Tesla Coil Build Thread

[Sigurthr]

Hey all, it's been a few years and this thread sure has grown! It's to the point now that information is a bit hard to find, so I wanted to retouch this post and put up some eternal links to/and important data. Also, I apologize for any dead links and images throughout the thread, especially on older pages. I used imageshack back then and when they went paid they wiped free user's content. Those of you who are new to the topic, feel free to post any questions or content, we won't hold it against you if you don't search through the thread manually.

SEE PAGE 99 FOR THE BEGINNING OF A NEW QUESTION & ANSWER SECTION!​

Sig's SSTC Logic Controller: The USSTCC http://sigurthrenterprises.blogspot.com/2015/01/the-usstcc-universal-sstc-logic.html
On this page is the download link to my master data file for SSTC building. It includes complete schematics, parts lists, and plans for building a fully functional SSTC even without my board.

Sig's DRSSTC Burst Calculator: http://sigurthrenterprises.blogspot.com/2015/01/sigurthrs-drsstc-calculator.html

A Treatise I wrote on Audio Modulation of SSTCs: http://sigurthrenterprises.blogspot.com/2014/02/a-treatise-on-audio-modulation-of-solid.html

A short video on how to phase a GDT without using a scope: http://sigurthrenterprises.blogspot.com/2014/01/how-to-phase-your-gate-drive.html

A simple digitally controlled interrupter for SSTCs and DRSSTCs: http://sigurthrenterprises.blogspot.com/2014/01/sigs-attiny85-sstc-interrupter.html

----------------------------------------------------------------------------------------------------------------------------------------​

Original First Post:


Hey all,
I've noticed we're lacking a central thread for TC's, not that there have been many lone threads so far (just a few throughout the forum's history), and figured I'd throw one together.

I'm currently working on my first Tesla Coil. Today I completed the NST chokes, main base, and the MMC Tank Capacitor array. Tomorrow I'm expecting my secondary and spark gap to arrive in the mail. I went with a pre-wound secondary because I don't have the space or tools to wind one myself without spending a dozen hours hand winding the coil directly on the form with no support/roller guide etc. Also, I can't get coilforms locally, lol. We have no real hardware store up here.

Pics:

You can see it dwarfs my work table and takes up half of the night tables I use as well.


Stats: (will be updated as I go along)
NST: 7.5kV 30mA nonGFCI
NST protection: 100T air-core 18AWG Chokes
Primary "Tank" Capacitor: 0.01764uF (17.6nF) 16kV MMC (24x Cornell Dubilier 942C20S47K-F 0.047uF 2kV Polypropylene Foil Caps)
Spark Gap: machined brass and steel
MMC Insulator/stand off: 1/4" Plexiglass
Primary Coil: TBD
Secondary Coil: TBD
Topload: 7.4" diameter ball of aluminium foil
Base: 3/4" CDX Plywood.
Hardware: 1/4-20 2" Bolts and Nuts

I'll be using the Series Capacitor / Parallel Spark Gap topology with a safety gap across the Tank cap. I don't plan on using Bleeder Resistors just yet, once I get the coil working well I'll add that in later if needed. They're just a safety precaution really and if you treat it as a lethal pulse cap there is no need for bleeders.

 

[Hiemal]

Oh, remember to keep the coil away from electronics. The spark gap will more or less be a constant EMP and will definitely make some things act funny...

Also, you know you have to tune the coil, correct? Otherwise you'll be off your resonant frequency and won't get the best sparks. o:

 

[Sigurthr]

Roger that. I ran the dedicated RF ground for it today (easy for a Ham) and I'm setting it up in the bedroom - only thing in there is an alarm clock.

Aye aye, I'm going with adjusting the tap on the primary for the tuning method. I'm not going to bother with other methods for tuning, I ran JAVATC to find the optimum topload. The hardest part will be adjusting the coupling factor though. I'm not quite sure whether to tune the inductance of the primary first and then adjust coupling or vise versa.

 

[Hiemal]

I wanna say spark gap tesla coils like loose coupling, but don't hold me to that one...

 

[Sigurthr]

Aye I read 0.05 to 0.2 coupling factor somewhere...

The secondary I got comes with a primary but it is like 15 turns of 14AWG stranded on a helical form a few centimeters larger than the secondary form... that's like 0.5 coupling factor! Way way too much, and far too many turns. They had it set up all wrong but the parts are good/brand new. They were using a 1.6nF cap for a 9/30 NST (WAY TOO SMALL, a 9/30 needs a 15nF cap) and a 2" topload (>8" needed) with the highly coupled and too many turns primary. Their pictures showed like 1.25" breakout, haha. I mostly bought it because the price I paid is less than a secondary from EVR + equivalent spark gap (from wherever) would cost. I get a 1.6nF 20kV cap for the junkbox and a bunch of 14AWG wire out of it (if I don't reuse the primary) for free. Of course I could make a secondary for cheaper, but again with no room or tools or supplies what is the point. In the future I might just order a secondary to my specs from EVR (but that takes 2 months production time) or order the stuff to make one online for any future coils.

