Introduction: Easy SSTC, Slayer Exciter on Steroids!
In this instructable, I'll show you how to build a REALLY powerful version of the slayer exciter circuit! This design is powerful enough to be considered an SSTC, or Solid State Tesla Coil! It can throw really hot electrical arcs of plasma over 3 inches long, and has the capability to be modulated! By modulating the circuit with a square wave with a frequency of around 60Hz and 10% duty cycle, you can achieve sparks that look like a small Vacuum tube tesla coil, and also greatly reduce power consumption!
If you prefer to watch my video tutorial, or just see what this thing is capible of, watch the video below:
Step 1: The Schematics & Plans!
Compared to the normal Slayer exciter circuit, this circuit has some major changes. The biggest one that drove the rest of the design was switching out the normal BJT NPN transistor for a MOSFET. I used 2 IRFP250 MOSFETs in parallel to improve the current handling capability and thermal characteristics compared to the comparatively wimpy MJT3055. However, using FETs presented an issue with the gate-drive, as the gate could no longer be driven directly by the feedback. (Most likely due to a impedance-matching issue with the gate acting like a capacitive load) To fix this, I simply used one of the very many DS0026 MOSFET driver inverted buffers I had in the parts bin.
I also added a resonant tank capacitor across the primary to make use of the EMF flyback and help the circuit achieve resonance, and a 10K --- 100K resistor across the input and output pins of the DS0026 to increase stability and reliability of the oscillation. Since the input to the DS0026 is susceptible to the uncontrolled voltage transients from the feedback, I also added a few 1N4148 diodes to shunt voltages that go 0.6V above or below the supply rail. If these are not used, the DS0026 will die within seconds of operation.
Step 2: Choosing a Suitable Power Supply
To power this circuit, there are many limitations on the type of power supply that can be used. For starters, Linear or unregulated power supplies MUST be used. Most modern Switch Mode Power Supplies (SMPS) will often not work as inrush current and other current peaks will cause them to go into foldback current limiting mode, which basically means that they turn the output off most of the way as a form of short circuit protection until you reset power to them or when the load is disconnected. Therefore unless they are modified/hacked, they will not be suitable. and as for the small little wall bricks, FORGET about those. Even the older unregulated transformer based ones are simply not big enough.
Now to achieve the results I did, you are going to need one beefy power supply. in my video, I used a few heavy transformer based power supplies configured to deliver between 30 and 50V. One of the transformer power supplies used was the 12V 10A beast shown in the picture, and this power supply was in series with another, smaller 12V 4A unregulated power supply, which was in series with yet another 24V 2A transformer. Under no load, I was achieving about 50V. However, under heavy load (peaks as high as 8A), the voltage would sag to 35V. and the smallest transformer would overheat, so I had to figure out how to make the circuit not draw so much current.
My solution was to only power the circuit 10% or so of the time. I did this by using a 555 timer to turn ON and OFF the power to the DS0026 very quickly. A small NPN transistor was connected on the HIGH side of the DS0026 so that the the 555 can directly control the voltage on the supply rail for the DS0026. To power the modulated circuit, you will need at least a 12V 4A power supply, but to achieve the same results as I did in my video, however you will need 50V 2A.
At voltages higher than 15V, however, you need to consider that you will either need to use separate power supplies with a common ground (it is OK to use a small 12V wall wart for this supply), or use some sort of voltage regulator or buck converter to step main supply voltage down to 12V. I learned the HARD way 4 minutes into the video that a 7812 is NOT suitable to regulate a 50V supply. Killing my one and only 7812, I ended up using a dodgy Zener Diode Shunt Regulator. This solution is not ideal, and certainly not the way I implemented it. Only use this method if you are desperate like I was at the time. Regardless if that is the solution you plan to use, take a look at the last picture for the schematic.
Step 3: Prototype the First Circuit on the Breadboard
There is not a lot to say here, Just build up the circuit on a breadboard, connect the power, and see if it works! If not, troubleshoot and see what is wrong.
I recommend to start working on the power supply section first, in my case, that's building up the Zener diode voltage regulator. Make sure that can deliver 12V or so, and at least 100mA. Using a wall adaptor for this instead is fine, just make sure you pay attention to polarity and that the negative of that supply is connected to the common for the entire circuit. Use a large filter capacitor of least 1000uF as well as some smaller film or ceramic capacitors for better filtering of higher frequency components.
