DIY High voltage Jacob's Ladder Build

[angryowl]

2 minutes ago, The Admiral said:

IIRC, the Tesla coil operates at high frequency, and so the current confines itself to the skin and not the vital organs. It is reputed to be 'safe'. Can't imagine a shock would comfortable though!

Ian

Right, I don't really know how they work, but I thought the shock would affect internal organs. Good to know, thanks.

[souls33k3r]

7 minutes ago, Davey-T said:

Just get some wellies and rubber gloves and you'll be fine 

Dave

Pinky promise?

[The Admiral]

But then again, having looked at the safety warning on the ZVS web page (http://www.instructables.com/id/ZVS-Driver/) I'm not so sure 

Ian

[angryowl]

I know there's several ways of measuring the voltage and current output of the two flybacks in series and may do it at some point. But I suspect, by the thickness and length of the arcs it outputs quite a few milliamps and volts which should be plenty to stop someone's heart.

So yes ZVS drivers in general are dangerous devices and should not be taken lightly and all safety measures should be applied when dealing with them. (Seen some videos on Youtube of people playing with ZVS flyback arcs, but they were holding one of the high voltage output wires with their bare hands! )

[angryowl]

Set up the circuit again, this time bolted down the flybacks and glued the two server fans on the wood board as can be see in the images. They are blowing cold air on both the ferrite cores/coils and the bodies of the flybacks themselves. The fans are generating so much flow that running the ladder anywhere next to them would be pretty much impossible so I've added some protection and it seems to work well. 

Had a stopwatch next to me, started the fans first, next the ladder then proceeded to leave them running for 1.5 minutes. Stopped everything down, the case of the flybacks was 42 degrees Celsius and the ferrite cores were 35 degrees. They were just warm to the touch so it does seem like the fans are doing their part, the only thing which concerns me is that after running it, I am getting an insulation burning smell if I get very close to the top of the flybacks. This is at the top of them, which makes me think either the primaries or secondaries are being overheated by the amount of power running through them and the smell I'm detecting might well be their insulation starting to burn off? Or this is completely normal and I'm overreacting? Any thoughts?

[angryowl]

What I'll try this morning is, removing the glue on top of the ferrite rods, and ensuring they are nice and flat. I think one of those metal clips holding it down is sticking out so will need to be filed down to have a perfectly flat surface between the two rods. By sheer change I've got a copper pipe heatsink which accepts 70mm fans which can be perfectly placed on top of the ferrite cores. 

Two options the way I see it:

1. Mount the heatsink with some silicone heat transfer compound which I have and with the fan on see if that cools the cores enough
2. I have a remaining 40mm x 40mm TEC device from a previous project which as a last resort I could mount on the heatsink then on top of the rods, cooling them to subzero temperatures. that should do the trick 

I know that on these types of flyback transformers the coils are deep embedded in resin. So even if I could grab my pliers and somehow take the encasing plastic off I'd still be left with a chunk of resin which would still be a pain cooling. What I am hoping to do is to try and cool them from the outside(server fans) and inside(ferrite rods) to the point I can safely run them for about 5 minutes without worrying they may short and catch on fire.

[angryowl]

Tried just the heatsink and fan and didn't help with cooling much at all.

So built an adaptor and did a few modifications to the flybacks and will use the heatsink with the 40mm square TEC. Due to the way it will be mounted, there will be two 1mm thick thermal silicone pads between the TEC and the top of the flyback ferrite rods. Which by the way have been smoothed out to a perfectly flat surface, not that it really matters as the silicone pads mould around grooves and other imperfections. 

Between the heatsink and the TEC there will be thermal paste applied as the surfaces are both very smooth and conducts heat much better than silicone pads.

This will be assembled tomorrow as I'm done for the day... 

[Alien 13]

Nice work, I am building a Tesla coil (actually a couple of varying sizes) at the moment all powered from 12V.

Alan

[angryowl]

2 hours ago, Alien 13 said:

Nice work, I am building a Tesla coil (actually a couple of varying sizes) at the moment all powered from 12V.

Alan

Thanks, yeah Tesla coils can be really cool devices and I would imagine fun to build. Are you documenting it anywhere? Never knew you could build them from 12V, always had the impression they run off the mains and draw huge amps!

[Alien 13]

1 minute ago, angryowl said:

Thanks, yeah Tesla coils can be really cool devices and I would imagine fun to build. Are you documenting it anywhere? Never knew you could build them from 12V, always had the impression they run off the mains and draw huge amps!

