Rolled capacitors taking ages to charge / Tesla Coils / Forums

Forums

4hv.org :: Forums :: Tesla Coils

« Previous topic | Next topic »

Rolled capacitors taking ages to charge

1 2 next

Move Thread

LAN_403

AndreiRS

Wed Feb 28 2018, 06:18PM

Registered Member #62109 Joined: Sun Jan 28 2018, 10:00PM
Location: Porto Alegre
Posts: 56

This is my first content post here. The other one was just a "hi" to everyone. I also searched for similar posts, I even found one, but then it goes to another possible problems. Some 10 years ago I tried to make a mall TC using a flyback, it didn't work. But probably because the power was very low on my setup and it was out of tune, since I didn't had anything to measure capacitor and coil inductance.

Recently I built a hv transformer, it is 13800V at 100mA. So I decided to make another Tesla coil. It already worked but it looks so weak. Nothing close to the 1.6kw transformer power. Ok I was using microwave oven capacitors, but not sure if that was the problem since I read about people using them for short time and saying it works, not sure... Then I made some modifications, raised my primary to put the base of the secondary inside it. I also removed a lot of turns from the secondary. It had ~1600 turns of 32awg. Now it has around 1200. Changed the rotary spark gap from stainless steel to copper.

The last changes were, tapping the microwave capacitors which were measuring 96nF to read 54nF. Same thing, didn't work well. But the spark gap was sparking like crazy. Then I just threw that thing on the table and got my 2 rolled aluminum foil capacitors. Each one is 215nF, too much for the given 50.1nF by Tesla map. I connected them in series and got 107nF.

When I turn on the transformer, I can only hear a small "bloop" and the transformer humming. But no sparks at the gap. I only let it like that for 2 seconds to avoid burning something, then I got one single spark and turned it off. Yesterday I unrolled both capacitors to 104nF and dipped them in electric motor varnish. I will have 52nF in series now, but not sure what to try next.

Sulaiman

Wed Feb 28 2018, 09:28PM

Registered Member #162 Joined: Mon Feb 13 2006, 10:25AM
Location: United Kingdom
Posts: 3140

I never managed to make a reliable roll-my-own SGTC primary capacitor bank.
Some died instantly, some after seconds, minutes, and once for many minutes :(

There is a reason why film capacitors are easily avaiable to 1 kV and progressively more rare above that.
Others may differ, but I found 3 kV about do-able, 5 kV or more eventually killed all of mine, due to corona I suspect.
At that time I did not have a vacuum pump so air was extremely difficult to exclude.
(if you are trying to make reliable dry high voltage film capacitors then all I can say is ha ha ha / good luck ! )
SO

If the spark gap does not fire, but does when you remove the capacitor(s), then you know where the problem is ;)

With EXTREME CAUTION observe your sgtc in the dark - internal capacitor faults shine.

AndreiRS

Wed Feb 28 2018, 10:44PM

Registered Member #62109 Joined: Sun Jan 28 2018, 10:00PM
Location: Porto Alegre
Posts: 56

Haha oh...

In the dark it doesn't emitt light. I forgot to mention it. Now that I dipped it into insulating varnish I have to wait. That thing takes some hours to dry. I have given up trying to make things with vacuum now. I just throw them inside oil or fiberglass resin depending on the purpose. But first I put them in a not so hot oven, for some time, to remove humidity. On my hv transformer, I used heat, then vacuum, then oil. After the crash test, they still give me the same capacitance... Hmmm. I can't find appropriate capacitors around here.

KUN

Thu Mar 01 2018, 08:19AM

Registered Member #61824 Joined: Mon Oct 02 2017, 05:38PM
Location:
Posts: 40

Why not make mmc bank caps, they are probabbly the most efficient and least likely to fail. I had my paralel plate caps, they worked for days but tend to heat real quick, so after some operation they melted, and arced trough. When i build my first mmc i noticed big increase in arc lenght, as they are more effitcient

Sulaiman

Thu Mar 01 2018, 12:45PM

Registered Member #162 Joined: Mon Feb 13 2006, 10:25AM
Location: United Kingdom
Posts: 3140

An MMC is the most reliable option,
but
at one time or other you gotta try roll-your-own capacitors, - it a coilers right-of-passage !

the_anomaly

Thu Mar 01 2018, 01:12PM

Registered Member #19 Joined: Thu Feb 02 2006, 03:19PM
Location: Jacksonville, FL
Posts: 168

I have not made caps but I have wound my own flyback secondaries using plastic sheet in between the layers as an insulator. I then put the transformer in mineral oil and put it in a vacuum jar for 15 minutes to remove the air. A lot of air comes out. I am convinced this is a crucial step with high voltage insulation. I've also read that ultrasound is sometimes used in addition to a vacuum to help disturb the air for high viscosity potting agents like some epoxies.

