Continue to Site

Welcome to our site!

Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

  • Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

Question concerning choice of NST for Spark Gap Tesla Coil

Not open for further replies.
I have been designing a spark gap tesla coil, so far trying to use what I have to save cost. I currently have a solid-state NST and I have read that they don't perform properly for Tesla coil applications (I imagine sudden current draws create back emf that is bad for the transistors?). I have disabled overvoltage protection by cutting one of the wires as shown. Ultimately I would be curious as to what problems arise with using a solid state NST with a SGTC before I look into getting a proper NST.
NST photo.jpg
From what I know of why they aren't recomended is because they are solid state switching supplies. Where as the old type was a linear supply.
I'm not sure what you mean as linear supply for the old NSTs as those are all transformers with a shunt put in to limit current. However the fact that the new NSTs are switching power supplies would mean that the waveform would have plenty of messy harmonics, I don't know if that is the main culprit. I would wonder if that could be compensated for by using some sort of band pass filter to have a truly sinusoidal waveform?
I think the two basic factors are:
A standard transformer has no fast protection, while any well-designed electronic equivalent will have protection circuitry.

The frequency - tens or hundreds of KHz for an electronic one vs. 50 / 60Hz for a basic transformer.

Looking at some schematics for tesla coils using neon transformers, the capacitor bank has to charge to peak voltage within the time of one half cycle as there is no rectification. It also has the primary in series with the capacitors, adding inductance which also limits the charge rate.

Once the spark finishes, the primary is open circuit again so not damping the secondary.

I just do not see how that can work with a high frequency electronic transformer; I think you would need to add high voltage, high speed rectifiers and let the cap bank charge over a number of cycles until the voltage builds to where it can bridge the spark gap?

Some of the schematics I found:

Or this, which could work with conventional or electronic transformers as it uses rectification.
Thanks for that assessment. I've found a deal on a traditional NST but once I've finished the typical setup I'll look into a HV rectified setup with the solid state NST and see how well that works. Removing the overvoltage stuff may have been sufficient for removing protections (NST continues to operate when shorted) but I guess I'll see.
you could also look at designs for spark gap transmitters, as the tesla coil primary circuit and spark gap transmitters have a lot in common. for instance, feeding high voltage DC to the primary tuning capacitor, and having the capacitor discharge through the primary winding across a rotary spark gap is a very common method to excite a tesla coil. you can get some good info from [here]. be aware some old schematics use zigzags for coils, so some schematics may look like there are resistors when they are actually inductors and transformers.
Build an RQ variable speed vacuum fan spark gap = RQVSVFSG. Put 9 copper pipes 3/4" x 2" long inside a 4" pvc pipe 6" long. Gap between each copper pipe is .030". Drill clearance holes in copper pipe 1" apart for 3/4" long 6-32 screws. Drill 1/4" holes in pvc pipe for the screws the over size holes allows you to adjust spark gap to .030" very easy. Buy a $3 vacuum cleaner from yard sale or goodwill store remove electric not with fan. Mount vacuum motor on 1 end of 4" pvc pipe. Put 4" pvc pipe in a pvc toilet flange to block air flow through pvc pipe and toilet flange acts like legs to hold spark up right. No air can flow through pipe so you need to drill 3/16" diameter holes in rows 2" long in the side of the pvc pipe so air is sucked in threw the holes at high speed to break all 8 spark gaps. My Harddrive crashed years ago I lots all my good pictures. You also need a variac to adjust the speed of the vacuum fan.

As you turn the variac dial vacuum motor turns faster and faster. As air is sucked into over the sparks gaps the gaps all break at higher and higher frequency. As vacuum motor picks up speed out put arcs and sparks of your tesla coil get longer and longer. What you have is a variable frequency spark gap controlled by the variac. I have used this on my 14KV 30ma TC and also my 14KW TC.

The guy that mastered rotory 20 years ago said, WHY would anyone want a rotor when you can have a RQVSVFSG. For some reason this never caught on everyone had rotor on the brain. A few people on the old TC forum built a RQVSVFSG and said the same thing, why would anyone want a rotor when you can have a RQVSVFSG. It still never caught on and I gave up telling people about it.

RQVSVFSG is so simple turn the variac dial it takes 5 seconds to adjust to maximum tesla coil output. Make changes to your TC turn the dial you get max output in a few seconds just turn the variace dial. Move the RQVSVFSG to other Tesla coils it takes 5 seconds to dial in maximum output.

You can see my RQVSVFSG in the first picture setting on the table next to the left TC. I had alligator clips on the RQVSVFSG I could move it to a different TC in a few seconds. I use to have 7 TCs all sizes from 6K 30ma to 14KW. I build all my caps to be rated 40,000. I wanted everything to be bullet proof and work perfect.

My big TC is 10" diameter it runs best at 12KW and produces 12 foot long arcs from the 3 ft diameter sphere. I can make a 27 ft diameter circle or spark & arcs in the back yard.

Flat wound primary with strike rail works best.

950 turns on the secondary works best.

Secondary diameter to length ratio 5 to 1 is best.

I tried, tested, experimented with everything.

I still have 2 rolls of #24 enamel coated copper wire each roll weight is about 80 lbs. Sold all my coils, spark gags, caps, other wire. It was FUN for many years.

I built double layer and triple layer secondary coils. Wind 1 coil 950 turns test it for max output then wind another 950 turn secondary over the first secondary connect the 2 secondary coils in parallel test it out put is about 5% better than before. 3 secondary coils over top each other gets better too not much about 7% better than 1 secondary.







HV = 14KV

Last edited:
I tried one initial setup with a rectified input from the switching power supply and a dc motor spinning a rotary spark gap. The first firing, there was a failure in insulating some of the capacitors, so I immersed them in oil. From there I tried it again and I didn't see any apparent shorts but the secondary was showing no excitation. So I got my oscilloscope and I was observing oscillations approximately 30% off of the targeted frequency 420khz vs 330khz So I imagine some of the capacitors were burnt through (Later it now appears the fault has worsened, it doesn't appear to be resonating properly). So I am looking into setting up a new capacitor bank.

Before I had gone with the cheaper high voltage ceramic capacitors out of reducing cost, though some have said they are considered lossy at high frequencies. I was wondering whether this is enough to warrant concern (I wouldn't imagine I would leave it on long enough to put in excessive heat)?
-Shipping will take some time, so while I wait, I'll try out homemade "beer bottle" setups and maybe look into 3D printed caps

I am currently looking into finding some MOTs for a separate/more traditional setup as NSTs are hard to find at cheap prices (hard to convince anyone a neon sign shop is "essential business").

This is an amazing setup, that topload is gargantuan.

I would be curious as to how you provided an RF ground for those circuits. So far I've just been setting up a large counterpoise on the ground with sheets of aluminum and whatever large pieces of metal were lying around.

If I get a proper HV AC supply, I'm sure I could find a fan lying around somewhere to advance my spark gap setup, though I lack a variac for proper tuning. I have a motor controller setup for the dc motor but I imagine phase is likely most important here?
Not open for further replies.

Latest threads

New Articles From Microcontroller Tips