Re: Royer SSTC
GrantX, Sun Mar 29 2015, 04:08AM

The typical ZVS oscillator circuit should be capable of switching at 500kHz, but I'm not sure how stable it would be. The IRFP260Ns are certainly fast enough though.

The frequency that the ZVS oscillates at is determined by the values of the tank capacitor and the center-tapped primary coil, so it will need to be tuned in a similar fashion to a spark-gap coil. Its easiest to tune a SGTC by winding a primary with many taps and using a fixed value HV capacitor. For the ZVS oscillator it would be much easier to wind a fixed value center-tapped primary and vary the number of capacitors in the MMC/tank. A bunch of cheap box-type polypropylene/MKP capacitors would be ideal.

Re: Royer SSTC
Teodor Zafiroski, Sun Mar 29 2015, 08:18AM

I was mainly concerned that the mosfets won't switch fast enough with 400Ohms per gate, I plan to use two of the 200Ohm resistors in series per gate, so a 400Ohm of gate resistance...
Another question, does the toroid look good enough, I have read that the pale green/blue toroids are iron powder ones and have a relative magnetic permeabilllity of around 75, is that correct? If it is, we're looking at ~110uH of inductance...
As for the caps, I plan to use FKP caps, mainly because of their high current rating and that they are readily available (in my junkbox).
As for the primary, what is the minimum number of turns I need to wind, also, I could lower the resonant frequuency of the secondary witha topload, no problem with that, but some folks have run Royer oscillators in the MHz, specifficaly the Mazzilli ZVS driver using RF mosfets. Around 6.5MHz to be exact, check it out!

Re: Royer SSTC
GrantX, Sun Mar 29 2015, 03:31PM

Thats a pretty neat little lamp. I'm currently building a desklamp using a big old 1970's xenon strobe designed for airport runways, but I've taken the easy way out by using a small 50Hz 12kV transformer.

As for the gate resistors, the gate capacitance and its series resistor can be seen as an RC circuit with a time constant. You can calculate how long the gate will take to charge and as long as the FETs don't spend more than approx 10-15% of each cycle in the linear region they shouldn't get too hot. Sorry, I'm currently a bit too drunk and tired to help you out with the math, most of it would probably be wrong haha. Online calculators should be good enough though.

The toroidal choke in series with the positive DC rail is fairly non-critical. A powdered iron core that can handle the amp-turns without saturating is all you need. Your core should be ok, depending on how much input current you're aiming for.

At 560kHz your primary wont need many turns. 1+1 to 4+4 turn center-tapped coil will be the most likely range, depending on the diameter, how many caps you want to used, desired input voltage and curent draw.

Hope that helps, sorry I couldn't get into specifics, starting to go to sleep in my computer chair haha.

Re: Royer SSTC
Teodor Zafiroski, Sun Mar 29 2015, 05:40PM

Thanks for the info, sure it helps, but wouln't 4+4 turns draw a lot of current? I aim to minimize current draw, wouldn't increasing the primary turns decrease current draw? Also, could I use a computer psu with the 12V rail to supply this circuit?

Re: Royer SSTC
GrantX, Mon Mar 30 2015, 04:21AM

Yes, more turns will result in a lower current draw. Impedance of an inductor increases with frequency. At ~30kHz a 4+4 turn primary (8 turns total) wrapped on the ferrite core of a flyback may result in a couple of hundred watts (ballpark figures). At 560kHz an air-core 4+4 turn primary shouldn't draw an extreme current. More turns means less capacitance required, and vice versa.

The PC supply might be good enough, it depends on how many amps the 12V rail can handle. Theres usually a sticker stating the rated wattage and output currents. You really want a supply that won't sag below 10V at full load, or your FETs will latch and explode. Test your supply before use by loading it to its maximum rating and measure the voltage drop across the dummy load.

EDIT: You mentioned minimizing current draw, do you have a specific power level you want to achieve? You will probably want at least 75-100W to produce some healthy plasma.

Re: Royer SSTC
Teodor Zafiroski, Mon Mar 30 2015, 01:25PM

I am aiming at around 200W power consumption, but you can see, in the "You Tesla Coil" thread I produced some "healthy" plasma with just 25W, (3cm to air, 7cm to ground) but that was a totally different design, about 200W would be nice...
Also, I have two 22nF 1kV WIMA FKP-1 caps and two 1nF 1,6kV WIMA FKP-1 caps, I can combine them to form various capacitances...

Re: Royer SSTC
GrantX, Tue Mar 31 2015, 04:31AM

Your relay-switched coil is very nice looking, I love the design of the primary and secondary assembly. Interrupted coils like your relay design, spark gaps and DRSSTCs can produce very long thin streamers. My old spark gap coil was making almost 20cm open-air sparks with an input of about 200VA, as seen here:

The ZVS/Royer oscillator will be similar to other CW SSTCs. At 200W you will get a very hot silent plasma discharge several cm long. Make sure to post some photos and videos when its finished, should be an interesting build.

