Small-ish 3D printed SGTC via cheap ZVS flyback build, humbly asking a couple ?s

[Michelle_]

The most expensive component of any SGTC is the HV capacitor bank. There really is no cheap way around it unless you resort to beer bottle capacitors but those are very lossy and frequently fail.

Yeah I see that now. Looks like I can use 15x 942C16P22K-F up to 12v for the current primary and secondary, then if I make a new primary and secondary with about 1/2 the resonant frequency I can use the same capacitors up to 24V or nearly. To run the current primary and secondary at 24v would take 52 of those capacitors so that's not gonna happen.

Question is what will happen if I try using the capacitors I have (beyond their rating), If I add all of my spares in series I can basically run at about 5500v with the current primary and secondary, not sure what it will do with that many volts, or what current the ZVS/flyback will provide then, any guesses?

[Michelle_]

Not sure what to make of the spark gap, my balls have to be really close together for it to reliably fire, and it doesn't really want to fire under 6kv. I guess that's how much power I'll try. I have a feeling this thing is gonna suck. Also you guys were right the primary design is dumb, when I redo it I'm going to make it more like a normal one with slots in the bottom and just use standoffs to zip tie the tube to, and space the turns closer. I won't bother printing a cover for this one since I'm changing it anyway and I doubt it will make sparks of any significant size at the lower voltage. I think the overall size is good for fitting the capacitors and all the electronics underneath though.

I'm going to have a do-over and use the 15x 942C16P22K-F at 24KV and make the resonant frequency 200KHZ. That should be as good as it can get with this power supply, and the secondary will also be larger and have more windings.

[alan sailer]

In my limited experience, if you turn on a tesla coil "first light" and you make any sparks at all
it counts as a rousing success.

Small sparks can be made bigger, no sparks, not so much.

Cheers.

[Michelle_]

In my limited experience, if you turn on a tesla coil "first light" and you make any sparks at all
it counts as a rousing success.

Small sparks can be made bigger, no sparks, not so much.

Cheers.

HOLY CRAP IT ACTUALLY WORKED ON THE FIRST TRY

Will post pics later, it's not going to set the world on fire but for 6 volts it seems pretty good to me!

[Michelle_]

I honestly can't believe the power this thing has from running on 6VDC input. I'm kind of glad I have to start with 6V because I have a new respect for how much power tesla coils can have. It really did fire up on the first try, I still can't believe it. I haven't increased the voltage over 6 yet but nothing got remotely warm including the spark gap and the capacitors according to the thermal camera. I'll post again after I have a V2.0

That's a 2 liter xenon filled globe in case anyone is wondering.

[markus]

Congratulations on first light! Also nice globe you got there. Small sparks are nice for playing around with stuff like that, bigger sparks just tend to shoot through everything after a certain point.

[NyaaX_X]

Congratulations

Sorry for my stupid..
I did not considered about the dV/dt and so on. And I don't want to use that much capacitors for the MMC then. So I use this http://www.eaco.com/uploads/soft/160630/MS.pdf
MS15000 & MS8000, even without bleeding/even-voltage resistors... So maybe the capacitors have been hurt a lot..  :'(
Or as I said, Lucky~ My SGTC is working at ~ 1MHz so maybe the dV/dt is OK? My circuit is really a crude set. And, The coil make a lot of noise so I only run it for seconds every time.

[Michelle_]

Congratulations

Sorry for my stupid..
I did not considered about the dV/dt and so on. And I don't want to use that much capacitors for the MMC then. So I use this http://www.eaco.com/uploads/soft/160630/MS.pdf
MS15000 & MS8000, even without bleeding/even-voltage resistors... So maybe the capacitors have been hurt a lot..  :'(
Or as I said, Lucky~ My SGTC is working at ~ 1MHz so maybe the dV/dt is OK? My circuit is really a crude set. And, The coil make a lot of noise so I only run it for seconds every time.

