Made my first Tesla Coil
I chose self-oscillating HFSSTC, easy looking at first sight, but little bit tricky to fine tune.
At first launch i got very little unstable flame, after tuning primary coil i got more or less good configuration that produce about 4 cm "brush flame", these results motivated me for future experiments.







Resonator has ~155 turns/0.7mm wire, L = 6 turns/1.13mm wire, C = 100pF, Lf = 9.5uH, mosfet - IRFP460.
Power supply is 230/110V transformer, after bridge rectifier no filtering capacitors present, only 22uF added to HFSSTC input, therefore, i think, flame become bushy and produce loud 100Hz sound.
Used "soft start", at beginning transformer is connected in series with 200W ballast(PTC heater), then pressing a button contactor shorts the ballast and connects transformer directly to AC mains.
I don’t measured DC voltage, AC voltage before bridge rectifier is 110V, power consumption is 330W.

If mosfet internal built-in diode is fast enough, there is real need to use external one, theoretically ?
For example, IXFH80N65X2, half bridge/bridge inverter, 35kHz, 230Vac before rectifier, Pmax=2kW, load - arc on ferrite transformer secondary output.



But in real situation i will add a APT30D60BG diode.

Just another ZVS driver, separate gate bias power supply to avoid startup issues.
Mosfets - 2*(2xIRFP250N), resonant capacitor 600nF.
Main power supply - 1.2kVA 50V DC(60V at low load), two iron transformers connected in series.
For primary coil I use copper foil(68mm wide, 0.1mm thick, 4 pieces soldered together), 2+2 turns.
After 30s of arc, primary is being very hot, looks like need use 3x parallel foils(like a foil litz).
In comparison ZVS/push-pull, with same primary, push-pull primary heating is about 10% of ZVS heating, ZVS produce large reactive currents in LC countur - this is most negative thing for me.




Mosfet connections


Short test video, power readings not synchronized.

Long time ago I assembled this half bridge, only low power test were made trough 100W lamp.
Both FETs heated too much, so decided freeze this project until I get oscilloscope, but forgot about it for 10 years 😊
Found that frequency is too high ~145kHz, IR2153 output is weak for this, replaced Rt timing resistor to correct value.
Test run through 200W lamp – FETs almost cold, all ok.
First direct AC run tripped home 20A circuit breaker, short circuit occurred on pcb, on AC in traces, cleaned pcb and no more fireworks.

Modifications -
Shorter trace from DC bus to FETs.
Changed burned out NTC to SCK-056 5R 6 А.
Used AlN insulator between FETs and heatsink.
Bootstrap capacitor 470uF film > 10uF tantalum.
Snubber film capacitor 2.2uF.

Connected my diy 10T/1000T transformer trough coupling transformer 15T/11T.
Running at 34kHz, max 350W(power factor 0.6) drawn from AC line, FETs heating very slow, with fan - cold.
Calculated primary turns is 12, now using 15.

Schematic, removed relay and indicator leds,  IR2153 power supply switched on manually.


Initial build -


After changes -


One turn over primary.


Low side FET D-S


High side FET D-S

How about using PVC insulated wire in diy ferrite core transformer ? (Cu 0.5mm^2, with insulation d = 2.1mm)
Using zero insulation between layers,  when done put in transformer oil + vacuum (aspirator vacuum pump) to remove air.