Tesla Coil Build Thread

[Hiemal]

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...

$1000?? What the hell, are you using gold plated mosfets or something?

I know that you can build a decent sized SSTC for a lot cheaper than that. The parts aren't THAT expensive, and you also get a plus of smaller size (less magnet wire to wind, therefore less expensive) and no tank cap for the primary side so that wipes a good chunk of change out too.

Just use a PLL controller, and you can play actual music music too, like how Fiddy's plasma speaker works.



Or, you may be referring to a DRSSTC (dual resonant solid state telsa coil)

Those are beasts. It's basically a SSTC with a primary capacitor. You actually have to interrupt the half bridge in it otherwise it would destroy itself from overcurrent and overvoltage due to resonant rise.

The arcs on these are insane. They're also expensive, and yes, would cost a good chunk of change for one... you have to use massive mosfet or IGBT bricks and heatsinks for these. You also need a nice bank of capacitors for the primary. And I mean a NICE tank. Otherwise they go boom boom.

 

[TacoHerder]

These are the guys im talking about. Now im pretty sure they are way out of my price range. Just think if someone thought they could walk in between those...
"Sweet Home Alabama" - Musical Tesla Coils - YouTube

 

[Sigurthr]

Yeah, those are DRSSTCs. Monsters in their own league.

I find the small size factor and high frequency of SSTCs appealing, and the charm of the SGTC (it is what Nikola himself used, after all). DRSSTCs are amazing but I can't see myself having enough money or space to actually build/own one.

Btw, you can build a very small DRSSTC or 4MHz Audio Modulated SSTC for only $300 from EasternVoltageResearch. A simple SSTC (800kHz, 2" streamers) will run you only $150. Or you can build a nice size SGTC for just under $250 (including MMC, NST, prewound Secondary, and machined spark gap). My SGTC is coming in at just under $300 so far, but I had to buy a lot of materials and hardware because I didn't have any (living in an apartment). I sold one of my unused flutes for $200 though so this project is actually only costing me $100.

Update: I was able to find Terry Fritz's papers on NST stresses and it seems that inductive filtering causes more problems than it solves when it comes to NST protection. It will effectively block the higher frequency oscillations but creates it's own oscillation at it's self resonant point, which is equally as damaging. Wish I found this before I hand wound 260 turns of chokes. Oh well, parts bin they go. His research also shows that RC filtering DOES has an effect on the rest of the circuit, but protecting your NST is usually worth it. So, I'll build my TC as planned, and just save up for a RC Filter for it before giving it any serious run time. Also, it has been shown that a 1/4" Acryllic shield over the Primary circuit with a strike rail along it's edge is effective at protecting from strikes with far more success than a standard strike rail.

Looks like a ready to go terry filter is only $15 more than the parts to make one, and it'l come from one place within the country. I know what to save up for now, heh.

 

[Seoul_lasers]

$1000?? What the hell, are you using gold plated mosfets or something?

I know that you can build a decent sized SSTC for a lot cheaper than that. The parts aren't THAT expensive, and you also get a plus of smaller size (less magnet wire to wind, therefore less expensive) and no tank cap for the primary side so that wipes a good chunk of change out too.

Just use a PLL controller, and you can play actual music music too, like how Fiddy's plasma speaker works.



Or, you may be referring to a DRSSTC (dual resonant solid state telsa coil)



Those are beasts. It's basically a SSTC with a primary capacitor. You actually have to interrupt the half bridge in it otherwise it would destroy itself from overcurrent and overvoltage due to resonant rise.

The arcs on these are insane. They're also expensive, and yes, would cost a good chunk of change for one... you have to use massive mosfet or IGBT bricks and heatsinks for these. You also need a nice bank of capacitors for the primary. And I mean a NICE tank. Otherwise they go boom boom.

You'd be correct. Small SSTCs are cheaper. $300 sounds right. I was actually referring to drsstcs. They're utilizing 100A igbts.

