Tesla Coil Build Thread

[Speedy78]

Did some work on it today. Have more tesla type results but still nothing spectacular. The only waveforms that looked normal were slightly out of tune by 10 to 20khz or so. Try to tap a different spot to move it more in tune and I could hear the gap sounding odd and the waveform was a bunch of zig zags all through the primary and secondary ringdown. Like the gap never quenched. After running for 10 seconds at the "opportune point" the sparks would begin to grow slightly and then go back to the thin, never growing, sparks you see in the video.

Would the lack of topload capacitance cause the sparks to not really grow? This toroid is from my onetesla that sadly produces better.

So whats the expert word? Seems like there's really no power still. Would a rewound secondary featuring less turns on a smaller former with slightly thicker wire be a good starting point so I can add that big topload back on?

The onetelsa also beats this with EM field. When the florescent bulb begins to glow im in strike range from streamers. Didnt have this problem with the big topload. Again pointing towards topload capacitance but i fear I'm just making up technical reasons as to why its not functioning right.

 

[Seoul_lasers]

Yeah I was planning on a quarter inch smaller and thicker wire. It was focused. It was almost like a purple skirt from the toroid to strike rail. No visible streamers and not focused on anything. Only other top load I have is the one tesla. Is that too small? I have one out of dryer ducting I made a while ago that's bigger than the one tesla but smaller than this one.

Having another stab at this....*
1) Your gap really sounds sputtery... try narrowing the gap a little like 1 or 2mm and see if the sputtering decreases.


2) Another thing I could think about causing that much detuning is "possibly" your strike rail location. Try a short run without it.
I am willing to bet that the corona is a sign of major loss of energy transfer to your secondary.


Give it a go! :can:

 

[Speedy78]

This is sounding a he'll of a lot better than before. The sound would change as I tapped more or less turns. Is that normal as I move away from resonance?

 

[Seoul_lasers]

This is sounding a he'll of a lot better than before. The sound would change as I tapped more or less turns. Is that normal as I move away from resonance?

Yup, that is completely correct. You can really hear the difference in the spark gap when you have de-tuned (away from resonance) your coil!

Though it might sound better, you're still loosing energy.
For a coil your size I'd opt the strike rail out.

In your case I believe it's acting as a spare turn and wasting energy even though it is not a complete turn due to it's position in the magnetic field.
:thinking::thinking::thinking::thinking:


Have a listen to my puny 3" dia coil. That's using a tungsten spark gap with 12Kv 180mA.
that's the noise of a fully in tune coil. (hissing/roaring)
Though you will not get anywhere near that output, you should be hitting at least 25" discharges

In case you are wondering, that is my classroom where I used to teach and no I didn't kill any students.
Though I did blow the tiles off my classroom room ceiling.

 

[Speedy78]

With the big toroid on there it seemed to favor it. I just don't want a wild arc to catch my primary. Next time i have a go at it I'll post my waveforms. They are very odd to say the least.

 

[Sigurthr]

The gap sounds very discontinuous, suggesting it is too wide still.

The lack of EM field size is due to a smaller topload capacitance. Nikola's personal coils used very large toroids which was conducive to extremely large EM fields with little to no breakout (as he was about transmission of power, not wasting it in pretty arcs).

The tuning does seem better and it doesn't look like there are any obvious flaws or issues that leap out. I have a few ideas you can run through but they all require that tuning not be an issue.

So, you mentioned scope waveforms, how are you scoping it currently? If you could get scope shots of the primary CURRENT that would be superb but even just capacitively coupled scope shots of the output from the EM field would be fine. How are you measuring the primary frequency btw? You can remove the secondary and place a dummy load (large soup can etc) in its place so you can measure the capacitively coupled primary EM field without the secondary's influence.

Steps to take:

1) make sure you're close in frequency, if you're tens of Kilohertz out then you're not going to get good results no matter what, SGTCs are very low bandwidth tank circuits because every stage is resonant. The standard procedure is to remove the NST and attach a signal generator to the gap terminals and then scope the output of the secondary for a signal peak. This peak happens when the primary is in tune. Manually tuning is best done by running the coil as normal with a scope a safe distance away and looking for a signal strength (amplitude) peak when adjusting the primary tap.