They can be powered from any small battery using a single transistor that self oscillates at the coils resonant frequency, even Amazon sell them as toys????  for less than £10. 

https://www.amazon.co.uk/JZK-DIY-Plasma-Speaker-Transmission/dp/B071DV9PZL/ref=sr_1_1?ie=UTF8&qid=1523051305&sr=8-1&keywords=tesla+coil+kit

Alan

[angryowl]

2 minutes ago, Alien 13 said:

They can be powered from any small battery using a single transistor that self oscillates at the coils resonant frequency, even Amazon sell them as toys????  for less than £10. 

https://www.amazon.co.uk/JZK-DIY-Plasma-Speaker-Transmission/dp/B071DV9PZL/ref=sr_1_1?ie=UTF8&qid=1523051305&sr=8-1&keywords=tesla+coil+kit

Alan

Ah I see, fair enough, thanks for that.

[angryowl]

So this morning tried the heat sink and the TEC solution and it indeed kept most of the ferrite rods cool during operation. However that has no effect whatsoever on the main problem which I'm facing, the cooling of the coils themselves.

With both fans at full power and the TEC on top managed to run for 3.5 minutes until the burning smell started to reappear at the top of the flybacks. This time I actually saw a bit of smoke coming out from the side of the flyback case just in the middle of the transformer. Upon inspection, it looks like it was one of the coils melted some of its enamel and a bit of the casing plastic. Everything still works just as fine and I can still run it perfectly fine reasonably for up to two minutes.

The main problem is I can't get easy access to the secondary coils to cool them properly so I can run this for more than 2 mins. I'm thinking of attempting a disassembly of one of them and see if there's any way of exposing the secondaries. I believe this type of flyback is the one encased in resin so the disassembly could prove very difficult.

If I do decide to disassemble one, and it ends up in total disaster, I could always find a ferrite core and wind my own secondary 

The bottom line is I'm pushing way too much power through the flybacks and they can't cope. 

I can leave it as it is and finish the construction and just run it for 2 mins at a time, or the other two options which I have. At the moment I don't know, I do know I'm keeping the configuration with the main 30V transformer and ZVS driver as I like it.

So depending how I feel, this project might be shelved until I decide what to do, unless someone on here has an idea on cooling the coils/flybacks differently, would be much appreciated.

In the meantime some pics of the TEC assembly which with the two silicone pads still cools down to about -7 degrees Celsius and draws quite a few amps!

[angryowl]

So the project is shelved until I can find a better solution. I may try taking one of the flybacks apart too see if anything can be done, if not I was thinking of winding my own secondary and doing it that way. Doing research at the moment so will see how that goes...

[angryowl]

So project is now resurrected. I opened up one of the transformers carefully with a dremel and got the coils out without much damage.

They where proper insulated with a transparent material which went over each set of winding twice! The thickness of the insulating material is 0.08mm and looks the real deal. Anyway the coli/core diameter as it was pulled out was 26mm.

Started off by removing the first insulating foil and reached the first layer of the high voltage windings. The total length of the winding section on the core is 28mm. Removed the first layer, managed to grab a single strand of wire and using my micrometer it turns out the wire is 0.035mm in diameter. Much thinner than I expected.  Anyway in AWG that's 47. Then proceeded to remove all of the insulation and layers and it was a total of 5 layers.

To count the number of wires per each layer I grabbed a random 4.72mm piece and counted the individual wires in that. Total wires were 96 and as you will see in the pictures there's spacing between the individual wires on the layers. 

That gives us 569.49 wires per full 28mm layer. Let's round that up to 570 times 5 layers in total gives us 2850 turns in total. Of course the actual length would be a bit more hard to calculate as the diameter of the core gets smaller and smaller with each layer, but there's no real need to calculate the length of wire.

Since I was using two identical transformers the total number of secondary turns would be 5700.

The remaining core seems to be made of some sort of ceramic material or something as it's incredibly strong. In the fourth picture you can see some, I assume, more layer of insulation, some type of of plastic maybe then the primaries encased.  

Some pictures to show things better:

 

[angryowl]

I still want a Jacob's Ladder so I'll wind my own secondaries. there's plenty of resources on the web on how to do it properly with insulation.

I'll join the two cores exactly as seen in the above picture then the ZVS windings will go in the middle. I'll print two cylinder casings out of ASA as it can withstand very high temperatures which will go in the two other cores.

I'm looking at NMN insulation material aka Nomex. They are easy to find and are great for high voltage insulation which should prevent arcing between the coils. Not sure about encasing it in resin or some other materials 
I've seen people use but I may have to. I'm thinking cooling wise what would be the best method to have the secondaries but at the same time have them properly insulated. More research needed...