 

[AUS]

Looks good, im jealous I always wanted a truley spectacular tesla coil!

I might look at doing this one day, I am fortunate to have a supply of HV capacitors, resistors and diodes at work rated to many KV. This year I am about to replace two 10kW AM TX's, these have 30kV unloaded on the HT supply (and we are talking amps here too). I've seen gecko lizards struck by lightening in the tube cavity, this will jump about 50mm / 2".

I would definately look at a bleeder resistor on the caps, you can use a series chain of hi ohm mains rated resistors. Sooner or later someone (you) may come along and touch them when its off. They'll stay charged for a surprisingly long time if there is no drain on them.

 

[Seoul_lasers]

just discharge the cap bank manually. You can simplify the build of the MMC without the resistors. Actually I ve never needed any bleeder resistors, even when building 100nF cap bank. Before working on the spkg, I just discharged any remaining charge via an insulated rod with a discharge terminal. Worked everytime.

 

[Sigurthr]

Update!

My Coils, Spark Gap, and discharge wand came in the mail today! I spent all of my free time building and rebuilding this TC, and I am VERY pleased with the result.

At first I tried it out as it came - as the seller intended it to be... meh. Total meh. 1" corona, no sparks, no streamers, disappointing. This was expected though as the capacitor included was a 1.6nF 20kV ceramic cap inside some PVC tubing. I couldn't find a single NST or OBIT that would be resonant or LTR with that value. I think it is the seller's method of making the TC "safe" and getting around the requirement for a specific NST value.

I removed the stock capacitor and wired up my MMC which is specially designed for my NST. I decided to go with a bleeder resistor after all, once I tested performance with and without to make sure there was no detriment. It is a 20kV 6.5MOhm 5W non-inductive resistor, and I placed it right across the high current buss from the MMC.

I decided not to fully wire in the NST and instead leave the leads flying and use jumpers to connect the power. This forms Safety Disconnect #1. If someone were to walk by and plug it in (won't happen but just IF) nothing would happen, no arcs, no corona, no HV, no caps charging at all.

I began by modeling the Tesla Coil with TeslaMap and JavaTC to see what needed to be done to the heart of the coil. I rewound the primary, using the wire supplied to the specs that TeslaMap and JavaTC suggested; 4.5 turns.

The computer modeling also suggested a topload to bring down the resonant frequency, so I made one. I made it 4" sphere, just as the modeling had suggested, and to my surprise... no real measureable increase in output. It was totally unexpected, but I kept on working.

I then removed the new topload and scrapped my models and remodeled it from the ground up. The new primary would be 2.24 turns and a break out point would be needed (streamers/sparks increase capacitance) to compensate. Done. Hey! the coil is starting to look pretty damn good! I managed 2.5" to 3" corona and sparks to air.

Now for fine tuning. I didn't really know where the issue was, so I started at the spark gap. I set and re-set it often but always being careful to make sure the gap fired without fail with the primary circuit removed - I was NOT going to chance a resonant rise blowing out my MMC! Modeling doesn't realy help in this area, it can tell you the maximum safe gap to test but it can't account for atmospherics, air currents, thermal effects on the gap, etc. I also played around with quenching, but found it to be of more detriment than anything. Cooling the electrodes did help extend run times and keep power levels up though. In the end I set my electrodes to fire at approximately 98% open circuit voltage (with primary circuit removed) and went with either a very gentle breeze for long run times or gentle suction to pull ionized gas out for shorter run times (this gives the best performance when thermal electrode issues are not a concern).

Now I should have been seeing better results than I was at this point, but I figured it was just inexperience and too high of expectations. I started looking for subtle changes between the model and the real thing and I discovered that I had neglected inputting the Primary Lead lengths. I had nearly 26 inches of combined lead length, all the same gauge as the primary. Stray Inductance galore!

Now one thing I noticed also was that the primary got very hot after only 30sec of run time, and the performance suffered. This was the primary wire which came with it. It turned out to be ~14AWG Stranded STEEL wire, the kind they use for bicycle brake lines! NO WONDER!