2) Verify your NST hasn't gone bad. As we discussed in text NST output sides are two windings in series where the antinodes are the output, one side can go bad through various failure routes resulting in insufficient loaded current or insufficient loaded voltage. These are hard to diagnose but since you have a second NST that is known to be good you can use it as a reference. Use the 7.5/30 in it's place and compare results with an identical tuning tap point. Note that the LTR cap value for a 12/30 IS the pure resonant value for a 7.5/30 so be sure to set your gap correctly or you risk frying the 7.5/30. If you get better results with the 7.5/30 then the 12/30 is likely damaged. In this case you can simply reconfigure your MMC to better suit the 7.5/30 AND this would result in a lower primary tuning range letting you use the BIG toroid again.

3) optimize the spark gap as this is the largest point of power loss in any classic tesla coil. Running with too small of a gap means the caps never charge to full voltage, but it poses no danger to the system. Running with too large of a gap means the voltage will stress the system and can result in failure, it also means the gap doesn't fire at the same rate it normally would meaning the duty cycle of the coil is lower, so less average power is delivered despite higher peak powers. Likewise, the wider the gap the more resistive plasma the energy has to go through and the more waste heat and light is produced, sapping energy from the system. Too small of a gap simply won't quench and when the gap is active (unquenched) no additional power from the transformer makes it to the primary. If you can ensure that the gap is well quenched then you can effectively lower the gap to eliminate power losses and increase output.

Having the coil out of tune you can get phase mismatches between the primary voltage peaks and the reflected power from the secondary which can cause erratic gap operation. Power flows in both directions in a SGTC, not just in one like a VTTC or SSTC. Energy is pumped into the secondary and it sloshes back into the primary where it SHOULD reinforce the newly incoming energy being pumped into the primary, but if it is out of phase due to being out of tune then it can destructively interfere with the energy being pumped into the primary.

4) Overcoupling is definitely not an issue here now, so you could very well be under-coupled. Try raising the primary or lowering the secondary. Another option is you could swap the tap points on the primary so that the outter turn is the fixed point and your tap point is the movable one more towards the center. This increases coupling slightly as the magnetic field is more centered around the bulk of the secondary, even though part of it is now further away from the secondary radially. This works sometimes on coils where raising or lowering the coils is not an option.

 

[Seoul_lasers]

With the big toroid on there it seemed to favor it. I just don't want a wild arc to catch my primary. Next time i have a go at it I'll post my waveforms. They are very odd to say the least.

Doubt you'll get an arc to the primary with that small an input.

A larger toroid is not going to help you in this case.

hmmm.... I wondering about that sparkgap too..

Maybe try making an adjustable temporary copper pipe gap an see if that doesn't
make a difference. Id do this ONLY if you still run into problems after checking all the things Sigurther suggested.


:can::thinking:

Actually, I completely agree with Sigurther on this. I think we've got a case both of 3# and 4# going on here.
That's why I suggested narrowing the gap as first before moving the coil form which will be a bit of a PITA.

 

[Sigurthr]

You may be right about the strike ring; it could detune the secondary a bit. Similar tests on solid state only show a 5% detuning effect though, not the 80 KHz he's experiencing, but right now there's low risk of a pri-sec arc anyway.

I'm still wary of that NST, second hand from a non-coiler is never verified quality in my book.

S_L, Speedy may want to upgrade to a mot ballasted dual or quad mot stack and ARSG at some point, I told him you'd have some input on conversion specs and details. I think his base, mmc, pri, sec, and topload could handle that power well.

 

[Seoul_lasers]

You may be right about the strike ring; it could detune the secondary a bit. Similar tests on solid state only show a 5% detuning effect though, not the 80 KHz he's experiencing, but right now there's low risk of a pri-sec arc anyway.

I'm still wary of that NST, second hand from a non-coiler is never verified quality in my book.

S_L, Speedy may want to upgrade to a mot ballasted dual or quad mot stack and ARSG at some point, I told him you'd have some input on conversion specs and details. I think his base, mmc, pri, sec, and topload could handle that power well.

