Tesla Coil Build Thread

[upaa27]

Oh crap, seriously? I can find a similar one and post it. It looks old with wires coming out of it. It has a sticker on the side saying not to use in construction because it doesn't have a GFI.

I'd have to say it looks just like this one, but is 9kv instead of 5kv.Gas Tube Sign Transformer 120 Volts Actown FG 4047 | eBay

Edit: Oh, now I know what your saying. No it is the iron core type not the ballast type. It costs a lot for shipping too, so it must be heavy.

And it is 60htz out I'm pretty sure.

Also, I think I know why yours is only 3ma. It's because the smps ones are designed to work with different tubes. They self determine how much current is needed and output that much. Also, they're high frequency. Imo, they're like a higher voltage electronic ballast.

That's a really lucky ebay find!

For the smps, yes, the load determine the draw. I think tha it is the resistance(or impedance if you don't rectify it) of the plasma that determines the current. I had one on hand from an old neon sign. I used it to power a discharge tube with 22 ma optimal but got only 3 out of it. Oh well.

For the iron core ones, I thought that the load also determines the draw on these? Although my limited experience is only with microwave oven transformers.

 

[Sigurthr]

The load determines the draw on ALL VOLTAGE SOURCE power supplies. But yes, the SMPS "NSTs" won't put out their rated power unless the load actually needs that much. The issue is that there is a standard convention in neon signs as far as ratings go; a 1foot long tube might require 7kV 15mA for example, and would be supplied with such a transformer in the old iron core days. However since modern SMPS units operate at high frequency it actually takes far less voltage and current to maintain full brightness because the ionization energy is lower for RF than it is for low frequency (50/60Hz). So the SMPS needed for that 1foot tube might only put out 3kV 5mA, but be labeled as 7kV 15mA so that an end user can see the iron core ratings the tube needed and match it with an appropriate SMPS. Since the false ratings are always going to be higher than the actual output, there is no legal issue. But for repurposing the SMPS it causes major problems.

Those ACtown iron core transformers are quite good though, looks like you got a nice one if it works as it should.

 

[USAbro]

Thanks. I'm hoping for the best. I'm also hope my 7 turn primary is enough, and my 14 inch 2inc diameter secondary is big enough.

Edit: I'm guess I'll at the very least get this performance right? http://www.angelfire.com/80s/sixmhz/trashy.html

What should I do to protect the NST? Sigurtur, are you sure chokes aren't good? The site linked says you need them.

 

[upaa27]

The load determines the draw on ALL VOLTAGE SOURCE power supplies. But yes, the SMPS "NSTs" won't put out their rated power unless the load actually needs that much. The issue is that there is a standard convention in neon signs as far as ratings go; a 1foot long tube might require 7kV 15mA for example, and would be supplied with such a transformer in the old iron core days. However since modern SMPS units operate at high frequency it actually takes far less voltage and current to maintain full brightness because the ionization energy is lower for RF than it is for low frequency (50/60Hz). So the SMPS needed for that 1foot tube might only put out 3kV 5mA, but be labeled as 7kV 15mA so that an end user can see the iron core ratings the tube needed and match it with an appropriate SMPS. Since the false ratings are always going to be higher than the actual output, there is no legal issue. But for repurposing the SMPS it causes major problems.

Those ACtown iron core transformers are quite good though, looks like you got a nice one if it works as it should.

Ah yes. You mentioned that to me earlier. My memory failed me there.

Any trustworthy smps companies that you know of? They are cheaper and the frequency makes them a bit easier to use for some purposes.

 

[Sigurthr]

Yes, I'm sure the chokes are useless. There's so much capacitive coupling going on that the RF completely bypasses the chokes unless you use RF hardline transmission cable for all the connections and ground all of the LV side in a faraday cage.

Think about it this way: an average 500W coil will transmit ~200W to your hand through the air at a distance of over three feet. The air between your hand and the topload has a resistance in the thousands of billions of ohms, yet tons of power flows because the capacitive impedance is low. You've got maybe two feet of air between the secondary and the NST... a dinky 100kOhm choke isn't going to do anything when there's a 100ohm capacitive path between the secondary and the NST AND the parasitic capacitance between turns of the choke itself would bypass any impedance the inductance of the choke poses.