I replaced the primary with 12AWG solid copper wire (it took every bit I had) and I had to make another jumper for the HV side of the Primary Lead out of 22AWG copper wire + aligator clip (just like I did for the LV side of the Primary Lead / Safey Disconnect #2). This turned out to be a very wise choice, both the 22awg jumper, and the new primary that is.

For the first time I saw actualy STREAMERS! Wiggly beautiful streamers! But wait... what is that... Racing Sparks????!!! NOOOOOOOO....

Coupling coefficient was too high. Two ways to fix this;
1) change to a flat sprial coil.
2) change the intercoil geometry (spacing/diameter).

I went with the second option. I figured, why not see if it can be a simple fix? I pushed the primary down to against the wooden base, the lowest 3/4" is completely below the secondary's first turns, it may be enough, at least try it, right? IT WORKED! Now the primary is 3/4" tall, from 0" to 0.75", and the secondary starts at 1" and goes to 7.5" in height above base. Racing sparks completely gone AND now the streamers are more persistant and LONGER! Ooooooo dancing lightning goodness.

End Result:
Operating Frequency (Fo): 1.1745MHz
Voltage peak @ Topload: 741,000V
Max Streamer Length: 4.5"
Skin Depth at Fo: 2.71mils
Coupling Coefficient: 0.114 (0.124 Recommended) [VERY Close to Ideal]
Primary Peak Current: 1518.9AMPS

-New primary runs cold for even 5 minutes.
-Spark gap does not need quenching, but does benefit from gas removal and cooling.
-NST runs cool and shows no signs of stress.
-MMC is Larger than Resonant, Bleeder Equipped, and only runs at 66% the rated voltage. No detectable temperature increase.
-Secondary runs cool. Interturn capacitance of secondary causes it to retain a charge well after disconnected and shorted. Damn thing makes for a painful but harmless shock.
-The electric field from the TC has doubled if not tripled in size in addition to the increase in streamer length. Now fluorescent tubes light from several feet away.

Pic:

Video: http://youtu.be/y2imD_4qr78

 

[Fiddy]

Wow nice work! +1 :beer:

 

[Meatball]

Anyone else go the DRSSTC route? I think that's the direction I'll continue.

 

[AUS]

Very cool!
Glad to see all the safety's in place too.
Seems pretty stable too. No melting on the spark gap or from the corona?
Much RFI - have you taken out every AM radio in a 3 block radius?

Yes +1, well done!

 

[Sigurthr]

I'd really like to try one and probably will at some point. The electronics behind it are a little beyond my skill level right now though, I have no experience with IGBT bricks and gate drivers / gate transformers, or interrupters. I've built H bridges before but for motor control, not for inverter technology, haha!

 

[Seoul_lasers]

Firstly, Good Job !
nice little build to show the concept of a working tesla coil. This one is safe for inside a classroom setting. Provided no-one comes into contact with the capacitors or the sparkgap.
I see several slight improvements you could make to really get some hotter sparks!

I have several improvements that will make your Tesla coil perform even better.

1) Get a thicker ground wire. The wire in the picture is FAR too thin and you are loosing transmission power from the top terminal. Try going car jumper cable. 2ga failing that try using a piece of copper strip. The more solid a ground the better! You've got to make the best connection for largest potential difference between earth and hot side.

2) The gap needs as much quenching as you can give it. (quenching allows the capacitors time to charge to maximum bang) (though it is a low power coil)
add 2 more gaps. For a coil running this voltage, ~8mm gap should be fine. 2mm per gap if 4 are used. ..etc

3) Add a top load. you'll have to re-tune the coil, add more turns, but you'll notice the sparks
are MUCH longer. 25% more possibly. I wouldn't go past a 9" dia toroid for something this small. Also, experiment by adding a single breakout point.