Yes, indeed I can help with those things. ah... :thanks:
I wasn't aware that the Neon XMRF hand't been tested out completely. Again, that strange behaviour could be a result of a blown winding set or inter-arcing. You might very well be right!! :undecided:

---> I have had a simular problem on my large 6.5" dia coil with the spark rail sucking away power. I did away with it and used a discharge point to control the spark's direction. Very rare, almost never had primary strikes. 96-100" or so sparks always managed to hit their intended victims. :e.g.:

Neons all 7 x 12Kv x 60mA of them are still working to this day! :evil:

I used 2 safety gaps (capacitor + transformer) as well as a ceramic cap 30Kv 500pF across the transformer safety gap to ground.
--0 o 0--

 

[Speedy78]

Plan for tomorrow is shrink the gap slightly and upgrade to 8ga car amp hookup wire. The 10ga is nice but by no means flexible. I also plan to have the main connection a clamp as well for testing purposes. Is this something that would work? Will also pull off the strike ring. May also try to raise the primary. What would be a good amount of lift at first. 1 inch?

Sucks testing during the day. Can't see any output and can barley see my laptop. Will also post the waveforms. Just sucks getting the delayed response. I need my coilers on speed dial for on the fly tips and info lol

yes once this is working i want to throw some real power at it! Feel free to link a mot source so I can start pricing things out.

edit. Any way to test this nst outside of a coil? Where do you find 180ma nst's? That's ridiculous. Yeah my gap sounds real choppy. It's set to charge the mmc to about 98% but it doesn't seem to be firing right.

 

[Sigurthr]

I need my coilers on speed dial for on the fly tips and info lol

Haha, that's easy for me, we're same time zone, but I think S_L might be an expensive call =P.

yes once this is working i want to throw some real power at it! Feel free to link a mot source so I can start pricing things out.

edit. Any way to test this nst outside of a coil? Where do you find 180ma nst's? That's ridiculous. Yeah my gap sounds real choppy. It's set to charge the mmc to about 98% but it doesn't seem to be firing right.

IIRC the output voltages can have a small swing in them (2200 is ok to use with a 2400) but the power ratings of the MOTs should be very close. Again, this is S_L's area of expertise so I defer to his reply.

Not easily. You have to suitably load the NST and monitor the current draw. You also need a reliable way of measuring the voltage between each terminal and ground. Internal arcing is very hard to test for since it requires the use of HV, but a blown winding will show up with a low voltage continuity test.

 

[Speedy78]

Just check between to two hv terminals for continuity with a dmm?

 

[Seoul_lasers]

Just check between to two hv terminals for continuity with a dmm?

1) There are no 180mA NSTs in existence except for a Neon bombarder units ( 10-30Kv variable pad mount XMFR).
The closest ever produced was a CCNST (Cold cathode neon) that was built by Franceformer
in the late 1980's -90s. these were rated for 12-15Kv at 100-120mA each. Due to their high output current no-one ever wanted to work around them. Actually they proved quite fatal!

You can parallel XMFRs for more current. that 3" dia coil had 3 x 60mA 12Kv neons in a bank for ~2.1KVA worth of input power.


2) Going MOT is not anywhere near as easy as using NSTs, but MOTs are quite easily obtainable for the cheap.
Rule of thumb for using MOTs are to get the older Sanyo Microwaves. They're rated for 2.2-2.5KVac @ at least 500mA.

To make an MOT bank you'll need at least 2-4 transformers in an aided setup (extremely DANGEROUS -- case = hot/lethal shock hazard). I've gone personally up to as high as 6 MOTs before I needed to seriously mod the transformers to output ~12Kv ~560mA with external ballasting and insulation. The filament windings have to be CAREFULLY removed to prevent case to ground arcing. The entire setup then
must be submerged in Shell Dia-lex non detergent transformer oil. Essentially what you have at the end of all that hard work is a dual phase ~9.5Kv (4 XMFR) to ~12.5Kv for (6 XFMR) Pole-pig.
If you have the experience and room to build one of these supplies, it is an option.
below is an example of 2-3 mots creating arcs.
Forums / High Voltage / Resonant MOT arcs - 4hv.org

Showing a Quad bank with TESLA COIL.
http://www.capturedlightning.org/hot-streamer/lulabs/Tesla.html

ALTERNATIVELY and used more frequently, you can choose to use a dual MOT setup and a voltage doubler or tripler circuit. While this works, it cuts down on the input current.