Once you verify the power supply regulation part of the circuit is working, you can move on to building up the DS0026 section and testing it. Once you make it, you can test it by connecting a LED to one of the outputs of the DS0026 and ground (paying attention to the LED polarity) and connect the input of the same buffer to ground and Vcc. The LED should light up when the input is grounded.
Step 4: Optional: Now Add the 555 Interrupter Portion!
Now just modify the circuit by adding the transistor between the power rail for the DS0026 MOSFET inverted buffer driver and the 555 circuitry to drive that transistor. Simple enough, right? (make sure there are no filter capacitors across the DS0026 itself, that defeats the whole purpose of modulating power to it on and off, since the capacitor will just smooth that out. But do keep the capacitors on the power rails themselves.)
Step 5: Testing, Troubleshooting, and Pictures!!!
The golden rule of troubleshooting electronics is Thou Shalt test voltages, according to Dave from the EEVblog! So make sure that the power supply's you are using are healthy and working as they should. Then make sure that the 12V output of the regulator or separate supply is also powering the circuit correctly. If you are using a regulated power supply, make sure you are not triggering some sort of short-circuit or overload protection. Many modern power supplies use foldback current limiting meaning that if the circuit pulls more than a certain amount of current for even a split second, the output voltage will drop to a tiny fraction of what it was before, and stop the functioning of this circuit.
Triple check all your connections are secure and good, and that you did not goof up anything. If you choose to use slightly different parts, make note of those assumptions and make sure they are not whats causing the problems by trying the parts that I used.
If you are using the 555 interrupted version of the circuit, make sure that the 555 is oscillating correctly. If you have an oscilloscope and a 10X probe, probe the output of the 555 while power is applied to just it, and make sure that it is in fact oscillating correctly. If you choose to use different parts for this, make sure that you are getting a square wave with a low duty cycle and about 10Hz-500Hz. If not, using a speaker connected to the output of the 555, you should hear some sharp annoying buzzing noise.
Try reversing the polarity of the primary coil, or the number of turns to see if that makes a difference. Again make sure connections are solid. I have spent over half a day collectively with simple mistakes like these. Even in my video I pointed out in an annotation that the polarity drawn in the picture is wrong! I did not have enough time left over to fix the mistake.
Make sure the tank capacitor is close to 500pF, do not assume that it is if you do not understand how to read values for capacitors. If the value is too low, the circuit will not work very well and will not give you spectacular results. Similarly if it is too high, then again the circuit will not work very well. There are some defining characteristics of the functioning of the circuit that you can really only learn by practice and building/messing with this circuit yourself.
If things are still not working, than it is likely you can killed some parts. I would recommend replacing the MOSFET(s), the DS0026, the NPN transistor, and the 555 with new ones, or testing the parts in question to see if they still work. You may find that the DS0026 and the 555 died, as I did at least 10 times in my development of this circuit. It is really easy to let out the magic smoke, but half the time when chips die, they do not show any physical signs of failure.
Hopefully by now your build is working, and you can achieve the same results as I have! All of the pictures are simply frames of the video, the last one having been modified by overlaying multiple frames to build up the spark density (so it looks more realistic) and reduce noise.
Question 9 months ago
Hey Max i tried to build this circuit and it work but after turn it on and then turn it off and play with it a few time the mosfet keep failing and shorting out (i use an irfp260 mosfet ) i run it with an 24v 15a supply can you help me please :( thank you
Answer 9 months ago
This circuit is not particularly efficient and your FET is under high stress with this circuit, both excessive voltage and excessive power dissipation in the FET. I used a computer heatsink and a 24V/48V switchable 96W power supply.
Also a lot of the common high voltage MOSFETs tend to be counterfeit when purchasing through non official distribution channels (digikey, mouser, jameco, LCSC). My current favorite IRFP260 equivelent is the HY1920W or HY1920P as it has similar ratings and was much cheaper at least for me.
Reply 9 months ago
do you suggest any other circuit? thank you :)
Reply 9 months ago
I have moved on to full-on SSTC and DRSSTC schematics, using a half-bridge. I tried full bridge but really there isn't that much of a difference for small coils when either one is designed to play into the advantages of the half or full bridge (higher current or higher voltage primary, respectively)
3 years ago
Nice circuit! I made it too, but instead of inverted MOSFET driver I used just regular non-inverting one as I found cheap source of these (TC4420CPA to be exact). I also use an class D audio amp as my interrupter with filter removed (so it outputs modulated square wave) and now my coil plays music! I think It's easiest way to get tesla coil playing music, the only drawback is probably the quality, but I'm quite happy with it (especially as I'm mostly playing 8bit songs). Currently I'm powering it from 31VDC 5.1A lab supply, and get not so long sparks (around 7cm). Well I'm more into playing music than sparks, but once I will exchange my IRFP260N to IRFP460 I think I'll finally increase supply voltage to get longer arcs, and louder sound (it's quite loud already actually!).