Not stupid, I made the same mistake as first. The capacitors you picked have good ratings but at that frequency which is high, you’re almost certainly exceeding the dv/dt rating on them. I was assuming your fly back is similar to mine which would mean 24v input = 24000 output. If you do the math you’ll see it would require extremely high dv/dt. My coil is quite loud too I think that’s just how they are. I don’t actually know what happens if you exceed the capacitors ratings for short bursts. Anyway I like your setup and it inspired me to build mine so don’t feel bad about it.

[NyaaX_X]

Thank you~

Now I'm thinking of this. Maybe they can help the circuit to against the harmful operation?

1_ The margin design of capacitor manufacturer.
2_ The self-healing of Film capacitor.
3_ The ESR of capacitor and primary coil.
4_ The ESR of spark gap.
5_ The voltage-limit by spark gap.
6_ The ESR of flyback transformer (high voltage until spark gap break down).
7_ The corona discharge and leakage.
8_ The damping of this circuit system.

I don't know.. Nonsense then.

[Michelle_]

Thank you~

Now I'm thinking of this. Maybe they can help the circuit to against the harmful operation?

1_ The margin design of capacitor manufacturer.
2_ The self-healing of Film capacitor.
3_ The ESR of capacitor and primary coil.
4_ The ESR of spark gap.
5_ The voltage-limit by spark gap.
6_ The ESR of flyback transformer (high voltage until spark gap break down).
7_ The corona discharge and leakage.
8_ The damping of this circuit system.

I don't know.. Nonsense then.

I don't know the answers to that, but I would be curious if anyone more knowledgeable can explain how something like your system works. I attached the calculation for dv/dt. With your system at 1MHZ and 24000V it ends up being around 150KV/uS, whereas you have 4 capacitors rated at around 2K each, so you're potentially exceeding the rating by 18.75x lol.

Again don't feel bad about it, apparently people are doing it. Here's a link to a youtube video with a similar tesla coil and the builder admits in the comments they did no calculations on the capacitors lol:

/>

[Michelle_]

You might be onto something on the losses idea though, with nothing attached I am putting out 6200V but across the spark gap when it's running it's only 2100V...

Maybe that means I can increase the voltage until it's 6000 across the gap actual hmm

Also I'm just going to button this one up and call it done, then build a new V2 and then be done with spark gaps. This one is doing everything I hoped it would do more or less especially if I can up the power, then it won't be a waste of capacitors and I'll have a low powered unit for tinkering which is kinda a good thing.

[Michelle_]

Just needs the top dust cover/coil insulator, and side "doors". It looks about like what I had in mind and everything fits inside as intended. Might mess with it later and see what voltages I can get across the gap if I raise the power supply voltage. As far as I am aware the voltage across the gap is what the capacitors are actually seeing so I should have a ton of room to increase the power. It would also explain why it doesn't spark under 6000 volts since it's probably actually more like 2000 volts across the spark gap? Not sure if this is normal or good.

[Michelle_]

I found out some interesting stuff in case anyone is curious, it turns out the static spark gap is EXTREMELY sensitive to adjustment. Fractions of a mm can change the voltage across the gap by thousands.

It seems that it needs more voltage to start than it will use for steady state. Also, interestingly, at a certain point increasing the input voltage will decrease the steady state voltage and increase the frequency of the sparking.

I think unless you're measuring the voltage across the gap it's difficult to have any idea what the real voltage is (at least for me lol). I think some people putting 24V through the flyback are not getting anywhere near 24000v across the gap, especially if it's a basic two part gap. Maybe it's different with a multi-contact gap but I'm not sure because I haven't tried one.

I calculated my MMC's maximum dv/dt at 5500v steady state (12000v/uS) and have it now set to run steadily at 5000v (with a no-load voltage of 7300). Now it's breaking out of the toroid on its own.

Does anyone know the implications of voltage vs repetition rate? It seems like it's possible to adjust it so it's repeating slower or faster but has the same voltage.

Again I don't really know what I'm doing here but I'm trying to learn.