Anyways...

 

[Sigurthr]

Just check between to two hv terminals for continuity with a dmm?

To check each winding individually you would check for continuity between one HV terminal and ground, and then the other and ground, and then between both. All should be continuous. This does NOT check for internal arcing, just checks for a LV-break.

1) There are no 180mA NSTs in existence except for a Neon bombarder units ( 10-30Kv variable pad mount XMFR).
The closest ever produced was a CCNST (Cold cathode neon) that was built by Franceformer
in the late 1980's -90s. these were rated for 12-15Kv at 100-120mA each. Due to their high output current no-one ever wanted to work around them. Actually they proved quite fatal!

You can parallel XMFRs for more current. that 3" dia coil had 3 x 60mA 12Kv neons in a bank for ~2.1KVA worth of input power.


2) Going MOT is not anywhere near as easy as using NSTs, but MOTs are quite easily obtainable for the cheap.
Rule of thumb for using MOTs are to get the older Sanyo Microwaves. They're rated for 2.2-2.5KVac @ at least 500mA.

To make an MOT bank you'll need at least 2-4 transformers in an aided setup (extremely DANGEROUS -- case = hot/lethal shock hazard). I've gone personally up to as high as 6 MOTs before I needed to seriously mod the transformers to output ~12Kv ~560mA with external ballasting and insulation. The filament windings have to be CAREFULLY removed to prevent case to ground arcing. The entire setup then
must be submerged in Shell Dia-lex non detergent transformer oil. Essentially what you have at the end of all that hard work is a dual phase ~9.5Kv (4 XMFR) to ~12.5Kv for (6 XFMR) Pole-pig.
If you have the experience and room to build one of these supplies, it is an option.
below is an example of 2-3 mots creating arcs.
Forums / High Voltage / Resonant MOT arcs - 4hv.org

Showing a Quad bank with TESLA COIL.
Lu Labs' Tesla Coil!

ALTERNATIVELY and used more frequently, you can choose to use a dual MOT setup and a voltage doubler or tripler circuit. While this works, it cuts down on the input current.


Anyways...

Those bombardment trannys are something to behold.

There are a few of those 15/120's around every blue moon, but I swear I've seen a 18/120 at one point. Nearly bought it but it got bid up too high.

I was thinking a 4 mot + 1mot-ballast where all primaries are in parallel and there are two series strings of two in parallel for 5k 1A output would be best for him. This way there is minimal case lethality (iirc all cases can remain grounded then as it's just a double dual-mot bank) and enough voltage to fire an ARSG with tons of current. I won't go near one of those case-live monsters, heh.

 

[Seoul_lasers]

To check each winding individually you would check for continuity between one HV terminal and ground, and then the other and ground, and then between both. All should be continuous. This does NOT check for internal arcing, just checks for a LV-break.



Those bombardment trannys are something to behold.

There are a few of those 15/120's around every blue moon, but I swear I've seen a 18/120 at one point. Nearly bought it but it got bid up too high.

I was thinking a 4 mot + 1mot-ballast where all primaries are in parallel and there are two series strings of two in parallel for 5k 1A output would be best for him. This way there is minimal case lethality (iirc all cases can remain grounded then as it's just a double dual-mot bank) and enough voltage to fire an ARSG with tons of current. I won't go near one of those case-live monsters, heh.

Yes, and that'd be what I suggest too.
However he'll want a doubler circuit /tripler to produce
enought potential to charge those caps.
With an amp or more of output even a tripler will give a manageable
300mA or so. Very economical /Decent option.

 

[Speedy78]

Is there a circuit diagram I can follow? Any way to see what type or look up mot specs before an oven purchase. I see a bunch of old ones on Craigslist for 5 to 20 bucks a piece.

Edit: re-ran the numbers in Java and I should be at 175khz on that secondary with the big toroid. Going to pull the strike rail and do the other steps previously mentioned and try it again tonight.