A Tesla coil with only three components! 

The problem is that the primary winding's inductance is so low that the circuit oscillates in the HFs - 10MHz or more. From tests I've made the frequency is very poorly determined and it's more like a spark gap jamming Device. Fortunately it's low power. The exciter I built used a BD433 transistor and a power supply voltage of 8 volts and about 200 mA current. The arc will burn a tiny pinhole in the skin of my finger. 😮

@@acmefixer1 maybe if there was a ferrite choke that would help with the 'jamming' and move the oscillation into a lower frequency?

Me and my monkey

👍🇬🇧

😱 omg.. I would of "NEVER" thought of that.. awesome 👍😊

@@young-salt oh I know but so what 😂

Instructions unclear, used fire to start pulse generator.

@@zeekjones1 The days when the 100th Slayer Exciter build started getting boring without some kind of twist, but was still too scared and unskilled to journey into different circuits.

Is the voice coil a single layer of windings? AFAIK all Tesla coils use single layers. The voltage gradient along the coil is very steep and I believe there is a risk of insulation breakdown if a second layer of turns is added, effectively laying down high voltage turns directly onto low voltage turns.

​@@williamfraserOn a larger scale, you're still using three components with a power transistor and a chunky HV capacitor, and you don't even need insulated wire, you can separate layers with graphene or nashua duct tape (just stretch the rubber to peel it off the metal first lol. Your circuit lights all fluorescent tubes in the room and reverberates for twenty-plus minutes using copper tubing for the primary and scrap HV city wire separated by nashua with the windings spaced as closely as possible without closing the circuit through the air. Portable harmonic oscillator has many applications. Most of them are kind of nefarious, but your circuit is quite the feat. You earned a subscriber with alerts turned on with this one. 😅

A little battery powered clock. They have a great coil in them for this stuff.

I too am new. Started in around 2018. But my god I wish I was showing basic when I was a kid.. I would have made a carrier out this.. 👍🇮🇪🙏

That was my initial intention, making it even more compact. It didn't work for me but there must be a way. A ferrite sleeve around the cap to provide some magnetic shielding perhaps, or simply adding proper insulation around the top of the cap housing.

Unlikely to work. The aluminum can of the capacitor acts like a shorted winding and robs power from the circuit.

lacing the cap inside the coil produces too much parasitic capacitance and reduces the Q of the coil.

Where do you think the video was filmed..

touching the lead probably would also trigger it especially if you are inside a building, or if you also touch the + wire. connecting the gound to the earth and then letting a kite with a rope wet in salt watter fly up and either getting close enough to it, or making it touch the middle terminal should also trigger it, if the kite is high enough and if the salt water is of the right concentration then you might not need to ground it.

The spark from the lighter simply induces the first pulse to trigger the transistor (if that is the correct term). The power for the oscillation comes from the supercap which acts as a low capacity rechargeable battery with low ESR. It still needs recharging from an external source.

OK so the Super capacitor is already charged up prior to the Spark Igniter ? This video looks extremely suspect to me.@@williamfraser

@@seditt5146 yes already charged. No free energy bs

Tried underdriving an unmodified slayer with a BD243C BJT, red LED, 1.2kOhm resistor, and 200nF of ceramic capacitance. Got down to 2.08V before the PL-S fluorescent tube extinguished and the circuit failed to relight with a piezo striker. The circuit further oscillates down to 1.4V without ionizing anything. Preheating at full power to a case temperature of 135ºC and then underdriving did NOT produce the results I expected (lower minimum input voltage operation). The circuit would extinguish at 2.5V and would not relight with emfs below that. Reducing case temperature to -18ºC, however, did allow me to reduce input voltage further to 2.03V with a base-emitter peak voltage of 1.08V and DC current of 140mA. So while higher temps _should_ offer higher gain and lower base-emitter voltage drop (as much as 200mV), this high power BJT performed better with even lower gain and higher Vbe. Actually, rather than these parameters being relevant, I noticed its transition frequency is 3MHz yet the circuit it was designed for exceeds it at 3.6Mhz. This means gain is affirmatively below unity (1) at my frequency, fractional even. By comparison, the 2N3904 has a transition frequency of 300MHz, so there wouldn't be significant attenuation of gain due to frequency and it may still benefit from some thermally driven tweaks to its properties before needing to change transistor technologies.

