Theory & FAQs
How can I make lightning?
At high enough voltages, electrons have a tendency to drift off path from a conductor, and go off-roading through insulators.
(aka how lightning exists)
. . . But why??
To understand this phenomenon, its best to first understand how conductors and insulators work:
Recall the following physics concepts, they are instrumental to understanding a tesla coil.
Resonance is the event that two systems are moving at the same frequency (Resonant frequency)
If the man pushes the swing at the right instance relative to the swing's motion, the swing will move farther. These two systems, the man pushing and the swing moving, are resonating.
In electronics, this translates to the frequency of charging and discharging of capacitors and inductors. (LC circuits)
A change in magnetic field produces a change in current, following the left-hand rule:
Note this can only be done with changing magnetic fields. A magnetic field with no change in intensity will not induce a current on a coil.
A change in current produces a change in magnetic field, following the right-hand rule. In practical aplication, a coiled wire will add up the magnetic fields from each turn like so:
This is essentially the complement of Faraday's law, where both concepts are instrumental for transmiting power and stepping up voltage wirelessly, using transformers.
Combining Faraday's and Ampere's law, we can use changing currents on the primary coil (left), to create a changing magntic field. Now with the secondary coil (right) coming in contact with that changing magnetic field of the primary coil, we can wirelessly form a changing current on the secondary coil.
Depending on the ratio between the coils, transformers can step-up or step-down voltage, inversely to current.
Let's run-down of how the circuit works:
The power supply takes in power from the outlet and steps it up to charge the capacitor in the primary circuit .
This power supply must step up the 120Vac 60Hz from the wall, to a higher AC voltage for the capacitor to quickly build charge.
It's common for these power supplies to reach around 6-10kVac as that is the highest voltage our wallets can typically survive. This is why microwave ovens are highly valuable amongst high voltage hobbyists.
2kV Microwave Oven Transformer
The power supply(60Hz) charges the capacitor. Once the capacitor is charged, there's enough voltage for the spark gap to conduct (Paschen's Law).
When the spark gap conducts, the capacitor and the inductor resonate back and forth at super high frequencies(~100kHz!), charging and discharging each other, until the losses are great enough to stop conducting through spark gap(~300Hz).
Note the primary circuit has 3 different frequencies at play.
60Hz from the power supply
~300Hz as the spark gap fires
~100kHz as the LC circuit oscilates
Exaggerated modulation example
Using the primary and secondary side coils as a transformer, the energy from the primary circuit is transfered to the secondary, increasing voltage at the cost of current. For optimal power transfer, it's desired to have the energy of primary circuit, equal to te energy of secondary. By doing so, the peak current in the secondary coil(inductor) will occur right at the peak change in magnetic field of the primary coil, thus maximizing the energy transfer.
Like the man pushing the swing at the right time to maximize the kid's momentum, the oscilations between charging and discharging of the primary and secondary coils must act the same.
This means the primary and secondary circuits must match frequencies, to push the most change-in-magnetic field, into electric current (and thus voltage)
This is where the lightning happens, and it's essentially a scaled up version of the primary circuit in reverse.
In the primary, a spark-gap allows a capacitor to discharge through an inductor in resonance.
In the secondary, a huge inductor charges a huge capacitor, allowing a huge spark-gap to occur between the capacitor's positive and negative terminals (top load and ground)
Note how I keep using the word "huge," thats because the size of the spark-gap (lightning) is largely subjected to how much charge and voltage the capacitor can hold.
Disclaimer: Tesla coils are not efficient. Although super cool, there are many complex physics involved, meaning cheap designs like mine, can easily source a multitude of losses. Here are some I found and aim to fix in the next revision:
Design & Manufacturing
There are 7 main components in a teslacoil. I'll work backwards, refferencing the schematic from right to left
7) BIG Spark Gap
Design Specs:
5ft long arcs of lightning
500kV, 100~500mA of current
Blue colors preffered
The spark gap should be over 5ft tall between electrodes. I can use the ground as the bottom electrode, and a pointy conductor for the for the positive electrode.
As for the negative electrode, I need a sharp metal to uneavenly concentrate charges at the tip. Something like a nail will do just fine considering the specs call for milliamps of current.
A steel nail will be ideal since steel emits blue light when ionized.
When dealing with high frequencies, it is best to use an RF ground (or Earth ground) as the negative electrode, as opposed to any other form of ground. This means, pounding metal rod, 6ft into the ground.
This step took the most mistakes out of all...