Another alternative would be to use this  

[angryowl]

Looked at the prices of similar gauge wire online and thy're ridiculous for the amount I need. Then I tried breaking a motor apart and getting some wire from it's windings, let's just say it didn't end well... 

Pulled the trigger on the large ferrite core and for the secondary will use standard Cat5 in-wall Ethernet cable wire. It's 0.515mm diameter SWG 25 copper insulated by some sort of plastic material. It's overall diameter including the plastic is 0.85mm.

Done a quick calculation off of the dimensions in the listing pictures and I should be able to wind approximately 4340 turns. This is taking into account the space for the primary but no insulation, so depending on what insulation I use, I was thinking regular PTFE Thread Seal Tape used for plumbing as it withstands high temps just fine. A bit worried about the wire's plastic material itself and how it would withstand heat. 

At 25 SWG this'll pull a lot of amps putting some pressure on the torroidal transformer and ZVS driver but we shall see what happens...

[angryowl]

IT HAS ARRIVED!!! (mind you I have small hands )

I'll be printing the primary and secondary "coil holders" from ASA material today. I'm thinking with the Cat5 SWG 25 wire as ste secondary this thing'll be capable of some massive current draw!

[angryowl]

The secondary coil holder which will be printed from ASA material. 

At the moment I'm trying to figure out if I have enough filament for a complete print as obviously whichever way it's flipped on the print bed it'll require quite a bit of support filament. I'd like to print it standing up like that as flipped I'd have to use more support so there's a very small difference in amount of material required. I'll do some checking as to how much filament I have left as this'll end up requiring between 40 to 45 meters of filament. 

 

[angryowl]

Well according to this I have approximately 46 meters of filament remaining. I can work with that, not much more left for the second coil holder but I think I'll just wind the ZVS wires directly on the core as they're only about 18x18 turns.

[angryowl]

The strands of Cat5 wire which'll form the secondary. I just need to find a way of straightening them without affecting their plastic coating too much or degrading it. I'll be doing some tests on small pieces using various methods. Anyone know a good way of achieving this? 

[angryowl]

Found a great method of straightening the wires; I've placed a piece of 3mm thick rubber on top of a very flat piece of glass, I've then got another piece of rubber, this time it's more slippery and much thinner and placed it on top of another piece of glass that's about 70mm wide. The wire is placed on the bottom rubber piece and then I simply apply a bit of pressure whilst moving the top glass with the rubber underneath in an up and down motion. The rubber ensures that the wire rotates whilst I move the upper plate and the pressure I apply straightens the wire out.

This technique seems to work very well and produces very straight wires and doesn't damage or affect the plastic insulation at all.

[Alien 13]

Looking forward to more updates, I have now built three low power Tesla coils the largest of which will light up a fluorescent bulb from a foot away...A Jacobs ladder is on my must do/need list along with a cloud chamber.

Alan

[angryowl]

Thanks @Alien 13. That Tesla coil sounds pretty nice. Cloud chambers are also amazing devices, they're so fun to observe.

However this project has now taken a more traditional DIY turn. My first print's support failed to stick to the layer and so that print had to be stopped about 10% through. That still consumed enough filament for me to now not have enough for another print. Anyway I've got more serious issues with my printer and I think my extruder stepper is failing, so that'll need some troubleshooting.

Anyway I've built sort of a coil holder out of some very heat resistant plastic, it's in layers with epoxy between them and will leave to dry through the night. For the ends I'll be using two pieces of 2.3mm plastic sheet and will just have to manually hack the hole through the middle. That'll be a job for tomorrow anyway...

[Ant]

That looks really cool

I notice that in the first half of the video the arcs pretty much go all the way up the rods? But from the half way point they don't in fact quite a lot don't get past 1/4 or 1/3 up!

I also noticed that the distance between the rods alters very slightly - is that related ?

Ant

[angryowl]

27 minutes ago, Ant said:

That looks really cool

I notice that in the first half of the video the arcs pretty much go all the way up the rods? But from the half way point they don't in fact quite a lot don't get past 1/4 or 1/3 up!

I also noticed that the distance between the rods alters very slightly - is that related ?

Ant

Hi Ant,

Not sure what to tell you. Noticed that too and I think in the second part of the video the coils in the flybacks were getting too hot and could not provide enough voltage anymore to allow the arc to travel all the way up? 

When you say "distance between the rods alters very slightly" you mean the attraction and of the rods when the arc is at the top? Yeah I noticed that too, quite interesting effect! Don't think it has an effect on the arc not going higher on the second part of the video. Unless I'm wrong?