The ceramic cap definitely makes a difference, at least when there is a resistor on the base. Another place where I tried two parallel but opposite facing LED's is between the positive terminal of the supercap and the resistor. Both LED's lit up so I assume the primary also creates a reverse pulse, perhaps induced by the secondary, that stores some energy in the ceramic cap before the new cycle starts. Just guessing here, if you have a scope you can test it! I also had success with a 2N7000 mosfet with nothing but the secondary on the gate. It was an easy substitute on the breadboard because the equivalent pinout is the same order as the 3904. Adding a gate pull-up resistor just resulted in overheating and no pulsing. Adding a pull down resistor worked well, even without a pull-up. The circuit can be started by simply touching the tip of the secondary, no emp required. What worked best was using an led between ground and gate and another between gate an positive (cathode goes to positive). Both light up strongly so the secondary reverse pulse must be pushing the gate well above positive. The LED's should also provide protection for the gate against Vgs limits.

@@williamfraser I'll find some time to do that, but rather than modify my slayer kit which is rather inconvenient to partake, I'll repeat the same breadboard layout you have and wind a new coil so we have similar attributes. Was there any reason for the 80T and diameter? Going for a specific coil inductance, ratio, time constant, or something else?

@@InfinionExperiments I made the 28mm coil years ago, I don't recall why I decided on such a low turn count, it would not have been based on any electronic property, more likely just to explore the limits of what might work. Small coils with high length to diameter ratios all seemed to require ferrite cores to work well while the short wide coils work well without ferrite.

@@williamfraser🤔 could you use this and step up the power and voltage with 2-car generators that diodes being used as the three leads on a AC generator. Run a distribution panel and about 90 amps and 120?

Trial and error! I have seen several topics on piezo crystals used in gimmick emp devices. Some other methods of induction also worked, all involving sparks, like shorting a large capacitor next to the coil. The lighter was just far more convenient.

The "trick with the lighter" would not be needed if the base bias resistor was not removed. The resistor is so inexpensive that it should not have been removed. 😮 🤨

@@acmefixer1 and how would it still be 3 components?

@@cornishcat11 Transistors are current operated devices; therefore the base should **always** have base bias current, no matter how small. The base bias could be as low as a microamp, through a 10 megohm resistor. You said it's "only 3 components" but it's *not!* The fourth component is the lighter - without it, the circuit won't start! 😈😳

@@acmefixer1 So you think a 'Start Cart' is a component of a jet fighter?

Lekker!

This is my first foray into high frequency so right now it is just an intriguing educational gadget.

I gave it a try, used various ceramic caps from 10nF and up to 220uF. It would not start, even when using the emp. When I added a 10k resistor in series it started (using the base-emitter LED as an indicator). A diode or LED in series it also worked (anode to positive). To start a second time the cap first had to be discharged.

Isn't that exactly what a Tesla coil is, a resonating tank circuit between the coil and the parasitic capacitance of the air around it?

@@williamfraser I think he meant like where you give it an initial "charge" and then "release" it and then it oscillates till it "settles down".

At this small scale, probably nothing more than an educational circuit or a party-trick gadget. The neon tube is just to show the field and gauge its output power. I have not heard of it being used in fluorescent lighting as such. According to Wikipedia (see "Tesla Coil") modern applications are in arc welding and for leak detection in vacuum vessels (the coil needs enough power to create a coronal discharge, the arcs are drawn to any tiny hole in the vessel, making it possible to visually identify it.)

Tesla coils are a technology developed by Nikola Tesla in his intellectual pursuit of abundance of energy from the natural world such as self-acting engines, transmitters and collectors of atmospheric electricity, as well as functional wireless power for the 19th and 20th century industrialists developing infrastructure for New York and other cities in the Americas. Today we have had the means to popularize some of the patents, essays, and ideas for tesla but not so much the big picture of its functional purpose. Yeah there practical applications for this specific device for education and testing, but there is a deeper problem or challenge to overcome that was left by the man.

The "ground" here is free air

@@ZomB1986 when he touches the 1 wire HE becomes ground. there is no ground. the FREE air as in the atmosphere is a capacitor. That is what lighting happens witch is a short. So your saying you are grounding in the middle of a capacitor?

Look for tutorials on basic electronics, then learn about more specific components and how they are used in various applications. Get a breadboard and a variety of components and just start building and experimenting.

I suppose it depends on the flux density at the coil. If the source is too far away it simply would not induce enough current to trigger the transistor. A frequency close to the resonant frequency will probably work better.

Yes it could do that, all it needs is a pulse strong enough to make the transistor turn on.

Wouldn't an EMP weapon simply burn out the transistor?

Well, with EMP close enough, the coil wire will likely burn through. Or induce enough current to fry the transistor.