You NEED to go use JavaTC. I don't understand why you haven't done it yet. It takes ten minutes at most and is free. It is the only way you are going to know if your primary has enough turns, if your primary needs to be raised or lowered in relation to your secondary, if your topload is too big (yes that's a thing! really big toploads produce NO STREAMERS), and approximately where to tap your primary coil at.

@Upaa27 You're not going to like the answer, haha. You get a CO2 laser power supply from china rated for at least 60W and employ the "DC Resonant Charging" method but skip out on the series diodes, since the supply outputs DC to start with. You essentially feed the HV into a series inductor and parallel capacitor to form a low pass filter to protect the SMPS from RF and limit inrush current to the parallel cap at gap discharge. It's messy and usually not worth pursuing.

 

[USAbro]

I don't understand JAVATC. It is too complicated. WTH is a ground plane radius, or a "Radius 2-LV end" or a "Toroid major diameter." :wtf:
They don't specify units etc.

 

[The Lightning Stalker]

I don't understand why you haven't done it yet.

Because he's trolling us.

 

[Sigurthr]

Because he's trolling us.

Given that he has posted photos of the parts in progress, I would sincerely hope it isn't just a ruse, as it would be going the extra mile in that regard. No, I think he's just young and inexperienced, and has never undertaken such a complicated endeavor before. Remember, LPF attracts a young crowd, it isn't all old men like us, haha.

I don't understand JAVATC. It is too complicated. WTH is a ground plane radius, or a "Radius 2-LV end" or a "Toroid major diameter." :wtf:
They don't specify units etc.

I've made it simple for you:

http://i.imgur.com/f6lnbEe.png

 

[USAbro]

Thank you very much! Now I can try it out. Yippee.

@Lightning Stalker. I certainly am not trolling you guys. I would have never gotten this far without you guys help.
I am reading ever blog or post I can on TC's and you guys still help. For example, I would have started to build chokes if Sigurthur didn't help me out.
I am not trolling.

And as Sigurthur said, I haven't built something like this before. I have built fairly complicated airguns, but they work by machanics, and if you can understand what you see, your good. But with a TC, knowledge of the invisible is involved.

 

[USAbro]

Why is Javatc taking so long? It's been crunching numbers for 30 minutes!

 

[Sigurthr]

It probably crashed. Either your device (perhaps you're not running it on a pc? Or don't have java installed? Or didn't allow the java applet?) has an issue or it bugged because an invalid variable was entered in somewhere. Close the browser and try it again, it happens sometimes.

 

[USAbro]

@Sigurthr-I sent you a pm but I'm not sure if I need it anymore.

I used another software called Tesla Coil Cad, and after inputting some specs, I got this.

Filename: new design.tes

Secondary Coil Design
Diameter of secondary coil : 2.00in
Winding height of secondary coil : 12.00in
Wire diameter for secondary coil : 0.0126in
Spacing between windings : 0.00in
Secondary turns : 952.00
Secondary wire length : 498.67ft
Secondary inductance : 7.03mH
Approximate resonant frequency : 882.88kHz
Secondary quarter wavelength resonant frequency : 493.44kHz
Secondary self capacitance : 4.62pF
Toroid capacitance required to form quarter wavelength coil : 10.17pF

Primary Coil Design
Primary capacitance : 0.015uF
Primary resonant frequency : 493.44kHz
Secondary coil diameter : 2.00in
Primary conductor diameter : 0.25in
Primary turn to turn spacing : 0.50in
Spacing between the secondary and the inside turn of the primary : 0.50in
The primary will need to be tapped between turn 6 and turn 7 to form a resonant circuit at 493.44kHz

The approximate inductance at each turn is :

Turn 1 0.17uH
Turn 2 0.59uH
Turn 3 1.33uH
Turn 4 2.44uH
Turn 5 4.02uH
Turn 6 6.13uH
Turn 7 8.84uH
Turn 8 12.23uH
Turn 9 16.38uH
Turn 10 21.37uH
Turn 11 27.26uH
Turn 12 34.13uH
Turn 13 42.06uH
Turn 14 51.12uH
Turn 15 61.38uH