Reply 1 year ago
Nice! With standard SSTC circuits, it helps to have a fat, short secondary. The field from the primary coil decays quite quickly (inverse square) and with the low drive voltage you need as much coupling as you can get. (SGTC's and DRSSTC's need less because they need the higher Q factor and risk of flash-over with getting the primary too close to secondary)
This particular circuit isn't super scalable because of how inefficient it is. A push-pull configuration works a lot better. I've been messing with HY1920P MOSFETs from LCSC and I can achieve this results:
Reply 1 year ago
That's awesome. Do you think you could make a circuit diagram showing how to implement the non-inverting driver?
Reply 3 years ago
That's pretty damn cool! One of the best recreations of my circuit I've seen yet!
I am trying to improve this circuit by using an H bridge to drive the primary with the primary coil in series with a capacitor. The intent is that when the primary tank circuit's natural frequency is matched to the secondary, then the coil can be driven with quite high power (beyond 100W) and by frequency modulating the H bridge with a PLL, I can audio modulate the output. Frequency modulation allows the coil to continually oscillate maximizing power. However it will be required to de-tune the coil slightly to reduce distortion because the air-core transformer itself performs the demodulation by means of slope detection (due to the high Q factor of the system)
So far I can get really high fidelity audio, but only half inch to 1 inch arcs. I need to change the architecture of my driver a reduce the dead-time of the MOSFET's and improve switching speed, reducing power losses at the cost of higher EMI and voltage transient issues. A gate driver transformer is pretty common on most designs I find online, I might give that a try. My current H bridge works well enough to overheat and fry most TV flyback transformers tho!!! I may make a video featuring an AC flyback operating at 40kHz instead.
2 years ago
Just found this, looks cool. One thought thou, shouldnt there be a resistor at the NPN base pin?
3 years ago
When i tried this circuit i, whenever i turned off the driver, the coil is still resonating and draw the same current as when the driver was turned on, the arcs seems to only shrink a bit but whenever and object come closer to the tesla coil the arcs get bigger, as big as when the driver was on, any idea?
Reply 3 years ago
Is this a problem you are looking to solve? Or just an observation? I am not sure what your question/observation is. If you can post a video of what you are trying to describe that would be helpful! I assume you are asking why the circuit continues to work briefly after powered off? If so,This is normal and expected. With very high Q circuits, (one's with very low losses), they can continue to ring for a substantial amount of time, on the order of several dozen milliseconds. Although more likely what you are observing is that the bulk capacitance on your PSU continues to power the system for several seconds as it discharges. While discharging, the voltage of the capacitor diminishs and this may change the characteristics of the oscillator cause it to oscillate at a different frequency, which ultimately changes the behavior of the output.
Reply 3 years ago
The main problem i was facing is the interruptor failed to turn off the main driver, whenever the main power supply of 12v is being cut off, after further investigation apparently, the driver seems to keep turning on because of the feedback signal, i am planning to interrupt the feedback signal instead of the 12v supply, because that is not working in my case
Reply 3 years ago
Have you tried a pull-down resistor on Vcc of the driver IC?
Reply 3 years ago
Why would I need to add a pulldown resistor? The chip I used draws a substantial current and the voltage at that node will quickly collapse to 0V as soon as the transistor is no longer turned on with the base pulled to 0V.
3 years ago
Hello,i'm building your tesla coil and i have one question,i didn't find the DS0026, do you now otherone that work properly???
Reply 3 years ago
Just do some research and find an alternative part. I only used that inverting mosfet gate driver IC because I had it on-hand.
Question 5 years ago
Why do you use an inverted buffer instead of just a regular buffer?
Answer 5 years ago
Because it was the convenient MOSFET driver chip I had available. I do not recommend anyone building this project to use this obsolete part. Its hard to source and expensive. You can replace it with a push-pull arrangement of BJTs which performs essentially the same function. (It amplifies current greatly and makes the impedance of the MOSFET gate appear 1+ beta times bigger. (Beta being the current gain of the transistors, which sbould be about 50 to 300)
Reply 3 years ago
Do you have a circuit of this?
Question 3 years ago
Will this circuit work and allow me to modulate the tesla coil with an audio signal? I'm afraid the optocoupler might not be powerful enough to switch voltage to the mosfet driver.