@@ZomB1986 After thousands of Amps flow through the ex-coil, yes.

Yes the 2N7000 does exactly that, it starts by simply touching the secondary with a finger. I haven't tried it with the smaller coil but it works as a direct replacement for the bjt on the breadboard (the pinout is compatible). Output (neon tube brightness) is less than bjt. It works well with an LED between gate and ground and another between gate an positive to act as clamps on the gate voltage swing. It is self starting with the pair of LEDs.

I would imagine so. Thin wire, higher resistance, less energy dumped every time the transistor switches on. But there could be all kinds of secondary effects to complicate things.

The frequency and voltages are out of my multimeter's range, I think. I tried making a capacitor-diode-coil "receiver" to measure rectified induced voltage on the loose coil. Max I ever saw was 75V with the two coils almost touching. Turns ratio was 5:8.

@@williamfraser , Thanks for the measurements! It's really interesting, I love such experiments )

I used a BD433 transistor. But most experimenters just use a 2N3055. It works well with the power transistor from an old AT power supply.

@@acmefixer1 basically any transistor with a decent amount of gain will work. The ZTX line of transistors requires very little base drive. In fact, if I remember correctly it oscillates down to about 0.2 or 0.1 Volt... If only my memory worked as well!! 🤔😂

@@gertbenade3082 Germanium transistors can oscillate down to 0.2 or 0.1 volt, but not silicon BJTs - they can't get below 0.5 volt. But the supply voltage is 8 volts or so. So to get enough power the transistor needs to handle several watts of power and thats why I said the BD433 would do it, but so would a 2N3055.

@@gertbenade3082 Wow, Zetex makes some great transistors! The datasheet surprised me just how good its attributes are. You're right, the VBE runs as low as 125mV at 175ºC thanks to its 200ºC max operating temperature, and the hfe can be as large as 1300 bordering a darlington. The main demerit to that one seems to be its toff times are a very sizeable 1300 ns, but the output capacitance is very low at 16pF. Might honestly not be an issue if it's small signal behavior that dominates here in the slayer. I did a bit of browsing and the ZTX series netted a bunch of other options. My favorite for this application would definitely be the ZTX1051A. While it has double the output capacitance, it more than makes up for it with 10A pulse current (from 6) and 300ns turn off times. Moreover, the transition frequency is slightly improved at 155 MHz (higher hfe at 10MHz) and Vce sat is as low as 25mV. With the knee of the IV curve being much further over, it allows 1/4 the Vce and power dissipation at 2A compared to the ZTE690. I think you found my newest favorite high performance through-hole transistor.

​@@acmefixer1 If the BJT is self-heated or externally heated so that its temperature is elevated, then the BJT will have lower than 0.5V Vbe. For gert's ZTX690, all that needs to be done is raise the junction temperature to 100ºC to break through to 0.4V Vbeon and 175ºC to break through to 0.2V Vbeon. This is only critical once the input voltage drops to 0.4V, so the application already naturally derates the transistor. The BD series is a really slow BJT class, hot garbage for small-signal operation. I bought a slayer kit that used a BD243C which is similar to Onsemi's BD433. It's more suitable for large-signal DC operation. In both the 433 and 243, the transition frequency is 3MHz. In the case of my slayer, the resonant frequency was 3.7MHz, so the hfe gain ended up being below unity at around 0.6. Had the coil been a lower inductance reaching higher resonances of 7-10MHz, the gain would be 0. The great thing about the ZTX690 is that you don't need heat sinking at all compared to the BD series, which makes perfect sense then to use a TO-92 package. At the frequencies and input voltages we're using here, it's better than MOSFETs too.

I'll have to try, sounds like it might work.

Got working just as you suggested. A series 10k resistor also worked instead of the diode. The cap had to be discharged first before I can start it a second time.

@William Fraser using the smallest capacitor possible, may help with auto start as it may drain by air resistance.

Replicate simple circuits to develop an understanding of how specific components work. Using a breadboard allows you to test "what-if" ideas in rapid succession. Every small change is an opportunity to learn, why it worked or didn't work. Circuits using a 555 timer chip is a fantastic place to start because they often form a crucial part of more complex circuits with real practical applications.

I don't know the HFE in operation, or how to measure it. The very proximity of test leads seem to alter the circuit's output (neon tube brightness). The supercap is a run of the mill 10F Kamcap I bought from Mantech.co.za.

don't need to measure it, it's well documented in the 2n3904 datasheet. hfe is 35 at 10MHz

28mm