6) Top Load
Design Specs:
Recomended ~30-40pF (at 100kHz)
r = 8" minor diameter
R = 32" major diameter
(Using inches for manufacturing simplicity)
5) Secondary Coil
Design Specs:
Length equal to the toroid diameter
~1220 turns
26AWG enamel magnetic wire
~9" diameter
Took me ~2lbs of 26awg enamel wire wound on a 12" PVC pipe.
Here's how I did it ⬇️
Found the resonant frequency of the secondary circuit ~117k Hz
Here's how I did it ⬇️
4) Primary Coil
Required Specs:
>7 turns
1/4" copper tubing
Same height as secondary
Sturdy assembly
Strike rail @ 2x spacing from coil
Separated strike rail from HV primary turn by 1.8" to ensure no 12-24kVpeak arcs occur.
Used ~27ft of 1/4" copper tubing.
Separated turns by .35" w/ 3D-printed(PLA) spacers.
Held parts together with nylon screws and zipties, to avoid attracting any un-welcomed HV arcs between the primary's turns, and arcs from the secondary coil.
Measured the full coil to be ~25.59µH at 100kHz
Allowed for tunable inductance by using high current alligator clips.
3) Main Capacitor
Required Specs:
100 - 130nF
24,000 - 30,000Vdc rated
Bleed resistors
Sturdy assembly
Finding a single 110nF 24kV capacitor is not easy. . . Nor cheap.
I took the traditional path instead, building a capacitor bank by matching 2x parallel strings of 8x .44uF 3kV caps in series.
(polarity is arbitrary)
Current discharge per cap at full capacity:
I = V/R = 3000Vdc_rated / 12Mohm = .25mA
Current rating of 1 Watt Ohmite resistors:
I = sqrt(P/R) = sqrt(1/12M) = .288mA
.288mA > .25mA ✅
Made sure to leave an inch (25.5mm) of separation between caps to prevent any arcing from the full 12kV of each terminal end.
2) Rotary Spark Gap
Required Specs:
Brass electrodes on bakelite (high temp. resistivity)
Safety gap to ground (to protect MOTs)
4 presentations per revolution
300 breaks(sparks) per second (4500 rpm / ~50Vac)
Using an angle grinder powered with a variac to control spark gap arcing frequncy.
Every rotation presents a shorter path for the spark to conduct through.
Peak currents can reach over 800A !
Meaning temperatures will get HOT.
These spikes would only occur for less than a millisecond however, as the energy peaks at the start of every LC ressonating frequency, so 1/117,000 ~.0000117s
On average, the wires will only experience around 40-50 amps, so 8awg wires will do the job nicely.
Chose bakelite phenolic for its high temp. and electrical insuative properties.
The first attempt at manufacturing the bakelite wheel was a total FAIL...
Taught me to never understimate the power of vibrations emerging from rotating objects with an imperfect center of mass.
⬅ Got it machined properly the second try
Ran a speed test on the angle grinder ➡
Taped a permanent magnet to the bakelite and set up a stationary hall effect sensor to track every turn. Coded an arduino to count everyturn and output the frequency as revs per minute (RPM).
Hit ~2000RPM when supplied by 25Vac from the variac. Assuming this trend is linear, I should receive ~4000RPM at 50Vac.
The angle grinder is also rated for 10kRPM at full power (120Vac), therefore by simple ratios, I should get around 4500RPM at 54Vac.
Safety gap spaced far enough to trigger with 12kV from tank cap, but not close enough to trigger with supply alone. ~1.5cm spacing.
1) Power Supply
Required Specs:
120Vac 60Hz input
8800Vac 60Hz output
~100-300mA
Properly insulated
Portable
⬆ Slide for the build montage ⬆
Read 220Vac RMS at ~3Vac input from the variac
220Vac / 3Vac = 73.333
Anticipating 8.2k-8.8kVac considering losses in core saturation, winding resistance, and leakage inductance.
120Vac * 73.333 = 8,800Vac
3Made a 100 : 1 voltage divider, using 10x 100kΩ, 10W ceramic resistors in series for the 1MΩ, 100W
Read 88.4Vac RMS at 120Vac input from the wall, meaning:
88.4 * 100 = 8840Vac ✅
A single 2kV MOT arcing with 120V input.
Notice the arc length of the 8kV supply is similar to that of a 2kV supply with the same input.
This is due to the capacitor ballast, limiting the current in the 8kV arc. Without these capacitors, the arc would trigger at the same proximity, but last for longer distances...
The transformers' small gauge secondary windings however, would not last quite long.
4x 2kV MOTs (8kV total) arcing at 120V input.
Video coming soon!