4) Double the transformers. Parellel for more current. ($$)


Your coil's output 120-150kV for your coil's output is possible, though remember we have two kinds of voltage here, The main one being HFREQ AC in the 100's of Khz-Mhz, but due to the potential from the top terminal you'll also have a HV-DC current building up as well. This is much more evident when using larger top loads. I personally have been the unwilling volunteer of a ~45cm long residual"spark" charge from a 42.3" spun topload from my 6KW coil after it had been turned off. Hurt like hell. Lots of current.

 

[Sigurthr]

Thanks!

Aye, I know the 16AWG ground wire was a no-no, haha, but it was thrown together quickly. This was mostly a "dry run" to see if I had what it took to become a Coiler. (I do.)

As for #2 I left the gap small to completely rule out any resonant rise problems (even though I am using a LTR MMC) and also because of such a low inductance/turns primary I didn't want to stress the caps too much. I figure that since the input impedance of the primary coil is so low the caps are going to be putting out incredible current levels, best to dial it back a bit. I did note better arcs with the gap widened though, you are right.

#3; I tried a 4" sphere and it degraded performance. This was with the old primary though but I don't think the conductor material was the problem. I'm fairly certain that with SGTCs once you get above a certain size you absolutely can not use a helical primary. I can't find any documentation on it though but it just seems to be the case. I mean it is certainly there if you look at coupling coefficient data; helical primaries are way too high in coupling for a SGTC.

Basically once you raise the topload size you have to increase the primary turns number. Doing so on a helical (vertical) coil increases the coupling factor. Raising the turns number on a spiral coil does not appear to significantly affect coupling. Thus if you are to avoid racing sparks and other signs of overcoupling you either have to stick to a small topload when using a helical coil or switch to a flat sprial coil. It should be noted here that this secondary coilform is epoxied to the base, and epoxied to the primary coilform, and cannot be separated. This relegates this secondary to helical coils only, which means you're stuck with a small topload.

It is starting to all make sense why the person who makes/sells these uses such an obviously inadequate capacitor. If he were to use a proper size capacitor he would have to change the primary from helical to sprial and increase the topload size to keep the number of turns needed to prevent primary overheating. It seems he isn't really out to make a superior product but rather an inexpensive to make product that does enough to sell and no more.

I am going to leave this coil alone as it is and build a new one from scratch in the future (I'm still unemployed and looking for work, can't afford to spend more money for a long time!) with a much larger secondary and a 1/4" copper tubing flat spiral primary. I'll use the same MMC and NST and sell this secondary on the cheap. I'm thinking of having EasternVoltageResearch make the secondary for me, it is cheap enough when you factor in that I cannot get PVC locally (50mi drive to nearest hardware store with any) and the price of wood and hardware locally. The Plywood I bought for this coil cost me over $30, and the nuts and bolts came to a little over $20.


BTW; no RFI problems detected at all. I ran this coil in my bedroom a good 20ft (through several walls) from anything sensitive. Also I live in the middle of no where, there are no neighbors within a mile of me.

 

[Seoul_lasers]

Thanks!

Aye, I know the 16AWG ground wire was a no-no, haha, but it was thrown together quickly. This was mostly a "dry run" to see if I had what it took to become a Coiler. (I do.)

As for #2 I left the gap small to completely rule out any resonant rise problems (even though I am using a LTR MMC) and also because of such a low inductance/turns primary I didn't want to stress the caps too much. I figure that since the input impedance of the primary coil is so low the caps are going to be putting out incredible current levels, best to dial it back a bit. I did note better arcs with the gap widened though, you are right.

#3; I tried a 4" sphere and it degraded performance. This was with the old primary though but I don't think the conductor material was the problem. I'm fairly certain that with SGTCs once you get above a certain size you absolutely can not use a helical primary. I can't find any documentation on it though but it just seems to be the case. I mean it is certainly there if you look at coupling coefficient data; helical primaries are way too high in coupling for a SGTC.