Note: The Primary Coil Design is not using the calculated Neon Transformer / Primary Capacitor Matching Value

Neon Transformer / Primary Capacitor Matching
Transformer secondary voltage : 9.00kV
Transformer secondary current : 30.00mA
Number of transformers : 1
Primary capacitor required to form a resonant circuit at 60Hz with neon transformer(s) : 0.0088uF

 

[Sigurthr]

It looks like that doesn't take into consideration the actual toroid dimensions you are using (critical) and just spits out the secondary coil self-resonance, and what toroid capacitance is needed to make it work with your primary circuit. Not bad, not great. Tesla Map (not free) does similar.

See if You can get JavaTC working, it's known to be the most accurate. I'll check my PM.

At the very least post your input data for all those fields I outlined in the how-to graphic I made, and I'll run JavaTC for you.

 

[USAbro]

Here you go. And I edited my post with the specs up their cause the secondary was a bit off.


Radius 1 LV end-1 inch
Radius 2 HV end-1 inch I'm assuming this is the radius of the form at each end
Height 1 LV end-1 inch
Height 2 HV end-1 inch I'm assuming that this height is where the coil starts and ends on the form and 1 inch is the distance from the wire to the end of the form on each side.
If you need coil height it is 12 inches. And the coil form is 14 inches.
Turns-952
AWG-28
Primary
Radius 1 LV end-2 inches
Radius 2 HV end-6.25 inches
Flat coil so both height is zero. It is on a flat plane.
Turns-7
Wire diameter-1/4 inch copper tubing.
Primary capacitor-15nf
Primary Lead Lenght-44 inches, +/- <10
Lead wire diameter is 12 gauge wire. 0.08080 inches fount out at http://www.tdiinternational.com/contents/en-us/d488_awg-to-inches.html\

Toroid

Minor Diameter-I don't know what this means but the outside of the doughnut is the entire diameter which is 14.75 inches.
Major Diameter-Is this the same minor diameter? Anyways, if this helps the duct is 3" flexible dryer duct, and it is coated with a lot of aluminum tape. The singe pie pan on the bottom, that is indented to about the middle of the toroid, is about 8 inches in diameter in the middle of the toroid. I'm using a metal bolt although I know I shouldn't, so for now, I'm covering it with electrical tape.

Center Toroid height is 14.50 inches.


Thanks.

 

[Sigurthr]

I think the reason it crashed on you is you have a few issues:

1)Secondary coilform diameter is 2 inches. If you used PVC, it goes by the inside diameter, but as a coilform you use the outside diameter. 2" pvc usually has an outside diameter of about 2.125", so the radius would be 1.0625.

2)952 turns won't fit on 12" (exactly) of winding, when using 28ga. The JavaTC calculator suggests 821 turns. This alone pops up an error when you input the numbers, so either it's less turns, or the winding length is more than 12".

3) *If Secondary LV height is 1, and your winding is 12", then Secondary HV height is 13. Picture from the bottom up: coilform edge is 0, winding start (lv) is 1, 12in of winding, winding stop (hv) is 13, coilform stop is 14.

4) Primary height when flat is equal to primary wire diameter; 0.25.

5) Toroid minor diameter is the diameter of the hollow tubing that comprises the outer rim, in your case 3" duct, which is usually around 3.25".

6) Not an error, but 44" of primary lead length represents a HUGE LOSS OF POWER, perhaps up to a 45% loss in power. I had 24" in my first coil, and reducing it to 12" increased my output power by about 28%.

That being said I adjusted your numbers to get you into the ballpark, and the output looks both good, and realistic.