I haven't really put much thought in to the primary yet to be honest, it's the easiest part to make. I have a couple feet (5?) of 12AWG solid copper wire that I ripped out of the walls when I ran new line (this stuff was CLOTH insulated!) which I can probably use for a nice self-supporting flat spiral. I also have a couple dozen feet of 18AWG solid enameled, and also a couple hundred feet of bare 14AWG stranded antenna wire if needed. Lots of choices. I want to get the precise measurements of the Secondary and re run it through JAVATC and TeslaMap before I muck about with the primary.

I'd just use 0.25" copper tubing if I could get my hands on it locally! Not paying shipping on that much copper.

I've been thinking that once I find full time work I may make my own SGTC Kits and sell them online for spare cash. I've always wanted to do it and now I have the know-how to actually pull it off. Counting chick's before the eggs though at this point, we'll see how this coil goes!

Also, I have an antenna analyzer, so if I ever do HF Tesla Coils (mini ones) I can just use that to find the resonant frequency and do all the tuning that way. Though... it is equally easy with a scope and there is no frequency limit then and I need to get a scope anyway.

 

[Pilgor]

The gap also produces a lot of UV light, so keep it covered by something.

 

[Sigurthr]

<- TIG Welder. UV scares me not (because I do not underestimate it). The thought is appreciated though =).

I may eventually build a faraday cage for it, but I can't find metal screening or chicken wire locally. Sometimes I hate living in the middle of no where.

 

[Seoul_lasers]

I'd be going Terry filter for the NST protection circuit. The MOVs can be removed.
The sparkgap to ground --0 o 0-- is essential.


I'd also be careful on the spacing of those caps. Make sure you've got at least 2cm between your rows on the MMC bank. Otherwise, you could have lead to lead interarcing!
You'd be surprised how much inductive kickback there can be on a TC circuit like this..

 

[Fiddy]

Oh, remember to keep the coil away from electronics. The spark gap will more or less be a constant EMP and will definitely make some things act funny...

Too right, operating my laptop as a audio source for my plasma speaker, it was ~2m away from the breakout, lost control of a couple of keys on the keyboard, had to reboot to reset them.

 

[Hiemal]

Ah, I have a really small tesla coil secondary. I think the resonant frequency of it is 2.56 mhz if I remember correctly.

Yeah, don't try to run that using semiconductors... it's just too fast for them and they start to run a lot into their linear regions and get very unhappy.

Tubes are much more adapted to high frequencies. Just gotta figure out how to make a 2.56 mhz tube coil.

Edit;

Yeah, Fiddy, that happens. When the spark gap is ignited a very powerful electromagnetic field is emitted outwards, and is induced into the millions of wires inside of your electronics...so they see some strange signals, and start to freak out as a result. Laptops are usually pretty well protected from this (keyword usually) due to RFI shielding.

I've also noticed that when using a plasma speaker with an AC output (AC flyback transformer... don't worry about this Sig, the flyback in yours is DC ) my iPhone would start to flip out. Induced currents are the worst.

 

[Sigurthr]

Terry Filters are a major expense though, out of my budget right now. I'd be more concerned about HV kickback or RF passthrough if I were using the parallel cap / series spark gap topology (which I'm not). From what I read when using the parallel spark gap/ series cap setup the two states are:

1) spark gap "off" - cap is charging, the only RF the NST can see is the reflected Secondary Ringdown, which with a low coupling factor and topload breakout losses/loading should be a very damped low current sine since it takes two transfers to get back to the NST and has to pass through the series cap which should resist rapid voltage changes.

2) spark gap "ON" - NST essentially shorted, cap discharging through primary. NST would see very little if any current/voltage reflection since the cap (acting power source) is in series with the load (primary) so the voltage is divided between the NST (now appearing as a load if your spark gap fires at above NST Vpk - worst case scenario) and current divided between it (and it has a high impedance - 250K ohms) and the primary (low impedance) and the spark gap (low impedance), which should not be much.

So in theory all you need is adequate choking to block out the reflected secondary ringdown which could hit when the spark gap is off and to make sure your spark gap is not set to fire at voltages above NST Vpk (Voc).

Again, this is all based on the reading I've done. I don't double Terry's expertise or experience and if he thought he needed to design a low pass filter for it, well then I'm sure there is a case where it is absolutely needed.

edit: Jared, I've seen some 4MHz SSTCs! Talk about getting quite high up there in frequency in relation to the semiconductors! Though tbh all you'd need to do is impedance match an 80m HF amp to the primary resonant circuit and you could forget the rest of the electronics. Key down an 80m CW transmitter and you've got a perfect sine wave at up to 1500W to drive your coil!

edit 2: I did a little more research on the Terry Filter and it seems the bulk of it (safety gap, MOVs, high power resistors) are primarily there to protect the NST from a primary strike (secondary voltage strikes the primary) and the parallel caps are mostly there to act as a low pass filter to shunt RF to ground. More reading yields that the RFI filering is to keep the RF out of the mains wiring, and not really to protect the NST. This makes sense as I had wondered how low current, low voltage RFI would damage the nst. Primary Strikes are indeed another matter all together... I will be using a strike rail and I've even considered protecting the primary with a mini faraday cage. Also, I'm not expecting streamers to exceed secondary height in length, so with a breakout point on top of the topload I would be very surprised if I fry my NST with a primary strike.

