bleed resistor for safety. The capacitors will hold charge long after the circuit is unplugged and forgotten about unless first drained through the bleed resistor.

The power from the full wave rectifier power supply is directed into the primary winding of the via some transistors. These transistors will turn on and off according to the signal being fed in on their gate pin. These transistors need to be able to operate at relatively high frequencies. The high freq requirement eliminates most of the transistors that can handle big currents or high voltages.

The physics says that I want as much current flowing through that primary coil as I can possible get so that I produce a large magnetic field. To get high current gains and still be controlled by the oscillator circuit that I built using integrated circuits and op-amps rated for maybe 100 mA

### Building the thing

The first thing I did was wind the secondary coil (with the invaluable help of my dad). The coil was about 1300 turns of 32 AWG wire on a 2 inch PVC pipe. The ends are held from unraveling by a few winds of electrical tape.

After the secondary coil was wound, the top torus and mount were constructed. This was a lot harder than it looks. We used a propane torch to help us bend 1 inch aluminum stock to hold the top toroidal cap on top of the secondary. The heat is not necessary, I just wanted to have an opportunity to burn my fingers -- I succeed. The toroid was made out of a 3 inch dryer ventilation conduit. The PVC helped make everything very modular so the whole assembly comes apart very easily and parts can be interchanged without much work at all.

The driving circuit consists of a pulse-width modulated low voltage driver that is connected to a parallel array of solid state transistors. These transistors dissipate a lot of heat so a heat sink is necessary. I didn't want to spend the money to buy heat sinks, so I decided to make one myself. The heat sink I made is more of a reservoir than it is a sink, but because I am not intending on using this thing for very long periods of time (at least initially), I think this will be OK.

### Saftey

Do not attempt this on your own without proper training. There are many hazards involved with this project. Not only are their extremely high voltages present, the arcs also produce ozone which is very bad for you -- it will corocorrode lung tissue for example. Ozone will begin effecting your health even if the concentration is to low to smell it. Always use tesla coils in a well ventilated.

Here is the bill of parts he has listed for the circuit (excluding the circuit for the power supply):

List of parts needed just for the circuit

Description
number needed
part number
provider
unit price
1k Ohm Resistor (R1, R4, and R7)
5
1.00KXBK-ND
Digi-Key
$0.84 10k Ohm resistor (R2, R3, and R8) 3 10.0KXBK-ND Digi-Key$0.84
100 Ohm resistor (R5)
1
100XBK-ND
Digi-Key
$0.84 2.7 Ohm resistor (R6) 1 RSA-2.7RCT-ND Digi-Key$1.85
multi-turn trim potentiometer (POT1)
2
A105870-ND
Digi-Key
$0.95 single turn 100k potentiometer (POT2) 1 987-1312-ND Digi-Key$0.67
100uF capacitor rated to 16V (C1 and C4)
2
493-10816-1-ND
Digi-Key
$0.66 10uF capacitor rated to 16V (C2, C5, and C9) 3 P807-ND Digi-Key$0.23
100nF capacitor rated to 16V (C3, C6, and C8)
3
478-4873-ND
Digi-Key
$2.18 100pF capacitor rated to 10V (C7) 1 338-3093-ND Digi-Key$1.52
10uH inductor rated to 2A (L1)
1
AIUR-03-100K-ND
Digi-Key
$1.72 7805 5V regulator (U1) 1 MC78L05BP-APMSCT-ND Digi-Key$0.49
dual comparator (U2 and U3)
2
MCP6562-E/SN-ND
Digi-Key
$0.71 MOSFET driver IC (U4) 1 576-1211-ND Digi-Key$2.59
Power MOSFET, 10A, 1200V (Q1 to Q4)
4
SCT2450KEC
Digi-Key
$4.89 OpAmp (U5) 1 LM358ANFS-ND Digi-Key$0.42

#### Acknowledgments

Thanks to Mehdi Sadaghdar for the initial design and thanks to my dad for letting me use his test equipment and for his advice.