J A V A T C version 13.2 - CONSOLIDATED OUTPUT
1/8/2015, 2:16:02 AM

Units = Inches
Ambient Temp = 68°F

----------------------------------------------------
Secondary Coil Inputs:
----------------------------------------------------

1.0625 = Radius 1
1.0625 = Radius 2
1 = Height 1
13 = Height 2
821 = Turns
28 = Wire Awg

----------------------------------------------------
Primary Coil Inputs:
----------------------------------------------------
Round Primary Conductor
2 = Radius 1
5.205 = Radius 2
0.25 = Height 1
0.25 = Height 2
5.2784 = Turns
0.25 = Wire Diameter
0 = Ribbon Width
0 = Ribbon Thickness
0.015 = Primary Cap (uF)
40 = Total Lead Length
0.08 = Lead Diameter

----------------------------------------------------
Top Load Inputs:
----------------------------------------------------
Toroid #1: minor=3.25, major=14.75, height=14.5, topload

----------------------------------------------------
Secondary Outputs:
----------------------------------------------------
494.39 kHz = Secondary Resonant Frequency
90 deg° = Angle of Secondary
12 inch = Length of Winding
68.4 inch = Turns Per Unit
0.00198 inch = Space Between Turns (edge to edge)
456.7 ft = Length of Wire
5.65:1 = H/D Aspect Ratio
29.3995 Ohms = DC Resistance
18746 Ohms = Reactance at Resonance
0.22 lbs = Weight of Wire
6.035 mH = Les-Effective Series Inductance
6.114 mH = Lee-Equivalent Energy Inductance
5.925 mH = Ldc-Low Frequency Inductance
17.173 pF = Ces-Effective Shunt Capacitance
16.95 pF = Cee-Equivalent Energy Capacitance
21.976 pF = Cdc-Low Frequency Capacitance
4.23 mils = Skin Depth
15.327 pF = Topload Effective Capacitance
89.6081 Ohms = Effective AC Resistance
209 = Q

----------------------------------------------------
Primary Outputs:
----------------------------------------------------
494.39 kHz = Primary Resonant Frequency
0 % = Percent Detuned
0 deg° = Angle of Primary
9.96 ft = Length of Wire
1.65 mOhms = DC Resistance
0.357 inch = Average spacing between turns (edge to edge)
1.069 inch = Proximity between coils
0 inch = Recommended minimum proximity between coils
5.545 µH = Ldc-Low Frequency Inductance
0.015 µF = Cap size needed with Primary L (reference)
1.392 µH = Lead Length Inductance
15.006 µH = Lm-Mutual Inductance
0.083 k = Coupling Coefficient
0.123 k = Recommended Coupling Coefficient
12.05 = Number of half cycles for energy transfer at K
12.13 µs = Time for total energy transfer (ideal quench time)

The take away is this: you'll need to tap your primary HV connection (outside connector) around 5.27 turns and adjust from there for best performance.

Potential improvements for AFTER the coil is running well and tuned:
1) Greatly reduce the primary lead length. The shorter the better. Copper strip 1" in diameter works better than any round wire.
2)increase coupling slightly by raising primary coil up about 1"; styrofoam under the insulating plastic you're using works great.

Note: at 500KHz this is a medium frequency coil, so there will be a reduction in spark length that is unavoidable. It shouldn't be that noticeable and you have realistic expectations (6"). If you can smooth out the toroid so only a single streamer forms, and if you can improve the ground connection as much as possible (tons of methods in this thread, go search em, too much to repost!!!!!) you'll see performance improvements beyond 6".

 

[USAbro]

Man you don't how much that helps. You deserve a billion rep! Thanks a lot. Would 6" streamers be to air or ground as I want air arcs. My calculation on that other software said the quarter wave frequency is about 500khz too.
I guess this coil is a good place to start.
The reason for the wire length is it goes in and out of everywhere, like connection to the inner primary tubing (10") connection to the spark gap, as I posted in my images, wire goes into the PVC supports and back out, then has to go to the caps on the table, then has to go the the secondary, about 10" goes to the safety gap, and it all adds up to a rough estimate of about 44" plus or minus 10".

I can't wait, the tracking numbers says the NST is arriving Saturday!

And I guess 7 turns is enough, as it requires 5.27 approximately.

Edit: Do you think output might be like this? https://www.youtube.com/watch?v=mcHFUzIvMtE