Edit: math

[NyaaX_X]

Hi,
Maybe you can find something in Richieburnett's webpage
https://www.richieburnett.co.uk/tesla.shtml

Best Regards,

[Michelle_]

Hi,
Maybe you can find something in Richieburnett's webpage
https://www.richieburnett.co.uk/tesla.shtml

Best Regards,

Thanks that’s the only page I’ve seen that goes into any kind of useful detail. Seems like all anyone cares about is making the most powerful drsstc they can afford.

[MRMILSTAR]

Voltage and rep rate are related. The relationship depends on the available current from your transformer. You can get an idea by looking at the output of JAVATC. Specifically, look at the capacitor charge level output. This will show you the voltage level of the capacitor charge at the end of each rep just before discharge.

[Michelle_]

Voltage and rep rate are related. The relationship depends on the available current from your transformer. You can get an idea by looking at the output of JAVATC. Specifically, look at the capacitor charge level output. This will show you the voltage level of the capacitor charge at the end of each rep just before discharge.

I saw that on there but I wasn't sure what to make of that since I have a DC power supply (and their calculator requires HZ and also throws out funny numbers if I just put 60). That's why I was thinking I could measure the voltage across the spark gap terminals but since it's reading lower than the voltage I would read with nothing connected to the power supply I'm not sure what the voltage on the capacitors actually is... Might just be missing a key fact here.

[NyaaX_X]

SSTC are more flexable for controlling and have better efficiency performance. But need more checking to avoid the over rated operation and transient rate to the switching components.

For SGTC, there are only transformer to transformer. If anything broken, it would be simply to be find out.

[Michelle_]

Yeah the spark gap does seem simpler for sure.

I'm going to totally redo this one though. There was a defect in my secondary and a small piece of the wiring burned out, I think it's been this way the whole time because I noticed an occasional flash but upping the power made it worse.

Upgrades:

- 3" non tapered secondary, taller (14" tall) with thicker wiring (30awg)
- Bigger toroid (3" donut with 9" OD)
- Redesigned primary coil holder with slots for tapping underneath and normal style wire standoffs for zip ties
- Lower overall resonant frequency
- Primary with closer spacing and more turns, range of adjustment encompasses ~15% deviation of designed secondary resonant freq
- Different capacitors with higher dV/dT
- Designed for 12,000V + 25% safety factor; should be able to use readily available industrial 12V din or chassis mount PSU versus my expensive bench top power supply. Voltage across gap will likely be under 12KV thus increasing the safety factor likely to 50%.
- Guard panels for chassis that snap onto the posts
- Strike ring
- Spark gap distance adjustable with precision threaded stage, possibly multi contact gap


This version, as far as I can say, is as "normal" as possible based on suggested ratios, sizes, etc...

[Michelle_]

I did it; here is V2 with improvements based on what I learned from V1. Thanks to everyone that has answered my questions and given me advice. It has been a great learning experience.

Disclaimer: I know aesthetically this will not be to everyone's tastes 

Material: Almost entirely translucent PETG printed from a bambu p1s (including secondary form, toroid form)
PSU: 12-14V (adjusted to 14v) 120W din rail mounted + ZVS flyback from Amazon/Chiner + 1.5A 120v circuit breaker
Primary: 10 turns (tapped to 8.5-ish) 1/8" copper tubing
Secondary: 3"x14" 30AWG 1300 turns 300KHZ designed 313KHZ actual
Toroid 3" ID 9" OD
Caps: 14x Cornell Dubilier 942C16P22K-F in series
Spark Gap: Single gap 1" steel balls with two fans

With initial tuning in daylight I'm getting 12" (30cm)+ sparks which I think is not bad for a 120W spark gap tesla coil built by a noob

Will continue to tinker with it later, the deep fried neon calculator claims it's possible to get 18" arcs from 120 watts but that seems optimistic, at any rate I'm happy with the improvement from the first iteration. I'm wondering if the toroid is too small because sparks are coming out of the sides even with a breakout point. Pictured is 3" neon globe. Overall unit is 27" (68.5cm) tall.

It could be better and more powerful, theoretically I can run 24v into this flyback but the capacitor situation would get out of hand quickly. Assuming this one doesn't die a sparky death like the first secondary I will move on to some kind of solid state unit next, eventually.

Cheers,

M