Re: Royer SSTC
Teodor Zafiroski, Tue Mar 31 2015, 03:17PM

Because of the small heatsink, maybe I should try a 555 timer circuit to turn on transistors that would ground the gates, that would look like the interrupted SSTC's, but I need to make it work first...
I will surely post photos, without a doubt, first, a couple of questions...
-Will a higher primary to secondary coupling coefficient work better, I want to figure out the diameter of the primary I will be making.
-I have the secondary with a long steel wire as the discharge terminal, what would be the capacitance of that piece of wire, it has 1mm diameter and is 180mm long. A couple of pF maybe? I need an approximation of the capacitance so I could calculate the resonant frequency. An update, this coil has a self-resonant frequency a bit lower than the previous mentioned 560kHz, due to that the real world differs from an ideal one, in which the calculations were made and due to extra space on the form that I wanted to finish off...
Here is a calculation of the coil itself and a picture of it:

Re: Royer SSTC
Teodor Zafiroski, Sat Apr 04 2015, 11:49AM

UPDATE: I have searched for a power supply that can give that output current at such voltage without sagging, and was unable to find any within my budget, but I have picked up one 12V 45Ah Lead-Acid battery from a car, it's not new but it hasn't been used a lot, only a couple of months. Could I use it as the power source?

Re: Royer SSTC
Mads Barnkob, Sat Apr 04 2015, 04:18PM


You would need two batteries in series, 12V is simply too low to drive the MOSFETs properly. You will not switch fast enough, thus spending almost the entire time in the linear region, dissipate so much heat that the MOSFET will eventually fail from heat or it will all fail from slow switching, both gates staying on, not getting switch completely off, etc.

Re: Royer SSTC
Teodor Zafiroski, Sat Apr 04 2015, 04:36PM

So why did I bother searching for a computer PSU in the first place? In the meantime I found another large lead acid, and It has almost the same characteristics, except that it is old.
My secondary diameter is 50mm, about how large primary diameter should I use, I'm going to wind it with 1.7mm copper wire...

Re: Royer SSTC
Sigurthr, Sun Apr 05 2015, 04:19AM


Why would 12V not be enough to switch a simple NFet? As long as there is adequate current supply to the gate, 12V is plenty. This isn't a DRSSTC with IGBTs where you benefit from overdriving the gates. Most fet gates reach saturation at under 8V anyway, and beyond that is just charging the miller capacitance.

My 3kW CW SSTC runs +-12V Vgs with no problem, and no signs of running in the linear region. I've used the same gate drive on coils up to 1MHz at lower continuous power levels too, all with only 2.5A of peak gate current (5ohm Rg).

Re: Royer SSTC
Mads Barnkob, Sun Apr 05 2015, 05:00AM

Re: Royer SSTC
Teodor Zafiroski, Sun Apr 05 2015, 11:08AM

Mads, three guys at thе macedonian forum for electronics have made an induction heater using the schematic at RMCybernetics, it is a simplified mazzilli driver, it has everything except the zeners and the resistor parralel to the zeners, and they've succeeded in that.

What could happen if I was to connect this circuit to a 12V car battery? Тhe gate voltages are limited to 12V with the zeners anyway...

Also, when using such a high current power source like a car battery, I am very afraid of a short circuit making the wires to vaporize and exploding the MOSFET's to pieces, could I connect a regular 15A house breaker to protect against short circuit?

Re: Royer SSTC
Sigurthr, Sun Apr 05 2015, 09:30PM

@Mads, ahh if sag is expected, yes, you're right. You'll get lockup in a traditional Royer. I wouldn't expect sag from a fully charged deep cycle SLA cell though.

I have a simple royer ZVS Flyback driver that I use on a high current supply or SLA, but it only draws 24A at full load on 13.8V.

Re: Royer SSTC
Teodor Zafiroski, Mon Apr 06 2015, 08:01PM

It's not even an SLA, it's the type where you can fill/refill acid and it has some sort of plastic screws so the acid can be refilled.
Can someone explain to me the problem with sag? I haven't met this term when talking about batteries...

Re: Royer SSTC
dexter, Mon Apr 06 2015, 09:01PM


if the voltage sags below 10V the mosfets will not fully turn on so it will dissipate alot of heat
in the case of a ZVS this also cause both mosfets to be partially on further worsen the situation
eventually (actually quite rapidly) one will pop

but for a big battery this isn't a problem

i ran mine from a 12V 4.5AH battery and i think is not even fully charged and i get no heat on mosfets
i use IRFP260N for which the datasheet suggest Vgs +/- 20V

Re: Royer SSTC
Sigurthr, Mon Apr 06 2015, 11:25PM

Voltage sag is when a power source cannot supply the current required by the load without the voltage falling to a lower than ideal level. For example, a 12V 1A power adapter for the wall outlet is designed to supply at most 1A at 12V, but at 1.2A it might only be capable of 10V, at 2A only 8V, etc. (and that's a rather nonlinear nonsense load curve of pseudo-random numbers to explain the concept, by the way).