I'll note it for future builds though to include a mains RFI filter at the very least . I wonder though... adding parallel capacitance like in the Terry Filter will likely throw off the impedance matching of the primary circuit as the 60Hz will see FAR more capacitance than the RF reflection will. Wouldn't you have to adjust your MMC if you use a Terry Filter?

 

[Seoul_lasers]

The gap also produces a lot of UV light, so keep it covered by something.

That would be in theory a good idea, but with the Ozone and heat created, you'll need as much air as possible blowing through the gap. A fast ducted fan blowing air through your gap should be enough to assist quench the gap well enough to provide excellent performance. DO NOT.. I repeat DO NOT USE A ROTARY spark gap on an NST based coil. doing so will cause enough stress on the NST secondary windings to lead to a failure.

Actually the terry filter is fairly cheap.. The MOVs can be safely removed as the sparkgap construction is designed to ground out anything above 20Kv kicking back to the NST side. I just used 2 10Kv ceramics 1nF each and 2 ceramic wireround 100ohm resistors. It worked well. The MOVs just provide that extra voltage protection. It's not totally needed. Ceramic caps =$4ea + the resistors $15 ea.

 

[Hiemal]

Terry Filters are a major expense though, out of my budget right now. I'd be more concerned about HV kickback or RF passthrough if I were using the parallel cap / series spark gap topology (which I'm not). From what I read when using the parallel spark gap/ series cap setup the two states are:

1) spark gap "off" - cap is charging, the only RF the NST can see is the reflected Secondary Ringdown, which with a low coupling factor and topload breakout losses/loading should be a very damped low current sine since it takes two transfers to get back to the NST and has to pass through the series cap which should resist rapid voltage changes.

2) spark gap "ON" - NST essentially shorted, cap discharging through primary. NST would see very little if any current/voltage reflection since the cap (acting power source) is in series with the load (primary) so the voltage is divided between the NST (now appearing as a load if your spark gap fires at above NST Vpk - worst case scenario) and current divided between it (and it has a high impedance - 250K ohms) and the primary (low impedance) and the spark gap (low impedance), which should not be much.

So in theory all you need is adequate choking to block out the reflected secondary ringdown which could hit when the spark gap is off and to make sure your spark gap is not set to fire at voltages above NST Vpk (Voc).

Again, this is all based on the reading I've done. I don't double Terry's expertise or experience and if he thought he needed to design a low pass filter for it, well then I'm sure there is a case where it is absolutely needed.

edit:

Jared, I've seen some 4MHz SSTCs! Talk about getting quite high up there in frequency in relation to the semiconductors! Though tbh all you'd need to do is impedance match an 80m HF amp to the primary resonant circuit and you could forget the rest of the electronics. Key down an 80m CW transmitter and you've got a perfect sine wave at up to 1500W to drive your coil!

Oh, semiconductors can definitely do it, it's just a ridiculous pain in the arse to get everything working properly. Parasitic capacitances, inductances, and all sorts of other stuff make working with HF circuits a pain in the butt.

 

[TacoHerder]

I envy how smart you all are and know all this cool stuff. Tesla coils are amazing, especially when they are setup to play music:beer:

 

[Seoul_lasers]

I envy how smart you all and know all this cool stuff. Tesla coils are amazing, especially when they are setup to play music:beer:

Sure are, and to build one of the Solid State coils you're looking at close to $1000 for a small 1KW sized one. multiply by 4 for those larger ones at least. You need some serious equipment for the professional SSTCs. Also in order to use a computer for input like midi, those large coils need to use SPDIF optical lines due to the massive EMF fields they produce.

Classic TC's like the one shown here are still just as impressive, just that they can't play Supermario Brothers or a StarWArs theme...

 

[Sigurthr]

Actually the terry filter is fairly cheap.. The MOVs can be safely removed as the sparkgap construction is designed to ground out anything above 20Kv kicking back to the NST side. I just used 2 10Kv ceramics 1nF each and 2 ceramic wireround 100ohm resistors. It worked well. The MOVs just provide that extra voltage protection. It's not totally needed. Ceramic caps =$4ea + the resistors $15 ea.

Hmm, that is WAY cheaper than I originally found for 100W 1K resistors, where did you find them?

I'm not certain but I think the main spark gap I have coming is a [O o O] grounded center design. Also I have a safety gap across the MMC (Just for the heck of it, you can see it without terminals in the photo) and will NOT be using a rotary gap (if it fails or misses, resonant rise WILL kill your transformer AND mmc).