Graham Armitage

Thu Mar 01 2018, 01:16PM

Registered Member #6038 Joined: Mon Aug 06 2012, 11:31AM
Location: Salado, TX
Posts: 248

I had success rolling capacitors for a 2kW NST bank. They worked for years producing impressive results. I used aluminum flashing with a double sheet of thick polythene (40kv breakdown). Once rolled, I inserted them into 4" PVC pipe, capped them and filled with vegetable oil (all I had at the time). Each one was around 10nF. With 3 in parallel was a perfect impedance match for the NST bank. I have made crude caps like this and they work, even if not very well or for long. If i were you I would check other aspects of the coil config. Is the impedance matching the transformer for maximum power transfer at 50/60Hz? MMC banks are a lot easier, more flexible, but definitely more expensive.

AndreiRS

Thu Mar 01 2018, 02:47PM

Registered Member #62109 Joined: Sun Jan 28 2018, 10:00PM
Location: Porto Alegre
Posts: 56

I'm always trying to find commercial capacitors around here. But I only find thing like 100nF 1600V. I would need 72 of them. And they are the el cheapo capacitors, very small. I have no idea what is the internal resistance for example, since the store don't even give a part number.

"you gotta try roll-your-own capacitors, - it a coilers right-of-passage !"
Hahahah, I think I like suffering lol.

Is it that feel, like, "I'm gonna try, maybe it works."

About the impedance, well I did what some people suggest. Put some wire on the topload going to a strike bolt I had near. Without electrical connection. Then I measured the ressonant frequency to be 160khz, using Terry Fritz tuner. Secondary is a 4" pipe, and winding is 13" in height. I did it 3:1 because I only had 32awg wire. Previously the coil had 1600 turns, now it has around 1200. That tuner didn't work for primary calibration. Then I measured capacitance and inductance and calculated the ressonant frequency. Since I had already 54nF on the microwave caps. I only had to find the spot with the inductance required on the primary. But is just worked the same, then I decided my own caps instead of the microwave oven ones. And after that it comes to my first post here. Today I will try the 52nF combination.

I'm just waiting to let them dry a bit more on my improvised oven, "capacitors on the roof on a sunny day".

Graham Armitage

Thu Mar 01 2018, 03:39PM

Registered Member #6038 Joined: Mon Aug 06 2012, 11:31AM
Location: Salado, TX
Posts: 248

So when I talk about impedance matching it is is important to match the capacitor to the transformer. Forget the secondary and primary coils of the tesla for a minute. When the reactance of the capacitor and transformer impedance match it will transfer maximum power to the tank circuit. If it's mismatched you will not get proper voltage buildup and spark gap wont fire at all. Use Xc = 1/2.pi.f.C for reactance calculation. Use 50 or 60Hz for the frequency in this calc. Once you have the capacitor matched to the transformer impedance, then that sets the parameters for frequency and dimensions of the secondary should be.

AndreiRS

Thu Mar 01 2018, 10:50PM

Registered Member #62109 Joined: Sun Jan 28 2018, 10:00PM
Location: Porto Alegre
Posts: 56

Got it, I mean maybe. Assuming C is in Farads, I got 53.805kohms after realizing that 1 divides by the whole rest of the equation. I was multiplying everything by 1/2.

What transformer impedance I need to measure, mains connections or hv secondary? My meter won't read the secondary inductance. Without the iron core it already goes over the scale of 20H. When I short circuit the secondary, it gives me 4.15mH, which at 60hz is only 1.565ohms. Nothing to do with the 53kohms.

Let me measure primary inductance and do some rule of 3 with turn numbers, maybe it will make more sense.

EDIT: Well, actually not over scale. My meter can't deal with the 2.7kohms of the secondary spool. By doing what I said earlier, I got 12.15H. I measured the primary inductance, 107.2mH. Divided 15076 turns by 133, equals 113.35. Multiply that by 107.2 and we have ~12150mH or 12.15H. Using your formula, the secondary impedance is 4.58kohms... Which is not even close to 53k from the begging of this post. I'm stuck. If I add 2.7k from pure winding resistance to that, then we have 7.28kohms.

1 2 next

Moderator(s): Chris Russell, Noelle, Alex, Tesladownunder, Dave Marshall, Dave Billington, Bjørn, Steve Conner, Wolfram, Kizmo, Mads Barnkob