Re: Royer SSTC
Teodor Zafiroski, Tue Apr 07 2015, 01:38PM

Mine is 12V 45Ah, well above yours which was 4.5Ah, so I shouldn't expect sag...
I use the IRFP260N's too, seems that it could work together nicely!
The only thing left to do is the primary, which I will wind with a couple of cm larger diameter than the primary, I just need to calculate the inductance and test it, I have some exams this week, but I will try to test this as soon as possible.

Re: Royer SSTC
Teodor Zafiroski, Sun Apr 12 2015, 07:30PM

I was able to fire up this circuit today, and got some incredibly small output, the circuit draws just over 500mA at 12V, I think that's because of the 3.5+3.5 turns on the primary, I need to reduce it to about 1.5+1.5 when I connect all of the caps in parallel, that should give me a greater power draw... Here's the schematic:


And here's a video of it in operation:

When I get my hand closer to the coil, the extra capacitance pulls it in tune and the tube lights up, the spark to my fingers is so hot that my fingers began smoking, the burns don't hurt though...

I need to reduce the primary turns, connect the tank caps in parallel instead of series, and that should result in larger sparks, and hopefully- streamers!
P.S. When I place the link to the youtube video in a youtube tag, the video won't load.

Re: Royer SSTC
hen918, Sun Apr 12 2015, 07:33PM

I think it would be a good idea to use a chicken stick and not put your fingers that close...

Re: Royer SSTC
Teodor Zafiroski, Sun Apr 12 2015, 07:35PM

We're talking about really low powers here... I use a chicken stick whenever I work with higher powers...

Re: Royer SSTC
dexter, Mon Apr 13 2015, 10:23AM

i tried this setup too but 2+2 turns and at 36V from a 150-200VA transformer and got 10mm streamers output
the primary was not in tune with the secondary ~200KHz primary 250KHz secondary
the mosfets got wery hot after a few second of operation

Re: Royer SSTC
Teodor Zafiroski, Mon Apr 13 2015, 11:44AM

2+2 turns at 36V? That explains a good part of the heating, and at that frequency with that much turns and at that voltage I expect the current draw to be large, have you measured it?

Re: Royer SSTC
dexter, Mon Apr 13 2015, 03:28PM

did no current measurements because i just slapped the parts together to see how it performs....

i used the same Royer to overdrive flybacks at 36V about 30A without much heating of the mosfets but the frequency was much lower 30-40KHz

Re: Royer SSTC
Sigurthr, Mon Apr 13 2015, 10:25PM

Unexpected or excessive FET heating almost always comes down to one or more of three causes:

1) Running in the Linear region from insufficient or incorrect gate drive.

2) Shoot Through / Crossconduction from inadequate gate drive dead time.

3) Avalanche Conduction from exceeding BVds. Inductive kicks on turn off of each leg can cause massive heating effects without instantly destroying the device. As such it often gets ruled out as a culprit unless measurements are done to watch the Vds and ensure no spikes are seen.

Re: Royer SSTC
hen918, Tue Apr 14 2015, 04:49PM


Or it could be due to the MOSFET being a re-named cheaper, bad quality or otherwise fake Chinese copy.
but, if bought from a reputable supplier, it is usually safe to assume that this is not the problem.

Re: Royer SSTC
dexter, Tue Apr 14 2015, 05:10PM


is not a fake mosfet i used the same setup (see picture few posts above) at 36V ~30A on a flyback, the power resistors and the windings got hot before the mosfets :)
IRFP260M's bought from a respected site

Re: Royer SSTC
Teodor Zafiroski, Wed Apr 22 2015, 05:36PM

I calculated the number of turns of the primary for the capacitance that I had, and tested it today. I got a small brush discharge, followed by a click when I made an arc to ground and the arc was gone. The breaker turned off, and when I turned it on again it immediately turned off every time. The MOSFET's were warm to the touch, and the inductor and tank caps as well. When I removed the MOSFET's from the circuit, the measurement had shown that they had failed all pins shorted. (one of them had 500 Ohm resistance gate to source, the other one had zero resistance gate to source, but they both were shorted drain to source)
I guess they became another addition to my silicon graveyard...
What may have blown them?

Re: Royer SSTC
dexter, Wed Apr 22 2015, 06:33PM

might be a dumb question but did your mosfets were isolated from the heatsink?

Re: Royer SSTC
Teodor Zafiroski, Wed Apr 22 2015, 07:56PM

Yes, of course they were... Why did it produce a discharge in the first place...