Basically once you raise the topload size you have to increase the primary turns number. Doing so on a helical (vertical) coil increases the coupling factor. Raising the turns number on a spiral coil does not appear to significantly affect coupling. Thus if you are to avoid racing sparks and other signs of overcoupling you either have to stick to a small topload when using a helical coil or switch to a flat sprial coil. It should be noted here that this secondary coilform is epoxied to the base, and epoxied to the primary coilform, and cannot be separated. This relegates this secondary to helical coils only, which means you're stuck with a small topload.

It is starting to all make sense why the person who makes/sells these uses such an obviously inadequate capacitor. If he were to use a proper size capacitor he would have to change the primary from helical to sprial and increase the topload size to keep the number of turns needed to prevent primary overheating. It seems he isn't really out to make a superior product but rather an inexpensive to make product that does enough to sell and no more.

I am going to leave this coil alone as it is and build a new one from scratch in the future (I'm still unemployed and looking for work, can't afford to spend more money for a long time!) with a much larger secondary and a 1/4" copper tubing flat spiral primary. I'll use the same MMC and NST and sell this secondary on the cheap. I'm thinking of having EasternVoltageResearch make the secondary for me, it is cheap enough when you factor in that I cannot get PVC locally (50mi drive to nearest hardware store with any) and the price of wood and hardware locally. The Plywood I bought for this coil cost me over $30, and the nuts and bolts came to a little over $20.

One of the reasons you lost output moving up toroid size is because you don't have enough turns on your primary. The effects the magnetic fields characteristics and as such fails to energize the full winding length. primary size will effect the coupling. The best overall design for a mini coil like yours is to flatten the primary out. You can use small gauge copper refrigerator copper tubing. 1 roll ($18) or so. with 1/2" spacing. 15 turns is more than enough. If you want I can show you how to make a flat primary for beans. Tubing works and actually so does 4ga electrical solid core (battery) cable. When your primary i beenalbe s tapped at 10 turns with that same 4" topload, you're going to notice something quite different in the spark length department.

You're talking to a seasoned TC'er who's built coils for galleries. Largest I've
been able to design with my space in Canada was a ~5-6KW coil with a 7" dia coil form and 1450 turns of #23 ga wire. It used a paralleled bank of NPFC Neons 12kV as the final input supply. 560mA @ 12KV. Never went past 600mA, due to our limited service.
Output exceeded well over 1MV.
11.5ft of spark is pretty decent.
I also, rigged up our 14.4kV PT and limited the input voltage to 100V at ~50A, using a variac.

 

[Sigurthr]

One of the reasons you lost output moving up toroid size is because you don't have enough turns on your secondary. The effects the magnetic fields characteristics and as such fails to energize the full winding length.

So I was right in thinking this secondary is too small then! Good to know. What size would you recommend for a larger coil using the same MMC/NST? Is 3"x15" going to be big enough to get ~1ft streamers? I'd rather not go so large that streamer length exceeds secondary height, because then I need to do serious primary side strike protection.

You're talking to a seasoned TC'er who's built coils for galleries. Largest I've
been able to design with my space in Canada was a ~5-6KW coil with a 7" dia coil form and 1450 turns of #23 ga wire. It used a paralleled bank of NPFC Neons 12kV as the final input supply. 560mA @ 12KV. Never went past 600mA, due to our limited service.
Output exceeded well over 1MV.
11.5ft of spark is pretty decent.
I also, rigged up our 14.4kV PT and limited the input voltage to 100V at ~50A, using a variac.

One word; AWESOME. How the heck do you wind such large secondaries? I assume you have some kind of motorized jig?

 

[brtaman]

Cool stuff. Nice build thread.

Tesla Coils are extremely fun. I had great times building them.

As far as costs, if you are frugal you can make a 2kw plus SSTC for less than the cost of a decent laser. I don't know where the several thousands of dollars came from.

Enjoy the hobby while you still have time, sadly I barely have time to breathe. Still have the parts laying around in my old lab for a monster DRSSTC though might get back into.

Anyone looking to get more into Tesla Coils just start browsing through 4HV and pretty soon you will be addicted.